As reported in the Journal of Clinical Oncology, Lo et al developed a novel risk calculator for sentinel node metastasis—the Melanoma Institute Australia Nomogram—that appears to more accurately predict risk of metastasis in patients with primary cutaneous melanoma than other commonly used risk models.
Study Details
The study involved analysis of data from 3,477 patients with melanoma who underwent sentinel node biopsy at Melanoma Institute Australia. A new nomogram was developed that replaced body site and Clark level from the Memorial Sloan Kettering Cancer Center model with mitotic rate, melanoma subtype, and lymphovascular invasion, while retaining patient age, tumor thickness, and ulceration in the risk nomogram.
Key Findings
- In the Melanoma Institute Australia cohort, 729 (21.0%) of 3,477 patients were sentinel node–positive.
- Compared with the Memorial Sloan Kettering model, the Melanoma Institute Australia model improved predictive accuracy by 6.2% (P < .001), from a receiver operating characteristic area under the curve C-statistic of 67.7% (95% confidence interval [CI] = 65.3%–70.0%) to 73.9% (95% CI = 71.9%–75.9%).
- National Comprehensive Cancer Network® (NCCN®) and ASCO/Society of Surgical Oncology (SSO) criteria recommend sentinel node biopsy on all patients with stage T2 disease and above, resulting in a C-statistic of 53.7% (95% CI = 52.7%–54.7%).
- Among the 2,748 sentinel node–negative patients in the Melanoma Institute Australia cohort, sentinel node biopsy would not have been offered to 22.1% using the Melanoma Institute Australia model, compared with 13.4% using the Memorial Sloan Kettering model and 12.4% using NCCN or ASCO/SSO criteria.
- In external validation in a cohort of 3,496 patients (28.4% sentinel node–positive) from The University of Texas MD Anderson Cancer Center, the Melanoma Institute Australia model had a C-statistic of 75.0% (95% CI = 73.2%–76.7%).
The investigators concluded: “A robust nomogram was developed that more accurately estimates the risk of sentinel node positivity in patients with melanoma than currently available methods. The model only requires the input of six widely available clinicopathologic parameters. Importantly, the number of patients undergoing unnecessary sentinel node biopsy would be significantly reduced compared with use of the Memorial Sloan Kettering Cancer Center nomogram or the NCCN or ASCO/SSO guidelines, without losing sensitivity. An online calculator is available at melanomarisk.org.au.”
Serigne N. Lo, PhD, of Melanoma Institute Australia, North Sydney, is the corresponding author for the Journal of Clinical Oncology article.
Disclosure: The study was supported by the Australian National Health and Medical Research Council, Medical Foundation of the University of Sydney, Melanoma Institute Australia, and others. For full disclosures of the study authors, visit ascopubs.org.
