In this article (Cancer Epidemiol Biomarkers Prev 2010;19:2479–87), which was published in the October 2010 issue of Cancer Epidemiology, Biomarkers & Prevention (1), the authors discovered errors in some of the data. The authors repeated the analyses of Tables 1–4 using the corrected data and revised accordingly. Descriptions of the revisions and the revised tables are shown below.
Characteristics of breast cancer cases diagnosed during 1997 to 2007 at KPNC and matched controls, by disease and hormone receptor status
| Characteristics . | All controls N = 208,640 . | All cases N = 20,864 . | HR-negative cases N = 3,669 . | HR-positive cases N = 17,195 . |
|---|---|---|---|---|
| Age at diagnosis/index date (y): mean ± SD | 61.88 ± 13.00 | 61.76 ± 13.00 | 58.59 ± 13.24 | 62.65 ± 12.84 |
| Years of prescription drug coverage: mean ± SD | 7.43 ± 3.18 | 7.43 ± 3.18 | 7.37 ± 3.15 | 7.44 ± 3.19 |
| Oral contraceptive usea: n, (col%) | 13,492 (6.5) | 1,618 (7.7) | 335 (9.1) | 1,283 (7.5) |
| Menopausal hormone therapy useb: n, (col%) | 68,347 (32.8) | 7,673 (36.8) | 1,203 (32.8) | 6,470 (37.6) |
| Lipophilic statin use: n, (col%) | ||||
| Never use | 174,537 (83.6) | 17,464 (83.7) | 3,147 (85.8) | 14,317 (83.3) |
| “Ever” use | 34,103 (16.3) | 3,000 (16.3) | 522 (14.2) | 2,878 (16.7) |
| “≥2 years” use | 18,343 (8.8) | 1,850 (8.9) | 257 (7.0) | 1,593 (9.3) |
| Years of lipophilic statin use: mean ± SD | ||||
| Among “Ever” users | 0.91 ± 1.72 | 0.87 ± 1.72 | 0.75 ± 1.53 | 0.90 ± 1.76 |
| Among “≥2 years” users | 4.48 ± 2.21 | 4.55 ± 2.29 | 4.35 ± 2.27 | 4.59 ± 2.29 |
| Race/ethnicityc: n, (col%) | ||||
| Non-Hispanic White | n/a | 15,283 (73.2) | 2,322 (63.3) | 12,961 (75.4) |
| Hispanic White | n/a | 1,386 (6.6) | 318 (8.7) | 1,068 (6.2) |
| African American | n/a | 1,597 (7.6) | 517 (14.1) | 1,080 (6.3) |
| Asian/Pacific Islander | n/a | 2,193 (10.5) | 419 (11.4) | 1,774 (10.3) |
| Other | n/a | 405 (1.9) | 93 (2.5) | 312 (1.8) |
| Characteristics . | All controls N = 208,640 . | All cases N = 20,864 . | HR-negative cases N = 3,669 . | HR-positive cases N = 17,195 . |
|---|---|---|---|---|
| Age at diagnosis/index date (y): mean ± SD | 61.88 ± 13.00 | 61.76 ± 13.00 | 58.59 ± 13.24 | 62.65 ± 12.84 |
| Years of prescription drug coverage: mean ± SD | 7.43 ± 3.18 | 7.43 ± 3.18 | 7.37 ± 3.15 | 7.44 ± 3.19 |
| Oral contraceptive usea: n, (col%) | 13,492 (6.5) | 1,618 (7.7) | 335 (9.1) | 1,283 (7.5) |
| Menopausal hormone therapy useb: n, (col%) | 68,347 (32.8) | 7,673 (36.8) | 1,203 (32.8) | 6,470 (37.6) |
| Lipophilic statin use: n, (col%) | ||||
| Never use | 174,537 (83.6) | 17,464 (83.7) | 3,147 (85.8) | 14,317 (83.3) |
| “Ever” use | 34,103 (16.3) | 3,000 (16.3) | 522 (14.2) | 2,878 (16.7) |
| “≥2 years” use | 18,343 (8.8) | 1,850 (8.9) | 257 (7.0) | 1,593 (9.3) |
| Years of lipophilic statin use: mean ± SD | ||||
| Among “Ever” users | 0.91 ± 1.72 | 0.87 ± 1.72 | 0.75 ± 1.53 | 0.90 ± 1.76 |
| Among “≥2 years” users | 4.48 ± 2.21 | 4.55 ± 2.29 | 4.35 ± 2.27 | 4.59 ± 2.29 |
| Race/ethnicityc: n, (col%) | ||||
| Non-Hispanic White | n/a | 15,283 (73.2) | 2,322 (63.3) | 12,961 (75.4) |
| Hispanic White | n/a | 1,386 (6.6) | 318 (8.7) | 1,068 (6.2) |
| African American | n/a | 1,597 (7.6) | 517 (14.1) | 1,080 (6.3) |
| Asian/Pacific Islander | n/a | 2,193 (10.5) | 419 (11.4) | 1,774 (10.3) |
| Other | n/a | 405 (1.9) | 93 (2.5) | 312 (1.8) |
aOral contraceptive “ever” use within 10 years before diagnosis/index date.
bHormone therapy “ever” use within 5 years before diagnosis/index date.
cRace information was captured by the KPNCCR and was available for cases only.
Risk of HR-negative or positive breast cancer associated with statin use
| Statin use . | HR-negative cases N = 3,669 n (col%) . | Matched controlsaN = 36,690 n (col%) . | OR [95% confidence interval (CI)]b . | HR-positive cases N = 17,195 n (col%) . | Matched controlscN = 171,950 n (col%) . | OR (95% CI)d . |
|---|---|---|---|---|---|---|
| Never | 3,147 (85.8) | 31,474 (85.8) | Reference | 14,317 (83.3) | 143,063 (83.2) | Reference |
| ≥2 years | 257 (7.0) | 2,720 (7.4) | 0.93 (0.81–1.08) | 1,593 (9.3) | 15,623 (9.1) | 1.02 (0.96–1.08) |
| Never | 3,147 (85.8) | 31,474 (85.8) | Reference | 14,317 (83.3) | 143,063 (83.2) | Reference |
| <1 y | 132 (3.6) | 1,257 (3.4) | 1.04 (0.87–1.25) | 654 (3.8) | 6,808 (4.0) | 0.95 (0.88–1.04) |
| ≥1–2 y | 133 (3.6) | 1,239 (3.4) | 1.07 (0.88–1.28) | 631 (3.7) | 6,456 (3.7) | 0.97 (0.89–1.06) |
| ≥2–3 y | 91 (2.5) | 829 (2.3) | 1.09 (0.87–1.36) | 465 (2.7) | 4,717 (2.7) | 0.98 (0.89–1.08) |
| ≥3–5 y | 96 (2.6) | 1,063 (2.9) | 0.89 (0.72–1.11) | 593 (3.4) | 5,990 (3.5) | 0.99 (0.90–1.08) |
| ≥5 y | 70 (1.9) | 828 (2.3) | 0.83 (0.65–1.08) | 535 (3.1) | 4,916 (2.9) | 1.08 (0.98–1.19) |
| Statin use . | HR-negative cases N = 3,669 n (col%) . | Matched controlsaN = 36,690 n (col%) . | OR [95% confidence interval (CI)]b . | HR-positive cases N = 17,195 n (col%) . | Matched controlscN = 171,950 n (col%) . | OR (95% CI)d . |
|---|---|---|---|---|---|---|
| Never | 3,147 (85.8) | 31,474 (85.8) | Reference | 14,317 (83.3) | 143,063 (83.2) | Reference |
| ≥2 years | 257 (7.0) | 2,720 (7.4) | 0.93 (0.81–1.08) | 1,593 (9.3) | 15,623 (9.1) | 1.02 (0.96–1.08) |
| Never | 3,147 (85.8) | 31,474 (85.8) | Reference | 14,317 (83.3) | 143,063 (83.2) | Reference |
| <1 y | 132 (3.6) | 1,257 (3.4) | 1.04 (0.87–1.25) | 654 (3.8) | 6,808 (4.0) | 0.95 (0.88–1.04) |
| ≥1–2 y | 133 (3.6) | 1,239 (3.4) | 1.07 (0.88–1.28) | 631 (3.7) | 6,456 (3.7) | 0.97 (0.89–1.06) |
| ≥2–3 y | 91 (2.5) | 829 (2.3) | 1.09 (0.87–1.36) | 465 (2.7) | 4,717 (2.7) | 0.98 (0.89–1.08) |
| ≥3–5 y | 96 (2.6) | 1,063 (2.9) | 0.89 (0.72–1.11) | 593 (3.4) | 5,990 (3.5) | 0.99 (0.90–1.08) |
| ≥5 y | 70 (1.9) | 828 (2.3) | 0.83 (0.65–1.08) | 535 (3.1) | 4,916 (2.9) | 1.08 (0.98–1.19) |
aControls matched to HR-negative cases bases on age at diagnosis/index date and years of prescription drug coverage.
bORs based on conditional logistic regression among HR-negative cases and their matched controls, adjusted for oral contraceptive and hormone therapy use.
cControls matched to HR-positive cases bases on age at diagnosis/index date and years of prescription drug coverage.
dORs based on conditional logistic regression among HR-positive cases and their matched controls, adjusted for oral contraceptive and hormone therapy use.
Risk of intrinsic breast cancer subtypes associated with statin use
| Study population . | Statin use ≥2 years . | Statin use “Never” . | OR (95% CI)a . |
|---|---|---|---|
| Luminal A | |||
| Controls | 8,889 (15.8) | 47,182 (84.1) | Reference |
| Cases | 931 (16.4) | 4,728 (83.5) | 1.05 (0.97–1.14) |
| Luminal B | |||
| Controls | 1,837 (13.6) | 11,658 (86.4) | Reference |
| Cases | 176 (12.9) | 1,185 (87.1) | 0.96 (0.81–1.15) |
| HER2+/ER− | |||
| Controls | 684 (12.2) | 4,916 (87.8) | Reference |
| Cases | 70 (12.3) | 498 (87.7) | 1.02 (0.77–1.36) |
| Triple negative | |||
| Controls | 1,234 (12.7) | 8,461 (87.3) | Reference |
| Cases | 111 (11.5) | 851 (88.5) | 0.88 (0.71–1.10) |
| Study population . | Statin use ≥2 years . | Statin use “Never” . | OR (95% CI)a . |
|---|---|---|---|
| Luminal A | |||
| Controls | 8,889 (15.8) | 47,182 (84.1) | Reference |
| Cases | 931 (16.4) | 4,728 (83.5) | 1.05 (0.97–1.14) |
| Luminal B | |||
| Controls | 1,837 (13.6) | 11,658 (86.4) | Reference |
| Cases | 176 (12.9) | 1,185 (87.1) | 0.96 (0.81–1.15) |
| HER2+/ER− | |||
| Controls | 684 (12.2) | 4,916 (87.8) | Reference |
| Cases | 70 (12.3) | 498 (87.7) | 1.02 (0.77–1.36) |
| Triple negative | |||
| Controls | 1,234 (12.7) | 8,461 (87.3) | Reference |
| Cases | 111 (11.5) | 851 (88.5) | 0.88 (0.71–1.10) |
aORs based on conditional logistic regression among cases and their matched controls, adjusted for oral contraceptive and hormone therapy use.
Tumor characteristics associated with statin use
| Tumor characteristics . | Statin use ≥2 y . | Statin use “Never” . | OR (95% CI)a . |
|---|---|---|---|
| ER status | |||
| ER+ | 1,588 (10.1) | 14,134 (89.9) | Reference |
| ER− | 262 (7.3) | 3,330 (92.7) | 0.90 (0.78–1.04) |
| PR status | |||
| PR+ | 1,276 (10.0) | 11,496 (90.0) | Reference |
| PR− | 564 (8.7) | 5,910 (91.3) | 0.92 (0.82–1.03) |
| HER2 statusb | |||
| HER2− | 1,042 (15.7) | 5,579 (84.3) | Reference |
| HER2+ | 247 (12.8) | 1,684 (87.2) | 0.94 (0.80–1.10) |
| Intrinsic subtypeb | |||
| Luminal A | 931 (16.4) | 4,728 (83.5) | Reference |
| Luminal B | 176 (12.9) | 1,185 (87.1) | 0.96 (0.88–1.05) |
| HER2+/ER− | 70 (12.3) | 498 (87.7) | 0.95 (0.83–1.10) |
| Triple-negative | 111 (11.5) | 851 (88.5) | 0.92 (0.82–1.03) |
| Tumor stage | |||
| Localized | 1,298 (10.2) | 11,377 (89.8) | Reference |
| Regional | 483 (8.2) | 5,410 (91.8) | 0.99 (0.94–1.05) |
| Metastasis | 59 (9.7) | 550 (90.3) | 1.01(0.87–1.17) |
| Tumor grade | |||
| Grade I | 452 (11.0) | 3,652 (89.0) | Reference |
| Grade II | 761 (10.2) | 6,711 (89.8) | 1.02 (0.95–1.09) |
| Grade III | 392 (7.7) | 4,686 (92.3) | 0.99 (0.92–1.06) |
| Tumor characteristics . | Statin use ≥2 y . | Statin use “Never” . | OR (95% CI)a . |
|---|---|---|---|
| ER status | |||
| ER+ | 1,588 (10.1) | 14,134 (89.9) | Reference |
| ER− | 262 (7.3) | 3,330 (92.7) | 0.90 (0.78–1.04) |
| PR status | |||
| PR+ | 1,276 (10.0) | 11,496 (90.0) | Reference |
| PR− | 564 (8.7) | 5,910 (91.3) | 0.92 (0.82–1.03) |
| HER2 statusb | |||
| HER2− | 1,042 (15.7) | 5,579 (84.3) | Reference |
| HER2+ | 247 (12.8) | 1,684 (87.2) | 0.94 (0.80–1.10) |
| Intrinsic subtypeb | |||
| Luminal A | 931 (16.4) | 4,728 (83.5) | Reference |
| Luminal B | 176 (12.9) | 1,185 (87.1) | 0.96 (0.88–1.05) |
| HER2+/ER− | 70 (12.3) | 498 (87.7) | 0.95 (0.83–1.10) |
| Triple-negative | 111 (11.5) | 851 (88.5) | 0.92 (0.82–1.03) |
| Tumor stage | |||
| Localized | 1,298 (10.2) | 11,377 (89.8) | Reference |
| Regional | 483 (8.2) | 5,410 (91.8) | 0.99 (0.94–1.05) |
| Metastasis | 59 (9.7) | 550 (90.3) | 1.01(0.87–1.17) |
| Tumor grade | |||
| Grade I | 452 (11.0) | 3,652 (89.0) | Reference |
| Grade II | 761 (10.2) | 6,711 (89.8) | 1.02 (0.95–1.09) |
| Grade III | 392 (7.7) | 4,686 (92.3) | 0.99 (0.92–1.06) |
aOR adjusted for age, prescription drug coverage, oral contraceptive use, hormone therapy use, and race.
bHER2 status was only evaluated in tumors 2002 to 2007.
In revised Table 1, there were fewer users of oral contraceptives and menopausal hormone therapy. There were also fewer ever users of lipophilic statins, with very little change in the number of women with 2 or more years of use. Mean duration of statin use was longer among ever users, with little change among women with 2 or more years of use.
In revised Table 2, the odds ratios (ORs) for hormone receptor (HR)-negative breast cancer remained nonsignificant for all categories of statin use, although the inverse association became stronger among women with 5 or more years of use. There was almost no change in the results for HR-positive breast cancers.
In revised Table 3, the ORs for risk of intrinsic breast cancer subtypes and statin use changed only slightly and remained nonsignificant for all 4 subtypes.
In revised Table 4, the case-only ORs for the association of various tumor characteristics with statin use remained nonsignificant for all examined parameters. The borderline significant association of progesterone receptor status and statin use became weaker and nonsignificant.
The overall conclusion remained unchanged: these findings do not confirm a preventive association between lipophilic statin use and breast cancer.
The authors sincerely regret these data errors and thank Joan Schwalbe for preparation of the corrected tables.
