Department of Epidemiology, School of Medicine, University of California, Irvine, 224 Irvine Hall, Irvine, CA 92697, USA.
Breast Cancer Res. 2010;12(1):R4. doi: 10.1186/bcr2467. Epub 2010 Jan 8.
Recent, international declines in breast cancer incidence are unprecedented, and the causes remain controversial. Few data sources can address breast cancer incidence trends according to pertinent characteristics like hormone therapy use history.
We used the prospective California Teachers Study to evaluate changes in self-reported use of menopausal hormone therapy (HT) between 1995 to 1996 and 2005 to 2006 and age-adjusted breast cancer incidence among 74,647 participants aged 50 years or older. Breast cancer occurrence was determined by linkage with the California Cancer Registry.
During 517,286 woman years of follow up, 565 in situ and 2,668 invasive breast cancers were diagnosed. In situ breast cancer incidence rates in this population did not change significantly from 2000 to 2002 to 2003 to 2005, whereas rates of invasive breast cancer declined significantly by 26.0% from 528.0 (95% confidence intervals (CI) = 491.1, 564.9) per 100,000 women in 2000 to 2002 to 390.6 (95% CI = 355.6, 425.7) in 2003 to 2005. The decline in invasive breast cancer incidence rates was restricted to estrogen receptor-positive tumors. In 1996 to 1999 and 2000 to 2002 invasive breast cancer incidence was higher for women who reported current HT use especially estrogen-progestin (EP) use at baseline than for never or past users; but by 2003 to 2005 rates were comparable between these groups. For women who were taking EP in 2001 to 2002,75% of whom had stopped use by 2005 to 2006, incidence had declined 30.6% by 2003 to 2005 (P = 0.001); whereas incidence did not change significantly for those who never took HT (P = 0.33).
Few data resources can examine prospectively individual HT use and breast cancer diagnosis. Stable in situ breast cancer rates imply consistent levels of screening and suggest recent declines in invasive breast cancer to be explained predominantly by changes in HT use.
最近，乳腺癌发病率在国际范围内呈前所未有的下降趋势，其原因仍存在争议。很少有数据源可以根据激素治疗使用史等相关特征来评估乳腺癌发病率趋势。
我们使用前瞻性加利福尼亚教师研究，评估了 1995 年至 1996 年和 2005 年至 2006 年期间绝经后激素治疗（HT）自我报告使用情况的变化，以及 74647 名年龄在 50 岁或以上参与者的年龄调整后乳腺癌发病率。通过与加利福尼亚癌症登记处的联系确定乳腺癌的发生情况。
在 517286 名女性随访年中，诊断出 565 例原位和 2668 例浸润性乳腺癌。在该人群中，原位乳腺癌发病率从 2000 年至 2002 年至 2003 年至 2005 年没有显著变化，而浸润性乳腺癌发病率从 528.0（95%置信区间（CI）=491.1，564.9）/100000 名女性显著下降至 390.6（95%CI=355.6，425.7）在 2003 年至 2005 年。浸润性乳腺癌发病率的下降仅限于雌激素受体阳性肿瘤。在 1996 年至 1999 年和 2000 年至 2002 年，与从未使用或过去使用过激素治疗的女性相比，报告当前 HT 使用（特别是雌激素-孕激素（EP））的女性的浸润性乳腺癌发病率更高；但到 2003 年至 2005 年，这些组之间的比率相当。对于 2001 年至 2002 年正在服用 EP 的女性，其中 75%的人在 2005 年至 2006 年期间停止使用，到 2003 年至 2005 年，发病率下降了 30.6%（P=0.001）；而从未使用过 HT 的女性发病率没有显著变化（P=0.33）。
很少有数据资源可以前瞻性地检查个体 HT 使用和乳腺癌诊断。原位乳腺癌稳定的发病率意味着筛查水平一致，并表明最近浸润性乳腺癌的下降主要归因于 HT 使用的变化。
