Abstract
Soft tissue tumors are rare tumors, and their histological examination remains a challenge. The establishment of the correct initial histopathologic diagnosis is critical. However, due to the rarity of soft tissue and bone tumors and the inherent difficulty of their classification and diagnostics, discrepancies may occur in up to one third of cases. For these reasons, several studies recommend the involvement of experienced pathologists frequently performing sarcoma diagnostics. Until now, there is only scarce information about how long it takes to establish a correct sarcoma diagnosis. We thus analyzed all consecutive patients presented to the Swiss Sarcoma Network Tumor Board (SSN-MDT/SB) with a primary diagnosis of a soft tissue tumor over a 2-year period (01/2019 to 12/2020) based on a tumor biopsy. We then compared the final histopathological diagnosis of two comparable institutions with similar case load, but different workflows: (i) institution A, with an initial diagnosis performed by a local pathologist, and reviewed by a reference pathologist, and (ii) institution B, with the final diagnosis performed directly by a reference pathologist. In addition, we analyzed the time from biopsy to establishment of the diagnosis. A total of 347 cases were analyzed, 196 from institution A, and 149 from institution B. In 77.6% of the cases, the diagnosis from the local pathologist was concordant with the expert review. Minor discrepancies were found in 10.2% of the cases without any consecutive changes in treatment strategy. In the remaining 12.2% of the cases, there were major discrepancies which influenced the treatment strategy directly. Establishing the final report took significantly longer in institution A (4.7 working days) than in institution B (3.3 working days; ). Our results confirm the importance of a pathological second review by a reference pathologist. We recommend direct analysis by experts, as diagnoses can be made more accurately and quickly. Within the SSN, establishing the sarcoma diagnosis is overall accurate and quick but still can be improved.
1. Introduction
Soft tissue tumors are rare tumors and histological examination remains a challenge [1]. The recently published WHO Classification of Soft Tissue and Bone Tumors [2] lists over 100 tumor entities including variants, often characterized by specific genetic aberrations, which can be detected by molecular diagnostic studies. Establishing the precise tissue diagnosis of a soft tissue or bone tumor is of utmost importance with respect to the choice of a correct treatment strategy for the patient. An incorrect histopathological diagnosis may lead to the initiation of an incorrect therapy with potentially severe or even lethal consequences for the patient [3–8].
Yet, due to the rarity of soft tissue and bone tumors and the inherent difficulty for a correct classification and diagnostic, discrepancies may occur in up to one third of cases [3–8]. For these reasons, several studies recommend the involvement of experienced pathologists who are involved in sarcoma diagnostics on a daily basis and who have access to auxiliary studies [3, 4, 9].
Various studies [4, 5, 7, 8] have shown that establishing the correct diagnosis for the treatment of soft tissue tumors is indeed a challenge, with 14% [10] to 43% [4] of all patients receiving an incorrect diagnosis, which could lead to incorrect treatment. Therefore, any multidisciplinary team (MDT) must assess these numbers constantly to compare with the reference benchmark for quality purposes.
Further, there is only scarce information on how long it takes to establish an expert review. Besides the correct diagnosis, the time from biopsy to establishing the diagnosis is an important quality indicator for the work-up of sarcoma patients. To the best of our knowledge, this factor has not yet been considered in published literature.
The patients treated in the Swiss Sarcoma Network (SSN) are either [1] referred directly to the member institutions prior to biopsy or [2] following a diagnosis of a mesenchymal tumor in an earlier outside biopsy. The current study concentrates on the first group in order to study the condition to optimize the diagnostic paths within the network. As the expansion of the network progresses in the future, there is hope that the percentage of the tissue studies outside the network (including “whoops” unintended resections) will diminish. Herein, we report first on the quality of accuracy in establishing the sarcoma diagnosis within the Swiss Sarcoma Network, and second, assess how long it takes to establish the diagnosis including expert review analysis.
2. Materials and Methods
All consecutive patients presented at the Swiss Sarcoma Network Board with a primary diagnosis of a soft tissue tumor from January 1, 2019, to December 31, 2020, were included in this study. Patients with incomplete records were excluded. A record was marked as incomplete when, for example, a case from institution A was missing an expert review, or when a case from institution B was initially diagnosed locally. The diagnoses were classified according to the WHO into benign, intermediate and malignant [11].
The biopsies of the two institutions were analyzed and compared. The samples of institution A were initially analyzed by the local pathology institute. This is a general pathology institute without specific subspecialization. Afterwards, the samples being reviewed and assessed by a reference institute pathologist specialized in soft tissue tumors. Conversely, institution B cases were assessed directly by the reference institute pathologist. These workflows are illustrated in Figure 1.
To determine the time from biopsy to the establishment of the diagnosis, the days between the arrival of the tissue specimen at the pathology institute until the date of the final report were calculated. Weekend days or holidays were not counted, unless the report was issued on one of these days. In the analysis of the current study only cases which can be diagnosed by conventional histopathologic staining, immunohistochemistry, and FISH were included, as these studies have a short turn-around-time of one to two days. The cases requiring PCR or NGS based analyses were excluded as they methodically require several days independently of the performance of the pathologist.
The accuracy of the diagnoses of the local histopathology institute A and the expert analysis was examined in a second step. Here, the diagnoses of institute A were compared with the expert opinion and divided into 3 groups according to the classification of Thway et al. [6]:(i)Cases without diagnostic discrepancy between local and reference institutions were classified into category A(ii)Category B includes cases with minor discrepancy in diagnosis but without therapeutic consequences(iii)Category C contains all cases where the diagnosis from the reference pathologist changed the treatment
In addition, all cases where the final report from institution A did not establish a diagnosis were consequently classified under category C [6].
The data were collected using the Adjumed ® -Database (www.adjumed.ch; Zurich, Switzerland) and analyzed with the statistical package “stats” of the open source software “R” [12].
The cantonal ethic commission has approved the application of the Swiss Sarcoma Network under the agreement number BASEC-NR 2019-01107. The study is also registered on https://climincaltrials.gov with the number NCT04300257 [13].
3. Results
3.1. Patient and Tumor Characteristics
A total of 347 cases were analyzed, 196 from Institution A and 149 from Institution B. 179 patients were female and 168 were male, and the median age was 55 (range 12–90) years. 163 cases were classified as benign (46.9%), 114 cases were malignant (32.8%), 66 cases were intermediate (19%), and 4 cases were unclassifiable (see Table 1).
The most common benign diagnosis was lipoma (69 cases, 42.3% of all benign tumors), followed by Schwannoma (11 cases, 6.7%). Regarding the malignant diagnosis, undifferentiated/unclassified sarcoma was the most common diagnosis (19 cases, 16.6% of all malignant tumors) followed by the dedifferentiated liposarcoma (14 cases, 12.2%, see Table 2).
3.2. Accuracy
Of the 196 tumors specimens from institution A (which underwent initial diagnosis by a local pathologist followed by specimen being reviewed by a reference pathologist, see Figure 1), 152 tumors (77.6%) were diagnostically concordant according to category A. Of the latter 152 tumors, 46.7% were benign, 18.4% were intermediate, 33.5% were malignant, and 1.4% unclassifiable. There were 20 cases (10.2%) with minor discrepancies, according to category B (Table 3). Of these, 70% were malignant, 15% intermediate, and 15% benign diagnoses. There were 24 tumors (12.2%) with major diagnostic discrepancies (Table 3) according to category C. 50% of these were malignant cases. From these major discrepancies, 12 cases were classified in this category because of a missing diagnosis in the final report from institution A. In one case, there was a reclassification from benign to malignant and one case was reclassified from malignant to benign. A summary of all original diagnoses, which were discordant from the expert review is shown in Table 3.
3.3. Analysis of Time to Diagnosis
Establishing the final report took on average 4.7 working days for institution A, which is significantly longer than the 3.3 days required by institution B (Figure 2). 10 cases were excluded from the analysis (7 from institution A and 3 from institution B) due to the necessity of NGS for the final diagnosis. We analyzed the data with a two-sided Wilcoxon t-test and found a value of .
If only malignant diagnoses were considered for analysis, establishing the diagnosis averaged 5.2 days in institution A, and 3.4 days, respectively, for institution B (, see Figure 2).
According to the most commonly diagnosed lesion of all, the diagnosis of a lipoma averaged 4.6 days at institution A and 3.2 days, respectively, at institution B (). Accordingly, and with respect to undifferentiated/unclassified sarcoma, institution A required 5.2 days, and institution B 3.0 days ().
4. Discussion
To the best of our knowledge, this is the first analysis comparing the duration of a histological review to establish a sarcoma diagnosis. Our results confirm the importance of a second pathological review by a reference pathologist. With an overall concordance of 77%, the results are comparable to the already published literature.
In 1986, Presant et al. [7] first reported on a histopathologic peer review of specimens from 216 consecutive patients with soft-tissue or bone sarcomas by a panel of three pathologists. Subtype of sarcoma, degree of confidence in diagnosis, and grade were compared with agreement or disagreement in pathologic opinion from the primary member institution versus the pathology review panel. There was a complete agreement between the primary pathologist and reviewer in 66% of cases. However, after the review, 12 cases (6%) were considered not to be sarcoma. In 27% of cases, the subtype of sarcoma was felt to be incorrect by reviewers.
In 2008, Lehnhardt et al. [5] reviewed 603 patients who were operated with the diagnosis of soft tissue sarcoma. They found a concordance in primary diagnostics of 28.3% for pathologists in private clinics, 29.6% for hospital affiliated pathologists, 36.8% for academic medical centers, and 70.5% for the department of pathology at their institution.
In 2010, Lurkin et al. [8] analyzed all histological data of all patients diagnosed with sarcoma in the Rhone-Alpes region between March 2005 and February 2006. Primary diagnoses were systematically compared with second opinions from regional and national experts. They included 366 patients; of these, 199 (54%) had full concordance between primary diagnosis and second opinion, 97 (27%) had partial concordance (identical diagnosis), and 70 (19%) had complete discordance.
Ray-Coquard et al. [4] reviewed the histological data of patients diagnosed with sarcoma in Rhone-Alpes (France), Veneto (Italy) and Aquitaine (France) over a 2-year period. Initial diagnoses were systematically compared with the second opinions from members of the group of pathologists of the GSF-GETO (French Unicancer Sarcoma Group). 1463 cases matched the inclusion criteria and were analyzed. Full concordance between primary and second diagnosis was observed in 824 (56%) cases, partial concordance in 518 (35%) cases and complete discordance in 121 (8%) cases.
A summary of the studies can be found in Table 4.
Interestingly, and specifically contrasting the analysis between benign and malignant lesions, the uncertainty to establish the correct diagnosis was greater in malignant lesions. Considering the analysis of minor discrepancies in the diagnosis comparing first line with expert review, the expert review delivers more diagnostic details or a supplement in the classification without obvious consequences regarding the treatment modality, specifically also for malignant diagnoses.
Our study has several limitations: The number of biopsies analyzed is still relatively small, and many diagnoses are benign, thereby not allowing further subgroup analysis. Also, considering the rarity of the disease and the 68 sarcoma entities included therein, further subtype analysis is not possible. The definition of diagnostic discordances is not always obvious and may skew the results. Arbitrarily, descriptive pathology reports without specification of dignity were classified as major discrepancies because adequate treatment can only be initiated when the final diagnosis is made.
Although there is a significant difference in the time to diagnosis, one may critically question to what extent this value has an influence on the time to diagnosis and further therapy. The time it takes to establish the histological examination is only one step on this path. It would therefore be interesting if a further study examines not only the duration of the biopsy, but the entire process from the suspected diagnosis to the initiation of the correct therapy. But from the point of view of the patient who must wait for a diagnosis, every day that is gained with a faster diagnosis is worth a lot. In addition, a rapid histological diagnosis is essential for a timely discussion at the multidisciplinary sarcoma board.
Any additional examination, especially if not done in the same institution, will lead to delays in the diagnostic process.
Several studies confirmed that a centralized pathological review improved the quality of the diagnosis. Lurkin et al. [8] support the direct analysis by an expert pathologist because of the multitude and complexity of sarcoma tumors. Also, the access to molecular biology analysis can be provided. Compared to the recommendation of the ECCO Essential Requirements for Quality Cancer Care, the pathway of Institution B is to be favored [14].
In a small country like Switzerland, and with sarcoma being a rare disease, establishing the correct pathological diagnosis is very challenging. The main reason is the small amount of cases per individual hospital. Compared with the volume of international sarcoma reference centers, the data of the entire country needs to be pooled and shared to reach high enough numbers for expert experience and teaching purposes. With the recently established Swiss Sarcoma Network, allowing real-world outcome analytics, there is the possibility to improve the precision, timeliness, and accuracy of sarcoma diagnosis in Switzerland in the near future. As of now, 7 central referral institutions joined the Swiss Sarcoma Network so far and benefit from a second opinion by an expert pathologist.
There is no clear definition in the literature on how a sarcoma expert is defined. As for the pathologists, the sarcoma experts within the Swiss Sarcoma Network are defined by their specific training, their specific sarcoma interest, defined by dedication of >30–50% of their duty time spent on treating sarcoma patients, their yearly scientific contributions, their number of cases reviewed and/or treated per year, and their participation of the weekly multidisciplinary tumor board including the number of discussed cases and strategic decisions.
5. Conclusions
The diagnosis of sarcoma remains challenging. According to our study and the current literature, an expert review by an experienced pathologist within a network such as the Swiss Sarcoma Network proves to be highly useful and beneficial for the patient both regarding accuracy and timeliness to establish the diagnosis. Establishing the sarcoma diagnosis as early as possible after biopsy is a critical quality indicator for a multidisciplinary team. Considering the rapidly rising health care costs, the potential increase in cost efficiency of such a process needs to be determined next.
Data Availability
The data used to support the findings of this study may be released from the Adjumed®-Database upon request to the Swiss Sarcoma Network ([email protected]).
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
The authors thank Dr. Julien Gérard Anet for his support with the statistical analysis.