A NEW MATHEMATICS, AND NEW COMPARATIVE EFFECTIVENESS CALCULATORS (CancerMath.net), FOR PREDICTING MELANOMA DEATH

Le résumé fourni par la source

BACKGROUND: Predicting the risk of melanoma death is both a challenge, and an essential step in indentifying the best treatment for each patient. METHODS: The binary-biological model of cancer metastasis captures, in mathematical terms, the most common mechanism of cancer death: the lethal spread of cancer cells, and contains equations that can tease out the independent impact on mortality of primary tumor size, nodal status, and other prognostic factors (the SizeAssessment and Prognostic Measurement methods) as well as combining information on primary tumor size, nodal status, and other prognostic factors into estimates of the risk of cancer death (the SizeOnly, Size+Nodes and Size+Nodes+PrognosticFactors [SNAP] methods). RESULTS: 13 prognostic factors were found to make marked independent contributions to melanoma lethality: thickness, nodal status, Clark's level II, Clark's level II, mixed epithelioid-spindle cell melanoma, desmoplastic melanoma, female, skin of upper limb & shoulder, Clark's level II, nodular melanoma, ulcerated, skin of scalp & neck, acral lentiginous melanoma. The SNAP method reveals that each millimeter of primary tumor thickness is associated with ~8% risk of death while each positive lymph node is associated ~23% extra risk of death. The accuracy of the SNAP method was confirmed with data on ~2700 melanoma patients seen at the MGH, and ~90,000 patients from the SEER national dataset. CONCLUSIONS: The equations of the binary-biological model were found to make highly accurate estimates of the risk of melanoma death, and provided a basis for web-based comparative effectiveness calculators (http://www.CancerMath.net) for estimating the risk of death for each patient.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.