癌症存活的存活年-年代-世代模型: 應用於臺灣肝癌，1997-2016年

A Survivorship-Period-Cohort Model for Cancer Survival: Application to Liver Cancer in Taiwan, 1997-2016

彭彥騰(Yan-Teng Peng)

指導教授 ： 李文宗

國立臺灣大學/公共衛生學院/流行病學與預防醫學研究所/碩士(2020年)

https://doi.org/10.6342/NTU202002891

摘要

癌症的預後是病患、醫師和公共衛生學者們所共同關心的議題。傳統監視癌症預後的世代方法有時間延遲的問題。本研究提出存活年-年代-世代模型解釋存活年、年代及世代效應對癌症預後的影響，並進行癌症存活的未來預測。我們以此模型分析臺灣肝癌相對存活資料(範圍涵蓋1997至2016年)。我們以此模型進行肝癌相對存活的未來預測，填補較新診斷病患世代未完整的追蹤。我們採用交叉驗證選擇外推方法，並利用拔靴法求得肝癌相對存活的95%信賴區間。我們發現，臺灣男性肝癌相對存活的長期趨勢(從1998至2016日曆年代或從1997至2015診斷年世代)是上升的。另外，隨著診斷後追蹤年的增加，相對存活有越明顯先升再降後又再升的現象。臺灣女性肝癌診斷後第一至五年的相對存活的上升趨勢較男性肝癌小。隨著診斷後追蹤年的增加，相對存活有越明顯先升再降後又再升然後呈現平穩的現象。臺灣男性肝癌的五年累積相對存活，隨著診斷年世代的越近現代而持續上升。臺灣女性肝癌的五年累積相對存活有先升再降後又再升的現象。臺灣男性和女性肝癌皆發現在2003年診斷世代後五年累積相對存活有明顯上升的趨勢。本研究所提出之存活年-年代-世代模型能夠協助釐清存活年、年代、世代效應對於癌症預後的影響。存活年-年代-世代模型也能夠即時監視癌症預後及推估未來趨勢。存活年-年代-世代模型值得推薦使用。

關鍵字

存活年-年代-世代模型 ； 相對存活 ； 肝癌 ； 癌症監視 ； 長期趨勢

並列摘要

The prognosis of cancer is a common concern for patients, physicians and public health researchers, alike. The traditional cohort approach of monitoring the prognosis of cancer has the problem of time delay. This study proposes a survivorship-period-cohort (SPC) model to elucidate the roles of survivorship, period, and cohort effects on cancer prognosis, and to predict the future trend in cancer survivals. We use the SPC model to analyze relative survivals of liver cancer in Taiwan (year of diagnosis from 1997 to 2016). We use the SPC model to predict future trend in relative survivals of liver cancer, by imputing the data of incomplete follow-up of some recently diagnosed patient cohorts. We use cross-validation to select the extrapolation method, and use bootstrapping method to estimate the 95% confidence intervals for the relative survivals of liver cancer. We found that long-term trend in relative survivals of male patients with liver cancer in Taiwan (calendar year from 1998 to 2016 or year of diagnosis from 1997 to 2015) is increasing. In addition, with more years since diagnosis, the more profoundly that relative survivals of male patients with liver cancer will rise first, fall and then rise again. The increase in relative survivals of female patients with liver cancer in Taiwan from the first year since diagnosis to the fifth year is smaller than that of male patients with liver cancer. With more years since diagnosis, the more profoundly that relative survivals of female patients with liver cancer will rise first, fall, rise again, and then level-off. The five-year cumulative relative survivals of male patients with liver cancer in Taiwan continues to increase as year of diagnosis becomes more recent. By contrast, the five-year cumulative relative survivals of female patients with liver cancer in Taiwan increase first, decrease, and then increase again. The five-year cumulative relative survivals for both male and female liver cancers in Taiwan increase after the 2003 year-of-diagnosis cohorts. The SPC model proposed in this study can help elucidate the roles of survivorship, period, and cohort effects on cancer prognosis. The SPC model can also monitor cancer prognosis in real time and predict future trends. The SPC model is recommended for use.

Keywords: survivorship-period-cohort model, relative survival, liver cancer, cancer surveillance, long-term trend

並列關鍵字

survivorship-period-cohort model ； relative survival ； liver cancer ； cancer surveillance ； long-term trend

參考文獻

1. Cutler SJ, Ederer F. Maximum utilization of the life table methodin analyzing survival. J Chron Dis. 1958; 8, 699-712.

2. Kaplan EL, Meier P. Nonparametric estimation from incompleteobservations. J Am Stat Assoc. 1958; 53, 457-481.

3. Brenner H, Gefeller O. An alternative approach to monitoring cancer patient survival. Cancer. 1996; 78(9): 2004-2010.

4. Brenner H, Long-term survival rates of cancer patients achieved by the end of the 20th century: a period analysis. Lancet. 2002; 360(9340): 1131-1135.

5. Brenner H, Hakulinen T. Up-to-date long-term survival curves of patients with cancer by period analysis. J Clin Oncol. 2002; 20(3): 826-832.