Outcomes of patients with relapsed or refractory classic Hodgkin lymphoma after frontline brentuximab vedotin

Abstract

Frontline therapy for classic Hodgkin Lymphoma (cHL) incorporating brentuximab vedotin (BV) improves outcomes compared with traditional chemotherapy, but up to 20% of patients relapse and need salvage treatment. Prior retrospective studies examining salvage therapies are mostly limited to patients who received chemotherapy-based treatment in frontline without novel agents. We evaluated outcomes in 116 patients with relapsed/refractory (R/R) cHL who received brentuximab and anthracycline-containing frontline treatment. High risk factors at relapse or progression were common, including primary refractory disease (62%), advanced stage (63%), and extranodal disease (46%). At first salvage, 73% of patients received PD-1 blockade (58% in conjunction with chemotherapy), and 81% received PD-1 blockade at any salvage line. Overall, 78% of patients proceeded to ASCT. With a median follow-up of 19 months, the 2-year PFS and OS from the start of salvage in all patients were 61% and 97% respectively. Among patients with ASCT, the 2-year post-transplant PFS (PFS_HCT) was 76% for patients who had PD-1 blockade as salvage before ASCT, compared with 59% for those who did not. In univariate analysis, PD-1 blockade use in first salvage was significantly associated with improved PFS_HCT, and this association remained statistically significant after adjusting for stage, extranodal involvement, and primary refractory disease (HR 0.31, p = 0.04). Primary refractory disease after BV-AVD emerged as an ongoing unmet need with a significantly inferior 2-year PFS_HCT compared with relapsed patients (58% vs 86%, p = 0.017). Among primary refractory patients who received ASCT, first salvage incorporating a PD-1 blockade showed a trend toward improved PFS_HCT compared with non-PD-1 blockade salvage. These results support PD-1 blockade incorporation as preferred first salvage in R/R cHL after BV-containing frontline therapy.

Introduction

Standard treatment of patients with relapsed or refractory (RR) classic Hodgkin lymphoma (cHL) is salvage therapy followed by autologous stem cell transplant (ASCT) in chemosensitive patients. Traditionally, salvage therapy comprised of non-cross-resistant regimens of cytotoxic chemotherapy that resulted in complete response (CR) in a majority of patients and, when followed by ASCT, were associated with long-term durable remission in about half of patients [1–5]. The incorporation of novel agents, brentuximab vedotin (BV) [6–8] and PD-1 blockade [9–13], into salvage regimens prior to ASCT improved PET CR rates. However, the available data on outcomes of RR cHL who received first line treatment regimens that included novel agents is limited. BV combined with doxorubicin, vinblastine, and dacarbazine (AVD) chemotherapy demonstrated significantly prolonged modified progression-free and overall survival over ABVD [14], which led to frontline utilization of BV for advanced stage cHL. Yet, up to 20% of patients will progress after BV-containing frontline therapy [14], and outcomes of patients with RR cHL after BV and anthracycline containing treatments have not been described.

Prior single-arm studies evaluating PD-1 blockade salvage regimens studies have demonstrated favorable outcomes in patients with RR cHL treated with PD-1 blockade–containing salvage therapy compared to BV-based or chemotherapy-only salvage therapy prior to ASCT, possibly attributed to chemosensitization associated with PD-1 blockade [15–17]. However, these studies largely included patients treated with conventional chemotherapy as initial therapy (e.g. ABVD), and optimal management of RR cHL after frontline BV is undefined at this time. We performed a multicenter retrospective study to evaluate outcomes of patients with RR cHL after frontline BV and anthracycline-containing regimen failure and to explore the impact of PD-1 blockade based versus other salvage regimens in these patients.

Methods

Consecutive patients with primary refractory or relapsed cHL after BV and anthracycline-containing frontline therapy were included from four participating institutions. Data on patient, disease, and treatment characteristics was collected by retrospective review of electronic medical records. The study was approved by IRB of each participating institution. Response to treatment was assessed locally by treating physicians using PET scans based on the response criteria at time of assessment [18]. Overall response was defined as achieving a CR or partial response (PR). PFS was calculated from the start date of (first) salvage to date of disease progression/relapse or death, censored at last available follow-up. Overall survival was calculated from the start date of (first) salvage to date of death (due to any cause) and censored at last available follow-up. Among patients who received transplant, PFS_HCT and OS_HCT were calculated similarly to PFS/OS except that the start date was the date of transplant. Event-free survival (EFS) was calculated as from start date of first salvage to date of disease progression/relapse, start of next treatment, or death, censored at last available follow-up. Early release is defined as relapse within 12 months from completion of front-line therapy. Baseline characteristics and response to therapy were evaluated using descriptive statistics. PFS and OS were assessed using Kaplan–Meier analysis. Chi-square test, log rank test, and Cox proportional hazards model were used to examine the outcomes in the different cohorts. The proportional hazards assumption was assessed using Schoenfeld residuals; no violations were detected across all models. All statistical analyses were performed with R version 4.4.2.

Results

116 patients who received BV-based curative intent anthracycline-containing initial therapy and had primary refractory or relapsed cHL were identified in the database. The median age at relapse of the study population was 35 years (range 17–82) and 51% were male. Patients received BV-containing upfront regimens between 2016 and 2023. Patients received BV-AVD (N = 88), combination of BV-AVD and ABVD (N = 21), and other brentuximab-anthracycline combinations (N = 7). Types of other brentuximab-anthracycline combinations included BV combined with AVEPC (in pediatric patients), BV-AVD combined with BV-darcabazine, BV-AVD combined with BEACOPP, and BV with doxorubicin and dacarbazine. Patients received a median of 5 cycles of BV during their frontline treatment. Patients with prior PD-1 blockade exposure were excluded. 62% had primary refractory disease (defined as progression during initial treatment or failure to achieve a complete remission after induction therapy), 27% relapsed within 12 months of completing initial treatment, and 11% had late relapse. At relapse, 63% patients presented with stage III/IV disease, 18% had B symptoms, 22% had bulk ≥5 cm, and 46% had extranodal disease. Patient characteristics are detailed in Table 1.

Table 1.

Patient characteristics.

Characteristic	N = 116a
Age at relapse	35 (17, 82)
Sex
Male	59 (51%)
Female	57 (49%)
Frontline therapy
AA(BV)VD	88 (76%)
AAVD/ABVD	21 (18%)
Other BV/anthracycline	7 (6%)
Timing of relapse
Primary refractory disease	72 (62%)
Early relapse	31 (27%)
Late relapse	13 (11%)
Extranodal sites at relapse
No	59 (54%)
Yes	51(46%)
Unknown	6
Stage at relapse
I	6 (5.6%)
II	32 (31%)
III	22 (21%)
IV	44 (42%)
Unknown	12
B symptoms at relapse
No	89 (82%)
Yes	19 (18%)
Unknown	8
Bulky at relapse
<5 cm	77 (78%)
≥5 cm	22 (22%)
Unknown	17
Salvage 1 treatment
Anti PD-1 + chemotherapy combination	67 (58%)
Anti PD-1 monotherapy	9 (7.8%)
BV + Nivolumab	8 (6.9%)
Chemotherapy alone	30 (26%)
Nivo+CD30 CAR	1 (0.8%)
RT	1 (0.8%)
PD-1 exposure in first salvage
NOT PD-1	31 (27%)
PD-1 exposed	85 (73%)
PD-1 exposure at any salvage
NOT PD-1	22 (19%)
PD-1 exposed	94 (81%)
ASCT
No	25 (22%)
Yes	91 (78%)
Transplant conditioning regimen
BEAM	89 (99%)
Other	1 (1.1%)
Unknown	1
Follow up in survivors (months)	19 (1, 104)

^aMedian (range); n (%).

The most frequently used first salvage regimens were PD-1 blockade + chemotherapy combinations (58%), followed by chemotherapy alone (26%), PD-1 blockade monotherapy (7.8%), BV+ nivolumab (6.9%), and others (1.8%). Thirty-one (26%) patients received 2 or more lines of salvage (maximum 6). In all patients, the overall response rate (ORR) to first salvage treatment was 90% (96 out of 107 patients with response assessment) and the CR rate was 65% (70/107). First salvage treatment that included PD-1 blockade (n = 85) led to an ORR of 94% with 70% CR versus ORR 80% and 53% CR for salvage that did not contain PD-1 blockade (n = 31). Overall, 94 (81%) patients received PD-1 blockade during any salvage line before transplant, and a total of 112 PD-1 blockade–containing treatment lines were administered. Of these, 83 consisted of PD-1 blockade + chemotherapy, 20 were PD-1 blockade monotherapy, and 9 comprised BV+ nivolumab. The response rates were 67% CR, 27% PR, 2.5% SD, 3.5% PD, among patients receiving PD-1 blockade + chemotherapy, 44% CR, 38% PR, 18% PD, among patients receiving PD-1 blockade monotherapy, and 62.5% CR, 25% PR, 12.5% PD, among patients receiving BV + nivolumab. Patients with primary refractory disease had a CR rate of 57% to first salvage treatment, which was significantly lower than a CR rate of 80% in patients with relapsed disease (p = 0.025).

With a median follow-up of 19 (range 1–104) months among survivors, the overall estimated 2-year PFS and OS from the start of salvage in all patients were 61% (95% CI: 51–73%) and 97% (95% CI: 92–100%) respectively (Fig. 1). Ninety-one pts (78%) underwent autologous stem cell transplantation (ASCT), with 77% receiving transplant in second line and 23% in third line or later. Among patients receiving transplant, 70 (77%) received at least one line of PD-1 blockade salvage prior to ASCT, 14% received post-transplant maintenance (PD-1 blockade, n = 4; BV, n = 6, PD-1 blockade + BV, n = 2), and 13% had peri-transplant radiotherapy. Disease control prior to autologous transplant was achieved in most patients, with 73.5% attaining CR, 24% attaining PR, and 2.5% attaining SD. Neither pre-transplant disease status (CR vs PR/SD) or post-transplant maintenance therapies were significantly associated with 2-year PFS_HCT (Supplementary Fig. 1A, B).

Fig. 1.

Progression-free survival (PFS) and overall survival (OS) of all patients.

The 2-year PFS_HCT and OS_HCT among transplanted patients was 70% (95% CI: 57–85%) and 93% (95% CI: 86–100%), respectively. Patients requiring more than one salvage therapy prior to ASCT had inferior PFS_HCT compared with those proceeding directly to transplant after first salvage (2-year PFS_HCT: 39%, 95% CI: 16–98%; versus 78%, 95% CI 66–94%; p = 0.0055) (Supplementary Fig. 2). In a univariate Cox proportional hazard model, receipt of PD-1 blockade first salvage was associated with significantly improved PFS_HCT (HR 0.30, 95% CI 0.10–0.85, p = 0.02). In contrast, several adverse clinical features were observed with inferior PFS_HCT, including stage IV disease at relapse (vs. stage I–III) (HR 3.21; 95% CI 1.16–8.93), presence of extranodal disease (HR 3.03, 95% CI 1.09–8.47), as well as primary refractory disease (HR 4.10, 95% CI 1.17–14.37; all p-values 0.03).

Use of PD-1 blockade–containing salvage therapy led to a trend toward improved 2-year PFS_HCT, compared with regimens without PD-1 blockade (76% [95% CI: 61–96%] vs. 59% [95% CI: 41–86%]; p = 0.12). When stratified by line of salvage, incorporation of PD-1 blockade–containing first salvage was associated with a significantly improved 2-year PFS_HCT of 82% (95% CI: 67–100%), compared with 46% (95% CI: 17–100%) when used only in second or later salvage, and 59% (95% CI: 41–86%) among those who did not receive PD-1 blockade (p = 0.043). Consistently, 2-year PFS_HCT was superior in patients treated with PD-1 blockade versus those without PD-1 blockade at first salvage (p = 0.017; Fig. 2A). The association remained statistically significant in multivariable analysis adjusting for stage, extranodal involvement, and primary refractory disease (HR 0.31; 95% CI 0.10–0.92, p = 0.04). Among patients receiving PD-1 blockade prior to ASCT, those exposed only in second or later lines had significantly inferior 2-year PFS_HCT (p = 0.019), underscoring the potential importance of early incorporation of PD-1 blockade in the salvage to transplant pathway.

Fig. 2. Post-transplant progression-free survival (PFS_HSCT) outcomes.

A PFS_HSCT of patients receiving PD-1 based initial salvage vs. non-PD-1 based first salvage. B PFS_HSCT in patients with primary refractory disease receiving first salvage with and without PD-1 blockade.

Among the 25 patients who did not proceed to transplant, seven were aged ≥75 years and were therefore likely ineligible for transplant. Of the remaining 18 potentially eligible patients, nine achieved a response to first salvage therapy (8 CR, 1 PR), two experienced disease progression, and the rest had no documented response. Most (96%) of these patients received at least one line of PD-1 blockade–based salvage, with 88% receiving a PD-1 blockade as first salvage, which precludes reliable stratified analysis on the impact of PD-1 blockade in this group. In the overall cohort, PD-1 blockade exposure in first salvage was also associated with significantly improved EFS, with a 2-year EFS of 63% (95% CI: 50–80%) versus 43% (95% CI: 28–65%) in patients not receiving PD-1 blockade first salvage (p = 0.019) (Supplementary Fig. 3), supporting a consistent treatment effect across endpoints.

Despite receiving salvage therapy and ASCT, the two-year PFS_HCT was significantly worse among patients with primary refractory disease versus relapsed disease (54%; 95% CI, 41–82% vs. 86%, 95% CI, 72–100%, p = 0.017) (Fig. 3), highlighting a persistent unmet need in patients without complete response after frontline BV-AVD. Among patients with primary refractory disease, while there was a consistent trend toward improved 2-year PFS_HCT rates in patients receiving first salvage with PD-1 blockade compared to patients who did not (83%, 95% CI: 70–100%, versus 44%, 95% CI: 25–77%, p = 0.087) in univariate analysis (Fig. 2B), the association was not statistically significant in multivariable modeling, likely due to limited statistical power.

Fig. 3.

Post-transplant progression-free survival (PFS_HCT)of patients with primary refractory vs. relapsed Hodgkin’s Lymphoma.

Discussion

Following the adoption of BV-AVD as initial treatment of advanced stage cHL, there remains limited evidence characterizing disease features, treatment patterns, and associated outcomes after BV-containing frontline therapy. In our multicenter cohort of patients with RR cHL previously treated with BV-containing frontline therapy, most presented with primary refractory and advanced stage disease at relapse. We found that the majority responded to first salvage treatment, with numerically higher ORR and CR rates observed with PD-1 blockade–based regimens. Among patients who underwent transplant, PD-1 blockade first salvage was significantly associated with improved PFS_HCT outcomes compared with non-PD-1 blockade regimens, and this association remained robust after adjusting for stage, extranodal involvement, and primary refractory disease. Although primary refractory disease continued to be associated with inferior outcomes, PD-1 blockade–containing salvage showed a consistent trend toward improved PFS_HCT within this high-risk subgroup, suggesting clinically meaningful benefit over other salvage regimens.

The 2-year PFS_HCT of 76% among patients who received PD-1 blockades before ASCT in our cohort appeared lower than the 88% 2-year PFS reported in a large retrospective series of 320 patients treated with PD-1 blockade salvage prior to ASCT [17]. Importantly, only 6% of patients in that study had received BV frontline treatment, underscoring the need for dedicated evaluations of PD-1 blockade salvage in BV-exposed populations. We also observed inferior PFS when PD-1 blockade was introduced only at later salvage lines, supporting the early integration of PD-1 blockade before transplant to derive maximal benefit. Notably, patients who received PD-1 blockade exclusively in later lines exhibited suboptimal responses to PD-1 blockade (4 PR, 1 SD, 1 PD, 1 unknown), suggesting that their inferior outcomes may also reflect underlying PD-1 blockade-refractory disease biology rather than the timing of PD-1 blockade exposure alone. Despite lower PFS estimates, 2-year OS remained favorable at 97%, reflecting advances in post-transplant salvage, as well as availability of active therapies in subsequent lines.

Recent phase II trials have reported excellent outcomes with PD-1 combined with chemotherapy salvage, including in patients with previous chemoresistant disease or high-risk prognostic variables such as primary refractory disease or extranodal disease [9, 10, 13]. In comparison, patients in our cohort who received PD-1+ chemotherapy as first salvage achieved CR rates of 69%, which was lower than the 89–95% CR rates observed in these trials. Correspondingly, the 2-year PFS_HCT rate of 78% (95% CI: 58–100%) after PD-1 blockade+chemotherapy first salvage in our study also appeared lower than the 2-year PFS_HCT rates of 87–97% reported in prior trials [9, 10, 13]. Although our study included a higher proportion of primary refractory (62%) patients compared to some PD-1+chemotherapy salvage trials (38–41%) [9, 13], which may partly explain the observed differences, outcomes in our relapsed (non-primary refractory) subgroup also appeared less favorable. Notably, Mei et al. reported high CR rates (86%) and a 2-year PFS of 88% with nivolumab-ICE in a comparably high-risk cohort that included 60% primary refractory patients, although only 17% had prior exposure to frontline BV [10]. Given that earlier PD-1 blockade salvage trials enrolled relatively few patients treated with frontline BV, one plausible explanation is that the less favorable outcomes observed in our cohort may reflect shared adverse biological features conferring resistance to both BV and PD-1 blockade.

In addition to its direct cytotoxic effects on Reed-Sternberg cells, BV targets CD30⁺ regulatory T cells and attenuates their immunosuppressive function [19]. The MMAE payload further promotes dendritic cell and CD8⁺ T-cell activation, leading to immunogenic cell death and amplifying anti-lymphoma immune responses [20]. Patients who fail BV-AVD may harbor an immunosuppressive tumor microenvironment that not only blunts BV-induced immunogenic cell-death but may also reduce responsiveness to PD-1 blockade. These observations highlight the need for future correlative studies delineating the mechanisms of sensitivity and resistance, with the goal of refining individualized salvage strategies and informing the development of rational, mechanism-based therapeutic combinations.

Although efficacy appeared lower than that reported in prior studies conducted in largely BV-naïve populations, the use of PD-1 blockade as part of first salvage was associated with significantly improved 2-year PFS_HCT, including significant benefit among patients with primary refractory disease. Taken together with the consistently favorable outcomes reported in prospective trials of PD-1 blockade–containing salvage followed by ASCT, these findings support PD-1 blockade as a preferred first salvage strategy for patients with RR cHL after upfront BV-based therapy, particularly in those without prior PD-1 blockade exposure.

A small proportion of patients received post-ASCT maintenance in our study. Although most of our cohort had primary refractory disease or early relapse, patients in the AETHERA trial were BV-naïve, which limits the applicability of those findings to our cohort of patients who received frontline BV-AVD. Emerging data suggest that the benefit of BV maintenance may be attenuated in patients previously exposed to novel agents, including BV and PD-1 blockade [21]. PD-1 blockade maintenance remains an area of ongoing investigation, with current evidence limited to small, single-arm phase II studies demonstrating favorable outcomes [22, 23]. In our cohort, maintenance therapy was not associated with improved PFS_HCT; however, the small number of patients receiving maintenance limits the conclusions that can be drawn.

Although BV-AVD use in patients with newly diagnosed advanced stage cHL may decline with the S1826 study demonstrating that N-AVD improved PFS compared to BV-AVD, BV-AVD remains a useful regimen for patients with active autoimmune disease or pre-existing interstitial lung disease, or who are otherwise not eligible for PD-1 blockade therapy [24]. In addition, BrECADD (BV, etoposide, cyclophosphamide, adriamycin, dexamethasone, dacarbazine), a BV-containing frontline regimen building on the escalated BEACOPP backbone with superior efficacy has also emerged as an important treatment option for patients with newly diagnosed advanced stage cHL [25]. It is not clear whether our findings may be generalizable to patients progressing after BrECADD, and thus, it will be important to understand specific outcomes in patients with RR cHL who progress after BrECADD. There are also ongoing clinical trials evaluating BV as part of frontline therapy of early stage cHL and BV-AVD is part of guidelines for unfavorable early stage cHL [26], for which salvage therapy containing PD-1 blockade may be recommended.

Limitations of our study include its retrospective design, which is subject to residual confounding, as well as the relatively small sample size and the heterogeneity of both BV-containing frontline regimens and salvage therapies used. Response assessments were not centrally reviewed, and data on important treatment variables, such as frequency of BV dose reductions or omissions and the use of consolidative radiotherapy during frontline therapy were not uniformly collected, limiting our ability to evaluate their potential impact on outcomes. Confidence intervals were wide, reflecting the limited precision of the estimates inherent to the small sample size. In addition, patients proceeding to ASCT represent a selected subgroup with treatment-sensitive disease, adequate functional status and organ function, and without advanced age precluding transplantation. As a result, a comparison of outcomes in patients receiving ASCT vs no ASCT is not possible. Additionally, as a retrospective study, it is not possible to infer the intent of the treatment program (e.g. intent to ASCT), which also limits comparison of treatment approaches across the transplanted and non-transplanted patients. Despite these limitations, this study represents the largest series to date examining salvage outcomes following failure of BV- upfront treatment and provides important insights into treatment patterns and efficacy across different salvage approaches.

In summary, patients with RR cHL after BV-AVD frontline treatment have improved outcomes with first salvage containing PD-1 blockade compared to salvage without. Overall, outcomes with PD-1 blockade+chemotherapy salvage were less favorable in our BV-anthracycline failure cohort than the CR rates and PFS reported in clinical trials, which were conducted largely in BV-naïve populations. With the evolving frontline landscape of frontline therapy, it will be important to define outcomes in patients relapsing after N-AVD and BrECADD, which were not addressed in our cohort. Nevertheless, a substantial proportion of patients worldwide continue to receive BV-AVD frontline therapy. In this context, our findings support PD-1 blockade combinations as an important first salvage option for patients with R/R disease following BV exposure, particularly in those without prior PD-1 blockade treatment.

Author contributions

SYO and AFH conceptualized the study and wrote the original draft; LC and JMM did the statistical analysis and reviewed and edited the draft; RM, HS, RS, SD, ALC, AF, KB, PA, MM, AM reviewed and edited the draft; NG, TC, UD, TO curated the database and reviewed and edited the draft

Competing interests

AM reports research support from ADC Therapeutics, Beigene, Miragen, Seattle Genetics, Merck, Bristol-Myers Squibb, Incyte, and SecuraBio. She reports honoraria from Affimed, Imbrium Therapeutics L.P./Purdue, Janpix Ltd., Merck, Seattle Genetics, and Takeda. PA reports Consultancy: Merck, BMS/Celgene, GenMab, Enterome, Genentech/Roche, ATB Therapeutics, Regeneron; Research funding: Kite; Research funding (institutional): Merck, BMS/Celgene, Adaptive, Genentech, IGM, Astra Zeneca, Pfizer. MM reports consultancy with Novartis, SeaGen, CTI, ADC Therapeutics, AstraZeneca, Synthekine; speakers’ Bureau with SeaGen, Incyte; and research funding (institutional) from BMS, Incyte, Beigene, Genentech. AFH reports research funding from BMS, Merck, Genentech,Inc/F. Hoffmann-La Roche Ltd, Gilead Sciences, SeaGen, AstraZeneca,and ADC Therapeutics, and consultancy for BMS, Merck, Genentech,Inc/F. Hoffmann-La Roche Ltd, Kite Pharma/Gilead, SeaGen,Karyopharm, Takeda, Tubulis, AstraZeneca, Genmab, ADC Ther-apeutics, and Regeneron. The remaining authors declare no competing financial interest

Ethics

The study was approved by the IRB of each participating institution. All IRBs waived the requirement for informed consent.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41408-026-01547-2.