A Matching-Adjusted Indirect Comparison of PET-Guided BrECADD Versus PET-Guided ABVD, BEACOPP, and eBEACOPP in Advanced Hodgkin Lymphoma

Abstract

Introduction

In the absence of head-to-head trial data, a matching-adjusted indirect comparison (MAIC) was conducted to compare the effectiveness, with respect to overall survival (OS) and progression-free survival (PFS), of PET-guided BrECADD versus PET-guided ABVD, BEACOPP, and eBEACOPP in advanced Hodgkin lymphoma.

Methods

Individual patient data (IPD) from patients in HD21 and aggregate data from the SWOG S0816, RATHL, HD9, and Mondello 2020 studies were utilized. Potential effect modifiers and prognostic variables for advanced Hodgkin lymphoma were identified via literature review, clinician input, and statistical analysis. As there was no common arm between HD21 and the comparator trials, unanchored MAICs were conducted. Data from patients in HD21 were weighted to align with published aggregate characteristics of the comparator trials.

Results

BrECADD was associated with significantly improved PFS and OS compared with PET-guided ABVD among patients aged 18–60 (SWOG S0816; PFS HR: 0.24, 95% CI 0.16, 0.35, p < 0.001; OS HR: 0.30, 95% CI 0.14, 0.63, p = 0.004) and patients aged 18 to 75 (RATHL; PFS HR: 0.24, 95% CI 0.16, 0.36, p < 0.001; OS HR: 0.47, 95% CI 0.25, 0.89, p = 0.01). In the comparison of BrECADD versus BEACOPP (per the Mondello et al. 2020 study cohort), there was a trend in PFS improvement [HR = 0.58 (95% CI 0.29, 1.19), p = 0.194)] in favor of BrECADD in adults aged 18–75 years of age. Additionally, a trend was observed among patients aged 66–75, favoring BrECADD for freedom from treatment failure [HR = 0.63 (95% CI 0.28, 1.44), p = 0.260] when compared to BEACOPP (HD9).

Conclusion

These findings provide robust evidence and further support for the use of BrECADD in adult patients with advanced Hodgkin lymphoma.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s12325-026-03695-0.

Key Summary Points

Why carry out this study?

Currently there are a lack of head-to-head trials comparing treatments.

The present study therefore utilized indirect treatment comparison to evaluate the comparative effectiveness of PETguided BrECADD (HD21) versus PET-guided ABVD (RATHL, S0816), BEACOPP (HD9; in patients over age 60), and eBEACOPP (per the Mondello et al. 2020 study cohort) in advanced Hodgkin lymphoma.

What was learned from the study?

BrECADD was associated with significantly improved progression-free survival and overall survival compared with PETguided ABVD, as well as a trend in improvement in eBEACOPP comparisons in the overall (18+) adult population with advanced Hodgkin lymphoma.

Overall survival results were also significantly in favor of BrECADD versus BEACOPP, specifically in adults over 60 years of age, with a similar trend observed for progression-free survival.

These results further support the use of BrECADD over PET-guided ABVD and BEACOPP-based treatment regimens in adult populations with advanced Hodgkin lymphoma, both under and over the age of 60.

Introduction

Hodgkin lymphoma is a rare disease with an estimated global incidence rate of 0.98 per 100,000 individuals in 2020 [1]. About 95% of Hodgkin lymphoma is classic Hodgkin lymphoma [2]. Staging of Hodgkin lymphoma determines treatment and can be conducted with different staging systems, including the Lugano classification and the German Hodgkin Study Group (GHSG) staging system [3, 4]. Of note, stage IIB with large mediastinal mass or extranodal involvement is specific to the GHSG staging system. In patients with early-stage disease (stage I or II), 5-year survival rates are 92.9% and 94.8%, respectively; survival in patients with stage III and stage IV Hodgkin lymphoma decreases to 87.1% and 82.0%, respectively [5].

The treatment pathway for Hodgkin lymphoma is complicated by the need to balance treatment intensity with the risk of toxicity. In the first-line setting, the 2026 National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology (NCCN Guidelines^®) recommend positron emission tomography (PET)-guided brentuximab vedotin, etoposide, cyclophosphamide, doxorubicin, dacarbazine, dexamethasone (BrECADD) and nivolumab with doxorubicin, vinblastine, and dacarbazine (nivolumab-AVD) for the management of advanced Hodgkin lymphoma [6]. In certain circumstances, PET-guided combination chemotherapy doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD) and brentuximab vedotin, doxorubicin, vinblastine, and dacarbazine are considered suitable treatment approaches. Escalated bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone (eBEACOPP) has demonstrated superior efficacy compared with ABVD in terms of disease control and is the standard of care (SOC) in some countries [7]; however, patients treated with eBEACOPP are exposed to considerable acute and long-term toxicities [8]. Additionally, although nivolumab-AVD is also recommended in the NCCN Guidelines^® and European Society for Medical Oncology (ESMO) guidelines [6, 9], in the absence of European Medicines Agency (EMA) and Food and Drug Administration (FDA) labels, treatment in many parts of the world does not yet include nivolumab-AVD. Thus, the other treatment approaches were considered more clinically relevant at the time of these analyses.

In the phase 3, open-label HD21 trial comparing PET-guided BrECADD versus PET-guided eBEACOPP in patients with advanced-stage (Ann Arbor stage III/IV or stage II with B symptoms, and either one or both risk factors of large mediastinal mass or extranodal lesions) first-line Hodgkin lymphoma, BrECADD demonstrated improved progression-free survival {PFS; hazard ratio [HR] = 0.66 [95% confidence interval (CI) 0.45, 0.97; p = 0.035]} and lower treatment-related morbidity [relative risk = 0.72 (95% CI 0.65, 0.80); p < 0·0001] compared with eBEACOPP among patients aged 18 to 60 years in the HD21 randomized cohort (NCT02661503) [10–13]. Overall survival (OS) rates were comparable at 98.6% (95% CI 97.7, 99.5) for BrECADD and 98.2% (95% CI 97.2, 99.3) for eBEACOPP. The BrECADD regimen also demonstrated a favorable safety profile. As a result, the HD21 trial was expanded to assess an additional phase 2, single-arm cohort of patients aged 61–75 years treated exclusively with PET-guided BrECADD [14].

There were clear limitations to using BEACOPP or PET-guided eBEACOPP from comparator trials to compare the effectiveness of BrECADD directly with other relevant treatment regimens through anchored analysis. Due to heterogeneity in the BEACOPP regimens across included studies, utilizing these regimens as a common comparator in an anchored analysis would violate the transitivity assumption and lead to biased results. Therefore, in the absence of a head-to-head trial with other first-line regimens, the objective of this study was to evaluate the comparative effectiveness of BrECADD versus PET-guided ABVD and BEACOPP-based regimens among three populations with advanced Hodgkin lymphoma using an unanchored approach: patients aged 18–60 years, patients aged 61–75 years, and the overall adult population (18–75 years), and versus BEACOPP in patients aged 66–75 years.

Methods

Data Sources

HD21 served as the index trial population for BrECADD, all patients randomized in HD21 were included in the initially pooled individual patient data (IPD) from the intent-to-treat (ITT) and the non-randomized cohorts [10–13]. Four studies were considered for inclusion in the indirect treatment comparison (ITC) based on available peer-reviewed publications, including SWOG S0816 (NCT00822120) [15, 16] and RATHL (NCT00678327) [17, 18] for PET-guided ABVD, HD9 (66- to 75-year-old cohort) [19, 20] for BEACOPP, and Mondello 2020 [8] for eBEACOPP. The main study cohorts from HD9 (patients aged 15–65 years) and HD18 were not assessed because patients in HD21 included a direct comparison of BrECADD and eBEACOPP, and the BEACOPP treatment of patients under 65 was outdated by treatment standards and by more recent studies using an escalated protocol in younger patients, while IPD for the eBEACOPP regimen was available in HD21. Select studies (HD9 older cohort and Mondello 2020 [8]) assessing BEACOPP without PET guidance were included due to lack of PET-guided data outside of HD-21 for certain populations.

An overview of the studies included is provided in Table 1. SWOG S0816 was a phase 2, open-label, non-randomized study conducted in the United States (US) and Canada of patients aged 18–60 years who had measurable stage III–IV Hodgkin lymphoma, no prior therapy for lymphoma, and an Eastern Cooperative Oncology Group (ECOG) performance status (PS) score of 0–2 [15, 16]. RATHL was a phase 3, prospective, randomized, controlled trial conducted in an international setting in previously untreated patients 18 years and older with advanced classic Hodgkin lymphoma (stage IIB–IV, or stage IIA with adverse features: bulky disease or at least three involved sites) [17, 18]. HD9, a prospective, randomized trial, included patients aged 66–75 years with previously untreated biopsy-proven Hodgkin lymphoma (stage III and IV, or stage IIB with large mediastinal mass or extranodal involvement) [19–21]. The Mondello 2020 [8] study was included as a retrospective analysis of patients aged 18 and older in Austria and Italy with newly diagnosed Hodgkin lymphoma stage III–IV who received either ABVD or eBEACOPP in a real-world setting.

Table 1.

Overview of studies considered for inclusion in the indirect treatment comparison

Study design	Country	Selection criteria	Age range (years)	Stage	Intervention	Comparator	Cycle length
HD21 [10–13] (NCT02661503)
Phase 3, randomized, multicenter, parallel, open-label	Europe, Australia, and New Zealand	Patients who were 18–60 years old, had histologically proven classical Hodgkin lymphoma that were stage IIB (with 1 or both risk factors: large mediastinal mass, extranodal disease), stages III or IV, had no previous treatment for Hodgkin lymphoma, and had an ECOG ≤2	18–60	IIB, III, IV	BrECADD: brentuximab vedotin 1.8 mg/kg (day 1), etoposide 150 mg/m2 (day 2–4), doxorubicin 40 mg/m2 (day 2), cyclophosphamide 1250 mg/m2 (day 2), dacarbazine 250 mg/m2 (day 3-4), dexamethasone 40 mg (day 2–5), G-CSF 6 mg (on day 5)	eBEACOPP: bleomycin 10 mg/m2 (day 8), etoposide 200 mg/m2 (day 1–3), doxorubicin 35 mg/m2 (day 1), cyclophosphamide 1250 mg/m2 (day 1), vincristine 1.4 mg/m2 (day 8), procarbazine 100 mg/m2 (day 1–7), prednisone 40 mg/m2 (day 1–14), G-CSF 6 mg (on day 4)	21 days
HD21 (61–75) [10–13] (NCT02661503)
Phase 2, single-arm cohort	Europe, Australia, and New Zealand	Patients aged 61–75 years	61–75	IIB, III, IV	BrECADD: brentuximab vedotin 1.8 mg/kg (day 1), etoposide 150 mg/m2 (day 2–4), doxorubicin 40 mg/m2 (day 2), cyclophosphamide 1250 mg/m2 (day 2), dacarbazine 250 mg/m2 (day 3–4), dexamethasone 40 mg (day 2–5), G-CSF 6 mg (on day 5)	NA	21 days
SWOG S0816 [15, 16] (NCT00822120)
Phase 2, open-label, non-randomized, intergroup	United States, Canada	Patients negative for HIV aged 18–60 years, who had measurable stages III–IV Hodgkin lymphoma, no prior therapy for lymphoma, an ECOG of 0–2, and no other serious medical ailments	18–60	III, IV	PET-guided ABVD ABVD: administered at standard doses (day 1 and 15) eBEACOPP: bleomycin 10 mg/m2 (day 8), etoposide 200 mg/m2 (day 1–3), doxorubicin 35 mg/m2 (day 1), cyclophosphamide 1250 mg/m2 (day 1), vincristine 1.4 mg/m2 (day 8), procarbazine 100 mg/m2 (day 1–7), prednisone 40 mg/m2 (day 1–14), G-CSF (day 8–13)	NA	ABVD: 28 days eBEACOPP: 21 days
RATHL [17, 18] (NCT00678327)
Phase 3, prospective, randomized (only for patients undergoing PET who received ABVD or AVD), controlled	United Kingdom, Italy, Australia, New Zealand, Norway, Sweden, and Denmark	Previously untreated patients aged ≥18 years with advanced classic Hodgkin lymphoma (stage III or IV, or stage IIA with adverse features: bulky disease or at least 3 involved sites) Previously untreated patients ≥18 years old with advanced classic Hodgkin lymphoma (stage III or IV, or stage IIA with adverse features: bulky disease or at least 3 involved sites)	18–79 (9.5% >60 years)	IIB, III, IV	PET-guided ABVD ABVD: administered at standard doses AVD: same as ABVD without bleomycin eBEACOPP: bleomycin 10,000 units/m2 (day 8), etoposide 200 mg/m2 (day 1–3), doxorubicin 35 mg/m2 (day 1), cyclophosphamide 1250 mg/m2 (day 1), vincristine 1.4 mg/m2 (day 8), procarbazine 100 mg/m2 (day 1–7), prednisone 40 mg/m2 (day 1–14), G-CSF (day 9 until count recovered) BEACOPP-14: bleomycin 10,000 units/m2 (day 8), etoposide 100 mg/m2 (day 1–3), doxorubicin 25 mg/m2 (day 1), cyclophosphamide 650 mg/m2 (day 1), vincristine 1.4 mg/m2 (day 8), procarbazine 80 mg/m2 (day 1–7), prednisone 80 mg/m2 (day 1–7), G-CSF (day 9–13)	NA	ABVD/AVD: 28 days eBEACOPP: 21 days BEACOPP-14: 14 days
HD9 (66–75) [19, 20] (NA)
Prospective randomized trial	NA	Patients aged between 66 and 75 years with previously untreated biopsy-proven Hodgkin’s disease. Patients with stage IIB and large mediastinal mass and/or extranodal involvement and/or massive spleen involvement and all patients in stages III and IV disease were eligible.	66–75	III, IV	BEACOPP: bleomycin 10 mg/m2 (day 8), etoposide 100 mg/m2 (day 1–3), doxorubicin 25 mg/m2 (day 1), cyclophosphamide 650 mg/m2 (day 1), vincristine 1.4 mg/m2 (day 8), provarbazine 100 mg/m2 (day 1–7), prednisone 40 mg/m2 (day 1–14) eBEACOPP: bleomycin 10 mg/m2 (day 8), etoposide 100 mg/m2 (day 1–3), doxorubicin 35 mg/m2 (day 1), cyclophosphamide 1200 mg/m2 (day 1), vincristine 1.4 mg/m2 (day 8), provarbazine 100 mg/m2 (day 1–7), prednisone 40 mg/m2 (day 1–14)	COPP-ABVD: cyclophosphamide 650 mg/m2 (day 1 and 8), vincristine 1.4 mg/m2 (day 1 and 8), procarbazine 100 mg/m2 (day 1–14), prednisone 40 mg/m2 (day 1–14), doxorubicin 25 mg/m2 (day 29 and 43), bleomycin 10 mg/m2 (day 29 and 43), vinblastine 6 mg/m2 (day 29 and 43), and dacarbazine 375 mg/m2 (day 29 and 43)	21 days
Mondello 2020 [8] (NA)
Retrospective analysis	Austria and Italy	Patients ≥18 years old with newly diagnosed Hodgkin lymphoma stages III–IV who received either ABVD or eBEACOPP outside of clinical trials	19–75 (3.3% >60 years)	III, IV	COPP-ABVD: cyclophosphamide 650 mg/m2 (day 1 and 8), vincristine 1.4 mg/m2 (day 1 and 8), procarbazine 100 mg/m2 (day 1–14), prednisone 40 mg/m2 (day 1–14), doxorubicin 25 mg/m2 (day 29 and 43), bleomycin 10 mg/m2 (day 29 and 43), vinblastine 6 mg/m2 (day 29 and 43), and dacarbazine 375 mg/m2 (day 29 and 43)	BEACOPP: bleomycin 10 mg/m2 (day 8), etoposide 100 mg/m2 (day 1–3), doxorubicin 25 mg/m2 (day 1), cyclophosphamide 650 mg/m2 (day 1), vincristine 1.4 mg/m2 (day 8), provarbazine 100 mg/m2 (day 1–7), prednisone 40 mg/m2 (day 1–14) eBEACOPP: bleomycin 10 mg/m2 (day 8), etoposide 100 mg/m2 (day 1–3), doxorubicin 35 mg/m2 (day 1), cyclophosphamide 1200 mg/m2 (day 1), vincristine 1.4 mg/m2 (day 8), provarbazine 100 mg/m2 (day 1–7), prednisone 40 v2 (day 1–14)	COPP-ABVD: 56 days BEACOPP and eBEACOPP: 21 days

ABVD doxorubicin, bleomycin, vinblastine, and dacarbazine; BrECADD brentuximab vedotin, etoposide, cyclophosphamide, doxorubicin, dacarbazine, dexamethasone; COPP cyclophosphamide, oncovin, procarbazine and prednisone; eBEACOPP escalated bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine and prednisone; ECOG Eastern Cooperative Oncology Group; G-CSF granulocyte colony-stimulating factor; HIV human immunodeficiency virus; NA not applicable; PET positron emission tomography; PS performance status

Across the HD21, SWOG S0816, and RATHL trials, the same criteria were used to determine PET status. Specifically, Deauville scores ranging from 1 to 3 were considered negative, whereas Deauville scores ranging from 4 to 5 were considered positive.

RATHL and SWOG S0816 stratified patient characteristics and outcome data by PET status, which allowed for analyses by subgroup. In RATHL and Mondello 2020 [8], only 9.5% and 3.3% of the populations, respectively, comprised patients older than 60 years. SWOG S0816 only included patients from the US and Canada, whereas HD21 and all other comparators included patients from Europe. HD21 and RATHL also included patients from Australia and New Zealand. SWOG S0816 and HD9 (66- to 75-year-old cohort) did not report information on ECOG PS. Extranodal site was only reported in HD9 (66- to 75-year-old cohort). In HD21, extranodal site information was collected, but the definition was restricted to direct extension into adjacent organs. Escalation to BEACOPP occurred across all studies, but in the SWOG S0186 and RATHL trials, ABVD treatment regimens differed. Other differences in treatment regimens are included in Table 1. The availability of patient characteristics is provided in Table 2.

Table 2.

Availability of patient characteristics in included studies

Study	Treatment of interest	Age	Sex	Race	Stage	IPS	B symptoms	ECOG	Bulky disease	Extranodal site
HD21 [10-13]	BrECADD	√+	√	√	√+	√+	√	√+		√
HD21 elderly [10-13]	BrECADD	√	√	√	√+	√+	√	√+
SWOG S0816 [15, 16]	PET-guided ABVD	√	√	√	√+	√+	√		√
RATHL [17, 18]	PET-guided ABVD	√+	√		√+	√+	√	√+	√
HD9 (66–75) [19, 20]	BEACOPP	√+	√		√+	√+	√		√	√
Mondello 2020 [8]	eBEACOPP	√+	√		√+	√+	√	√+	√

The + icon indicates that results by subgroup are available. However, across studies, specific subgroup categories varied.

ABVD doxorubicin, bleomycin, vinblastine, and dacarbazine; BrECADD brentuximab vedotin, etoposide, cyclophosphamide, doxorubicin, dacarbazine, dexamethasone; eBEACOPP escalated bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone; ECOG Eastern Cooperative Oncology Group; IPS International Prognostic Score; PET positron emission tomography; PS performance status

Effect Modification

Covariates identified as potential effect modifiers (EM) and prognostic variables (PV) for advanced Hodgkin lymphoma in previous literature reviews were considered in extensive consultation exercises with two focus groups composed of hematologists/oncologists from France, Germany, Italy, the Netherlands, Norway, and the United Kingdom. Physicians evaluated variables for three populations: patients aged 18–60 years, patients older than 60 years, and the overall adult population (18 years and older). Physicians voted on the covariates as EMs and PVs. Full agreement or partial agreement was reached when a variable received six votes or four votes as a PV or EM, respectively. Age, sex, International Prognostic Score (IPS), ECOG PS, stage, and B symptoms all reached partial or full agreement as EMs or PVs. Of note, although age and sex are included in IPS, they are also recommended for adjustment per the National Institute for Health and Care Excellence (NICE) Decision Support Unit (DSU) Technical Support Document (TSD) 18 [22].

EMs and PVs were further identified through statistical analyses using IPD from HD21, the HD21 61- to 75-year-old extension cohort, and pooled data from HD21 (i.e., randomized and non-randomized cohorts). Based on Cox regression analyses, age, sex, IPS, and ECOG PS were identified as PVs. No variables were identified statistically as EMs.

Outcomes

The efficacy outcomes of interest included PFS and OS, as both outcomes were available for HD21 [10–13], SWOG S0816 [15, 16], RATHL [17, 18], and Mondello 2020 [8]. Although HD21 and RATHL included patients with stage IIB Hodgkin lymphoma, stage III/IV subgroup results were available for PFS and OS.

The HD9 (66- to 75-year-old cohort) trial used freedom from treatment failure (FFTF) as a survival endpoint in lieu of a traditional PFS definition [19, 20]. Therefore, modified PFS was generated using IPD from HD21 to match the definition of FFTF in HD9, defined as the time from registration to occurrence of death from any cause, progressive disease, no complete response at the end of protocol treatment, relapse, or non-study treatment. FFTF and OS outcomes were not available by subgroup for the HD9 trial. The efficacy outcomes reported in the included studies, including the definitions, are provided in Table 3.

Table 3.

Efficacy outcomes reported in included studies

	HD21/HD21 (61- to 75-year-old cohort)	SWOG S0816 [15, 16]	RATHL [17, 18]	HD9 (66- to 75-year-old cohort) [19, 20]	Mondello, 2020 [8]
PFS definition	Time from randomization until progression, relapse, or death from any cause, or censored at the date of last information on the disease status	Date of registration to the first observation of PD, relapse, or death	Date of registration to the date of first appearance of PD, relapse, or death from any cause	FFTF: time from registration to occurrence of HD-specific events including PD, no CR after primary treatment, non-study treatment, relapse, and death from Hodgkin lymphoma	Time from entry onto a study until lymphoma progression or death as a result of any cause
OS definition	Time from randomization until death from any cause or censored at the date of last information on the patient being alive	Date of registration to the date of death	NR	The time from registration to death from any cause	Time from entry onto a study until death as a result of any cause
Median follow-up (months)	HD21: 53.0 HD21 (61- to 75-year-old cohort): 27.1	70.8	87.6	80	104.4
KM - PFS	IPD	√	√	√	√
KM - PFS subgroup	IPD	PET	PET, age, stage		PET
KM - OS	IPD	√	√	√	√
KM - OS subgroup	IPD	PET	PET, age, stage		PET

CR complete response; FFTF freedom from treatment failure; IPD individual patient-level data; KM Kaplan–Meier; NR not reported; OS overall survival; PD progressive disease; PET positron emission tomography; PFS progression-free survival

Statistical Methods

IPD were available for HD21 (i.e., index trial), and published aggregate data were available for the comparator studies. Of note, IPD from HD21 were pooled for the randomized and non-randomized cohorts to obtain an adult population with the full age range treated with BrECADD, whereas only randomized data were available for eBEACOPP, given that eBEACOPP was not studied in the HD21 non-randomized cohort. To obtain the time-to-event IPD for SWOG S0816, RATHL, HD9 (66- to 75-year-old cohort), and Mondello 2020 [8], Engauge Digitizer was employed to convert the image files of the Kaplan–Meier survival curves from the study publications into numeric values for time and corresponding survival probabilities [23]. The coordinates were then used to generate pseudo-IPD (i.e., time and censoring status) for each curve in R version 4.3.1 within the R Studio environment using a validated algorithm by Guyot [24].

Since there was no common arm between HD21 and the comparator trials, an unanchored matching-adjusted indirect comparison (MAIC) was conducted as opposed to other alternative methods in accordance with the NICE DSU TSD 18 [22].

HD21 IPD were weighted to align with published aggregate characteristics of the comparator trials using a method of moments approach [25, 26]. These weights were then used to calculate the effective sample size (ESS) achieved as ${\left(\sum ,w_i\right)}^2/\left(\sum ,w_i^2\right)$, where a small value indicated an irregular distribution of weights across patients, suggesting that only a small fraction of patients drove the treatment effect [27]. The analysis was deemed infeasible when the resulting ESS was less than 25% of the original sample size [28], or the regression model for indirect comparison did not converge (normally due to zero events or 100% events in at least one category). The population was restricted to stages III and IV to facilitate a more similar population as compared to that of the Mondello 2020 [8] cohort, among other adjustments in age, sex, functional status, and B symptoms, resulting in an ESS of 339.93 for HD21, which is 47.7%, maintaining the premise for executing this comparison.

Patients in HD21 with missing values for any of the covariates included in the analysis were excluded to ensure that complete data were available for all patients involved in the comparison. Weighted Cox regression with a robust variance estimator was used to generate hazard ratios (HRs) and their 95% CIs, reflecting the expected relative effect between BrECADD and the comparator. Naïve comparisons were also conducted using Cox regression without any reweighting or variable adjustment. For Cox regression, the proportional hazard (PH) assumption was evaluated based on Schoenfeld residuals [29]. In situations where the PH assumption was violated (Schoenfeld p < 0.05), restricted mean survival time (RMST) analysis was performed to provide an absolute, time‑bounded measure of effect that is not bound to the PH assumption. RMST time points were selected based on the minimum duration out of the entire reported follow-up in the two study arms of the index population and comparator study, as they closely relate to a clinically relevant time horizon [30, 31].

Base Case, Sensitivity Analyses, and Subgroup Analysis

Given the differences in trial design (e.g., disease staging and geography) and patient characteristics, the authors planned a robust set of analyses. The base case analysis included full adjustment of variables identified by the clinical experts as PVs/EMs. BrECADD in HD21 was compared with PET-guided ABVD (SWOG S0816 and RATHL) for various age, stage, and PET2 status (i.e., positive or negative) subgroups, and with BEACOPP and eBEACOPP [HD9 (66- to 75-year-old cohort) and Mondello 2020 [8]] for age- and stage-specific subgroups.

Where feasible, three sensitivity analyses were conducted to adjust for (1) variables shown to be statistically significant in the PV/EM analysis; (2) variables that received full agreement from physicians (variables receiving only partial agreement were not included); and (3) the inclusion of age as a continuous variable to address the differences in variable reporting in comparator trials. The study follow-up time differed between the HD21 randomized and HD21 single-arm populations, which could lead to censoring bias since most HD21 patients over age 60 in the single-arm cohort would be censored earlier than patients under age 60. Therefore, in cases where the two cohorts were pooled, further sensitivity analyses were conducted to adjust the censoring time cutoff to 24 months, 36 months, and no cutoff (no censoring based on follow-up time), in addition to the base case censoring at 30 months. All analyses performed are listed in Table 4.

Table 4.

Base case and sensitivity analyses

Analysis	Set conditions	Variables and rationale
Base case	All/full variable adjustment Base case age variable definition: age as categorical variable (50%/50% cut at median), or using RATHL category (RATHL only) [17, 18]	Age, sex, IPS, ECOG PS, stage, B symptom The variables for full adjustment (available in both HD21 [10–13] and comparator trials) matched the PV/EM full and partial agreement by clinical experts The age distribution is skewed, therefore, converting median age to a categorical variable would not bias the results
Sensitivity 1	Significant variables only	Age, sex, IPS, ECOG PS (based on HD21 IPD analysis) [10–13]
Sensitivity 2	Adjustment for variables that received full agreement from physicians as PV or EM	Age, ECOG PS, stage, B symptom
Sensitivity 3	Adjustment for continuous age (mean = median)	Explore the impact of adjusting age as continuous variable, all other variables adjusted are the same as base case
Time cutoff (pooled cohort only)	Base case: censoring at 30 months	In the pooled analysis (HD21 + HD21 age >60 years), [10–13] considering the potential censoring bias due to different follow-up time of the two cohorts, different time cutoffs may be tested to assess the robustness of results
	Sensitivity 4: censoring at 24 months
	Sensitivity 5: censoring at 36 months
	Sensitivity 6: no time cutoff

ECOG Eastern Cooperative Oncology Group; EM effect modifier; IPS International Prognostic Score; PS performance status; PV prognostic variable

All analyses were performed using R version 4.3.1 within the R Studio environment [32].

This article reports a retrospective analysis using individual patient data from the HD21 trial and aggregate data extracted from published studies (SWOG S0816, RATHL, HD9, and Mondello 2020 [8]). Review boards at participating institutions had previously approved all procedures for the HD21, SWOG S0816, RATHL, and HD9 trials. As the present study is a retrospective analysis of existing trial and published data and did not involve new patient recruitment or intervention, no additional institutional review board review was required for this analysis.

Results

In all analyses performed and presented below, the ESS were consistently greater than 45% of the original sample sizes in the HD21 population after MAIC reweighting. The matching variables and ESS before and after matching for HD21 versus the comparator trials are provided in Table 5. The results of the fully adjusted, base case comparisons are presented in Fig. 1.

Table 5.

Matching variables and effective sample sizes before and after matching for HD21 versus comparators

BrECADD, HD21 subgroup [10-13]	Comparator		Matching variables	HD21 ESS
	Intervention (study)	Subgroup (n)		Before matching	After matching
Age 18–60, stage III and IV	PET-guided ABVD (SWOG S0816) [15, 16]	ITT (n = 336)	Age, sex, IPS, stage, B symptoms	631	619.64 (98.2%)
Age 18–60, stage IIB, III, and IV	PET-guided ABVD (RATHL) [17, 18]	Age ≤60 (n = 1103)	Sex, IPS, ECOG PS, stage, B symptoms	748	440.93 (58.9%)
Age 61–75, stage IIB, III, and IVb	PET-guided ABVD (RATHL) [17, 18]	Age >60 (n = 98)	Sex, IPS, ECOG PS, stage, B symptoms	85	Not applicablea
Age 18–75, stage IIB, III, and IV, HD21 pooled ITTb	PET-guided ABVD (RATHL) [17, 18]	ITT (n = 1201)	Age, sex, IPS, ECOG PS, stage, B symptoms	833	438.54 (52.6%)
Age 18–75, stage III and IV, HD21 pooledb	PET-guided ABVD (RATHL) [17, 18]	Stages III and IV (n = 702)	Age, sex, IPS, ECOG PS, B symptoms	713	632.20 (88.7%)
Age >65, stage III and IV	BEACOPP (HD9) [19, 20]	Age 66–75 (n = 42)	Age, sex, IPS, stage, B symptoms	47	25.84 (55.0%)
Age 18–75, stage III and IV, HD21 pooled	eBEACOPP (Mondello 2020) [8]	ITT (n = 121)	Age, sex, IPS, ECOG PS, stage, B symptoms	713	339.93 (47.7%)

^aOnly naïve (unweighted) comparisons were possible since patient characteristics were not available for the RATHL age >60 years subgroup.

^bFor the comparison of BrECADD and PET-guided ABVD in the 18-75 and 61-75 age groups, patients up to age 79 were included for RATHL, as data on the  < 75 year age group were not available.

ABVD doxorubicin, bleomycin, vinblastine, and dacarbazine; BrECADD brentuximab vedotin, etoposide, cyclophosphamide, doxorubicin, dacarbazine, dexamethasone; eBEACOPP escalated bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine and prednisone; ECOG Eastern Cooperative Oncology Group; IPS International Prognostic Score; ITT intent to treat; PET positron emission tomography; PS performance status

Fig. 1.

Fully adjusted Cox regression results for progression-free survival and overall survival. *Naïve (unweighted) results are presented since patient characteristics were not available for the RATHL age 61–75 subgroup. ESS*/N is the effective sample size over the total sample size. For the comparison of BrECADD and PET-guided ABVD in the 18–75 and 61–75 age groups, patients up to age 79 were included for RATHL, as data on the  < 75-year age group were not available. ABVD doxorubicin, bleomycin, vinblastine, and dacarbazine; BrECADD brentuximab vedotin, etoposide, cyclophosphamide, doxorubicin, dacarbazine, dexamethasone; CI confidence interval; eBEACOPP escalated bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine and prednisone; CI confidence interval; HR hazard ratio; ITT intent-to-treat; OS overall survival; PFS progression-free survival

BrECADD Versus PET-Guided ABVD Regimens (SWOG S0816 and RATHL)

In the base case MAIC-weighted analysis, BrECADD was associated with significantly improved PFS and OS compared with PET-guided ABVD among patients aged 18–60 (SWOG S0816) and patients aged 18–75 years (RATHL). BrECADD was associated with significantly improved outcomes compared to PET-guided ABVD per the SWOG S0816 protocol, with an HR for PFS of 0.24 (95% CI 0.16, 0.35, p < 0.001). In the OS analyses, the HR was 0.30 (95% CI 0.14, 0.63, p = 0.004) in favor of BrECADD. Similarly, in the comparison against PET-guided ABVD per the RATHL protocol, the HR for PFS was 0.24 (95% CI 0.16, 0.36, p < 0.001); in the OS analyses, the HR was 0.47 (95% CI 0.25, 0.89, p = 0.01) in favor of BrECADD.

Among patients who were PET2-positive, BrECADD was associated with significantly improved PFS compared with PET-guided ABVD regimens in the weighted comparison [SWOG S0816: HR = 0.28 (95% CI 0.13, 0.62), p = 0.005; RATHL: HR = 0.17 (95% CI 0.09, 0.31), p < 0.001]. In the analyses of OS in patients who were PET2-negative, BrECADD was significantly superior to PET-guided ABVD [SWOG S0816: HR = 0.13 (95% CI 0.03, 0.54), p < 0.015; RATHL: HR = 0.15 (95% CI 0.05, 0.46), p < 0.001].

Among patients who were PET2-negative, BrECADD demonstrated significantly improved PFS compared with PET-guided ABVD regimens in the weighted comparison [SWOG S0816: HR = 0.18 (95% CI 0.10, 0.32), p < 0.001; RATHL: HR = 0.23 (95% CI 0.12, 0.43), p < 0.001]. Of note, for patients with negative PET results, the treatment regimen was AVD in the RATHL and Mondello 2020 [8] studies. A trend favoring BrECADD was observed in the weighted OS comparisons, but the results were not statistically significant [SWOG S0816: HR = 0.39 (95% CI 0.14, 1.14), p = 0.089; RATHL: HR = 0.77 (95% CI 0.31, 1.89), p = 0.550].

Only naïve comparisons of BrECADD versus PET-guided ABVD in patients aged 61–75 were feasible, since patient characteristics were not available for this age group in RATHL, and patients over 60 were excluded from SWOG S0816. The PFS results were consistent with those for patients aged 18 to 60 years (vs. RATHL or SWOG S0816) and 18–75 years (vs. RATHL ITT), where BrECADD demonstrated a reduced risk of progression compared with PET-guided ABVD [HR = 0.45 (95% CI 0.21, 0.97), p = 0.036]. The naïve comparison of BrECADD versus PET-guided ABVD in patients aged 61 to 75 years showed an OS trend favoring BrECADD, but statistical significance was not reached [HR = 0.77 (95% CI 0.32, 1.82), p = 0.544].

RMST was calculated to account for PH assumption violations. RMST results referencing either a 30-month tau or 36-month tau for adults over age 60 and all-adults comparisons, or 60 months for comparisons with adults between 18 and 60 years of age (guided by the minimum of the average follow-up time of the studies being compared), were consistent with the Cox regressions for the provided analyses.

BrECADD Versus BEACOPP and eBEACOPP in Patients up to 75 Years of Age (HD9 and Mondello et al. 2020 [8])

In adults aged 18–75 years, there was a trend in PFS improvement [HR = 0.58 (95% CI 0.29, 1.19), p = 0.194] in favor of BrECADD (HD21 pooled, stages III and IV subgroup) compared with real-world use of eBEACOPP (Mondello 2020 [8]). The eBEACOPP Kaplan–Meier curve for OS was flat until around 30 months, whereas the HD21 pooled data were censored at 30 months in the base case; therefore, the OS comparison was not feasible, as the Cox regression did not converge.

Among patients aged 66–75, BrECADD was associated with a 71% reduced risk of death when compared with BEACOPP [HD9 (66- to 75-year-old cohort) trial]. A trend favoring BrECADD for FFTF was observed [HR = 0.63 (95% CI 0.28, 1.44), p = 0.260], but the results were not statistically significant.

The sensitivity analyses to explore the inclusion/exclusion of covariates in the model and various censoring cutoff time points were consistent with the base case results. The results for all naïve, base case, and sensitivity analyses are summarized in Table 6.

Table 6.

Unanchored MAIC results for BrECADD versus PET-guided ABVD and BEACOPP/eBEACOPP

Age	Stage	Comparator population	Outcome	HR (95% CI)
				Naïve	PH test	BC	PH test	S1a	PH test	S2b	PH test	S3c	PH test	S4	PH test	S5	PH test	S6	PH test
vs. PET-guided ABVD
18–60	III, IV	SWOG S0816 ITT [15, 16]	PFS	0.25 (0.17, 0.36)		0.24 (0.16, 0.35)		0.24 (0.17, 0.36)		0.24 (0.16, 0.35)		0.21 (0.14, 0.32)
			OS	0.33 (0.16, 0.69)		0.30 (0.14, 0.63)		0.31 (0.15, 0.65)		0.30 (0.14, 0.64)		0.22 (0.11, 0.48)
18–60	IIB, III, IV	RATHL (age ≤60) [17, 18]	PFS	0.32 (0.23, 0.45)		0.24 (0.16, 0.37)		0.29 (0.21, 0.41)		0.24 (0.16, 0.36)
			OS	0.36 (0.19, 0.68)		0.32 (0.14, 0.70)		0.34 (0.18, 0.65)		0.31 (0.14, 0.67)
18–60	III, IV	SWOG S0816 (PET2+) [15, 16]	PFS	0.25 (0.13, 0.47)	V	0.28 (0.13, 0.62)	V	0.28 (0.13, 0.57)	V	0.23 (0.11, 0.46)	V	0.24 (0.10, 0.57)	V
			OS	0.14 (0.04, 0.48)		0.13 (0.03, 0.54)		0.11 (0.03, 0.39)		0.14 (0.04, 0.49)		0.08 (0.02, 0.31)
18–60	III, IV	SWOG S0816 (PET2-) [15, 16]	PFS	0.18 (0.10, 0.31)		0.18 (0.10, 0.32)		0.18 (0.10, 0.32)		0.18 (0.10, 0.32)		0.17 (0.10, 0.30)
			OS	0.37 (0.13, 1.09)		0.39 (0.14, 1.14)		0.37 (0.13, 1.08)		0.39 (0.14, 1.13)		0.34 (0.11, 1.00)
61–75d	IIB, III, IV	RATHL (age >60) [17, 18]	PFS	0.45 (0.21, 0.97)		0.38 (0.15, 0.97)		0.39 (0.15, 0.97)		0.37 (0.15, 0.87)
			OS	0.77 (0.32, 1.82)		0.68 (0.25, 1.84)		0.68 (0.25, 1.83)		0.65 (0.25, 1.67)
18–75d	IIB, III, IV	RATHL ITT [17, 18]	PFS	0.30 (0.21, 0.42)		0.24 (0.16, 0.36)		0.27 (0.19, 0.38)		0.24 (0.16, 0.36)		0.24 (0.15, 0.37)		0.23 (0.15, 0.36)		0.23 (0.16, 0.35)		0.24 (0.17, 0.35)
			OS	0.55 (0.31, 0.97)		0.47 (0.25, 0.89)		0.51 (0.28, 0.92)		0.46 (0.25, 0.88)		0.33 (0.17, 0.64)		0.45 (0.21, 0.96)		0.39 (0.21, 0.73)		0.36 (0.20, 0.66)
18–75d	III, IV	RATHL (stage III and IV) [17, 18]	PFS	0.27 (0.19, 0.38)		0.24 (0.16, 0.35)		0.24 (0.16, 0.35)		0.25 (0.17, 0.35)		0.20 (0.14, 0.30)		0.23 (0.15, 0.34)		0.24 (0.17, 0.35)		0.27 (0.19, 0.38)
			OS	0.52 (0.28, 0.96)		0.45 (0.24, 0.84)		0.45 (0.24, 0.85)		0.48 (0.26, 0.88)		0.34 (0.17, 0.65)		0.41 (0.20, 0.87)		0.37 (0.20, 0.69)		0.31 (0.18, 0.54)
18–75d	IIB, III, IV	RATHL (PET2+) [17, 18]	PFS	0.22 (0.13, 0.36)		0.17 (0.09, 0.31)		0.20 (0.12, 0.33)		0.18 (0.10, 0.32)		0.15 (0.08, 0.29)		0.18 (0.09, 0.33)		0.17 (0.09, 0.31)		0.19 (0.11, 0.33)
			OS	0.21 (0.08, 0.52)		0.15 (0.05, 0.46)		0.17 (0.07, 0.44)		0.16 (0.05, 0.48)		0.08 (0.03, 0.23)		0.17 (0.04, 0.63)		0.13 (0.04, 0.39)		0.21 (0.08, 0.56)
18–75d	IIB, III, IV	RATHL (PET2-) [17, 18]	PFS	0.24 (0.15, 0.40)		0.23 (0.12, 0.43)		0.22 (0.13, 0.37)		0.22 (0.12, 0.40)		0.23 (0.11, 0.48)		0.21 (0.10, 0.42)		0.21 (0.11, 0.40)		0.21 (0.12, 0.38)
			OS	0.86 (0.37, 1.98)		0.77 (0.31, 1.89)		0.82 (0.35, 1.92)		0.75 (0.31, 1.82)		0.56 (0.21, 1.52)		0.58 (0.20, 1.65)		0.56 (0.24, 1.33)		0.37 (0.17, 0.80)
vs. BEACOPP/ eBEACOPP
61–75	III, IV	HD9 (61- to 75-year-old cohort) ITT [19, 20]	FFTF	0.62 (0.33, 1.17)		0.70 (0.34, 1.44)		0.60 (0.29, 1.22)		0.75 (0.39, 1.43)		0.74 (0.35, 1.55)
			OS	0.22 (0.10, 0.51)	V	0.36 (0.14, 0.98)		0.27 (0.10, 0.71)		0.28 (0.12, 0.69)		0.41 (0.15, 1.14)
66-75	III, IV	HD9 (66- to 75-year-old cohort) ITT [19, 20]	FFTF	0.70 (0.34, 1.43)		0.63 (0.28, 1.44)		0.59 (0.25, 1.37)		0.68 (0.33, 1.42)		0.61 (0.26, 1.39)
			OS	0.24 (0.09, 0.65)		0.29 (0.09, 0.92)		0.20 (0.07, 0.61)		0.25 (0.09, 0.69)		0.26 (0.08, 0.82)
18–75	III, IV	Mondello 2020 ITT[8]	PFS	0.58 (0.30, 1.13)		0.58 (0.29, 1.19)		0.62 (0.31, 1.24)		0.62 (0.30, 1.29)		0.52 (0.25, 1.08)		0.52 (0.25, 1.08)		0.50 (0.26, 0.97)		0.61 (0.33, 1.13)
			OS													1.44 (0.33, 6.16)	V	0.79 (0.31, 2.01)	V

For the comparison of BrECADD and PET-guided ABVD in the 18–75 and 61–75 age groups, patients up to age 79 were included for RATHL, as data on the < 75-year age group were not available.

^aAdjustment for significant variables only

^bAdjustment for variables that received full agreement from physicians as PV or EM

^cAdjustment for continuous age (mean = median)

^d

The PHA test is denoted with a “V” when the PHA was violated (p < 0.05).

ABVD doxorubicin, bleomycin, vinblastine, and dacarbazine; BC base case; BEACOPP bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine and prednisone BrECADD brentuximab vedotin, etoposide, cyclophosphamide, doxorubicin, dacarbazine, dexamethasone; CI confidence interval; eBEACOPP escalated bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine and prednisone; HR hazard ratio; ITT intent to treat; OS overall survival, PET2 positron emission tomography (scan conducted after two cycles of chemotherapy); PFS progression-free survival; S sensitivity analysis

RMST was calculated to account for any violation of the PH assumption. RMST results referencing either a 30-month tau or 36-month tau for adults over age 60 and all-adults comparisons, or 60 months for comparisons with adults between 18 and 60 years of age (guided by the minimum of the average follow-up time of the studies being compared), were consistent with the Cox regressions for the provided analyses.

Discussion

The comparative effectiveness of BrECADD versus PET-guided ABVD, eBEACOPP, and BEACOPP (in patients over age 60) as first-line treatment for patients with advanced Hodgkin lymphoma was assessed using an unanchored MAIC examining OS and PFS. The analyses were stratified by age (i.e., 18–60 years, older than 60 years, and 18 years and older) to provide an in-depth examination of the clinical value of treatments in populations of different ages who are at increased risk for the disease. As the MAIC approach can only account for differences in patient-level characteristics that affect outcomes, the HD21 population was restricted to age subgroups and disease severity to match the criterion in the comparator trials. Only unanchored ITCs were feasible due to the lack of common arms. Of note, unanchored comparisons are more susceptible to bias and systematic error from improper model specification when there is residual confounding [27].

In the base case MAIC-weighted analysis, BrECADD was associated with significantly improved PFS and OS compared with PET-guided ABVD (SWOG S0816 and RATHL) among patients aged 18–60 and patients aged 18–75. This trend was also observed when the groups were stratified by PET2 status, even among patients who were PET2-positive and escalated from ABVD to BEACOPP. The exception was the analysis of OS among patients who were PET2-negative, in which a numerical improvement was observed for BrECADD over PET-guided ABVD. It can be noted that in HD21, the cycle 2 PET-positive rate was observed at 236/751 (31.4%) and 234/749 (31.2%) for BrECADD and eBEACOPP, respectively, where such patients were recommended to receive successive rounds of treatment (4 additional vs. 2 additional). The PET positivity in patients treated with BECADD was high due to strict criteria for residual metabolic activity imposed by central review and can be observed as part of the increasing rate of PET2 positivity in GHSG trials over the years that is likely due to the increasing quality of the scanners detecting smaller residual fields. Recent real-world evidence has observed a PET-positive rate of 17% for patients receiving BrECADD [36]. In the 61- to 75-year-old group, BrECADD demonstrated significantly improved PFS and a trend toward improved OS compared with PET-guided ABVD (RATHL) based on naïve comparison. These findings add to the ITC literature in advanced Hodgkin lymphoma, which to date has not included comparisons of BrECADD with other interventions [33–35].

The relatively small number of OS events analyzed due to follow-up time may have limited the power to detect statistical significance. Similarly, the HD21 non-randomized cohort (aged 61–75 years) had a relatively short median follow-up time (27.1 months), limiting the generalizability and interpretability of any analysis involving these data. Thus, the long-term benefit of BrECADD among patients aged 61–75 could not be assessed using these data. In addition, derivation of modified PFS from the phase 2, single-arm cohort IPD of patients aged 61–75 in HD21 was used to match FFTF in the HD9 (66- to 75-year-old cohort) study. Assumptions were required for some patients who did not achieve complete response by the end of treatment or those who did not have this endpoint assessed, leading to potential bias. Additionally, reduced ESS following matching could also contribute to bias, but reductions in the present study ranged from 1.8% to 52.3%, which are in line with a 2025 review of MAIC in oncology, which found an average reduction of 44.9% (range of 0–88%) [37].

Authors identified all statistically significant variables through direct analysis as well as all clinically relevant prognostic variables and effect modifiers with external clinical experts from across the EU which were adjusted for in various scenarios (see Table 4 for variables identified and scenarios ran) to support a robust set of analyses informing interpretation. Where comparator studies did not report one or more potential PVs, these variables could not be adjusted for in the corresponding comparisons (e.g., ECOG PS was not reported in SWOG S0816; extranodal site was not reported in SWOG S0816 and RATHL). Additionally, studies were conducted in geographic regions with variable overlap, followed different escalation protocols, and imposed varied follow-up durations. Other limitations of comparator studies included unavailable patient characteristics for the stages III and IV subpopulations in RATHL, a treatment regimen that is no longer the SOC in HD9, and low resolution of the Kaplan–Meier curves and inconsistency between the main text and supplementary material in Mondello 2020 [8]. Another limitation of the Mondello 2020 analysis is its retrospective design. Despite robust scenario analyses, there remains a risk of residual unmeasured confounding given the heterogeneity across study designs and limitations in data availability. Additionally conducting analyses with Kaplan– Meier curves may impact the precision of the results but is still an acceptable method for conduct in the absence of comparator IPD.

Where the proportional hazards assumption was violated (at p < 0.05; see Table 6) RMST results were consistent with the primary analysis findings for each population and outcome across scenarios tested.

Despite potential limitations related to data availability and comparability, patient characteristics between trials, and the ITC methodology applied, our findings support the use of BrECADD over PET-guided ABVD, PET-guided eBEACOPP, and eBEACOPP in treating patients with advanced Hodgkin lymphoma across age groups. The current results augment the clinical evidence base from HD21, which only evaluated PET-guided BrECADD versus PET-guided eBEACOPP. The results from this MAIC are particularly relevant in informing treatment decision-making, especially given updated ESMO guidelines and NCCN Guidelines^®, which include BrECADD as a recommended regimen in this indication [6, 9]. These analyses can be revisited as evidence develops for new emerging internationally recommended regimens with additional consideration for expanded efficacy and safety outcomes where definitions align and data availability allows.

Conclusion

This set of analyses brings forward robust findings related to the comparative efficacy of the included regimens with considerations of the related necessary adjustments made and acknowledgement of limitations. Across the collection of indirect comparisons, BrECADD was associated with significantly improved PFS and OS compared with PET-guided ABVD in the adult population with advanced Hodgkin lymphoma. OS results were also significantly in favor of BrECADD versus BEACOPP in adults over age 60, with a similar trend observed for PFS. These results further support the use of BrECADD over PET-guided ABVD and BEACOPP-based treatment in adult populations with advanced Hodgkin lymphoma, both under and over age 60, and can be used to inform shared treatment decision-making between clinicians and patients.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Conceptualization: Fjoralba Kristo, Alexa Molinari, Victoria Paly, Vanessa Stache, Zhiyi Lan, Hoora Moradian. Data curation: Fjoralba Kristo, Alexa Molinari, Zhiyi Lan, Athanasios Zomas. Formal analysis: Zhiyi Lan, Zijiao Yuan. Investigation: Fjoralba Kristo, Alexa Molinari, Vanessa Stache, Victoria Paly. Methodology: Fjoralba Kristo, Alexa Molinari, Zhiyi Lan, Victoria Paly, Athanasios Zomas, Zijiao Yuan, Thomas Melchardt. Project administration: Alexa Molinari, Fjoralba Kristo. Resources: Fjoralba Kristo, Alexa Molinari, Victoria Paly. Software: Zhiyi Lan, Zijiao Yuan. Validation: Fjoralba Kristo, Alexa Molinari, Zhiyi Lan, Victoria Paly, Zijiao Yuan, Hoora Moradian. Visualization: Fjoralba Kristo, Alexa Molinari, Zhiyi Lan, Abbey Poirier, Christian Jaworek, Hishan Tharmaseelan, Thomas Melchardt. Writing—review and editing: Fjoralba Kristo, Alexa Molinari, Zhiyi Lan, Victoria Paly, Ramsha Khan, Abbey Poirier, Zijiao Yuan, Athanasios Zomas, Christian Jaworek, Hishan Tharmaseelan, Hoora Moradian, Thomas Melchardt. All authors have reviewed and approved the final version of this manuscript to be submitted and take responsibility for the content of the article.

Funding

The study as well as the journal’s rapid service and open access fees were funded by Takeda Pharmaceuticals. Study design, execution, analysis, and reporting was completed by the sponsor, Takeda Pharmaceuticals, in conjunction with input from co-authors.

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Conflict of Interest

Fjoralba Kristo, Alexa Molinari, Victoria Paly, Athanasios Zomas, and Vanessa Stache report employment and ownership of stock/shares with Takeda. Zhiyi Lan, Zijiao Yuan, Abbey Poirier, and Hoora Moradian were employed by Cytel Inc., at the time of this study, which received consulting fees from Takeda associated with the study and manuscript. Christian Jaworek and Hishan Tharmaseelan report employment by the Department of Internal Medicine, Faculty of Medicine and University Hospital Cologne, Center for Integrated Oncology Aachen Bonn Cologne Duesseldorf (CIO ABCD), German Hodgkin Study Group (GHSG), University of Cologne, Cologne, Germany. Thomas Melchardt reports employment by Paracelsus Medical University, Salzburg, Austria.

Ethical Approval

This article reports a retrospective analysis using individual patient data from the HD21 trial and aggregate data extracted from published studies (SWOG S0816, RATHL, HD9, and Mondello 2020 [8]). Review boards at participating institutions had previously approved all procedures for the HD21, SWOG S0816, RATHL, and HD9 trials. As the present study is a retrospective analysis of existing trial and published data and did not involve new patient recruitment or intervention, no additional institutional review board review was required for this analysis.