Phase II Study of Gemcitabine and Split-Dose Cisplatin Plus Pembrolizumab as Neoadjuvant Therapy Before Radical Cystectomy in Patients With Muscle-Invasive Bladder Cancer

PURPOSE

To evaluate the safety and efficacy of gemcitabine and cisplatin in combination with the immune checkpoint inhibitor pembrolizumab as neoadjuvant therapy before radical cystectomy (RC) in muscle-invasive bladder cancer.

METHODS

Patients with clinical T2-4aN0/XM0 muscle-invasive bladder cancer eligible for RC were enrolled. The initial six patients received lead-in pembrolizumab 200 mg once 2 weeks prior to pembrolizumab 200 mg once on day 1, cisplatin 70 mg/m² once on day 1, and gemcitabine 1,000 mg/m² once on days 1 and 8 every 21 days for four cycles. This schedule was discontinued for toxicity and subsequent patients received cisplatin 35 mg/m² once on days 1 and 8 without lead-in pembrolizumab. The primary end point was pathologic downstaging (< pT2N0) with null and alternative hypothesis rates of 35% and 55%, respectively. Secondary end points were toxicity including patient-reported outcomes, complete pathologic response (pT0N0), event-free survival, and overall survival. Association of pathologic downstaging with programmed cell death ligand 1 staining was explored.

RESULTS

Thirty-nine patients were enrolled between June 2016 and March 2020 (72% cT2, 23% cT3, and 5% cT4a). Patients received a median of four cycles of therapy. All patients underwent RC except one who declined. Twenty-two of 39 patients (56% [95% CI, 40 to 72]) achieved < pT2N0 and 14 of 39 (36% [95% CI, 21 to 53]) achieved pT0N0. Most common adverse events (AEs) of any grade were thrombocytopenia (74%), anemia (69%), neutropenia (67%), and hypomagnesemia (67%). One patient had new-onset type 1 diabetes mellitus with ketoacidosis related to pembrolizumab and no patients required steroids for immune-related AEs. Clinicians consistently under-reported AEs when compared with patients.

CONCLUSION

Neoadjuvant gemcitabine and cisplatin plus pembrolizumab met its primary end point for improved pathologic downstaging and was generally safe. A global study of perioperative chemotherapy plus pembrolizumab or placebo is ongoing.

INTRODUCTION

Patients with muscle-invasive bladder cancer (MIBC) have suboptimal outcomes after definitive management. Standard of care for MIBC includes neoadjuvant cisplatin-based chemotherapy followed by radical cystectomy (RC). Despite this aggressive treatment, more than 50% of patients develop recurrent disease within 2 years and die of complications related to bladder cancer.¹

CONTEXT

• Key Objective

• Muscle-invasive bladder cancer (MIBC) has poor outcomes despite aggressive treatment with cisplatin-based neoadjuvant chemotherapy followed by radical cystectomy (RC). This phase II trial examines the addition of the immune checkpoint inhibitor pembrolizumab to gemcitabine and split-dose cisplatin as neoadjuvant therapy before RC.

• Knowledge Generated

• Neoadjuvant pembrolizumab plus gemcitabine and split-dose cisplatin had a high rate of pathologic downstaging (56%) in MIBC, meeting its primary end point for improvement over historical controls. Toxicity was manageable and there was no association of pathologic downstaging with programmed cell death ligand 1 status.

• Relevance

• Pembrolizumab plus gemcitabine and cisplatin is promising as neoadjuvant therapy before RC in MIBC. Randomized studies to confirm efficacy compared with gemcitabine and cisplatin alone are ongoing.

Randomized trials of neoadjuvant chemotherapy have used cisplatin-based regimens, including methotrexate, vinblastine, doxorubicin, and cisplatin.² Compared with methotrexate, vinblastine, doxorubicin, and cisplatin, the combination of gemcitabine and cisplatin (GC) has similar efficacy with fewer side effects in the metastatic setting³ and similar pathologic downstaging rates in the neoadjuvant setting.⁴ GC has become a widely used neoadjuvant regimen for MIBC. More recently, immune checkpoint inhibitors (ICIs) have been Food and Drug Administration–approved for advanced or metastatic urothelial cancer (UC), and maintenance ICI has been approved after first-line platinum-based chemotherapy in metastatic disease, suggesting earlier introduction of ICI improves survival in advanced disease.^5,6

Regimens that combine platinum-based chemotherapy and ICI have been tested in randomized phase III trials in advanced or metastatic UC. Combination treatment resulted in slightly higher complete response rates than chemotherapy alone, but neither the Keynote-361 nor the IMvigor130 studies met their primary end point for survival outcomes, despite a modest improvement in progression-free survival in IMvigor130.^7,8 In the MIBC setting, higher response rates with combination approaches may be of greater importance than in the metastatic setting since pathologic downstaging (< pT2) is associated with improvement in long-term survival in patients with MIBC receiving cisplatin-based neoadjuvant chemotherapy.²

The current phase II study evaluated GC in combination with the anti–programmed death-1 ICI pembrolizumab as neoadjuvant therapy before RC in patients with MIBC.

METHODS

Eligibility

Patients with clinical stage T2-T4a N0/X M0 were enrolled. Eligible patients had histologically confirmed UC; mixed histology, including a component of UC, were eligible. Patients were medically appropriate to undergo RC and eligible for cisplatin (serum creatinine ≤ 1.5 times upper limit of normal or estimated glomerular filtration rate ≥ 60 mL/min). Patients had Eastern Cooperative Oncology Group performance status of 0 or 1, no prior systemic chemotherapy for UC, and no hearing impairment greater than grade 2. Patients were excluded if they had a diagnosis of immunodeficiency, required systemic steroids or other immunosuppressive therapy, or had active autoimmune disease requiring systemic treatment in the past 2 years.

Study Design

This was a multicenter, phase II, open-label, single-arm study evaluating pembrolizumab in combination with GC as neoadjuvant therapy before RC in patients with MIBC (LCCC1520). The primary end point was pathologic downstaging (< pT2N0M0) at RC. Secondary end points included pathologic complete response (pCR) (pT0N0M0), event-free survival (defined as day 1 of neoadjuvant treatment to progression, recurrence after cystectomy, or death from any cause), overall survival (defined as day 1 of treatment until death from any cause), and safety or toxicity. Recurrence-free survival (RFS) defined as time from cystectomy to progression or death from any cause is also reported.

The study was reviewed and approved by the institutional review boards at each participating institution and conducted in accordance with ethical principles per the Declaration of Helsinki. The trial was registered on ClinicalTrials.gov (NCT02690558), and all participants provided written informed consent.

Procedures and Treatments

All patients underwent diagnostic transurethral resection of the bladder tumor before enrollment to confirm ≥ T2 UC. Archival formalin-fixed paraffin-embedded tumor tissue and peripheral blood was collected. All patients had imaging with computerized tomography or magnetic resonance imaging of the abdomen and pelvis, and computerized tomography of the chest.

The first six patients received a single dose of pembrolizumab 200 mg intravenous (IV) 2 weeks before combination pembrolizumab 200 mg IV once on day 1, cisplatin 70 mg/m² once on day 1, and gemcitabine 1,000 mg/m² once on day 1 and day 8 every 21 days for four cycles. After the initial six patients, the Protocol (online only) was amended because of poor tolerability with removal of the lead-in pembrolizumab and change to pembrolizumab 200 mg IV once on day 1, split-dose cisplatin 35 mg/m² once on days 1 and 8, and gemcitabine 1,000 mg/m² once on day 1 and 8 every 21 days for four cycles. A diagram of enrolled patients is included in Appendix Figure A1 (online only).

Use of growth factor support was permitted per institutional practice or required based on hematologic toxicity. Dexamethasone 8 mg was administered before cisplatin on days 1 and 8 with option to continue on days 2, 3, 9, and 10. Cisplatin was administered if estimated glomerular filtration rate ≥ 50 mL/min (per Chronic Kidney Disease Epidemiology Collaboration [CKD-EPI] equation).⁹ If either cisplatin or gemcitabine was held or delayed for adverse events (AEs), pembrolizumab was also held until treatment was resumed. If pembrolizumab was held because of immune-related AEs, GC was continued.

RC with lymph node dissection was completed within 4-8 weeks of chemotherapy, when possible. All patients underwent a standard template bilateral lymphadenectomy including obturator, internal iliac, and external iliac lymph nodes. Fresh frozen and formalin-fixed paraffin-embedded tumor tissue and peripheral blood were collected at the time of RC.

Clinical Assessment

Repeat imaging was performed after neoadjuvant therapy before RC. All patients who received any protocol treatment were evaluable for toxicity. Toxicity was assessed via National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE v4.03). Toxicity was monitored continuously with Pocock-type sequential boundaries.

Patient-reported outcomes (PROs) were assessed using relevant PRO-CTCAE items (v1.0) for toxicity monitoring of 11 selected AEs potentially associated with therapy. For each AE, up to three items were scored for frequency, severity, and interference with daily activities, and assigned a composite PRO-CTCAE score to compare with clinician-based CTCAE assessments.¹⁰

Correlative Assessments

Programmed cell death ligand 1 (PD-L1) expression was scored by a board-certified pathologist at QualTek Molecular Laboratories (Santa Barbara, CA) using the PD-L1 IHC 22C3 pharmDx (Agilent Technologies, Santa Clara, CA) assay and reported as a modified proportion score (MPS) defined as the proportion of cells within tumor nests (including tumor cells and intercalated mononuclear inflammatory cells) with membranous staining of PD-L1 divided by the total number of viable tumor cells × 100. The MPS excludes PD-L1 expression of distal immune cells and therefore either estimates or underestimates the companion diagnostic assay Combined Positive Score.¹¹ A cutoff of ≥ 10 was considered positive to best approximate the Combined Positive Score cutoff used across UC studies.

Statistical Analysis

This study enrolled 39 patients with a primary end point of pathologic downstaging at cystectomy. The null pathologic downstaging rate of 35% (based on prior studies of neoadjuvant chemotherapy^1,12,13) was tested against a one-sided alternative using a Simon's two-stage design. In the first stage, 21 patients were accrued. If there were > 8 patients with pathologic downstaging, the study would proceed to the second stage with 18 additional patients. The null hypothesis was to be rejected and the regimen considered worthy of further investigation if ≥ 19 of the 39 patients had pathologic downstaging. This design yields a type I error rate of 0.05 and power of 80% when the true pathologic downstaging rate is 55%, representing a 20% improvement in pathologic downstaging rate. The study was monitored regularly by the Institutional Data Safety and Monitoring Committee at the University of North Carolina. All statistical analyses were performed using SAS v9.4 and R v3.6.0.

RESULTS

Patient Characteristics

Thirty-nine patients were enrolled between June 2016 and March 2020. Baseline patient characteristics are listed in Table 1.

TABLE 1.

Patient Demographics and Clinical Characteristics (N = 39)

Efficacy

All 39 patients were evaluable for the primary end point. Twenty-two of 39 patients had < pT2N0M0 at cystectomy with an overall pathologic downstaging rate of 56% (95% CI, 40 to 72). The pCR rate was 36% (14 of 39 patients; 95% CI, 21 to 53). One patient with clinical T3b MIBC at study enrollment was diagnosed with metastatic disease and 10 patients (26%) had node-positive disease at the time of cystectomy. One of six patients who received the pembrolizumab lead-in had pathologic downstaging. Nineteen of 27 patients who received all four cycles of treatment had pathologic downstaging (70%). The rate of pathologic downstaging of patients with clinical T3 or T4 disease was 45% (5 of 11 patients). Pathologic response rates are summarized in Figure 1 and Appendix Table A1 (online only).

FIG 1.

(A) Pathologic response and (B) pathologic downstaging at the time of cystectomy (pathologic stage compared with pretreatment clinical stage).

No patients had clinical or radiographic progression before RC. Thirty-eight patients underwent cystectomy. One patient declined cystectomy because of personal preference and was counted as a nonresponder for the primary end point. Median time to cystectomy was 46 days after completion of protocol therapy (range, 13-154 days) and was completed within 12 weeks in all but two patients (one because of an infected laceration unrelated to treatment, and one because of a treatment-related arterial thromboembolism to the lower extremity). Two patients (5%) had margins positive for invasive UC and two had margins positive for noninvasive UC. The median number of lymph nodes removed was 14 (range, 4-36).

With a median follow-up of 15.7 months (interquartile range, 11.5-21.9 months), eight patients (21%) have relapsed. Six of these eight patients had node-positive disease at the time of cystectomy. Nine patients (23%) have died while on follow-up: seven deaths were disease-related and no deaths were treatment-related. Median event-free survival, RFS, and overall survival have not been reached, with 1 year rates of 89%, 75%, and 91%, respectively. Patients with pathologic downstaging had significantly longer RFS compared to those without pathologic downstaging (median not reached v 10.9 months, hazard ratio 0.13 [95% CI, 0.03 to 0.59], log-rank P = .009; Fig 2).

FIG 2.

Recurrence-free survival after cystectomy for (A) all patients and (B) stratified by pathologic response.

Dose Reductions and Modifications and Discontinuation

Six patients received lead-in pembrolizumab followed by full-dose cisplatin and gemcitabine in combination with pembrolizumab before Protocol amendment. Four of the six patients in this cohort stopped treatment early because of AEs including grade 3 or 4 neutropenia in four patients, grade 4 thrombocytopenia in three patients, and febrile neutropenia in three patients. Two of these patients had a negative workup for hemolysis, no patients required bone marrow biopsy, and all recovered counts in a period consistent with chemotherapy toxicity. All four patients who came off treatment early had worsening renal function felt to be related to cisplatin that precluded further protocol therapy. Thirty-three subsequent patients were enrolled and treated without the lead-in pembrolizumab dose and with split-dose cisplatin. Of those, 25 patients (76%) completed four cycles of neoadjuvant treatment. Overall, nine patients (23%) had dose reductions because of treatment-related AEs.

Ten of the 12 patients who came off treatment early did so because of treatment-related AEs. Six were because of acute kidney injury attributed to cisplatin, two were because of pancytopenia (one in the setting of sepsis and reduced ejection fraction in a patient with underlying heart disease that was determined to be not immune-related), one because of an arterial thromboembolism to the lower extremity, and one related to new-onset diabetic ketoacidosis secondary to pembrolizumab. Pembrolizumab was otherwise never held for an immune-related AE in any patient. Two patients discontinued protocol treatment early because of reasons felt unrelated to protocol treatment: one patient had an infected leg laceration after a fall and one patient withdrew consent after cycle 3 but completed the fourth cycle of neoadjuvant GC and cystectomy off protocol (patient included in primary end point per Protocol with informed consent for tissue and outcomes data).

Adverse Events

All 39 patients were evaluable for toxicity. All patients experienced at least one grade 1 AE, and 74% of patients experienced grade 3 or higher treatment-related AEs (Table 2). The most common AEs of any grade were hematologic, including thrombocytopenia (74%), anemia (69%), and neutropenia (67%). One patient had new-onset type 1 diabetes mellitus with ketoacidosis related to pembrolizumab, and no patients required steroids for immune-related AEs. No other grade 3-4 AEs related to pembrolizumab were reported (Appendix Table A2, online only). One patient developed myelodysplastic syndrome 1 year after cystectomy, thought to be possibly treatment-related. One patient had a cardiac arrest in the setting of a pulmonary embolus and myocardial infarction approximately 1 month after cystectomy that was not thought to be treatment related, and the patient recovered. There were no treatment-related deaths.

TABLE 2.

Treatment-Related Adverse Events in > 10% of All Participants and All Grade 3-4 AEs (worst grade per patient)

Patients reported the following symptoms as PRO-CTCAE composite grade 3 or higher: fatigue (31%), aching muscles (18%), pain (15%), nausea (13%), and shortness of breath (13%) (Appendix Table A3, online only). Clinicians reported a lower grade of AEs compared with patients 62% of the time, and only over-reported AE grading 1% of the time (Fig 3). Longitudinal trends in composite scores of nausea and fatigue (two of the more common AEs) for the duration of protocol therapy are in Appendix Figure A2 (online only).

FIG 3.

Direct comparison of clinician-reported and patient-reported AEs. CTCAE clinician-reported maximum grade and PRO-CTCAE patient-reported maximum composite score were compared for each AE for every patient. The difference between maximum AE grade reported by clinicians and patients was calculated. The number of patient-AE pairs for each possible difference value are shown for each AE, demonstrating a consistent under-reporting of AE grade by clinicians. AE, adverse event; CTCAE, Common Terminology Criteria for Adverse Events; PRO-CTCAE, Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events.

PD-L1 Immunohistochemistry

Eighteen out of 33 patients with available baseline tumor tissue were positive for PD-L1 per the MPS ≥ 10. The rate of pathologic downstaging was not significantly different between PD-L1–positive and –negative patients (67% v 47%, P = .25).

DISCUSSION

This phase II trial of neoadjuvant pembrolizumab plus cisplatin and gemcitabine met its primary end point with a pathologic downstaging rate of 56% in patients with MIBC. This rate of pathologic downstaging is consistent with prior studies of neoadjuvant chemotherapy-ICI combinations in MIBC, including the HCRN GU14-188, SAK 06/17, and BLASST-1 studies, which reported pathologic downstaging rates of 50%-66% for pembrolizumab, durvalumab, and nivolumab, respectively, in combination with GC.^14-16 Our study is the first to demonstrate a similarly robust rate of pathologic downstaging in the setting of split-dose cisplatin, a regimen that is more tolerable and accessible to patients with limited renal function.¹⁷

The 56% rate of pathologic downstaging in the current study is higher than with neoadjuvant GC alone with reported rates of pathologic downstaging in trials of approximately 45%-50%.^4,18 There has been a lack of clear benefit to combination ICI-chemotherapy in metastatic UC; however, increased pathologic downstaging may lead to better outcomes in the neoadjuvant setting with micrometastatic disease.⁷ This may be similar to the improvement in survival seen with maintenance ICI in metastatic UC.⁵

The 36% pCR rate in our study is in line with prior reports of neoadjuvant GC in MIBC (30% and 36% in the COXEN and VESPER trials, respectively) and with other neoadjuvant chemotherapy-ICI studies (pCR rate of 30%-44%).^4,14-16,18 The use of pathologic downstaging in addition to pCR as an intermediate end point in MIBC has been well validated after neoadjuvant chemotherapy, and pathologic downstaging is predictive of disease-free survival in this setting.^2,19,20 Some studies have suggested that pathologic downstaging is a similar, if not a superior, predictor of long-term survival compared with pCR in MIBC.²¹

Additionally, the pCR rate in our study of 36% suggests that more than a third of patients may not require cystectomy after receiving chemotherapy-ICI, but more work is needed to identify the patients who can safely avoid cystectomy. Several ongoing trials investigate a bladder-sparing approach after a clinical complete response to neoadjuvant treatment and include molecular testing for the presence of defects in DNA repair genes to optimally select patients, including the RETAIN study, which reported an 89% bladder-intact survival for the active surveillance group.²² Other studies also use this approach with the neoadjuvant regimens of dose-dense GC (NCT03609216) and combination chemotherapy-ICI (NCT03558087, NCT04506554). Biomarkers will be critical to select patients who may be able to avoid RC, and biomarker analysis is ongoing in our study. Pathologic downstaging did not correlate with PD-L1 positivity by the MPS in the current study, which needs validation in larger, randomized studies; however, it is unlikely that one single biomarker will comprehensively predict benefit.

Several studies have investigated a short course of ICI monotherapy before RC, including the PURE-01 study with a pCR rate of 42% after three cycles of pembrolizumab and the ABACUS study with a pCR rate of 31% after two cycles of atezolizumab.^23,24 PURE-01 also demonstrated longer RFS in patients with pCR than in non-pCR patients. Our data similarly validate pathologic downstaging as predictive of RFS and demonstrate that patients who do not have pathologic downstaging have poor outcomes with a median RFS of only 10.9 months. The 26% rate of lymph node positivity in our study is slightly higher than that in prior studies of GC alone (18% in the VESPER trial) and underscores the high risk and aggressive pathology of the patients in this single-arm study using split-dose cisplatin. It remains unclear which patients benefit from chemotherapy, ICI, or both in the neoadjuvant setting, especially given the similar pCR rates across single-arm studies including those using ICI alone. Adjuvant ICI may provide a similar benefit to patients, given the recent report of an improvement in disease-free survival in the Checkmate-274 trial,²⁵ again highlighting the critical importance of biomarker-based patient selection.

Overall, the clinician-reported toxicity seen in our study was similar to other neoadjuvant studies.¹⁸ Our rate of 41% grade 3-4 neutropenia and 33% grade 3-4 thrombocytopenia is on the higher end of rates previously reported in other studies of neoadjuvant GC (20%-46% and 11%-17%, respectively).^4,18 Our Protocol initially included a lead-in dose of pembrolizumab designed to allow for the expansion of effector T cells without the cytotoxic effects of chemotherapy. However, based on the hypothesis that the rapid expansion of immune cells followed by cytotoxic chemotherapy was leading to increased myelosuppression, the lead-in dose was omitted. Subsequently, other studies have been reported with lead-in ICI before combination chemotherapy-ICI regimens, including a randomized study in metastatic UC, suggesting better tolerability of lead-in dosing than demonstrated in our study.^26,27 Overall, AEs related to pembrolizumab in our study were low grade and consistent with prior studies. Other studies of chemotherapy-ICI have demonstrated preserved efficacy with routine steroid use for nausea prophylaxis,²⁸ but we cannot rule out the fact that our low rate of immune-related AEs was related to routine steroid use in our study.

The need for incorporation of PROs in toxicity reporting in clinical trials is highlighted in our study, given the 62% rate of under-reporting of subjective toxicities by clinicians compared with patients. This rate of under-reporting is consistent with prior studies, which report rates of 41%-74%, and we extend this work to the treatment of MIBC.^29,30 Ongoing randomized studies should include PROs to comprehensively assess the impact of combination therapies on AE rates.

In summary, pembrolizumab plus gemcitabine and split-dose cisplatin demonstrates promising activity in the neoadjuvant treatment of MIBC. A randomized phase III trial to evaluate perioperative pembrolizumab plus GC is ongoing to validate our findings (Keynote-866; NCT03924856).

APPENDIX

FIG A1.

Flow diagram of the trial. AE, adverse event; IV, intravenous.

FIG A2.

Distribution of composite PRO-CTCAE scores at each protocol timepoint for (A) fatigue and (B) nausea. PRO-CTCAE, Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events.

TABLE A1.

Pathologic Response at Radical Cystectomy

TABLE A2.

Pembrolizumab-Related AEs (ie, attributed as probably or definitely related to pembrolizumab per investigator assessment)

TABLE A3.

Patient-Reported AEs and Maximum PRO-CTCAE Composite Score for Each PRO-CTCAE Symptom (n = 39)

DATA SHARING STATEMENT

Clinical trial data will be uploaded to clinicaltrials.gov and data may be shared with external parties upon request to the corresponding author.

AUTHOR CONTRIBUTIONS

Conception and design: Tracy L. Rose, Allison M. Deal, Angela B. Smith, Eric Wallen, William Y. Kim, Matthew I. Milowsky

Administrative support: Tracy L. Rose, Anthony Drier

Provision of study materials or patients: Tracy L. Rose, Michael R. Harrison, Sundhar Ramalingam, Young E. Whang, Blaine Brower, Mary Dunn, Marc A. Bjurlin, Angela B. Smith, Matthew E. Nielsen, Hung-Jui Tan, Eric Wallen, Michael E. Woods, Daniel George, Tian Zhang, Anthony Drier, William Y. Kim, Matthew I. Milowsky

Collection and assembly of data: Tracy L. Rose, Michael R. Harrison, Allison M. Deal, Sundhar Ramalingam, Young E. Whang, Blaine Brower, Mary Dunn, Marc A. Bjurlin, Daniel George, Tian Zhang, Anthony Drier, Matthew I. Milowsky

Data analysis and interpretation: Tracy L. Rose, Michael R. Harrison, Allison M. Deal, Sundhar Ramalingam, Chelsea K. Osterman, Hillary M. Heiling, Matthew E. Nielsen, Hung-Jui Tan, Michael E. Woods, Daniel George, William Y. Kim, Matthew I. Milowsky

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Phase II Study of Gemcitabine and Split-Dose Cisplatin Plus Pembrolizumab as Neoadjuvant Therapy Before Radical Cystectomy in Patients With Muscle-Invasive Bladder Cancer

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).