This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Garbe, C. Articles by Leiter, U. Search for Related Content PubMed Articles by Garbe, C. Articles by Leiter, U. Related Articles Related Article Social Bookmarking                            What's this?

In Reply:

Department of Dermatology, Eberhard Karls University, Tübingen, Germany

Dr Autier and his colleagues report on the conclusions of our article,¹ stating that we claimed intense follow-up examinations according to the guidelines of the German Society of Dermatology would improve survival because melanoma recurrences are diagnosed at an earlier stage. In their letter-to-the-editor, Dr Autier and colleagues also discussed a number of methodological points as to why our study is not suitable to demonstrate this survival benefit.

If Dr Autier and his colleagues actually did read our paper, they did not seem to read carefully. Our conclusion is that the survival benefit of such a surveillance strategy has to be established in comparative prospective trials and that the present data may serve as a solid basis for planning and conducting future randomized trials. Furthermore, we did not suggest that the surveillance strategy of the German guidelines is the most favorable follow-up schedule. On the contrary, we suggested that the intensity of clinical and technical examinations can be reduced during the follow-up of patients in the primary tumor stages and may be intensified in locoregional disease.

We demonstrated in our article¹ that such a follow-up strategy is suitable to primarily detect recurrences during follow-up examinations in 84% of all patients, and that the first hint on detection of developing metastasis came from the patient in only 17% and not — as in retrospective studies in the literature — in about 50% of cases. Additionally, we demonstrated that about 50% of all recurrences were detected in such an early stage that R₀ resection seemed appropriate, and that this subgroup had a much more favorable prognosis than those with late detection. It is imaginable that an intense follow-up strategy may contribute to early detection of recurrences and to a survival benefit. Nevertheless, we were very cautious in our conclusions and stated in the discussion: "It is important to ascertain whether the effort necessary to implement this quality of follow-up protocol is actually of benefit to the patient in prolonging their survival time. The present study design is not suitable to provide a definitive answer to this question since comparative cohorts with different follow-up schedules are required."

Although Dr Autier and his colleagues did not report our main points and conclusions correctly, it may be of interest for planning and conducting future follow-up studies to discuss the methodological points they raised.

The first point to discuss would be how to define a cohort or study group for a study on surveillance of cancer patients. The definition of such a cohort or study group obviously depends on the purpose and main end points of the study. The main end points of our study were the diagnostic sensitivity of different examination techniques for detection of recurrences and their potential impact on the course of the disease. Therefore, definition of the cohort or study group does not necessarily require only inclusion of newly diagnosed melanoma patients. The disadvantage of this design would be a recruitment time of several years and a follow-up time as long as suggested by the respective surveillance schedule (10 years in Germany). Our cohort consisted of all patients who were under surveillance at the Department of Dermatology in Tübingen at the beginning of 1996 and, additionally, all patients who were newly diagnosed between 1996 and 1998. For every patient, the first visit during the study period was used as the start time for the calculation of the prospective follow-up time. Therefore, calculation of the number of person-years spent by each person in the cohort is possible, as is the estimation of the incidence of recurrences and of their timing. Survival curves were only calculated for those patients who developed recurrences during the study period, and the time of diagnosis of recurrences was used as start time (recurrences before start of the study were not included in this analysis; the patients, however, were classified in their respective stage of the disease at time of inclusion).

This definition of a cohort for follow-up investigations resembles the actual practice of patient’s surveillance and allows for the inclusion of patients in different time spans of their follow-up. Thus, patients from the first to the tenth year of their follow-up have been included, as have patients whose follow-up was prolonged as a result of previous recurrences. This study design means that results can be received within a few years of follow-up. In the special case of our study, all patients were already documented prospectively (since 1983) in a computerized database with their complete tumor data and course of the disease. Therefore, inclusion of previous data on recurrences for the calculation of hazard rates was possible, starting at primary diagnosis of the tumor. Since the 1980’s, the follow-up schedule in Tübingen did not change, and later on, the German guidelines were based on these recommendations.²

Secondly, all patients under surveillance in our department at the start of the study and those newly diagnosed within the study period were included into the study cohort. It is well known that hospital-based study groups may differ from the entire disease population of a given geographic area. In our case, the broad majority of melanoma patients were referred to the Department of Dermatology in Tübingen. The department organizes the follow-up in a cooperative manner with dermatologists and general practitioners in their private offices. The composition of our study group shows that 66% of patients had primary melanoma with less than one millimeter tumor thickness, which probably closely resembles the situation in Germany. Nevertheless, we would suggest in every case to perform follow-up studies with hospital-based study groups, because a population-based study group seems to be impractical for such an investigation (we would be interested how the European Organization of Research and Treatment of Cancer [EORTC] melanoma group would address this topic).

Thirdly, the question we addressed by our survival calculation was whether our classification of early or late detected recurrences translates into a survival advantage. The idea behind this question is to use this kind of classification as a potential surrogate marker for future studies on the surveillance of melanoma. Naturally, we were aware of the possibilities of lead-time and length-time bias and stated the following: "The objection that the later appearance of subsequent death is a result of an earlier detection of the first metastasis cannot be totally invalidated."

Furthermore, adjustment of our survival analysis for additional potentially prognostic factors would not add important information to answer our question as outlined above.

In addition, Dr Autier and colleagues did not seem to understand the purpose of the survival analysis as described above. We regret if we have not pointed out this purpose clearly enough in our article. In order to answer the question whether early or late detection of the recurrent disease translates into survival benefit, the calculation has to be done from the time point of detection of relapse.

Lastly, our study "employed a prospective, descriptive design to examine the effectiveness of a specific follow-up strategy". We did not intend to perform a comparative study. However, the results of our study may provide valuable data for planning and conducting future randomized trials.

In conclusion, we reported extensive data on the results of a prospective, descriptive study on surveillance of melanoma patients that may serve to define a more rational approach to follow-up examinations. We stressed that future randomized trials are required. We feel that it is not sufficient just to claim that most follow-up examinations are unnecessary.³ Therefore, we would appreciate any contribution of the EORTC Melanoma Group to improve the empirical basis for the design of optimized schedules for the surveillance of melanoma patients.

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[Abstract/Free Full Text]

2. D’Hoedt B, Stroebel W, Stutte H, Rassner G. Nachsorge des malignen Melanoms an der Tübinger Hautklinik. In Orfanos CE, Garbe C (eds.) Das maligne Melanom der Haut. Munich, Zuckschwerdt, 1990, pp 304–311

3. Eggermont AM: Monitoring of patients with stage I melanoma after excision of the primary tumor. Ann Dermatol Venereol 122:292–297, 1995[Medline]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

Related Article

• Management of Melanoma Patients: Benefit of Intense Follow-Up Schedule Is Not Demonstrated
  Philippe Autier, Jan Willem Coebergh, Mathieu Boniol, Jean-François Doré, Esther de Vries, and Alexander M.M. Eggermont
  JCO 2003 21: 3707 [Full Text]

This article has been cited by other articles:

				C. Voit, M. Kron, J. Rademaker, M. Schwurzer-Voit, W. Sterry, L. Weber, C. Ozdemir, T. Proebstle, and U. Keilholz Molecular Staging in Stage II and III Melanoma Patients and Its Effect on Long-Term Survival J. Clin. Oncol., February 20, 2005; 23(6): 1218 - 1227. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Garbe, C. Articles by Leiter, U. Search for Related Content PubMed Articles by Garbe, C. Articles by Leiter, U. Related Articles Related Article Social Bookmarking                            What's this?