Phase II Study of Two Weeks on, One Week off Sunitinib Scheduling in Patients With Metastatic Renal Cell Carcinoma

Abstract

Purpose

Standard frontline treatment of patients with metastatic renal cell carcinoma currently includes sunitinib. A barrier to long-term treatment with sunitinib includes the development of significant adverse effects, including diarrhea, hand-foot syndrome (HFS), and fatigue. This trial assessed the effect of an alternate 2 weeks on, 1 week off (2/1) schedule of sunitinib on toxicity and efficacy in previously untreated patients with metastatic renal cell carcinoma.

Methods

Patients started with oral administration of 50 mg sunitinib on a 2/1 schedule and underwent schedule and dose alterations if toxicity developed. The primary end point was < 15% grade ≥ 3 fatigue, diarrhea, or HFS. With 60 patients, the upper bound of the CI would fall below the published 4/2 schedule grade ≥ 3 toxicity rate of 25% to 30%.

Results

Fifty-nine patients were treated between August 2014 and March 2016. Seventy-seven percent were intermediate or poor risk per Memorial Sloan Kettering Cancer Center criteria. With a median follow-up of 17 months, 25% of patients experienced grade 3 fatigue, HFS, or diarrhea; 37% required a dose reduction, and 10% discontinued because of toxicity. The overall response rate was 57%, median progression-free survival was 13.7 months, and median overall survival was not reached. At 12 weeks, Functional Assessment of Cancer Therapy–General scores dropped between 0% and 10% from baseline, with less reduction in patients who continued treatment longer.

Conclusion

The primary end point of decreased grade 3 toxicity was not met; however, treatment with a 2/1 sunitinib schedule is associated with a lack of grade 4 toxicity, a low patient discontinuation rate, and high efficacy.

INTRODUCTION

The von Hippel-Lindau tumor suppressor–hypoxia-inducible factor pathway is pivotal to the pathobiology of clear cell renal cell carcinoma (RCC) and has been crucial in the development of targeted agents that include vascular endothelial growth factor (VEGF) receptor inhibitors and antibodies to circulating VEGF.¹ Despite their efficacy, these agents continue to be the focus of efforts to maximize response and minimize toxicity.

Sunitinib, an oral inhibitor of VEGF receptors 1, 2, and 3 and platelet-derived growth factor receptor, is administered on a 4 weeks on, 2 weeks off (4/2) schedule. Sunitinib demonstrated a prolonged median progression-free survival (PFS) in a large phase III study of patients with metastatic RCC (mRCC),² received US Food and Drug Administration approval in 2006, and became a standard of care in frontline treatment of mRCC. In real-world practice, maintenance of dose intensity of sunitinib can be challenging because of treatment-related adverse effects, including fatigue, hand-foot syndrome (HFS), and diarrhea. In the original phase III trial, 38% and 32% of patients in the sunitinib arm underwent dose interruptions and reductions, respectively, secondary to toxicities.² The effect of treatment-related toxicities on patients can be significant. Wong et al³ quantified preferences in patients with RCC who completed a survey with hypothetical treatment choice questions. Fatigue and diarrhea were found to be the most troublesome toxicities for patients, and to reduce severe fatigue to mild to moderate fatigue, patients were willing to forego 4.4 months of PFS. On the standard 4/2 schedule, a cyclic scoring pattern is observed in patient-reported quality-of-life (QOL) indicators, with improvement in mean scores after the 2-week treatment break (ie, between day 28 and 42 scores).⁴ Thus, optimization of sunitinib administration could maintain dose intensity but decrease the incidence of troublesome toxicity.

Although early preclinical studies administered a continuous dose of sunitinib, the current 4/2 schedule was chosen for clinical drug development to allow for recovery from bone marrow and adrenal toxicities observed in animal models.⁵ In clinical practice, adverse effects increase in intensity as the treatment cycle progresses and tend to be worse in the final 2 weeks of the treatment cycle. Several alternative schedules of sunitinib have been explored in clinical studies with the goal of improving drug tolerance and delivered dose intensity.^4,6-9 According to dynamic microbubble ultrasound analysis in patients administered sunitinib, the tumor blood volume at days 7 and 14 were reduced compared with baseline, with no additional reduction in blood flow from days 14 to 28.⁹ These findings suggest that the putative antitumor mechanism of sunitinib (ie, reduction in tumor-associated blood flow) is optimized after 14 days, and dosing of sunitinib from days 15 to 28 may increase toxicity without meaningfully contributing to efficacy. In our prior retrospective analysis,¹⁰ of 187 patients, 13% started on a 2/1 schedule and the rest on a 4/2 schedule. During treatment, 53% of patients continued on the traditional schedule and 47% began or were transitioned to alternative schedules as a result of common adverse effects, which included fatigue (64%), HFS (38%), and diarrhea (32%). Median overall survival (OS) was 17.7 months (95% CI, 10.8 to 22.2 months) on the 4/2 schedule versus 33.0 months (95% CI, 29.3 months to not estimable) on the 2/1 schedule (P < .001). Several other retrospective studies also have reported reduced toxicity and comparable outcomes in patients who switched from a 4/2 schedule to a 2/1 schedule as a result of toxicity.^6,11-13 In 31 patients who transitioned from a 4/2 schedule to a 2/1 schedule, Najjar et al⁶ showed that two of the most common toxicities, fatigue and HFS, were significantly less frequent on the 2/1 schedule. Moreover, the RAINBOW (Ramucirumab Plus Paclitaxel Versus Placebo Plus Paclitaxel in Patients With Previously Treated Advanced Gastric or Gastroesophageal Junction Adenocarcinoma) study retrospectively analyzed 249 patients and compared them in three groups (4/2 schedule, 4/2→2/1 schedule, and upfront 2/1 schedule).¹¹ In the 4/2→2/1 group, the overall incidence of grade ≥ 3 toxicities was significantly reduced (from 45.7% to 8.2%; P < .001) after the switch to a 2/1 schedule. The incidence of common grade 3 to 4 toxicities, such as fatigue (10.1% in the 4/2 phase v 0% in the 2/1 phase; P < .001), hypertension (9.1% v 2.4%; P = .007), HFS (10.1% v 3.4%; P = .003), and thrombocytopenia (7.7% v 0.5%; P < .001), also was reduced.

Taken together, these data provide additional impetus to explore modified schedules of sunitinib. We conducted this phase II study with the hypothesis that administering sunitinib on a 2/1 schedule leads to a reduced incidence of grade ≥ 3 fatigue, HFS, and diarrhea compared with historical data while maintaining clinical efficacy.

METHODS

Eligibility Criteria

The institutional review boards of all participating institutions (The University of Texas MD Anderson Cancer Center, Cleveland Clinic Foundation, Fox Chase Cancer Center, and University of North Carolina Lineberger Cancer Center) approved the protocol. All patients provided written informed consent before registration. Eligible patients had histologically or cytologically confirmed treatment-naïve metastatic clear cell RCC; were age ≥ 18 years; and had measurable or evaluable metastatic disease per Response Evaluation Criteria for Solid Tumors (RECIST) version 1.0,¹⁴ an Eastern Cooperative Oncology Group performance status of 0 to 1, and relatively normal organ and bone marrow function. The study was a registered clinical trial. Participating institutions followed the guidelines of the MD Anderson Multicenter Management Plan.

Study Design and Treatment

In this single-arm, multicenter phase II study, patients were treated with sunitinib 50 mg daily on a 2/1 schedule (ie, 2 weeks on followed by 1 week off). Schedule change and dose reduction for toxicity occurred according to the following algorithm: level 0, 50 mg 2/1 schedule; level -1, 50 mg 1 week on/3 days off alternating with 1 week on/4 days off; level -2, 37.5 mg 2/1 schedule; level -3, 37.5 mg 1 week on/3 days off alternating with 1 week on/4 days off; level -4, 25 mg 2/1 schedule; and level -5, 25 mg 1 week on/3 days off alternating with 1 week on/4 days off. Dose levels -1, -3, and -5 were included to determine whether increasing the frequency of breaks could keep patients on a particular dose level. Patients were treated until investigator-assessed RECIST-defined progressive disease or unacceptable toxicity was recorded. One cycle was equal to 6 weeks. Adverse events were graded and categorized by National Cancer Institute Common Terminology Criteria for Adverse Events (version 4.03). Staging studies were assessed per RECIST version 1.0.

Patient Evaluation

Baseline screening evaluations included computed tomography (CT) imaging of the chest, abdomen, and pelvis; bone scan; and CT or magnetic resonance imaging of the brain. CT imaging was repeated every 6 weeks before the start of the next cycle. The Functional Assessment of Cancer Therapy–General (FACT-G) questionnaire was administered to study patients at baseline and after 12, 24, and 36 weeks of treatment by the study team before physician encounter. The FACT-G determines physical (28 points), societal/family (28 points), emotional (24 points), and functional well-being (28 points). These four domains provide an overall denominator of 108. A minimally important difference for the FACT-G is reported to range from 4% to 7%.^15,16

Statistical Considerations

The primary end point was the incidence of grade ≥ 3 treatment-related fatigue, HFS, and diarrhea in patients with mRCC administered sunitinib on a 2/1 schedule and was hypothesized to occur in ≤ 15% of patients compared with a historical rate of 25% to 30% for patients on a 4/2 schedule. Enrollment of 60 patients was planned to estimate the likelihood of toxicity using a 95% CI with a half-width of < 10%. An upper bound of the CI < 25% indicated that the 2/1 schedule has a better toxicity profile than the standard 4/2 schedule. Secondary end points were patient-reported outcomes on the basis of the FACT-G. The percentage of patients with the primary toxicity end point as well as response and dose reductions were presented with 95% CIs. FACT-G scores were calculated according to the validated methods at each time point.¹⁷ Scores for patients with available information were reported descriptively and plotted over time. OS, PFS, and duration of primary toxicity were estimated and plotted using Kaplan-Meier method. SAS 9.3 software (SAS Institute, Cary, NC) was used for the analyses and adverse event and FACT-G score plots.

RESULTS

Patient Characteristics

Sixty patients were enrolled in the study between August 2014 and March 2016. Fifty-nine received treatment and included in the outcome analysis because one patient was not treated as a result of declining mental status before the first dose. Table 1 lists patient demographics and clinical characteristics and shows that patients primarily were male (62%) and had a median age of 65.5 years (range, 45 to 92 years). Patients had a median of two disease sites (range, one to four sites). Metastases most commonly were seen in the lungs (65%), and the remaining target sites were lymph node (17%), liver (13%), bone (5%), and other (48%). Sixty-seven percent of patients had Memorial Sloan Kettering Cancer Center (MSKCC) intermediate-risk features, 10% were in the MSKCC poor-risk category, and 22% showed favorable MSKCC risk features.

Table 1.

Patient Demographics and Clinical Characteristics

Safety

Twenty-five percent of patients had grade ≥ 3 fatigue, diarrhea, or HFS (95% CI, 15.0% to 38.4%) during the study. The median time to development of these toxicities was 3.8 months (95% CI, 1.8 months to not reached; Appendix Fig A2, online only). These selected grade 3 toxicities were of relatively short duration (Fig 1), with a median of 8 days (range, 1 to 48 days). The median (range) number of days of grade ≥ 3 diarrhea, fatigue, and HFS was 5 (1 to 48), 8 (3 to 27), and 7 (2 to 16), respectively. Two patients left the study 3 and 27 days into unresolved grade 3 fatigue, with no date of resolution available. In addition, one episode of grade 4 neutropenia was observed, and no other grade 4 toxicities were reported. Twenty-nine patients (49%) required a schedule change or dose reduction, seven of whom (12%) continued with 50 mg after a schedule change. Ten percent left the study for toxicity (Table 2). Overall, the toxicities patients were most likely to experience were fatigue (78% all grade, 14% grade 3), diarrhea (76% all grade, 8% grade 3), and HFS (54% all grade, 5% grade 3). A number of hematologic adverse effects were observed (Fig 2), with relatively few grade 3 events. Of the six patients who left the study as a result of toxicity, the reasons for treatment discontinuation were osteonecrosis, proteinuria, neutropenia, congestive heart failure, gout, and general intolerance, with multiple concomitant grade 2 toxicities.

Fig 1.

Duration of selected grade 3 toxicities. Each bar represents time on treatment for one study patient with primary toxicity. HFS, hand-foot syndrome.

Table 2.

Toxicity Outcome Summary

Fig 2.

Treatment-emergent toxicities seen in at least 10% of patients. GERD, gastroesophageal reflux disease.

Efficacy

The median follow-up was 17 months. Of the 59 evaluable patients, one stopped treatment before the first assessment and was counted as a nonresponder. Thirty-three patients (56%) responded (95% CI, 42.4% to 68.8%); one (2%) with a complete response and 32 (54%) with a partial response. Seven patients (12%) had progressive disease as their best response (Table 3). At the time of the analysis, the median PFS was 13.7 months (95% CI, 10.9 to 16.3 months) and median OS had not been reached (Appendix Fig A1, online only).

Table 3.

Efficacy Summary

Patient-Reported Outcomes

Fifty-eight patients responded to the FACT-G questionnaire of whom 54 had baseline information and 28 completed information for all four time points. Patients started with a median score of 88.3 of a possible 108 (range, 52.7 to 108). No difference was seen in baseline scores between the patients who did or did not experience grade 3 toxicities. The 38 patients who continued with treatment for at least 36 weeks showed the least decline in FACT-G scores (Fig 3). Eight and seven patients continued with treatment for 24 to 36 and 12 to 24 weeks, respectively. An additional five patients continued with treatment for < 12 weeks, four patients without a baseline FACT-G score are not presented. Median FACT-G scores in later weeks remained numerically higher among patients who did not experience grade 3 toxicity (Appendix Table A1, online only).

Fig 3.

Change in Functional Assessment of Cancer Therapy–General scores as a function of time on treatment with 95% confidence intervals.

DISCUSSION

Despite the rapidly changing number of approved agents, anti-angiogenic agents remain a mainstay of treatment of patients with advanced RCC. Anti-angiogenic agents possess class-specific adverse effects, which include diarrhea, HFS, and fatigue. These toxicities affect patients’ ability to continue with treatment, even in the controlled environment of a clinical trial, as shown by the > 20% discontinuation rate of sunitinib and pazopanib, another commonly used anti-angiogenic agent, in phase III trials.^2,18,19 Modification of the administration of these agents to maintain dose intensity but reduce toxicity could improve patient QOL; reduce the burden on the patient, caregivers, and health care providers; and reduce the cost of treatment as a result of reduced use of concomitant medications, clinic and emergency department visits, and hospitalizations. In addition, efficacy outcomes may improve for these agents by virtue of continuing treatment for longer periods of time with adequate drug exposure. Because anti-angiogenic agents will continue to be important for the treatment of RCC as both monotherapy and, potentially, in combination regimens for the foreseeable future, optimization of their administration is important.

The current study tested the hypothesis that administration of sunitinib on a 2/1 schedule would decrease toxicity for patients relative to historical control data. The primary end point of occurrence of grade 3 HFS, diarrhea, and fatigue in ≤ 15% of patients was not met (Table 2). Twenty-five percent of patients experienced at least one of these three grade 3 toxicities, and 37% required a dose reduction. The dose reduction rate in the phase III Pazopanib Versus Sunitinib in the Treatment of Locally Advanced and/or Metastatic Renal Cell Carcinoma (COMPARZ) study was 51% for sunitinib-treated patients.¹⁸ Of note, in the original phase III sunitinib study, only 32% of sunitinib-treated patients had a dose reduction.² Nevertheless, no significant grade 4 toxicities were observed in the current study. In addition, the discontinuation rate of 10% is considerably lower than the 20% discontinuation rate for adverse events seen with sunitinib in the COMPARZ study.¹⁸ Grade 3 hypertension occurred in 16%, thrombocytopenia in 3%, and neutropenia of 8% of patients in the current study, which compares favorably with grade 3/4 hypertension of 16%, thrombocytopenia of 18%, and neutropenia of 19% in COMPARZ.¹⁸ The 5% grade 3 HFS rate in the current study also was lower than the 11% rate seen in COMPARZ.¹⁸ These findings suggest that although patients still experienced grade 3 toxicities, the initial 2/1 schedule and subsequent schedule and dose alterations ensured that 90% could continue treatment and avoid protracted high-grade toxicities.

The study was designed to provide more granular information on treatment-driven QOL changes. Assessment of QOL with the FACT-G questionnaire showed that patients who experienced grade 3 toxicities had a numerical decline in their aggregate score, but this difference was not significant (Fig 3). Studies that assessed the minimally important difference for QOL tools indicated that a 4% to 7% change would have some effect on patients,^15,16 which was not reached by patients on treatment for > 36 weeks in the current study but was broached by patients who discontinued treatment by 24 weeks. Assessment of the relationship between time on treatment and change in QOL showed that those individuals who left the study earlier had the greatest decline in FACT-G scores. This finding may indicate that schedule and dose adjustment were insufficient in the patients who discontinued treatment early or that inherent disease characteristics associated with early disease progression were manifested in a declining QOL. In addition, the overall recovery of QOL scores, especially in patients who continued treatment for > 24 weeks, seems superior to data reported for the sunitinib arm of the COMPARZ study, where the decline in QOL for sunitinib-treated patients persisted.¹⁸

The 57% response rate and median 13.7-month PFS are particularly striking given the predominantly intermediate- and unfavorable-risk study population and should be viewed in the context of data from several other retrospective analyses and prospective trials. An objective response rate (ORR) of 12% and a PFS of only 5.3 months for previously untreated, intermediate- and poor-risk patients in the sunitinib arm of the randomized phase II CABOSUN trial²⁰ of cabozantinib versus sunitinib as initial targeted therapy for patients with mRCC may be due to suboptimal drug scheduling but is influenced by the study population itself. A more comparable control population may be found in the Checkmate 214 study, which had a similar risk group distribution.²¹ Sunitinib-treated patients exhibited an ORR of 32% (95% CI, 28% to 36%) and a PFS of 12.3 months (95% CI, 9.8 to 15.2 months). Similarly, a retrospective analysis by Kondo et al¹³ showed that the median PFS was longer in patients on a 2/1 schedule than those on a 4/2 schedule (18.4 v 9.1 months; P = .13), although the ORR tended to be higher with a 4/2 schedule (50% v 32%; P = .14). The RAINBOW analysis showed a median PFS of 10.4 months in the 2/1 group and 9.7 months in the 4/2 group.²² In addition, a recent randomized phase II trial compared 36 patients who received sunitinib on a 4/2 schedule with 38 patients on a 2/1 schedule. The ORR was 47% for the 2/1 schedule and 36% for the 4/2 schedule, Failure-free survival rates at 6 months were 44% with the 4/2 schedule and 63% with the 2/1 schedule.²³ In summary, these studies showed that efficacy is at least maintained and perhaps improved on alternate sunitinib schedules, with indications that toxicity is meaningfully mitigated.

This study has several limitations. It was conducted at academic centers, which could bias the results relative to practices with less experience with either the disease or this class of agent. Whether these observations are a reflection of a real-world application of a 2/1 sunitinib schedule requires additional investigation. Furthermore, the lack of randomization and a relatively small sample size may limit the effect of these observations. Larger randomized studies would more conclusively prove the superiority of a 2/1 sunitinib schedule over the Food and Drug Administration–approved 4/2 schedule. Nonetheless, the acceptable toxicity and high efficacy shown here support the adoption of a 2/1 option in the management of patients with mRCC.

In conclusion, sunitinib administered to a predominantly intermediate- and poor-risk frontline RCC population results in low treatment discontinuation rates, a maintained QOL, and robust efficacy. These nonrandomized data also support the use of alternate sunitinib scheduling as a means to maintain patient QOL and to extend duration of treatment.

Appendix

Fig A1.

(A) Progression-free survival (PFS). (B) Overall survival (OS). Tick marks indicate censoring.

Fig A2.

Time to first primary end point toxicity. Tick marks indicate censoring.

Table A1.

FACT-G Results

AUTHOR CONTRIBUTIONS

Conception and design: Eric Jonasch, Rebecca S. Slack, Brian I. Rini

Provision of study materials or patients: Brian I. Rini

Collection and assembly of data: Daniel M. Geynisman, Elshad Hasanov, Matthew I. Milowsky, W. Kimryn Rathmell, Summer Stovall, Troy R. Gilchrist, Elizabeth R. Plimack, Nizar M. Tannir, Brian I. Rini

Data analysis and interpretation: Eric Jonasch, Rebecca S. Slack, Daniel M. Geynisman, Elshad Hasanov, Matthew I. Milowsky, W. Kimryn Rathmell, Donna Juarez, Lisa Pruitt, Moshe C. Ornstein, Elizabeth R. Plimack, Nizar M. Tannir, Brian I. Rini

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 Two Weeks on, One Week off Sunitinib Scheduling in Patients With Metastatic Renal Cell Carcinoma

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. 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/site/ifc.

Eric Jonasch

Consulting or Advisory Role: Pfizer, Novartis, Genentech, Roche, Exelixis, Eisai

Research Funding: Exelixis, Pfizer, Novartis, Peloton Therapeutics

Travel, Accommodations, Expenses: Pfizer

Rebecca S. Slack

No relationship to disclose

Daniel M. Geynisman

Consulting or Advisory Role: Pfizer, Novartis, Exelixis

Research Funding: Pfizer (Inst), Genentech (Inst), Merck (Inst)

Elshad Hasanov

No relationship to disclose

Matthew I. Milowsky

Research Funding: Pfizer (Inst), Merck (Inst), Acerta Pharma (Inst), Roche (Inst), Genentech (Inst), Bristol-Myers Squibb (Inst), Incyte (Inst), Agensys (Inst)

Travel, Accommodations, Expenses: Roche, Genentech

W. Kimryn Rathmell

No relationship to disclose

Summer Stovall

No relationship to disclose

Donna Juarez

No relationship to disclose

Troy R. Gilchrist

No relationship to disclose

Lisa Pruitt

No relationship to disclose

Moshe C. Ornstein

Consulting or Advisory Role: Pfizer, Eisai

Elizabeth R. Plimack

Consulting or Advisory Role: AstraZeneca, MedImmune, Bristol-Myers Squibb, Clovis Oncology, Exelixis, Genentech, Roche, Horizon Pharma, Inovio Pharmaceuticals, Merck Sharp & Dohme, Novartis, Pfizer, Eli Lilly

Research Funding: Bristol-Myers Squibb (Inst), AstraZeneca (Inst), Pfizer (Inst), Merck Sharp & Dohme (Inst), Peloton Therapeutics (Inst), Astellas Pharma (Inst), Genentech (Inst), Roche (Inst)

Patents, Royalties, Other Intellectual Property: US Patent No. 14/588,503 filed January 2, 2015 (Inst), US Patent No: 15/226,474 filed July 1, 2015 (Inst)

Nizar M. Tannir

Honoraria: Pfizer, Novartis, Bristol-Myers Squibb, Exelixis, Nektar, Argos Therapeutics, Calithera Biosciences

Consulting or Advisory Role: Novartis, Exelixis, Bristol-Myers Squibb, Nektar, Pfizer, Argos Therapeutics

Research Funding: Bristol-Myers Squibb, Novartis, Exelixis, Epizyme, Mirati Therapeutics

Travel, Accommodations, Expenses: Pfizer, Novartis, Exelixis, Nektar, Bristol-Myers Squibb, Argos Therapeutics, Calithera Biosciences

Brian I. Rini

Consulting or Advisory Role: Pfizer

Research Funding: Pfizer (Inst)

Travel, Accommodations, Expenses: Pfizer