Abstract
Background:
While HPV vaccination and Pap screening have advanced cervical cancer (CCa) prevention, high-grade squamous intraepithelial lesions (HSIL) remain common, particularly among individuals with metabolic comorbidities like diabetes and hypercholesterolemia. Statins, commonly used for lipid control, possess anti-inflammatory and antiproliferative properties that may offer protective effects against cervical dysplasia. We explored the association between statin use and lesion grade in a population of dysplasia patients, and whether effects vary by comorbidity and race.
Methods:
Cross-sectional, observational retrospective analysis of electronic health records and billing data for 2,378 non-Latina/e (nL) Black and nL white patients diagnosed with LSIL or HSIL between 2014 – 2021 at a large academic medical center. Logistic regression assessed associations between statin use, comorbidity profiles (diabetes, hypercholesterolemia), race, and HSIL vs. LSIL, adjusting for potential confounders. Interaction terms were tested to evaluate effect modification.
Results:
Statin users had significantly lower odds of HSIL than nonusers (adjusted OR=0.48, p<0.0001), despite being older, with higher comorbidity rates. Predicted HSIL probabilities ranged from 4% - 20% in statin users versus 13% to 29% in nonusers. The lowest risk was observed among diabetic patients on statins, particularly among nL Black patients, suggesting a possible synergistic protective effect in metabolically vulnerable populations. Only 35% of patients with a hypercholesterolemia diagnosis listed were on statins.
Conclusions:
Statin use was associated with substantially lower HSIL risk, particularly among nL Black patients with diabetes.
Impact:
These findings support further investigation of statins as a potential low-cost chemopreventive tool for cervical dysplasia, especially in populations with metabolic dysfunction.
Keywords: Cervical dysplasia, Statin Therapy, Metabolic comorbidities, Health disparities, Diabetes, Hypercholesterolemia, Chemoprevention
INTRODUCTION
Cervical cancer (CCa) incidence and mortality in the United States have declined, largely due to widespread human papillomavirus (HPV) vaccination and increased Papanicolaou (Pap) screening uptake.1,2 Despite these gains, precancerous cervical lesions remain common, with high-grade squamous intraepithelial lesions (HSIL) occurring at approximately 1.4 cases per 1000 person-years.3 HSIL is strongly linked to persistent high-risk HPV infections (e.g., HPV 16 and 18) and has a lower likelihood of spontaneous regression compared with low-grade squamous intraepithelial lesions (LSIL).4–6 Current clinical guidelines recommend aggressive management of HSIL, often involving colposcopic evaluation and excisional procedures.7 While effective, these interventions carry risks of second-trimester miscarriage, preterm birth, and recurrent HSIL, posing substantial physical, emotional, and financial burdens.8–10 Consequently, identifying effective preventive strategies, developing alternative interventions, and elucidating HSIL etiology are therefore critical public health imperatives.1
Emerging evidence suggests that cardiovascular disease (CVD) risk factors, including metabolic syndrome and chronic inflammation, may contribute to the pathogenesis of cervical dysplasia and CCa.11,12 Metabolic syndrome—characterized by central obesity, hypertension, hypertriglyceridemia, and insulin resistance—has been linked to heightened carcinogenesis in diverse malignancies, including prostate, colon, and pancreatic cancers.13–15 Dysregulated lipid metabolism also appears relevant, with elevated serum cholesterol associated with adverse outcomes in CCa.16 Chronic low-level inflammation and oxidative stress, both key features of metabolic syndrome, are implicated in various carcinogenic processes, potentially exacerbated by higher levels of free estrogen in hypercholesterolemic states.14,17,18
Statins — which are among the most frequently prescribed lipid-lowering agents — have drawn interest for their potential anti-cancer effects, which are mediated by antiproliferative, immunomodulatory, and pro-apoptotic properties.19–21 In both preclinical research and some clinical studies, statins have been associated with reduced cancer incidence and improved outcomes in breast, colon, and other malignancies.20–22 However, these findings are inconclusive; multiple meta-analyses and systematic reviews have found no consistent association between statin use and cancer risk or survival.19,23 Despite the mixed evidence, further inquiry is warranted — particularly in preinvasive conditions such as HSIL, where statin therapy could be leveraged as a low-cost chemopreventive strategy.
Racial disparities further underscore the need for novel preventive approaches. Non-Latinx (nL) Black individuals in the United States have higher CCa incidence (8.4 vs. 7.0 per 100,000) and mortality (3.2 vs. 2.1 per 100,000) than their nL-white (nLW) counterparts according to the SEER*Stat database [RRID:SCR_003293]. Disproportionate rates of persistent HPV infection and HSIL in nL-Black (nLB) individuals often translate into more frequent invasive procedures, which may contribute to the poorer reproductive outcomes experienced by this population.13 Although HPV vaccination may mitigate differences in infection rates by vaccine-preventable genotypes, knowledge gaps persist regarding non-vaccine-preventable HPV strains that disproportionately affect nLB individuals.4
Meanwhile, the burden of CVD — and its associated risk factors — remains strikingly high in nLB populations, despite a lower reported prevalence of diagnosed hypercholesterolemia (21.5% vs. 26.9%).24–27 Underdiagnosis and undertreatment likely contribute to persistently higher CVD mortality. Moreover, statin prescribing is nearly 50% lower in nLB populations as compared to nLWs.28–30 Women are also less likely than men to be prescribed statins.30 This in turn raises the possibility that discrepancies in CVD and statin may also influence HSIL progression through metabolic and inflammatory mechanisms, presenting opportunities to both improve cardiovascular outcomes and reduce HSIL risk.
If statins demonstrate even partial chemopreventive effects for HSIL, they could represent a noninvasive strategy with significant implications for both patients at greatest risk and overburdened healthcare systems. Accordingly, we conducted a cross-sectional study to explore the association between statin use and lesion grade among a diverse cohort of cervical dysplasia patients. Though observational data cannot definitively elucidate biologic mechanisms, detecting associations that point to protective effects can generate new hypotheses regarding metabolic or inflammatory pathways in cervical dysplasia, and inform potential strategies for CCa interception.
MATERIALS AND METHODS
Study Design:
We conducted a cross-sectional study using electronic medical records (EMRs) and billing data (Cerner Millennium, RRID:SCR_013581) from the Virginia Commonwealth University Health System (VCUHS) to investigate whether statin use is associated with a lower likelihood of HSIL at diagnosis. We further assessed whether race (nLB vs nLW) ) modifies this potential association. As this was an observational study using existing data, no randomization or blinding was performed. The study protocol was reviewed and approved by the VCU Institutional Review Board (IRB #HM20021592), with a waiver of informed consent due to minimal risk and the retrospective nature of data collection.
Data Source:
Hospital and physician billing claims for LSIL or HSIL diagnosed between January 1, 2014, and December 4, 2021, were identified at VCUHS, a large academic center serving a racially diverse population (47% nLW-, 32% nLB, 15% other/unknown, 6% Latine/x). To ensure accurate classification, we cross-referenced claims with pathology reports from the Department of Pathology, thereby enhancing case ascertainment and mitigating potential misclassification.
Case Selection and Inclusion/Exclusion Criteria:
Eligible patients included those diagnosed with LSIL or HSIL at VCUHS or those diagnosed externally (e.g. at smaller community clinics) but treated at VCUHS. We excluded individuals of racial identities other than nLB or nLW,, and those identifying as Latine/x (regardless of race), in order to best explore differential exposures and outcomes between nLB and nLW patients,. Cervical lesions were classified via the Bethesda System31 with LSIL representing CIN1, and HSIL representing CIN2, CIN3, or carcinoma in situ. Lesion diagnoses were initially identified by International Classification of Diseases (ICD-9/10) codes (International Classification of Diseases Version 10 - Procedure Coding System, RRID:SCR_010351) and confirmed via pathology records.
Exposure of Interest and Covariates:
The primary exposures were statin use(defined as having an active prescription at the time of lesion diagnosis or treatment, as extracted from the clinical record), and race (nLB vs. nLW). Additional covariates extracted from EMRs (Cerner Millennium, RRID:SCR_013581) included age, marital status, insurance status, hypercholesterolemia, diabetes, HIV status, smoking history, excessive alcohol use, bacterial vaginosis, pulmonary and heart conditions, and HPV vaccination status. Covariates were coded from ICD-9/10 (International Classification of Diseases Version 10 - Procedure Coding System, RRID:SCR_010351) encounters closest to the date of cervical dysplasia diagnosis. Notably, individuals were categorized as having the above listed comorbidities if it was listed in the captured diagnoses or problem list entries at the time of cervical lesion diagnosis or treatment; we did not retrieve historical lipid values, prior diagnoses or past statin use.
Statistical Analysis:
We first conducted bivariate analyses (chi-square tests for categorical variables; t-tests or Wilcoxon rank-sum for continuous variables) to identify covariates significantly associated (p<0.05) with both HSIL and at least one key exposure (current statin use or race). Our final logistic regression models considered variables meeting these criteria as potential confounders. In building the model, we began with a crude logistic regression estimating the odds of HSIL among current statin users relative to non-users. Covariates were systematically added using a forward selection approach, retaining those that met a 10% change-in-estimate threshold or that were deemed clinically relevant. Likelihood ratio and Hosmer-Lemeshow tests were used to evaluate model fit and calibration, respectively.
Effect modification was tested through interaction terms involving race, statin use, and relevant covariates. Significant interactions at p<0.05 prompted stratified analyses. To facilitate clinical interpretation, we computed adjusted predicted probabilities of HSIL across relevant subgroups (e.g., diabetic vs. non-diabetic, by race) using marginal estimates from STATA 18 (Stata, RRID:SCR_012763).
Data Availability.
The de-identified data generated in this study are available from the corresponding author upon reasonable request.
RESULTS
Overall Sample Characteristics:
From January 1, 2014, through December 4, 2021, we identified 2,761 cases of squamous intraepithelial lesions (SIL). After excluding 383 patients of other races/ethnicities or missing race data, the analytic sample totaled 2,378 cervical dysplasia patients: 53% nLB (n=1,264) and 47% nLw (n=1,114). LSIL accounted for 75% (n=1,793) of cases, while HSIL represented 25% (n=585). Statin prescriptions were recorded in 6.8% (n=162) of the sample, of whom only 40% had a formal hypercholesterolemia diagnosis; conversely, 36% of hypercholesterolemia-diagnosed patients (n=188) were on statins.
Differences by Race (Table 1):
Table 1.
Demographic and Clinical Characteristics By Participant Race
| Total (N=2378) |
nL-Black (n=1264) |
nL-White (n=1114) |
||
|---|---|---|---|---|
| n (%) | n (%) | n (%) | p-value | |
|
| ||||
| Age: Mean (SD) | 38.7 (13.7) | 38.2 (13.5) | 39.2 (13.8) | 0.0617 |
|
| ||||
| Lesion Type | 0.0004 | |||
| HSIL | 585 (24.6) | 274 (21.7) | 311 (27.9) | |
| LSIL | 1793 (75.4) | 990 (78.3) | 803 (72.1) | |
|
| ||||
| High Cholesterol at Diagnosis | 0.0677 | |||
| Yes | 181 (7.6) | 108 (8.5) | 73 (6.6) | |
| No | 2197 (92.4) | 1156 (91.5) | 1041 (93.4) | |
|
| ||||
| Statin Use | 0.1067 | |||
| Yes | 162 (6.8) | 96 (7.6) | 66 (5.9) | |
| No | 2216 (93.2) | 1168 (92.4) | 1048 (94.1) | |
|
| ||||
| BMI | <.0001 | |||
| <30 | 823 (34.6) | 326 (25.8) | 497 (44.6) | |
| ≥30 | 896 (37.7) | 608 (48.1) | 288 (25.9) | |
| Unknown | 659 (27.7) | 330 (26.1) | 329 (29.5) | |
|
| ||||
| Marital Status | <.0001 | |||
| Single | 1383 (58.2) | 889 (70.3) | 494 (44.3) | |
| Separated/Divorced | 339 (14.3) | 144 (11.4) | 195 (17.5) | |
| Married/Life Partner | 575 (24.2) | 188 (14.9) | 387 (34.7) | |
| Widow | 67 (2.8) | 38 (3.0) | 29 (2.6) | |
| Missing | 14 (0.6) | 5 (0.4) | 9 (0.8) | |
|
| ||||
| Insurance Status | <.0001 | |||
| Private | 1080 (45.4) | 418 (33.1) | 662 (59.4) | |
| Public | 530 (22.3) | 344 (27.2) | 186 (16.7) | |
| Uninsured/Other | 768 (32.3) | 502 (39.7) | 266 (23.9) | |
|
| ||||
| Smoking | 0.0109 | |||
| Never | 1764 (74.2) | 921 (72.9) | 843 (75.7) | |
| Current | 550 (23.1) | 303 (24.0) | 247 (22.2) | |
| Other | 16 (0.7) | 14 (1.1) | 2 (0.2) | |
| Missing | 48 (2.0) | 26 (2.1) | 22 (2.0) | |
|
| ||||
| Excessive Alcohol Use | <.0001 | |||
| Never | 1026 (43.1) | 617 (48.8) | 409 (36.7) | |
| Former | 587 (24.7) | 273 (21.6) | 314 (28.2) | |
| Current | 679 (28.6) | 332 (26.3) | 347 (31.1) | |
| Missing | 86 (3.6) | 42 (3.3) | 44 (3.9) | |
|
| ||||
| HPV Vaccination Status | <.0001 | |||
| Yes | 121 (5.1) | 88 (7.0) | 33 (3.0) | |
| No | 2257 (94.9) | 1176 (93.0) | 1081 (97.0) | |
|
| ||||
| HIV | <.0001 | |||
| Yes | 75 (3.2) | 65 (5.1) | 10 (0.9) | |
| No | 2303 (96.8) | 1199 (94.9) | 1104 (99.1) | |
|
| ||||
| Bacterial Vaginosis | <.0001 | |||
| Yes | 783 (32.9) | 577 (45.6) | 206 (18.5) | |
| No | 1595 (67.1) | 687 (54.4) | 908 (81.5) | |
|
| ||||
| Arthritis | 0.0249 | |||
| Yes | 76 (3.2) | 50 (4.0) | 26 (2.3) | |
| No | 2302 (96.8) | 1214 (96.0) | 1088 (97.7) | |
|
| ||||
| Heart Conditions | <.0001 | |||
| Yes | 204 (8.6) | 137 (10.8) | 67 (6.0) | |
| No | 2174 (91.4) | 1127 (89.2) | 1047 (94.0) | |
|
| ||||
| Pulmonary Conditions | <.0001 | |||
| Yes | 436 (18.3) | 299 (23.7) | 137 (12.3) | |
| No | 1942 (81.7) | 965 (76.3) | 977 (87.7) | |
|
| ||||
| Diabetes | <.0001 | |||
| Yes | 269 (11.3) | 190 (15.0) | 79 (7.1) | |
| No | 2109 (88.7) | 1074 (85.0) | 1035 (92.9) | |
HSIL was more common among nLW dysplasia patients (27.9%) than nLB patients (21.7%) (p=0.0004). Relative to nLW patients, nLB patients were more frequently single (70.3% vs 44.3%), had higher rates of public or no insurance (39.7% vs 23.9%), and had elevated prevalence of obesity, bacterial vaginosis, HIV, diabetes, heart disease, and pulmonary conditions (all p<0.0001). Notably, nLB patients reported a higher rate of current smoking (24.0% vs 22.2%) than nLW patients, with t lower rates of past smoking (21.6% vs 28.2%) and lower excessive alcohol use (26.3% vs 31.1%).
Differences by Lesion Severity (Table 2):
Table 2.
Demographic and Clinical Characteristics By HSIL/LSIL Status
| Total (N=2378) |
HSIL (n=585) |
LSIL (n=1793) |
||
|---|---|---|---|---|
| n (%) | n (%) | n (%) | p-value | |
|
| ||||
| Age: Mean (SD) | 38.7 (13.7) | 37.0 (13.2) | 39.2 (13.8) | 0.0007 |
|
| ||||
| Race | 0.0004 | |||
| nL-White | 1114 (46.8) | 311 (53.2) | 803 (44.8) | |
| nL-Black | 1264 (53.2) | 274 (46.8) | 990 (55.2) | |
|
| ||||
| High Cholesterol at Diagnosis | 0.0009 | |||
| Yes | 181 (7.6) | 26 (4.4) | 155 (8.6) | |
| No | 2197 (92.4) | 559 (95.6) | 1638 (91.4) | |
|
| ||||
| Statin Use | <.0001 | |||
| Yes | 162 (6.8) | 18 (3.1) | 144 (8.0) | |
| No | 2216 (93.2) | 567 (96.9) | 1649 (92.0) | |
|
| ||||
| BMI | <.0001 | |||
| <30 | 823 (34.6) | 215 (36.8) | 608 (33.9) | |
| ≥30 | 896 (37.7) | 160 (27.4) | 736 (41.0) | |
| Unknown | 659 (27.7) | 210 (35.9) | 449 (25.0) | |
|
| ||||
| Marital Status | 0.0503 | |||
| Single | 1383 (58.2) | 349 (59.7) | 1034 (57.7) | |
| Separated/Divorced | 339 (14.3) | 63 (10.8) | 276 (15.4) | |
| Married/Life Partner | 575 (24.2) | 151 (25.8) | 424 (23.6) | |
| Widow | 67 (2.8) | 17 (2.9) | 50 (2.8) | |
| Missing | 14 (0.6) | 5 (0.9) | 9 (0.5) | |
|
| ||||
| Insurance Status | 0.0615 | |||
| Private | 1080 (45.4) | 241 (41.2) | 839 (46.8) | |
| Public | 530 (22.3) | 141 (24.1) | 389 (21.7) | |
| Uninsured/Other | 768 (32.3) | 203 (34.7) | 565 (31.5) | |
|
| ||||
| Smoking | 0.0011 | |||
| Never | 1764 (74.2) | 397 (67.9) | 1367 (76.2) | |
| Current | 550 (23.1) | 166 (28.4) | 384 (21.4) | |
| Other | 16 (0.7) | 3 (0.5) | 13 (0.7) | |
| Missing | 48 (2.0) | 19 (3.2) | 29 (1.6) | |
|
| ||||
| Excessive Alcohol Use | 0.5847 | |||
| Never | 1026 (43.1) | 243 (41.5) | 783 (43.7) | |
| Former | 587 (24.7) | 140 (23.9) | 447 (24.9) | |
| Current | 679 (28.6) | 175 (29.9) | 504 (28.1) | |
| Missing | 86 (3.6) | 27 (4.6) | 59 (3.3) | |
|
| ||||
| HPV Vaccination Status | 0.1768 | |||
| Yes | 121 (5.1) | 36 (6.2) | 85 (4.7) | |
| No | 2257 (94.9) | 549 (93.8) | 1708 (95.3) | |
|
| ||||
| HIV | 0.9023 | |||
| Yes | 75 (3.2) | 18 (3.1) | 57 (3.2) | |
| No | 2303 (96.8) | 567 (96.9) | 1736 (96.8) | |
|
| ||||
| Bacterial Vaginosis | 0.0094 | |||
| Yes | 783 (32.9) | 167 (28.5) | 616 (34.4) | |
| No | 1595 (67.1) | 418 (71.5) | 1177 (65.6) | |
|
| ||||
| Arthritis | 0.0186 | |||
| Yes | 76 (3.2) | 10 (1.7) | 66 (3.7) | |
| No | 2302 (96.8) | 575 (98.3) | 1727 (96.3) | |
|
| ||||
| Heart Conditions | 0.0006 | |||
| Yes | 204 (8.6) | 30 (5.1) | 174 (9.7) | |
| No | 2174 (91.4) | 555 (94.9) | 1619 (90.3) | |
|
| ||||
| Pulmonary Conditions | 0.0019 | |||
| Yes | 436 (18.3) | 82 (14.0) | 354 (19.7) | |
| No | 1942 (81.7) | 503 (86.0) | 1439 (80.3) | |
|
| ||||
| Diabetes | <.0001 | |||
| Yes | 269 (11.3) | 32 (5.5) | 237 (13.2) | |
| No | 2109 (88.7) | 553 (94.5) | 1556 (86.8) | |
Compared with LSIL, HSIL was associated with younger age (mean 37.0 vs 39.2 years, p=0.0007), lower prevalence of hypercholesterolemia (4.4% vs 8.6%, p=0.0009), and significantly lower statin use (3.1% vs 8.0%, p<0.0001). Notably, diabetes (5.5% vs 13.2%, p<0.0001) and obesity (34.6% vs 45.0%, p<0.0001) were also more common among LSIL patients, whereas current smoking was more frequent in HSIL patients (28.4% vs 21.4%, p=0.0011).
Differences by Statin Use (Table 3):
Table 3.
Demographic and Clinical Characteristics By Statin Use at Time of SIL Diagnosis
| Total (N=2378) |
Yes (n=162) |
No (n=2216) |
||
|---|---|---|---|---|
| n (%) | n (%) | n (%) | p-value | |
|
| ||||
| Age: Mean (SD) | 38.7 (13.7) | 54.6 (12.2) | 37.5 (13.0) | <.0001 |
|
| ||||
| Race | 0.1067 | |||
| nL-White | 1114 (46.8) | 66 (40.7) | 1048 (47.3) | |
| nL-Black | 1264 (53.2) | 96 (59.3) | 1168 (52.7) | |
|
| ||||
| Lesion Type | <.0001 | |||
| HSIL | 585 (24.6) | 18 (11.1) | 567 (25.6) | |
| LSIL | 1793 (75.4) | 144 (88.9) | 1649 (74.4) | |
|
| ||||
| High Cholesterol at Diagnosis | <.0001 | |||
| Yes | 181 (7.6) | 65 (40.1) | 116 (5.2) | |
| No | 2197 (92.4) | 97 (59.9) | 2100 (94.8) | |
|
| ||||
| BMI | <.0001 | |||
| <30 | 823 (34.6) | 44 (27.2) | 779 (35.2) | |
| ≥30 | 896 (37.7) | 114 (70.4) | 782 (35.3) | |
| Unknown | 659 (27.7) | 4 (2.5) | 655 (29.6) | |
|
| ||||
| Marital Status | <.0001 | |||
| Single | 1383 (58.2) | 52 (32.1) | 1331 (60.1) | |
| Separated/Divorced | 339 (14.3) | 44 (27.2) | 295 (13.3) | |
| Married/Life Partner | 575 (24.2) | 45 (27.8) | 530 (23.9) | |
| Widow | 67 (2.8) | 21 (13.0) | 46 (2.1) | |
| Missing | 14 (0.6) | 14 (0.6) | ||
|
| ||||
| Insurance Status | <.0001 | |||
| Private | 1080 (45.4) | 56 (34.6) | 1024 (46.2) | |
| Public | 530 (22.3) | 29 (17.9) | 501 (22.6) | |
| Uninsured/Other | 768 (32.3) | 77 (47.5) | 691 (31.2) | |
|
| ||||
| Smoking | 0.4040 | |||
| Never | 1764 (74.2) | 115 (71.0) | 1649 (74.4) | |
| Current | 550 (23.1) | 45 (27.8) | 505 (22.8) | |
| Other | 16 (0.7) | 1 (0.6) | 15 (0.7) | |
| Missing | 48 (2.0) | 1 (0.6) | 47 (2.1) | |
|
| ||||
| Excessive Alcohol Use | 0.0043 | |||
| Never | 1026 (43.1) | 90 (55.6) | 936 (42.2) | |
| Former | 587 (24.7) | 27 (16.7) | 560 (25.3) | |
| Current | 679 (28.6) | 42 (25.9) | 637 (28.7) | |
| Missing | 86 (3.6) | 3 (1.9) | 83 (3.7) | |
|
| ||||
| HPV Vaccination Status | 0.0522 | |||
| Yes | 121 (5.1) | 3 (1.9) | 118 (5.3) | |
| No | 2257 (94.9) | 159 (98.1) | 2098 (94.7) | |
|
| ||||
| HIV | 0.0228 | |||
| Yes | 75 (3.2) | 10 (6.2) | 65 (2.9) | |
| No | 2303 (96.8) | 152 (93.8) | 2151 (97.1) | |
|
| ||||
| Bacterial Vaginosis | 0.1847 | |||
| Yes | 783 (32.9) | 61 (37.7) | 722 (32.6) | |
| No | 1595 (67.1) | 101 (62.3) | 1494 (67.4) | |
|
| ||||
| Arthritis | <.0001 | |||
| Yes | 76 (3.2) | 14 (8.6) | 62 (2.8) | |
| No | 2302 (96.8) | 148 (91.4) | 2154 (97.2) | |
|
| ||||
| Heart Conditions | <.0001 | |||
| Yes | 204 (8.6) | 57 (35.2) | 147 (6.6) | |
| No | 2174 (91.4) | 105 (64.8) | 2069 (93.4) | |
|
| ||||
| Pulmonary Conditions | <.0001 | |||
| Yes | 436 (18.3) | 52 (32.1) | 384 (17.3) | |
| No | 1942 (81.7) | 110 (67.9) | 1832 (82.7) | |
|
| ||||
| Diabetes | <.0001 | |||
| Yes | 269 (11.3) | 79 (48.8) | 190 (8.6) | |
| No | 2109 (88.7) | 83 (51.2) | 2026 (91.4) | |
Statin users (n=162) were older (mean age 54.6 vs 37.5 years; p<0.0001) and had a markedly lower HSIL prevalence than nonusers (11.1% vs 25.6%; p<0.0001). Comorbidities such as hypercholesterolemia (40.1% vs 5.2%), obesity (70.4% vs 35.3%), diabetes (48.8% vs 8.6%), and heart disease (35.2% vs 6.6%) were significantly more common among statin users than nonusers. Excessive alcohol use was more common among statin nonusers (p=0.0043). No significant racial differences in statin prescriptions were observed (p=0.1067).
Explored Confounders:
Based on the descriptive analyses, several variables were identified as potential confounders for the logistic regression model, as they were significantly associated with both HSIL (the outcome) and at least one of the key exposures of interest (race or statin use). These include age, marital insurance, HPV vaccination and smoking status, obesity, bacterial vaginosis, diabetes, heart conditions, and pulmonary conditions. These variables were thus considered potential confounders and systematically assessed during multivariable model development.
Logistic Regression (Table 4):
Table 4.
Multivariate Logistic Regression Model Building Evaluating the Association between Statin Use and HSIL (CIN2+)
| Model | Variables Included | N | OR (95% CI) | p-value (variable added) | % Change in OR | LR χ2 | LR Test p-value |
|---|---|---|---|---|---|---|---|
|
| |||||||
| 1 | Statin (Crude) | 2378 | 0.36 (0.22–0.60) | <0.0001 | — | 19.94 | N/A |
|
| |||||||
| 2 | + Age | 2378 | 0.42 (0.25–0.70) | 0.022 | 16% | 25.25 | — |
|
| |||||||
| 3 | + Race | 2378 | 0.43 (0.26–0.73) | <0.0001 | 3% (keep) | 37.60 | — |
|
| |||||||
| 4 | + BMI | 2378 | 0.48 (0.28–0.80) | <0.0001 | 10% | 52.44 | — |
| 9 | + Diabetes | 2378 | 0.57 (0.34–0.97) | 0.001 | 10% | 64.62 | — |
| 10 | – Age (final model) | 2378 | 0.52 (0.31–0.88) | — | 1% (drop) | 62.13 | 0.110 |
OR: Odds ratio; CI: Confidence interval; LR χ2: Likelihood ratio chi-square. Percent (%) change indicates the change in the odds ratio upon adding each subsequent variable. Variables that did not achieve a ≥10% change were removed to ensure parsimony. Those variables (all added to the model with age, race and bmi) were: marital, insurance, smoking and HPV-vaccination status, hypercholesterolemia, heart disease, and pulmonary disease) However, regardless of percent change, race was retained in all models, as it represented one of our primary exposures of interest in this analysis. Model 9 represents the most parsimonious final model (variables retained: statin, race, BMI, diabetes). Upon adding diabetes, age was no longer significant; therefore, we used the likelihood ratio test to confirm that the reduced model (excluding age) did not significantly differ from the full model.
In crude models, statin use was associated with a 64% reduction in HSIL odds (OR=0.36, 95% CI 0.22–0.60, p<0.0001). Sequentially adjusting for age, race, body mass index (BMI), and diabetes attenuated the association somewhat (final adjusted OR=0.48, p<0.0001). Of note, although adding race minimally changed the estimate (3% change), we retained it due to our primary interest in racial differences. Neither marital status nor cardiovascular/pulmonary conditions appreciably influenced the effect estimate or improved model fit. Thus, they were excluded for parsimony. No significant statin-race or statin-diabetes interaction was detected at p<0.05, but stratified analyses by race and diabetes, motivated by a priori interest, revealed stronger protective signals among certain subgroups.
Stratified Estimates:
Although interaction terms were not statistically significant, we stratified our estimates to detect any potential differences by subgroup that interaction terms might have missed. Table 5 depicts the estimates stratified by subgroups. Subgroup analyses showed the most robust reductions in HSIL odds among nLB patients with diabetes who were on statins (OR=0.10, 95% CI 0.02–0.40, p=0.001). A similarly protective effect was suggested in nLW diabetic patients using statins (OR=0.26, 95% CI 0.06–1.13, p=0.072), though statistical significance was borderline. Predicted probabilities further corroborated these trends, with the lowest HSIL risk among diabetic patients on statins (4%–10%, depending on race), in contrast to roughly 24%–29% among those not on statins.
Table 5.
Stratified Odds Ratios and Predicted Probabilities for HSIL (CIN2+) by Statin Use, Diabetes Status, and Race
| Statin Use | Diabetes Status | Race | Odds Ratio (95% CI) | OR p-value |
Predicted Probabilities (95%CI) | Predicted Probabilities p-value |
|---|---|---|---|---|---|---|
|
| ||||||
| No | Not Diabetic | White (Ref) | 1.00 (—) | 0.29 (0.27, 0.32) | <0.001 | |
| Black | 0.74 (0.60–0.90) | 0.002 | 0.24 (0.21, 0.26) | <0.001 | ||
| Diabetic | White | 0.35 (0.16–0.78) | 0.010 | 0.13 (0.04, 0.22) | 0.004 | |
| Black | 0.46 (0.28–0.75) | 0.002 | 0.16 (0.10, 0.23) | <0.001 | ||
| Yes | Not Diabetic | White | 0.59 (0.27–1.29) | 0.188 | 0.20 (0.08, 0.32) | 0.001 |
| Black | 0.47 (0.20–1.14) | 0.096 | 0.17 (0.05, 0.29) | 0.007 | ||
| Diabetic | White | 0.26 (0.06–1.13) | 0.072 | 0.10 (-0.03, 0.23) | 0.134 | |
| Black | 0.10 (0.02–0.40) | 0.001 | 0.04 (-0.01, 0.09) | 0.148 | ||
DISCUSSION
Our findings in this diverse cohort suggest a significant inverse association between statin use and HSIL, independent of key confounding variables such as BMI, diabetes status, and race. Although lower HSIL prevalence among statin users could theoretically reflect better healthcare engagement or an inherently different risk profile, the association remained robust after multivariable adjustment. Stratified analyses indicated the strongest protective trends among nLB dysplasia patients with diabetes. These findings, while exploratory, are biologically plausible given well-established links among metabolic dysfunction, systemic inflammation, and cervical carcinogenesis.
The anticancer properties of statins are supported by preclinical evidence demonstrating decreased oncogenic signaling (e.g., c-Myc inhibition), increased reactive oxygen species, immunomodulatory effects, and reduced cellular proliferation.18,32–34 Through HMG-CoA reductase inhibition, statins may alter lipid raft composition, disrupting membrane-bound growth factor signaling.35 Although direct causality cannot be inferred from these data, the observed association aligns with emerging mechanistic models suggesting that addressing metabolic and inflammatory pathways may be key for controlling early-stage cervical dysplasia.18,32,36
Our results highlight a clinically meaningful gap: only 36% of hypercholesterolemia-diagnosed patients in our sample were prescribed statins, echoing previous research documented underutilization of guideline-based lipid-lowering therapy, particularly among women and racial/ethnic minority patients.24,26,29,30,37 This shortfall poses a missed opportunity to address both CVD and, potentially, cervical dysplasia progression. Interestingly, while racial and ethnic disparities in statin prescribing are well documented in the literature, we did not observe any significant differences in statin prescription rates by race within our study.24,26,29,30,37 However, our sample, comprised of all VCUHS patients with a cervical dysplasia diagnosis was 53% nLB, a disproportionately large percentage given that only 32% of the overall VCUHS patient population is nLB, a finding that warrants further study. Taken together, our findings underscore the continuing need to improve care access and quality, particularly among populations at heightened risk for both metabolic disease and cervical dysplasia.
Clinical and Research Implications:
If corroborated, these findings suggest that improved screening and treatment of hypercholesterolemia could serve as an adjunct strategy for reducing cervical dysplasia progression. Given statins’ noninvasive nature and favorable safety profile, their chemopreventive potential warrants further validation in randomized controlled trials. Integrating molecular studies focusing on how statins modulate HPV-related pathways and cervical epithelial biology is equally vital. Furthermore, addressing structural barriers to statin therapy among women and nLB populations could provide dual benefits, enhancing both cardiovascular and CCa prevention efforts.
Strengths and Limitations:
The strengths of our study include the large, racially diverse patient population with pathology-confirmed cervical lesions, facilitating robust epidemiologic inference. However, certain limitations warrant caution. Unmeasured confounding, such as differences in HPV genotypes or sexual behavior, cannot be ruled out. Furthermore, our cross-sectional design included statin usage and comorbid diagnoses documented the time of lesion diagnosis, only, restricting inferences about temporality; prospective studies are needed to clarify causal direction.
Reliance on billing and EMR data may introduce misclassification of statin exposure, especially regarding duration and adherence. Statin use was defined based on prescription status at or near the time of HSIL diagnosis, which may not reflect exposure prior to lesion onset. Similarly, although approximately only 40% of statin users had a formal diagnosis of hypercholesterolemia recorded in the extracted records, this likely reflects the effect of statins in normalizing LDL levels, rather than true absence of the condition. Because our dataset captured diagnoses present at the time of lesion diagnosis, earlier hypercholesterolemia diagnoses may not have been retained if the condition was controlled. This limitation may underestimate the proportion of patients with metabolic risk factors and should be considered when interpreting comorbidity prevalence.
Additionally, our use of LSIL as the reference group rather than individuals without cervical lesions introduces further limitations. While this approach allowed for a more clinically comparable sample — all patients had received pap screening and undergone evaluation for abnormal cytology — it may have also attenuated the strength of associations, since LSIL is itself an early-stage lesion with potential for progression. However, many LSIL cases resolve on their own, and may not have progressed to HSIL, regardless of statin use. Patients with LSIL may also differ from those with HSIL in ways not fully captured in our data, such as healthcare and screening access and utilization patterns, that could potentially correlate with statin use, thus contributing to residual confounding. Future studies including patients without cytologic abnormalities could better characterize the full risk spectrum and more accurately assess statins’ potential protective effects.
Future Directions:
Given these promising but preliminary findings, further research is warranted. Prospective longitudinal studies are needed to confirm temporal associations between statin use and cervical dysplasia progression. Similarly, our cross-sectional design limits the ability to evaluate longitudinal outcomes such as recurrence or progression of HSIL. Future studies should investigate whether statin initiation or continued use following an HSIL diagnosis is associated with a reduced risk of lesion recurrence or persistence. Longitudinal cohort analyses stratified by hypercholesterolemia status and statin use patterns are needed to better define the potential role of statins in secondary prevention of cervical dysplasia. Although we adjusted for key demographic and clinical variables, residual confounding cannot be ruled out, and future studies using propensity score matching or prospective cohort designs may help clarify causal associations.
Our analysis was limited to patients with LSIL or HSIL, allowing for a more clinically comparable cohort with confirmed dysplasia and similar healthcare engagement. However, this restricts generalizability to the broader population, particularly individuals without cervical lesions or those not screened. Future studies including broader populations, such as screened patients without cytologic abnormalities, will be needed to assess whether similar associations hold in the general population. Likewise, given that LSIL lesions often regress spontaneously, they may be less influenced by statin use compared to HSIL. Future research should explore statins’ role in the natural history of LSIL and examine effects stratified by HPV genotype, especially oncogenic subtypes, to better understand potential mechanisms of protection. Moreover, mechanistic studies exploring molecular pathways through which statins impact cervical epithelial tissue could yield valuable insights into potential chemopreventive targets. Likewise, given emerging evidence that statin use may improve outcomes in other cancers, including those unrelated to HPV21–24,38, similar studies of other cancerous and pre-cancerous conditions would be valuable, and could help clarify whether the observed effects are cervical-specific or reflect broader systemic mechanisms or interactions.
Importantly, interventional trials testing the efficacy of statins in preventing HSIL progression among high-risk populations could substantiate clinical practice changes. Finally, targeted strategies are needed to expand access to metabolic screening and treatment—particularly in medically underserved populations—to help reduce persistent differences in cervical dysplasia and cancer outcomes.
Conclusions:
In summary, we observed a pronounced and independent inverse association between statin use and HSIL risk in a large, diverse patient population. These preliminary data support the hypothesis that statins may confer chemopreventive benefits by modulating metabolic and inflammatory pathways central to cervical carcinogenesis. Future research should investigate the timing, dose-response relationships, mechanisms underlying this effect, and strategies to increase accessible statin utilization. If confirmed, statin therapy may represent a cost-effective adjunct for preventing HSIL progression and reducing the physical and financial burdens of invasive procedures, particularly in populations at greatest risk of both cervical cancer and cardiovascular disease morbidity and mortality.
Acknowledgements
We are deeply grateful to the Massey Comprehensive Cancer Center for their steadfast support and the resources that made this research possible. We also extend heartfelt thanks to the members of the COQUI Lab, whose thought partnership, dedication, and collaboration were vital to shaping and advancing this work. Your contributions have left an indelible mark on this project—and on us.
Funding:
Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under award numbers P30CA016059 and K01CA285946, and by the Victory Over Cancer Grant #V2023-005. K.Y. Tossas was the recipient of these awards.
Footnotes
Conflicts of Interest: The authors report no conflict of interest.
Declaration of Generative AI and AI-assisted technologies in the writing process: The author acknowledges the use of ChatGPT, a generative AI tool, for editorial assistance in sentence refinement, clarity, and conciseness. The author affirms that all intellectual content, critical analysis, and interpretations are solely their own.
REFERENCES
- 1.World Health Organization. Cervical cancer. [cited 2025 Feb 16]. Available from: https://www.who.int/news-room/fact-sheets/detail/cervical-cancer
- 2.Centers for Disease Control and Prevention. Impact of the HPV Vaccine. Human Papillomavirus (HPV). 2024. Jul 15. [cited 2025 Mar 7]. Available from: https://www.cdc.gov/hpv/vaccination-impact/index.html
- 3.Hughes KC, Herring TA, Song JN, Huh WK, Wright JD, Stoler MH, et al. Cervical high-grade squamous intraepithelial lesion burden and standard of care treatment effectiveness and safety in the United States, 2008–2018: The EACH-WOMAN Project. J Low Genit Tract Dis. 2023;27(2):105–12. doi: 10.1097/LGT.0000000000000719 [DOI] [PubMed] [Google Scholar]
- 4.Hirth J, McGrath CJ, Kuo YF, Rupp RE, Starkey JM, Berenson AB. Impact of human papillomavirus vaccination on racial/ethnic disparities in vaccine-type human papillomavirus prevalence among 14–26 year old females in the US. Vaccine. 2018;36(50):7682–8. doi: 10.1016/j.vaccine.2018.10.075 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Meggiolaro A, Unim B, Semyonov L, Miccoli S, Maffongelli E, La Torre G. The role of Pap test screening against cervical cancer: a systematic review and meta-analysis. Clin Ter. 2016;167:124–39. doi: 10.7417/CT.2016.1942 [DOI] [PubMed] [Google Scholar]
- 6.Trimble CL, Piantadosi S, Gravitt P, Ronnett BM, Pizer E, Elko A, et al. Spontaneous regression of high-grade cervical dysplasia: effects of human papillomavirus type and HLA phenotype. Clin Cancer Res. 2005;11(13):4717–23. doi: 10.1158/1078-0432.CCR-04-2599 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.American College of Obstetricians and Gynecologists. Updated guidelines for management of cervical cancer screening abnormalities. 2020. Oct. [cited 2025 Mar 26]. Available from: https://www.acog.org/clinical/clinical-guidance/practice-advisory/articles/2020/10/updated-guidelines-for-management-of-cervical-cancer-screening-abnormalities
- 8.Bruno MT, Cassaro N, Mazza G, Cicciarelli A, Vecchio GM, Caruso G, et al. Spontaneous regression of cervical intraepithelial neoplasia 3 in women with a biopsy-cone interval of greater than 11 weeks. BMC Cancer. 2022;22:1072. doi: 10.1186/s12885-022-10179-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jin G, Zhang L, Li C, Zhang D. Pregnancy outcome following loop electrosurgical excision procedure (LEEP): a systematic review and meta-analysis. Arch Gynecol Obstet. 2014;289(1):85–99. doi: 10.1007/s00404-013-2955-0 [DOI] [PubMed] [Google Scholar]
- 10.Kyrgiou M, Mitra A, Arbyn M, Paraskevaidis E, Athanasiou A, Martin-Hirsch PP, et al. Fertility and early pregnancy outcomes after treatment for cervical intraepithelial neoplasia: systematic review and meta-analysis. BMJ. 2014;349:g6192. doi: 10.1136/bmj.g6192 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jovani M, Liu EE, Paniagua SM, Fong MW, Lee H, Bhatt A, et al. Cardiovascular disease related circulating biomarkers and cancer incidence and mortality: is there an association? Cardiovasc Res. 2022;118(10):2317–28. doi: 10.1093/cvr/cvab282 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Centers for Disease Control and Prevention. FastStats. 2024. Oct 26. [cited 2025 Mar 7]. Available from: https://www.cdc.gov/nchs/fastats/leading-causes-of-death.htm
- 13.Wudtisan J, Tantipalakorn C, Charoenkwan K, Sreshthaputra RA, Srisomboon J. Factors associated with development of high-grade squamous intraepithelial lesions of the uterine cervix in women younger than 30 years. Asian Pac J Cancer Prev. 2019;20(4):1031–6. doi: 10.31557/APJCP.2019.20.4.1031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lee DY, Lee TS. Associations between metabolic syndrome and gynecologic cancer. Obstet Gynecol Sci. 2020;63(3):215–24. doi: 10.5468/ogs.2020.63.3.215 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Moore JX, Chaudhary N, Akinyemiju T. Metabolic syndrome prevalence by race/ethnicity and sex in the United States, National Health and Nutrition Examination Survey, 1988–2012. Prev Chronic Dis. 2017;14:E24. doi: 10.5888/pcd14.160287 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Lin F, Zheng R, Yu C, Su Y, Yan X, Qu F. Predictive role of serum cholesterol and triglycerides in cervical cancer survival. Int J Gynecol Cancer. 2021;31(2):171–6. doi: 10.1136/ijgc-2020-001333 [DOI] [PubMed] [Google Scholar]
- 17.Kajabwangu R, Ngonzi J, Izudi J, Mpiima DP, Atuhairwe C, Atuhaire C, et al. Association between metabolic syndrome and cervical cancer among women in Southwestern Uganda: a case-control study. Gynecol Oncol Rep. 2024;54:101465. doi: 10.1016/j.gore.2024.101465 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Crescencio ME, Rodríguez E, Páez A, Masso FA, Montaño LF, López-Marure R. Statins inhibit the proliferation and induce cell death of human papillomavirus positive and negative cervical cancer cells. Int J Biomed Sci. 2009;5(4):411–20. [PMC free article] [PubMed] [Google Scholar]
- 19.Chen Z, Wu P, Wang J, Chen P, Fang Z, Luo F. The association of statin therapy and cancer: a meta-analysis. Lipids Health Dis. 2023;22(1):192. doi: 10.1186/s12944-023-01955-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Xie L, Zhu G, Shang J, Yu H, Zhang X, Li C, et al. An overview on the biological activity and anti-cancer mechanism of lovastatin. Cell Signal. 2021;87:110122. doi: 10.1016/j.cellsig.2021.110122 [DOI] [PubMed] [Google Scholar]
- 21.Bjarnadottir O, Romero Q, Bendahl PO, Jirström K, Rydén L, Loman N, et al. Targeting HMG-CoA reductase with statins in a window-of-opportunity breast cancer trial. Breast Cancer Res Treat. 2013;138(2):499–508. doi: 10.1007/s10549-013-2473-6 [DOI] [PubMed] [Google Scholar]
- 22.Nielsen SF, Nordestgaard BG, Bojesen SE. Statin use and reduced cancer-related mortality. N Engl J Med. 2012;367(19):1792–802. doi: 10.1056/NEJMoa1201735 [DOI] [PubMed] [Google Scholar]
- 23.Dale KM, Coleman CI, Henyan NN, Kluger J, White CM. Statins and cancer risk: a meta-analysis. JAMA. 2006;295(1):74–80. doi: 10.1001/jama.295.1.74 [DOI] [PubMed] [Google Scholar]
- 24.Kenik J, Jean-Jacques M, Feinglass J. Explaining racial and ethnic disparities in cholesterol screening. Prev Med. 2014;65:65–9. doi: 10.1016/j.ypmed.2014.04.026 [DOI] [PubMed] [Google Scholar]
- 25.Centers for Disease Control and Prevention. Products – Data Briefs – Number 36 – April 2010. 2019. Jun 10. [cited 2025 Mar 7]. Available from: https://www.cdc.gov/nchs/products/databriefs/db36.htm
- 26.Kalra DK. Bridging the racial disparity gap in lipid-lowering therapy. J Am Heart Assoc. 2021;10(1):e019533. doi: 10.1161/JAHA.120.019533 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Nelson K, Norris K, Mangione CM. Disparities in the diagnosis and pharmacologic treatment of high serum cholesterol by race and ethnicity: data from the Third National Health and Nutrition Examination Survey. Arch Intern Med. 2002;162(8):929–35. doi: 10.1001/archinte.162.8.929 [DOI] [PubMed] [Google Scholar]
- 28.Centers for Disease Control and Prevention. Heart disease deaths – Health, United States. 2024. Aug 5. [cited 2025 Mar 7]. Available from: https://www.cdc.gov/nchs/hus/topics/heart-disease-deaths.htm
- 29.Amrock SM, Duell PB, Knickelbine T, Knowles JW, Sturm AC, Ahmad ZS, et al. Health disparities among adult patients with a phenotypic diagnosis of familial hypercholesterolemia in the CASCADE-FH™ patient registry. Atherosclerosis. 2017;267:19–26. doi: 10.1016/j.atherosclerosis.2017.10.006 [DOI] [PubMed] [Google Scholar]
- 30.Jackson EA, Ruppert K, Derby CA, Manson JE, Carson AP, Lewis CE, et al. Is race or ethnicity associated with under-utilization of statins among women in the United States: The study of women’s health across the nation. Clin Cardiol. 2020;43(12):1388–97. doi: 10.1002/clc.23448 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Pangarkar MA. The Bethesda System for reporting cervical cytology. Cytojournal. 2022;19:28. doi: 10.25259/CMAS_03_07_2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Jones HM, Fang Z, Sun W, Tang Y, Wu J, Liu L, et al. Atorvastatin exhibits anti-tumorigenic and anti-metastatic effects in ovarian cancer in vitro. Am J Cancer Res. 2017;7(12):2478–90. [PMC free article] [PubMed] [Google Scholar]
- 33.Kim DS, Ahn HS, Kim HJ. Statin use and incidence and mortality of breast and gynecology cancer: a cohort study using the National Health Insurance claims database. Int J Cancer. 2022;150(7):1156–65. doi: 10.1002/ijc.33869 [DOI] [PubMed] [Google Scholar]
- 34.Mei Z, Liang M, Li L, Zhang Y, Wang Q, Yang W. Effects of statins on cancer mortality and progression: a systematic review and meta-analysis of 95 cohorts including 1,111,407 individuals. Int J Cancer. 2017;140(5):1068–81. doi: 10.1002/ijc.30526 [DOI] [PubMed] [Google Scholar]
- 35.Zhuang L, Kim J, Adam RM, Solomon KR, Freeman MR. Cholesterol targeting alters lipid raft composition and cell survival in prostate cancer cells and xenografts. J Clin Invest. 2005;115(4):959–68. doi: 10.1172/JCI19935 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Liang J, Yu D, Luo C, Guo W, Zhang Y, Xu J, et al. Epigenetic suppression of PGC1α (PPARGC1A) causes collateral sensitivity to HMGCR-inhibitors within BRAF-treatment resistant melanomas. Nat Commun. 2023;14(1):3251. doi: 10.1038/s41467-023-38968-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Dorsch MP, Lester CA, Ding Y, Joseph M, Brook RD. Effects of race on statin prescribing for primary prevention with high atherosclerotic cardiovascular disease risk in a large healthcare system. J Am Heart Assoc. 2019;8(22):e014709. doi: 10.1161/JAHA.119.014709 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Lebo NL, Griffiths R, Hall S, Dimitroulakos J, Johnson-Obaseki S. Effect of statin use on oncologic outcomes in head and neck squamous cell carcinoma. Head Neck. 2018;40(8):1697–706. doi: 10.1002/hed.25152 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The de-identified data generated in this study are available from the corresponding author upon reasonable request.