Soft tissue sarcomas are exceedingly rare tumors with a mesenchymal origin [
It has been suggested by multiple authors that percutaneous soft tissue biopsies should be performed at sarcoma referral centers under the care of experienced musculoskeletal oncologists, citing decreased accuracy and potential alterations in the clinical course when biopsies are performed in the community [
This retrospective chart review was performed under Institutional Review Board (IRB) approval with waiver of informed consent. The initial list of patients was generated using ICD and CPT codes related to percutaneous needle biopsy. All data were collected from the electronic medical records (EMRs) of one community orthopedic clinic and one large academic medical center.
We reviewed the charts of all patients who underwent ambulatory percutaneous needle biopsy of a soft tissue mass (Figure
The image shows a 40-year-old male with a painless, superficial popliteal mass seen on (a) physical examination and (b) magnetic resonance imaging. In-office biopsy results showed intermediate-grade soft tissue sarcoma of indeterminate type. Final surgical pathology results showed grade 2 fibrosarcoma.
All procedures of office-based percutaneous needle biopsy and final surgical excision of soft tissue lesions were performed by a single board-certified and fellowship-trained musculoskeletal oncologist. Informed consent was obtained following a thorough discussion of risks, benefits, and expectations prior to the completion of any procedures. Needle biopsies were performed with Tru-Cut© needles (Allegiance, Illinois, USA), taking multiple cores to maximize the amount of tissue biopsied, thus increasing the diagnostic ability [
Data collection from the final study population ( Patient demographic information (Table Lesion location Lesion diameter Lesion depth relative to fascia (superficial or deep) Result of percutaneous biopsy (benign or malignant, exact diagnosis) Result of final surgical pathological analysis (benign or malignant, exact diagnosis) Documentation of biopsy-related complications Documentation of the propriety of patients’ ultimate treatment
Characteristics | Values ( |
---|---|
Age (years, SD) | 54.5 (18.5) |
|
|
Gender ( |
|
Female | 31 (44.9) |
Male | 38 (55.1) |
|
|
Location of lesion ( |
|
Lower extremity | 51 (73.9) |
Upper extremity | 18 (26.1) |
|
|
Size of lesion ( |
|
<5 cm | 19 (27.5) |
≥5 cm | 46 (66.7) |
Cyst/abscess | 4 (5.8) |
|
|
Depth ( |
|
Superficial | 6 (8.7) |
Deep | 63 (91.3) |
|
|
Biopsy results ( |
|
Benign | 42 (62.3) |
Malignant | 27 (37.7) |
|
|
Final pathology ( |
|
Benign | 38 (55.1) |
Malignant | 31 (44.9) |
The primary outcome examined was the diagnostic accuracy of the office-based percutaneous needle biopsy. This was obtained through examination of the level of agreement between the results of the initial office-based biopsy and pathological analysis of the surgically excised soft tissue lesion. Agreement between the two biopsies was determined based on both the status of the lesion (benign versus malignant) and the exact diagnosis. All pathology reports were reviewed by a third-party board-certified pathologist in order to determine the accuracy of ambulatory percutaneous biopsy when compared to the final pathological analysis of the surgically excised masses. In order to assess safety, we included documentation of any complications determined to be related to the ambulatory biopsy. Clinical utility of the clinic-based biopsy as a diagnostic test was assessed through determination of whether appropriate clinical treatment for each patient’s final diagnosis would have been initiated with office-based CNB as the only diagnostic tool.
All statistical analysis was performed using Stata (StataCorp, Texas, USA). Accuracy was determined through the percent agreement between the exact diagnoses given by each biopsy method. Spearman’s rank correlation coefficient (Spearman’s rho) was used to quantify the level of agreement between the two biopsy techniques (benign versus malignant) and determine statistical significance of the primary outcome (
Demographic information of the 69 patients included in our final data analysis is described in Table
In our study, 4 patients had malignant lesions that were reported benign by CNB, and thus these biopsies were counted as incorrect for both exact diagnosis and determining benign versus malignant. These patients were as follows: A 45-year-old female with a left distal lateral arm mass. Initial CNB was read as “adipose and fibrous tissue with focal myxoid change.” Due to a high clinical suspicion and the fact that this mass could be excised with adequate margins without increased morbidity, a wide resection was performed. Final pathology reported the mass as a grade 1 myxoid liposarcoma. A 60-year-old male with a mass about the left proximal posterior calf, enveloping the distal popliteal neurovascular bundle. Initial biopsy at the margins of the lesion reported “spindle cell neoplasm.” Due to high clinical suspicion for malignancy, the nonspecific nature of the biopsy report, and involvement of critical neurovascular structures, the patient was given the option of open biopsy for further diagnosis. However, the patient declined this and opted for above knee amputation. The final pathology report diagnosed a “malignant solitary fibrous tumor.” A 55-year-old female with a mass about the right shoulder. Initial biopsy reported a “low-grade spindle cell lesion.” Due to high clinical suspicion and the nonspecific results of the biopsy, an open biopsy was performed at the time of planned excision. This frozen section was read as high-grade sarcoma, and thus a wide excision was performed. Final pathology was reported as a high-grade pleomorphic sarcoma. A 56-year-old male with a mass to the right anterior thigh. Initial biopsy reported skeletal muscle and mature adipose tissue. Following marginal excision, the final pathology report indicated the mass was a well-differentiated liposarcoma.
Additional 5 biopsies did not report the exact diagnosis, but were correct with regards to benign versus malignant. These patients were as follows: A 46-year-old female with a left shoulder mass. Initial CNB was read as “fragments of skeletal muscle and adipose tissue”; however, final pathology after marginal resection reported the mass as a desmoplastic fibroblastoma. A 45-year-old male with a mass about the left arm. Initial CNB reported “benign skeletal muscle.” After marginal resection, final pathology diagnosed a benign vascular malformation. A 55-year-old male with a mass to the posterior elbow. Initial biopsy reported benign skeletal and adipose tissue. This mass was marginally excised, and the final pathology report diagnosed a “benign vascular malformation with thrombosis and papillary endothelial hyperplasia.” A 50-year-old male with a mass to the right lateral thigh. Initial biopsy reported a benign lipomatous lesion. Due to clinical suspicion and the ease of wide resection without increased morbidity, wide resection was performed. The final pathology report diagnosed the mass as a “hibernoma with myxoid features.” A 84-year-old male with a mass to the right posterior thigh. Initial biopsy reported a “bland fibromyxoid spindle cell neoplasm.” Due to high clinical suspicion and the ease of wide resection without increased morbidity, wide resection was performed. The final pathology report diagnosed the mass as a perineuroma.
Upon retrospective review of all of these cases, the use of office-based core needle biopsy, in combination with clinical suspicion and the optional use of an open biopsy, leads to the correct treatment in 100% of cases.
We also assessed the clinical utility of our office-based biopsy in terms of a diagnostic test for malignancy. Overall sensitivity, specificity, PPV, and NPV were found to be 87.1%, 100.0%, 100.0%, and 90.5%, respectively. Using the results of the pathological analysis of the final surgical excision, we determined that all patients received correct treatment following the use of the office-based CNB.
Percutaneous needle biopsy has been shown by numerous studies to be safe and effective for diagnosis of soft tissue tumors when compared to open biopsy [
Currently, there is no consensus in the literature regarding the accuracy and clinical utility of office-based CNB of soft tissue masses. Skrzynski et al. examined the diagnostic accuracy and financial burden of musculoskeletal tumor biopsy, comparing clinic-based percutaneous needle techniques to open surgical biopsies. Their results showed substantially lowered cost associated with percutaneous biopsy and decreased but acceptable diagnostic accuracy when compared to open biopsy [
In this study, we determined the accuracy of the outpatient percutaneous biopsy through comparison with the pathology report from the final surgical excision. We found that percutaneous biopsy was 87.0% accurate in determining the exact diagnosis of soft tissue masses and 94.2% (
Using our data, we also investigated clinic-based CNB as a diagnostic tool for detecting malignancy. We determined the sensitivity, specificity, PPV, and NPV by comparing the results of the outpatient biopsy to the pathology of the final excision. The overall sensitivity was found to be 87.1% while the overall specificity was 100.0%. Overall PPV and NPV were 100.0% and 90.5%, respectively. The high level of specificity meant that we were likely to detect all benign masses as benign with our outpatient biopsy. However, our sensitivity and NPV indicate that some false negatives did exist and some malignant lesions were initially detected as benign. The specificity of our outpatient biopsy demonstrates that there were no false-positive results, and that all truly benign masses were identified as such. Our PPV shows that all masses identified as malignant by initial clinic-based CNB were identified as malignant by subsequent pathologic analysis of the excised masses. We determined these measures to be acceptable for using clinic-based biopsy as a diagnostic test, particularly when compared to results previously reported in the literature [
Limitations of this study centered on its retrospective nature. Retrospective chart review studies do not allow for proper randomization of subjects or control of confounding variables. Additionally, there is no potential for standardization of procedures or data collection, but this was likely mitigated through the use of a single physician’s patient population. Since this chart review covered multiple years of patient visits, multiple pathologists were involved in the reporting of biopsy results. Our study design likely minimized this impact through the use of a third-party experienced pathologist in interpreting the level of agreement between the two biopsy methods. We also recognized that this study only included those patients who underwent surgical excision of their soft tissue masses following the clinic-based biopsy. This could have led to our population including a disproportionate number of patients with malignant diagnoses. However, final surgical excision pathology was needed in our study design to be used as a gold standard with which we compared our results. The best approach to this research question would involve a double-blind, randomized clinical trial in which patients were placed into one of the two biopsy techniques. However, our study design allowed for comparison of two validated biopsy techniques in the same masses, essentially using the subjects as their own control.
Many orthopedic surgeons believe that percutaneous biopsies of soft tissue masses should only be done in sarcoma referral centers, with concerns of decreased accuracy, potential complications, and alterations in the clinical course if biopsies are performed in the community [
This investigation was performed at the University of Arizona College of Medicine, Phoenix, which is affiliated with the Banner Health system. This project was approved by the Banner Health’s Institutional Review Board.
This project was previously presented by Erin Stockwell as a scholarly project at the University of Arizona College of Medicine, Phoenix.
Each author certifies that he or she has no commercial associations (e.g., consultancies, stock ownership, equity interest, and patent/licensing arrangements) that might pose conflicts of interest in connection with the submitted article.