Prevalence of Soft Tissue and Bone Tumor Diagnostic Discrepancies from Initial to Referral Sarcoma Center

Khomsit Thongthammachat, M.D.¹, Sorranart Muangsomboon, M.D.², Akarin Nimmannit, M.D.3, Charuwan Akewanlop, M.D.1, Chandhanarat Chandhanayingyong, M.D.4, Rapin Phimolsarnti, M.D.4, Jomjit Chantharasamee,

M.D.¹,*

¹Division of Medical Oncology, Department of Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand, ²Department of Pathology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand, 3Department of Research, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand, 4Department of Orthopedic Surgery, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand.

*Corresponding author: Jomjit Chantharasamee E-mail: jomjit025@gmail.com

Received 31 July 2024 Revised 15 October 2024 Accepted 24 October 2024 ORCID ID:http://orcid.org/0000-0002-1237-331X https://doi.org/10.33192/smj.v77i3.270439

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

INTRODUCTION

Soft tissue and bone sarcomas are rare types of cancer of mesenchymal origin. The incidence in Asian populations is approximately 2.8 per 100,000 people per year.1 Given the existence of more than 170 soft tissue sarcoma subtypes (STS) according to the WHO 2020 classification2, it is difficult to make an accurate diagnosis. The delay in correct diagnosis causes delay in treatment and may result in the loss of a curative opportunity. Studies that examine the rate of diagnostic discordances among different pathologists have reported variations ranging from 25% to 40%.3,4 Due to the potential for incorrect diagnoses and the rarity of the disease, improper treatment is a significant concern.1-3 Therefore, we systematically compared an initial histopathological diagnosis provided by a pathologist (first opinion, FO) with a diagnosis by a second specialized sarcoma pathologist (second opinion, SO). We analyzed patients suspected of having soft tissue or bone tumors during the years 2014 to 2019.

MATERIALS AND METHODS

Objectives

The main objective of this study was to assess the rate of pathological diagnostic discordance in soft tissue and bone tumors. The secondary objective was to investigate the duration between the date of the pathological diagnosis and the start of specific treatments, such as

surgery, chemotherapy, or radiation (individually or in combination). Furthermore, our objective was to explore the factors associated with diagnostic discordance.

Study design

We conducted a retrospective study involving patients diagnosed with soft tissue or bone tumors between 2014 and 2019 at Siriraj Hospital. The initial pathological diagnosis, provided by an external pathologist (referred to as the “first opinion’s diagnosis), was subsequently reviewed by a specialized sarcoma pathologist at Siriraj Hospital (referred to as the “second opinion diagnosis”). The revised pathological diagnosis was based on the WHO 2013 classification of Tumours of Soft Tissue and Bone (4th edition). We collected data on patient demographics, the date of the initial pathological diagnosis by the first opinion, the date of pathological diagnostic confirmation by the second opinion at Siriraj Hospital, and the date of the initiation of specific treatments, such as surgery, radiotherapy, chemotherapy, or other systemic therapies prescribed at Siriraj Hospital. Furthermore, we retrieved the results of histopathological diagnosis and pathological information, including tumor type (histological category), subgroup (family, subtype, variant), mitotic rate, size, and margin status, from the Department of Pathology database of the Faculty of Medicine Siriraj Hospital, Mahidol University.

Definition of outcomes

We categorized and evaluated the discordances between the initial pathological diagnosis provided by the ‘first opinion (FO)’ and the revised diagnosis given by the sarcoma specialized pathologist, called the ‘second opinion (SO)’, using a three-point scale.

‘Zero concordance’ was defined when the initial pathological diagnosis (by the FO) indicated a benign condition, but the revised diagnosis indicated malignancy (sarcoma), or vice versa. It was also applied when the tumor was classified into different histological types, such as a change from synovial sarcoma to liposarcoma. ‘Partial concordance’ was defined when both pathologists (FO and SO) agreed on the same histological type. However, there was a discordant result regarding either grading (including grades not initially reported) or a different subtype, for instance, a change from dedifferentiated liposarcoma to myxoid round cell liposarcoma).

‘Full concordance’ was when there was a perfect agreement between the FO and SO, encompassing histological type, subtype and grading.

The ‘Time from initial report to initiation of treatment’ was defined as the interval between the date of the first report (FO), including the preliminary diagnosis, and the date of the beginning of specific cancer treatment, such as surgery, systemic therapy, radiation or ablation.

Statistical analysis

The sample size was calculated using the nQuery Advisor program with the formula for estimating a population proportion, considering a tolerance of 40% for the prevalence of diagnostic differences11, and a two-sided 95% confidence interval of 5%. We collected data for 369 cases, accounting for approximately 10% of patients with incomplete information, resulting in a total of 400 patient data points.

To evaluate the characteristics and diagnostic concordance, we analyzed categorical data using Pearson’s χ² test or Fisher’s exact test, as appropriate for parametric data. The Kolmogorov-Smirnov test was used to assess normal and non-normal data distributions. The Mann- Whitney U test (also known as the Wilcoxon rank-sum test) was used to compare a non-parametric data. Continuous data were analyzed using the Student’s t test. The level of statistical significance was established at P = 0.05 for a two-sided test. The χ2 test was used to determine the level of concordance and the type of discordance. All analyzes were performed using SPSS® (version 23.0). A multivariate Cox proportional hazards regression analysis was performed using factors associated with concordance.

RESULTS

Baseline characteristics

All values reported subsequently for grade (graded using the Fédération Nationale des Centres de Lutte Contre le Cancer (FNCLCC) Sarcoma Group), histopathology, and the type or site of sarcoma are based on findings obtained from the second review.

Of all cases, a total of 185 patients were initially collected by pathology laboratories. Among these, 69 were reviewed by other pathologists, and 9 were not initially diagnosed with sarcoma. After applying the inclusion criteria, the final analysis included 107 patients. The initial diagnosis of soft tissue or bone tumor was made by private laboratories in 24 (24.3%) cases and by public pathological laboratories in 81 (75.5%). The characteristics of the patients are summarized in Table 1. Among the 107 patients analyzed, 54 were male (50.5%) and 53 were female (49.5%). The median age was 49 years (range, 1-97), with 24 cases (22.4%) involving visceral sarcomas. Initial immunohistochemical (IHC) results were available for 49 cases (45.7%), while 25 cases received a complete diagnosis without planned further IHC. Only one case in our study was confirmed by fluorescent in situ hybridization (FISH) testing for the rearrangement of EWSR1.

Concordance analysis

Concordance analysis was performed on 63 (58.9%) biopsied and 44 (41.1%) surgical samples. Table 2 shows the histopathological grade, type of tumor sample (biopsied sample vs. surgical specimen), and type of pathological laboratory (private vs. public). The full concordance between the diagnoses of the first and second opinion was observed in 28 (26.1%) cases, partial concordance (identical diagnosis of soft tissue or bone tumor but different in either grade or subtype) in 18 (16.8%) cases, and grade was not initially reported in 61 cases. Zero concordance (from benign to malignant tumor or vice versa or different histopathological type) was found in 61 cases (57%). Of the 64 specimens for which the classification was not initially reported, 44 were given an FNCLCC classification by the second pathologist, while 20 specimens could not be classified. The complete concordance rate was higher in patients 18 years or younger (p 0.002), in those with initial available IHC (yes vs. no = 36.7% vs. 17.2%) (p 0.027), and those with tumor size of 100 mm or more (yes vs. no = 33.3 vs. 15.3) (p = 0.04). The main discrepancies were related to histopathological type (n = 53, 49.5%) and subtype (n = 8, 7.5%), as shown in Table 1. The most common zero-concordance histopathology diagnosed by a second

TABLE 1. Patient characteristics.

TABLE 1. Patient characteristics. (Continue)

**MPNST = malignant peripheral nerve sheath tumor

# one patient with Kaposi’s sarcoma on HAART, and one with closed observation

TABLE 2. Concordance analysis of clinical variables and degree of concordance.

pathologist was an undifferentiated sarcoma (n=17). Details of the confirmed diagnoses made by the second opinion among the zero concordance group are reported in Table 3. Among 25 patients with complete initial IHC, 12 cases (48%) had complete concordance, while 6 cases had major discrepancies including 1 from sarcoma was revised to be carcinoma, 1 medullary carcinoma to rhabdomyosarcoma, 1 lymphoma to Ewing sarcoma, 1 neuroendocrine tumor to Ewing sarcoma, 1 lipoma to myxofibrosarcoma, 1 liposarcoma to no residual tumor. (Table 3)

Analysis of factors associated with a shorter time from pathological diagnosis to initiation of specific treatment (DDT).

Of 82 de novo patients, 66 had treatment records available. The mean duration from diagnosis to treatment was 68 (range: 0-272). In particular, most cases of visceral sarcoma (15 of 16, 93.7%) had a time from initial pathological diagnosis to initiation of specific treatment of less than 68 days. No significant differences were found according

to the availability of initial IHC, the type of laboratory and the complete concordance to achieve a treatment time shorter than 68 days.

Details on the time from the initial pathological diagnosis to the initiation of specific treatment are presented in Table 4.

Univariate analysis of concordance was conducted using previously established prognostic factors. Factors that demonstrated statistically significant associations with better concordance in this analysis included age less than 18 years (p = 0.004), complete initial IHC (p = 0.005) and tumor size 100 mm or more (p = 0.012). Subsequently, a multivariate Cox proportional hazards regression analysis of concordance was performed using the factors mentioned above. This analysis revealed that age less than 18 years, complete initial IHC and tumor size 100 mm or more were associated with better concordance (HR 5.48, 95% CI 1.49-20; p = 0.01, HR 4.17

and 95% CI 1.43-12.12; p = 0.009, HR 0.32 and 95%CI

0.10-0.99; p = 0.04, respectively).

TABLE 3. Diagnostic discrepancies in cases with zero concordance.

TABLE 3. Diagnostic discrepancies in cases with zero concordance. (Continue)

TABLE 4. Time from initial pathological diagnosis to the initiation of specific treatment in 66 de novo cases.

*DDT = time from initial pathological diagnosis to the initiation of specific treatment

** 16 patients had only review of slides without treatment record.

TABLE 5. Univariate and multivariate analysis of concordance.

DISCUSSION

The primary objective of this study was to assess the prevalence of diagnostic discordance and the role of centralized histological review. An accurate diagnosis is crucial for prognosis and appropriate treatment. Recent literature reports diagnostic discrepancies in the histopathological diagnosis of soft tissue sarcomas ranging from 14% to 47%.4-8 We analyzed the concordance between the initial evaluation and the second opinion, along with various factors associated with clinical and pathological conditions. Using the criteria established by the WHO 2013 Classification of Tumors of Soft Tissue and Bone, our findings indicate a full concordance rate of 33%. Additionally, the observed discordance rate of 67%, with a major discrepancy of 56% and a minor discrepancy of 11%, is higher than that reported in other sarcoma studies.5,9-11

This study confirms that establishing a sarcoma diagnosis is highly challenging, as more than 74% of initial diagnoses changed upon the second pathological review. The main result suggests that concordance appears to depend on age, initial IHC, and tumor size. These results emphasize that if the first-opinion pathologist performs immunohistochemical (IHC) studies completely prior to referral, the rate of concordance can be higher and prevent delay in diagnosis. The rationale for younger age being associated with higher concordance is unclear, possibly because most pediatric sarcoma subtypes have distinct histomorphology that most pathologists are familiar with. The most frequent discrepancies were related to

histological type. Major discrepancies were based on hematoxylin and eosin- (H&E-) stained slides, with diagnostic discrepancies associated with the difficulty of rare and unusual neoplasms. Such discrepancies were often related to undifferentiated sarcomas diagnosed as spindle cell sarcomas, malignant solitary fibrous tumors, unclassified malignant tumors, pleomorphic spindle cell sarcoma, leiomyosarcomas, endometrial stromal sarcomas, and unclassified malignant tumor. Among the cases with ‘full concordance’, osteosarcoma was the most common histopathology. However, only 25 patients completed the initial IHC before referral, possibly because most patients were referred soon after the preliminary diagnosis without waiting for a final diagnosis. Of these 25 cases, 6 had zero concordance, including a case where leoimyosarcoma changed to undifferentiated sarcoma, a case where a solitary fibrous tumor became undifferentiated pleomorphic sarcoma, 1 case where neurofibrosarcoma became spindle cell sarcoma, NOS, 1 case of rhabdomyosarcoma to medullary carcinoma, 1 case of sarcoma to carcinoma and 1 case of undifferentiated sarcoma to dermal sarcoma. These discrepancies occurred due to disagreements in IHC interpretation. Among the 61 specimens, 4 received the FNCLCC grade by the second pathologist, while 20 remained ungraded due limited amount of tissue or slide quality. Our study revealed higher histological and classification discrepancies compared to those reported in the SSG sarcoma group, where 25% of cases had their sarcoma histologic type reclassified, and 40% saw a change

in malignancy grade.12 However, it is important to note that most of the specimens referred had incomplete IHC. We also identified 24 cases of soft tissue sarcoma arising from visceral sites. The most common histological type was leiomyosarcoma (n=5), consistent with previous studies.14-16 However, one case changed its diagnosis from leiomyosarcoma to undifferentiated stromal sarcoma. Among the histopathologies with full concordance, leiomyosarcoma was the most common (4 of 5), followed

by hemangiopericytoma (2 of 2).

Our study also examined the time from diagnosis to the initiation of specific treatment. The mean duration from the initial diagnosis date to the start of definitive treatment, including surgery, chemotherapy or radiation, for newly diagnosed patients was 68 days. Patients with visceral primary tumors had a shorter time from diagnosis to initial specific treatment compared to those with extremity sites, possibly due to a higher rate of upfront surgery without prior biopsy, driven by abdominal symptoms. The maximum duration from diagnosis to treatment in our study was 272 days, primarily due to a lengthy referral process. However, we did not find any significant associations between factors such as complete initial IHC, histological type, or full concordance and the time to definitive treatment. We were unable to establish a clear association between concordance (full versus zero/partial, or full/partial versus zero) and the time to treatment. Several confounding factors influence the time to treatment, including the time required for imaging studies, the duration of retrieving pathology slides from the original hospital, and the time from the initial visit to the specialist.

The introduction of molecular genetic testing can significantly impact diagnostic accuracy, even in centers with established expertise in sarcoma.13,14 However, unlike previous studies reporting confirmation rates of 10-14% of samples diagnosed by a second pathologist through additional tests or molecular tests2,10,11,15, our study only performed such testing in one case initially diagnosed as Ewing’s sarcoma. This limitation was largely due to reimbursement problems during the time of diagnosis, as many patients were unable to afford the costs associated with these tests. However, in cases with pathognomonic histomorphology and IHC results suggestive of small round cell sarcoma, a preliminary diagnosis may be sufficient to conclude the diagnosis without the need for additional ancillary tests.

This study has several limitations. First, a significant number of specimens had incomplete IHC prior to review, which could introduce bias in the diagnostic capability

of the initial center and impact the rate of diagnostic discrepancy of sarcomas. Second, patients with ultrarare sarcomas or certain histopathological types are usually referred to specialized sarcoma centers, whereas common sarcomas with pathognomonic or specific morphology features can be diagnosed by the first pathologist without the need for referral. Third, pathologists who work in referral centers may already receive some guidance from external reports, such as indications of “suspicion of sarcoma” or “unlikely subtype”. Last, there are uncontrolled factors in the treatment process, including the time of referral, the time of specialist consultations, and the time of complete staging scans before surgery, all of which can potentially lead to delays in initiating specific treatment.

CONCLUSION

The diagnostic discrepancy rate of 74% observed in second opinion cases highlights the importance of obtaining opinions from soft tissue pathologists. However, a lingering question remains: is a centralized pathological diagnosis necessary before referral? Although a quick and accurate diagnosis can expedite the initiation of treatment in centers with access to sarcoma pathologists, we were unable to establish a clear association between diagnosis concordance and the timing of the initiation of treatment for patients initially diagnosed with sarcoma from outside hospitals.

Data Availability Statement

The data supporting this study are available upon request from the corresponding author

ACKNOWLEDGMENTS

The authors gratefully acknowledge Miss Khemajira Karaketklang of the Department of Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University for assistance with statistical analysis.

DECLARATION

Grants and Funding Information

This research was not funded.

Conflict of Interest

The authors declare that they have no known competing financial interests or personal relationships that could appear to influence the work reported in this paper.

Registration Number of Clinical Trial

None.

Author Contributions

Conceptualization and methodology, J.C, K.T, and

A.N. ; Pathology consultation, S.M; Formal analysis, J.C. and K.T. ; Visualization and writing – original draft, K.T. ; Writing – review and editing, J.C., and K.T ; Supervision, C.A., C.C, R.P. All authors have read and agreed to the final version of the manuscript.

Ethics Approval

This study was carried out according to the guidelines established in the Declaration of Helsinki and approved by the Institutional Review Board of Siriraj Hospital, Mahidol University. (approval number; COA no. Si 465/2021)

Use of Artificial Intelligence

None

REFERENCES

1. Kallen ME, Hornick JL. The 2020 WHO Classification: what’s new in soft tissue tumor pathology? Am J Surg Pathol. 2021; 45(1):e1-e23.

2. Coindre J, Nguyen B, Goussot J, de Mascarel I, Maree D, De Mascarel A, Trojani M, editors. Histological changes after chemotherapy of soft tissue sarcomas in the adult. Annales de Pathologie; 1985.

3. Harris M, Hartley AL, Blair V, Birch JM, Banerjee SS, Freemont A, et al. Sarcomas in north west England: I. Histopathological peer review. Br J Cancer. 1991;64(2):315-20.

4. Arbiser ZK, Folpe AL, Weiss SW. Consultative (expert) second opinions in soft tissue pathology: analysis of problem-prone diagnostic situations. Am J Clin Pathol. 2001;116(4):473-6.

5. Lehnhardt M, Daigeler A, Hauser J, Puls A, Soimaru C, Kuhnen C, Steinau HU. The value of expert second opinion in diagnosis of soft tissue sarcomas. J Surg Oncol. 2008;97(1):40-3.

6. Lurkin A, Ducimetière F, Vince DR, Decouvelaere A-V, Cellier

   D, Gilly FN, et al. Epidemiological evaluation of concordance between initial diagnosis and central pathology review in a comprehensive and prospective series of sarcoma patients in the Rhone-Alpes region. BMC Cancer. 2010;10:150.

7. Al-Ibraheemi A, Folpe AL. Voluntary second opinions in pediatric bone and soft tissue pathology: a retrospective review of 1,601 cases from a single mesenchymal tumor consultation service. Int J Surg Pathol. 2016;24(8):685-91.

8. Perrier L, Rascle P, Morelle M, Toulmonde M, Ranchere Vince D, Le Cesne A, et al. The cost-saving effect of centralized histological reviews with soft tissue and visceral sarcomas, GIST, and desmoid tumors: The experiences of the pathologists of the French Sarcoma Group. PLoS One. 2018;13(4):e0193330.

9. Ray-Coquard I, Montesco M, Coindre J, Dei Tos A, Lurkin A, Ranchère-Vince D, et al. Sarcoma: concordance between initial diagnosis and centralized expert review in a population-based study within three European regions. Ann Oncol. 2012;23(9): 2442-9.

10. Thway K, Fisher C. Histopathological diagnostic discrepancies in soft tissue tumours referred to a specialist centre. Sarcoma. 2009;2009:741975.

11. Thway K, Wang J, Mubako T, Fisher C. Histopathological diagnostic discrepancies in soft tissue tumours referred to a specialist centre: reassessment in the era of ancillary molecular diagnosis. Sarcoma. 2014;2014:686902.

12. Alvegård T, Berg N. Histopathology peer review of high-grade soft tissue sarcoma: the Scandinavian Sarcoma Group experience. J Clin Oncol. 1989;7(12):1845-51.

13. Schaefer I-M, Cote GM, Hornick JL. Contemporary sarcoma diagnosis, genetics, and genomics. J Clin Oncol. 2018;36(2):101- 10.

14. Asavamongkolkul A, Lektrakul N, Suntornpong N, Akewanlop C, Ruangsetakit C, Benjarassamerote S. Deadly Soft Tissue Sarcoma: Curable Using a Multidisciplinary Approach. Siriraj Med J. 2003;55(6):324-35.

15. Italiano A, Di Mauro I, Rapp J, Pierron G, Auger N, Alberti L, et al. Clinical effect of molecular methods in sarcoma diagnosis (GENSARC): a prospective, multicentre, observational study. Lancet Oncol. 2016;17(4):532-8.