HSD3B1 Genotype and Clinical Outcomes in Metastatic Castration-Sensitive Prostate Cancer

Key Points

Question  Does inheritance of the adrenal-permissive HSD3B1(1245C) allele indicate probable worse clinical outcomes in men treated with androgen deprivation therapy with or without docetaxel for metastatic castration-sensitive prostate cancer?

Findings  In this study including 475 genotyped white men with metastatic prostate cancer treated in the E3805 Chemohormonal Therapy vs Androgen Ablation Randomized Trial for Extensive Disease in Prostate Cancer trial, the adrenal-permissive genotype (ie, inheritance of ≥1 HSD3B1[1245C] allele) was associated with significantly shorter time to castration-resistant disease and significantly lower overall survival in men with low-volume disease. There was no association between genotype and outcomes in men with high-volume disease.

Meaning  The adrenal-permissive genotype is associated with inferior outcomes in low-volume metastatic prostate cancer and may help identify patients who could benefit from escalated therapy.

Importance  The adrenal-restrictive HSD3B1(1245A) allele limits extragonadal dihydrotestosterone synthesis, whereas the adrenal-permissive HSD3B1(1245C) allele augments extragonadal dihydrotestosterone synthesis. Retrospective studies have suggested an association between the adrenal-permissive allele, the frequency of which is highest in white men, and early development of castration-resistant prostate cancer (CRPC).

Objective  To examine the association between the adrenal-permissive HSD3B1(1245C) allele and early development of CRPC using prospective data.

Design, Setting, and Participants  The E3805 Chemohormonal Therapy vs Androgen Ablation Randomized Trial for Extensive Disease in Prostate Cancer (CHAARTED) was a large, multicenter, phase 3 trial of castration with or without docetaxel treatment in men with newly diagnosed metastatic prostate cancer. From July 28, 2006, through December 31, 2012, 790 patients underwent randomization, of whom 527 had available DNA samples. In this study, the HSD3B1 germline genotype was retrospectively determined in 475 white men treated in E3805 CHAARTED, and clinical outcomes were analyzed by genotype. Data analysis was performed from July 28, 2006, to October 17, 2018.

Interventions  Men were randomized to castration plus docetaxel, 75 mg/m², every 3 weeks for 6 cycles or castration alone.

Main Outcomes and Measures  Two-year freedom from CRPC and 5-year overall survival, with results stratified by disease volume. Patients were combined across study arms according to genotype to assess the overall outcome associated with genotype. Secondary analyses by treatment arm evaluated whether the docetaxel outcome varied with genotype.

Results  Of 475 white men with DNA samples, 270 patients (56.8%) inherited the adrenal-permissive genotype (≥1 HSD3B1[1245C] allele). Mean (SD) age was 63 (8.7) years. Freedom from CRPC at 2 years was diminished in men with low-volume disease with the adrenal-permissive vs adrenal-restrictive genotype: 51.0% (95% CI, 40.9%-61.2%) vs 70.5% (95% CI, 60.0%-80.9%) (P = .01). Overall survival at 5 years was also worse in men with low-volume disease with the adrenal-permissive genotype: 57.5% (95% CI, 47.4%-67.7%) vs 70.8% (95% CI, 60.3%-81.3%) (P = .03). Hazard ratios were 1.89 (95% CI, 1.13-3.14; P = .02) for CRPC and 1.74 (95% CI, 1.01-3.00; P = .045) for death. There was no association between genotype and outcomes in men with high-volume disease. There was no interaction between genotype and benefit from docetaxel.

Conclusions and Relevance  Inheritance of the adrenal-permissive HSD3B1 genotype is associated with earlier castration resistance and shorter overall survival in men with low-volume metastatic prostate cancer and may help identify men more likely to benefit from escalated androgen receptor axis inhibition beyond gonadal testosterone suppression.

Advanced prostate cancer treatment with androgen deprivation therapy (ADT) via medical or surgical castration depletes circulating gonadal testosterone levels and nearly always produces initial clinical responses. However, castration-resistant prostate cancer (CRPC) eventually develops, typically by revived signaling through the androgen receptor pathway.¹ A major mechanism of androgen receptor restimulation is tumor synthesis of testosterone or dihydrotestosterone from extragonadal precursor steroids, including adrenal dehydroepiandrosterone and its respective sulfate.² The critical role of precursor steroids has been clinically validated by the survival benefit conferred by abiraterone, which blocks extragonadal androgen synthesis by inhibiting 17α-hydroxylase/17,20-lyase.^3,4

The enzyme 3β-hydroxysteroid dehydrogenase-1, encoded by the gene HSD3B1 (OMIM 109715), catalyzes the rate-limiting step in the metabolic conversion from dehydroepiandrosterone to testosterone and dihydrotestosterone in prostatic tissues.⁵ A common missense-encoding germline variant affects steady-state levels of the enzyme and results in a divergence of metabolic phenotypes.^6,7 HSD3B1(1245A) is the adrenal-restrictive allele and encodes for a more rapidly degraded enzyme that limits conversion from dehydroepiandrosterone to dihydrotestosterone. HSD3B1(1245C) is the adrenal-permissive allele and encodes for a stable enzyme that allows for more robust conversion from dehydroepiandrosterone to dihydrotestosterone.⁶ The population frequency of the adrenal-permissive HSD3B1(1245C) allele varies tremendously by race and is disproportionally high in white men (eg, carried in approximately 70% of Italian and Spanish men and only about 9% of Yoruba Nigerian men).⁸ Retrospective studies in at least 7 cohorts of men with nonmetastatic and metastatic prostate cancer suggest that inheritance of the adrenal-permissive HSD3B1(1245C) allele is associated with decrements in meaningful clinical end points, including time to development of CRPC, metastasis-free survival, and overall survival.^9-13 However, to our knowledge, these findings have not been prospectively validated.

The clinical landscape for treatment of advanced prostate cancer has shifted in recent years. One of the notable findings is the survival benefit conferred by redirecting agents previously reserved for treatment of metastatic CRPC and incorporating them into upfront therapy. For example, compared with use of ADT alone, a survival benefit was noted when docetaxel, abiraterone, enzalutamide, or apalutamide was added at the time of ADT initiation.^14-19 However, there is still marked variation in outcomes, reflecting underlying biologic heterogeneity, and the most appropriate treatment for an individual patient is not clear. Accordingly, sound biomarkers based on resistance mechanisms are needed. The E3805 Chemohormonal Therapy vs Androgen Ablation Randomized Trial for Extensive Disease in Prostate Cancer (CHAARTED) randomized men with metastatic castration-sensitive prostate cancer to ADT alone or in combination with docetaxel.¹⁶ Patients were prospectively stratified by a low or high volume of metastatic disease. We genotyped men treated in this trial and hypothesized that accentuated extragonadal dihydrotestosterone synthesis conferred by inheritance of the adrenal-permissive HSD3B1(1245C) allele would be associated with more rapid development of CRPC and lower overall survival for men treated with ADT with or without docetaxel. Given the aforementioned racial variation in germline frequency of the HSD3B1(1245C) allele together with the potential for more genomic heterogeneity in high-volume disease (ie, competing disease drivers may mitigate the effect of the HSD3B1[1245C] allele), we analyzed outcomes in white men enrolled in E3805 CHAARTED and hypothesized that this allele would have the greatest consequences for those with low-volume disease.

The E3805 CHAARTED phase 3 randomized trial was designed in 2005 by the Eastern Cooperative Oncology Group (ECOG; now merged with the American College of Radiology Imaging Network to form ECOG-ACRIN) and was approved by the institutional review board at each participating institution. The study was directed by the ECOG-ACRIN Cancer Research Group. Patients were enrolled by ECOG-ACRIN, the Southwest Oncology Group, the Alliance for Clinical Trials in Oncology, NRG Oncology (a merged group that includes the National Surgical Adjuvant Breast and Bowel Project, the Radiation Therapy Oncology Group, and the Gynecologic Oncology Group), and through the Clinical Trials Support Unit. The ECOG-ACRIN Statistical Center collected the data and was the primary coordinating center. The E3805 CHAARTED trial design and primary analysis have been published¹⁶ and the full study protocol is available in Supplement 1. Written informed consent was obtained from all patients. The present study was approved by the E3805 principal investigator, ECOG, and Cancer Therapy Evaluation Program. For the present report, we obtained updated clinical data and DNA samples to enable further analysis according to HSD3B1 genotype.

Eligible patients had pathologically confirmed prostate cancer or a clinical scenario compatible with prostate cancer with an elevated prostate-specific antigen level, evidence of metastatic disease on imaging, and an ECOG performance status level of 0, 1, or 2 (on a scale from 0 to 5, with higher scores indicating greater impairment; patients with a score of 2 were eligible if the decrement in functional status was secondary to prostate cancer). Patients already receiving ADT were eligible if treatment had been initiated within 120 days of randomization and if there was no evidence of progression. Prior ADT was allowed if administered for adjuvant purposes if the duration of therapy was less than or equal to 24 months and the progression-free interval after such therapy was more than 12 months. Patients were required to have adequate organ function to permit treatment with docetaxel (Supplement 1). We determined germline HSD3B1 genotype from DNA extracted from blood samples collected as part of the trial using previously described methods.⁹ Investigators blinded to clinical data performed the genotyping.

Patients were randomized in a 1:1 ratio to either ADT alone or with docetaxel. The method of permuted blocks was used for assignment. Neither patients nor enrolling physicians were blinded to treatment allocation.

All patients received ADT via surgical castration (orchiectomy) or medical castration with either a luteinizing hormone-releasing hormone agonist or antagonist with or without an antiandrogen (bicalutamide or flutamide). Patients were randomized to receive ADT plus docetaxel at a dose of 75 mg/m² every 3 weeks for 6 cycles or ADT alone. Details regarding supportive medications and dose modifications are available in Supplement 1. Patients were stratified by ECOG performance status (0-1 vs 2), age (<70 vs ≥70 years), planned use of combined androgen blockade for more than 30 days (yes vs no), planned use of zoledronic acid or denosumab for prevention of skeletal-related events (yes vs no), duration of any prior adjuvant ADT (<12 vs ≥12 months), and the extent of metastatic disease (high volume [defined as the presence of visceral metastases or ≥4 bone metastases with ≥1 lesion beyond the pelvis and vertebral bodies] vs low volume).

Patients assigned to receive ADT plus docetaxel were seen every 3 weeks during the period they were receiving docetaxel, after which they were seen every 3 months. Patients assigned to ADT alone were seen every 3 months. Prostate-specific antigen levels were assessed at each visit. Imaging consisted of computed tomography of the abdomen and pelvis, technetium-99m bone scanning, and computed tomographic or radiographic imaging of the chest. Patients underwent imaging at baseline and at the time of castration resistance or as clinically indicated. Radiographic disease progression was determined by the Response Evaluation Criteria in Solid Tumors (RECIST), version 1.0.²⁰ Serologic progression was defined as an increase in the prostate-specific antigen level of more than 50% above the nadir reached after ADT began, with 2 consecutive elevations at least 2 weeks apart. The date of a first recorded increase of more than 50% above the nadir was documented as the date of progression. If the nadir was less than 2 ng/mL, an increase of more than 2 ng/mL was required.

We analyzed clinical outcomes in white patients enrolled in E3805 CHAARTED given the previously established racial variation in HSD3B1(1245C) allelic frequency, with a much higher prevalence in white individuals. We analyzed 2-year freedom from CRPC (an early end point in metastatic prostate cancer) and 5-year overall survival (primary end point). Overall survival included death from any cause. Castration-resistant prostate cancer was established by clinical or serologic progression with a testosterone level less than 50 ng/dL (to convert to nanomoles per liter, multiply by 0.0347) or source documentation of medical or surgical castration.

For the main analysis, patients were combined across study arms according to genotype to assess the outcomes associated with the HSD3B1 genotype when treated with ADT with or without docetaxel. Secondary analyses were conducted according to study arm to investigate whether the outcome of docetaxel varied with genotype. Time-to-event end points were determined from the time of randomization and were evaluated using Kaplan-Meier methods and Cox proportional hazards regression models. Patients were censored if they were lost to follow-up or had not experienced the event by the specified point. Clinical outcomes were compared according to HSD3B1 genotype using log-rank and Wilcoxon tests. Demographic and treatment characteristics were compared across genotypes to assess for confounders using Kruskal-Wallis analysis of variance, Pearson χ² test, and Fisher exact test. Analyses were stratified by volume of disease (high volume vs low volume). Multivariable Cox proportional hazards regression models were used to assess the outcomes associated with HSD3B1 genotype and account for known prognostic factors, including age, baseline prostate-specific antigen level, Gleason Score, treatment arm (ADT with or without docetaxel), and ECOG performance status. All tests were 2-sided, and P values ≤.05 were interpreted as statistically significant. Data analysis was conducted from July 28, 2006, to October 17, 2018. Analyses were performed with the use of SAS, version 9.4 (SAS Institute Inc).

From July 28, 2006, through December 31, 2012, 790 men underwent randomization. Ten patients were ineligible, 7 had incomplete information to assess eligibility, and 6 patients in the combination group did not start the assigned therapy (Figure 1).²¹ All patients were followed up and analyzed according to their assigned group. A total of 527 patients had available DNA and consented for inclusion in germline genetic studies, enabling determination of HSD3B1 genotype. The adrenal-restrictive genotype was defined as no HSD3B1(1245C) alleles (ie, only the HSD3B1[1245A] allele was present), whereas the adrenal-permissive genotype was defined as greater than or equal to 1 HSD3B1(1245C) allele. The adrenal-permissive genotype was present in 270 of 475 white men (56.8%) compared with 7 of 52 nonwhite men (13.5%) (P < .001). To isolate the effect of the HSD3B1 genotype and eliminate potential confounders from other unmeasured genomic factors that may vary by race, we focused subsequent analyses on white patients. Although the main analyses in this study focus on men with at least 1 HSD3B1(1245C) allele compared with those with none, additional data are available regarding the frequency with which men specifically inherited 0 vs 1 vs 2 HSD3B1(1245C) alleles (eTable 1 in Supplement 2) and their clinical outcomes (eFigure 1 in Supplement 2).

Clinical and treatment characteristics are reported for men with low-volume (Table 1) and high-volume (Table 2) disease. Mean (SD) age of the overall population was 63 (8.7) years. There were no statistically significant differences in demographic or treatment characteristics by genotype in men with low-volume disease. In those with high-volume disease, ECOG performance status was lower (better) in men with the adrenal-restrictive genotype. The median length of follow-up was 64.4 (interquartile range [IQR], 43.4-82.6) months for the low-volume group (80.4 [IQR, 68.8-84.5] months for surviving patients) and 42.6 (IQR, 23.3-68.9) months for the high-volume group (71.8 [IQR, 59.7-84.8] months for surviving patients).

Freedom from CRPC at 2 years was significantly lower in men with low-volume disease (Figure 2A) with the adrenal-permissive genotype vs the adrenal-restrictive genotype: 51.0% (95% CI, 40.9%-61.2%) vs 70.5% (95% CI, 60.0%-80.9%) (P = .01). The adrenal-permissive genotype remained associated with a higher risk of CRPC after multivariable analysis adjusting for known prognostic factors, with a hazard ratio of 1.89 (95% CI, 1.13-3.14; P = .02) (eTable 2 in Supplement 2). Although death prior to development of CRPC was rare (1/174 men), competing risk analyses using cumulative incidence functions were also performed and corroborated these results (eFigure 2 in Supplement 2). In men with high-volume disease (Figure 2B), there was no significant difference in freedom from CRPC at 2 years based on adrenal-permissive vs adrenal-restrictive genotype: 26.6% (95% CI, 19.7%-33.4%) vs 27.3% (95% CI, 19.4%-35.3%) (P = .89). The corresponding hazard ratio was 1.10 (95% CI, 0.82-1.47; P = .52). There was no evidence of an interaction between treatment arm and genotype for patients with either low-volume or high-volume disease (Figure 3A and B). Patients with low-volume disease did not significantly benefit from docetaxel, whereas those with high-volume disease did regardless of HSD3B1 genotype.

Overall survival at 5 years was significantly worse in men with low-volume disease (Figure 2C) who had the adrenal-permissive genotype vs those with the adrenal-restrictive genotype: 57.5% (95% CI, 47.4%-67.7%) vs 70.8% (95% CI, 60.3%-81.3%) (P = .03). The adrenal-permissive genotype remained associated with a higher risk of death after multivariable analysis adjusting for known prognostic factors, with a hazard ratio of 1.74 (95% CI, 1.01-3.00; P = .045) (eTable 3 in Supplement 2). In men with high-volume disease (Figure 2D), there was no significant difference in overall survival at 5 years according to adrenal-permissive vs adrenal-restrictive genotype: 37.3% (95% CI, 29.8%-44.7%) vs 33.1% (95% CI, 24.7%-41.4%) (P = .65). The corresponding hazard ratio was 0.89 (95% CI, 0.65-1.22; P = .48). There was no evidence of an interaction between treatment arm and genotype for patients with either low-volume or high-volume disease (Figure 3C and Figure 3D). Men with low-volume disease did not benefit from docetaxel therapy, whereas men with high-volume disease did regardless of HSD3B1 genotype.

This expanded analysis of a large randomized trial appears to demonstrate that the adrenal-permissive genotype (ie, inheritance of at least 1 copy of the HSD3B1(1245C) allele that enhances dihydrotestosterone synthesis from adrenal precursor steroids) is associated with early development of CRPC and worse overall survival in white patients with low-volume metastatic prostate cancer treated with ADT regardless of the use of docetaxel. Inheritance of the adrenal-permissive allele affected overall survival despite any subsequent therapies received after the development of CRPC. To our knowledge, these results represent the first prospective apparent support of the importance of the HSD3B1 genotype with respect to clinical outcomes, and they are concordant with the findings of several retrospective cohorts^9-13 as well as the preclinical findings that led to this clinical evaluation.⁶

In contrast to the findings in men with low-volume disease, we found that the HSD3B1 genotype was not associated with clinical end points in men with high-volume disease. It is well documented that men with high-volume disease have clinical courses distinct from those with low-volume disease.²² There are also increasing genomic alterations across the spectrum of prostate cancer from locoregional disease to metastatic castration-sensitive or castration-naive disease to metastatic CRPC.^23,24 It is plausible that high-volume disease may be further along this spectrum than low-volume disease and that competing molecular mechanisms may attenuate the role of HSD3B1-mediated biologic factors. There may be less reliance on extragonadally driven androgen receptor stimulation and more access to alternative pathways,²⁵ which might override the advantage afforded to tumor cells possessing the HSD3B1(1245C) allele. Two recent trials demonstrating improvement in overall survival with the addition of the potent androgen receptor antagonists enzalutamide or apalutamide to ADT in metastatic castration-sensitive prostate cancer support the possibility of a different degree of dependence on androgen receptor signaling inasmuch as the hazard ratios for progression end points and death with the addition of enzalutamide or apalutamide were lower (effect size was larger) for men with low-volume disease.^18,19

Although our primary objective was to evaluate the association between HSD3B1 genotype and clinical outcomes in men treated with ADT regardless of the use of docetaxel, we also investigated whether the benefit observed with docetaxel in the E3805 CHAARTED trial varied with genotype. We found that the HSD3B1 genotype did not appear to be predictive of differential benefit with docetaxel. Regardless of genotype, men with high-volume disease achieved better outcomes when docetaxel was added to ADT, whereas outcomes in men with low-volume disease varied only by genotype, with no clear benefit from docetaxel. This finding suggests the possibility that men with low-volume metastatic prostate cancer and the adrenal-permissive genotype may require alternative treatment strategies, particularly with agents that inhibit extragonadal androgen synthesis or the androgen receptor ligand-binding domain.²⁶ However, the enzyme encoded by HSD3B1 regulates biochemically active steroidal metabolites of abiraterone,^27,28 and the HSD3B1(1245C) allele is associated with the generation of a specific metabolite, 5α-abiraterone, that has partial androgen receptor agonist activity. Further studies are warranted to assess whether potent androgen receptor antagonists, which have proven efficacy in metastatic castration-sensitive prostate cancer^18,19 as well as metastatic and nonmetastatic CRPC,^29-32 might directly overcome the tumor growth advantage imparted by the HSD3B1(1245C) allele and yield particular benefit for patients with low-volume disease who have the adrenal-permissive genotype.

This study has limitations. One limitation of this work is that our analysis was restricted to white men enrolled in E3805 CHAARTED. However, white patients accounted for 85% of all trial participants and 90% of those who had DNA available for genotyping. Moreover, by focusing the analysis on white patients, we sought to reduce potential confounders that might mask the effect of the HSD3B1 genotype.

Taken together, our findings suggest that the HSD3B1 genotype can be used to risk stratify white men with low-volume metastatic prostate cancer. Those with the adrenal-permissive genotype have a worse prognosis inasmuch as they develop CRPC sooner and have shorter overall survival than men with the adrenal-restrictive genotype. This information could assist clinicians in counseling patients and guide researchers in identifying those for whom escalated androgen receptor axis inhibition beyond mere gonadal testosterone suppression is most warranted. Accordingly, HSD3B1 genotype could be used as a stratification factor for patients with low-volume disease in future clinical trials.

Accepted for Publication: November 20, 2019.

Corresponding Author: Nima Sharifi, MD, GU Malignancies Research Center, Cleveland Clinic, 9500 Euclid Ave, Cleveland, OH 44195 (sharifn@ccf.org).

Published Online: February 13, 2020. doi:10.1001/jamaoncol.2019.6496

Author Contributions: Drs Hearn and Sharifi had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Hearn, Sweeney, Li, Jarrard, Sharifi.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Hearn, Sweeney, Li, Hobbs, Jarrard, Carducci, Sharifi.

Critical revision of the manuscript for important intellectual content: Hearn, Sweeney, Almassi, Reichard, Reddy, Li, Hobbs, Jarrard, Chen, Dreicer, Garcia, Carducci, DiPaola, Sharifi.

Statistical analysis: Almassi, Reddy, Li, Hobbs.

Obtained funding: Hearn, Sweeney, Sharifi.

Administrative, technical, or material support: Hearn, Sweeney, Jarrard, Dreicer, Carducci, DiPaola, Sharifi.

Supervision: Hearn, Jarrard, Garcia, Dipaola, Sharifi.

Conflict of Interest Disclosures: Dr Hearn reported receiving grants from the US Department of Defense, the Prostate Cancer Foundation, and the Conquer Cancer Foundation of the American Society of Clinical Oncology (ASCO) Merit Award during the conduct of the study. Dr Sweeney reported receiving grants from Sanofi during the conduct of the study and grants and personal fees from Astellas, Bayer, Sanofi, Janssen, and Dendreon outside the submitted work. In addition, Dr Sweeney had a patent to parthenolide issued and a patent to dimethylaminoparthenolide issued and licensed. Dr Hobbs reported receiving grants from Amgen outside the submitted work and serving as a paid consultant for SimulStat and scientific advisor and owner of Presagia. Dr Dreicer reported receiving personal fees from Janssen, Pfizer, Orion, Vizuri, and Novartis outside the submitted work. Dr Garcia reported receiving grants from Astellas, Clovis, Genentech, Merck, and Janssen and personal fees from Bayer, Clovis, Sanofi, AAA, Janssen, and Merck outside the submitted work. Dr Sharifi reported receiving personal fees from Janssen and Pfizer outside the submitted work. In addition, Dr Sharifi had a patent to HSD3B1 issued. No other disclosures were reported.

Funding/Support: This study was coordinated by the Eastern Cooperative Oncology Group–American College of Radiology Imaging Network (ECOG-ACRIN) Cancer Research Group (Peter J. O'Dwyer, MD, and Mitchell D. Schnall, MD, PhD, group cochairs) and supported by the National Cancer Institute of the National Institutes of Health under the following award numbers: U10CA180820, U10CA180794, UG1CA233160, UG1CA233180, UG1CA233234, UG1CA233270, U10CA180888, and UG1CA233196. This work was also supported by a grant from the US Department of Defense Congressionally Directed Medical Research Programs (Dr Hearn), a Conquer Cancer Foundation of ASCO Merit Award (Dr Hearn), a grant from the Prostate Cancer Foundation (Drs Sharifi and Hearn), and additional grants from the National Cancer Institute (R01CA172382, R01CA190289, R01CA236780, and R01CA168899 [Dr Sharifi]).

Role of the Funder/Sponsor: Sanofi donated docetaxel, provided funding to ECOG-ACRIN and provided a grant for sample collection (Dr Sweeney). Sanofi otherwise had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health, nor does mention of trade names, commercial products, or organizations imply endorsement by the US government.

Meeting Presentation: This study was presented, in part, at the ASCO annual meeting; Chicago, Illinois; June 1, 2019.