SMJ

E-ISSN 2228-8082

Volume 75, Number 12, December 2023

Siriraj Medical Journal

The world-leading biomedical science of Thailand

Indexed by

THAILAND SECTION 1954

https://he02.tci-thaijo.org/index.php/sirirajmedj/index

E-mail: sijournal92@gmail.com

MONTHLY

ORIGINAL ARTICLE REVIEW ARTICLE

SMJ

Siriraj Medical Journal

The world-leading biomedical science of Thailand

ORIGINAL ARTICLE Volume 75 Number 12

December 2023

838 Systematic Review and Meta-Analysis on

Role of Adiponectin to Leptin Ratio in Women with

Polycystic Ovarian Syndrome

Archie Fontana Iskandar, Nicolas Daniel Widjanarko,

Felicia Grizelda Suryatenggara, Leonardo Leonardo, Rosalia Sylfiasari, Nadhea Debrinita Surya, Christian Ardianto

851 A Systematic Review and Meta Analysis of Non-Randomized Interventional Studies on the Pamidronate Treatment Efficacy for Patients with Bone Fibrous Dysplasia

Nicolas Daniel Widjanarko, Anthony Ekaputra, Jessica Felicia Ang

864 Plasma Alpha Synuclein as a Potent Biomarker of Diseases with Synucleinopathies

Chaisak Dumrikarnlert, Lertchai Wachirutmangur, Suthipol Udomphanthurak, Chatchawan Rattanabannakit,

Prachaya Srivanitchapoom, Vorapun Senanarong

871 Mobile Device Digital Photography for Teledermatology Consultation: Real-Life Situations

Sumanas Bunyaratavej, Pattriya Jirawattanadon, Chudapa Sereeaphinan, Supisara Wongdama,

Sanchai Sombatmaithai, Charussri Leeyaphan

880 Effect of Wearing a Face Mask on the 6-Minute Walk Test in Healthy Volunteers

Simaporn Promsarn, Kanokwan Rattanasaengloet, Sutat Pipopsuthipaiboon, Nongnoot Panitchatchawal,

Patharapan Lersritwimanmaen

887 Efficacy and Safety of Topical 5% Azelaic Acid Solution Versus 2% Minoxidil Solution in the Treatment of Female Pattern Hair Loss

Kanchalit Thanomkitti, Chutipon Pruksaeakanan, Chanika Subchookul, Norramon Charoenpipatsin, Daranporn Triwongwaranat, Supenya Varothai, Rattapon Thuangtong, Tanyalak Chumnumrat

894 Factors Related Pain Catastrophizing in Hospitalized Patients with Trauma

Prampree Nantawong, Thitipong Tankumpuan, Ketsarin Utriyaprasit, Natthida Owattanapanich

902 Assessing Low-Concentration Atropine in Myopia Progression: A Systematic Review

Stella Nathania, Jovita Jutamulia, Gabriella Hafidha Badruddin

909 Efficacy of Pregabalin, Solifenacin, or Combination therapy for Ureteral Stent Related Symptoms: A Systematic Review and Meta-Analysis

Nicholas Andrian Singgih, Jacinda Risha Oktaviani, William Adipurnama, Cecilia Noviyanti Salim, Kevin Tandarto, Athaya Febriantyo Purnomo, Egi Edward Manuputty

SMJ

SIRIRAJ MEDICAL JOURNAL

https://he02.tci-thaijo.org/index.php/sirirajmedj/index

Executive Editor: Apichat Asavamongkolkul Editorial Director: Aasis Unnanuntana

Editor-in-Chief: Thawatchai Akaraviputh, Mahidol University, Thailand

Associate Editors

Adisorn Ratanayotha, Mahidol University, Thailand Chenchit Chayachinda, Mahidol University, Thailand Pornprom Muangman, Mahidol University, Thailand Phunchai Charatcharoenwitthaya, Mahidol University, Thailand

Varut Lohsiriwat, Mahidol University, Thailand

Andrew S.C. Rice, Imperial College London, UK

International Editorial Board

Morris Solomon Odell, Monash University, Australia

Anusak Yiengpruksawan, The Valley Robotic Institute, USA Barbara Knowles, The Jackson Laboratory, USA Christopher Khor, Singapore General Hospital, Singapore Ciro Isidoro, University of Novara, Italy

David S. Sheps, University of Florida, USA

David Wayne Ussery, University of Arkansas for Medical Sciences, USA Davor Solter, The Jackson Laboratory, USA

Dennis J. Janisse, Medical College of Wisconsin, USA

Dong-Wan Seo, University of Ulsan College of Medicine, Republic of Korea Folker Meyer, Argonne National Laboratory, USA

Frans Laurens Moll, University Medical Center Ultrecht, Netherlands

G. Allen Finley, Delhousie University, Canada

George S. Baillie, University of Glasgow, United Kingdom

Gregory Bancroft, London School of Hygiene of Tropical Medicine, United Kingdom Gustavo Saposnik, St. Michael’s Hospital, Canada

Harland Winter, Harvard Medical School, USA

Hidemi Goto, Nagoya University Graduate School of Medicine, Japan Ichizo Nishino, National Institute of Neuroscience NCNP, Japan Intawat Nookaew, University of Arkansas for Medical Sciences, USA James P. Doland, Oregon Health & Science University, USA

John Damian Smith, Texas A&M University-San Antonio, USA John Hunter, Oregon Health & Science University, USA

Juri Gelovani, Wayne State University, USA

Karl Thomas Moritz, Swedish University of Agricultural Sciences, Sweden Kazuo Hara, Aichi Cancer Center Hospital, Japan

Keiichi Akita, Tokyo Medical and Dental University Hospital, Japan Kym Francis Faull, David Geffen School of Medicine, USA

Kyoichi Takaori, Kyoto University Hospital, Japan Marcela Hermoso Ramello, University of Chile, Chile Marianne Hokland, University of Aarhus, Denmark

Matthew S. Dunne, Institute of Food, Nutrition, and Health, Switzerland Mitsuhiro Kida, Kitasato University & Hospital, Japan

Moses Rodriguez, Mayo Clinic, USA

Nam H. CHO, Ajou University School of Medicine and Hospital, Republic of Korea Nima Rezaei, Tehran University of Medical Sciences, Iran

Noritaka Isogai, Kinki University, Japan

Paul James Brindley, George Washington University, USA

Pauline Mary Rudd, National Institute for Bioprocessing Research and Training Fosters Avenue Mount Merrion Blackrock Co., Dublin, Ireland

Peter Hokland, Aarhus University Hospital, Denmark

Philip A. Brunell, State University of New York At Buffalo, USA Philip Board, Australian National University, Australia

Richard J. Deckelbaum, Columbia University, USA Richard W. Titball, University of Exeter, USA Robert W. Mann, University of Hawaii, USA

Robin CN Williamson, Royal Postgraduate Medical School, United Kingdom Sara Schwanke Khilji, Oregon Health & Science University, USA

Seigo Kitano, Oita University, Japan

Shomei Ryozawa, Saitama Medical University, Japan Shuji Shimizu, Kyushu University Hospital, Japan

Stanlay James Rogers, University of California, San Francisco, USA Stephen Dalton, University of Georgia, USA

Sue Fletcher, Murdoch University, Australia

Tai-Soon Yong, Yonsei University, Republic of Korea Tomohisa Uchida, Oita University, Japan

Victor Manuel Charoenrook de la Fuente, Centro de Oftalmologia Barraquer, Spain Vincent W.S. Chan, University of Toronto, Canada

Wen-Shiang Chen, National Taiwan University College of Medicine, Taiwan Wikrom Karnsakul, Johns Hopkins Children’s Center, USA

Yasushi Sano, Director of Gastrointestinal Center, Japan Yik Ying Teo, National University of Singapore, Singapore Yoshiki Hirooka, Nagoya University Hospital, Japan

Yozo Miyake, Aichi Medical University, Japan Yuji Murata, Aizenbashi Hospital, Japan

Ampaiwan Chuansumrit, Mahidol University, Thailand Anuwat Pongkunakorn, Lampang Hospital, Thailand Jarupim Soongswang, Mahidol University, Thailand Nopphol Pausawasdi, Mahidol University, Thailand Nopporn Sittisombut, Chiang Mai University, Thailand Pa-thai Yenchitsomanus, Mahidol University, Thailand Pornchai O-Charoenrat, Mahidol University, Thailand Prapon Wilairat, Mahidol University, Thailand Puttinun Patpituck, Mahidol University, Thailand Rungroj Krittayaphong, Mahidol University, Thailand Saranatra Waikakul, Mahidol University, Thailand

Editorial Board

Sayomporn Sirinavin, Mahidol University, Thailand Suneerat Kongsayreepong, Mahidol University, Thailand Supakorn Rojananin, Mahidol University, Thailand Surapol Issaragrisil, Mahidol University, Thailand

Suttipong Wacharasindhu, Chulalongkorn University, Thailand Vasant Sumethkul, Mahidol University, Thailand

Vitoon Chinswangwatanakul, Mahidol University, Thailand Watchara Kasinrerk, Chiang Mai University, Thailand Wiroon Laupattrakasem, Khon Kaen University, Thailand Yuen Tanniradorn, Chulalongkorn University, Thailand

Journal Manager: Nuchpraweepawn Saleeon, Mahidol University, Thailand

Medical Illustrator: Nuchpraweepawn Saleeon, Mahidol University, Thailand

Proofreaders: Noochpraweeporn Saleeon, Mahidol University, Thailand, Amornrat Sangkaew, Mahidol University, Thailand

Office: His Majesty the King’s 80th Birthday Anniversary 5th December 2007 Building (SIMR), 2nd Fl., Room No.207 Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand Tel: 02-419-2888 Fax: 02-411-0593 E-mail: sijournal92@gmail.com

Systematic Review and Meta-Analysis on Role of Adiponectin to Leptin Ratio in Women with Polycystic Ovarian Syndrome

Archie Fontana Iskandar, M.D., Nicolas Daniel Widjanarko, M.D., Felicia Grizelda Suryatenggara, M.D., Leonardo Leonardo, M.D., Rosalia Sylfiasari, M.D., Nadhea Debrinita Surya, M.D., Christian Ardianto, M.D.

Faculty of Medicine and Health Sciences, Atma Jaya Catholic University of Indonesia.

ABSTRACT

Objective: PCOS or Polycystic Ovarian Syndrome, a multifaceted disorder marked by disruptions in endocrine and metabolic processes, influences reproductive age women. The most commonly used criteria for diagnosing this condition are the Rotterdam 2003 and the National Institutes of Health Consensus 1990 guidelines. Recent studies are currently focusing on novel biomarkers, such as adiponectin and leptin to gain deeper insights on the intricate pathophysiology of PCOS. Therefore, this review aimed to consolidate the importance of the A:L (adiponectin to leptin) ratio as a potential and promising biomarker for PCOS.

Materials and Methods: The method followed the PRISMA 2020 guidelines. Furthermore, MEDLINE, Proquest, and EBSCOhost databases were used to obtain eligible studies published up to February 2023. This study was registered in PROSPERO on April 2, 2023 with registration number CRD42023411754. ESHRE/ASRM or Rotterdam Guideline was used as the diagnosis criteria for women with PCOS. To examine the studies' heterogeneity, the I2 statistic and Cochran's Q test were utilized. Meanwhile, the evaluation on publication bias visually employed a funnel plot and was confirmed through Egger's test and rank correlation test. Data analysis was conducted with JASP 0.17.1, and statistical significance was characterized as a p-value below 0.05.

Results: In the systematic review, a total of nine studies were incorporated, and seven studies were used in the

subsequent meta-analysis. Each paper showcased a reduced A:L ratio in women with PCOS, with a standardized mean difference (SMD) among PCOS and control groups of 0.49 (CI: 0.37 - 0.61). The residual heterogeneity test yielded a p-value of 0.069, and no publication bias indication both pre and post intervention (p=0.002).

Conclusion: Referring to the findings, the A:L ratio was notably lower in PCOS patients. Consequently, the A:L

ratio holds promise as a novel and potential biomarker for PCOS.

Keywords: Polycystic ovarian syndrome; adiponectin to leptin ratio; biomarker (Siriraj Med J 2023; 75: 838-850)

INTRODUCTION

Polycystic Ovarian Syndrome (PCOS) stands as a familiar and diverse endocrine disruption that impacts reproductive age women. This condition significantly impacts the endocrine, reproductive, and metabolic systems.1 In numerous studies, the occurrence differs based on the particular used criteria for diagnosis. The

worldwide prevalence can vary, spanning from 4% to 21% when considering the Rotterdam 2003 criteria and the National Institutes of Health Consensus 1990 criteria.2 Over the years, PCOS has shown a rapidly increasing trend and is frequently associated with abdominal adiposity, infertility, obstetrical problems, and cardiovascular disease, leading to a deprived life quality.3

Corresponding author: Nicolas Daniel Widjanarko E-mail: nicolaswidjanarko310@gmail.com

Received 20 September 2023 Revised 21 October 2023 Accepted 24 October 2023 ORCID ID:http://orcid.org/0000-0003-4093-0782 https://doi.org/10.33192/smj.v75i12.265452

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

Recent analysis revealed that this disease is linked to a range of metabolic disorders, emphasizing the need for further in-depth investigation to explore the relationship between the two entities.4 This field of study aims to uncover novel biomarkers to provide enhanced comprehension of the intricate pathophysiological mechanisms at play in PCOS, including factors like adiponectin (A) and leptin (L).5 These adipokines have been associated to a variety of metabolic disorders, such as obesity, insulin resistance, hyperandrogenism, as well as dyslipidemia.6 Leptin, originating from adipose tissue, serves as

an indicator of body fat levels and plays role in energy homeostasis and insulin resistance.7 Adiponectin is additionally generated by adipocytes, controlling lipid metabolism and glucose uptake. Furthermore, it functions by increasing hepatic glycolysis and fatty acid oxidation, as well as decreasing gluconeogenesis.8 Numerous investigations have documented the connection between leptin and adiponectin and their relevance to cardiometabolic conditions, suggesting that the A:L ratio could act as a predictor of metabolic risk factors.9 A:L ratio also has better diagnostic accuracy in identifying the risk of insulin resistance compared to these adipokines alone.10 In PCOS patients, it has shown potential as a promising biomarker for diagnosis, as PCOS shares several common metabolic disturbances.11 Despite previous reports have investigated the connection between the A:L ratio and PCOS, a better understanding of these adipokines in the course of disease is still needed. Findings also show fewer studies on the significance of the A:L ratio as a biomarker, thus we performed a meta-analysis to synthesize its magnitude and obtain the effect size (ES). Therefore, this article summarizes the final results of the A:L ratio as a promising reliable biomarker for PCOS.

MATERIALS AND METHODS

The design was planned and executed as the guideline in the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) 2020 protocol.12

Variable of interest and aim of the study

This study examines the differences in the A:L ratio between PCOS and non-PCOS groups as a novel diagnosis biomarker.

Eligibility criteria

This article encompassed all previously published observational studies that examined the A:L ratio's role in marking inflammation and insulin resistance in PCOS cases.

All articles published dating up to 2023 were also

Original Article SMJ

included. Furthermore, the 2003 European Society for Human Reproduction and Embryology and American Society for Reproductive Medicine (ESHRE/ASRM or Rotterdam Guideline)13,14 were used as diagnosis criteria for women with PCOS. Studies falling under categories, such as review, case reports, case series, conference abstracts, book sections, commentaries/editorials, and studies consisting of non-human subjects were omitted. Articles lacking complete text and those unrelated to the relevant subject matter were also disregarded.

Literature search and information sources

The PRISMA 2020 guidelines were used during the literature search. MEDLINE, Proquest, and EBSCOhost search engines were used to find relevant cases, dating up to February 2023. All studies obtained were screened for duplicates, while the titles and abstracts were independently reviewed by the authors. Furthermore, the studies were omitted from consideration if the titles and/or abstracts did not align with the focus of this review. The full text was read by three authors and the eligible papers were included in this review. Conflicts were solved by consensus and the opinion of all reviewers.

Data collection process

Six reviewers individually carried out the process of data extraction. Any disagreements that arose were resolved through common consent among the reviewers and, when necessary, by seeking the input of the sixth reviewer. Comprehensive data was documented concerning the name of the primary author, study year, the country, the study's type, demographic characteristic of the patients (number of participants and their age), PCOS diagnosis criteria, population matching, adjusted confounding factors/exclusion criteria, and their outcome of interest. For bivariate data extraction, yielded studies were further classified based on the L:A or A:L ratio from each group (PCOS and non-PCOS). This primary outcome was measured by mean difference (MD) as the parameter of ES.

Data and outcome measures

Adiponectin and leptin serum level was presented as μg/L or ng/mL, consecutively. Values were displayed as mean ± standard deviation (SD) for normally distributed data, or as median (interquartile range) for not normally distributed data. A:L ratio was reported or calculated by dividing the amount of adiponectin by the amount of leptin, leading to a single numerical value of ratio. When the study reported the L:A ratio, the value was reversed to obtain the A:L ratio.

Synthesis of result and summary measures

The main outcomes were tabulated in univariate and bivariate results. Data regarding the number of participants and patient’s age were included as demographic characteristics in each study. The types of study, PCOS diagnosis criteria, population matching, and other possible outcome biases were also presented in an univariate table. These baseline characteristics were further discussed to elaborate on the confounding factors that could interfere with the results of the A:L ratio. The bivariate table presents the A:L ratio for each group, and also a p-value and 95% Confidence Interval (CI).

Quality assessment

Each article underwent evaluation utilizing the Newcastle-Ottawa Scale (NOS) for case-control studies, and an adjusted version of NOS was employed for cross- sectional studies.15,16 These tools consisted of three main domains, namely (a) Selection, (b) Comparability, (c) Outcome/exposure. According to the NOS for case- control studies, the quality was classified as follows:

(1) Good (3-4 stars in the selection domain, 1-2 stars in comparability, and 2-3 stars in exposure), (2) Fair (2 stars in selection, 1-2 stars in comparability, and 2-3 stars in exposure), and (3) Poor (0-1 star in selection or 0 stars in comparability or 0-1 stars in exposure). In relation to the modified NOS for cross-sectional studies, the categorization of study quality was conducted as outlined below: (1) Very good (9-10 stars), (2) Good

(7-8), (3) Satisfactory (5-6), (4) Unsatisfactory (0-4). Each study was independently assessed by two authors. Any discrepancies were resolved through deliberation among all the authors until an unanimous agreement was achieved. The findings of the assessment of study quality are displayed in Table 1.

Data synthesis and statistical analysis

The standardized mean difference (SMD) and the 95% CI of the A:L ratio in participants of all included studies was calculated using the “Meta-Analysis Effect Size Calculator”.17 The standard error (SE) was obtained by dividing the length of the CI by 3.92. The ES and SE were then entered into a Microsoft Excel spreadsheet and used to perform meta-analysis. I2 statistic and Cochran's Q test were employed to examine the heterogeneity among articles. In cases where heterogeneity was identified (p < 0.10 or I2 > 75%)18,19, It was used the random effects model for the analysis, and generated a forest plot to illustrate the combined effect size. Subsequently, a visual assessment on the publication bias was performed through a funnel plot. The statistical confirmation was accomplished

by conducting a rank correlation and Egger's tests. The trim and fill method16 was employed to correct the bias found. However, it was unable to ascertain the presence of publication bias when the total number of included studies was fewer than 10. All analyses were conducted using JASP 0.17.1 and p <0,05 was assumed as statistically significant.

RESULTS

Literature search

Fig 1 presents a flowchart summarizing the study selection process and its outcomes. Furthermore, the searches strategy yielded 561 potentially relevant studies. Following the specified criteria for selection, a total of 12 studies were pinpointed for a more thorough full- text evaluation, one was considered not suitable, and another two had no control group for comparison. In total, nine papers were involved in systematic review and seven studies eligible for data extraction were used in meta-analysis. The PCOS groups exhibited sample sizes that spanned from 31 to 241, while the control groups had sample sizes ranging from 22 to 216.

Characteristics of included studies

A total of nine studies, consisting of 1,598 women (865 PCOS and 733 non-PCOS) met the inclusion criteria. The characteristics, including the number of participants (N), age (years), population matching, exclusion criteria, and area under the curve (AUC) represented the sensitivity and specificity of A:L ratio were extracted from each study and reported in Table 1. Furthermore, six of them were case-control studies5,20–24 and the other three were cross- sectional studies.11,25,26 All the papers used the Rotterdam ESHRE/ASRM 2003 criteria for diagnosing PCOS. The mean ages of PCOS and non-PCOS populations ranged from 23.4 ± 6.1 to 32.85 ± 4.25 years and 23.9 ± 3.6 to

32.5 ± 4.9 years, respectively. The results showed that only five studies5,11,20,22,26 had carried out age population matching and two studies11,26 had performed BMI-matching population. Among the nine studies, two came from India5,11, one from Bangladesh24, one from Ghana21, two from Bahrain22,23, one from Australia25, one from Italy26 and one from Brazil.20

Quality Assessment

Quality assessment for obtained studies were performed using the NOS for case-control (Table 2A) and cross- sectional studies (Table 2B). A total of three, five, and one papers were in the very good,11,25,26 good,5,20–23 and fair quality categories, respectively.24

In the nine studies included in the qualitative

Original Article SMJ

TABLE 1. Study Characteristics.

thyroid, or lung disease, or severe mental illness.

https://he02.tci-thaijo.org/index.php/sirirajmedj/index Volume 75, No.12: 2023 Siriraj Medical Journal 841

Iskandar et al.

TABLE 1. Study Characteristics. (Continue)

*Area Under the Curve

Fig 1. PRISMA flow diagram of involved studies.

TABLE 2A. Risk of bias assessment for case-control studies.

Note: S1: adequate case definition, S2: case representativeness, S3: control selection, S4: control definition, C1: Comparability, E1: Exposure Ascertainment, E2: identical method of ascertainment for cases and controls, and E3: non-Response Rate.

TABLE 2B. Bias assessment risk for cross-sectional studies.

Note: S1: sample representativeness, S2: sample size, S3: non-respondents, S4: exposure ascertainment, C1: comparability, O1: assessment of outcome, and O2: statistical test.

synthesis, the A:L ratio exhibited a significant decrease in PCOS patients, with p-values of <0.0001 to 0.05. The meta-analysis findings from the seven incorporated studies were visualized in both forest and funnel plots (Figs 2A and 2B). The accumulation diagrams of forest plots showed the SMD in each study and the ultimate ES derived from the amalgamation of all the studies. As displayed by the figure, the ultimate weighting of the combined value was depicted in the form of a rhombus shape, while a square shape indicated a weight for each study. The dimensions of each square were determined based on the study's weight in the meta-analysis and calculated referring to the population samples.

Sarray et al.,22 constituted the larger proportion of samples (N= 457) and Mohana et al.,24 accounted for

the smallest proportion (N=40). All studies had an ES favoring PCOS groups, which was marked by the positive ES on the right side of the plot. A study by Shorakae et al.,25 had a 95% CI border coinciding with the vertical axis on the left. Based on meta-analysis results, SMD between PCOS and non-PCOS groups was 0.49 (CI: 0.37 - 0.61). Table 5 showed the model for meta-analysis with each Q-statistics and their p-values.

The test results of residual heterogeneity showed a p-value of 0.069. This indicated that the heterogeneity for all studies was not statistically affecting the result, leading to the selection of the fixed effect model. When heterogeneity was attributed to random variation, the random effects model was employed, and it could not be explained because the model presupposed that the

Fig 2A. Meta-analysis results for A:L Ratio (Forest Plot)

TABLE 3. Results of studies in meta-analysis.

Fig 2B. Meta-analysis results for A:L Ratio (Funnel Plot).

treatment impact was distributed over certain populations and gave each study a more equal weighting. Furthermore, these effects were varying but connected to each other. The results of the I2 test showed an estimate of 46.287% (<75%), indicating that the random effect model was not considerable.

Results for the Omnibus test of Model Coefficients showed a p-value of <0.001, and the association of the A:L ratio was considered significant. Furthermore, significant results were observed between SMD in PCOS and non-

PCOS groups (with intercept result of p<0.001), as shown in Table 4B.

The Egger test was used to assess the bias observed and the results showed no publication bias before and after the intervention (p=0.002). Regarding the correlation test for detecting asymmetry in the funnel plot, Kendall's T value was 0.524, but this correlation did not reach statistical significance (p = 0.136), as detailed in Table 4C.

TABLE 4A. Test for fixed and random effects model.

TABLE 4C. Rank correlation test and regression test for funnel plot asymmetry (Egger’s test).

Calculation on SE for both groups from all included studies was performed employing the formula below: SE = (upper CI limit – lower CI limit) / 3.92, or by calculating the square root of the calculated error variance (v). A total of six studies were inside the 95% confidence triangle, with five being located near the vertical axis with 0.49 as the symmetrical line, and one located far from the triangle at the right side of the bottom. Furthermore, only one study was located outside the triangle area. The funnels plot diagram showed that the distribution of studies was asymmetry, with 1 study lying outside the triangle area with SE between 0.05 to 0.1. Others were plotted inside the triangle area, with three studies having an SE between

0.1 to 0.2, two had values between 0.2 to 0.3, and one was between 0.3 to 0.35. The higher the SE, the lower the position of the papers in the inverted funnel, indicating low power compared to others. Accompanying the plot was the 'Rank Correlation Test' for assessing funnel plot

asymmetry, and the results showed non-significance with a p-value of 0.136.

DISCUSSION

Meta-analysis results suggested that PCOS patients had a lower A:L ratio, with a statistically significant ES (MD=0.49, 95%CI: 0.37 -0.61, p<0.001). Similarly, Lin

et al.,27 found that in patients with PCOS, adiponectin levels fell statistically significant, but leptin levels rose significantly. Leptin was a hormone derived from adipocytes and encoded by the human obese (ob) gene, which regulated glucose homeostasis. Several studies had recognized its role in several metabolic and endocrine disorders, including PCOS. Prolonged elevation of circulating leptin levels in obese individuals can lead to the development of resistance and decreased receptor sensitivity to leptin.28 Decreased leptin sensitivity could lead to excessive triglyceride accumulation in multiple

organs and impaired insulin sensitivity.29 Moreover, the expression of leptin receptors had also been documented in granulosa cells, indicating that it had direct regulatory action in maintaining ovarian folliculogenesis. Leptin resistance could alter the process and this explained the contribution of leptin in PCOS pathophysiology.30

Based on two previous meta-analyses,31,32 increased concentrations of leptin were linked with an increased probability of developing PCOS compared to control participants. According to previous studies, higher levels were also linked to an increased risk of insulin resistance, metabolic disturbance, and cardiovascular disease, all of which could contribute to the development of PCOS. Leptin was assumed to have a stimulatory impact on LH secretion while exerting an inhibitory influence on the actions of FSH and insulin-like growth factor (IGF)-1.33 Moreover, hyperleptinemia conditions can inhibit ovarian response to gonadotropin stimulation. Based on these findings, raised leptin levels could potentially disrupt hormonal balance and ovarian function in women, elucidating its contribution to PCOS pathophysiology.34 As analyzed by Zheng et al.,32 hormone levels were somewhat elevated in non-obese PCOS patients than controls with matching BMI, but this difference was not statistically significant. Hence, leptin levels were elevated in individuals with PCOS, regardless of their BMI.

Adiponectin hormone released by adipose tissue contributes to mitigating atherogenic harm and insulin resistance, and it also has an effect on other tissues. Given adiponectin's insulin-sensitizing properties, its reduced levels in obesity, and the capacity of testosterone to diminish adiponectin levels,34 one could hypothesize that women with PCOS might have lower concentrations.34 Among non-PCOS women, the levels exhibited a decline as BMI increased.36 Meanwhile, in women with PCOS, there was an independent impact on insulin sensitivity, which was not associated with obesity and had been observed in lean individuals as well (BMI < 25 kg/m2).37 Adiponectin receptor 1 (adipoR1) and adiponectin receptor 2 (adipoR2) have been discovered as the membrane receptors responsible for mediating its glucose-lowering and anti-inflammatory actions. Previous research indicated that the expression of these receptors was diminished in individuals with obesity. Nevertheless, in women with PCOS, there was an upregulation observed in both visceral fat and subcutaneous tissue.35

According to meta-analysis by Toulis et al.,38 adiponectin was reduced in PCOS women in contrast with control group with a comparable BMI. The levels were descending in obese individuals with PCOS than those without. Also, it was related to insulin sensitivity,

where patients with higher insulin resistance had more inferior amounts of the hormone.

The A:L ratio has been considered as an indication of adipose tissue malfunction and a potentially valuable biomarker for metabolic disorders.39 Insulin resistance (IR) had been associated with a decrease in A:L ratio. Even though IR is considered an intrinsic feature to PCOS, yet it is not included in most established diagnostic criteria. Therefore, A/L ratio come up as a promising novel biomarker for IR in PCOS patients, regarding their independent association with PCOS. Besides, it might help to identify PCOS women who are at risk of developing IR, which marked by an alteration in the A:L ratio and subsequently be used as a screening tools for the risk factor of IR in PCOS.11

Several studies showed that A:L ratio considered as the best predictor of IR in PCOS, compared to other adipokines (adiponectin (A) alone, leptin (L) alone, resistin (R) alone, or L:R ratio).22,39,40 Supported by its AUC values in the ROC curve, the A:L ratio exhibited superior sensitivity and specificity in five studies (with AUC values of 0.78711, 0.868,24 0.83021, 0.861,23 and 0.65022).

An AUC between 0.8 and 0.9 as reported in three of the studies21,23,24 indicated a good diagnostic performance.41 The ROC curve was unaffected by disease prevalence since it depended on the sensitivity and specificity. Therefore, samples could be calculated independent of the incidence of disease among the general population.42

For decades, IR is considered as a major factor for the development of the metabolic syndrome.43 With the purpose to prevent further metabolic and cardiovascular complications, A:L ratio emerge as a propitious biomarker of IR, thus generate a requirement of an accurate and personalized clinical assessment. Treatment plans should be customized for each PCOS woman based on her risk profile for metabolic disease, as well as any complaints she may have regarding infertility, hirsutism, or menstruation disturbances.44

Heterogeneity and publication bias analysis

The I2 test for heterogeneity showed a value of 46.287%, which could be classified as “represent moderate heterogeneity”.45 Considering the low to moderate heterogeneity of this result, further subgroup analyses were not necessary. The heterogeneity of the result could be observed from clinical, methodological, or statistical perspectives. From a clinical perspective, the differences in participants or outcomes could lead to high levels. This study involved 40 to 457 women as participants. The larger the sample, the more likely or unlikely the ES could occur. From a methodological

perspective, the differences in study design and population matching could lead to high heterogeneity. Apart from the matching effort, all studies excluded those with pregnancy, diabetes, liver disease, renal dysfunction, history of oral contraceptives, hormonal imbalance (Cushing syndrome, androgen deficiency), and all medical history encompassing the use of medications such as hormone replacement therapy, as well as any history of viral, bacterial, allergic, or inflammatory conditions. Considering the possibility of controlling the confounding factors at the beginning of the study, heterogeneity from methodological perspectives was unlikely. Ultimately, from a statistical perspective, variation in intervention effects or results contributed to increased heterogeneity. In this systematic review, three studies reported the L/A ratio and four other studies reported the A:L ratio. Four studies reported ratio in median and interquartile range values,21,22,25 and three other papers reported in mean and SD values. Despite the various reporting results, all studies were calculated using SMD in meta-analysis, thus minimizing the statistical heterogeneity.

The source of potential bias in this study included publication bias, where there was a tendency among authors and publishers to predominantly release papers that showcased significant results. In this systematic review, results for the A:L ratio driven from all papers were reportedly significant. Other sources of bias included data variances among methodological designs, where two and five studies had cross-sectional and case-control designs.

Strengths and limitations of the study

Our review had several strengths, including being the first meta-analysis that investigated the potential role of the A:L ratio in determining the presence of inflammation in PCOS. This review also included all studies that either reported an A:L ratio or an inverted L:A ratio. All studies were considered to have good quality, implying more reliable conclusions. The results varied based on the difference in the A:L ratio in PCOS and non-PCOS groups, with several papers being reported in mean and median. These varying results could be solved by normalizing the data distributions to obtain more parametric findings. Several confounding variables, particularly insulin resistance and BMI could not be adjusted. This condition could potentially influence the outcome and exposure relationship, leading to a discrepancy in the summary ES.

Future directions

A threshold for A:L ratio had been proposed, with values

of >1.0, 0.5-1, and <0.5 being deemed normal, moderately increased risk, and severe increase in cardiometabolic risk, respectively.39 Furthermore, there was no established cut-off value for the A:L ratio as a biomarker for PCOS. Future studies with larger populations and interventional study designs are advised to depict the appropriate values of the A:L ratio limit in PCOS patients.

CONCLUSION

The findings revealed a significant reduction in the A:L ratio among PCOS patients than non-PCOS individuals. Consequently, the A:L ratio holds promise as a potential novel biomarker for PCOS in the forthcoming.

ACKNOWLEDGEMENTS

Acknowledgments are delivered to all partners from Atma Jaya Catholic University of Indonesia for the support and contributions provided.

Conflict of interest

No conflict of interest in this study.

Funding

There was no specific grant or funding from any institutions or sponsors.

Registration of review protocol

This review was registered in PROSPERO on April 2nd, 2023 with the registration number CRD42023411754.

REFERENCES

1. Louwers YV, Laven JSE. Characteristics of polycystic ovary syndrome throughoutlife.TherAdvReprodHealth.2020;14:2633494120911038.

2. Chang S, Dunaif A. Diagnosis of Polycystic Ovary Syndrome: Which Criteria to Use When? Endocrinol Metab Clin North Am. 2021;50(1):11-23.

3. Shi N, Ma Hbo. Global trends in polycystic ovary syndrome research: A 10-year bibliometric analysis. Front Endocrinol. 2023;13:1027945.

4. Chen W, Pang Y. Metabolic Syndrome and PCOS: Pathogenesis and the Role of Metabolites. Metabolites. 2021;11(12):869.

5. Gupta V, Mishra S, Mishra S, Gupta V. L:A ratio, Insulin resistance and metabolic risk in women with polycystic ovarian syndrome. Diabetes Metab Syndr. 2017;11 Suppl 2:S697-701.

6. Zorena K, Jachimowicz-Duda O, Ślęzak D, Robakowska M, Mrugacz M. Adipokines and Obesity. Potential Link to Metabolic Disorders and Chronic Complications. Int J Mol Sci. 2020;21(10): 3570.

7. Picó C, Palou M, Pomar CA, Rodríguez AM, Palou A. Leptin as a key regulator of the adipose organ. Rev Endocr Metab Disord. 2022;23(1):13-30.

8. Choi HM, Doss HM, Kim KS. Multifaceted Physiological Roles of Adiponectin in Inflammation and Diseases. Int J Mol Sci. 2020;21(4):1219.

9. Frühbeck G, Catalán V, Rodríguez A, Ramírez B, Becerril S,

   Salvador J, et al. Adiponectin-leptin Ratio is a Functional Biomarker of Adipose Tissue Inflammation. Nutrients. 2019; 11(2):454.

10. Agostinis-Sobrinho C, Vicente SE de CF, Norkiene S, Rauckienė- Michaelsson A, Kievisienė J, Dubey VP, et al. Is the Leptin/ Adiponectin Ratio a Better Diagnostic Biomarker for Insulin Resistance than Leptin or Adiponectin Alone in Adolescents? Child Basel Switz. 2022;9(8):1193.

11. Mishra P, Mittal P, Rani A, Bharti R, Agarwal V, Suri J. Adiponectin to Leptin Ratio and its Association with Insulin Resistance in Women with Polycystic Ovarian Syndrome. Indian J Endocrinol Metab. 2022;26(3):239-44.

12. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Syst Rev. 2021;10:89.

13. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril. 2004;81(1):19-25.

14. Wang F fang, Pan J xue, Wu Y, Zhu Y hang, Hardiman PJ, Qu F. American, European, and Chinese practice guidelines or consensuses of polycystic ovary syndrome: a comparative analysis. J Zhejiang Univ Sci B. 2018;19(5):354-63.

15. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol. 2010;25(9):603-5.

16. Mamikutty R, Aly AS, Marhazlinda J. Selecting Risk of Bias Tools for Observational Studies for a Systematic Review of Anthropometric Measurements and Dental Caries among Children. Int J Environ Res Public Health. 2021;18(16):8623.

17. Effect Size Calculator [Internet]. [cited 2023 May 18]. Available from: https://www.campbellcollaboration.org/escalc/ html/EffectSizeCalculator-Home.php

18. Mikolajewicz N, Komarova SV. Meta-Analytic Methodology for Basic Research: A Practical Guide. Front Physiol. 2019;10:203.

19. Chapter 10: Analysing data and undertaking meta-analyses [Internet]. [cited 2021 Jan 26]. Available from: /handbook/ current/chapter-10

20. Lecke SB, Mattei F, Morsch DM, Spritzer PM. Abdominal subcutaneous fat gene expression and circulating levels of leptin and adiponectin in polycystic ovary syndrome. Fertil Steril. 2011;95(6):2044-9.

21. Obirikorang C, Owiredu WKBA, Adu-Afram S, Acheampong E, Asamoah EA, Antwi-Boasiakoh EK, et al. Assessing the variability and predictability of adipokines (adiponectin, leptin, resistin and their ratios) in non-obese and obese women with anovulatory polycystic ovary syndrome. BMC Res Notes. 2019; 12(1):513.

22. Sarray S, Madan S, Saleh LR, Mahmoud N, Almawi WY. Validity of adiponectin-to-leptin and adiponectin-to-resistin ratios as predictors of polycystic ovary syndrome. Fertil Steril. 2015; 104(2):460-6.

23. Golbahar J, Das NM, Al-Ayadhi MA, Gumaa K. Leptin-to- adiponectin, adiponectin-to-leptin ratios, and insulin are specific and sensitive markers associated with polycystic ovary syndrome: a case-control study from Bahrain. Metab Syndr Relat Disord. 2012;10(2):98-102.

24. Mohana CA, Hasanat MA, Rashid EU, Jahan IA, Morshed MS, Banu H, et al. Leptin and Leptin adiponectin ratio may be promising markers for polycystic ovary syndrome and cardiovascular

    risks: Leptin and LAR in PCOS. Bangladesh Med Res Counc Bull. 2021;47(3):266-72.

25. Shorakae S, Abell SK, Hiam DS, Lambert EA, Eikelis N, Jona E, et al. High-molecular-weight adiponectin is inversely associated with sympathetic activity in polycystic ovary syndrome. Fertil Steril. 2018;109(3):532-9.

26. Savastano S, Valentino R, Di Somma C, Orio F, Pivonello C, Passaretti F, et al. Serum 25-Hydroxyvitamin D Levels, phosphoprotein enriched in diabetes gene product (PED/ PEA-15) and leptin-to-adiponectin ratio in women with PCOS. Nutr Metab. 2011;8:84.

27. Lin K, Sun X, Wang X, Wang H, Chen X. Circulating Adipokine Levels in Nonobese Women With Polycystic Ovary Syndrome and in Nonobese Control Women: A Systematic Review and Meta-Analysis. Front Endocrinol [Internet]. 2021 [cited 2023 May 19];11. Available from: https://www.frontiersin.org/articles/10.3389/ fendo.2020.537809

28. Sitticharoon C, Klinjampa R, Souvannavong-Vilivong X, Chatree S, Boonpuan P, Sripong C, et al. Serum Neuropeptide Y and Leptin Levels compared between Non-pregnant and Pregnant Women in Overall, Non-obese, and Obese Subjects. Siriraj Med J. 2018;70(3):204-12.

29. Obradovic M, Sudar-Milovanovic E, Soskic S, Essack M, Arya S, Stewart AJ, et al. Leptin and Obesity: Role and Clinical Implication. Front Endocrinol. 2021;12:585887.

30. Wołodko K, Castillo-Fernandez J, Kelsey G, Galvão A. Revisiting the Impact of Local Leptin Signaling in Folliculogenesis and Oocyte Maturation in Obese Mothers. Int J Mol Sci. 2021;22(8):4270.

31. Seth MK, Gulati S, Gulati S, Kumar A, Rawat D, Kumari A, et al. Association of Leptin with Polycystic Ovary Syndrome: a Systematic Review and Meta-Analysis. J Obstet Gynaecol India. 2021;71(6):567.

32. Zheng SH, Du DF, Li XL. Leptin Levels in Women With Polycystic Ovary Syndrome: A Systematic Review and a Meta- Analysis. Reprod Sci Thousand Oaks Calif. 2017;24(5):656-70.

33. Childs GV, Odle AK, MacNicol MC, MacNicol AM. The Importance of Leptin to Reproduction. Endocrinology. 2020;162(2):bqaa204.

34. Houjeghani S, Pourghassem Gargari B, Farzadi L. Serum Leptin and Ghrelin Levels in Women with Polycystic Ovary Syndrome: Correlation with Anthropometric, Metabolic, and Endocrine Parameters. Int J Fertil Steril. 2012;6(2):117-26.

35. Michalakis KG, Segars JH. The role of adiponectin in reproduction: from polycystic ovary syndrome to assisted reproduction. Fertil Steril. 2010;94(6):1949-57.

36. Sitticharoon C, Souvannavong-Vilivong X, Klinjampa R, Churintaraphan M, Nway NC, Keadkraichaiwat I, et al. Serum Adiponectin, Visfatin, and Omentin Compared between Non- pregnant and Pregnant Women in Overall, Non-obese, and Obese subjects. Siriraj Med J. 2018;70(3):219-26.

37. Svendsen PF, Nilas L, Nørgaard K, Jensen JEB, Madsbad S. Obesity, body composition and metabolic disturbances in polycystic ovary syndrome. Hum Reprod Oxf Engl. 2008;23(9): 2113-21.

38. Toulis KA, Goulis DG, Farmakiotis D, Georgopoulos NA, Katsikis I, Tarlatzis BC, et al. Adiponectin levels in women with polycystic ovary syndrome: a systematic review and a meta- analysis. Hum Reprod Update. 2009;15(3):297-307.

39. Frühbeck G, Catalán V, Rodríguez A, Gómez-Ambrosi J. Adiponectin-leptin ratio: A promising index to estimate adipose tissue dysfunction. Relation with obesity-associated cardiometabolic

    
    

    risk. Adipocyte. 2018;7(1):57-62.

40. Yun JE, Won S, Mok Y, Cui W, Kimm H, Jee SH. Association of the leptin to high-molecular-weight adiponectin ratio with metabolic syndrome. Endocr J. 2011;58(9):807-15.

41. Nahm FS. Receiver operating characteristic curve: overview and practical use for clinicians. Korean J Anesthesiol. 2022;75(1): 25-36.

42. Hajian-Tilaki K. Receiver Operating Characteristic (ROC) Curve Analysis for Medical Diagnostic Test Evaluation. Casp J Intern Med. 2013;4(2):627-35.

43. Sangaraju SL, Yepez D, Grandes XA, Talanki Manjunatha R,

    Habib S. Cardio-Metabolic Disease and Polycystic Ovarian Syndrome (PCOS): A Narrative Review. Cureus. 2022;14(5):e25076.

44. Spritzer PM. Primary and secondary prevention of metabolic and cardiovascular comorbidities in women with polycystic ovary syndrome. Rev Bras Ginecol E Obstet Rev Fed Bras Soc Ginecol E Obstet. 2015;37(1):1-4.

45. van Doorn J, van den Bergh D, Böhm U, Dablander F, Derks K, Draws T, et al. The JASP guidelines for conducting and reporting a Bayesian analysis. Psychon Bull Rev. 2021;28(3): 813-26.

A Systematic Review and Meta Analysis of Non-Randomized Interventional Studies on the

Pamidronate Treatment Efficacy for Patients with Bone Fibrous Dysplasia

Nicolas Daniel Widjanarko, M.D., Anthony Ekaputra, M.D., Jessica Felicia Ang, M.D.

Faculty of Medicine and Health Sciences, Atma Jaya Catholic University of Indonesia.

ABSTRACT

Objective: Pamidronate is one of the main therapies for Fibrous Dysplasia (FD), with documented enhancements in patients' clinical characteristics. Nevertheless, its usage has yielded inconclusive results. Therefore, this review aimed to investigate pamidronate’s impact on several clinical and biochemical outcomes in FD patients.

Materials and Methods: This review was conducted under the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. MEDLINE, ProQuest, Wiley, and EBSCO search databases were used to search the literature. Risk of Bias In Non-Randomized Studies of Interventions (ROBINS-I) was applied for quality assessment of the included studies and Review Manager (RevMan) 5.4 was employed in performing the meta-analysis. Results: There are eight and seven studies used in the meta-analysis and systematic review, respectively. The results showed there are two studies with a low risk of bias and six with a moderate category. All papers included in this meta-analysis showed significant differences in the reduction of bone pain (p<0.00001) and serum alkaline phosphatase (SAP) (p=0.04) after pamidronate treatment compared to the before-treatment groups.

Conclusion: The findings of this study indicated that pamidronate therapy had been proven to significantly reduce

bone pain and increase SAP in FD patients. However, trials with more age-specific samples and a lower risk of bias should be carried out to determine the statistical significance of overall results.

Keywords: Fibrous dysplasia; pamidronate; bone pain; serum alkaline phosphatase; bone mineral density (Siriraj Med J 2023; 75: 851-863)

INTRODUCTION

Fibrous Dysplasia (FD) is a rare skeletal disorder characterized by an abnormal fibrous tissue development within bone which leads to deformities and functional impairments. This condition exhibits an estimated prevalence of approximately 1 in 20,000 individuals worldwide.1 Furthermore, FD typically manifests during childhood or adolescence, with the majority of cases being diagnosed before the age of 30.2 The common

manifestations are bone pain, skeletal deformities, and an increased propensity for fractures. Craniofacial FD can lead to facial asymmetry and vision/hearing impairments, while long bone involvement often causes limb-length discrepancies and pathologic fractures.3 This indicates that early intervention is essential in managing FD-related complications and preventing functional disabilities. Several studies have also identified various timely treatment strategies, such as surgical interventions

Corresponding author: Nicolas Daniel Widjanarko E-mail: nicolaswidjanarko310@gmail.com

Received 21 September 2023 Revised 21 October 2023 Accepted 26 October 2023 ORCID ID:http://orcid.org/0000-0003-4093-0782 https://doi.org/10.33192/smj.v75i12.265453

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

and pharmacological therapies, which aim to alleviate pain, mitigate bone deformities, and enhance the overall quality of life for affected individuals.4

Bisphosphonates are therapeutically valuable analogs of naturally occurring pyrophosphate. These compounds find utility in conditions related to accelerated bone resorption, like Paget's disease, bone metastases, osteoporosis, and FD. Furthermore, previous reports have shown that chemical alterations on their side chain can substantially increase their efficacy.5 Bisphosphonates have a complex molecular structure for their action, with a cumulative inhibitory impact on pathologic bone loss. These compounds prevent bone resorption by being absorbed and transported to the mineral surfaces which interferes with osteoclast function. Bisphosphonates also exert an anti-osteoclast effect by inducing osteoblasts to create an osteoclast inhibitory factor, thereby inhibiting bone resorption and stimulating formation.6

Pamidronate, a biphosphonate, was derived from inorganic pyrophosphate (PPi) and exhibits a strong affinity for hydroxyapatite crystals found in bone remodeling zones. Furthermore, pamidronate as a second-generation bisphosphonate includes R-2 side chains that contain nitrogen, compared to early variants. The key distinction lies in the farnesyl pyrophosphate synthase inhibition, an enzyme that is expressed in mammalian cells and is essential for the formation of lipids. This indicates that cellular apoptosis only appears in osteoclasts, due to the capacity of the compound to be maintained in bone before endocytosis during osteoclast-mediated matrix digestion and bone mineral breakdown.7 Pamidronate also has intravenous (IV) preparations, thereby reducing the number of doses as well as gastrointestinal side effects experienced by certain patients treated with oral bisphosphonates.8

Pamidronate as one of the primary therapy for FD, has been shown to develop the patients’ clinical features by reducing pain, bone turnover markers, and raising bone mineral density (BMD). Furthermore, its intravenous administration for polyostotic and monostotic FD has also been proven to be effective in reducing fracture rates in the afflicted regions.9 BMD increase was observed after pamidronate treatment, but the response varied among patients.10 This treatment also has an initial pain-relieving effect from its direct analgesic properties, but the exact pain-relieving mechanism is still being searched for.11 The use of pamidronate contributes to the reduction of serum alkaline phosphatase (SAP). SAP level is a vital laboratory finding correlated with osteoblastic activity which is often applied as an important prognostic marker for FD and bone-related malignant tumors, such as

osteosarcoma.12 A study by Park et al., showed high SAP is related to the recurrence and progression into a more severe condition of FD, indicating that the marker can be used as a reliable predictor for the disease evolvement.13 Individualized pamidronate administration for FD management has shown inconclusive results.14 Hence, this review aims to analyze the pamidronate treatment effect on several clinical and biochemical parameters in

patients with FD.

MATERIALS AND METHODS

This review was completed under the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) 2020 statement.15

Variable of interest

This study aimed to evaluate the pamidronate treatment effect on bone pain, serum alkaline phosphatase (SAP), and bone mineral density (BMD) in patients suffering from FD.

Eligibility Criteria Types

This review consisted of both published and unpublished studies examining the impact of pamidronate treatment on patients diagnosed with bone FD. Meanwhile, studies falling under the categories of reviews, cross-sectional analyses, cohort investigations, case reports, case series, conference abstracts, book sections, commentaries/ editorials, and papers entailing non-human subjects were omitted. Articles lacking complete text and those unrelated to the relevant subject matter were disregarded.

Participants

Inclusion criteria were polyostatic or monostatic FD patients with normal renal and hepatic function, who were not taking drugs that affect bone metabolism. The FD diagnosis was according to clinical history, biochemical examination, and radiographic findings. Furthermore, there were no limitations for age, gender, and race.

Outcome of interest

The interest outcome of this study was the bone pain assessment for patients receiving pamidronate treatment, which was measured and reported as proportion (%) data, while SAP and BMD z-scores were measured and reported in numerical (continuous) data.

Search strategy and study selection

MEDLINE, EBSCO-Host, Wiley, and ProQuest electronic databases were used to search for eligible

studies by the year 2023. Furthermore, EBSCO-Host and ProQuest databases were screened for grey literature to identify potential unpublished studies with suitable PICO criteria. Papers were identified by three independent authors by using the following keywords: (("polyostotic fibrous dysplasia"[All Fields] OR “bone fibrous dysplasia” [All Fields] OR “bone dysplasia”[All Fields]) AND ("pamidronate"[All Fields]) AND ("bone mineral density"[All Fields] OR “bone pain”[All Fields] OR “serum alkaline phosphatase”[All Fields])).

All acquired studies were imported into the Mendeley reference manager program. They were then checked for duplication, followed by titles and abstracts screening. The papers were assessed separately by the authors and eliminated if the title and/or abstract were not appropriate for this review. The selected papers were reviewed in full-text assessment using the aforementioned eligibility criteria, and suitable articles were included in the review. The differences observed were settled among the review team members.

Data collection process

The included studies were analyzed and the data extracted were first author, country, design, sample size, baseline characteristics (age and sex), FD diagnosis criteria, FD severity assessment, Pamidronate treatment protocol, Pamidronate regimen follow-up duration, and pamidronate treatment side effects.

Summary measures

The numerical data were in the form of mean ± standard deviation (SD) for normally distributed, or median (interquartile range) for non-normally distributed. The p-value and 95% Confidence Interval (CI) were also included for each item to show the results’ significance. A p-value less than or equal to 0.05 was statistically significant.

Assessment of risk bias/ Quality assessment

Each study was evaluated using the Cochrane Risk of Bias In Non-randomized Studies of Interventions (ROBINS-I) for non-randomized controlled trials. The tool consisted of seven main domains, which were divided into 3 main categories: (1) Pre-intervention, consisting of (a) Bias because of confounding, (b) Bias in the participants’ selection, (2) intervention, consisting of (c) Bias in interventions classification, (3) Post-intervention, comprising (d) Bias because intended interventions deviations, (e) Bias because of missing data, (f) Bias in outcomes measurement, and (g) Bias in reported result selection. From each domain, the bias risk was

considered as low, moderate, serious, critical risk, and no information. Each trial's overall quality was divided into five groups based on the degree of bias present: (1) low risk of bias (low for all domains), (2) moderate risk of bias (low or moderate for all domains), (3) serious risk (serious at least one domain, but not at critical risk in any domain), (4) critical risk (critical in at least one domain), (5) no information (lack of information in one or more key domains where judgment was needed). All authors separately evaluated each article, and any disagreements were addressed among the whole review team until agreement was obtained.

Synthesis of results and Statistical analysis

For all outcomes, the odds ratios (ORs) for bone pain and the standardized mean differences (SMDs) for SAP and BMD were calculated as the effect size. The ORs were computed based on the proportion differences between the two groups, and the SMDs were computed based on the mean changes from baseline to the end of each group. Statistical analyses were carried out for between-group comparison, before and after pamidronate treatment. For some studies, the SAP values from individual patients were added to obtain the mean value, and the SD was calculated using the SD Calculator.16 Considering that some showed primary results applying different methods, the meta-analyses were performed using a random effect model. This model presupposed that the treatment impact was distributed over certain populations and gave each study a more equal weighting. The combined effect measures of the direct comparisons from an individual intervention were compared using the inverse variance method for continuous data and the Cochran Mantel Haenszel test (CMH) for proportional data.

Heterogeneity across trials was assessed using the I2 statistic. An I2 value of < 25%, 25-50%, and >50 was placed in the low, moderate substantial, and high categories, respectively. When heterogeneity was present, possible causes were investigated through sensitivity analyses. A proportional result was obtained as weighted ORs and a continuous result as weighted SMDs. Differences across studies were calculated based on population sample sizes. Furthermore, publication bias was analyzed visually through a funnel plot, where each trial impact was plotted with its inversed SE. All studies were conducted using RevMan software version 5.4.

RESULTS

PRISMA

Fig 1 provides a selection process overview of the study and its outcomes. The initial search strategy yielded 269

Fig 1. PRISMA 2020 flow diagram of included studies.

potentially relevant studies from the keywords searching using the Medical Subject Heading (MeSH). Among these papers, 252 were selected for comprehensive full-text evaluation, and only 74 full-access articles were retrieved. Furthermore, 62 studies were excluded according to the predefined criteria, namely the same study center (n=4), the combination of other bisphosphonate treatments (n=28), and different measurements for the outcome of interest (n=34). Eight studies were included in the systematic review and seven were included in the meta- analysis which was eligible for data extraction. Among the included papers, there was no specific publication date. Unpublished studies with appropriate inclusion criteria were not found, indicating that it was likely not to affect the conclusions of the review.

Quality assessment

Among the eight studies that were evaluated using ROBINS-I, six of them6,14,17–20 had a low risk of bias, while two21,22 had a high risk. A study by Chapurlat et al.,

200414 was the continuation of Chapurlat et al., 199717 with a larger population number. According to Cochrane’s recommendations, the Robvis tool was used to summarize the bias risk (Figs 2A and 2B), which was rated as “low”, “moderate”, “serious” and “critical” across various domains.23 A total of eight studies consisting of 143 children and adults with FD fulfill the criteria. The included studies’ characteristics comprised of number of participants (N), mean age (years), sex (%), FD diagnosis criteria, FD severity assessment, treatment protocol, follow-up duration, and pamidronate treatment side effects, as shown in Table 1. All the papers were open trials with non-randomized study designs. At the time of diagnosis, the youngest age of the participant was one year and the oldest aged 63 years. Meanwhile, at the treatment onset, the youngest aged three years and the oldest was 69 years. Seven studies6,14,17,19–22 utilized clinical evaluation as the basis of FD diagnosis, and one study18 relied on radiologic findings to uphold the FD diagnosis. All papers administered pamidronate via an intravenous route with

Fig 2A. Results of risk of bias assessment within studies illustrated with the Robvis tool.

Fig 2B. Risk of bias domains summary in included studies.

a standard dose of 60mg/day or 0.5–1 mg/kgBW/day for three consecutive days. The regimen intervals ranged from four months to one year, with most of them having 6-month intervals for two years of treatment duration. The most common reported side effects included fever (hyperthermia/hyperpyrexia),6,14,17–22 bone pain,6,17–21 hypocalcemia,6,17,19–21 and osteomalacia.14,17

Final Results

The seven studies included in the quantitative synthesis showed that bone pain and SAP were lower in the after- treatment group than in the before-treatment group,

with a p-value ranging from 0.001 to 0.05. The meta- analysis results were described as a forest plot (Figs 3A and 3B) and the subsequent publication bias as a funnel plot (Figs 4A and 4B). The accumulation diagrams of forest plots showed the pooled OR for bone pain and SMD for SAP, as the final effect size obtained from the combination of all papers. Furthermore, the final weight of the combined value was shown in a rhombus shape, and a square shape indicated a weight for each study. The size of each square was determined by its weight in the meta-analysis, which was calculated based on the study population samples.

Widjanarko et al.

TABLE 1. Study Characteristics.

Original Article SMJ

TABLE 1. Study Characteristics. (Continue)

https://he02.tci-thaijo.org/index.php/sirirajmedj/index Volume 75, No.12: 2023 Siriraj Medical Journal 857

TABLE 2A. Bone Pain.

*N/A: Not available.

**considered as significant

TABLE 2B. Serum Alkaline Phosphatase (SAP).

*N/A: Not available.

**NS: Not significant.

TABLE 2C. Bone Mineral Density (BMD).

*FDa = Fibrous Dysplasia Areas

**CL = Contralateral Side

Fig 3A. Data Results in the Meta Analysis (Forest Plot) for Bone Pain Before and After Treatment of Pamidronate.

Fig 3B. Data Results in the Meta Analysis (Forest Plot) for Serum Alkaline Phosphatase (SAP) Before and After Treatment of Pamidronate.

Fig 4A. Publication Bias as Funnel Plot Diagram for Bone Pain Assessment Before and After Treatment of Pamidronate in Patients with FD.

Fig 4B. Publication Bias as Funnel Plot Diagram for Serum Alkaline Phosphatase (SAP) Before and After Treatment of Pamidronate in Patients with FD.

All data in this meta-analysis showed significant differences in bone pain and SAP reduction after pamidronate treatment compared to the before-treatment groups, as shown in Figs 3A and 3B, with a p-value of <0.00001 and

0.04 respectively. The two outcomes were considered to have a low heterogeneity, with I2 results of 0% and 17%, respectively. The publication bias represented as a Funnel Plot Diagram was shown in Figs 4A and 4B for bone pain reduction and SAP, respectively.

DISCUSSION

FD was a benign bone lesion characterized by altered osteogenesis, leading to intramedullary fibro osseous proliferation. The excessive fibro osseous tissue subsequently replaced the normal bone tissue. This phenomenon could lead to an increased risk of fracture, leading to pathologic fracture, specifically in weight- bearing bones on the lower extremities. Along with the poor FD bone quality, this could cause deformities and increased severity of pain.24 Notwithstanding, McCune Albright syndrome (MAS) was a condition, where FD simultaneously occurred with endocrine dysfunction and skin pigmentation.25

Pamidronate was a second-generation bisphosphonate that inhibited osteoclast-mediated bone resorption and was widely used as a treatment among patients with FD, osteogenesis imperfecta, cerebral palsy, and chronic neuropathy.26 The outcomes typically included clinical, radiological, and biochemical measurements for FD diagnosis. Bone pain, SAP, and BMD were the most reported outcomes to be analyzed.

An FD lesion could precipitate normal bone resorption and create a phosphaturic factor, leading to a deficiency of phosphate and consequent osteomalacia. This condition then led to lower BMD values, particularly in the trabecular-predominant bone locations.27 Treatment with bisphosphonates could prevent the process, thereby enabling rapid restoration of BMD in the affected regions. The efficacy of pamidronate to increase BMD was proven by Lee et al. when the BMD Z-scores and lumbar spine (mg/cm2) from the pamidronate group increased after the therapy during their follow-up (both p<0.001). All patients also experienced improvements in bone pain after the use of pamidronate.28 These findings were consistent with the results of this current study. Zacharin et al.,22 reported a significant increase in BMD z-score (0.82

± 0.06 to 0.94 ± 0.07, p=<0.05), and Chapurlat et al.,14 showed a significant elevation of BMD z-score (0.831 to 0.973, p=0.003) after several cycles of Pamidronate treatments.

Pain was one of the major complaints in patients

with FD. Furthermore, thinly myelinated sensory nerve fibers (A-delta) and CGRP+ nerve fibers, which were rich in peptides, were the main nerve fibers found in innervate bone. The periosteum, mineralized bone, and marrow all exhibited this pattern of innervation. Patients with FD experienced excessive bone growth, pathological bone remodeling, and ectopic sensory plus sympathetic nerve fiber sprouting in the marrow and mineralized bone. These conditions led to alterations in the sensory nerve fibers, thereby causing bone pain formation.29 Majoor et al. stated that pain was the most frequent complaint in FD patients regardless of age, bone involvement, and types of FD (monostotic or polyostotic).30

Compared to other forms of pain, FD pain could be well managed to retain functional status and life quality. The cumulative OR of 6.13 and the I2 value of 0% from the meta-analysis indicated a substantial impact of pamidronate in reducing bone pain, with no significant heterogeneity among the included studies. This suggested that the results of this study were consistent and similar, thereby increasing the reliability of the meta-analysis findings.6,14,19–21 After the first round of pamidronate therapy, pain intensity was decreased, and an additive effect was shown after numerous cycles of therapy. Lala et al., (2000) reported a gradual decrease alongside pamidronate treatment cycles. Furthermore, low pain (1–3) gone after the first therapeutic cycle, and moderate or severe pain (4–9) gone after the third cycle.21 Isaia et al. observed that only 1/11 individuals had the sensation at the end of treatment (2000), down from 3/11 patients two years before (1998), and 8/11 patients at the start of treatment cycles in 1994.19 These findings suggested that cumulative doses of pamidronate could reduce bone resorption and increase bone formation, leading to pain depletion.

Serum alkaline phosphatase (SAP) was a plasma membrane enzyme as a marker for bone formation and as a predictor for disease activity in metabolic bone diseases.31 Adenyl cyclase in FD patients was constantly active due to the GNAS 1 mutation, and this led to elevated cAMP activity, which promoted aberrant osteoblasts and hyperfunction of skeletal progenitor cells.32 Ma et al., revealed that the increase in ALP level was due to elevated levels of calcitonin through the action of cAMP.33 Bone SAP was directly inhibited by three BPs that contained nitrogen atoms in their structure (Pamidronate, Alendronate, and Zolendronate), in a time- and dose- dependent manner. SAP activity was enhanced by low dosages of BPs (10-10-10-5 M), but it was inhibited by high doses (10-4 M). Since SAP was a metalloenzyme that was dependent on zinc and magnesium, the BP-induced SAP

inhibition could be the consequence of metal chelation by the drug's phosphonate groups. This hypothesis was supported by the findings that an increase of Zn2+ or Mg2+ could reverse SAP inhibition.34

Decreased levels of SAP indicated a decline in bone turnover during the use of Pamidronate, indicating that the enzyme could be used to monitor the treatment response.35 Based on the forest plot results, pamidronate treatment significantly and positively affects SAP levels among the included studies (95%CI: 3.72 – 204.17, p=0.04). The pooled mean difference of 103.94, along with a low to moderate heterogeneity suggested a consistent treatment effect. Studies from Parisi et al., and Chapurlat et al., showed a significant reduction of SAP levels after pamidronate treatment, from 315±184 to 194±150 (p=0.05)27 and 185±176 to 151.7±460.6 (p=0.015), respectively.14 Although a marked reduction of the enzyme levels was observed in Zacharin et al., (818.5 ± 181.7 to 654.0 ± 171), the result was not significant. This occurred in some patients six months after the second pamidronate therapy round, which indicated time between pamidronate infusions needed to be shorter in more severe cases, leading to a more notable SAP reduction.22

Pamidronate treatment was associated with a variety of side effects, including acute phase response, musculoskeletal discomfort, jaw osteonecrosis, hypocalcemia, and various ophthalmic events,36 as reported in most of the included studies. However, all side effects were relieved by supportive therapies, only limited to the initial treatment, and did not persist after the second cycle of pamidronate infusion. The safety of Pamidronate treatment was also supported by Jjuszczak et al., who found that 39 out of 42 (93%) patients agreed Pamidronate is effective for NSAIDs- refractory chronic recurrent multifocal osteomyelitis in children, without reported prominent side effects.37 Demographic characteristics from each study significantly differed from others. The age in some studies varied greatly, while those in others varied considerably. The majority of cases were identified in children, while some studies focused on pediatric populations,14,17,27 and others included a wider range of ages.18–22 However, it was challenging to determine the extent of the correlation between age and FD. Hart et al, reported that clinically significant bone lesions often became visible by the age of five, and nearly no substantial lesions were developed after the age of fifteen. Adult FD lesions could become less active, due to the apoptosis of bone marrow mesenchymal stromal cells (BMSCs) carrying mutations.38 In terms of ethnicity, the correlation between FD and the patient’s race from different countries (France,14,17 Italy,19–22 Argentina,27 Canada18) was not readily evident, taking into account

that FD was generally considered to be unrelated to a specific race or ethnicity. This diversity could represent various racial groups, thereby allowing the conclusions to be generalized to broader populations.

Although GNAS mutation testing was the "gold standard" for FD diagnosis, it could be complicated by the sensitivity of the procedure and the degree of mosaicism in the afflicted tissue.39 Other diagnostic methods were standard PCR and NGS, but both had limited value for evaluating mutations.40 Therefore, several knowledgeable medical professionals could frequently rely on clinical findings and specific FD characteristics on a radiographic exam (ground-glass appearance),41 making a diagnosis without the necessity for biopsy or molecular approach, as mentioned in most of the included studies.

Some studies had adjusted confounding factors, such as disease severity by classifying using Feuillan’s score.19–21 Patients with varying degrees of disease severity could respond differently to treatment. This indicated that including disease severity as a covariate in analyses could help account for this. The variation in gender distribution across studies showed that there might be differences in the proportion of females and males among participants, but FD had an equal sex distribution throughout all populations.42

Strengths and limitations of the study

Based on previous findings, this was the first systematic review and meta-analysis on pamidronate treatment in FD of the bone, both in children and adult populations. Our research comprised several primary studies that follow the development of Pamidronate usage from time to time, from the first published study in 1997 until the latest in 2006. Thus, we have exhibited a broad view regarding this matter. An additional novelty of our studies found that SAP, aside from BMD, was proven to be significantly reduced following Pamidronate treatment, furthermore demonstrating that this enzyme might be used to surveil the effectiveness of the treatment in the future. The grey literature was also included to minimize the potential publication bias.

All studies were open and non-randomized trials. While these could provide initial insights into the potential effectiveness of pamidronate for FD, their findings were considered to be less robust compared to randomized controlled trials (RCTs). Therefore, further RCTs were required to enhance the evidence and better understand the outcomes. The small sample size and the absence of a control group can potentially undermine the robustness of our findings, as it may not have provided adequate statistical power to detect more subtle effects.

CONCLUSION

In conclusion, the reduction in bone pain and the decrease in SAP levels appeared to be favorable treatment outcomes associated with Pamidronate treatment in individuals with FD. The meta-analysis results supported the conclusion that Pamidronate was effective in improving these conditions.

ACKNOWLEDGEMENTS

The authors are grateful to all colleagues from Atma Jaya Catholic University of Indonesia for the support and contributions provided.

Conflict of interest

No conflict of interest.

Funding

No specific grant nor funding from any agencies or sponsors.

REFERENCES

1. Chapurlat RD, Meunier PJ. Fibrous dysplasia of bone. Baillieres Best Pract Res Clin Rheumatol. 2000;14(2):385-98.

2. Lietman SA, Levine MA. Fibrous dysplasia. Pediatr Endocrinol Rev PER. 2013;10 Suppl 2:389-96.

3. Berglund JA, Tella SH, Tuthill KF, Kim L, Guthrie LC, Paul SM, et al. Scoliosis in Fibrous Dysplasia/McCune-Albright Syndrome: Factors Associated With Curve Progression and Effects of Bisphosphonates. J Bone Miner Res Off J Am Soc Bone Miner Res. 2018;33(9):1641-8.

4. Collins MT, Kushner H, Reynolds JC, Chebli C, Kelly MH, Gupta A, et al. An instrument to measure skeletal burden and predict functional outcome in fibrous dysplasia of bone. J Bone Miner Res Off J Am Soc Bone Miner Res. 2005;20(2):219-26.

5. Rudge ES, Chan AHY, Leeper FJ. Prodrugs of pyrophosphates and bisphosphonates: disguising phosphorus oxyanions. RSC Med Chem. 2022;13(4):375-91.

6. Parisi MS, Oliveri B, Mautalen CA. Effect of intravenous pamidronate on bone markers and local bone mineral density in fibrous dysplasia. Bone. 2003;33(4):582-8.

7. Drake MT, Clarke BL, Khosla S. Bisphosphonates: Mechanism of Action and Role in Clinical Practice. Mayo Clin Proc Mayo Clin. 2008;83(9):1032-45.

8. Yoon JH, Choi Y, Lee Y, Yoo HW, Choi JH. Efficacy and safety of intravenous pamidronate infusion for treating osteoporosis in children and adolescents. Ann Pediatr Endocrinol Metab. 2021;26(2):105-11.

9. Brendan Chan, Margaret Zacharin. Pamidronate Treatment of Polyostotic Fibrous Dysplasia: Failure to Prevent Expansion of Dysplastic Lesions During Childhood. J Pediatr Endocrinol Metab. 2006;19(1):75-80.

10. Parisi MS, Oliveri B. Long-term pamidronate treatment of polyostotic fibrous dysplasia of bone: A case series in young adults. Curr Ther Res Clin Exp. 2009;70(2):161-72.

11. Liu J, Zhao C, Liu B, Liu H, Wang L. Analgesia and curative effect of pamidronate disodium combined with chemotherapy

    on elderly patients with advanced metastatic bone cancer. Oncol Lett. 2019;18(1):771-5.

12. Ren HY, Sun LL, Li HY, Ye ZM. Prognostic Significance of Serum Alkaline Phosphatase Level in Osteosarcoma: A Meta- Analysis of Published Data. BioMed Res Int. 2015;2015:160835.

13. Park BY, Cheon YW, Kim YO, Pae NS, Lee WJ. Prognosis for craniofacial fibrous dysplasia after incomplete resection: age and serum alkaline phosphatase. Int J Oral Maxillofac Surg. 2010;39(3):221-6.

14. Chapurlat RD, Hugueny P, Delmas PD, Meunier PJ. Treatment of fibrous dysplasia of bone with intravenous pamidronate: long-term effectiveness and evaluation of predictors of response to treatment. Bone. 2004;35(1):235-42.

15. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Syst Rev. 2021;10:89.

16. Standard Deviation Calculator [Internet]. [cited 2023 Sep 2]. Available from: https://www.calculator.net/standard-deviation- calculator.html

17. Chapurlat RD, Delmas PD, Liens D, Meunier PJ. Long-term effects of intravenous pamidronate in fibrous dysplasia of bone. J Bone Miner Res Off J Am Soc Bone Miner Res. 1997;12(10): 1746-52.

18. Plotkin H, Rauch F, Zeitlin L, Munns C, Travers R, Glorieux FH. Effect of pamidronate treatment in children with polyostotic fibrous dysplasia of bone. J Clin Endocrinol Metab. 2003;88(10): 4569-75.

19. Isaia GC, Lala R, Defilippi C, Matarazzo P, Andreo M, Roggia C, et al. Bone turnover in children and adolescents with McCune- Albright syndrome treated with pamidronate for bone fibrous dysplasia. Calcif Tissue Int. 2002;71(2):121-8.

20. Lala R, Matarazzo P, Andreo M, Marzari D, Bellone J, Corrias A, et al. Bisphosphonate treatment of bone fibrous dysplasia in McCune-Albright syndrome. J Pediatr Endocrinol Metab JPEM. 2006;19 Suppl 2:583-93.

21. Lala R, Matarazzo P, Bertelloni S, Buzi F, Rigon F, de Sanctis C. Pamidronate treatment of bone fibrous dysplasia in nine children with McCune-Albright syndrome. Acta Paediatr 2000;89(2): 188-93.

22. Zacharin M, O’Sullivan M. Intravenous pamidronate treatment of polyostotic fibrous dysplasia associated with the McCune Albright syndrome. J Pediatr. 2000;137(3):403-9.

23. Igelström E, Campbell M, Craig P, Katikireddi SV. Cochrane’s risk of bias tool for non-randomized studies (ROBINS-I) is frequently misapplied: A methodological systematic review. J Clin Epidemiol. 2021;140:22-32.

24. Hartley I, Zhadina M, Collins MT, Boyce AM. Fibrous Dysplasia of Bone and McCune-Albright Syndrome: a Bench to Bedside Review. Calcif Tissue Int. 2019;104(5):517-29.

25. Spencer T, Pan KS, Collins MT, Boyce AM. The clinical spectrum of McCune-Albright syndrome and its management. Horm Res Paediatr. 2019;92(6):347-56.

26. Laine J, Kadado A, James C, Novotny S. The Role of Bisphophonates in Pediatric Orthopaedics: What Do We Know After 50 Years? Current Concept Review. J Pediatr Orthop Soc N Am [Internet]. 2019 Nov 1 [cited 2023 Sep 1];1(1). Available from: https:// www.jposna.org/index.php/jposna/article/view/33

27. Parisi MS, Oliveri MB, Mautalen CA. Bone mineral density response to long-term bisphosphonate therapy in fibrous dysplasia. J Clin Densitom Off J Int Soc Clin Densitom.

    2001;4(2):167-72.

28. Lee JM, Kim JE, Bae SH, Hah JO. Efficacy of pamidronate in children with low bone mineral density during and after chemotherapy for acute lymphoblastic leukemia and non- Hodgkin lymphoma. Blood Res. 2013;48(2):99-106.

29. Chapurlat RD, Gensburger D, Jimenez-Andrade JM, Ghilardi JR, Kelly M, Mantyh P. Pathophysiology and medical treatment of pain in fibrous dysplasia of bone. Orphanet J Rare Dis. 2012; 7(1):S3.

30. Majoor BCJ, Traunmueller E, Maurer-Ertl W, Appelman-Dijkstra NM, Fink A, Liegl B, et al. Pain in fibrous dysplasia: relationship with anatomical and clinical features. Acta Orthop. 2019;90(4): 401-5.

31. Shu J, Tan A, Li Y, Huang H, Yang J. The correlation between serum total alkaline phosphatase and bone mineral density in young adults. BMC Musculoskelet Disord. 2022;23:467.

32. Wang J, Du Z, Li D, Yang R, XiaodongTang, Yan T, et al. Increasing serum alkaline phosphatase is associated with bone deformity progression for patients with polyostotic fibrous dysplasia. J Orthop Surg. 2020;15(1):583.

33. Ma J, Liang L, Gu B, Zhang H, Wen W, Liu H. A retrospective study on craniofacial fibrous dysplasia: preoperative serum alkaline phosphatase as a prognostic marker? J Cranio-Maxillo-fac Surg Off Publ Eur Assoc Cranio-Maxillo-fac Surg. 2013;41(7): 644-7.

34. Vaisman DN, McCarthy AD, Cortizo AM. Bone-specific alkaline phosphatase activity is inhibited by bisphosphonates: role of divalent cations. Biol Trace Elem Res. 2005;104(2):131- 40.

35. Greenblatt MB, Tsai JN, Wein MN. Bone Turnover Markers

    in the Diagnosis and Monitoring of Metabolic Bone Disease. Clin Chem. 2017;63(2):464-74.

36. Chilbule SK, Madhuri V. Complications of pamidronate therapy in paediatric osteoporosis. J Child Orthop. 2012;6(1): 37-43.

37. Juszczak B, Sułko J. Patient-reported effectiveness and safety of Pamidronate in NSAIDs-refractory chronic recurrent multifocal osteomyelitis in children. Rheumatol Int. 2022;42(4):699-706.

38. Hart ES, Kelly MH, Brillante B, Chen CC, Ziran N, Lee JS, et al. Onset, progression, and plateau of skeletal lesions in fibrous dysplasia and the relationship to functional outcome. J Bone Miner Res Off J Am Soc Bone Miner Res. 2007;22(9): 1468-74.

39. Lee SE, Lee EH, Park H, Sung JY, Lee HW, Kang SY, et al. The diagnostic utility of the GNAS mutation in patients with fibrous dysplasia: meta-analysis of 168 sporadic cases. Hum Pathol. 2012;43(8):1234-42.

40. Elli FM, de Sanctis L, Bergallo M, Maffini MA, Pirelli A, Galliano I, et al. Improved Molecular Diagnosis of McCune– Albright Syndrome and Bone Fibrous Dysplasia by Digital PCR. Front Genet. 2019;10:862.

41. Kushchayeva YS, Kushchayev SV, Glushko TY, Tella SH, Teytelboym OM, Collins MT, et al. Fibrous dysplasia for radiologists: beyond ground glass bone matrix. Insights Imaging. 2018;9(6):1035-56.

42. Bhadada SK, Bhansali A, Das S, Singh R, Sen R, Agarwal A, et al. Fibrous dysplasia & McCune-Albright syndrome: An experience from a tertiary care centre in north India. Indian J Med Res. 2011;133(5):504-9.

Plasma Alpha Synuclein as a Potent Biomarker of Diseases with Synucleinopathies

Chaisak Dumrikarnlert,*,**, Lertchai Wachirutmangur,*, Suthipol Udomphanthurak,*, Chatchawan Rattanabannakit,*, Prachaya Srivanitchapoom,*,***, Vorapun Senanarong, M.D.*

*Department of Neurology, Siriraj Hospital, Mahidol University, Bangkok, Thailand, **Neuroscience Center, Bangkok International Hospital, Bangkok, Thailand, ***Brain Center, Ramkhamhaeng Hospital, Bangkok, Thailand.

ABSTRACT

Objective: We explored whether plasma α-syn be used as a potential biomarker for synucleinopathies.

Materials and Methods: α-syn levels in plasma from 54 Parkinson’s disease dementia (PDD) and dementia with Lewy bodies (DLB) patents, 31 Alzheimer’s disease dementia (AD), and 29 controls were measured by enzyme- linked immunosorbent assay (ELISA).

Results: The mean age of the synucleinopathies group, the AD group, and the normal controls was 72.70, 74.26, and 62 years old. The median plasma α-syn levels in the synucleinopathies group, AD group and controls were 9.72 (4.41-25.30), 16.78 (7.68-51.41) and 16.65 (10.37-32.72) ng/ml, respectively (Independent-Samples Kruskal-Wallis test, p = 0.026). The α-syn levels in the synucleinopathies group were lower than those of AD and controls. There was a fair correlation between plasma α-syn levels and the sum of the Unified Parkinson’s Disease Rating Scale (UPDRS) part 3 (spearman correlation coefficient r = -0.261, p = 0.021) but not with cognition measured by Thai Mental Status Examination (TMSE). The area under the receiver operating characteristic curve (ROC) was 0.710 between the PDD and DLB vs non synucleinopathies group (AD and normal controls) (SE = 0.052, p ≤ 0.001). At the cut-off levels of 11.4 ng/ml indicated a sensitivity of 58% (95% CI 43.21-71.81%), specificity of 84.78% (95% CI 71.13-93.66%), positive predictive value (PPV) of 80.56%, a negative predictive value (NPV) of 65% and a precision of 70.83%.

Conclusion: The present results suggest that plasma α-syn could be a potential biomarker to differentiate synucleinopathies

from Alzheimer’s disease and the elderly with normal cognition.

Keywords: Alpha synuclein; biomarker; plasma; Parkinson’s disease dementia; Lewy body dementia; Alzheimer’s disease (Siriraj Med J 2023; 75: 864-870)

INTRODUCTION

Alpha-synuclein (α-synuclein) is a presynaptic neuronal protein that is neuropathologically related to synucleinopathies, including Parkinson’s disease (PD), dementia with Lewy bodies (DLB) and multiple system atrophy (MSA).1,2 This protein had been explored as a diagnostic biomarker for synucleinopathies in cerebrospinal fluid (CSF), blood plasma, serum, and skin.3-10 Many

studies have shown that CSF alpha-synuclein can be used to differentiate between synucleinopathies and Alzheimer’s disease, the most common neurodegenerative disease.6,7,11-12 However, fewer studies have investigated alpha-synuclein in peripheral blood and its utility to differentiate between synucleinopathies and other diseases.13-16 Furthermore, data on the plasma alpha-synuclein level in patients with PD are still inconclusive because previous studies had

Corresponding author: Vorapun Senanarong E-mail: vorapun.sen@mahidol.ac.th

Received 22 September 2023 Revised 3 November 2023 Accepted 18 November 2023 ORCID ID:http://orcid.org/0000-0002-2774-4187 https://doi.org/10.33192/smj.v75i12.265475

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

shown that blood alpha-synuclein level in patients with PD can be higher7,17 or lower18,19 compared to normal control subjects. Some studies have also found that there is no difference in the blood level of alpha-synuclein between PD patients and control.13,20-22 Therefore, we investigated whether plasma alpha-synuclein levels can be used to differentiate between synucleinopathies, Alzheimer’s disease, and control.

MATERIALS AND METHODS

Study design and population

The patients were recruited from the memory clinic at Siriraj Hospital, Mahidol University. The diagnosis of PD and Parkinson’s disease dementia (PDD) was based on the clinical diagnostic criteria of the Movement Disorder Society for Parkinson’s disease23 and the recommendations of the MDS Task Force for the diagnosis of PDD.24 Dementia with Lewy bodies (DLB) was diagnosed by using consensus criteria for clinical diagnosis developed by the DLB Consortium25, and probable Alzheimer’s disease (AD) was defined using the criteria of the National Institute on Aging and the Alzheimer’s Association (NIA-AA).26

The sample size was calculated using the mean ± standard deviation (SD) from the reference literature.27 Inclusion criteria were individuals whose age was more than 40 years, without a minimum year of education and diagnosed with Alzheimer’s disease, synucleinopathies, or normal cognition. Exclusion criteria were individuals who were less than 40 years old or did not meet criteria for the diagnosis of AD or synucleinopathies. We also excluded patients currently using medications that can cause parkinsonism, such as antipsychotic medications. Acetylcholine esterase inhibitor and Parkinson medications (e.g., levodopa, dopamine agonist) were allowed. Therefore, we recruited 114 patients, consisting of 54 PDD and DLB patents, 31 AD patients, and 29 control participants.

We collected clinical information including age, sex, diagnosis, Thai version Mental Status Examination (TMSE) score, Addenbrooke’s Cognitive Examination-Revised (ACE-R), Thai Activities of Daily Living Scale (ADL)28, Unified Parkinson’s Disease Rating Scale (UPDRS), and result of dopamine transporter scan (using 99mTc- TRODAT-1 SPECT image).

Blood sample and measurement of plasma alpha- synuclein concentration

Venous blood (5 ml) was drawn from the participants in the morning, 9.00 to 12.00 am, and the samples were processed within 30 minutes of collection.

Plasma was prepared after collection of whole blood in an ethylenediaminetetraacetic acid treated tube. The processed samples were treated by centrifugation for 15 minutes at 1,500 g at room temperature. Following centrifugation, plasma was immediately transferred into clean and low residue polypropylene tube using a low residue tip. Plasma was stored at -80°C for less than 3 months prior to examination. No hemolyzed, icteric, or lipemic samples used. The levels of alpha-synuclein in the blood were tested by immunosorbent assay (ELISA) using the Human Phosphorylated Alpha Synuclein (PSNCA) ELISA Kit of MyBioSource, Inc, United States. The test was performed concurrently after 8-10 samples had been collected.

Statistics

For the statistical analyses, IBM SPSS Statistics 18 software was used. The blood α-synuclein data was not normally distributed and assessed by the Kolmogorov- Smirnov test. Mann-Whitney U was used to compare the results between two groups and the Kruskal-Wallis test was used to compare the results between more than two groups with adequate correction for multiple comparisons (Bonferroni). To analyze frequency difference of dichotomous variables, the chi-square test was used. The Spearman rank-order correlation coefficient was used to assess the correlations between variables. The receiver operating characteristic (ROC) curve and the area under the curve (AUC) were used to determine the cutoff value for α-synuclein.

RESULTS

Of the 114 participants in our study, there were 28 men (51.85%) in the synucleinopathies group, 15 men (45.45%) in the AD group and 3 men (11.11%) in the control group. The median age of all participants was

70.4 ± 10.463 years. Age frequencies showed median synucleinopathies (IQR) = 74 (66-77), AD = 75 (64-81),

NC 61 (54.5-67.0); Independent-Samples Kruskal-Wallis Test p<0.001. The pairwise comparison of group diagnosis found that between synucleinopathies vs control: test statistic 32.987, standard error (SE) 7.604, standard deviation (SD) 4.338, p<0.0001; between AD vs control: test statistic 36.503, SE 8.533, SD 4.278, p<0.0001; between synucleinopathies vs AD: test statistic 3.516, SE 7.443, SD 0.472, p=0.637. The score of TMSE and ACE-R was lower in both disease groups compared to control and the ADL score in both disease groups was higher. The details of the score in each group are shown in Table 1.

TABLE 1. Patient’s characteristics.

*p<0.001 between AD vs control, and between synucleinopathies vs control.

¥p<0.001 between synucleinopathies vs control, and synucleinopathies vs AD.

Data are presented as n (%), mean ± standard deviation, or median (interquartile range). p-value corresponds to one-way and Kruskal-Wallis

For plasma alpha-synuclein level, median in synucleinopathies groups, AD groups, and controls was 9.72 (4.41-25.30), 16.78 (7.68-51.41), and 16.65

(10.37-32.72) ng/ml, respectively (Independent-Samples Kruskal-Wallis test, p = 0.026). Pairwise comparisons of group diagnosis in plasma α-syn levels revealed that between PDD and DLB vs control: test statistic -15.99, Standard Error (SE) 7.61, SD -2.10, p = 0.036; between PDD and DLB vs AD: test statistic 17.39, SE 7.45, SD 2.34, p = 0.02; and between controls vs AD: test statistic 1.407, SE 8.54, SD 0.17, p = 0.869. The plasma level of alpha-synuclein in the synucleinopathies group was significantly lower than in both the AD and the control group. We look for correlation between plasma alpha- synuclein and other factors, such as age, education, cognitive score (TMSE) and UPDRS score, but found a fair correlation only between plasma α-syn levels and the sum of UPDRS part 3 (spearman correlation coefficient r = -0.261, p = 0.021). Data on the correlation of other factors are shown in Table 2. The area under the receiver operating characteristic curve (ROC) was 0.710 between the PDD and DLB group vs. non-synucleinopathies (AD

and normal controls)(SE = 0.052, p ≤ 0.001) as shown in Fig 1.

At the cutoff levels of 11.4 ng/ml indicated a sensitivity of 58% (95% CI 43.21-71.81%), specificity of 84.78% (95% CI 71.13-93.66%), positive predictive value (PPV) of 80.56%, a negative predictive value (NPV) of 65% and a precision of 70.83%.

In our study, 15 participants were scanned using 99mTc-TRODAT-1 SPECT imaging. Eleven participants had a positive scan, five of whom were in the synucleinopathies group and six were in the AD group. Comparing the positive and negative groups in age, blood alpha-synuclein, UPDRS part 3, or cognitive score, did not show any significant differences.

DISCUSSION

In synucleinopathies (PD, DLB and MSA), we diagnosed mainly by clinical criteria while neuroimaging biomarkers, structural and functional, were used only to support and exclude other possible causes.1,29 Previous studies have shown the potential of cerebrospinal fluid (CSF) alpha-synuclein to differentiate synucleinopathies

TABLE 2. Spearman’s rank correlation coefficient between each factor.

* Correlation is significant at the 0.05 level (2-tailed).

** Correlation is significant at the 0.01 level (2-tailed).

a Spearman's rank correlation coefficient

Alphasynuclein (ng/mL)

100

80

60

40

20

AUC = 0.710

P < 0.001

0

0

20

40

60

80

100

100-Specificity

Sensitivity

Fig 1. Receiver operating characteristic (ROC) curve for plasma α-synuclein levels to detect synucleinopathies (PDD and DLB). ROC curve of plasma α-synuclein levels for distinguishing synucleinopathies group from non-synucleinopathies group. AUC, area under ROC curve. Cut-off value = 11.4 ng/ ml

from AD or normal cognitive control.12, 30-31 However, due to difficult, invasive and costly of CSF collection compared to blood draws, so blood-based biomarkers are promising methods to use in clinical practice for the evaluation of neurodegenerative disease.32-34

Plasma alpha-synuclein in a previous study shows that it can be used to differentiate PD from AD or normal control13-16, and most of the study found that plasma level of alpha-synuclein is higher than AD or control.7,16-17

But in our study, plasma level of alpha-synuclein in synucleinopathies was lower than AD or control, and there are some other studies that reported similar results.18-19 These conflicting results could be due to the different assay or method used to measure plasma alpha-synuclein35 or because the main source of alpha-synuclein is red blood cells (RBC), so even low RBC contamination could affect the results.36 Therefore, the handling method and the preparation of the sample are other factors that cause

inconsistency in the plasma alpha-synuclein level. In addition, disease severity and disease duration, which are different in each study, can also affect the results.13 According to published data, commonly used technologies for evaluation included the bead-based multiplexed immunoassay system (Luminex), sandwiched ELISA or ImmunoMagnetic Reduction (IMR).16 If we look studies that use the same technique that we used7,13,20-21,37-38, sandwiched ELISA, most had either higher or equal of plasma alpha-synuclein in the disease group compare to the control. It may be that, first, our studies inclusion criteria were synucleinopathies disease group, not just Parkinson’s disease, which is different from previous research. Synucleinopathies consist of Parkinson’s disease, Dementia with Lewy Body, and Multisystem Atrophy, which each of them had clinically and pathologically heterogeneous. For example, pathological hallmark of MSA was glial cytoplasmic inclusions (GCIs) predominantly in striatum, midbrain, pons, medulla, and cerebellum whereas for DLB, pathological hallmark was widespread of Lewy bodies or Lewy neurites in cerebral cortex and limbic system.39-41 Due to different pathological seeding locations of alpha synuclein, clinical manifestations, criteria diagnosis, and prognosis of disease also different and that’s explained why cut off value of blood alpha synuclein in our study different diagnosis of patient from previous studies, show different results. Second, our recruiting population was in an earlier stage and less severe than others with respect to the UPDRS score. Normally, plasma alpha-synuclein levels will increase over time along with disease severity13 so our studies that patient were still in early stage, which mean alpha synuclein still didn’t spread that much in central nervous system, measurement of alpha synuclein will be lower

compare to other studies.

Plasma alpha-synuclein alone may not be suitable to differentiate parkinsonism from other neurodegenerations or controls and may need other biomarkers to increase accuracy. There are many recent studies that use multiple biomarkers, in CSF or blood, to produce greater sensitivity, specificity, and precision to differentiate between parkinsonism and control.42-48 The most recent published studies48 found that using combination of α-synuclein, Aβ42, Aβ40, Aβ42/40, and NfL could achieve a best diagnostic value in differentiating parkinsonian syndrome from healthy control with AUC 0.98. In future research, we may need to use multiple biomarkers, including plasma alpha- synuclein, to discriminate parkinsonism from other neurodegenerative diseases.

For imaging modality, we know for a long time that brain perfusion single photon emission computed tomography (SPECT) can be used for aiding diagnosis of

Alzheimer’s disease in early stage.49 Because of difference abnormal perfusion area in each type of dementia, brain SPECT can help in differential diagnosis of dementia.50 More specific type of SPECT, 99mTc-TRODAT-1 SPECT (TRODAT), is dopamine transporter imaging that can be used for differential diagnosis of Parkinsonism, between synucleinopathies and secondary Parkinsonisms.51 It can also differentiate between Dementia with Lewy Bodies and Alzheimer’s disease, which scan should be positive if patient had DLB.52 We tried to analyze data between the positive and negative TRODAT SPECT group or between the positive TRODAT SPECT group in synucleinopathies and AD groups, but small sample sizes prevented us from detecting significance differences when comparing between them.

CONCLUSION

Plasma α-synuclein is a new biomarker for the diagnosis of Parkinson’s disease and other synucleinopathies that may be useful to distinguish them from other diseases. Our study showed that the plasma level of α-synuclein is lower in synucleinopathies compared to Alzheimer’s disease and normal control participants. At the cutoff levels of 11.4 ng/ml indicated a sensitivity of 58% (95% CI 43.21-71.81%), specificity of 84.78% (95% CI 71.13-

93.66%), positive predictive value (PPV) of 80.56%, a negative predictive value (NPV) of 65% and a precision of 70.83%. The area under ROC was 0.710 between the PDD and DLB vs. the group without synucleinopathies (AD and normal controls) (SE = 0.052, p ≤ 0.001). Plasma α-synuclein level correlates well with the motor sign of parkinsonism, measured by the sum score of UPDRS part 3, but not with cognition, evaluated using the TMSE score. Using multiple biomarkers, both fluid and imaging, will give more benefit to differentiate synucleinopathies from Alzheimer’s disease.

Statement of Ethics

The protocol for this study was approved by the Siriraj Institutional Review Board (SIRB) of the Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand (COA no. Si 667/2018). Written informed consent was obtained from all participants in this study. All authors confirm that the research was conducted in accordance with the Declaration of Helsinki. Abstract and some content of this article was presented as Poster at Alzheimer’s Association International Conference in 2020, Chicago but only abstract and figure was published online in supplementary issue of Alzheimer’s & Dementia Journal. Content and figure in this article were all different from previous publication to avoid issue of plagiarism.

Conflict of interest statement

All authors declare no personal or professional conflicts of interest relating to any aspect of this study.

Funding sources

This was an unfunded study.

Author contributions

CD was involved in data and statistical analysis, interpretation of data, and manuscript writing. LW was involved in taking blood samples and processing them. SU was involved in data and statistical analysis. CR and PS participated in data collection. VS participated in design of study, interpretation of data, and manuscript revision. All authors approved the protocol.

Data availability statement

All data generated for this study are included in the article. There are no other datasets generated during the current study.

REFERENCES

1. Martı´ MJ, Tolosa E, Campdelacreu J. Clinical Overview of the Synucleinopathies. Mov Disord. 2003;18 Suppl 6:S21-7.

2. Goedert M. Alpha-synuclein and neurodegenerative diseases. Nat Rev Neurosci. 2001;2:492-501.

3. Eller M, Williams DR. α-Synuclein in Parkinson Disease and other neurodegenerative disorders. Clin Chem Lab Med. 2011;49(3): 403-8.

4. Kasuga K, Nishizawa M, Ikeuchi T. α-Synuclein as CSF and Blood Biomarker of Dementia with Lewy Bodies. Int J Alzheimers Dis. 2012;2012:437025.

5. Tokuda T, Qureshi MM, Ardah MT, Varghese S, Shehab SA, Kasai T, et al. Detection of elevated levels of α-synuclein oligomers in CSF from patients with Parkinson disease. Neurology. 2010;75:1766-70.

6. Hong Z, Shi M, Chung KA, Quinn JF, Peskind ER, Galasko D, et al. DJ-1 and α-synuclein in human cerebrospinal fluid as biomarkers of Parkinson’s disease. Brain. 2010;133:713-26.

7. Lee PH, Lee G, Park HJ, Bang OY, Joo IS, Huh K. The plasma alpha-synuclein levels in patients with Parkinson’s disease and multiple system atrophy. J Neural Transm (Vienna). 2006;113(10): 1435-9.

8. Lin CH, Yang SY, Horng HE, Yang CC, Chieh JJ, Chen HH, et al. Plasma α-synuclein predicts cognitive decline in Parkinson’s disease. J Neurol Neurosurg Psychiatry. 2017;88:818-24.

9. Kim JY, Illigens BM, McCormick MP, Wang N, Gibbons CH. Alpha-Synuclein in Skin Nerve Fibers as a Biomarker for Alpha-Synucleinopathies. J Clin Neurol. 2019; 15(2):135-42.

10. Wang Z, Becker K, Donadio V, Siedlak S, Yuan J, Rezaee M, et al. Skin α-Synuclein Aggregation Seeding Activity as a Novel Biomarker for Parkinson Disease. JAMA Neurol. 2021;78(1): 30-40.

11. Kasuga K, Tokutake T, Ishikawa A, Uchiyama T, Tokuda T, Onodera O, et al. Differential levels of alpha-synuclein, beta- amyloid42 and tau in CSF between patients with dementia

    with Lewy bodies and Alzheimer’s disease. J Neurol Neurosurg Psychiatry. 2010;81:608-10.

12. Tateno F, Sakakibara R, Kawai T, Kishi M, Murano T. Alpha- synuclein in the Cerebrospinal Fluid Differentiates Synucleinopathies (Parkinson Disease, Dementia with Lewy Bodies, Multiple System Atrophy) From Alzheimer Disease. Alzheimer Dis Assoc Disord. 2011;26(3):213-6.

13. Foulds PG, Diggle P, Mitchell JD, Parker A, Hasegawa M, Masuda-Suzukake M, et al. A longitudinal study on α-synuclein in blood plasma as a biomarker for Parkinson’s disease. Sci Rep. 2013;3:2540.

14. Koehler NK, Stransky E, Meyer M, Gaertner S, Shing M, Schnaidt M, et al. Alpha-Synuclein Levels in Blood Plasma Decline with Healthy Aging. PLoS One. 2015; 10(4):e0123444.

15. Chang CW, Yang SY, Yang CC, Chang CW, Wu YR. Plasma and Serum Alpha-Synuclein as a Biomarker of Diagnosis in Patients With Parkinson’s Disease. Front Neurol. 2020;10:1388.

16. Bougea A, Stefanis L, Paraskevas GP, Emmanouilidou E, Vekrelis K, Kapaki E. Plasma alpha-synuclein levels in patients with Parkinson’s disease: a systematic review and meta-analysis. Neurol Sci. 2019;40:929-38.

17. Duran R, Berrero FJ, Morales B, Luna JD, Ramirez M, Vives F. Plasma alpha-synuclein in patients with Parkinson’s disease with and without treatment. Mov Disord. 2010;25(4):489-93.

18. Li QX, Mok SS, Laughton KM, McLean CA, Cappai R, Masters CL, et al. Plasma alpha-synuclein is decreased in subjects with Parkinson’s disease. Exp Neurol. 2007; 204(2):583-8.

19. Gorostidi A, Bergareche A, Ruiz-Martinez J, Martí-Massó JF, Cruz M, Varghese S, et al. Alpha-synuclein levels in blood plasma from LRRK2 mutation carriers. PLoS One. 2012;7(12):e52312.

20. Caranci G, Piscopo P, Rivabene R, Traficante A, Riozzi B, Castellano AE, et al. Gender differences in Parkinson’s disease: focus on plasma α-synuclein. J Neural Transm (Vienna). 2013;120(8):1209-15.

21. Park MJ, Cheon SM, Bae HR, Kim SH, Kim JW. Elevated levels of alpha-synuclein oligomer in the cerebrospinal fluid of drug- naïve patients with Parkinson’s disease. J Clin Neurol. 2011;7(4): 215-22.

22. Mata IF, Shi M, Agarwal P, Chung KA, Edwards KL, Factor SA, et al. SNCA variant associated with Parkinson disease and plasma alpha-synuclein level. Arch Neurol. 2010;67(11):1350-6.

23. Postuma RB, Berg D, Stern M, Poewe W, Olanow CW, Oertel W, et al. MDS clinical diagnostic criteria for Parkinson’s disease. Mov Disord. 2015;30(12):1591-601.

24. Dubois B, Burn D, Goetz C, Aarsland D, Brown RG, Broe GA, et al. Diagnostic procedures for Parkinson’s disease dementia: recommendations from the movement disorder society task force. Mov Disord. 2007;22(16):2314-24.

25. McKeith IG, Boeve BF, Dickson DW, Halliday G, Taylor JP, Weintraub D, et al. Diagnosis and management of dementia with Lewy bodies: Fourth consensus report of the DLB Consortium. Neurology. 2017;89(1):88.

26. McKhann GM, Knopman DS, Chertkow H, Hyman BT, Jack CR Jr, Kawas CH, et al. The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 2011;7(3):263.

27. Wennström M, Surova Y, Hall S, Nilsson C, Minthon L, Boström F, et al. Low CSF levels of both α-synuclein and the α-synuclein

    cleaving enzyme neurosin in patients with synucleinopathy. PLoS One. 2013;8(1):e53250.

28. Senanarong V, Harnphadungkit K, Prayoonwiwat N, Poungvarin N, Sivasariyanonds N, Printarakul T, et al. A new measurement of activities of daily living for Thai elderly with dementia. Int Psychogeriatr. 2003;15(2):135-48.

29. Saeed U, Lang AE, Masellis M. Neuroimaging Advances in Parkinson’s Disease and Atypical Parkinsonian Syndromes. Front Neurol. 2020;11:572976.

30. Gao L, Tang H, Nei K, Wang L, Zhao J, Gan R, et al. Cerebrospinal fluid alpha-synuclein as a biomarker for Parkinson’s disease diagnosis: a systematic review and meta-analysis. Int J Neurosci. 2015;125(9):645-54.

31. van Steenoven I, Majbour NK, Vaikath NN, Berendse HW, van der Flier WM, van de Berg WDJ, et al. α-Synuclein species as potential cerebrospinal fluid biomarkers for dementia with lewy bodies. Mov Disord. 2018;33(11):1724-33.

32. Ashton NJ, Hye A, Rajkumar AP, Leuzy A, Snowden S, Suárez- Calvet M, et al. An update on blood-based biomarkers for non-Alzheimer neurodegenerative disorders. Nat Rev Neurol. 2020;16(5):265-84.

33. Leuzy A, Mattsson-Carlgren N, Palmqvist S, Janelidze S, Dage JL, Hansson O. Blood-based biomarkers for Alzheimer’s disease. EMBO Mol Med. 2022;14(1): e14408.

34. Teunissen CE, Verberk IMW, Thijssen EH, Vermunt L, Hansson O, Zetterberg H, et al. Blood-based biomarkers for Alzheimer’s disease: towards clinical implementation. Lancet Neurol. 2022; 21(1):66-77.

35. Chiu MJ, Leu LF, Sabbagh MN, Chen TF, Chen HH, Yang SY. Long-Term Storage Effects on Stability of Aβ1-40, Aβ1-42, and Total Tau Proteins in Human Plasma Samples Measured with Immunomagnetic Reduction Assays. Dement Geriatr Cogn Dis Extra. 2019;9(1):77-86.

36. Barbour R, Kling K, Anderson JP, Banducci K, Cole T, Diep L, et al. Red blood cells are the major source of alpha-synuclein in blood. Neurodegener Dis. 2008;5:55-59.

37. Ding J, Zhang J, Wang X, Zhang L, Jiang S, Yuan Y, et al. Relationship between the plasma levels of neurodegenerative proteins and motor subtypes of Parkinson’s disease. J Neural Transm (Vienna). 2017;124(3):353-60.

38. Malec-Litwinowicz M, Plewka A, Plewka D, Bogunia E, Morek M, Szczudlik A, et al. The relation between plasma alpha- synuclein level and clinical symptoms or signs of Parkinson’s disease. Neurol Neurochir Pol. 2018;52(2):243-51.

39. Outeiro TF, Koss DJ, Erskine D, Walker L, Kurzawa-Akanbi M, Burn D, et al. Dementia with Lewybodies: an update and

    outlook. Mol Neurodegeneration. 2019;14:5.

40. Campese N, Fanciulli A, Stefanova N, Haybaeck J, Kiechl S, Wenning GK. J Neural Transm. 2021;128:1481-94.

41. Koga S, Sekiya H, Kondru N, Ross OA, Dickson DW. Mol Neurodegeneration. 2021;16:83.

42. Parnetti L, Chiasserini D, Bellomo G, Giannandrea D, De Carlo C, Qureshi MM, et al. Cerebrospinal fluid Tau/α-synuclein ratio in Parkinson’s disease and degenerative dementias. Mov Disord. 2011;26(8):1428-35.

43. Kang JH, Irwin DJ, Chen-Plotkin AS, Siderowf A, Caspell C, Coffey CS, et al. Association of cerebrospinal fluid β-aymloid 1-42, T-tau, P-tau 181, and α-synuclein levels with clinical features of drug-naïve patients with early Parkinson disease. JAMA Neurol. 2013;70(10):1277-87.

44. Hanchcliffe C. Blood and cerebrospinal fluid markers in Parkinson’s disease: current biomarker findings. Current Biomarker Findings. 2015;5:1-11.

45. Lin CH, Yang SY, Horng HE, Yang CC, Chieh JJ, Chen HH, et al. Plasma Biomarkers Differentiate Parkinson’s Disease From Atypical Parkinsonism Syndromes. Front Aging Neurosci. 2018; 10:123.

46. Chen NC, Chen HL, Li SH, Chang YH, Chen MH, Tsai NW, et al. Plasma Levels of α-Synuclein, Aβ-40 and T-tau as Biomarkers to Predict Cognitive Impairment in Parkinson’s Disease. Front Aging Neurosci. 2020;12:112.

47. Ren J, Pan C, Wang Y, Xue C, Lin H, Xu J, et al. Plasma α-synuclein and phosphorylated tau 181 as a diagnostic biomarker panel for de novo Parkinson’s disease. J Neurochem. 2022;161(1): 506-15.

48. Li Q, Li Z, Han X, Shen X, Wang F, Bai L, et al. A Panel of Plasma Biomarkers for Differential Diagnosis of Parkinsonian Syndromes. Front Neurosci. 2022;16:805953.

49. Ratanamart V, Senanarong V, Chulakadabba S, Chiewvit P. SPECT as An Aid for Clinicians in the Diagnosis of Alzheimer’s Disease: A case report and review of current diagnostic approaches and the need for early accurate diagnosis to optimize treatment. Siriraj Med J. 2003;55(1):31-41.

50. Ferrando R, Damian A. Brain SPECT as a Biomarker of Neurodegeneration in Dementia in the Era of Molecular Imaging: Still a Valid Option? Front Neurol. 2021;12:629442.

51. Fabiani G, Camargo CHF, Filho RM, Froehner GS, Teive HAG. Evaluation of Brain SPECT with 99mTc-TRODAT-1 in the Differential Diagnosis of Parkinsonism. Parkinsons Dis. 2022;2022:1746540.

52. Hung GU, Chiu PY. The Value of 99mTc-Trodat-1 SPECT for Discriminating Dementia with Lewy Bodies and Alzheimer’s disease. J Nucl Med. 2017;58 (Suppl 1):1279.

Mobile Device Digital Photography for Teledermatology Consultation: Real-Life Situations

Sumanas Bunyaratavej, M.D.*, Pattriya Jirawattanadon, M.D.*, Chudapa Sereeaphinan, M.D.*, Supisara Wongdama, M.D.*, Sanchai Sombatmaithai, M.D.**, Charussri Leeyaphan, M.D.*

*Department of Dermatology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand, **Department of Medicine, Faculty of Medicine, Mahasarakham University, Mahasarakham 44150, Thailand.

ABSTRACT

Objective: The use of mobile phones for teledermatology consultations is increasing. In this study, we aimed to describe photographic problems in teledermatology performed via mobile phones and their effects on diagnostic decision-making. Materials and Methods: Three dermatologists independently reviewed the medical histories and photographs of patients taken by primary-care physicians for teledermatology between January 2018 and August 2020. The consensus of the dermatologists’ decision-making was categorized into “definite diagnoses given,” “probable diagnoses given,” and “unable to provide any diagnosis.” Relationships between photographic errors and dermatologist decision-making were investigated. Factors related to photographic problems were evaluated. Results: In all, 899 images from 220 patients were reviewed. The most common purpose of teledermatology was to make a diagnosis. The most frequent diagnoses were eczema, infection, and autoimmune diseases. Consultants gave definite diagnoses for 63.2% of patients and probable diagnoses for another 29.5%. However, diagnoses were not made in 7.3% of cases. Defocusing and non-eczematous lesions were significantly associated with the inability to give diagnoses (P = 0.002 and 0.037, respectively). Pictures from peripheral areas showed higher frequencies of distortion errors, improper framing, wasted space, and improper background, while truncal regions tended to have lighting problems. The outpatient department setting was associated with a lack of overview and defocusing. Conclusion: Focusing was the central factor for making diagnoses in teledermatology. Lighting should be more concerned in truncal regions. While using smartphone cameras, distortion should be aware. These factors should be considered to improve the effectiveness of teledermatology.

Keywords: Consult; dermatology; smartphone; teledermatology; telemedicine (Siriraj Med J 2023; 75: 871-879)

INTRODUCTION

Telemedicine is increasingly being used in dermatology. Diagnoses are based mainly on inspection, with additional information from palpation and patient history.1 Two types of teledermatology are currently in use: (1) store and forward techniques, in which clinical data are sent electronically to dermatologists for evaluation; and (2) live interaction techniques, involving real-time synchronous communication between the patient and dermatologist, typically facilitated through videoconferencing technology,

enabling direct visual and audio communication.2 In the store and forward technique, high-quality images can replace primary-care physician descriptions of skin lesions, which are susceptible to describer bias.1

It is undisputed that clinical photographs are beneficial for educational purposes, research, and the management of dermatological conditions, especially in healthcare facilities without dermatologists.1-4 Moreover, teledermatology in outpatient settings enables the immediate diagnosis of complicated cases by general physicians

Corresponding author: Charussri Leeyaphan E-mail: Charussrilee@gmail.com

Received 4 August 2023 Revised 27 September 2023 Accepted 10 October 2023 ORCID ID:http://orcid.org/0000-0001-8430-376X https://doi.org/10.33192/smj.v75i12.264488

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

consulting with dermatologists located elsewhere. This joint approach avoids the referral delays and travel costs that would otherwise be incurred if patients have to visit dermatologists for in-person examinations. These benefits were demonstrated by Zakaria et al. (2010) and

S. Paradela de la Morena et al. (2015), who found that approximately two-thirds of patients could be treated without a clinic-based evaluation after the implementation of teledermatology.5,6 The diagnostic accuracy reported for teledermatology was approximately 80%, compared with face-to-face diagnoses.3,7-9

In addition to the complex nature of certain diseases, the evolution of lesions, and incomplete clinical data, poor image quality negatively affects the accuracy and reliability of teledermatological diagnoses.6,7,10 Inadequate pictures result from the photographic techniques used rather than any shortcomings of the technology or the camera options.1 The photographic technique standard mainly comprises light and shadow, background, the field of view, orientation and framing, distortion, focus and resolution, scale, color calibration, and patient confidentiality.1,11-13 From previous studies, approximately 80% to 90% of images obtained with store-and-forward teledermatology were considered adequate or excellent.6,8,14

Smartphones are used daily for dermatological consultations. Advancements in related camera technologies have resulted in a tool that can capture high-quality images and is easy to use.15 Nevertheless, few articles have discussed the quality of pictures obtained from mobile phones in practice. Thus, this study evaluated photographic problems in mobile-phone teledermatology, the factors associated with inadequate photographs, and their effects of the various shortcomings on diagnostic decision- making. Therefore, these findings could be beneficial for physicians seeking consultations in teledermatology.

MATERIALS AND METHODS

Ethics consideration

This retrospective study was conducted at the Faculty of Medicine Siriraj Hospital, Mahidol University, Thailand. Before the research began, its protocol was approved by the Siriraj Institutional Review Board (COA no. Si 801/2020).

Data collection and evaluation

Photographs of patients taken by primary-care physicians between January 2018 and August 2020 were collected from a private social-media group for dermatology teleconsultation. The clinical information collected from the group consisted of baseline characteristics, consultation setting, and purpose of consultation. Three independent

board-certified dermatologists, each with over a decade of teaching and consulting experience, including their roles as committee members of monthly dermatologic photographic assessment conferences, reviewed the patient histories and images to describe lesion morphology and provide diagnoses. For each patient, the dermatologists’ diagnoses were categorized into “definite diagnoses given,” “probable diagnoses given,” and “unable to provide any diagnosis.” In terms of image quality assessment, photographs of the same body area of the patients were compiled and placed into a corresponding body-region group. We further evaluated these groupings by drawing upon an image-quality checklist we had adapted from various established recommendations11,12,16 and adjusted to suit mobile phone photography (Supplementary Table 1). Assessments were made of the domains of photographic techniques (the presence of overview photographs and close-up views, focus, lighting, background, framing, orientation, wasted space, perspective distortion, color saturation, and white balance). Any disagreements on clinical diagnoses and photographic assessments were resolved through discussion and consensus. Data related to patient confidentiality issues were also evaluated.

Statistical analysis

Data analyses were performed using PASW Statistics for Windows, version 18 (SPSS Inc., Chicago, IL, USA). With each patient, we selected photographs that had the best quality of photographic technique from representative regions. Subsequently, these were used to evaluate the relationships between image quality and dermatologist diagnostic decisions. As appropriate, Chi2 tests or Fisher’s exact tests were used. In addition, the worst quality photographs from each body region were chosen to identify factors associated with inadequate images.

RESULTS

We collected 899 images from 220 patients and grouped them into 385 body regions. As shown in Table 1, most patients were older than 18 years (81.6%). The primary source of consultation was an outpatient department (OPD; 82.3%), with 17.7% consulted in an inpatient department (IPD) setting. In most of the cases (55%), general physicians were seeking the diagnosis of a skin condition. In another 25% of the cases, general physicians required advice on diagnosing and managing the presenting condition. In the remaining cases (20%), general physicians had formed a diagnosis and were only seeking advice on managing the condition. Most skin lesions were erythematous and eczematous (31.8% and 24.5% of the patients, respectively). In 42 cases (19.1%),

TABLE 1. Patient demographic data.

Demographic data

Number of patients

n = 220 (%)

Mean age at the onset ± SD (years) 41.64 ± 22.47 Age group

0–5 years 15/207 (7.2)

6–17 years 23/207 (11.1)

18–64 years 134/207 (64.7)

> 65 years 35/207 (16.9)

Median duration of disease (IQR; months) 1.00 (0.22–3.00)

Gender

Male 105/209 (50.2)

Consultation setting

Outpatient Inpatient

181/220 (82.3)

39/220 (17.7)

Female 104/209 (49.8)

Purpose of consultation

For diagnosis 122/220 (55.5)

For proper management 42/220 (19.1)

Provided data in teleconsultation Disease duration

Underlying disease Previous treatment Current medication Occupation

Pet or animal exposure Drug allergy

Family history

Environmental exposure

186/220 (84.5)

148/220 (67.3)

134/220 (60.9)

133/220 (60.5)

33/220 (15.0)

4/220 (1.8)

1/220 (0.5)

1/220 (0.5)

0/220 (0.0)

For diagnosis and proper management 56/220 (25.5)

Diagnosis category

Definite diagnosis given (spot diagnosis) 116/220 (52.7)

Definite diagnosis given (requiring provided history) 23/220 (10.5)

Probable diagnoses given along with proper management 65/220 (29.5)

Disease morphology Erythematous lesions Eczematous lesions

Tumor, nodules, and plaques Vesiculobullous lesions Pustular lesions

Hair and nail lesions Ulcer and erosion Papulosquamous lesions

Hyperpigmentation

70/220 (31.8)

54/220 (24.5)

29/220 (13.2)

17/220 (7.7)

14/220 (6.4)

11/220 (5.0)

9/220 (4.1)

8/220 (3.6)

8/220 (3.6)

Unable to provide any diagnosis 16/220 (7.3)

Concurrent skin disease 34/216 (15.7)

Abbreviations: IQR, interquartile range; SD, standard deviation

a lack of patient confidentiality was an issue. Specifically, hospital names or logos, patient faces, patient tattoos, and name tags were visible in images associated with 29, 8, 5, and 4 cases, respectively. Patient consent for photography from referring general practitioners was explicitly mentioned in only 11 of the 220 cases (5%), while the verbal consent was obtained in the remaining cases. The scale measurement was found in only 1 patient. The diagnostic concordance rate between the three consultant dermatologists was 184 of 220 cases (83.6%). No significant associated factors with the discordance were found, including sex, age, lesion type, location of the lesion, and image quality. Nevertheless, for the final diagnosis, disagreements were resolved through discussions and consensus among three experienced dermatologists. In 63.2% of all cases, the three consulting dermatologists could provide definite diagnoses. With 29.5% of cases, a definitive diagnosis could not be determined, but lists of probable diagnoses were given and case management plans were given. The three consultant dermatologists could not provide diagnoses for only 7.3% of the patients.

The diseases of the patients receiving definite diagnoses are listed in Supplementary Table 2. Eczema was the most common disorder in 33.8% of the cases, followed by cutaneous infection (14.5%) and autoimmune and connective tissue diseases (8.6%). The main reason given by the dermatologists for not making a diagnosis (Table 2) was problems with the photographic technique used. Predominant were the lack of a close-up shot (31.3%) and defocusing (25.0%).

As shown in Table 3, there were no statistically

significant differences in the age at onset, duration of symptoms, or consultation setting of the different dermatologist decision-making groups. However, in terms of disease morphology, eczematous lesions were significantly more frequent in patients with a definitive diagnosis (30.2%) than in those with probable (15.4%) and no (12.5%) diagnoses (probability value [P] = 0.037). Additionally, we demonstrated a statistically significant difference between the focus of images and the decision- making of the dermatologists. Defocus was found in 37.5% of the cases without any diagnosis, 24.6% of the probable diagnosis group, and 10.1% of the definite diagnosis group (P = 0.002).

According to Table 4, we evaluated 385 body region images captured by primary-care physicians. The common technical errors with the photographs were the lack of a close-up view (61.3%) and inadequate lighting (55.6%). Relative to the other areas of the body, the photographs of peripheral areas had a significantly higher proportion with distortion (P < 0.001), improper framing (P < 0.001), wasted space (P < 0.001), and improper background (P = 0.005). However, photographs from the truncal region had a significantly higher proportion with poor lighting than those from other regions (P < 0.001). Head and neck photographs lacked an overview shot (P = 0.049). Moreover, defocusing and lack of overview photographs were more common in the OPD setting than the inpatient setting (37.8% versus 25.9% [P = 0.046]; and 14.8% versus 4.9% [P = 0.018]). The photographs of patients under 5 years of age tended to have more wasted space than those of children over 5 (61.5% versus 38.0 %; P = 0.018).

TABLE 2. Reason for inability to provide diagnosis, as given by the dermatologists.

Reason for ambiguous diagnosis†

Number of patients

n = 16 (%)

Photographic techniques 11/16 (68.8)

No close-up shot 5/16 (31.3)

Defocusing 4/16 (25.0)

Need additional photos from specific area 3/16 (18.8)

Distortion 2/16 (12.5)

Disease-related factors

Evolution of the lesions

Need more physical examination

7/16 (43.8)

5/16 (31.3)

2/16 (12.5)

Inadequate lighting 1/16 (6.3)

†One case could have more than one cause of ambiguous diagnosis.

TABLE 3. Factors effecting dermatologists’ decision-making to give diagnoses.

* P < 0.05 indicates statistical significance, Chi2 test.

Abbreviations: IQR, interquartile range; SD, standard deviation

Bunyaratavej et al.

TABLE 4. Factors associated with photographic technical errors.

* P < 0.05 indicates statistical significance, Chi2 test.

Abbreviations: IPD, inpatient department; OPD, outpatient department

DISCUSSION

In the situation of a limited number of dermatologists, waiting times for face-to-face consultations are likely to become extended. Moreover, with a steadily growing population, dermatologists will likely face pressure to expand their services.9 Teleconsultation is a promising solution in these scenarios. Our data confirmed its benefits as our three consulting dermatologists could provide definite or probable diagnoses and management plans in most cases.

From our literature review, the quality of photographs affects the accuracy of diagnoses.6,7 We found that focus was the most critical factor since good focus was the only domain significantly associated with the ability to make a definitive diagnosis in our study. This finding underscores the need to have high-quality images for teleconsultation. It is recommended to use the flash function to avoid shadows, improve exposure, and obtain better focused and more detailed photographs than otherwise. The need for the flash function is especially critical in low-light areas such as the oral cavity (Fig 1).12,16,17 However, a loss of morphological features could occur if the flashlight is placed too close to lesions.16-18 Furthermore, photographs should reveal the distribution of the lesions and their associated anatomical structures. The morphological details of the lesions should also be obtained through a well-lit and focused close-up shot.16,17

Other techniques that should be considered are lighting, background, wasted space, framing, distortion, white balance, and color saturation. Diffused and broad- spectrum lighting is appropriate, and the light source should be oblique to the skin surface.11,16 Natural light can vary in intensity and color. While the light from a camera flash is more consistent, it can cause reflection, especially in scalp and nail photographs.11,19 Regarding background, patient-identifiable objects and distracting elements such as jewelry should be removed, and solid non-reflective backgrounds are preferable.11,12,16-18 In terms of framing, an orientation following the direction of the body area is recommended to eliminate wasted space in the background (i.e., horizontal and vertical orientation for the chin and legs, respectively).12 To take a picture of fingernails, we suggest that patients flex all fingers to bring the fingertips together (Fig 2).20 Next, distortion is expected in the photograph from a mobile phone due to the automatic wide-angle camera setting. To prevent distortion, it is recommended to zoom in slightly and move backward until the image fits the frame, with the camera placed perpendicular to the photographic plane.12 For color temperature, the skin tone might appear red in warm light, whereas a cool tone will decrease redness. Therefore, a non-neutral white balance can interfere with the diagnosis of erythematous lesions. Finally, a preset filter that affects color tone and saturation should be avoided.

Fig 1. Using a camera flash for the photographs avoids shadows and provides improved exposure results in more detailed photos, especially in low-light areas such as the oral cavity. Fig 1A demonstrates a blurry photo with poor light exposure, in which the lesion could not be identified. Fig 1B shows a focused and appropriately exposed image with the use of the camera flash.

Fig 2. Nail photography. Flexing the fingers at the proximal interphalangeal joints in order to bring all of the fingertips toward the palm will demonstrate the distribution of lesions in nail disorders.

We also analyzed factors associated with poor photographic techniques. Body regions seemed to have the most effect on photographic quality. The peripheral areas were more likely to have distortion, improper framing, wasted space, and improper background. On the other hand, truncal regions tended to have lighting problems. The photographs from the inpatient consultations frequently had inappropriate backgrounds. This finding was partly because of the more distracting elements found in the inpatient setting versus the OPD setting. However, the difference was nonsignificant. The OPD setting was associated with a lack of overview and defocusing. Possibly, this was due to the enormous number of patients visiting the OPD of primary care hospitals in Thailand, making it challenging to take high-quality images. Finally, wasted space was more frequent in pictures of patients under 5 years of age than in older patients. This finding would have resulted from the smaller body sizes of the younger patients. All of the factors above should be considered when taking photographs for teledermatology.

Finally, previous studies reported dermatitis, psoriasis, tumors, and onychomycosis as common consultation problems, with a lower frequency of cases with erythematous morphology (such as urticarial lesions).21-23 However, the typical morphology in our subjects was erythematous, followed by eczematous and tumoral or nodular lesions. As a result, more education on eczematous, tumoral, and especially erythematous lesions would be advantageous.

In terms of patient confidentiality, patient consent for photography from referring general practitioners was mentioned in only 5% of cases. In most Thai hospitals, there were no standardized forms for obtaining patient consent for photography. Given the existence of the Personal Data Protection Act in Thailand, physicians should be more conscientious about obtaining documented consent for photography.

The main limitation of this study was related to the image-quality checklist. Even though we based it on various established guidelines, there was a lack of standardized criteria for assessing photographic quality. More specifically, there were no prescribed criteria to determine whether individual photographs are acceptable for each domain (overview photographs, close-ups, focus, lighting, background, framing, and orientation, wasted space, perspective distortion, color saturation, and white balance). Another limitation was the broad spectrum of the information with various dermatologic conditions. Consultant experience also was a factor influencing the success of teledermatology but this study faced challenges in collecting relevant data. Collecting such data could prove beneficial in future research.

In conclusion, teleconsultation can increase patient access to dermatologists, especially in facilities without specialists. When mobile phones are used for teledermatology, good camera focus is the most important photographic technique. An acceptable image quality can enhance the diagnostic accuracy of dermatologists.

ACKNOWLEDGMENTS

We thank Assistant Professor Chulaluk Komoltri for help with the data analyses. We are also indebted to Mr. David Park for the English-language editing of this paper.

Conflict of Interest

All authors declare that there are no conflicts of interest related to this study.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

REFERENCES

1. Sidoroff A. Chapter 2 - The Role of Clinical Photography in Dermatology. In: Hamblin MR, Avci P and Gupta GK (eds) Imaging in Dermatology. Boston: Academic Press, 2016.p.5-11.

2. Trettel A, Eissing L and Augustin M. Telemedicine in dermatology: findings and experiences worldwide - a systematic literature review. J Eur Acad Dermatol Venereol 2018;32:215-24.

3. Warshaw EM, Hillman YJ, Greer NL, Hagel EM, MacKonald R, Rutks IR, et al. Teledermatology for diagnosis and management of skin conditions: a systematic review. J Am Acad Dermatol 2011;64(4): 759-72.

4. Kunde L, McMeniman E and Parker M. Clinical photography in dermatology: ethical and medico-legal considerations in the age of digital and smartphone technology. Australas J Dermatol 2013;54:192-7.

5. Zakaria A, Maurer T, Su G, Amerson E. Impact of teledermatology on the accessibility and efficiency of dermatology care in an urban safety-net hospital: A pre-post analysis. J Am Acad Dermatol 2019;81(6):1446-52.

6. Paradela-De-La-Morena S, Fernandez-Torres R, Martínez-Gómez W, Fonseca-Capdevila E. Teledermatology: diagnostic reliability in 383 children. Eur J Dermatol 2015;25(6):563-9.

7. Weingast J, Scheibböck C, Wurm EM, Ranharter E, Porkert S, Dreiseitl S, et al. A prospective study of mobile phones for dermatology in a clinical setting. J Telemed Telecare 2013;19(4): 213-8.

8. Krupinski EA, LeSueur B, Ellsworth L, Levie N, Hansen R, Silvis N, et al. Diagnostic accuracy and image quality using a digital camera for teledermatology. Telemed J 1999;5(3): 257-63.

9. Rubin CB, Kovarik CL. Teledermatologic care, the Affordable Care Act, and 20 million new patients: picturing the future. JAMA Dermatol 2014;150(3):243-4.

10. Marwaha SS, Fevrier H, Alexeeff S, Crowley E, Haiman M, Pham N, et al. Comparative effectiveness study of face-to-face

    and teledermatology workflows for diagnosing skin cancer. J Am Acad Dermatol 2019;81(5):1099-106.

11. Finnane A, Curiel-Lewandrowski C, Wimberley G, Caffery L, Katragadda C, Halpern A, et al. Proposed Technical Guidelines for the Acquisition of Clinical Images of Skin-Related Conditions. JAMA Dermatol 2017;153(5):453-7.

12. Muraco L. Improved Medical Photography: Key Tips for Creating Images of Lasting Value. JAMA Dermatol 2020;156(2): 121-3.

13. Dermatologists BAo. UK guideance on the use of mobile photographic devices in dermatology. 2021.

14. Dusendang JR, Marwaha S, Alexeeff SE, Crowley E, Haiman M, Pham N, et al. Association of teledermatology workflows with standardising co-management of rashes by primary care physicians and dermatologists. J Telemed Telecare 2022;28(3): 182-7.

15. Koh U, Betz-Stablein B, O’Hara M, Horsham C, Curiel- Lewandrowski C, Soyer HP, et al. Development of a Checklist Tool to Assess the Quality of Skin Lesion Images Acquired by Consumers Using Sequential Mobile Teledermoscopy. Dermatology 2022;238(1):27-34.

16. Association AT. Quick Guide to Store-Forward Teledermatology

    for Referring Providers. 2012.

17. Ashique KT, Kaliyadan F, Aurangabadkar SJ. Clinical photography in dermatology using smartphones: An overview. Indian Dermatol Online J 2015;6(3):158-63.

18. Kaliyadan F, Manoj J, Venkitakrishnan S, Dharmaratnam AD. Basic digital photography in dermatology. Indian J Dermatol Venereol Leprol 2008;74(5):532-6.

19. Quigley EA, Tokay BA, Jewell ST, Marchetti MA, Halpern AC. Technology and Technique Standards for Camera-Acquired Digital Dermatologic Images: A Systematic Review. JAMA Dermatol 2015;151(8):883-90.

20. Kaliyadan F, Ashique KT. Nail Photography: Tips and Tricks. J Cutan Aesthet Surg 2016;9(4): 254-7.

21. Kjærsgaard Andersen R, Jemec GBE. Teledermatology management of difficult-to-treat dermatoses in the Faroe Islands. Acta Dermatovenerol Alp Pannonica Adriat 2019;28(3):103-5.

22. Giavina-Bianchi M, Santos AP, Cordioli E. Teledermatology reduces dermatology referrals and improves access to specialists. EClinicalMedicine 2020;29-30:100641.

23. Cutler L, Ross K, Withers M, Chiu M, Cutler D. Teledermatology: Meeting the Need for Specialized Care in Rural Haiti. J Health Care Poor Underserved 2019;30(4):1394-406.

Effect of Wearing a Face Mask on the 6-Minute Walk Test in Healthy Volunteers

Simaporn Promsarn, B.Sc.*, Kanokwan Rattanasaengloet, M.A.**, Sutat Pipopsuthipaiboon, M.Sc.*, Nongnoot Panitchatchawal, B.Sc.*, Patharapan Lersritwimanmaen, M.D., Ph.D**

*Pulmonary Function Test Unit, Department of Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand, **Division of Respiratory Disease and Tuberculosis, Department of Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand.

ABSTRACT

Objective: This study aimed to examine the influence of wearing different types of face masks on the results of the 6-Minute Walk Test (6MWT) in a cohort of healthy volunteers.

Materials and Methods: Volunteers were partitioned into three groups (each comprising 36 individuals) with different mask-wearing conditions: NIOSH-approved N95 mask, and double-mask scenarios featuring two layers of surgical mask, and a combination of a surgical mask covered by a cloth face mask. Each participant performed two rounds of the 6MWT, one while wearing a mask and another without. Various metrics, such as the six-minute walk distance (6MWD), oxygen saturation (SpO2), and dyspnea and fatigue scores, were measured.

Results: No significant differences were found in the 6MWD results between the mask-wearing and non-mask- wearing scenarios across all the mask types. Strong positive correlations were also established between the 6MWD results in the mask-wearing and non-mask-wearing conditions. However, a small cohort experienced dyspnea significantly more when wearing double surgical masks compared to not wearing a mask. Additionally, there were no major deviations in SpO2 levels or fatigue scores regardless of the type of mask used.

Conclusion: The study indicated that wearing an N95 mask, or double surgical masks, or cloth-over-surgical face masks did not significantly influence the outcomes of the 6MWT in healthy individuals. However, caution is advised in the case of wearing double surgical masks, as this may induce a greater sense of dyspnea. This suggests that face masks can be worn safely during physical fitness and pulmonary function assessments, aligning with their essential role in viral spread prevention in daily life.

Keywords: 6-Minute walk test; face mask; oxygen saturation; dyspnea; fatigue (Siriraj Med J 2023; 75: 880-886)

INTRODUCTION

The 6-Minute Walk Test (6MWT) is a method used to assess functional capacity1 and illness severity. It is often used for comparative purposes before and after various treatments, such as lung transplantation and lung surgery, pulmonary rehabilitation, chronic obstructive pulmonary disease (COPD), pulmonary hypertension, heart failure, peripheral arterial disease, and fibromyalgia, and to assess elderly patients’ functional

capacity. In the test, the distance walked in 6 minutes (6-minute walk distance, 6MWD), oxygen saturation in the blood (SpO2), heart rate (HR), respiratory rate (RR), blood pressure (BP), dyspnea score, and fatigue score are measured. The test is easy to perform utilizing simple equipment and technology, and is cost-effective, making it popular in healthcare settings.

Since the beginning of the COVID-19 pandemic caused by the Severe Acute Respiratory Syndrome Coronavirus 2

Corresponding author: Simaporn Promsarn E-mail: simaporn.but@mahidol.edu

Received 24 August 2023 Revised 3 November 2023 Accepted 4 November 2023 ORCID ID:http://orcid.org/0009-0001-5769-1610 https://doi.org/10.33192/smj.v75i12.264856

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

(SARS-CoV-2), the World Health Organization (WHO) has recommended wearing face masks as a preventive measure in public places.2 This measure has been effective in helping prevent the transmission and spread of the virus.3 However, the use of a face mask may have an impact on the 6-Minute Walk Distance (6MWD), but there is currently no established testing protocol for the 6-Minute Walk Test (6MWT) to be performed during a pandemic situation according to the standards set by the American Thoracic Society4 or the European Respiratory Society.5 Existing studies have only explored the effects of wearing face masks on the 6MWD results during the 6MWT in sample groups, considering various types of masks, such as surgical masks or N95 masks, for individuals who have previously been infected with COVID-196, while wearing the oronasal surgical mask for patients with end-stage lung diseases7, and the use of surgical masks with healthy volunteers8, as well as the effects of wearing multiple layers of masks, such as the double-mask technique9 involving the use of two surgical masks or a combination of a surgical mask as the first layer and a cloth face mask as the second layer. These studies aimed to provide supplementary information for developing testing guidelines to be used during the COVID-19 pandemic and for other respiratory conditions. Therefore, this research aimed to add to the field by investigating the impact of wearing a single layer N95 standard face mask, the double-mask technique involving two surgical masks or a combination of a surgical mask and a cloth face mask on the 6MWD during the 6MWT. The study also involved healthy volunteers to provide additional reference test data for walking assessments during a pandemic situation and for other respiratory

conditions.

MATERIALS AND METHODS

Healthy volunteers were included in the study who were aged over 18-years old, with a body mass index (BMI) between 18.5–22.90, non-smokers, and without a history of tuberculosis, coronavirus disease (COVID-19), asthma, chronic obstructive pulmonary disease, cystic fibrosis, and no contraindications to perform the 6MWT). The sample size calculation involved randomizing volunteers into six groups, each with 18 individuals. The calculation was based on data from a literature review, specifically from Swiatek et al.8 study on the Healthy group. The average 6MWD before wearing a mask was 610.45 meters, and after wearing it, it was 605.36 meters. Since the standard deviation of the difference was unavailable, it was set to 39.62 meters to maximize the sample size, with a maximum expected difference of 30 meters.10 The

highest expected average 6MWD without a mask was

635.36 meters. An independent comparison of means between two population groups was planned using a paired t-test, with a significance level of 5% (standard value = 1.96) and a test power of 80%.

α/2 β

σ 2 (Z + Z )2

μ

2

n/group = d

d

n =

(1.96+0.84)2(39.62)2

(605.36-635.36)2

n = 16

The quantitative analysis conducted necessitated careful consideration of the sample size for non-inferiority assessment. Initially, a sample size of 16 was contemplated, but subsequently augmented to mitigate potential data loss. This augmentation constitutes a 10% increment relative to the originally computed sample size, equating to an additional two participants. Consequently, the revised sample size now stands at 18. It is imperative to elucidate that this research encompasses a comparative investigation involving two distinct cohorts: individuals who initially abstained from mask usage and subsequently commenced wearing masks, and those who initially donned masks but subsequently ceased usage. In both groups, a standardized resting period of 30 minutes was observed before mask alteration. Notably, during testing, no masks were worn by participants in accordance with the testing protocols prescribed by the American Thoracic Society/European Respiratory Society.

In the first group, volunteers were equipped with NIOSH-approved N95 masks devoid of valves, specifically the 3M™ 1860S model. Group 2 participants utilized a double-layer medical mask, with each layer comprising 3-sublayers. Group 3, wore a 3-tiered medical surgical mask covered by a cloth face mask. The cloth masks, known as Siri Masks, were tested for dust filtration according to the Thai Industrial Standards Institute (TISI) 2424-2552 / EN143:2000 (European Union Personal Protective Equipment: EU PPE) standards.

This study was approved by the Institutional Review Board of the Faculty of Medicine Siriraj Hospital (COA no. Si 1015/2021).

Statistical analysis

The study findings were reported using descriptive statistics, including frequency, percentage, standard deviation, maximum, minimum, mean and median values. The study explored various health variables, such

Fig 1. Three mask wearing scenarios tested: (A) NIOSH-approved N95 mask without a valve, specifically the 3M™ 1860S, (B) medical surgical mask, double-layered, and (C) three-tiered medical surgical mask covered by a cloth face mask.

as age, gender, weight, BMI, 6MWD, SpO2, HR, dyspnea score, and fatigue score. The statistical software SPSS® Version 18 was employed to analyze the data, utilizing inferential statistics to investigate the relationships between variables or factors, including the independent t-test and Pearson correlation coefficient. Statistical significance was determined using a p-value threshold of less than 0.05.

RESULTS

Table 1 presents the volunteer characteristics. In total, 108 individuals were included in the study, comprising 13 males and 95 females, with an average age of 37 (±9.08) years old, mean weight of 52.97 (±5.49) kg, average height of 158.70 (±6.08) cm, and mean body mass index (BMI) of 21.0 (±1.29) kg/m2.

From Table 2 and Fig 2, it can be found that the 6MWD in the sample group wearing masks and not wearing masks during the three types of tests did not show statistically significant differences at a significance level of 0.05. Additionally, the 6MWD results when wearing masks and not wearing masks showed a strong positive correlation (rsp = 0.95, p < 0.001) in the same direction. The mean difference in walking distance was

1.431 (±16.64) meters, and the concordance correlation coefficient (CCC) between the meters walked during the 6MWT with and without masks was 0.98. Furthermore, the difference in walking distance was not dependent on the type of mask used (p = 0.252). It was also found that the sample group did not have a difference in the 6MWD greater than 30 meters when wearing masks versus not wearing masks for all types of masks in all the groups.

TABLE 1. General characteristics of the study participants.

TABLE 2. 6MWT differences with or without an N95 face mask, double surgical face mask, or cloth mask over a surgical face mask.

w = With face mask; w/o = Without face mask; 6MWD = 6-minute walking distance; 6MWT = 6-minute walking test; HR = Heart rate; SpO2 = Oxygen saturation measured by pulse fingertip oximeter, Mean±SD or median(Minimum–Maximum); 1-minute and 3-minute scores correspond with measurements at 1 minute and 3 minutes after the 6MWT. Dyspnea and fatigue scores were measured with the Borg scale.

Fig 2. (A) Spearman correlation (rsp = 0.95, p < 0.001) and (B) concordance correlation coefficient (CCC) between the meters walked during the 6-minute walking test (6MWT) with and without a face mask (CCC = 0.98).

w/wo = with/without masks. ▲ = N95 face mask, ● = Double surgical face masks, ☐ = Cloth mask over a surgical face mask (n = 36).

No significant decrease in SpO2 of more than 4% was found. However, among the sample group of 34 individuals (62.03%), there was a difference in SpO2 levels when wearing masks versus not wearing masks, with a mean difference of 0.07% (±0.92) after completing the test.

There was no significant difference between wearing masks and not wearing masks in terms of the SpO2, and dyspnea and fatigue scores for the N95 masks and the combination of a cloth face mask over a medical surgical mask. However, among the sample group wearing the double-surgical masks, featuring a 3-layer surgical mask (medical face mask) with two overlapping layers, and not wearing the masks, a significant difference was found in dyspnea after completion of the test (p = 0.010) and at 1 minute (p = 0.016) and 3 minutes (p = 0.036) after completion of the test.

DISCUSSION

Six-minute walk distance

There was no significant difference in the 6MWD when wearing the N95, double surgical face masks, or cloth mask over a surgical face mask combination during the 6MWT test (p = 0.932, 0.806, and 0.990, respectively). There was also a strong positive correlation in the same direction (rsp = 0.937, p < 0.001) between the 6MWD

results and the use of any of the three mask-type scenarios.

These findings are consistent with other studies that showed there were no statistically significant differences in the 6MWD when wearing an N9510-13, surgical face mask11-15, or cloth mask.11,13 Furthermore, it was found

that wearing or not wearing a mask did not result in a difference of more than 30 meters in the 6MWD, which is clinically significant for diagnosing, predicting, and monitoring individuals with chronic lung diseases.10,16-17

Oxygen saturation

There was no significant difference in the SpO2 levels when wearing the N95, double surgical face masks, or cloth mask over a surgical face mask combination during the 6MWT test. These findings align with the study conducted by Fukushi et al.18, who found that mask-wearing does not impact oxygen saturation levels in red blood cells during any level of exercise load.19 It is possible that there may be a temporary decrease in blood oxygen saturation after performing activities or exercise, but this effect is short term and clinically insignificant.

Dyspnea and fatigue

The sample group who wore the double surgical face masks experienced significantly higher levels of dyspnea compared to when not wearing a mask during the test and after completing the test, at a statistically significant level of 0.05 (p = 0.010) one minute after the test (p = 0.016), and three minutes after the test (p = 0.036). These findings are consistent with the study conducted by Person et al.15, who found that wearing a surgical mask resulted in greater dyspnea compared to not wearing one, and with the study conducted by Kusalin et al.11, who found that surgical masks caused more dyspnea compared to N95 cloth PM2.5 masks. However, Cabanillas-Barea et al.20 found that N95 masks

resulted in more dyspnea compared to surgical masks. Additionally, Fukushi et al.18 found that wearing surgical masks or cloth masks did not worsen dyspnea during light to moderate exercise but worsened dyspnea during strenuous exercise. No significant difference in fatigue was observed between wearing or not wearing any of the three types of masks in the present study.

Limitations

This study primarily included middle-aged, healthy individuals with normal BMI and a female predominance. As such, our findings may not be readily extrapolated to diverse populations with varying demographics, obesity levels, and comorbidities, particularly pulmonary and cardiovascular disorders. Moreover, our results are specifically relevant to healthy individuals without underlying cardiac conditions. Additionally, the study’s focus was on short-term mask effects, and long-term implications were not explored. Future research should consider these limitations and investigate broader demographic profiles and long-term outcomes.

CONCLUSION

This study provides valuable insights into the impact of various mask types on 6MWT in healthy individuals. It is important to emphasize that the safety considerations related to different masking strategies discussed herein primarily apply to a cohort of healthy participants without underlying cardiopulmonary diseases.

The findings underscore that, for individuals without pre-existing cardiac or pulmonary conditions, wearing N95 masks, double surgical masks, or a cloth mask over a surgical mask during the 6MWT does not appear to significantly affect the distance walked in six minutes (6MWD), oxygen saturation (SpO2) levels, or fatigue. These results offer reassurance regarding the continued use of face masks as a crucial measure in mitigating the spread of viral infections during daily life and within the context of physical fitness and pulmonary function assessments.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the volunteered to participate in this study; Mr. Piyanat Promsarn and Research Department, Faculty of Medicine Siriraj Hospital, Mahidol University for their assistance with language editing.

Funding

This study is supported by Siriraj Research

Development Fund (Managed by Routine to Research: R2R) IO: R016535020

REFERENCES

1. ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories. ATS statement: guidelines for the six- minute walk test. Am J Respir Crit Care Med. 2002;166(1):111-7.

2. World Health Organization. WHO Coronavirus Disease (COVID-19) Dashboard. [Internet]. 2020. [cited 2020 August 18]; Available from: https://covid19.who.int/

3. Mishra A, Biswal S, Das S, Padhy RN, Pradhan BB. Social Transmission of Corona Virus: An overview. Siriraj Med J. 2020;72(6):508-11.

4. Holland A, Spruit M, Troosters R, Puhan MA, Pepin V, Saey D, et al. An official ERS/ATS technical standard: field walking tests in chronic respiratory disease. Eur Respir J. 2014;44: 1428-46.

5. ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories. ATS Statement: Guidelines for the six- minute walk test. Am J Respir Crit Care Med. 2002; 166(1): 111-7.

6. Salles-Rojas A, Guzmán-Valderrábano C, Madrid WA, Gonzalez- Molina A, Silva-Ceron M, Rodriguez-Hernandez C, et al. Masking the 6-Minute Walking Test in the COVID-19 Era. Ann Am Thorac Soc. 2021;18(6):1070-4.

7. Just IA, Schoenrath F, Passinger P, Stein J, Kemper D, Knosalla C, et al. Validity of the 6-Minute Walk Test in Patients with End-Stage Lung Diseases Wearing an Oronasal Surgical Mask in Times of the COVID-19 Pandemic. Respiration. 2021;100(7): 594-9.

8. Swiatek KM, Lester C, Ng N, Golia S, Pinson J, Grinnan D. Impact of Face Masks on 6-Minute Walk Test in Healthy Volunteers. Pulm Circ. 2021;11(1):2045894020988437.

9. Centers for Disease Control and Prevention. Maximizing Fit for Cloth and Medical Procedure Masks to Improve Performance and Reduce SARS-CoV-2 Transmission and Exposure, 2021. [Internet]. 2020. [cited 2021 October 4]; Available from: https:// www.cdc.gov/mmwr/volumes/70/wr/mm7007e1.htm

10. Singh SJ, Puhan MA, Andrianopoulos V, Hernandes NA, Mitchell KE, Hill CJ, et al. An official systematic review of the European Respiratory Society/American Thoracic Society: measurement properties of field walking tests in chronic respiratory disease. Eur Respir J. 2014;44(6):1447-78.

11. Khamtue K, Sitilertpisan P, Dacha S. Comparative effects of wearing N95, surgical, cloth, and PM2.5 masks during six- minute walk test on dyspnea, breathing effort, oxygen saturation, and functional capacity in pre-aging individuals. Journal of Associated Medical Sciences. 2023;56(2):8-17.

12. Salles-Rojas A, Guzmán-Valderrábano C, Madrid WA, Gonzalez- Molina A, Silva-Ceron M, Rodriguez-Hernandez C, et al. Masking the 6-Minute Walking Test in the COVID-19 Era. Ann Am Thorac Soc 2021;18(6):1070-4.

13. Dacha S, Chuatrakoon B, Sornkaew K, Sutthakhun K, Weeranorapanich P. Impact of wearing different facial masks on respiratory symptoms, oxygen saturation, and functional capacity during six-minute walk test (6MWT) in healthy young adults. European Respiratory Journal. 2021;58:OA1183. DOI: 10.1183/13993003.congress-2021.OA1183

    
    

14. Radtke KL, Porcari JP, Foster C, Miller M, Helget A. Evaluation of Six-Minute Walk Test (6MWT) Performance with and without a Facemask. Int J Res Ex Phys 2021;16(2):67-80.

15. Person E, Lemercier C, Royer A, Reychler G. Effect of a surgical mask on six minute walking distance. Rev Mal Respir. 2018; 35(3):264-8.

16. Miyamoto S, Nagaya N, Satoh T, Kyotani S, Sakamaki F, Fujita M, et al. Clinical correlates and prognostic significance of six-minute walk test in patients with primary pulmonary hypertension. Comparison with cardiopulmonary exercise testing. Am J Respir Crit Care Med. 2000;161(2 Pt 1):487-92.

17. Lacasse Y, Goldstein R, Lasserson TJ, Martin S. Pulmonary rehabilitation for chronic obstructive pulmonary disease.

    Cochrane Database Syst Rev. 2006;(4):CD003793.

18. Fukushi I, Nakamura M, Kuwana SI. Effects of wearing facemasks on the sensation of exertional dyspnea and exercise capacity in healthy subjects. PLoS One. 2021;16(9):e0258104.

19. Mairbäurl H. Red blood cells in sports: effects of exercise and training on oxygen supply by red blood cells. Front Physiol. 2013;4:332.

20. Cabanillas-Barea S, Rodríguez-Sanz J, Carrasco-Uribarren A, Lopez-de-Celis C, Gonzalez-Rueda V, Zegarra-Chavez D, et al. Effects of Using the Surgical Mask and FFP2 during the 6-Min Walking Test. A Randomized Controlled Trial. Int J Environ Res Public Health. 2021;18(23):12420.

Efficacy and Safety of Topical 5% Azelaic Acid Solution Versus 2% Minoxidil Solution in the Treatment of Female Pattern Hair Loss

Kanchalit Thanomkitti, M.D., Chutipon Pruksaeakanan, M.D., Chanika Subchookul, M.D., Norramon Charoenpipatsin, M.D., Daranporn Triwongwaranat, M.D., Supenya Varothai, M.D., Rattapon Thuangtong, M.D., Tanyalak Chumnumrat, B.Sc.

Department of Dermatology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand.

ABSTRACT

Objective: To determine the efficacy and safety of 5% azelaic acid solution in comparison with 2% minoxidil solution in the treatment of FPHL.

Materials and Methods: Twenty-six FPHL patients with Ludwig grade I or II were randomly treated with 5% azelaic acid solution or 2% minoxidil solution twice daily for 6 months. At baseline, 2, 4, and 6 months, hair density and hair shaft diameter were assessed at the targeted fixed area. At 6 months, patient and investigator assessments of hair growth were performed using a 7-point scale.

Results: Hair density and hair shaft diameter in the patients treated with 5% azelaic acid and 2% minoxidil solution were significantly increased compared to the baseline in all cases and visits (P < 0.05). There were no statistically significant differences in hair density and hair shaft diameter changes between both groups (P > 0.05). Both the investigator and patient assessments were comparable between both groups at 6 months. Pruritus was the major adverse effect reported in both groups, but only mild and all could be tolerated.

Conclusion: 5% Azelaic acid solution might be an effective treatment for FPHL, comparable with 2% minoxidil, and could be an alternative treatment for FPHL in minoxidil-allergic patients and pregnant women.

Keywords: Androgenetic alopecia; azelaic acid; female pattern hair loss; minoxidil allergy; pregnancy (Siriraj Med J 2023; 75: 887-893)

INTRODUCTION

Androgenetic alopecia (AGA), also known as pattern hair loss, is the most common form of nonscarring alopecia, affecting up to 80% of men and 50% of women throughout their lifetime.1 It is characterized by progressive hair loss due to miniaturization of the hair follicles, resulting in vellus transformation of the terminal hairs.2 Androgens and genetic predisposition appear to play important roles in etiopathogenesis.2,3 Current treatments of AGA that

are approved by the US Food and Drug Administration (FDA) consist of oral finasteride (5α-reductase inhibitor) in men, and topical minoxidil and low-level light therapy (LLLT) in both men and women.4,5

Concerning female pattern hair loss (FPHL), the treatment options are limited. Although the roles of androgens and genetic susceptibility are less apparent than in male AGA, oral finasteride and other antiandrogens appear to be helpful in FPHL. However, these medications are

Corresponding author: Chutipon Pruksaeakanan E-mail: chutipon.pruksa@gmail.com

Received 24 October 2023 Revised 16 November 2023 Accepted 16 November 2023 ORCID ID:http://orcid.org/0000-0002-3475-6741 https://doi.org/10.33192/smj.v75i12.266001

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

off-label and restricted, especially in women of childbearing age due to their teratogenicity and the increase in risk of breast cancer.4 Therefore, topical minoxidil is the only FDA-approved first-line medication for FPHL prescribed in general practice. However, concern has been raised in FPHL patients with pregnancy, and so minoxidil should be avoided for pregnant women.4,6 There have been reports of neonatal hypertrichosis and fetal malformations (heart, brain, and vascular) related to topical minoxidil use during pregnancy.7,8 Although topical minoxidil usage by mothers offers no harm to breastfed children according to expert consensus, minoxidil-induced hypertrichosis in a breastfed infant was recently reported.9 Therefore, minoxidil should be prescribed under supervision, especially when a substantial maternal dosage is part of the therapy and while nursing a premature baby.9

Elevated estrogen levels during pregnancy slow down the hair follicles’ normal cycle of shedding. As a result, most pregnant women with FPHL can actually have less hair loss.10 However, some of them might also suffer from ongoing hair loss during the gestational period, which may happen as a result of iron deficiency anemia, stopping the oral contraceptive pill, stress, or an imbalance of essential vitamins.11,12 Further hair thinning usually has a negative impact on quality of life and self- esteem.

Azelaic acid is an effective topical treatment for various dermatologic conditions, such as rosacea, acne, and hyperpigmentation, owing to its anti-inflammatory, antioxidant, and antibacterial properties.13 In addition, azelaic acid is also an inhibitor of 5α-reductase, which is a key enzyme in the pathogenesis of AGA.14 A recent study also showed that azelaic acid could protect hair bulge cells from ultraviolet B damage via an increase in catalase activity, and upregulate Gli1 and Gli2 expression, which could enhance telogen to anagen transition and promote hair growth.15 Therefore, azelaic acid is believed to be beneficial in the treatment of AGA, and there are many commercial topical preparations containing minoxidil solutions in combination with 5% azelaic acid. Moreover, topical azelaic acid is considered to be safe to apply during pregnancy (pregnancy category B) and breastfeeding.16 However to the best of our knowledge, there has been no previous study on the efficacy of azelaic acid topical solution monotherapy or a controlled study comparing the efficacy of azelaic acid topical solution and the standard treatment of AGA. Thus, the objective of this study was to determine the efficacy and safety of 5% azelaic acid solution in comparison with 2% minoxidil solution in the treatment of FPHL.

MATERIALS AND METHODS

Study design

This prospective, randomized, double-blinded comparative study was conducted at the Department of Dermatology, Faculty of Medicine Siriraj Hospital, Mahidol University. The study was approved by the Siriraj Institutional Review Board and informed consent was obtained from all participants. Overall, 26 FPHL women were enrolled and randomized to receive 5% azelaic acid topical solution or 2% minoxidil topical solution by block randomization (1:1 allocation ratio). The patients were instructed to apply the solution all over their thinning scalp twice daily. The patients were evaluated at months 0 (baseline), 2, 4, and 6.

Participants

FPHL patients aged ≥ 18 years, with a Ludwig classification grade I or II, were recruited in the study. They must not have received any topical or systemic hair loss treatment for at least 6 months prior to the enrollment. The exclusion criteria included patients who had other scalp, systemic, or psychiatric conditions that could be the cause of alopecia. Pregnant or breastfeeding women were also excluded.

Azelaic acid and minoxidil solution

Both 5% azelaic acid and 2% minoxidil solution were formulated by the Pharmacy Department, Faculty of Medicine Siriraj Hospital, Mahidol University. Both solutions used the same vehicles, consisting of 50% ethyl alcohol, 25% propylene glycol, and 25% purified water.

Outcome assessment

Hair density and diameter

On the day of enrollment, each patient was tattooed with 4 dots forming a 1x1 cm square on the vertex area of the scalp, using a brown cosmetic tattoo ink (Micro Pigments, Biotouch Inc., Los Angeles, CA, USA) that would gradually disappear after 6 months. At each visit, hairs in the target area were cut to approximately 1 mm in length and collected. Macrophotographs of the target area on the scalp were taken using a dermoscopic device (Dino-Lite DermaScope®, Dino-Lite, Naarden, the Netherlands) and DinoCapture software. All hairs in the target area were manually counted and reported as the total hair and terminal hair counts. The hair diameter at each visit was calculated by the mean hair diameter of ten representative hairs from the target area. Each hair was measured using an electronic external micrometer (RS PRO External Micrometer, RS Components Ltd., Corby, UK). Both the hair density and hair diameter at

each visit of all the participants were evaluated by the same author (C.P.), who was blinded to the treatment.

Global photographic review (GPR)

Standardized photographs were taken using a digital camera (Digital Canon PowerShot G15, Canon Inc., Tokyo, Japan) at baseline and 6 months. All the photographs at baseline and 6 months were evaluated by two blinded expert dermatologists (K.T. and D.T.) for assessing the improvement in the patient’s global hair volume using a 7-point scale (-3 = significant worsening; -2 = moderate worsening; -1 = slight worsening; 0 = no change; +1 = mild improvement; +2 = moderate improvement; +3 = significant improvement).

Patient’s own evaluation

After 6 months, patients evaluated their improvement of FPHL compared with the baseline using a 7-point scale, as in the GPR assessment.

Safety assessments

At every patient visit, safety evaluations of 5% azelaic acid solution and 2% minoxidil solution were conducted utilizing the combined information from the history taking, physical examination, and photoimaging. The majority of recorded adverse events were related to scalp irritation, including erythema, itching, scaling, and pruritus. Subjects rated the severity as none, mild,

moderate, or severe. Inspection of the returned container of the designated preparation allowed for the assessment of patient drug compliance.

Statistical analysis

All the data were analyzed using PASW Statistics, version 18.0 (SPSS Inc, Chicago, IL, USA). Descriptive statistics were demonstrated as the frequency, percentage, mean ± standard deviation (SD), median, and range. All the continuous data were evaluated for normality by the Shapiro–Wilk test. The paired t-test and repeated measured ANOVA were used to compare the mean hair density and hair diameter of each visit. The independent t-test was used to compare the means of the two groups. Nonparametric data were compared using the Mann– Whitney U test. Pearson chi-square test and Fisher’s exact test were used to compare the categorical data between the two groups. A P-value of <0.05 was considered to be statistically significant.

RESULTS

Demographic data of the patients

This trial enrolled a total of 26 FPHL patients with a mean age of 38.7 years. At baseline, the patients’ demographics and hair loss characteristics were similar between both treatment groups (Table 1). Most of them had grade I hair loss severity on the basis of the Ludwig classification.

TABLE 1. Demographic data of patients with female pattern hair loss treated with 5% azelaic acid and 2% minoxidil topical solution.

Hair density and hair diameter

The hair density and hair diameter of the participants treated with 5% azelaic acid and 2% minoxidil solution at each follow-up visit are shown in Table 2. Compared to the baseline, the hair density and hair diameter of both groups significantly increased after two months of treatment. There were no significant differences in the mean percentage of improvement in hair density and diameter between both groups during six months. At the end of the study, the hair density of the 5% azelaic acid solution group and 2% minoxidil solution group increased by 20.4% and 21.1%, respectively. The diameter of the hair improved by 5.3% in the group receiving 5% azelaic acid solution and 7.6% in the group receiving 2% minoxidil solution.

Global photographic review

The investigators assessment of both groups at six months were comparable. Five patients (38.5%) treated with 5% azelaic acid solution showed a mild to moderate improvement in their FPHL (4 mild improvement, 1 moderate improvement), compared with four patients (30.8%) treated with 2% minoxidil solution (1 mild improvement, 3 moderate improvement). Clinical photographs demonstrating the treatment responses

after a 6-month period of using 5% azelaic acid solution and 2% minoxidil solution are shown in Fig 1 and Fig 2, respectively.

Patient’s own assessment

Regarding the patient’s own assessment, 12 patients (92.3%) treated with 5% azelaic acid solution reported an improvement in their hair loss (5 mild improvement, 7 moderate improvement), while 11 patients (84.6%) treated with 2% minoxidil solution reported an improvement (3 mild improvement, 4 moderate improvement, 4 significant improvement) at the end of the study.

Adverse effects

The major reported adverse effect of both 5% azelaic acid and 2% minoxidil topical solution was pruritus, comprising 46.2% of patients in the azelaic acid group and 23.1% in the minoxidil group (Table 3). Most patients in both groups rated only a mild degree of pruritus, which tended to improve over time. One patient treated with 2% minoxidil solution complained of dryness, while one patient treated with 5% azelaic acid solution reported scaling. No patients reported erythema, or a burning or stinging sensation.

TABLE 2. Comparison of hair density and hair diameter at each follow-up visit after treatment with 5% azelaic acid and 2% minoxidil topical solution

Fig 1. Female pattern hair loss patients (Ludwig grade I) who received 5% azelaic acid solution twice daily for 6 months. Patient No. 1 (A baseline, B 6 months) and patient No. 2 (C baseline, D 6months).

Fig 2. Female pattern hair loss patients (Ludwig grade I) who received 2% minoxidil solution twice daily for 6 months. Patient No. 3 (A baseline, B 6 months) and patient No. 4 (C baseline, D 6 months).

TABLE 3. Reported adverse effects from 5% azelaic acid and 2% minoxidil topical solution.