A. Nadler1, K. J. Ruth2, J. M. Farma1, S. S. Reddy1 1Fox Chase Cancer Center,Surgical Oncology,Philadelphia, PA, USA 2Fox Chase Cancer Center,Biostatistics,Philadelphia, PA, USA
Introduction: The time from initial biopsy (bx) to definitive excision, or surgical interval (SI), does not appear to affect outcomes for melanoma. However, the impact of a prolonged SI where residual melanoma (RM) may be present in the final specimen is less clear. This study was undertaken to assess whether SI as it relates to the presence of RM in the final specimen affects prognosis.
Methods: A retrospective review of our institutional melanoma database from 2009 and 2011 was performed. Fisher exact tests and Kruskal Wallis tests were used to compare characteristics by RM status. Survival was estimated with Kaplan Meier methods and compared with the log-rank test. Cox proportional hazards regression was used to adjust for covariates.
Results: Of 240 patients in the database, 179 treated for non-metastatic cutaneous melanoma were included. The median age was 61 years and 51% were male (n=92). At initial bx, 60% (n=108) had a shave bx, 14% (n=25) had a punch bx, and 26% (n=46) had an excisional bx. All patients underwent a radical excision of the primary lesion and 74% (n=132) underwent a sentinel lymph node bx. The median SI was 41 days (range 8-1280) and it did not differ by biopsy type (p=0.36). On final pathology, 45% (n=81) had RM present. RM was more likely to be found in punch bx (n=19 of 25) compared to shave bx (46 of 108) and excisional bx (16 of 46) (76% vs. 43% vs. 35%, p=0.003). The presence of RM did not differ by Breslow depth (p=0.32). Median follow-up was 12.4 months (range 0.2-35.7), during which 13 patients had disease recurrence or died. Recurrence free survival (RFS) was significantly lower in patients with a longer SI (split at median ≤ 41 days for shorter SI vs. > 41 days for longer SI) (p=0.035). RFS at 12 months was 97.3% (CI 89.6-99.3) for a shorter SI and 89.7% (CI 78.1-98.3) for a longer SI; at 24 months, the RFS was 94.4% (CI 82.3-98.3) and 83.5% (CI 68.6-91.8), respectively. Differences in overall survival (OS) were borderline significant (p=0.069) with OS at 12 months was 98.7% (CI 91.2-99.8) for shorter SI and 94.7% (CI 84.3-98.3) for longer SI; at 24 months, the OS was 95.6% (CI 81.8-99.0) and 82.2% (CI 64.1-91.7), respectively. RFS and OS did not differ by RM status (p=0.20 and p=0.19, respectively). However, for SI and RM status in combination, patients who had RM and a longer SI had the lowest RFS with a 12 month RFS of 81.0% (CI.55.9-92.7) compared to 92.9% (CI 84.3-96.9) for the other groups combined (p=0.022). In multivariable analysis, with adjustment for stage, age, and Breslow depth, patients who had RM and a longer SI had borderline significance for RFS (p=0.074, hazard ratio = 2.99, 95% CI=0.90-9.98).
Conclusion: Longer SI, especially in combination with the presence of RM following initial bx, is associated with worse RFS. Patients anticipated to have RM should be triaged appropriately to avoid delays to definitive excision.