(P 281) THE UTILITY OF INTRAOPERATIVE MARROW MARGIN FROZEN SECTION IN EXTREMITY BONE SARCOMA RESECTION

Objective: Achieving negative margins in the surgical resection of bone sarcomas is of paramount importance. Positive surgical margins are the more important predictor of local disease recurrence and are associated with worse overall survival. Intraoperative frozen section assessment of marrow margins remains the most widely utilized tool to assess bony margins in patients undergoing surgical resections of extremity long bone sarcomas.

Recent evidence has suggested that intraoperative frozen section assessment of bony margins is redundant in the modern era of magnetic resonance imaging. Additionally, frozen sections of marrow margins are technically challenging to prepare and interpret due to the fatty and bony nature of the tissue sample. However, the current literature evaluating the efficacy of intraoperative frozen section margin assessment rely on an intraoperative gross split specimen. Splitting and examining a gross split specimen is a technically challenging and time-consuming endeavour that may not be feasible at all centers.

The purpose of the current study was to 1) evaluate the accuracy of intraoperative frozen section marrow margins compared to final pathologic assessment, 2) identify how well frozen section analysis compares to gross intraoperative margin assessment, and 3) determine how positive intraoperative frozen sections impact surgical decision making.

Methods: Consecutive patients who underwent limb-salvage surgical resection of bone sarcomas arising from the long bone from January 2010 to January 2022 were included in the study. Patients with benign tumors, primary soft tissue tumors, and patients presenting with metastatic bone disease were excluded. Patients undergoing amputation and those who did not undergo an intraoperative frozen section assessment of marrow margins were excluded. All surgical resections were performed at a single centre.

At our institution, the standard practice is to plan the intraoperative osteotomy site based on intramedullary extent of disease visualized on the preoperative MRI. We generally aim for a margin of 1-2 cm beyond the intramedullary extent of the tumour. For patients undergoing neoadjuvant chemotherapy, the post-chemo MRI was used. Intraoperatively, the osteotomy site is marked utilizing a ruler based on joint palpation and/or fluoroscopy. Immediately following the osteotomy, the intramedullary canal is evaluated for any evidence of gross disease (Figure 1). An intraoperative pathology consultation is then initiated, and a sample of the bone marrow is curetted from the unaffected residual bone and sent for frozen section. The final intramedullary margin distance is determined by the pathologist after full review of the specimen and reported as part of the final synoptic pathology report.

Each intraoperative pathologic frozen section assessment of the bone marrow margin was counted independently. The frozen section, intraoperative gross findings and final pathologic diagnoses of marrow margins were all classified as indeterminant/positive or negative. Indeterminant cases referred to frozen sections which did not have overtly malignant cells but had abnormal pathologic features that prompted uncertainty in which the surgeon treated the margin as potentially positive.

The demographic data was reported using descriptive statistics with percentages, mean and standard deviation (SD) or median with ranges, depending on the data distribution. The final pathologic evaluation was considered the gold standard and sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated.

Results: There were 158 patients with 166 intraoperative frozen section marrow margins that were included (Table 1). The mean marrow margin on final pathologic specimen was 4.3cm (±2.6). The mean age was 45 (±21.3) and 59% of patients were male. The most common histologic subtypes were osteosarcoma (46%), chondrosarcoma (32%) and Ewing’s sarcoma (10%) with the most common anatomical location being the femur (56%), followed by the humerus (20%) and tibia (20%).

There were four indeterminant/positive frozen section margins (2.2%) in this cohort (Table 2). Two of the positive frozen sections were corroborated on final pathologic review. Both were also identified intraoperatively on gross evaluation of the cut end of the specimen. Two of the indeterminant/positive frozen sections were discordant with the final pathologic review. Both cases underwent additional bony resection and demonstrated negative intramedullary tumor margins on final pathologic review. Two patients had negative intraoperative frozen section marrow margins with positive marrow margins on final pathologic specimen.

Based on these findings, intraoperative frozen section marrow margin assessment yielded a sensitivity and specificity of 50% and 98.8% respectively (Table 3). The corresponding PPV and NPV were 50% and 98.8%, respectively. Gross examination of the marrow margin yielded a sensitivity and specificity of 50% and 98.8% respectively. The corresponding PPV and NPV were 100% and 98.8%, respectively.

Conclusion: The current study aimed to evaluate the clinical utility and diagnostic accuracy of intraoperative frozen section marrow margin in patients with extremity bone sarcomas. When compared to the final pathologic assessment, frozen section demonstrated a PPV of 50% and NPV of 99% compared to a PPV of 100% and NPV of 99% for gross intraoperative margin assessment. Positive frozen section marrow margins led to additional bony resections in three of four cases.

Frozen section marrow margin assessment did not provide additional clinical value beyond gross marrow margin assessment intraoperatively in extremity bone sarcoma resections. Frozen section assessment demonstrated a low PPV and led to unnecessary additional bone resections. The results of the current study demonstrate that examination of gross marrow margins intraoperatively is an adequate method for margin assessment and may save time and resources.