|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 25, No 28 (October 1), 2007: pp. 4503-4504 © 2007 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.12.2895
Regarding Adjuvant Radiation Therapy in Merkel Cell Carcinoma: Selection Bias and Its Affect on Overall SurvivalDepartment of Radiation Oncology, Memorial Sloan-Kettering Cancer Center, New York, NY
Department of Therapeutic Radiology, School of Medicine, Yale University, New Haven, CT To the Editor: We read with great interest the article by Mojica et al1 regarding adjuvant radiation therapy and overall survival in Merkel cell carcinoma (MCC) of the skin. Several issues should be considered before accepting the conclusion that radiation therapy enhances overall survival in patients with MCC based on a retrospective analysis of a heterogenous population. The analysis by Mojica et al suffers from a significant selection bias. The authors apparently included all cases from diagnosis until death. Because the Survival, Epidemiology, and End Results (SEER) registry offers little information regarding comorbidity and complications related to surgery, some investigators have censored patients who died within several months of surgery or diagnosis from analysis.2-4 Patients with significant postoperative complications, poor performance status, and substantial comorbidity may not have been considered appropriate candidates for adjuvant radiation therapy, and many of these patients would not have been able to complete a typical course of adjuvant radiation therapy. Of the SEER MCC patients who died within 4 months of diagnosis (n = 58); 39 received surgery alone and had a 34% survival at the third month. Of the remaining 19 patients who received surgery plus radiation therapy, 68% were alive at the third month. Including this early death population introduces a strong selection bias in favor of the radiation therapy group. Once this potential bias is removed, adjuvant radiation therapy no longer significantly improves overall survival (P > .08). Furthermore, according to the 1988 SEER Extent of Disease coding manual, the information pertaining to extent of disease (ie, size of primary, lymph node involvement, local extension) was reported to SEER within a 4-month period after the initial diagnosis. Therefore if a patient died within this 4-month time period, his/her SEER information may not be complete and as such should be excluded from analysis. Within the survival analysis, lymph node dissection did not seem to be statistically significant,1 which is in stark contrast to the work by Allen et al.5 In the Mojica et al analysis, of the 603 patients analyzed in their survival analysis, 291 did not have data for the cancer directed therapy. Thus, survival data for those 291 patients were included in the reference group, which may explain why no statistically significant benefit was demonstrable. This variable should have been excluded from the survival analysis because the data set was incomplete. Although SEER historic staging may be reliable, many investigators, and certainly the largest treatment analyses, have evaluated patients based on either the Memorial Sloan-Kettering Cancer Center (MSKCC) or the American Joint Committee on Cancer staging systems, which are considered by many to be more robust than the SEER historical system.2,5-7 Although MCC was first described by Toker in 1972, very few instances had been described in the literature by 1980. However, MCC remained a challenging cancer to diagnose even by the most skilled pathologists until the mid-1980s when immunohistochemical techniques were refined to specifically identify MCC and differentiate it from other neuroendocrine tumors. Although the reported data set spans from 1973 to 2002, it is not clear why patients before 1988 were considered in the analysis, as pathological confirmation of MCC was not coded in SEER until 1988. We investigated the SEER MCC data and presented the findings at the same meeting as the authors of the current study, but our analysis revealed no overall survival benefit for adjuvant radiation therapy.2 When we performed a multivariate analysis of the data (n = 478), radiation therapy was not associated with an improvement in overall survival.2 The statistical model included MSKCC stage, age, and adjuvant radiation therapy, excluding those patients who died within 4 months of diagnosis.2 Adjuvant radiation therapy was not associated with a significant mortality hazard ratio (HR) in the Cox proportional hazard multivariate analysis. The mortality HR for radiation therapy was 0.83 (95% CI, 0.63 to 1.09).2 The mortality HR for radiation therapy by MSKCC stage was: stage I HR, 1.01 (95% CI, 0.65 to 1.56), stage II HR, 0.82 (95% CI, 0.46 to 1.45), and stage III HR, 0.73 (95% CI, 0.46 to 1.15).2 The analysis offered by Mojica et al is limited by selection bias. Once correcting for the selection bias, adjuvant radiation therapy does not significantly improve overall survival based on our independent analysis. SEER is an excellent resource, but because of the inability to access and evaluate potentially confounding data based on the limitations of the SEER registry (as mentioned by the authors), any definitive results related to therapeutic maneuvers must be considered with extraordinary caution. Variables such as chemotherapy use, margin status, radiation therapy dose, technical radiation therapy detail, comorbidities, and local control are not readily available from the SEER registry. These factors and those we described earlier, may have substantial influence over survival findings. Definitive therapeutic conclusions should be tempered by the lack of access to these potentially important pieces of information. AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST The author(s) indicated no potential conflicts of interest. REFERENCES
1. Mojica P, Smith D, Ellenhorn JD: Adjuvant radiation therapy is associated with improved survival in Merkel cell carcinoma of the skin. J Clin Oncol 25:1043-1047, 2007 2. Housman DM, Smith BD, Wilson LD: Staging, prognosis, and treatment of merkel cell carcinoma: A population-based study abstract of the 88th Annual Meeting of the American Radium Society, Maui, Hawaii, May 6-10, 2006. Oncology (Williston Park) 20:1-62, 2006[Medline] 3. Housman DM, Smith BD, Detterbeck FC, et al: The role of radiation therapy in malignant thymomas: A population-based study. Int J Rad Oncol Biol Phys 66:S150, 2006 4. Smith BD, Smith GL, Cooper DL, et al: The cutaneous B-cell lymphoma prognostic index: A novel prognostic index derived from a population-based registry. J Clin Oncol 23:3390-3395, 2005 5. Allen PJ, Bowne WB, Jaques DP, et al: Merkel cell carcinoma: Prognosis and treatment of patients from a single institution. J Clin Oncol 23:2300-2309, 2005 6. Medina-Franco H, Urist MM, Fiveash J, et al: Multimodality treatment of Merkel cell carcinoma: Case series and literature review of 1024 cases. Ann Surg Oncol 8:204-208, 2001[Medline] 7. McAfee WJ, Morris CG, Mendenhall CM, et al: Merkel cell carcinoma: Treatment and outcomes. Cancer 104:1761-1764, 2005[CrossRef][Medline]
Related Reply
Related Article
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
