Combining brentuximab vedotin with dexamethasone, high-dose cytarabine and cisplatin as salvage treatment in relapsed or refractory Hodgkin lymphoma: the phase II HOVON/LLPC Transplant BRaVE study

Abstract

Achieving a metabolic complete response (mCR) before high-dose chemotherapy (HDC) and autologous peripheral blood stem-cell transplant (auto-PBSCT) predicts progression free survival (PFS) in relapsed/refractory classical Hodgkin lymphoma (R/R cHL). We added brentuximab vedotin (BV) to DHAP to improve the mCR rate. In a Phase I dose-escalation part in 12 patients, we showed that BV-DHAP is feasible. This Phase II study included 55 R/R cHL patients (23 primary refractory). Treatment consisted of three 21-day cycles of BV 1.8 mg/kg on day 1, and DHAP (dexamethasone 40mg days 1-4, cisplatin 100mg/m2; day 1 and cytarabine 2x2g/m2; day 2). Patients with a metabolic partial response (mPR) or mCR proceeded to HDC/auto-PBSCT. Based on independent central FDG-PET-CT review, 42 of 52 evaluable patients (81% [95% CI: 67-90]) achieved an mCR before HDC/auto-PBSCT, five had an mPR and five had progressive disease (three were not evaluable). After HDC/auto-PBSCT, four patients with an mPR converted to an mCR. The 2-year PFS was 74% [95% CI: 63-86], and the overall survival 95% [95% CI: 90-100]. Toxicity was manageable and mainly consisted of grade 3/4 hematological toxicity, fever, nephrotoxicity, ototoxicity (grade 1/2) and transiently elevated liver enzymes during BV-DHAP. Eighteen patients developed new onset peripheral neuropathy (maximum grade 1/2) and all recovered. In conclusion, BV-DHAP is a very effective salvage regimen in R/R cHL patients, but patients should be monitored closely for toxicity. ClinicalTrials.gov identifier: NCT02280993.

Introduction

Salvage chemotherapy followed by high-dose chemotherapy (HDC) and autologous peripheral blood stem cell transplant (auto-PBSCT) has been the standard of care for patients with relapsed or refractory classical Hodgkin lymphoma (R/R cHL) for decades.^1,2With this treatment, cure rates of 40-60% can be achieved. Patients failing this treatment and those relapsing after second-line treatment generally have a very poor prognosis.^3-5

Response to salvage treatment is one of the most important predictors of outcome after auto-PBSCT, with metabolic active residual disease, as assessed by [¹⁸F]fluorodeoxyglucose (FDG) - positron emission tomography (PET) - computed tomography (CT) scan, before HDC/auto-PBSCT conferring an inferior prognosis.^6-8 Therefore, higher cure rates may be achieved by improving the metabolic complete response (mCR) rate before HDC/auto-PBSCT. Conventional salvage chemotherapy regimens result in mCR rates of about 50-60%.^6,9-11 DHAP (dexamethasone, high-dose cytarabine, cisplatin) is one of the most commonly used salvage regimens for R/R cHL in Europe.¹²

Brentuximab vedotin (BV) is targeted high-dose intracellular chemotherapy, consisting of an anti-CD30 antibody conjugated to the potent anti-microtubule agent monomethyl auristatin-E.^13,14 Several phase II studies have shown promising clinical activity of BV in R/R cHL, both as monotherapy and combined with chemotherapy.^15-20 Toxicities of BV include infusion-related reaction (IRR), myelosuppression, and peripheral neuropathy, the latter being reversible in the majority of patients.^{15,16,18,20,21}

In the current prospective, multicenter, international phase I/II Transplant BRaVE study we investigated the efficacy and safety of BV-DHAP followed by HDC (BEAM: carmustine, etoposide, cytarabine, melphalan) and auto-PBSCT in R/R cHL patients.

Results of the phase I part of this study in 12 patients have been published previously and showed that the combination of BV-DHAP is feasible with acceptable toxicity.²² The recommended dose level was established at full dose of all drugs with BV dosed at 1.8 mg/kg.²² The primary endpoints for the phase II single arm part were the fraction of patients achieving an mCR as judged by independent review of PET-CT scan after the third cycle of BV-DHAP, and the rate of grade 3/4 non-hematologic adverse events (AE), including neurotoxicity, during BV-DHAP.

Methods

Patients

The study enrolled patients aged ≥18 years with histologically confirmed CD30 positive cHL by local pathology assessment, either having primary refractory disease or a first relapse after first-line chemotherapy. Online Supplementary Table S1 shows the complete list of inclusion and exclusion criteria. Central pathology review was performed by two experienced hematopathologists (DDJ, AD).

All patients provided written informed consent. The study protocol was approved by the Ethical Review Committee (ERC) of all participating centers. The study was carried out in accordance with the principles of the Declaration of Helsinki.

Study design and treatment

Transplant BRaVE (clinicaltrials.gov identifier: NCT02280993) is a prospective, open-label study conducted at eight centers in the Netherlands (n=5), France (n=3) and Denmark (n=1). An independent Data Safety Monitoring Board (DSMB) evaluated the general progress and safety aspects of the study at predefined intervals.

Baseline assessment included a lymph node and bone marrow biopsy, and a PET-CT scan. Patients filled in a neurotoxicity questionnaire at study entry, prior to each cycle, and at three months after auto-PBSCT.

Patients were treated with three 21-day cycles of BV (1.8 mg/kg, i.v., day 1), dexamethasone (40 mg orally or intravenous [i.v.], days 1-4), cisplatin (100 mg/m², continuous i.v. [24 hours], day 1) and cytarabine (2x2 g/m² q12 hours, 3 hours for each infusion, day 2). After cycle 2, stem cells were mobilized and harvested using granulocyte colony-stimulating factor (G-CSF). A PET-CT scan was performed after cycle 3. Patients with progressive disease (PD) went off study, whereas patients with a partial response (mPR) or mCR proceeded to BEAM (carmustine, 300 mg/m², day -7; etoposide, 100 mg/m² and cytarabine, 100 mg/m², 2x/day, days -6, -5, -4 and -3; and melphalan, 140 mg/m², day -2), followed by auto- PBSCT (on day 0). Six weeks after auto-PBSCT, response evaluation was performed by PET-CT. G-CSF was recommended to prevent long-lasting neutropenia.

Endpoints

All endpoints and their definitions are described in Online Supplementary Table S2. Responses were determined according to the 2014 Lugano criteria.²³ All PET-CT scans were centrally reviewed by two independent nuclear medicine physicians (AA, RV) and a third adjudicator (OH) in case of discrepancies. Visual assessment was performed using the Deauville score (DS), assessing DS1-3 as mCR. Toxicity was reported according to the Common Terminology Criteria for Adverse Events (CTCAE) version 4.03.

Statistical analysis

Details about the study design and statistical analysis are provided in Online Supplementary Appendix 1. Efficacy analysis was performed among all evaluable patients. Primary safety analysis was performed among all patients who received at least one dose of study medication. Response rates and their corresponding 95% two-sided exact confidence intervals (CI) were calculated. AE were analyzed descriptively. The Kaplan-Meier method was used for time-to-event analysis. An exploratory analysis with a Cox proportional hazards regression was performed on all phase II patients, and six patients from the phase I part of the study who were treated at the recommended dose level. The Kaplan-Meier method and log-rank test were used to analyze univariable associations with progression-free survival (PFS). All statistical analyses were performed using R software version 3.6.1 and SAS software version 9.4.

Results

Patients and treatment

Between May 2014 and July 2017, a total of 67 patients with R/R cHL were enrolled for the entire Transplant BRaVE phase I/II study (n=12 in phase I and n=55 in phase II). Due to withdrawal of consent of two patients after one cycle of BV-DHAP and three patients not completing all BV-DHAP cycles, five more patients were enrolled in phase II than planned according to the sample size calculations (n=50) to ensure a sufficient number of evaluable patients in the primary analysis.

Patients' characteristics for the phase II patients are summarized in Table 1. Median age was 29 years, and 27 patients were female (49%). Twenty-three patients (43%) had primary refractory disease and 16 patients (29%) had relapsed within one year of first-line treatment. Among the first 20 patients of phase II (stage 1), enough responses were observed (16 mCR) with acceptable toxicity (7 patients experienced significant toxicity) for the DSMB to approve proceeding to stage 2.

Of the 55 enrolled patients, 49 (89%) completed all three cycles of BV-DHAP, and 47 (85%) underwent BEAM and auto-PBSCT (Figure 1). Two patients withdrew consent after cycle 1 due to psychological issues, and two patients had PD after cycle 2. In cycle 3, two patients did not receive BV due to toxicity. One of these patients received VIM (ifosfamide, mitoxantrone and etoposide) in cycle 3 because of hepatotoxicity and was not evaluable for response. However, this patient still proceeded successfully to BEAM and auto-PBSCT. The other patient received DHAP without BV because of an anaphylactic shock following BV infusion in cycle 2. This patient went off study thereafter because of toxicity and a mixed response by local PET-CT assessment (which was eventually considered mCR by central PET-CT review) and proceeded to auto-PBSCT after additional treatment with miniBEAM.

Besides the two patients who did not receive BV in cycle 3, dose reductions or delays included three delays of cycle 2 due to infection (n=1), venous thrombosis (n=1), or neutropenia (n=1), and three delays of BV infusion due to IRR (grade 1/2). Cycle 3 was delayed in two patients (malaise and neutropenia), and there were two delays of BV infusion (IRR: one grade 2 and one grade 3). Furthermore, eight patients switched from cisplatin to carboplatin due to ototoxicity (n=7; grade 1/2) or nephrotoxicity (n=1; grade 3, recovered completely), and one patient received no cisplatin and cytarabine in cycle 3 due to electrolyte disorder and sepsis.

Efficacy and stem cell harvest

Three patients were not evaluable for response after three cycles of BV-DHAP: two patients withdrew consent after cycle 1, and one patient did not have a PET-CT scan after cycle 3. By independent central PET-CT review, 42 of 52 evaluable patients achieved an mCR (81% [95%CI: 67-90]) and five patients an mPR (10%), resulting in an overall response rate of 90% (95%CI: 79-97). A total of five patients had PD (10%) and did not proceed to BEAM. Two of those patients showed PD on a CT scan after cycle 2 and three had PD on the PET-CT scan after cycle 3 (Figure 1). After auto-PBSCT, four out of five patients with mPR converted to mCR. One patient had a persisting mPR and received additional radiotherapy according to the local physician’s decision, and is still in mCR thereafter.

There were no significant differences in baseline characteristics (i.e., age, time to relapse and first-line treatment) between patients with mCR or mPR. The mCR rate was lower for patients with primary refractory disease compared to patients with a later relapse, but this was not statistically significant (mCR rate 73% [95%CI: 69-96] vs. 86% [95%CI: 50-89]; P=0.29, respectively).

Stem cell harvest after cycle 2 was successful using GCSF in all patients, with one apheresis session in 43 patients and two apheresis sessions in nine patients, of whom two patients received plerixafor (3 patients went off study before apheresis). The median yield was 5.3x10⁶ CD34⁺/kg (range: 1.8-22.7).

Table 1.Patients' characteristics.

Safety

During BV-DHAP treatment, 20 patients (36%) experienced one or more AE that met the dose-limiting toxicity criteria (considered significant toxicity).

Grade 3/4 neutropenia and thrombocytopenia were common (Online Supplementary Table S3). After BEAM/auto-PBSCT, the median recovery time to an absolute neutrophil count (ANC) ≥0.5x10⁹/L was 12 days (range: 8-29), and the median recovery time to platelets ≥20x10⁹/L was 15 days (range: 6-46) (Online Supplementary Table S3).

During BV-DHAP, febrile neutropenia (n=14) was the most common non-hematologic grade 3/4 toxicity, followed by elevated liver enzymes (n=10) and electrolyte disorders (n=6) (Table 2). After BEAM/auto-PBSCT, one patient developed veno-occlusive disease (VOD) that was fatal. This patient already had elevated levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST) and gamma-glutamyl transferase (GGT) during BV-DHAP and very high levels of AST (2400 U/L), ALT (970 U/L), lactate dehydrogenase (LDH; 1,400 U/L), GGT (900 U/L) and direct bilirubin (660 mol/L) during the VOD after BEAM/auto-PBSCT.

Figure 1.Consort diagram. Number of patients in the full analysis set going through the protocol treatment including reasons for exclusion. BV: brentuximab vedotin; DHAP: dexamethasone, high-dose cytarabine, cisplatin; C: cycle; N: number; CT: computed tomography; SC: stem cell; PD: progressive disease; VIM: ifosfamide, mitoxantrone, etoposide; PET: positron emission tomography; mCR: metabolic complete response; BEAM: carmustine, etoposide, cytarabine, melphalan; auto-PBSCT: autologous peripheral blood stem-cell transplant.

Peripheral neuropathy grade 1/2 was present before study entry in 11 patients (n=1 grade 2) but did not worsen during BV-DHAP treatment. During BV-DHAP treatment, 15 (27%) and three (5%) patients developed novel onset grade 1 and 2 peripheral neuropathy, respectively, but all recovered. Of all patients, regardless of the presence of peripheral neuropathy at baseline, 12 patients reported transient muscle weakness (grade 1/2) in the neurotoxicity questionnaire, of whom 11 recovered without sequelae. No grade 3/4 neuropathy has occurred (Online Supplementary Table S4).

In total, seven patients experienced ototoxicity (n=3 grade 1, n=4 grade 2) and switched from cisplatin to carboplatin in cycle 2 or 3. Three patients recovered without sequelae, and three patients had continuing ototoxicity (hearing loss or tinnitus) 6 months after auto-PBSCT (1 patient unknown).

Serious AE (SAE) grade 3/4 following BV-DHAP treatment are described in Table 3. In total, 18 (33%) patients experienced one or more SAE during BV-DHAP. SAE that occurred in more than one patient were febrile neutropenia (n=9), infections (n=2), and renal function disorder (n=2). Most SAE recovered, except for the two renal function disorders which recovered with sequelae (persisting grade 1 or 2 nephrotoxicity, e.g., decreased glomerular filtration rate or persisting high levels of creatinine). One additional nephrotoxicity grade 3 was not considered an SAE because of rapid recovery without hospitalization.

Survival

After a median follow-up of 27 months, the 2-year PFS by intention-to-treat for all 55 patients was 73.5% (95%CI: 62.6-86.4) (events=14/55) and the 2-year overall survival (OS) was 94.9% (95%CI: 89.5-00.0) (events=3/55) (Figure 2A and B).

Three patients died during the study period: one patient died of encephalitis (exact cause remained unknown despite a brain autopsy, the patient did not recover from seizures; brain autopsy did not show cerebral localization of lymphoma or infection), and one patient died due to VOD. Both deaths occurred within four months after BEAM/auto-PBSCT. The third patient died of an unrelated head trauma, nine months after BEAM/auto-PBSCT while in mCR. One patient who withdrew consent after cycle 1 went off study and later died from PD; this patient was censored at the time of withdrawal of consent.

Patients with progression after treatment in this study received salvage treatment according to the treating physician’s choice. Four patients received BV monotherapy, two of whom had a complete response, but all progressed again and needed a third salvage regimen.

Exploratory analysis of survival

For an exploratory analysis of PFS, six patients from phase I who were treated at the recommended dose level were added to the analysis, to a total of 61 patients.²²

Patients with mPR after three cycles showed a significantly lower PFS compared to patients with mCR. Twoyear PFS rates of patients with mPR (n=5) versus patients with mCR (n=48) were 40% (95%CI: 14-100) versus 87% (95%CI: 78-97): log-rank P=0.004, hazard ratio (HR) 6.02 (95%CI: 1.50-24.2; P=0.011), respectively (Figure 3A and Online Supplementary Table S5). A multivariable Cox analysis showed that patients with an mPR had a significantly increased risk of progression, independently of primary refractory status (Online Supplementary Table S5). Patients with relapsed disease (n=37) had a lower risk of progression compared to patients with primary refractory disease (n=24), with 2-year PFS rates of 86% (95%CI: 75-98) versus 63% (95%CI: 46-85): log-rank P=0.036, HR 0.33 (95%CI: 0.11-0.98; P=0.046), respectively (Figure 3B and Online Supplementary Table S5). Univariable analysis did not show significant associations for other baseline risk factors (i.e., B symptoms, age, stage and first-line treatment regimen) (Online Supplementary Table S5).

Central pathology review

Based on morphology, immunophenotype, and molecular clonality analysis if needed, central pathology review confirmed cHL (according to the World Health Organization classification 2016²⁴) in 59 of all 67 patients (88%) of the complete phase I (cHL confirmed in 10 of 12 patients in total) and phase II (cHL confirmed in 49 of 55 patients in total) part of the study. In all cases with equivocal morphological and/or immunohistochemical features, including cases with high numbers of Epstein Barr virusencoded RNA (EBER)-positive atypical large cells and/or small lymphocytes (n=16), extensive immunohistochemical and molecular T-cell receptor and immunoglobulin heavy and light chain gene rearrangement assays (BIOMED) were performed (Online Supplementary Table S6). In eight patients, cHL could not be confirmed. Of these, five patients were diagnosed with peripheral T-cell lymphoma (PTCL) not otherwise specified (NOS), one patient with angioimmunoblastic T-cell lymphoma (AITL), and one patient with immunodeficiency-associated B-lymphoproliferative disorder (IA-B-LPD).²⁵ In one patient, a classifying diagnosis could not be made due to lack of representative material in the biopsy sample. Additionally, in one patient, a composite lymphoma of cHL and lymphoplasmacytic lymphoma (LPL) was diagnosed. In all cases, high CD30 expression was present. Of the seven patients with PTCL, AITL or IA-B-LPD, six had an mCR after three cycles of BVDHAP. One patient with PTCL had PD after cycle 2, one with AITL had PD after auto-PBSCT, and one patient with PTCL died due to unrelated head-trauma. When excluding the patient with unrelated death, there was no significant difference in PFS between patients with confirmed cHL versus patients with another diagnosis (2-year PFS 81% vs. 67%, log-rank P=0.36).

Figure 2.Kaplan-Meier survival analysis. Kaplan-Meier survival analysis for all 55 phase II patients by intention-to-treat, including the number of patients at risk at 1, 2 and 3 years with regard to (A) progression-free survival and (B) overall survival, measured from enrollment.

Discussion

In this international, prospective phase II study, we investigated the efficacy and safety of BV-DHAP as first salvage treatment for patients with R/R cHL. This study is the first to investigate this combination. Treatment with BV-DHAP resulted in a high proportion of patients with an mCR prior to HDC/auto-PBSCT, and toxicity was mostly reversible.

Data on FDG-PET-CT results following treatment with DHAP are scarce, but generally only about 25% of patients achieved a CR as assessed by CT scan.^4,26 Other trials have recently investigated BV in combination with other salvage chemotherapy combinations, such as bendamustine, ICE (ifosfamide, carboplatin, etoposide) or ESHAP (etoposide, methylprednisolone, high-dose cytarabine and cisplatin), and have shown mCR rates up to 76% prior to HDC/auto-PBSCT.^15-18,27 The administration schedule of BV differed among these studies, and most studies used more than three administrations of BV in total.^15-18,27 In the current study, three cycles of BV-DHAP resulted in a high mCR rate with only three administrations of BV. This makes it a less ‘financially toxic’ therapy than using BV in first line for all patients or to use it as consolidation therapy after auto-PBSCT.

In R/R cHL patients treated with salvage chemotherapy followed by HDC and auto-PBSCT, historical studies demonstrate a 5-year PFS of approximately 50%.^1-4,26,28,29 In 97 patients treated with ICE, 2-year event-free survival was 70%.⁶ Another regimen consisting of bendamustine, gemcitabine and vinorelbine (in 59 patients) resulted in a 2- year PFS of 63%.³⁰ With the present treatment protocol, we have been able to achieve a high 2-year PFS rate of 74%. A total of 14 events occurred (including 3 deaths), and at the present median follow-up of 27 months, no relapses have occurred beyond 18 months from enrollment. Longer follow- up is needed to confirm that the majority of patients in remission after 2 years are indeed cured.^3,28,31

The unprecedented high response rate and prolonged PFS of this treatment regimen were achieved at the cost of higher toxicity in comparison to other salvage regimens. However, most of the observed toxicities, including neutropenia, thrombocytopenia, fever, nausea/vomiting, ototoxicity and nephrotoxicity are toxicities of specific concern during treatment with DHAP.^4,26,32,33 Other regimens of BV with bendamustine, nivolumab, ICE or ESHAP seem to induce fewer AE, with most toxicities consisting of hematologic toxicity.^{15,16,18,19,34} While the occurrence of grade 3/4 non-hematologic toxicity was low with BV-bendamustine, a substantial proportion of the patients (25%) did not undergo auto-PBSCT, resulting in a lower 2-year PFS of 62.6%.¹⁶ Another recent study with BV-bendamustine in 40 patients had a 3-year PFS of 67.3%, and 82.5% of patients underwent auto-PBSCT.¹⁹ The combination of BV with nivolumab resulted in an mCR rate of 61% with almost all patients experiencing grade 1/2 toxicity and 31% having grade 3/4 toxicity; however, these AE were also manageable.³⁴

Table 2.Adverse events grade 3 or 4 during brentuximab vedotin and dexamethasone, high-dose cytarabine and cisplatin, high-dose cytarabine, cisplatin.

A sequential approach of BV monotherapy followed by chemotherapy in PET-positive patients is interesting, since some patients could be spared the toxicity of salvage chemotherapy without losing efficacy. However, only a minority of patients achieved a PET-negative response after BV monotherapy.¹⁵ The ESHAP regimen is similar to DHAP, except that it contains methylprednisolone instead of dexamethasone, and cisplatin is given over 4 days of 25 mg/m²/day compared to 100 mg/m² in one day with the DHAP regimen.¹⁸ Hematologic toxicity was comparable between BV-ESHAP and BV-DHAP with about 50% of patients experiencing grade 3/4 thrombocytopenia and neutropenia. For BV-ESHAP, grade 3 fever and mucositis were the most frequent non-hematologic grade 3/4 toxicities whereas DHAP was also associated with fever, but not with mucositis. In contrast, only grade 1/2 renal dysfunction occurred with BV-ESHAP, and no cases of elevated liver enzymes or ototoxicity are described.¹⁸

In ten patients, a transient grade 3/4 increase in liver enzymes was observed during BV-DHAP treatment (n=1 grade 4), which was reversible in all patients. One patient developed a fatal VOD after BEAM/auto-PBSCT. Additionally, one patient treated in the phase I part of this study also developed a grade 3 VOD, which, however, recovered without sequelae. Both patients already had elevated liver enzymes during BV-DHAP treatment. This complication has previously been described in patients receiving high-dose alkylating agents such as melphalan or cyclophosphamide.³⁵

BV as consolidation therapy has been shown to prolong PFS in high-risk R/R cHL patients who have undergone HDC/auto-PBSCT.³⁶ Whether BV before auto-PBSCT in combination with chemotherapy, or as consolidation after auto-PBSCT will be more effective is unknown. Of note, with BV consolidation, peripheral neuropathy occurred in 67% of patients, including 13% (n=22) grade 3 peripheral neuropathy. With BV-DHAP, the incidence of peripheral neuropathy was lower, was mostly reversible, and no grade 3/4 occurred, probably because only three administrations of BV were given.

In-depth pathology workup and reclassification were performed to exclude lymphomas that are known as cHL mimickers such as AITL and PTCL (with follicular helper T-cell immunophenotype with secondary cHL-like blasts), as well as immunodeficiency-associated B-cell lymphoproliferative disorders (IA-B-LPD).^37-39 In retrospect, seven cases were identified as cHL-mimickers with central pathology review. Awareness for cHL-mimickers is important because patients with T-cell lymphoma generally have a worse prognosis.⁴⁰ In this cohort of patients, no significant differences in response rates or PFS were observed between patients with confirmed or unconfirmed cHL, although the number of patients is too small to validate this finding.

Figure 3.Kaplan-Meier exploratory analysis. Kaplan-Meier exploratory analysis for all 55 phase II patients and six patients from phase I who were treated at the same dose level, including the number of patients at risk at 1, 2 and 3 years with regard to (A) progression-free survival (PFS) stratified for patients with a metabolic complete response (mCR; n=48) or partial response (mPR; n=5) on the positron emission tomography-computed tomography (PET-CT) scan after three cycles of BVDHAP, measured from the time of that PET-CT scan, and (B) PFS stratified for relapsed patients (n=37; defined as recurrent disease after having reached a complete response on first-line treatment) versus patients with primary refractory disease (n=24; no complete response on first-line treatment), measured from enrollment. BV: brentuximab vedotin; DHAP: dexamethasone, high-dose cytarabine, cisplatin.

Table 3.Serious adverse events grade 3 or 4 during brentuximab vedotin and dexamethasone, high-dose cytarabine and cisplatin.

An exploratory analysis on PFS showed that patients with an mPR prior to BEAM/auto-PBSCT have a higher risk of relapse, despite conversion to an mCR after auto- PBSCT. This finding is in line with other trials investigating risk factors for relapse after auto-PBSCT.^5-7 PET-adapted therapy could probably further improve outcome by intensifying treatment for high-risk patients with new agents, such as checkpoint inhibitors in addition to BV. Moreover, a group of patients at low-risk for relapse, might possibly be cured with a combination of new drugs only, without the toxic consequences of HDC and auto-PBSCT. Risk stratification based on the PET-CT scan at relapse could also be further improved by quantitative analysis and the assessment of metabolic tumor volume.^41,42

The addition of BV to salvage treatment has not yet been investigated in a randomized phase III trial. However, several phase II studies have now shown that BV in combination with chemotherapy results in high mCR rates prior to HDC/auto-PBSCT. A combined pooled analysis of all of these studies is planned to give more insight into the effect of BV on response rates and toxicity in this setting.

In conclusion, in R/R cHL, three cycles of BV-DHAP is a highly effective salvage regimen resulting in an mCR rate of 81% prior to HDC/auto-PBSCT as shown by independent central PET-CT review. Patients should be monitored closely for toxicity, especially hematologic toxicity, nephrotoxicity and liver toxicity.

Footnotes

• Received November 18, 2019
• Accepted March 19, 2020

Correspondence

*MJK and JD contributed equally as co-first authors

MARIE JOSÉ KERSTEN m.j.kersten@amsterdamumc.nl

Disclosures

The study drug (BV) was provided for the study and the study was funded by Takeda. Takeda did not have any influence on the analysis of the data or the interpretation of the results. AH received consultancy fees, honoraria, and research funding from Millennium/Takeda. MJK: Millennium/Takeda: Honoraria, Research Funding; Celgene: Honoraria, Research Funding; Roche: Honoraria, Research Funding; Gilead: Honoraria; Kite Pharma: Honoraria; Novartis: Honoraria. FM: Janssen: Scientific Lectures; BMS: Membership on an entity's Board of Directors or advisory committees; Epizyme: Consultancy; Gilead: Consultancy, Membership on an entity's Board of Directors or advisory committees; Roche: Membership on an entity's Board of Directors or advisory committees; Celgene: Consultancy, Membership on an entity's Board of Directors or advisory committees. PJL: Millennium/Takeda: Consultancy, Research Funding; Servier: Consultancy, Research Funding; Roche: Consultancy; BMS: Consultancy; Sandoz: Consultancy; Genmab: Consultancy. AD: Millennium/Takeda: Consultancy, Honoraria, Research Funding. PB: Millennium/Takeda: Honoraria, Research Funding, Scientific Advisory Board; Roche: Honoraria; BMS: Honoraria, Scientific Advisory Board; MSD: Honoraria, Scientific Advisory Board; Jansen: Honoraria. MH: Consultant/advisor: Roche, Takeda, Celgene, Genmab; Research support: Roche, Takeda, Celgene, Genmab, Novartis, Janssen, Incyte, Genentech. TG: Millennium/Takeda: Honoraria, Gilead, Roche, MSD. JMZ: Consultant/advisor: Gilead, Roche, Takeda; Honoraria: Gilead, Roche, Takeda, Janssen. DDJ: Consultant/advisor: Takeda. All remaining authors have declared no conflicts of interest.

Contributions

MJK and AH designed the study; all authors collected the data; JD, MLY and HvT analyzed the data; JD and MJK wrote the manuscript with contributions from all authors, who also interpreted the data, read, commented on, and approved the final version of the manuscript; DdJ and AD performed the central pathology review; JZ, CB, AA and RV organized and performed the central FDG-PET-CT review; MJK and AH supervised the study.

Funding

This work was supported by research funding from Takeda.

Acknowledgments

The authors would like to thank all patients who participated in the trial, the Transplant BRaVE-trial team of the Trial Office of the Amsterdam UMC, location AMC, for their efforts in trial management and central data management, and the members of the Data Safety and Monitoring Board. The authors thank Marjolein Spiering, Edith van Dijkman, the data managers, trial nurses, lab and pharmacy personnel for their essential assistance with collecting and managing the study data. The authors thank Prof. Dr. Otto S. Hoekstra and Drs. Gerben J.C. Zwezerijnen for reviewing discrepancies in the central FDG-PET-CT review and Nathalie Hijmering, HOVON Pathology Facility and Biobank for biopsy collection and support of central pathology review.