Preoperative Statin Use at the Time of Radical Prostatectomy Is Not Associated With Biochemical Recurrence or Pathologic Upgrading

Accepted Manuscript Title: Preoperative Statin Use at the Time of Radical Prostatectomy is Not Associated with Biochemical Recurrence or Pathological Upgrading Author: Timothy D. Lyon, Robert M. Turner II, Jonathan G. Yabes, Elen Woldemichael, Benjamin J. Davies, Bruce L. Jacobs, Joel B. Nelson PII: DOI: Reference:

S0090-4295(16)30498-8 http://dx.doi.org/doi: 10.1016/j.urology.2016.08.004 URL 19956

To appear in:

Urology

Received date: Accepted date:

1-3-2016 1-8-2016

Please cite this article as: Timothy D. Lyon, Robert M. Turner II, Jonathan G. Yabes, Elen Woldemichael, Benjamin J. Davies, Bruce L. Jacobs, Joel B. Nelson, Preoperative Statin Use at the Time of Radical Prostatectomy is Not Associated with Biochemical Recurrence or Pathological Upgrading, Urology (2016), http://dx.doi.org/doi: 10.1016/j.urology.2016.08.004. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1 Preoperative Statin Use at the Time of Radical Prostatectomy is Not Associated with Biochemical Recurrence or Pathological Upgrading Timothy D. Lyon, MD1, Robert M. Turner II, MD1, Jonathan G. Yabes, PhD2, Elen Woldemichael, BS1, Benjamin J. Davies, MD1, Bruce L. Jacobs, MD, MPH1, Joel B. Nelson, MD1 1

Department of Urology, University of Pittsburgh Medical Center, 3471 Fifth Ave., Suite 700, Pittsburgh, PA, 15213, USA 2

Department of Medicine, University of Pittsburgh Medical Center, 200 Meyran Ave., Pittsburgh, PA, 15213, USA Corresponding Author:

Timothy D. Lyon, MD 3471 Fifth Ave., Suite 700 Pittsburgh, PA 15213 Ph: 412-692-4091 Fax: 412-692-4101 [email protected]

Additional Authors:

[email protected] [email protected] [email protected] [email protected] [email protected] [email protected]

Running Head Statin Use at RRP Keywords Prostate cancer; statins, HMG-CoA; recurrence; Gleason grading Funding/Disclosures Robert M. Turner II, MD is supported in part by the National Institutes of Health Institutional TL1 award (5TL1TR000145-10). Bruce L. Jacobs, MD is supported in part by the National Institutes of Health Institutional KL2 award (KL2TR000146-08), the GEMSSTAR award (R03AG048091), the Jahnigen Career Development Award, and the Tippins Foundation Scholar Award.

Page 1 of 20

2 ABSTRACT Objective: To determine the association of statin use with oncologic outcomes and risk of pathologic upgrading following radical prostatectomy. Materials and Methods: Using a prospectively populated database of 3,042 men who underwent open radical prostatectomy, patients were grouped according to reported statin use at the time of surgery. The primary outcome was time to biochemical recurrence. The secondary outcome was risk of pathological upgrading among a subset of 1,256 patients with Gleason pattern 3+3=6 on biopsy. A multivariable Cox model was used to assess risk of biochemical recurrence, and multivariable logistic regression was used to assess risk of pathological upgrading. Results: 824 men (27%) reported statin use at the time of radical prostatectomy. Statin users were older and had higher body mass index (BMI), higher Charlson comorbidity index, and lower pretreatment PSA values than statin nonusers. Over a median follow-up of 70 months (IQR 36-107), a total of 455 men (15%) experienced biochemical recurrence. Statin use was not associated with biochemical recurrence (adjusted HR 1.06, 95% CI 0.86-1.31). Of those men with biopsy Gleason 3+3=6 disease, 647 (52%) were upgraded to higher-grade disease following radical prostatectomy, however statin use was not associated with pathologic upgrading (adjusted OR 0.78, 95% CI 0.58-1.04). Conclusion: Preoperative statin use at the time of radical prostatectomy was not associated with biochemical recurrence or risk of pathologic upgrading in this cohort. These data add to the existing body of literature suggesting that statin use is not associated with more favorable clinical outcomes following radical prostatectomy.

INTRODUCTION

Page 2 of 20

3 Statins may inhibit the growth of prostate cancer by decreasing serum cholesterol, which is a known precursor for androgen synthesis. Both in vitro and animal studies have demonstrated not only that cholesterol promotes survival for prostate cancer cells,1 but also that decreasing cholesterol reduces androgen levels and slows prostate cancer growth.2-4 Whether statin use translates into a clinical benefit for men with prostate cancer, however, is controversial. The existing data regarding the impact of preoperative statin use on recurrence following radical prostatectomy is mixed. Two meta-analyses have concluded that preoperative statin use has no effect on the rate of biochemical recurrence (BCR) after radical prostatectomy.5,6 Conversely, work by several others has found that both preoperative7 and postoperative statin use8,9 is associated with a decreased risk of BCR. Preoperative statin use has also been linked to more favorable pathologic features, including lower risk of non-organ-confined disease10 as well as lower prostate-specific antigen (PSA) levels at diagnosis, smaller tumor volume, lower percentage of tumor in gland, and decreased risk of positive surgical margins.11,12 These data suggest there may also be a decreased risk of pathologic upgrading among statin users diagnosed with Gleason 3+3=6 disease on biopsy, which could impact men considering active surveillance. In this study, the association of preoperative statin use with biochemical recurrence and risk of pathologic upgrading following radical prostatectomy was examined. METHODS Study Population

Page 3 of 20

4 A total of 3,084 men who underwent radical prostatectomy by a single surgeon (JBN) at the University of Pittsburgh Medical Center between November 1999 and July 2015 comprise the study cohort. These men were included in a prospectively populated and continuously maintained database. After excluding 42 patients who received neoadjuvant androgen deprivation therapy, the study population was 3,042 men. Next, a subset of patients with Gleason 3+3=6 disease on prostate biopsy was identified. For this subset, patients who were diagnosed from prostatic tissue obtained from transurethral resection (n = 6), had fewer than six biopsy cores (n = 31), or were missing data regarding the number of cores (n = 184), number of positive cores (n = 9), biopsy grade (n = 1), and pathological grade (n = 1) were excluded. The final study population for the patients with Gleason 3+3=6 disease on biopsy was 1,256 men. Specialty-trained, genitourinary pathologists reviewed all biopsy and pathologic specimens. Postoperatively, patients were monitored with periodic clinical and prostate-specific antigen assessments. Outcomes The primary study outcome was time from radical prostatectomy to biochemical recurrence, defined as the time of a second postoperative PSA value of 0.2 ng/mL or greater. The secondary study outcome was the risk of pathological upgrading among the subset of patients with Gleason 3+3=6 disease on biopsy. Pathologic upgrading was defined as a pathologic Gleason score of 3+4=7 or greater. The primary predictor was reported statin use at the time of radical prostatectomy. For purposes of comparison, patients were grouped as statin users or nonusers. On subset analysis, reported statin use was tested along with other demographic and clinical

Page 4 of 20

5 characteristics known to be predictors of pathologic upgrading including age, BMI, race, Charlson comorbidity index, preoperative PSA, year of surgery, clinical stage, number of biopsy cores taken, highest positive percentage of a single biopsy core, prostate size, time between biopsy and RP and whether or not biopsy pathology was reviewed at the University of Pittsburgh Medical Center by a dedicated genitourinary pathologist. Statistical Analysis Demographic, clinical and pathologic characteristics were compared between statin users and nonusers using t-test (parametric) or Wilcoxon rank sum test (nonparametric) for continuous variables, and chi square or Fisher’s exact tests for categorical variables. Time to biochemical recurrence was calculated from the date of radical prostatectomy to the date of recurrence. Patients were censored at the time of death or last follow up. Death was not treated as a competing event due to a small proportion of deaths without recurrence (3%). The Kaplan-Meier method was used to estimate biochemical recurrence-free survival curves for each statin group and compared using the log-rank test. Cox proportional hazards models were used to calculate unadjusted and adjusted hazards ratios for biochemical recurrence. Variables used in the adjusted models included those with p < 0.10 in univariable analysis as well as other clinicopathological variables known to be associated with biochemical recurrence including pathologic stage and grade, perineural invasion, margin status and tumor size.5-10 The subset analysis to determine predictors of pathological upgrading was performed using multivariable logistic regression. All clinical variables in the univariable analysis have previously been associated with upgrading and thus all were included in the

Page 5 of 20

6 multivariable model.13-15 Both unadjusted and adjusted odds ratios for the risk of upgrading were calculated, and all variables used in the unadjusted model were included in the adjusted model. All continuous covariates were tested for linearity, and the percent of positive biopsy cores was incorporated in the final model as a non-linear term. Statistical analyses were carried out using R (version 13.2)16 using the packages dpylr17 for data manipulation, compareGroups18 for descriptive tables, ggplot219 for graphics, survival20 for survival analysis. Statistical significance was defined as p < 0.05. The University of Pittsburgh institutional review board approved the study (IRB0304058). RESULTS Demographic and clinicopathologic characteristics of the study population are summarized in Table 1. Median follow-up was 70 months (IQR 36-107 months). A total of 824 (27%) men reported statin use at the time of radical prostatectomy. Compared to nonusers, statin users were significantly older (p < 0.001), had higher body mass index (BMI, p <0.001), higher Charlson comorbidity index (p < 0.001), and lower pretreatment PSA values (p = 0.02). There was no difference between groups with regard to pathologic grade, pathologic stage, or positive margin status. Median time from prostate biopsy to RP was not different between groups (2.3 months [IQR 2.7-3.2] vs 2.3 months [IQR 2.73.2], p = 0.90). During the study period, 455 men (15%) experienced biochemical recurrence. In statin users, the 5-year and 10-year estimates of biochemical recurrence-free survival were 85.7% and 79.5 %, respectively; in statin nonusers, the 5-year and 10-year estimates

Page 6 of 20

7 of biochemical recurrence-free survival were 86.9% and 79.5%, respectively (p = 0.61) (Figure 1). Univariable and multivariable analyses examining predictors of biochemical recurrence are shown in Table 2. On multivariable analysis, preoperative PSA, pathologic Gleason score, pathologic T stage, node positive disease, perineural invasion, a positive surgical margin, and increasing diameter of the dominant tumor nodule were found to be independently associated with biochemical recurrence (all p < 0.01); however, statin use was not found to be a significant predictor of biochemical recurrence (adjusted HR 1.06, 95% CI 0.86-1.31). A subset of 1256 men with Gleason 3+3=6 disease on biopsy were included in analysis of predictors of pathologic upgrading. Of these men 1150 (92%) had their biopsy pathology reviewed by a genitourinary pathologist prior to RP. Statin users were found to be older (p < 0.001), had higher BMI (p = 0.02), and greater Charlson comorbidity index scores (p < 0.001) than statin nonusers (Supplementary table). Results of the multivariable logistic regression model examining predictors of pathologic upgrading are summarized in Table 3. Statin use was not associated with pathologic upgrading (adjusted OR 0.78, 95% CI 0.58-1.04) after adjusting for age, race, BMI, Charlson comorbidity index, preoperative PSA, year of surgery, clinical stage, percentage of positive cores, highest percentage of cancer in a single core, prostate weight, time between biopsy and RP and whether biopsy pathology was reviewed by a genitourinary pathologist. Significant predictors of upgrading included increasing age, pretreatment PSA, year of surgery, percentage of cores positive, smaller prostate size,

Page 7 of 20

8 increasing time between biopsy and RP, and not having pathologic re-review of the biopsy specimen (all p < 0.05). DISCUSSION Preoperative statin use was not found to be associated with a decreased risk of biochemical recurrence in this cohort. Statin users were significantly older, had higher BMIs, and more comorbidities than their counterparts. Despite this, no demographic factors were found to be significant predictors of BCR; only disease characteristics including higher PSA, higher grade or stage, perineural invasion, positive margins, or increasing diameter of tumor nodule were associated with an increased risk of BCR. Additionally, this is the first study to our knowledge to investigate a potential relationship between preoperative statin use and the rate of pathologic upgrading for patients with Gleason 3+3=6 disease on biopsy. Statin use was not associated with a decreased risk of upgrading in this cohort (adjusted OR 0.78, 95% CI 0.58-1.04). Significant predictors of upgrading from Gleason 3+3=6 included increasing age, higher preoperative PSA, increasing percentage of positive cores, smaller prostate weight, increasing time from biopsy to RP, and lack of pathologic review of the biopsy specimen, as have been reported previously.13-15 Statins have been widely studied for their effects on prostate cancer. In vitro, statins can inhibit prostate cancer cell growth and development by inhibiting the mevalonate pathway, reducing angiogenesis and inhibiting androgen receptor expression.2,21,22 Several large, population-based studies investigated the association between statin use and the risk of being diagnosed with prostate cancer; of the 5 metaanalyses published on the topic, four concluded that there was no association, 23-26 and

Page 8 of 20

9 one found a 7% reduction of prostate cancer incidence among statin users.27 On the basis of these data, statins have not been widely recommended for the prevention of prostate cancer. Many investigators have since shifted focus to whether statins can be beneficial to men who already have prostate cancer. The evidence linking statin use to a decreased risk of biochemical recurrence following radical prostatectomy is mixed, although two large meta-analyses refute this hypothesis. A series of 1146 patients who underwent radical prostatectomy were combined in a meta-analysis including 6 other series to evaluate the association of preoperative statin use to BCR. The authors concluded that preoperative statin use was not associated with BCR, with a pooled HR of 1.00 (95% CI 0.80-1.19).5 Similarly, a meta-analysis of 5 studies found that preoperative statin use was not associated with the risk of BCR, adjusted HR 1.02 (95% CI 0.80-1.29).6 Although these meta-analyses are limited by the retrospective nature of comprising studies, no prospective series are available. The results of the present study support these data. There is, however, also some evidence to support a link between statin use and a decreased risk of biochemical recurrence. In a population of 1,319 men who underwent prostatectomy from 1998-2008 at Veteran’s Affairs hospitals in the United States, preoperative statin use was not associated with an increased risk of BCR on univariable analysis, HR 0.9 (95% CI 0.68-1.19); however, in the adjusted analysis statin use became a significant predictor, adjusted HR 0.7 (95% CI 0.51-0.98).7 On secondary analysis of dose, statin use remained a significant predictor of BCR only if a minimum dose of 20mg simvastatin (or equivalent) was used. An analysis of 2,137 Korean men who underwent radical prostatectomy failed to implicate preoperative statin use to the risk of BCR, but

Page 9 of 20

10 did find that postoperative statin use was associated with a decreased risk of BCR for men with Gleason 7 or greater disease, adjusted HR 0.27 (95% CI 0.13-0.59).9 Undergrading of prostate cancer on biopsy poses a significant clinical challenge, as men who harbor more significant disease than is found on biopsy may be inappropriately recommended for active surveillance. Contemporary series suggest that upgrading from 3+3=6 on biopsy to 3+4=7 or higher on radical prostatectomy specimen occurs in 20-50% of cases.13-15,28 Statins have been associated with favorable pathologic features at radical prostatectomy, including lower risk of non-organ-confined disease, lower PSA at diagnosis, smaller tumor volume, lower percentage of tumor in gland, and decreased risk of positive surgical margins,10-12 leading to the hypothesis that statin use might confer a decreased risk of pathologic upgrading. However, these data failed to demonstrate a significant relationship between statin use and the risk of pathologic upgrading in men with Gleason 3+3=6 disease on biopsy. Instead, the data suggest that clinicians should continue to utilize clinical variables including patient age, prostatespecific antigen level, volume of disease, gland size, time from biopsy to RP and lack of review of biopsy pathology to assess a patient’s risk of upgrading at radical prostatectomy. The reported overall upgrading rate of 52% is at the high end of literature reported rates.13-15,28 To investigate potential confounders in this relationship, time from biopsy to RP and rate of pathologic re-review of biopsy specimens were included in the multivariable model for upgrading. The median time from biopsy to RP was not different between statin users and nonusers (p = 0.9), and there was also no difference in the proportion of men in each group who had their biopsy pathology reviewed by a

Page 10 of 20

11 University of Pittsburgh genitourinary pathologist (p = 0.33). Both of these variables were found to be significant predictors of upgrading at the time of RP (Table 3) and may account for the rate of upgrading seen in the present series. Results of the current study should be interpreted in light of its limitations. As with all retrospective studies, the potential for bias from unmeasured confounders exists. Data regarding statin exposure was limited, as dosage, type and duration of preoperative statin use was not collected prospectively and therefore is not analyzed. There is potential for significant variability within the statin user group in terms of dosage, duration of use and adherence to the medication regimen which could not be controlled for due to the absence of data, and therefore all persons classified as statin users do not necessarily have equal exposure. Further, it is unknown whether any statin non-users had previously used statins, which could further contaminate the makeup of the study groups and affect the validity of the results. Although these concerns limit the strength of our findings, they are likely reflective of real world patient medication adherence; in one study of 541,221 patients with high cardiovascular risk, median time to self-discontinuation of statin therapy was 15 months.29 Previous work has shown a potential benefit of postoperative statin use on the risk of BCR.8,9 Although this was not analyzed, statins are usually intended as lifelong medications, with as many as 44% of patients using them 10 years after initial prescription,30 and therefore a proportion of statin users likely continued to use them in the postoperative setting. The study population was primarily Caucasian and all operations were performed by a single high-volume surgeon, which limits the generalizability of results; however, these concerns are in part mitigated by the large

Page 11 of 20

12 sample size, long duration of follow-up, and the absence of variation in surgical technique or surgical quality that may bias previous studies on this topic. Despite these limitations, the reported results are important as they add to a growing body of literature suggesting there is no beneficial relationship between preoperative statin use and BCR following radical prostatectomy, and is the first study to our knowledge to report on the relationship between preoperative statin use and the risk of pathologic upgrading for patients diagnosed with Gleason 3+3=6 disease on biopsy. Multivariable analysis did not reveal a significant relationship between reported statin use and pathologic upgrading in low risk men (p = 0.09), however the adjusted odds ratio (95% CI) of 0.78 (0.58-1.04) suggests the potential for as much as a 40% decrease in upgrading risk and may be an area for further research, particularly in a cohort of men with more detailed information on statin exposure. Taken in conjunction with prior work, these data decrease enthusiasm that preoperative statin use should be considered beneficial for cancer-specific outcomes in men undergoing radical prostatectomy. CONCLUSION In a large cohort of men who underwent radical prostatectomy, reported preoperative statin use was not associated with risk of biochemical recurrence. Among a subset of these men with Gleason 3+3=6 disease on biopsy, preoperative statin use was not associated with an increased risk of pathologic upgrading. These data are important as they add to a mixed body of literature regarding the impact of statin use in men with prostate cancer, and add further credence to the conclusion that statin does not positively influence oncologic outcomes following radical prostatectomy. Statins may have some

Page 12 of 20

13 benefit but until a clearer consensus is reached, clinicians should not recommend their use for the purpose of improving prostate cancer-specific outcomes.

REFERENCES 1.

Zhuang L, Lin J, Lu ML, et al.: Cholesterol-rich lipid rafts mediate aktregulated survival in prostate cancer cells. Cancer research 2002; 62: 22272231.

2.

Hoque A, Chen H and Xu XC: Statin induces apoptosis and cell growth arrest in prostate cancer cells. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology 2008; 17: 88-94.

3.

Mostaghel EA, Solomon KR, Pelton K, et al.: Impact of circulating cholesterol levels on growth and intratumoral androgen concentration of prostate tumors. PloS one 2012; 7: e30062.

4.

Pelton K, Freeman MR and Solomon KR: Cholesterol and prostate cancer. Current opinion in pharmacology 2012; 12: 751-759.

5.

Mass AY, Agalliu I, Laze J, et al.: Preoperative statin therapy is not associated with biochemical recurrence after radical prostatectomy: our experience and meta-analysis. The Journal of urology 2012; 188: 786-791.

6.

Scosyrev E, Tobis S, Donsky H, et al.: Statin use and the risk of biochemical recurrence of prostate cancer after definitive local therapy: a meta-analysis of eight cohort studies. BJU international 2013; 111: E71-77.

Page 13 of 20

14 7.

Hamilton RJ, Banez LL, Aronson WJ, et al.: Statin medication use and the risk of biochemical recurrence after radical prostatectomy: results from the Shared Equal Access Regional Cancer Hospital (SEARCH) Database. Cancer 2010; 116: 3389-3398.

8.

Allott EH, Howard LE, Cooperberg MR, et al.: Postoperative statin use and risk of biochemical recurrence following radical prostatectomy: results from the Shared Equal Access Regional Cancer Hospital (SEARCH) database. BJU international 2014; 114: 661-666.

9.

Song C, Park S, Park J, et al.: Statin use after radical prostatectomy reduces biochemical recurrence in men with prostate cancer. The Prostate 2015; 75: 211-217.

10.

Mondul AM, Han M, Humphreys EB, et al.: Association of statin use with pathological tumor characteristics and prostate cancer recurrence after surgery. The Journal of urology 2011; 185: 1268-1273.

11.

Krane LS, Kaul SA, Stricker HJ, et al.: Men presenting for radical prostatectomy on preoperative statin therapy have reduced serum prostate specific antigen. The Journal of urology 2010; 183: 118-124.

12.

Loeb S, Kan D, Helfand BT, et al.: Is statin use associated with prostate cancer aggressiveness? BJU international 2010; 105: 1222-1225.

13.

Capitanio U, Karakiewicz PI, Valiquette L, et al.: Biopsy core number represents one of foremost predictors of clinically significant gleason sum upgrading in patients with low-risk prostate cancer. Urology 2009; 73: 10871091.

Page 14 of 20

15 14.

Epstein JI, Feng Z, Trock BJ, et al.: Upgrading and downgrading of prostate cancer from biopsy to radical prostatectomy: incidence and predictive factors using the modified Gleason grading system and factoring in tertiary grades. European urology 2012; 61: 1019-1024.

15.

Pinthus JH, Witkos M, Fleshner NE, et al.: Prostate cancers scored as Gleason 6 on prostate biopsy are frequently Gleason 7 tumors at radical prostatectomy: implication on outcome. The Journal of urology 2006; 176: 979-984; discussion 984.

16.

Team RDC: R: A language and environment for statistical computing, Vienna, Austria, R Foundation for Statistical Computing, 2010.

17.

Wickham H, Francois R: dplyr: A Grammar of Data Manipulation, R package version 0.4.1, 2015.

18.

I. S SH, Villa J: Building Bivariate Tables: The compareGroups Package for R. Journal of Statistical Software 2014; 57: 1-16.

19.

H W: ggplot2: elegant graphics for data analysis: New York, Springer, 2009.

20.

T T: A Package for Survival Analysis in S. version 2.38, 2015.

21.

Yang L, Egger M, Plattner R, et al.: Lovastatin causes diminished PSA secretion by inhibiting AR expression and function in LNCaP prostate cancer cells. Urology 2011; 77: 1508 e1501-1507.

22.

Wang C, Tao W, Wang Y, et al.: Rosuvastatin, identified from a zebrafish chemical genetic screen for antiangiogenic compounds, suppresses the growth of prostate cancer. European urology 2010; 58: 418-426.

Page 15 of 20

16 23.

Bonovas S, Filioussi K and Sitaras NM: Statin use and the risk of prostate cancer: A metaanalysis of 6 randomized clinical trials and 13 observational studies. International journal of cancer Journal international du cancer 2008; 123: 899-904.

24.

Dale KM, Coleman CI, Henyan NN, et al.: Statins and cancer risk: a metaanalysis. Jama 2006; 295: 74-80.

25.

Kuoppala J, Lamminpaa A and Pukkala E: Statins and cancer: A systematic review and meta-analysis. European journal of cancer 2008; 44: 2122-2132.

26.

Browning DR and Martin RM: Statins and risk of cancer: a systematic review and metaanalysis. International journal of cancer Journal international du cancer 2007; 120: 833-843.

27.

Bansal D, Undela K, D'Cruz S, et al.: Statin use and risk of prostate cancer: a meta-analysis of observational studies. PloS one 2012; 7: e46691.

28.

Louie-Johnsun M, Neill M, Treurnicht K, et al.: Final outcomes of patients with low-risk prostate cancer suitable for active surveillance but treated surgically. BJU international 2009; 104: 1501-1504.

29.

Lin I, Sung J, Sanchez RJ, et al.: Patterns of Statin Use in a Real-World Population of Patients at High Cardiovascular Risk. Journal of managed care & specialty pharmacy 2016; 22: 685-698.

30.

Helin-Salmivaara A, Lavikainen P, Korhonen MJ, et al.: Long-term persistence with statin therapy: a nationwide register study in Finland. Clinical therapeutics 2008; 30 Pt 2: 2228-2240.

Page 16 of 20

17 Figure 1. Kaplan-Meier biochemical recurrence-free survival curve after radical prostatectomy

Table 1. Demographic, clinical, and pathological characteristics of the study population Characteristics Statin User Statin Nonuser n = 824 n = 2218 Age, years, mean (SD) 61 (6) 59 (6) Race (%) White 795 (97) 2096 (95) Non-white 29 (4) 122 (6) BMI, kg/m2, median [IQR] 27.9 [25.8-30.5] 27.3 [25.1-30.0] Charlson Comorbidity Index (%) 0 615 (75) 1892 (85) 1 172 (21) 242 (11) 2 or more 37 (5) 84 (3.8) Preoperative PSA, ng/mL, mean (SD) 6.0 (4.9) 6.5 (5.4) Year of RRP 1999-2004 209 (25) 658 (30) 2005-2009 348 (42) 857 (39) 2010-2015 267 (32) 703 (32) Pathologic Gleason score (%) 3+3=6 or less 215 (26) 610 (28) 3+4=7 414 (50) 1121 (51) 4+3=7 118 (14) 286 (13) 8 33 (4) 89 (4) 9-10 43 (5) 110 (5) Pathologic T stage (%) T2 or less 602 (73) 1613 (73) T3/T4 222 (27) 605 (27) Pathologic N stage (%) N0/Nx 797 (97) 2136 (96) N1 27 (3) 82 (4) Perineural invasion (%) Negative 171 (21) 510 (23) Positive 653 (79) 1708 (77) Margin status (%) Negative 767 (93) 2069 (93) Positive 57 (7) 149 (7) Diameter of dominant cancer nodule, cm, 1.38 (0.65) 1.37 (0.64) mean (SD) Biochemical recurrence (%)

P value* < 0.001 0.03 < 0.001 < 0.001

0.02 0.05

0.84

0.89 0.66 0.20 0.91 0.77 0.89

Page 17 of 20

18 No 699 (85) 1888 (85) Yes 125 (15) 330 (15) Abbreviations: SD, standard deviation; BMI, body mass index; PSA, prostatespecific antigen; RRP, radical retropubic prostatectomy; IQR, interquartile range * P values for continuous and categorical variables generated from t tests (or Wilcoxon Rank Sum) and chi-square tests, respectively Table 2. Cox proportional hazards model examining predictors of biochemical recurrence after radical prostatectomy Predictor Univariable analysis Multivariable analysis* Hazard Ratio [95% CI] P Value Hazard Ratio [95% CI] P Value Statin Use 0.61 0.59 No Ref. Ref. Yes 1.05 [0.86-1.30] 1.06 [0.86-1.31] Age 1.02 [1.01-1.04] 0.004 0.99 [0.98-1.01] 0.56 Race 0.19 0.91 Non-white Ref. Ref. White 0.77 [0.52-1.14] 0.98 [0.65-1.46] BMI, kg/m2 1.02 [1.00-1.04] 0.08 1.02 [0.98-1.04] 0.12 Charlson Comorbidity Index 0.04 0 Ref. Ref. 1 1.37 [1.07-1.76] 1.17 [0.90-1.51] ≥2 1.17 [0.73-1.87] 1.35 [0.84-2.19] Preoperative PSA, ng/mL 1.07 [1.06-1.08] < 0.001 1.03 [1.02-1.04] < 0.001 Year of RRP 0.01 0.16 1999-2004 Ref. Ref. 2005-2009 1.13 [0.90-1.41] 1.03 [0.82-1.3] 2010-2015 1.50 [1.15-1.97] 0.80 [0.60-1.07] Pathologic Gleason Grade < 0.001 < 0.001 3+3=6 Ref. Ref. 3+4=7 4.87 [3.04-7.80] 2.81 [1.71-4.62] 4+3=7 23.90 [14.9-38.2] 8.07 [4.84-13.46] 8 32.40 [19.4-54.2] 9.64 [5.51-16.87] 9-10 77.70 [47.8-126] 16.75 [9.76-28.74] Pathologic T Stage < 0.001 < 0.001 pT2 or less Ref. Ref. pT3/T4 8.78 [7.18-10.7] 2.61 [2.06-3.31] Pathologic N Stage < 0.001 < 0.001 N0/Nx Ref. Ref. N1 22.5 [17.7-28.7] 5.07 [3.80-6.75] Perineural invasion < 0.001 0.001 No Ref. Ref. Yes 10.9 [6.37-18.5] 2.55 [1.46-4.45] Margin Status < 0.001 < 0.001 Negative Ref. Ref.

Page 18 of 20

19 Positive 5.13 [4.13-6.37] 1.59 [1.25-2.02] Diameter of dominant cancer 3.16 [2.81-3.55] < 0.001 1.28 [1.09-1.49] nodule, cm Abbreviations: CI, confidence interval; BMI, body mass index; PSA, prostate specific antigen; RRP, radical retropubic prostatectomy * P values for hazard ratios are computed using the log rank or Wald test under a Cox model for categorical or continuous variables, respectively. Multivariable model included all listed covariates.

0.002

Table 3. Results of multivariable logistic regression model examining predictors of pathologic upgrading from biopsy Gleason 3+3=6. P Value* Characteristics Odds Ratio (95% CI) Statin Use 0.09 No Ref. Yes 0.78 (0.58-1.04) Age 1.07 (1.04-1.09) <0.001 BMI 1.01 (0.97-1.04) 0.78 Race 0.06 White Ref. Non-white 1.92 (0.97-3.80) Charlson comorbidity index 0.41 0 Ref. 1 1.17 (0.79-1.75) ≥2 0.71 (0.37-1.37) Preoperative PSA 1.11 (1.06-1.16) <0.001 Year of radical prostatectomy <0.001 1999-2005 Ref. 2006-2015 3.26 (2.45-4.34) Clinical stage 0.82 T1c Ref. T2a 1.13 (0.80-1.60) T2b 1.10 (0.55-2.20) T2c-T4 1.34 (0.61-2.97) Number of biopsy cores 0.96 6-11 Ref. ≥12 0.99 (0.73-1.35) Percentage of cores positive** n/a <0.001 Highest % of CaP in a single core 0.07 <5% Ref. 5-24% 1.47 (1.07-2.01) 25-49% 1.65 (1.05-2.57) >50 1.80 (0.93-3.48) Prostate weight 0.98 (0.98-0.99) <0.001 Months between Biopsy and RP 1.05 (1.004-1.10) 0.03 Biopsy Pathology Reviewed

Page 19 of 20

20 No Ref. Yes 0.61 (0.38-0.99) 0.047 Abbreviations: CI confidence interval; BMI body mass index; PSA prostate specific antigen; CaP cancer of the prostate * P values for odds ratios generated from multivariable logistic regression model adjusting for all listed covariates **Unable to estimate odds ratio as the relationship is non-linear

Page 20 of 20