The relationship between age andmelanoma prognosis is growingmore apparent and presentsinteresting scientific and social questions.My colleagues and I publishedtwo papers analyzing melanoma patientsfrom our institution. Our firstpaper examined a population of 620patients during a 26-year period, andour most recent paper analyzed 1,018melanoma patients over 30 years.[1,2]In both of these studies, age remainedan important prognostic predictor ofdisease-free and disease-specific survivalbased on multivariate analysis(Cox proportional hazard). We alsoapplied a novel classification and regressiontree (CART) evaluation ofthe data that showed age maintaininga significant influence on disease-freesurvival. Age maintained importancein disease-specific survival when genderwas used as the first parameter tosegregate the entire patient populationbefore applying tree-structuredstatistics.
The relationship between age and melanoma prognosis is growing more apparent and presents interesting scientific and social questions. My colleagues and I published two papers analyzing melanoma patients from our institution. Our first paper examined a population of 620 patients during a 26-year period, and our most recent paper analyzed 1,018 melanoma patients over 30 years.[1,2] In both of these studies, age remained an important prognostic predictor of disease-free and disease-specific survival based on multivariate analysis (Cox proportional hazard). We also applied a novel classification and regression tree (CART) evaluation of the data that showed age maintaining a significant influence on disease-free survival. Age maintained importance in disease-specific survival when gender was used as the first parameter to segregate the entire patient population before applying tree-structured statistics.
Depending on the thickness of the melanoma, age break points of 57.5 and 76.5 years old were important parameters of disease-free survival for melanomas less than or equal to 2.18 mm, with the older patients faring worse. For melanomas over 2.18-mm thick, the age split at 71.5 years was significant, again with older patients demonstrating a worse 5- and 10-year disease-free survival. In disease-specific survival for lesions less than 1.63 mm, females had better survival if they were 58.5 years of age or younger. Men with melanoma over 1.25 mm thick faired worse if they were older than 57.5 years old. We have also seen that the proportion of patients with nodular melanoma increased directly with age, and we have been under the impression that, compared with younger patients, a greater number of elderly patients (over 60) present late in the course of their disease.
The importance of the paper by Drs. Swetter, Geller, and Kirkwood is that it highlights the need to establish more rigorous evaluation and screening for melanoma in the elderly. A national screening program would be a significant proactive, preventive measure that would likely save lives and prevent morbidity. One simple explanation for the increased incidence in the elderly is that this is an expanding population, thanks to improved health care and longevity. Also, in their youth, the current elderly were not aware of the risk factors related to ultraviolet light, and many of our patients admit to unprotected sun exposure with burns as children and young adults. It is possible that this contributes to the incidence in the elderly.
It is important to remember that surgery remains the primary treatment for melanoma and plays a significant role in treating metastatic melanoma when deemed resectable. Overall, surgery for melanoma in the elderly is well tolerated. With the advent of sentinel lymph node biopsy, the morbidity of regional lymph node evaluation is diminished and feasibility is increased. Such assessment should be offered for melanomas 1 mm or more in thickness or lesions less than 1 mm if ulcerated or Clark's level IV. Besides the problem of early detection in the elderly, the authors raise several other important concerns about age. The increasing medical comor bidities that arise with age may affect the clinician's decision as to administering adjuvant or supplemental therapeutic treatments. Indeed, we really don't know the effects of or tolerance to a variety of biologic or chemotherapeutic agents in the elderly because they have not been thoroughly studied. This is frequently because of agebased exclusion criteria in clinical studies. In addition, comorbid medical conditions are often part of clinical trial exclusions and such medical conditions are more common in older patients, again making older patients more frequently ineligible to enroll in a clinical study.
The authors have appropriately raised the issue of undertreatment in the elderly. It is possible that one reason why older patients fare worse is that they are not receiving the same treatment as their younger counterparts. It is possible that some physicians may accept narrower surgical margins and not be as aggressive with staging or follow-up studies; however, this possibility is hard to assess.
The contribution of immune response to melanoma survival in the elderly is in need of serious evaluation. In a recent article by Sondak et al, examining the correlation between mitotic index and the rate of positive sentinel lymph nodes, the investigators found a continuum of results, with fewer positive sentinel lymph nodes in elderly patients compared to younger patients. Chao et al with Kelly McMaster's group reported similar findings showing that the frequency of sentinel lymph node metastasis decreased with increasing age. However, our data and others' seems to show that increasing age is a marker for a worse prognosis.
If Sondak's and McMaster's findings hold consistent, then what is the meaning of fewer positive sentinel lymph nodes in older patients, and how does this correlate with the risk of disseminated disease? Does hematogenous spread occur more frequently in the elderly, as these authors hypothesize? Are lymph nodes in the elderly less able to physically "capture" tumor and control dissemination? Chao et al report that as age increased, melanoma thickness, ulceration, and regression were found to increase. They also found that more of the older patients were male. This further supports the data and concepts presented in the manuscript by Swetter and coauthors.
Since elderly patients apparently fare worse than their younger counterparts and given that nodal assessment in the elderly may eventually prove not to hold the same significance as in younger patients, we must focus on the two key parameters common to both: thickness and ulceration. A corollary to this thought is that screening may carry an even greater impact in the elderly than in the young. The potential impact of a nationwide screening program targeting our older generations, as recommended by the authors, may greatly improve early diagnosis in this expanding population, which in turn may decrease surgical morbidity, improve survival, and decrease overall medical care costs.
Financial Disclosure:Dr. Averbook is a speaker for Schering Oncology/Biotech.
Averbook BJ, Russo LJ, Mansour EG: Along-term analysis of 620 patients with malignantmelanoma at a major referral center. Surgery124:746-756, 1998.
Averbook BJ, Fu F, Rao S, et al: A longtermanalysis of 1018 patients with melanomaby classic Cox regression and tree-structuredsurvival analysis at a major referral center:Implications on the future of cancer staging.Surgery 132:589-604, 2002.
Sondak VK, Taylor JM, Sabel MS, et al:Mitotic rate and younger age are predictors ofsentinel lymph node positivity: Lessons learnedfrom the generation of a probabilistic model.Ann Surg Oncol 11:247-258, 2004.
Chao C, Martin RCG II, Ross MI, et al:Correlation between prognostic factors and increasingage in melanoma. Ann Surg Oncol11:259-264, 2004.
Balch CM, Soong S-J, Gershenwald JE,et al: Prognostic factors analysis of 17,600melanoma patients: Validation of the AmericanJoint Committee on Cancer melanoma stagingsystem. J Clin Oncol 19:3622-3634, 2001.