statin use after breast cancer dec mortality

 a recent article found that in patients with early breast cancer, statin use after the diagnosis of breast cancer was associated with lower all-cause and breast cancer-specific mortality particularly among patients with hormone receptor-positive subtypes: see breast cancer statin use helps JAMAonc2026 in dropbox, or doi:10.1001/jamanetworkopen.2026.16375

Details:

-- 7389 female patients with early invasive breast cancer (BC) in Finland were included in this retrospective population-based cohort study,  based on the Finnish national registries from 1995 and 2013

    -- 2313 who were on statins after their BC diagnosis and outcomes were compared to 5076 who were not on statins

-- patient characteristics, comparing the groups with breast cancer diagnosis who were on statins after BC diagnosis vs those women having no statins after BC diagnosis:

    -- median age: 64 vs 58 yo

    -- median follow-up time: 7.4 vs 6.0 years

    -- deaths: total deaths 407 (17.6%) vs 1266 (24.9%), and BC-related deaths 128 (5.5%) vs 576 (11.3%)

    -- subtypes off breast cancer:

        -- luminal A-like: 702 (30.3%) vs 1422 (28.0%)

        -- luminal B-like (HER2-negative): 1170 (50.6%) vs 2499 (49.2%)

        -- luminal B-like (HER2-positive): 141 (6.1%) vs 422 (8.3%)

        -- HER2-poitive (non-luminal): 109 (4.7%) vs 265 (5.2%)

        -- Triple-negative: 191 (8.3%) vs 468 (9.2%)

    -- no statin use before BC diagnosis: 1254 (54.2%) had vs 4881 (96.2%) did not have statins after BC diagnosis;  individuals who had statin use before BC diagnosis: 1059(45.8%) had vs 195 (3.8%) did not have statins after BC diagnosis

    -- BC histology:

        -- ductal: 1745 (75.8%) vs 3975 (78.3%)

        -- lobular: 442 (19.1%) vs 826 (16.3%)

        -- other: 117 (5.1%) vs 273 (5.4%)

   -- tumor extent at diagnosis:

        -- localized: 1453 (62.8%) vs 3130 (61.7%)

        -- locally advanced:  860 (37.2%) vs 1946 (38.3%)

    -- any mammography screening before BC diagnosis: 1692 (73.2%) vs 3226 (63.6%)

    -- primary treatment:

        -- surgery with curative intent: 1900 (82.1%) vs 4120 (81.2%)

        -- other or unknown: 413 (17.9%) vs 956 (18.8%)

    -- comorbidities, per the Charlson Comorbidity Index (CCI):

        -- median 0 (0-2) vs 0 (0-2)

        -- mean: 0.8 (0-13) vs 0.8 (0-13)

    -- diabetes before BC diagnosis: 330 (14.3%) vs 205 (4.0%)

    -- hypertension before BC diagnosis: 1443 (62.4%) vs 1886 (37.2%)

-- statin use in Finland is only by prescription and is partially reimbursed by the Social Insurance institution

    -- information was available by package size and drug strength at every purchase, and the total annual use of each statin in milligrams for each patient was calculated and the defined daily dose (DDD) was determined, as by per the World Health Organization.  data was from 1995-2015

    -- patients were considered to be statin users prior to BC diagnosis if they had any purchases of statins between 1995 and the year of their BC diagnosis.

        -- and all patients were considered nonusers until the year of their first statin purchase. this included every year with a purchase of statins, irrespective of the amount, and this was recorded as a year with statin use

        -- after the first statin purchase, patients were not allowed to transfer back to being nonusers, in order to limit the bias regarding discontinuation of the statin use if in palliative care. 

-- results were analyzed with age and multivariable adjustment (adjusting factors: age at BC diagnosis, number of mammography screening rounds attended before BC diagnosis, diabetes [all types] before BC diagnosis, primary treatment for BC [surgery, or other or unknown]), anticoagulative and antihypertensive drug use after BC diagnosis, and CCI (Charlson Comorbidity Index).

    -- In the first sensitivity analysis, they adjusted their multivariable-adjusted model also for blood cholesterol level after BC diagnosis

    -- in the second sensitivity analysis, they performed a standard multivariable adjusted Cox regression model without CCI

    -- in the third sensitivity analysis, they excluded all patients who were on statins prior to the BC diagnosis in order to see if this would alter the results of the post-BC diagnoses

--MAIN OUTCOMES AND MEASURES: all-cause and breast cancer–specific mortality during follow-up

Results:

-- see above patient characteristics for the number of deaths, subtypes of breast cancer, histology, tumor extent, primary treatment and assessment of comorbidities, comparing those on statin use after BC diagnosis vs those without statin use after BC diagnosis

-- associations between statin use before BC diagnosis and all-cause and breast cancer-specific mortality:

    -- All patients: no statistical difference if on or off statins

    -- Luminal A-like: no statistical difference if on or off statins

    -- Luminal B-like (HER2-negative): no statistical difference if on or off statins

    -- Luminal B-like (HER2-positive): only statistical difference if on or off statins, 67% decrease, aHR 0.33 (0.12-0.87); this difference was only significant for BC-specific and not for all-cause mortality

    -- Luminal B-like (HER2-positive (non-luminal)): none statistically significant

    -- Triple-negative): none statistical difference if on or off statins

-- multivariable-adjusted associations between statin use after BC diagnosis and all-cause mortality that are statistically significant, comparing those who did vs didn't take statins (the latter being the reference):

    -- All patients: 32% decrease, aHR 0.68 (0.61-0.76)

    -- statin dose:

        -- first tertile (lowest consumption): 31% decrease, aHR 0.69 (0.58-0.82)

        -- second tertile (middle consumption): 38% decrease, aHR 0.62 (0.52-0.75)

        -- third tertile (highest consumption): 35% decrease, aHR 0.65 (0.53-0.79)

    -- Luminal A-like: 41% decrease, aHR 0.59 (0.47-0.74)

        -- first tertile: 35% decrease, aHR 0.65 (0.47-0.92)

        -- second tertile: 54% decrease, aHR 0.46 (0.32-0.68)

        -- third tertile: 44% decrease, aHR 0.56 (0.36-0.85)

    -- Luminal B-like (HER2-negative): 28% decrease, aHR 0.72 (0.61-0.83)

        -- first tertile: 33% decrease, aHR 0.67 (0.53-0.85)

        -- second tertile: 36% decrease, aHR 0.64 (0.50-0.83)

        -- third tertile: 27% decrease, aHR 0.73 (0.56-0.95)

    -- Luminal B-like (HER2-positive): 49% decrease, aHR 0.51 (0.32-0.79)

        -- first tertile: 33% decrease, not significant

        -- second tertile: 57% decrease, aHR 0.43 (0.20-0.92)

        -- third tertile: 66% decrease, aHR 0.34 (0.15-0.75)

    -- Luminal B-like (HER2-positive (non-luminal)): none statistically significant

    -- Triple-negative): 30% decrease, aHR 0.70 (0.49-0.99)

        -- none significant by statin dose

-- multivariable-adjusted associations between statin use status after BC diagnosis and BC-specific mortality that are statistically significant, comparing those who did vs didn't take statins (the latter being the reference):

    -- All patients: 41% decrease, aHR 0.59 (0.48-0.71)

    -- statin dose:

        -- first tertile (lowest consumption): 27% decrease, aHR 0.73 (0.55-0.96)

        -- second tertile (middle consumption): 48% decrease, aHR 0.52 (0.38-0.73)

        -- third tertile (highest consumption): 55% decrease, aHR 0.45 (0.30-0.66)

    -- Luminal A-like: 48% decrease, aHR 0.52 (0.33-0.83)

        -- first tertile: 35% decrease, not significant

        -- second tertile: 65% decrease: aHR 0.35 (0.15-0.81)

        -- third tertile: 66% decrease, aHR 0.34 (0.12-0.94)

    -- Luminal B-like (HER2-negative): 38% decrease, aHR 0.62 (0.47-0.80)

        -- first tertile: not significant

        -- second tertile: 38% decrease, aHR 0.62 (0.41-0.96)

        -- third tertile: 64% decrease, aHR 0.36 (0.20-0.65)

    -- Luminal B-like (HER2-positive): 71% decrease, aHR 0.29 (0.14-0.62)

        -- first tertile: not significant

        -- second tertile: 76% decrease, aHR 0.24 (0.06-0.99)

        -- third tertile: not significant

    -- Luminal B-like (HER2-positive (non-luminal): none statistically significant

    -- Triple-negative: none statistically significant

-- the 1-year and 3-year lag-times were checked to make sure that the statin use was really associated with the BC results and not coincidental: these lag-time results were basically the same as in the above findings

sensitivity analyses:

-- the first sensitivity analysis: for the 3244 patients with at least 1 measurement of blood cholesterol level after BC diagnosis, statin use was associated with 33% higher survival in BC-specific mortality and a 34% higher survival in all-cause mortality (though, of note, there was no information on the actual lipid levels in both groups; see below)

    -- in this analysis, statin use also lowered BC-specific mortality in the luminal B-like (HER2-negative) subtype while all-cause mortality was lowered in luminal A-like and luminal B-like (HER2-negative) subtypes

-- the second sensitivity analysis evaluated the whole population for post-diagnostic statin use when the Charlson Comorbidity Index was not used in the multivariable-adjusted analyses. this did not affect the results

    -- though the CCI was very low in this study with median of 0 and mean of 0.68: a score of 0 means no formal comorbidity, and 1-2 points (Mild) means low risk of short-term mortality (a score of 5 means high risk of severe outcome)

--the third sensitivity analysis excluded all patients who were on statins prior to the BC diagnosis to see if this would alter the results of the post-BC diagnoses. The exclusion of pre-BC diagnostic statin users did not affect the results

Commentary:

-- Breast cancer (BC) was the most frequently diagnosed cancer and the leading cause of cancer-related deaths in female individuals globally in 2022, with nearly 2,300,000 new cases and 660,000 deaths

    --In Finland in 2022, the site of this study, BC was the most common new cancer and the second leading cause of cancer-related deaths in female individuals

-- as is well-known and reinforced by the European Society for Medical Oncology (ESMO) clinical practice guidelines and those of others, tumors are routinely managed in clinical practice according to their intrinsic subtypes, which provide both prognostic and predictive information

    -- subtyping is still mostly performed using surrogate definitions based on immunohistochemical or in situ hybridization methods used to evaluate hormone receptors (estrogen and progesterone receptors [ER and PR]), HER2, and proliferation factor Ki-67 statuses to categorize tumors into the 5 different intrinsic subtypes: luminal A-like, luminal B-like (HER2-negative), luminal B-like (HER2-positive), HER2-positive (nonluminal), and triple-negative

    -- the moniker "luminal" refers to cancers where the cells resemble the luminal (inner) cells that line the milk ducts of the breast; luminal B cancers are less common than luminal A ones but are typically more aggressive

-- the medical literature is mixed on the benefit of statins in cancer prevention, with a couple of negative observational trials, but:

    -- a detailed review of 17 observational studies found that statin use was consistently associated with reduced hepatocellular carcinoma and hepatic decompensation in patient with chronic liver disease: https://pubmed.ncbi.nlm.nih.gov/42285810/

    -- the anti-inflammatory effects of pitavastatin may prevent pancreatic cancer by decreasing pancreatic inflammation: https://gmodestmedblogs.blogspot.com/2024/06/pitavastatin-decreases-pancreatic.html

-- meds that decrease systemic inflammation (eg colchicine) may well have a role in preventing diseases associated with inflammation (eg MI’s), but ??cancers which are also associated with chronic inflammation): https://gmodestmedblogs.blogspot.com/2024/03/colchicine-decreases-cardiovasc-events.html

-- this large, retrospective cohort study found that statin usage after a BC diagnosis was associated with better survival, in particular a lower mortality in hormone receptor-positive subtypes (luminal A-like, luminal B-like [HER2-negative] and luminal B-like [HER2-positive]), but not in hormone receptor negative subtypes (HER2-positive [nonluminal] and triple-negative).

-- the authors do comment that "postdiagnostic statin use was associated with lower BC-specific mortality even after adjustment for blood cholesterol levels". However, even after looking at the Supplemental material as well as the article itself, there were no data on actual lipid levels in either the individuals on statins

-- though this study found that the benefit of post-diagnosis statin use was highest in those with luminal B-like (HER2 positive) subtypes, this involved a small cohort (41 patients, but a 76% decrease); however this result is consistent with statins enhancing the effects of HER2-targented adjuvant therapy

-- they also found that the higher the dose of statins taken was associated with better BC-specific survival, though statin dosing was largely not significantly association with all-cause survival (one would expect that the cardiovascular protection from statins would help decrease all-cause mortality)

-- Potential mechanisms by which statins could have anti-cancer effects, per the authors:

    -- statins affect the biosynthesis of cholesterol in hepatocytes, but also in cancer cells; lowering cholesterol biosynthesis might be associated with antitumoral effects

    -- cholesterol metabolites, particularly 27-hydroxycholesterol, have been shown to stimulate the growth of ER-positive BC cells in animal studies; lowering blood cholesterol levels with statins could be particularly beneficial for patients with ER-positive breast cancer

    -- by blocking the mevalonate-pathway, statins affect the synthesis of isoprenoid-proteins that impact tumor growth and development

    -- statins could also reduce the risk of adjuvant treatment-related cardiotoxic effects of trastuzumab (used in HER2 positive breast cancers), and especially anthracyclines (eg doxorubicin), which might increase all-cause survival in these patients

    -- statins also might strengthen the effect of aromatase inhibitors (AI, used in breast cancer treatment) and was associated with lower risk of BC recurrence in a Danish study among postmenopausal patients with early BC receiving AI adjuvant therapy

   -- and, I would add, statins are anti-inflammatory, and chronic inflammation can lead to DNA damage and cancer development and growth; for example, there are studies suggesting that “inflammation can influence breast cancer progression, metastasis, and therapeutic outcome by establishing a tumor supportive immune microenvironment”: https://pmc.ncbi.nlm.nih.gov/articles/PMC6861749/

    -- also, it has been known for decades that statins have "pleiotropic" effects (ie, effects beyond the beneficial lipid ones), such as being anti-inflammatory, anti-oxidation, reducing matrix metalloproteinase production, effects on T-cell differentiation, blocking BC apoptosis by upregulated tumor necrosis-factor-related apoptosis-inducing ligand receptors on T-lymphocytes: https://www.ahajournals.org/doi/10.1161/circresaha.116.308537 . in addition, in the oncology literature, statins have been found to be antiproliferative, pro-apoptotic, anti-invasive,  and radiosensitizing. statin studies have found that around 50% of the statin benefit in cardiovascular outcomes is through the array of pleiotropic effects

Limitations:

-- we do not have any information on the lipid levels on the individuals in the study, either in the statin or non-statin use categories. this is important because hyperlipidemia itself has been found to be associated with breast cancer: https://pmc.ncbi.nlm.nih.gov/articles/PMC5857015/pdf/PJMS-34-209.pdf

    -- the authors do state that they "adjusted the multivariable-adjusted model also for blood cholesterol level after BC diagnosis", but there was no information about this, including in the article's Supplement

-- this was a pretty healthy group of women, with very low Charlson Comorbidity Indices. That might well limit the generalizability to a sicker group of women

-- this was a very long-term retrospective study (1995-2013). there was no information of potential confounders such as obesity, smoking, physical activity, specifics of breast cancer stage, tumor grade, use of hormone replacement therapy, medications taken that might have effects on the breast cancer or associated mortality. These include important risk factors for cancer and may well have changed over this long-term study and affected the results. Also, in the few comorbidities assessed, it is unclear what the level of control was achieved for the diabetes or hypertension or the medications taken for them

-- there could also be very different breast cancer associated exposures in Finland than in the US or other countries, such as endocrine disrupting chemicals (eg bisphenol A (BPA), or phthalates or parabens that have estrogenic effects, or PFAS (the "forever chemicals") that are associated with cancer, or air pollution, or organic pesticides, or alcohol consumption, or radiation exposure or lack of physical activity or differences in diet.... These types of differences could affect the risk of breast cancer, the BC-specific mortality and morbidity, and the all-cause mortality, making the study results less generalizable to other countries with perhaps very different exposures/risk factors

so, 

-- an interesting and provocative though retrospective study finding that statins taken after the diagnosis of early breast cancer improved breast cancer related mortality

-- as noted above, there is documentation of many statin physiologic effects that are consistent with decreasing cancer morbidity/mortality, and these would support the beneficial use of post-cancer statins

    -- as with GLP-1 receptor agonists, it seems that some commonly used medications (eg statins) may have more and more non-target benefits (eg breast cancer benefit for statins)...

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

to get access to the dropbox: https://www.dropbox.com/scl/fo/pcf4vymtpo77f49wd2ljr/ANyJqB40CuBd-DMEr67XDoI?rlkey=5dsalmuhf8ekd0svsgvn93xm1&dl=0

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them