Queensland Skin and Cancer Foundation, Queensland Institute of Dermatology, South-Brisbane, Queensland, Australia.
QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia; Faculty of Medicine, The University of Queensland, Brisbane, Queensland, Australia.
J Am Acad Dermatol. 2020 Nov;83(5):1406-1414. doi: 10.1016/j.jaad.2020.03.076. Epub 2020 Apr 1.
The literature surrounding survival of patients with multiple primary melanomas (MPM) yields variable and opposing findings, constrained by statistical challenges.
To critically examine the available literature regarding survival of patients with MPM compared with a single primary melanoma and detail statistical methods used.
Electronic searches were performed of PubMed, Embase, Web of Science, and Scopus, with cross-checking of references, for the period January 1956 to June 2019. Studies published in English examining survival in patients with multiple melanomas were included. Case studies and small case series were excluded.
There were 14 studies eligible for inclusion. Conclusions on survival varied markedly depending on the statistical method used. Four studies that accounted for survival bias by partitioning the survival time were included in the quantitative review, with 3 of these reporting a survival disadvantage for MPM, whereas the fourth showed no difference in survival. The pooled hazard ratio was 1.39 (95% confidence interval, 1.07-1.81) but with significant heterogeneity (I = 96.8%, P < .001).
Studies showed significant heterogeneity in methodology.
When data were analyzed with robust statistical methods, patients with MPM had a survival disadvantage compared with patients with a single primary melanoma.
有关多发性原发性黑色素瘤（MPM）患者生存的文献得出的结果存在差异且相互矛盾，这受到统计挑战的限制。
批判性地检查关于 MPM 患者与单一原发性黑色素瘤相比的生存情况的现有文献，并详细说明所使用的统计方法。
对 PubMed、Embase、Web of Science 和 Scopus 进行了电子检索，并对参考文献进行了交叉核对，检索时间为 1956 年 1 月至 2019 年 6 月。纳入了发表在英文期刊上的研究，这些研究检查了多发性黑色素瘤患者的生存情况。排除了病例研究和小病例系列研究。
有 14 项研究符合纳入标准。根据使用的统计方法，对生存的结论存在显著差异。有 4 项研究通过分割生存时间来考虑生存偏差，并将其纳入定量综述，其中 3 项研究报告 MPM 的生存劣势，而第 4 项研究则表明生存无差异。合并的危险比为 1.39（95%置信区间，1.07-1.81），但存在显著的异质性（I = 96.8%，P < 0.001）。
研究在方法学上存在显著的异质性。
当使用稳健的统计方法进行数据分析时，与单一原发性黑色素瘤患者相比，多发性原发性黑色素瘤患者的生存情况较差。
