Li Ming, Morrell Stephen, Creighton Nicola, Tervonen Hanna, You Hui, Roder David, Currow David
Cancer Epidemiology and Population Health, University of South Australia, GPO Box 2471, Adelaide, SA, 5001, Australia.
Cancer Information and Analysis Unit, Cancer Institute NSW, GPO Box 41, Alexandria, Sydney, NSW, 1435, Australia.
Cancer Epidemiol. 2018 Aug;55:23-29. doi: 10.1016/j.canep.2018.04.014. Epub 2018 May 25.
Cancer survival has improved markedly in Australia for all ages but it is still lower in older patients. We hypothesize that the survival gap by age has increased. Our rationale is that treatment constraints in older people and potentially their limited participation in trials may have limited opportunities for survival gain.
Post-diagnostic five-year cancer-specific mortality rates were analysed by age group for cancers recorded on the NSW Cancer Registry. Live cases were censored on December 31st, 2012. Hazards ratios (HRs) were obtained from proportional hazards regression for 1990-99 and 2000-12 diagnostic periods, using 1980-89 as the reference, adjusting for socio-demographic factors, degree of cancer spread, and for all cancers combined, for cancer sites.
Five-year mortality reduced by diagnostic period for all cancers collectively from 53% in 1980-89 to 33% in 2000-12, with decreases for separate cancer sites. Adjusted HRs (95% confidence intervals) were 0.78 (0.77, 0.80) for 1990-99 and 0.61 (0.58, 0.63) for 2000-12 for all cancers combined. The downward trend in HRs was smaller for the 80+ year age group, leading to significantly higher HRs of 0.83 (0.81, 0.87) and 0.73 (0.70, 0.76) for 1990-99 and 2000-12 respectively. Results were similar using competing risk regression and 5-year rather than 10-year age strata.
The reduction in cancer mortality was smaller in older people, as seen in the USA. Research is needed to achieve the best trade-offs between cancer control and harm avoidance in older people. Multidisciplinary teams have an important contribution to make.
澳大利亚所有年龄段的癌症生存率均有显著提高，但老年患者的生存率仍较低。我们推测年龄导致的生存差距有所扩大。我们的理由是，老年人的治疗限制以及他们可能有限地参与试验，可能限制了生存获益的机会。
对新南威尔士州癌症登记处记录的癌症病例，按年龄组分析诊断后五年的癌症特异性死亡率。存活病例于2012年12月31日进行截尾。使用1980 - 89年作为参照，通过比例风险回归得出1990 - 99年和2000 - 12年诊断期的风险比（HRs），并对社会人口学因素、癌症扩散程度以及所有癌症合并情况、癌症部位进行调整。
所有癌症总体的五年死亡率从1980 - 89年的53%降至2000 - 12年的33%，各癌症部位均有下降。所有癌症合并的调整后HRs（95%置信区间）在1990 - 99年为0.78（0.77, 0.80），在2000 - 12年为0.61（0.58, 0.63）。80岁及以上年龄组的HRs下降趋势较小，导致1990 - 99年和2000 - 12年的HRs分别显著更高，为0.83（0.81, 0.87）和0.73（0.70, 0.76）。使用竞争风险回归以及5年而非10年年龄分层时，结果相似。
如在美国所见，老年人群中癌症死亡率的降低幅度较小。需要开展研究，以在老年人群中实现癌症控制与避免伤害之间的最佳权衡。多学科团队可发挥重要作用。
