|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology
Management of Melanoma Patients: Benefit of Intense Follow-Up Schedule Is Not Demonstratedfor the European Organization of Research and Treatment of Cancer, Melanoma Group–Epidemiology Committee, Luxembourg Health Institute, Luxembourg Garbe et al1 claim to have performed a prospective cohort study in patients diagnosed with cutaneous melanoma, from which they conclude that intense follow-up examinations according to guidelines of the German Society of Dermatology improve survival because melanoma recurrences are diagnosed at an earlier stage. We believe that their conclusions are not legitimate because their study suffers from major methodological imperfections. Firstly, a prospective cohort design requires the constitution of a group (cohort) of patients with newly diagnosed melanoma that are followed by recording occurrence and timing of events of interest (ie, second primary, recurrence, death). Ideally, to avoid selection biases, that cohort should include all newly diagnosed melanoma patients during a defined period of time in a given study area. Prospective follow-up will allow for calculation of the numbers of person-years spent by each patient in the cohort, as well as estimation of incidence and timing of events. Accordingly, the reported study was not a prospective cohort design, but rather a cross-sectional study that expanded over a 25-month period, during which elaborate medical examinations were performed in the selected patients still alive after a diagnosis of melanoma given between January 1983 and August 1996. As a consequence, we are a bit confused by the reporting of results in the published article. Although follow-up procedures were not similar before and during the study, the estimated recurrence rates displayed in Figure 1 of Garbe et al1 were derived from the merging of recurrences detected before and during the study period. In contrast, survival curves after detection of recurrence (see Fig 2 in Garbe et al1) were drawn from patients whose recurrence was detected during the study period only, and followed after the end of the study in August 1998. Secondly, data on years of diagnosis of first primary melanoma suggest that recruited patients had been selected from a larger group of patients diagnosed with melanoma up to 13 years before the start of the study. No information is provided on factors that could have influenced the referral to the Department of Dermatology of the University of Tuebingen (Tuebingen, Germany). Melanoma patients referred to the University could have been at higher risk for disease recurrence than nonreferred melanoma patients from the same source population. As a result, proportions of patients with disease recurrence reported in the study may have been inflated. Thirdly, the only real prospective component in the reported study is the time between disease recurrence and death. But the survival advantage (see Fig 2 in Garbe et al1) of patients in whom disease recurrence was found at an earlier stage is misleading since it does not take into account the two biases related to earlier detection of a disease:2,3 lead time bias by earlier detection which increases the time between diagnosis and death artificially; and slow progressing lesions with less aggressive potential will preferably be detected (length time bias). Furthermore, we were a bit surprised that statistical testing of differences in survival curves was not adjusted for key confounding factors such as age, sex, and exact Breslow thickness of primary melanoma, as is usually done in studies on predictors of death from melanoma.4 Additionally, in an observational study, to appreciate the influence of a particular action (therapeutic, clinical, early detection) on mortality from melanoma, survival should be estimated from the date of diagnosis of the first melanoma, and not from the date of diagnosis of disease recurrence. Specific mortality data are preferable to survival, as mortality does not suffer from lead-time bias. Lastly, the suitability of intense follow-up schedule for the detection of second primary melanomas is not supported by the comparison with a control group or with reports on different follow-up protocol. AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST The authors indicated no potential conflicts of interest. REFERENCES
1. Garbe C, Paul A, Kohler-Späth H, et al: Prospective evaluation of a follow-up schedule in cutaneous melanoma patients: recommendations for an effective follow-up strategy. J Clin Oncol 21:520–529, 2003 2. Cole P, Morrison AS: Basic issues in population screening for cancer. J Natl Cancer Inst 64:1263–1272, 1980[Medline] 3. Feinstein AR. Clinical epidemiology. The architecture of clinical research. W. B. Saunders Cy, Philadelphia, 1985
4. Balch CM, Soong SJ, Gershenwald JE, et al: Prognostic factors analysis of 17,600 melanoma patients: validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 19:3622–3634, 2001
Related Article
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
