Brentuximab Vedotin Plus Chemotherapy in North American Subjects with Newly Diagnosed Stage III Or IV Hodgkin Lymphoma Radhakrishnan Ramchandren1, Ranjana H

Published OnlineFirst January 7, 2019; DOI: 10.1158/1078-0432.CCR-18-2435

Clinical Trials: Targeted Therapy Clinical Cancer Research Brentuximab Vedotin plus Chemotherapy in North American Subjects with Newly Diagnosed Stage III or IV Hodgkin Lymphoma Radhakrishnan Ramchandren1, Ranjana H. Advani2, Stephen M. Ansell3, Nancy L. Bartlett4, Robert Chen5, Joseph M. Connors6, Tatyana Feldman7, Andres Forero-Torres8, Jonathan W. Friedberg9, Ajay K. Gopal10, Leo I. Gordon11, John Kuruvilla12, Kerry J. Savage6, Anas Younes13, Gerald Engley14, Thomas J. Manley14, Keenan Fenton14, and David J. Straus13

Abstract

Purpose: To evaluate safety and efficacy outcomes for sub- among subjects who received AþAVD compared with ABVD jects on the ECHELON-1 study treated in North America (NA). (HR ¼ 0.60; P ¼ 0.012). For PFS, the risk of progression or Experimental Design: ECHELON-1 is a global, open-label, death was also reduced (HR ¼ 0.50; P ¼ 0.002). Subsequent randomized phase III study comparing doxorubicin, vinblas- anticancer therapies were lower in the AþAVD arm. Grade 3 or tine, and dacarbazine in combination with brentuximab vedo- 4 adverse events (AEs) were more common, but there were tin (AþAVD) versus ABVD (AVD þ bleomycin) as first-line fewer study discontinuations due to AEs in the AþAVD arm therapy in subjects with stage III or IV classical Hodgkin as compared with ABVD. Noted differences between arms lymphoma (cHL; NCT01712490). Subjects were randomized included higher rates of febrile neutropenia (20% vs. 9%) 1:1 to receive AþAVD or ABVD intravenously on days 1 and and peripheral neuropathy (80% vs. 56%), but lower rates of 15 of each 28-day cycle for up to 6 cycles. pulmonary toxicity (3% vs. 10%) in subjects treated with Results: The NA subgroup consisted of 497 subjects in the AþAVD versus ABVD. AþAVD (n ¼ 250) and ABVD (n ¼ 247) arms. Similar to the Conclusions: The efficacy benefit and manageable toxicity primary analysis based on the intent-to-treat population, profile observed in the NA subgroup of ECHELON-1 support the primary endpoint [modified progression-free survival AþAVD as a frontline treatment option for patients with (PFS) per independent review] demonstrated an improvement stage III or IV cHL.

Introduction 1Department of Hematology/Oncology, Barbara Ann Karmanos Cancer Cen- ter, Detroit, Michigan. 2Department of Medicine, Stanford University, Palo Approximately 8,200 cases of classical Hodgkin lymphoma Alto, California. 3Division of Hematology, Mayo Clinic, Rochester, Minnesota. (cHL) are diagnosed annually in the United States (1). For over 4 Division of Oncology, Siteman Cancer Center, Washington University School 40 years, the first-line standard of care for cHL in North America of Medicine, St. Louis, Missouri. 5Department of Hematology and Hemato- (NA) has been chemotherapy with ABVD (2–4). Despite high poietic Cell Transplantation, City of Hope National Medical Center, Duarte, California. 6University of British Columbia and the Department of Medical response rates, approximately 30% of patients with advanced Oncology, British Columbia Cancer Centre for Lymphoid Cancer, Vancouver, stage Hodgkin lymphoma are refractory or relapse following British Columbia, Canada. 7John Theurer Cancer Centre, Hackensack Uni- first-line treatment with ABVD (5–7). versity Medical Center, Hackensack, New Jersey. 8Department of Medicine, ECHELON-1 is an international, phase III trial comparing 9 University of Alabama at Birmingham, Birmingham, Alabama. James P doxorubicin, vinblastine, and dacarbazine (AVD) in combination Wilmot Cancer Institute, University of Rochester Medical Center, Rochester, 10 with brentuximab vedotin (ADCETRIS; AþAVD) versus ABVD New York. Fred Hutchinson Cancer Research Center, University of þ fi Washington, Seattle, Washington. 11Robert H. Lurie Comprehensive Cancer (AVD bleomycin) as rst-line therapy in subjects with stage III Center, Northwestern University, Chicago, Illinois. 12Division of Medical or IV cHL (8). The primary endpoint was modified progression- Oncology and Hematology, Princess Margaret Hospital, Toronto, Ontario, free survival (PFS) per independent review facility (IRF), defined Canada. 13Lymphoma Service, Department of Medicine, Memorial Sloan as progression, death, or the receipt of additional treatment for 14 Kettering Cancer Center, New York, New York. Seattle Genetics, Inc., subjects not achieving complete response (CR) at the completion Bothell, Washington. of first-line therapy. It showed that AþAVD was superior to ABVD Note: Supplementary data for this article are available at Clinical Cancer (HR ¼ 0.77, P ¼ 0.035) with 2-year modified PFS rates of 82.1% Research Online (http://clincancerres.aacrjournals.org/). and 77.2%, respectively, with no significant difference in overall Corresponding Author: Radhakrishnan Ramchandren, The University of survival (OS) at the OS interim analysis. Treatment with AþAVD Tennessee, 1926 Alcoa Hwy, Knoxville, TN 37920. Phone: 865-305-9162; E-mail: was associated with higher rates of febrile neutropenia and [email protected] peripheral neuropathy, and lower rates of pulmonary-related doi: 10.1158/1078-0432.CCR-18-2435 toxicity compared with ABVD. AþAVD is the first treatment 2019 American Association for Cancer Research regimen utilizing a targeted agent to show superior efficacy in

www.aacrjournals.org OF1

Ramchandren et al.

another cancer within 3 years before the first dose, or clinically Translational Relevance relevant cardiovascular conditions. For over 40 years, the standard of care for first-line Hodgkin lymphoma in North America has been chemotherapy with Disease assessments ABVD or an ABVD-like regimen. The novel AþAVD regimen Modified PFS by independent review facility (IRF) was the incorporates a targeted agent, the CD30-directed antibody- primary objective of the ECHELON-1 trial. Regional analysis of drug conjugate brentuximab vedotin (ADCETRIS), into a modified PFS per IRF, including North America, was a prespecified backbone of AVD chemotherapy. The phase 3, ECHELON-1 sensitivity analysis. All additional analyses for the regional sub- trial established AþAVD as the first frontline regimen with a groups were exploratory. Disease progression was evaluated in targeted agent to demonstrate a modified progression-free accordance with the Revised Response Criteria for Malignant survival benefit in comparison with ABVD for stage III or IV Lymphoma (Cheson 2007; ref. 10) by both a blinded IRF and classical Hodgkin lymphoma (cHL). A preplanned subgroup investigator (INV) assessment. analysis of ECHELON-1 revealed regional variation with a Modified PFS per IRF was defined as time from randomization trend toward greater efficacy in North America; we present the to first documentation of progressive disease (per Cheson 2007), efficacy, safety, and disease management of subjects in death due to any cause, or confirmed nonresponse [End of North America to better understand AþAVD as a first-line Treatment (EOT) Deauville score 3) for subjects who received treatment option for patients with stage III or IV cHL. Although additional anticancer therapy. Modified PFS per INV was defined these findings are hypothesis-generating, they may provide as time from randomization to first documentation of progressive meaningful information for managing patients with AþAVD disease (per Cheson 2007), death due to any cause, or confirmed and influence future cHL studies. nonresponse (per Cheson 2007) for subjects who received additional anticancer therapy. PFS per INV was defined as time from randomization to first documentation of progressive disease (per Cheson 2007) or death due to any cause. terms of modified PFS compared with ABVD while also elim- An additional secondary endpoint was rate of CR defined as inating the need for bleomycin in subjects with previously best overall response achieved at the end of randomized regimen untreated stage III or IV cHL. For the ECHELON-1 primary per IRF assessment (10). Subject disease status, following cycle 2 endpoint of modified PFS, assessment by region of the world and at EOT, by PET scan (using the Deauville criteria) was also was among the preplanned subgroup analyses, and consistent assessed by IRF. with the primary analysis showed an improvement in modified All survival endpoints were summarized using the Kaplan– PFS for subjects randomized to AþAVD compared with those Meier method and evaluated with the use of a log-rank test. A Cox randomized to ABVD within the NA region (Canada and the regression model was used to estimate the HR and the 95% United States). Here, we present additional safety and efficacy confidence interval (CI) for the treatment effect. outcomes for subjects on the ECHELON-1 study treated in NA. Safety evaluations included incidence of adverse events (AEs), as well as severity and type of AE defined according to the Medical Dictionary for Regulatory Activities (MedDRA), version 19.0, with Materials and Methods grading of AEs defined according to the National Cancer Institute Trial design Common Terminology Criteria for Adverse Events [NCI-CTCAE], ECHELON-1 is a global, open-label, randomized phase III version 4.03. study of AþAVD versus ABVD as first-line therapy in subjects with stage III or IV cHL (NCT01712490). Subjects were randomly Oversight assigned in a 1:1 ratio, stratified by region [Americas (39%), The ECHELON-1 trial was conducted in accordance with reg- Europe (50%), and Asia (11%)] and international prognostic ulatory requirements; the protocol (previously described ref. 8) score (IPS), to receive AþAVD or ABVD intravenously on days 1 was approved by Institutional Review Boards and ethics commit- and 15 of each 28-day cycle for up to 6 cycles. A detailed study tees at individual sites, and adhered to Good Clinical Practice design and the results of the entire intent-to-treat (ITT) popula- guidelines (as defined by the International Council for Harmo- tion have been published previously (8). The NA subset consisted nisation). A steering committee and an independent data and of subjects enrolled at 85 sites across the United States and safety monitoring committee oversaw the conduct of the trial, and Canada. all the subjects provided written informed consent. Data were analyzed by sponsor statisticians and interpreted by academic Subjects authors and sponsor representatives. All the authors vouch for the Subjects 18 years of age or older with histologically confirmed completeness and accuracy of the data and adherence of the trial stage III or IV cHL, according to the World Health Organization to the protocol. pathology classification system (9), who had not been previously treated with systemic chemotherapy or radiotherapy, were eligi- ble. Subjects were required to have an Eastern Cooperative Results Oncology Group (ECOG) performance status of 0, 1, or 2, and Demographics satisfactory lab results (absolute neutrophil count, platelet count, The NA subgroup of the ECHELON-1 study consisted of 497 hemoglobin level, markers of liver and kidney function). Subjects subjects (37% of the entire study population), 250 in the AþAVD were ineligible for the study if they had nodular lymphocyte arm and 247 in the ABVD arm. In the AþAVD arm, 214 subjects predominant HL, peripheral sensory or motor neuropathy, any (86%) were from the United States and 36 (14%) were from evidence of residual disease from another cancer, diagnosis of Canada, whereas in the ABVD arm, 225 subjects (91%) were from

OF2 Clin Cancer Res; 2019 Clinical Cancer Research

Brentuximab Vedotin for Hodgkin Lymphoma in North America

Table 1. Demographics and disease characteristics for North American European region was 0.83 (95% CI, 0.59–1.17; P ¼ 0.281) and subgroup 0.91 (95% CI, 0.43–1.93; P ¼ 0.810) in the Asian region (8). þ A AVD ABVD Modified PFS per investigator was generally consistent with (n ¼ 250) (n ¼ 247) these results by region (Supplementary Table S1). Subject demographics fi Sex, n (%) Modi ed PFS per IRF showed a consistent trend toward þ Male 146 (58) 148 (60) improvement among NA subjects treated with A AVD versus Female 104 (42) 99 (40) ABVD across all subgroups that were prespecified in the pri- Age, mean (SD) 38.2 (16) 40.5 (15) mary analysis (Fig. 2). In addition, there was an advantage a Age categories (years) , n (%) for AþAVD versus ABVD in subjects who were PET-negative at < 45 171 (68) 162 (66) cycle 2 (HR ¼ 0.60; 95% CI, 0.37–0.97; P ¼ 0.034), with a 45–59 50 (20) 51 (21) fi – 60–64 11 (4) 14 (6) 2-year modi ed PFS of 87.9% (82.5% 91.6%) versus 78.3% – 65 18 (7) 20 (8) (71.7% 83.5%), respectively. Subjects who were PET2-positive Country, n (%) showedatrendtowardagreaterbenefitwithAþAVD (HR ¼ United States 214 (86) 225 (91) 0.52; 95% CI, 0.21–1.27; P ¼ 0.139), with a 2-year modified Canada 36 (14) 22 (9) PFS rate on the AþAVD arm of 58.2% (33.5%–76.5%) versus n Race, (%) 36.6% (16.7%–56.9%) on the ABVD arm. White 213 (85) 204 (83) fi Black or African American 15 (6) 16 (6) Although the prespeci ed event count for assessing the key Not reported 9 (4) 15 (6) secondary endpoint of OS has not been reached for the ITT Asian 7 (3) 7 (3) population, an interim analysis of OS in subjects treated in Other 6 (2) 5 (2) NA showed a trend in favor of the AþAVD arm versus the Disease characteristics, n (%) ABVD arm (HR ¼ 0.51; 95% CI, 0.23–1.14; P ¼ 0.094). The Ann Arbor stage at initial diagnosis interim 2-year OS rates (95% CI) were 97.0% (93.7%–98.6%) Stage III 99 (40) 115 (47) þ Stage IV 150 (60) 132 (53) in the A AVD arm and 93.2% (88.8% to 95.8%) in the IPS score ABVD arm. 0–1 50 (20) 55 (22) Modified PFS was the predefined primary endpoint of the 2–3 123 (49) 119 (48) ECHELON-1 trial; however, an assessment of traditional PFS 4–7 77 (31) 73 (30) defined as time to either progression or death, was also per- ECOG performance status formed in NA subjects. The results from this analysis demon- 0 118 (47) 118 (48) fi þ 1 121 (48) 118 (48) strate a similar bene tforA AVDversusABVDaswasobserved fi 2 11 (4) 11 (4) for the modi ed PFS endpoint. At a median follow-up of Extranodal involvement at initial diagnosis 24.7 months, there was an improvement in PFS per investigator None 74 (30) 92 (37) assessment in subjects treated with AþAVD versus ABVD, 1 site 74 (30) 79 (32) corresponding to a 50% reduction in the risk of progression > 1 site 78 (31) 63 (26) or death (HR ¼ 0.50; 95% CI, 0.32–0.79; P ¼ 0.002; Fig. 1B). B symptomsb 146 (58) 139 (56) The 2-year PFS per investigator (95% CI) was 88.1% (83.1%– aAge at date of informed consent. 91.7%) in the AþAVD arm and 76.4% (70.1%–81.5%) in bSubjects who present with B symptoms [general malaise (fever, night sweats, weight loss, etc.)] for at least 1 visit prior to the start of study drug administration. ABVD arm, an absolute difference of 11.7 percentage points. A benefit was also seen in NA subjects who were PET2-negative at cycle 2 (HR ¼ 0.439; 95% CI, 0.25–0.77; P ¼ 0.003), with a 2-year PFS rate of 91.9% (87.0%–95.0%) in the AþAVD arm the United States and 22 (9%) were from Canada. Baseline versus 81.1% (74.7%–86.0%) with ABVD; for PET2-positive demographics and disease characteristics were consistent between subjects, 2-year PFS rates of 54.5% (32.1%–72.4%) versus treatment arms (Table 1), and were generally comparable with the 48.0% (26.0%–67.0%) were seen (HR ¼ 0.901; 95% CI, 0.38– overall ITT population. 2.12; P ¼ 0.810). Response rates as determined by IRF demonstrated a Efficacy consistent benefitforAþAVD compared with ABVD and are At a median follow-up of 24.7 months, there was a clinically summarized in Supplementary Table S2. The CR rates fol- meaningful improvement in modified PFS per IRF in subjects lowing first-line therapy (per the Revised Response Criteria treated with AþAVD versus ABVD, corresponding to a 40% for Malignant Lymphoma; ref. 10) were 72% in the AþAVD reduction in the risk of progression, death, or modified progres- arm and 67% in the ABVD arm. At the cycle 2 PET scan, 88% sion events (HR ¼ 0.60; 95% CI, 0.40–0.90; P ¼ 0.012; Fig. 1A). of subjects in the AþAVD arm and 83% of subjects in the The 2-year modified PFS per IRF (95% CI) was 84.3% (78.7%– ABVD arm had a Deauville score 3; at the EOT PET scan, 88.5%) in the AþAVD arm and 73.7% (67.3%–79.1%) in the the rates of Deauville Score 3 were 86% and 78%, respec- ABVD arm, an absolute difference of 10.6 percentage points. The tively. A summary of the rates of Deauville Scores per IRF at AþAVD and ABVD arms reported a total of 38 and 57 events, the cycle 2 and EOT PET scans is presented in Supplementary respectively. The majority of events reported in both arms were Table S2. disease progression (28 in the AþAVD arm vs. 39 in the ABVD Subjects treated with AþAVD required fewer subsequent arm) followed by deaths (6 events in the AþAVD arm vs. 10 events anticancer therapies than subjects treated with ABVD. In the in the ABVD arm). Modified progression accounted for 4 events in AþAVD arm, 30 subjects (12%) received at least 1 subsequent the AþAVD arm versus 8 events in the ABVD arm (Table 2). As anticancer therapy (some subjects received more than 1 previously reported, the HR for modified PFS per IRF in the subsequent treatment): 12 subjects received 1 or more cycles

www.aacrjournals.org Clin Cancer Res; 2019 OF3

Ramchandren et al.

A 1.0

0.8

0.6 PFS

0.4 Probability of modified 0.2 Events N Progression Death Modified Total A+AVD 250 28 6 4 38 Log-rank test P value: 0.012 ABVD 247 39 10 8 57 HR (95% CI): 0.596 (0.395−0.899) 0.0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 Time (months) from randomization Number at risk: A+AVD 250 240 238 227 208 203 196 185 173 165 129 123 117 72 68 62 35 30 30 13 11 10 3 3 3 0 0 ABVD 247 231 226 210 185 173 164 156 147 141 109 102 99 63 57 54 28 24 23 8 7 7 1 1 1 0 0 B 1.0

0.8

0.6

PFS 0.4 Probability of

0.2 Events N Progression Death Total DVA+A 250 24 5 92 Log-rank test P value: 0.002 ABVD 247 43 10 53 HR (95% CI): 0.500 (0.318−0.786) 0.0

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 Time (months) from randomization Number at risk: A+AVD 25024123822921320720218917617113712912477726538313114121133300 ABVD 247 233 228 216 191 182 174 167 159 155 123 113 109 66 60 58 31 27 25 8 7 7 2 2 2 0 0

Censored subjects are represented by a solid circle (A+AVD) or triangle (ABVD).

Figure 1. Kaplan–Meier analysis of modiﬁed PFS per IRF (A) and PFS per investigator assessment (B) by treatment group in North America.

of chemotherapy, 12 received high-dose chemotherapy plus Dose delivery transplant, 9 received radiation, and 3 received a checkpoint There were noted differences in the management of drug inhibitor. In the ABVD arm, 52 subjects (22%) received at dosing and administration for each treatment regimen; the least 1 subsequent anticancer therapy. Of these, 41 subjects greatest differences were with brentuximab vedotin and bleo- received 1 or more cycles of chemotherapy, 22 received high- mycin (Table 3). Subjects who received at least 1 dose of study dose chemotherapy plus transplant, 11 received radiation, 10 drug were included in the safety population, which comprised subjects received a checkpoint inhibitor alone or in combi- 489 subjects (249 in the AþAVD arm and 240 in the ABVD nation, and 3 received radiation with chemotherapy. arm). Alteration of the dose or administration of any individual

OF4 Clin Cancer Res; 2019 Clinical Cancer Research

Brentuximab Vedotin for Hodgkin Lymphoma in North America

Table 2. Summary of end of treatment Deauville scores and subsequent For the ECHELON-1 safety population, there were noted dif- therapies for subjects modified PFS events per IRF ferences between treatment arms in select TEAEs of interest, þ A AVD ABVD including febrile neutropenia, peripheral neuropathy, and pul- n ¼ 250 n ¼ 247 monary toxicity; similar differences were noted in the NA Number with events per IRF, n (%) 38 (15) 57 (23) Reason leading to modified PFS event, n (%) population. þ Progressive disease 28 (11) 39 (16) Neutropenia occurred in 62% of subjects receiving A AVD Death due to any cause 6 (2) 10 (4) and 54% of subjects receiving ABVD (Table 4). The incidence of Receipt of additional therapy after non-CRa, n (%) 4 (2) 8 (3) grade 3 or higher neutropenia was higher in subjects treated Salvage chemotherapy 3/4 (75) 7/8 (88) with AþAVD (59%) compared with subjects receiving ABVD b Deauville score at end of treatment (45%). The incidence of febrile neutropenia was also higher in 300þ 4 2/3 (67) 2/7 (29) A AVD arm (20%) compared with the ABVD arm (9%). þ 5 1/3 (33) 5/7 (71) Among subjects in the A AVD arm, 35 (14%) received G-CSF Radiation 1/4 (25) 1/8 (13) primary prophylaxis; of these, 3 subjects (9%) experienced Deauville score at end of treatmentb febrile neutropenia on-study. Among subjects in the ABVD 3 1/1 (100) 0 arm, 11 (5%) received G-CSF primary prophylaxis; none expe- 4 0 1/1 (100) rienced febrile neutropenia on-study. 500 Peripheral neuropathy was reported by the investigators in 80% aConfirmed noncomplete responders (Deauville 3, 4, or 5) with receipt of of subjects receiving AþAVD and 56% of subjects receiving ABVD additional anticancer treatment after completion of first-line therapy. bEvidence of a non-CR required an EOT Deauville score 3. (Table 4). Although the incidence of grade 1 peripheral neuropathy was similar between treatment arms, the incidence of grade 2 and grade 3 peripheral neuropathy was higher in subjects treated drug was most common for brentuximab vedotin (61%) in the with AþAVD (21% and 17%, respectively) compared with sub- AþAVD arm and bleomycin (53%) in the ABVD arm. The most jects treated with ABVD (12% and <1%, respectively). The inci- frequent alterations of brentuximab vedotin in the AþAVD arm dence of peripheral neuropathy in the AþAVD arm was similar were dose delay and dose reduction; 35% and 32% of subjects between NA (79.5%) and Asia (75.7%) and was lower in Europe in the AþAVD arm had at least 1 brentuximab vedotin dose (55.4%; Supplementary Table S6). The incidence of peripheral delay or dose reduction, respectively. The most frequent altera- neuropathy in the ABVD arm was lower in Europe (35.1%) and tions for bleomycin in the ABVD arm were dose delay and dose Asia (42.3%), compared with NA (55.8%). discontinuation; 27% and 26% of subjects in the ABVD arm Of the 198 NA subjects in the AþAVD arm who experienced had at least 1 bleomycin dose delay or dose discontinuation, peripheral neuropathy during the study, 149 subjects (75%) respectively. showed improvement or resolution at last follow-up: 42% had Regional rates of dose modifications for brentuximab vedotin complete resolution and 33% showed improvement by at least and bleomycin are presented in Supplementary Table S3. 1 grade. Ongoing events at last follow-up related to peripheral Notably, dose reductions for brentuximab vedotin were more neuropathy were grade 1 (37%), grade 2 (15%), or grade 3 (6%). common in NA than in Europe or Asia (32%, 21%, 26%, Similarly, 74% of the 134 subjects who experienced peripheral respectively); however, dose delays were less common (35%, neuropathy in the ABVD arm showed improvement or resolution: 55%, 59% respectively). In the ABVD arm, bleomycin discon- 60% had complete resolution and 13% showed improvement tinuation was more common in NA (26%) compared with by at least 1 grade. Ongoing events at last follow-up related Europe (9%) or Asia (18%). to peripheral neuropathy were grade 1 (27%), grade 2 (12%), In all, full 6 cycles of brentuximab vedotin were received by or grade 3 (<1%). 70% of subjects in the AþAVD arm, whereas 6 cycles of bleo- Pulmonary-related toxicity was reported in 3% of subjects in mycin were received by 63% of subjects in the ABVD arm. For the the AþAVD arm versus 10% of subjects in the ABVD arm (Table 4). remaining drugs in both regimens (doxorubicin, vinblastine, Pulmonary-related toxicity events of grade 3 or higher were and dacarbazine), 83% to 90% of subjects received all 6 cycles in reported in 2% of subjects in the AþAVD arm and 6% of subjects NA (Table 3); similar rates of doxorubicin, vinblastine, and in the ABVD arm. In subjects who received primary prophylaxis dacarbazinedosedeliverywereobservedinEuropeandAsia with G-CSF, pulmonary-related toxicity was observed in 1 of the (data not shown). 35 subjects in the AþAVD arm (grade 3) and 3 of the 11 subjects in the ABVD arm (2 were grade 3). Safety There were 9 deaths that occurred within 30 days of last Treatment-emergent adverse events (TEAEs) are summa- treatment (i.e., on-study deaths). In the AþAVD arm, there were rized in Supplementary Table S4. In the AþAVD arm, all 2 on-study deaths; 1 of these was related to febrile neutropenia in subjects in the NA safety population experienced at least 1 a subject who did not receive primary prophylaxis with G-CSF. TEAE, 81% had grade 3 or higher TEAEs, 77% had drug-related There were 7 on-study deaths in the ABVD arm; 6 of these subject grade 3 or higher TEAEs, and 15% had TEAEs that resulted in deaths were associated with pulmonary-related toxicity. study drug discontinuation or dose alterations. Similarly, with ABVD, all subjects experienced at least 1 TEAE, 67% had grade 3 or higher TEAEs, 56% had drug-related grade 3 or higher Discussion TEAEs, and 24% had TEAEs that resulted in study drug or dose The international phase III ECHELON-1 trial demonstrated an discontinuation. The most common AEs in both arms were improvement in modified PFS for subjects with previously nausea, constipation, fatigue, and neutropenia (Supplemen- untreated stage III or IV cHL treated with AþAVD as compared tary Table S5). with ABVD. Prespecified subgroup analyses of the primary

www.aacrjournals.org Clin Cancer Res; 2019 OF5

Ramchandren et al.

Event n (%) HR DVBADVA+ApuorgbuS %59( )IC llarevO 052/83 742/75)2.51( )1.32( 695.0 593.0( − )998.0 Age <60 years 30/221 (13.6) 49/213 (23.0) 0.534 (0.339−0.841) ≥60 years 8/29 (27.6) 8/34 (23.5) 0.974 (0.360−2.634) <45 years 22/171 (12.9) 33/162 (20.4) 0.565 (0.330−0.970) ≥45 years 16/79 (20.3) 24/85 (28.2) 0.668 (0.355−1.257)

IPS score 0–1 9/50 (18.0) 12/55 (21.8) 0.789 (0.332−1.874) 2–3 17/123 (13.8) 20/119 (16.8) 0.733 (0.384−1.400) 4–7 12/77 (15.6) 25/73 (34.2) 0.396 (0.199−0.789)

Baseline cancer stage Stage III 14/99 (14.1) 24/115 (20.9) 0.640 (0.331−1.237) Stage IV 24/150 (16.0) 33/132 (25.0) 0.554 (0.327−0.937)

Baseline B symptoms Present 24/146 (16.4) 32/139 (23.0) 0.664 (0.391−1.127) Absent 14/104 (13.5) 25/108 (23.1) 0.509 (0.265−0.980)

Baseline extra nodal sites 0 12/74 (16.2) 24/92 (26.1) 0.556 (0.278−1.112) 1 12/74 (16.2) 11/79 (13.9) 0.966 (0.426−2.190) >1 14/78 (17.9) 20/ 63 (31.7) 0.569 (0.287−1.127)

Baseline ECOG status 0 16/118 (13.6) 24/118 (20.3) 0.618 (0.328−1.163) 1 19/121 (15.7) 30/118 (25.4) 0.554 (0.312−0.984) 2 3/ 11 (27.3) 3/11 (27.3) 0.692 (0.139−3.451)

Gender Male 20/146 (13.7) 34/148 (23.0) 0.537 (0.309−0.934) Female 18/104 (17.3) 23/ 99 (23.2) 0.679 (0.366−1.258)

0.1 0.5 1 Favors A+AVD HR Favors ABVD Subgroups analyses presented are those that were conducted for the primary analysis

Figure 2. Forest plot of modiﬁed PFS per IRF by baseline risk factor subgroups.

Table 3. Dose modiﬁcation, delay, or discontinuation of study drug in North America AþAVD ABVD (n ¼ 249) (n ¼ 240) Action, n (%) BV Doxorubicin Vinblastine Dacarbazine Bleomycin Doxorubicin Vinblastine Dacarbazine No action taken 98 (39) 144 (58) 134 (54) 147 (59) 114 (48) 156 (65) 149 (62) 158 (66) Action on study druga 151 (61) 105 (42) 115 (46) 102 (41) 126 (53) 84 (35) 91 (38) 82 (34) Dose reduced 79 (32) 9 (4) 26 (10) 10 (4) 2 (<1) 8 (3) 20 (8) 6 (3) Dose heldb 28 (11) 1 (<1) 4 (2) 0 22 (9) 1 (<1) 4 (2) 1 (<1) Dose delayedc 86 (35) 93 (37) 92 (37) 88 (35) 65 (27) 70 (29) 73 (30) 70 (29) Dose discontinued 30 (12) 14 (6) 18 (7) 14 (6) 63 (26) 4 (2) 9 (4) 4 (2) No alternative frontline regimen 243 243 243 243 238 238 238 238 6 cycles randomized regimen 175 (70) 213 (86) 207 (83) 216 (87) 150 (63) 215 (90) 211 (88) 212 (88) Abbreviation: BV, brentuximab vedotin. aSubjects recording the same action on study drug multiple times within a cycle or overall will be counted only once in the respective summary. bHold indicates a dose that was skipped before restarting in the subsequent treatment round. cDelay indicates doses administered after the protocol-speciﬁed treatment window for a cycle.

OF6 Clin Cancer Res; 2019 Clinical Cancer Research

Brentuximab Vedotin for Hodgkin Lymphoma in North America

Table 4. Neutropenia, neuropathy, and pulmonary toxicity in North America context with regards to historical efficacy assessments. Thus, AþAVD ABVD an analysis of PFS, defined as time to disease progression or n n ¼ n ¼ Adverse event category, (%) ( 249) ( 240) death from any cause per investigator assessment, was per- a Neutropenia formed. Similar to the modified PFS results, a PFS difference of Incidence of neutropenia (any grade) 154 (62) 130 (54) P ¼ Grade 3 or higher neutropenia 146 (59) 109 (45) 11.7 percentage points at 2 years and a HR of 0.50 ( 0.002) þ Incidence of febrile neutropeniab 51 (20) 22 (9) in favor of A AVD was observed among subjects treated Peripheral neuropathyc in NA. Although differences in response across regions were Incidence of peripheral neuropathy (any grade) 198 (80) 134 (56) noted, these differences were not significant from those observ- Grade 1 peripheral neuropathy 102 (41) 104 (43) ed in NA. Grade 2 peripheral neuropathy 53 (21) 28 (12) In the primary analysis of the ECHELON-1 trial, there was an Grade 3 peripheral neuropathy 43 (17) 2 (<1) Pulmonary toxicityd increased incidence of febrile neutropenia and peripheral neu- þ Incidence of pulmonary toxicity (any grade) 7 (3) 25 (10) ropathy in the A AVD arm, and an increase in pulmonary- Grade 3 or higher pulmonary toxicity 4 (2) 14 (6) related toxicity in the ABVD arm. In the NA region, febrile aNeutropenia includes preferred terms of neutropenia and neutrophil count neutropenia rates were similar to those observed in the entire decreased. safety population. Rates in the AþAVD arm were reduced and bA total of 35 subjects received AþAVD and G-CSF primary prophylaxis. Of comparable (9%) with the rates in the ABVD population among these, 3 subjects (9%) experienced febrile neutropenia. the 35 subjects in the AþAVDarmwhoreceivedG-CSFprimary cIncludes the preferred terms of peripheral motor neuropathy, peripheral prophylaxis beginning with cycle 1. These results highlight the sensorimotor neuropathy, peroneal nerve palsy, muscular weakness, hypotonia, or muscle atrophy. importance of G-CSF primary prophylaxis for all subjects treated dIncludes adverse events in the Interstitial lung disease Standardized MedDRA with AþAVD. Query, MedDRA dictionary Version 19.0. The incidence of peripheral neuropathy was higher in the NA subset than in Europe for both treatment arms. Per the ECHELON-1 protocol (Supplementary Table S7), management endpoint of the study, modified PFS per IRF, showed evidence of of grade 2 neuropathy required brentuximab vedotin dose benefit (HR of less than 1) across the majority of subgroups. The reduction, whereas grade 3 neuropathy required withholding analyses presented here further assessed the safety and efficacy brentuximab vedotin until symptoms improved to grade 2 or observed in subjects treated within the NA region of the phase III better, followed by dose reductions. Rates of resolution or ECHELON-1 trial. improvement by last follow-up visit were also higher in NA ECHELON-1 was prospectively designed to compare the (approximately 75% of subjects in both arms) than rates in the safety and efficacy of AþAVD to ABVD with subject random- overall safety population (approximately 66% in both arms). ization stratified by region, as geographic differences in treat- Recently, several HL trials have assessed the safety and effi- ment of patients with Hodgkin lymphoma (such as the appli- cacy of response-adapted strategies intended to maintain cation of consolidative radiotherapy) and access to supportive efficacy and minimize the risk of both short-term (i.e., bleo- care therapies vary globally. Regional variations in the out- mycin-induced pulmonary toxicity) and long-term toxicities comes of large multi-national randomized trials have been (i.e., secondary malignancies, infertility). Although direct reported in other settings (11–13). As baseline demographics comparisons between studies are difficult due to protocol and disease characteristics were comparable between the NA differences in subject population and treatment approaches, subset and other regions, other potential factors for the regional relating the data from ECHELON-1 in the context of the PET- differences need to be considered, including regional variation adapted RATHL trial may be an important consideration (14). in patient management. Dose reductions for brentuximab In addition to subjects with stage III or IV disease, RATHL also vedotin were more common in NA than in Europe or Asia; enrolled 42% of subjects with stage II disease, which differed however, dose delays were less common. In the ABVD arm, from the ECHELON-1 trial, which only enrolled subjects with bleomycin discontinuation was more common in NA com- stage III or IV disease. Subjects treated in RATHL received 2 pared with Europe and Asia. It is possible that these regional cycles of ABVD followed by deescalation to AVD versus differences in dosing of brentuximab vedotin and bleomycin continuation of ABVD for the PET2-negative patients or esca- may have contributed to the differential results in the NA lation to BEACOPP for the PET2-positive patients. Both ECH- subgroup. However, other factors including the potential for ELON-1 and RATHL reduced pulmonary toxicity by complete- population differences in drug metabolism and/or sensitivity ly eliminating or limiting the exposure to bleomycin. In the arepossiblealternative considerations. subset of the RATHL subjects with stage III or IV disease who For subjects treated in NA, modified PFS by IRF and inves- were 60 years old, the 3-year PFS rates for the PET2-negative tigator assessment demonstrated a consistent improvement in subjectswerethesamefortheAVDandABVDsubsetsat favor of AþAVD versus ABVD, with differences in 2-year rates 82.1%. For subjects who were PET-positive at cycle 2, the of 10.6 and 12.8 percentage points and corresponding HRs of 3-year PFS rate was 63.9% following BEACOPP. In the ECH- 0.60 (P ¼ 0.012) and 0.52 (P ¼ 0.002), respectively. For ELON-1 NA subset (which included 12.7% of subjects 60 patients with incomplete response at EOT, additional therapy years old), the 2-year PFS for PET2-negative AþAVD versus for active disease before radiographic evidence of progression ABVD subjects was 91.9% versus 81.1%, respectively; 2-year (new lesions or tumor increase >50%) was considered an event PFS for PET2-positive AþAVDversusABVDsubjectswas for the modified PFS endpoint. Thus, modified PFS may be a 54.5% and 48%, respectively. When focusing on PET2-nega- more clinically relevant surrogate for primary treatment failure. tive subjects, the ECHELON-1 NA subset demonstrated an An analysis of a more traditional PFS endpoint can provide increase in PET2-negative rate at 88% for AþAVD compared further confirmation of the primary endpoint and further with 83.7% in RATHL. Furthermore, this higher PET2-negative

www.aacrjournals.org Clin Cancer Res; 2019 OF7

Ramchandren et al.

rate in the ECHELON-1 NA subset was also paired with an L.I. Gordon reports receiving commercial research grants from Juno, and increase in 2-year PFS to 91.9%, which compares favorably to is a consultant/advisory board member for Juno and Bayer. J. Kuruvilla the 3-year PFS of 82.1% in RATHL for subjects treated with reports receiving speakers bureau honoraria from Seattle Genetics, Bristol- Myers Squibb, Merck, Roche, Amgen, Astra Zeneca, Janssen, and Karyo- ABVD or ABVD deescalated to AVD. Longer follow-up will pharm, and is a consultant/advisory board member for Seattle Genetics, determine if the 2-year estimates remain stable at later time Bristol-Myers Squibb, Merck, Roche, Abbvie, Gilead, Janssen, and Karyo- points. Overall, the ECHELON-1 NA subset data supports that pharm. K.J. Savage is a consultant/advisory board member for Seattle AþAVD compares favorably with the treatment approach of Genetics, Bristol-Myers Squibb, Merck, Verastem, and Abbvie. A. Younes RATHL without the need for additional treatment decisions reports receiving speakers bureau honoraria from Takeda, Roche, Epizyme, (including escalation to BEACOPP) predicated on real-time Bristol-Myers Squibb, Merck, Curis, Janssen, Accerta, and Abbvie, and is a consultant/advisory board member for Xynomics. G. Engley, T.J. Manley, PET2 disease assessment, which can be challenging in certain and K. Fenton hold ownership interest (including patents) in Seattle clinical settings. Genetics. D.J. Straus reports receiving speakers bureau honoraria from Identifying specific drivers of regional variation in randomized Oncotracker and Medical Crossfire, and is a consultant/advisory board multinational trials has often proven challenging. Trials are typ- member for Seattle Genetics and InPractice (Elsevier). No potential conflicts ically not designed with an objective of detecting potential of interest were disclosed by the other authors. regional differences in efficacy; however, results from specific regions help establish the reliability of primary results and are Authors' Contributions important to better inform patient care decisions made within the Conception and design: S.M. Ansell, R. Chen, J.M. Connors, J.W. Friedberg, regions of interest. Furthermore, the availability of this regional A.K. Gopal, A. Younes, G. Engley, T.J. Manley data will better inform cost-effectiveness analysis models, where Development of methodology: S.M. Ansell, R. Chen, J.M. Connors, A.K. Gopal, understanding regional safety and efficacy results, as well as G. Engley Acquisition of data (provided animals, acquired and managed patients, transitions in care, are important modeling considerations. fi provided facilities, etc.): R. Ramchandren, R.H. Advani, N.L. Bartlett, Although the improvements in modi ed PFS and PFS were not R. Chen, J.M. Connors, T. Feldman, A. Forero-Torres, A.K. Gopal, as pronounced in Europe and Asia when compared with NA, it is J. Kuruvilla, K.J. Savage, A. Younes, D.J. Straus notable, that there were more brentuximab vedotin and bleomy- Analysis and interpretation of data (e.g., statistical analysis, biostatistics, cin dose delays and less bleomycin dose discontinuations in computational analysis): R. Ramchandren, R.H. Advani, S.M. Ansell, Europe and Asia, which may be important considerations. N.L. Bartlett, R. Chen, J.M. Connors, A.K. Gopal, J. Kuruvilla, K.J. Savage, G. Engley, T.J. Manley, K. Fenton, D.J. Straus The results of the ECHELON-1 NA subgroup analyses are consis- fi Writing, review, and/or revision of the manuscript: R. Ramchandren, tent with those of the primary analyses, indicating a bene t R.H. Advani, S.M. Ansell, N.L. Bartlett, R. Chen, J.M. Connors, A. Forero- in the subjects who received AþAVD compared with those who Torres, J.W. Friedberg, A.K. Gopal, L.I. Gordon, J. Kuruvilla, K.J. Savage, received ABVD. G. Engley, T.J. Manley, K. Fenton, D.J. Straus Administrative, technical, or material support (i.e., reporting or organizing Disclosure of Potential Conflicts of Interest data, constructing databases): T. Feldman Study supervision: R. Ramchandren, R.H. Advani, S.M. Ansell, A. Younes, R. Ramchandren and N.L. Bartlett is a consultant/advisory board member G. Engley for Seattle Genetics. R.H. Advani is a consultant/advisory board member for Seattle Genetics and Takeda. R. Chen and J.M. Connors report receiving speakers bureau honoraria from and are consultant/advisory board mem- Acknowledgments bers for Seattle Genetics. T. Feldman reports receiving commercial research The authors wish to acknowledge Matt Blahna of Seattle Genetics, Inc. and grants from Amgen, Bristol-Myers Squibb, Cell Medica, Eisai, Genentech, Elizabeth O'Connor of MMS Holdings, Inc. for assistance in manuscript Idera Pharmaceutical, Janssen, Kyowa, Millennium, Pfizer, Portola, Roche, preparation. Research support was provided by Millennium Pharmaceuticals, Seattle Genetics, Viracta, and Trillium, speakers bureau honoraria from Inc., a wholly owned subsidiary of Takeda Pharmaceutical Company Limited, Takeda (Millenium), Seattle Genetics, Abbvie, Pharmacyclics, Janssen, and Seattle Genetics. Celgene, and Kite Pharma, is a consultant/advisory board member for Bristol-Myers Squibb and Seattle Genetics. A. Forero is an employee of The costs of publication of this article were defrayed in part by the payment of and holds ownership interest (including patents) in Seattle Genetics. page charges. This article must therefore be hereby marked advertisement in J.W. Friedberg is a consultant/advisory board member for Bayer and Astellas. accordance with 18 U.S.C. Section 1734 solely to indicate this fact. A.K. Gopal reports receiving commercial research grants from Seattle Genetics, Teva, Bristol-Myers Squibb, Pfizer, Merck, Takeda, Janssen, and Gilead, and is a consultant/advisory board member for Seattle Genetics, Received September 4, 2018; revised November 12, 2018; accepted January 4, Janssen, Takeda, Pfizer, Amgen, Sanofi, Gilead, Acerta, and Aptevo. 2019; published first January 7, 2019.

References 1. Siegel R, Ward E, Brawley O, Jemal A. Cancer statistics, 2011: the impact of bleomycin, vinblastine, and imidazole carboxamide versus MOPP. eliminating socioeconomic and racial disparities on premature cancer Cancer 1975;36:252–9. deaths. CA Cancer J Clin 2011;61:212–36. 5. Gordon LI, Hong F, Fisher RI, Bartlett NL, Connors JM, Gascoyne RD, 2. Carde P, Johnson P. Principles of chemotherapy in Hodgkin lymphoma. et al. Randomized phase III trial of ABVD versus Stanford V with or In: Engert A, Horning SJ, editors. Hodgkin lymphoma a comprehensive without radiation therapy in locally extensive and advanced-stage update on diagnostics and clinics. Berlin/Heidelberg, Germany: Springer- Hodgkin lymphoma: an intergroup study coordinated by the Eastern Verlag; 2011. Cooperative Oncology Group (E2496). J Clin Oncol 2013;31:684–91. 3. NCCN. RETIRED NCCN Guidelines Version 1.2017: Hodgkin Lympho- 6. Canellos GP, Anderson JR, Propert KJ, Nissen N, Cooper MR, Hender- ma. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines). son ES, et al. Chemotherapy of advanced Hodgkin's disease with MOPP, Plymouth Meeting, PA: National Comprehensive Cancer Network; 2017. ABVD, or MOPP alternating with ABVD. N Engl J Med 1992;327: p1–83. 1478–84. 4. Bonadonna G, Zucali R, Monfardini S, De Lena M, Uslenghi C. 7. Carde P, Karrasch M, Fortpied C, Brice P, Khaled H, Casasnovas O, Combination chemotherapy of Hodgkin's disease with adriamycin, et al. Eight cycles of ABVD versus four cycles of BEACOPPescalated

OF8 Clin Cancer Res; 2019 Clinical Cancer Research

Brentuximab Vedotin for Hodgkin Lymphoma in North America

plus four cycles of BEACOPPbaseline in stage III to IV, international tional, phase 3 trial of decitabine vs treatment choice of supportive care prognostic score >/¼3, high-risk Hodgkin lymphoma: first results of or cytarabine in older patients with newly diagnosed acute myeloid the phase III EORTC 20012 Intergroup trial. J Clin Oncol 2016;34: leukemia and poor- or intermediate-risk cytogenetics. BMC Cancer 2028–36. 2014;14:69. 8. Connors JM, Jurczak W, Straus DJ, Ansell SM, Kim WS, Gallamini A, et al. 12. Ohtsu A, Shah MA, Van Cutsem E, Rha SY, Sawaki A, Park SR, et al. Brentuximab vedotin with chemotherapy for stage III or IV Hodgkin's Bevacizumab in combination with chemotherapy as first-line therapy in lymphoma. N Engl J Med 2018;378:331–44. advanced gastric cancer: a randomized, double-blind, placebo-controlled 9. Campo E, Swerdlow SH, Harris NL, Pileri S, Stein H, Jaffe ES. The phase III study. J Clin Oncol 2011;29:3968–76. 2008 WHO classification of lymphoid neoplasms and beyond: 13. Swain SM, Im YH, Im SA, Chan V, Miles D, Knott A, et al. Safety profile of evolving concepts and practical applications. Blood 2011;117: pertuzumab with trastuzumab and docetaxel in patients from Asia with 5019–32. human epidermal growth factor receptor 2-positive metastatic breast 10. Cheson BD, Pfistner B, Juweid ME, Gascoyne RD, Specht L, Horning SJ, cancer: results from the phase III trial CLEOPATRA. Oncologist 2014; et al. Revised response criteria for malignant lymphoma. J Clin Oncol 2007; 19:693–701. 25:579–86. 14. Johnson P, Federico M, Kirkwood A, Fossa A, Berkahn L, Carella A, et al. 11.MayerJ,ArthurC,DelaunayJ,Mazur G, Thomas XG, Wierzbowska A, Adapted treatment guided by interim PET-CT scan in advanced Hodgkin's et al. Multivariate and subgroup analyses of a randomized, multina- lymphoma. N Engl J Med 2016;374:2419–29.

www.aacrjournals.org Clin Cancer Res; 2019 OF9

Brentuximab Vedotin plus Chemotherapy in North American Subjects with Newly Diagnosed Stage III or IV Hodgkin Lymphoma

Radhakrishnan Ramchandren, Ranjana H. Advani, Stephen M. Ansell, et al.

Clin Cancer Res Published OnlineFirst January 7, 2019.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-18-2435

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2019/01/07/1078-0432.CCR-18-2435.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2019/02/03/1078-0432.CCR-18-2435. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.