Radiotherapy for Advanced Hodgkin Lymphoma with Initial Bulk: A Combined Analysis of Two Randomized Trials

Abstract

Purpose

The role of consolidative radiation therapy (RT) in patients with advanced Hodgkin lymphoma with initial bulk is unclear. GITIL/FIL HD0607 and FIL HD0801, 2 randomized controlled trials with similar design and methodologies, did not identify a benefit to consolidative RT after a metabolic complete response to 6 cycles of doxorubicin, bleomycin, vinblastine and dacarbazine. However, their limited sample sizes reduced statistical power to detect a small but clinically meaningful benefit to RT.

Methods and Materials

In a secondary analysis of these 2 phase 3 trials, reconstructed patient data were used to compare outcomes for early and complete responders randomized to no RT or RT to the site(s) of initial bulk. Estimates of progression-free survival (PFS) in the intent-to-treat (ITT) and per-protocol (PP) analyses were generated using the combined data and compared between groups using the log-rank test.

Results

A total of 412 patients were included in the ITT analysis, and 373 patients were included in the PP analysis. Median age was 30 to 32 years, 42% of patients were stage IIB, and 73% of bulky sites were located in the mediastinum. For the no RT versus RT groups, 5-year ITT PFS estimates were 90.1% versus 90.1%, respectively (P = .81). Five-year PP PFS rates were 90.9% versus 92.9%, respectively (P = .31). There was no observed difference between no RT and RT groups in subgroups according to size of bulky disease: 5 to 7 cm (P = .78), 7 to 10 cm (P = .25), and >10 cm (P = .69).

Conclusions

In this combined analysis of 2 randomized phase 3 clinical trials, consolidative RT to initial sites of bulky nodal involvement was not associated with a PFS benefit in patients with advanced Hodgkin lymphoma in metabolic complete response after 2 and 6 cycles of doxorubicin, bleomycin, vinblastine and dacarbazine.

Introduction

Radiation therapy (RT) plays an integral role in the management of early stage Hodgkin lymphoma (HL), but its role in advanced disease is uncertain.^1,2 While consolidative RT to the sites of partial response after chemotherapy improves progression-free survival (PFS), current data do not support consolidative RT to sites of initial bulk that have achieved metabolic complete response (CR).3, 4, 5

The GITIL and FIL cooperative groups executed 2 randomized trials for adults with advanced HL to examine treatment intensification in patients with residual positron emission tomography (PET)-positive disease after 2 cycles of doxorubicin, bleomycin, vinblastine and dacarbazine (ABVD).^6,7 Both trials also randomized patients with bulky disease (nodal mass >5 cm) at diagnosis that achieved a metabolic CR after 2 and 6 cycles of ABVD to RT or observation.^8,9 Although these studies attempted to address the utility of RT in the setting of bulky disease, each study suffered from limitations including absence of a predefined statistical design⁸ and small, underpowered sample size.⁹

The similarities between the 2 studies’ design, treatment era, inclusion criteria, chemotherapy regimen, use of PET imaging, RT dose, and prespecified endpoints provide a unique opportunity to combine data and further examine the efficacy of RT in the setting of bulky disease. Through a combined analysis of reconstructed patient data from these published trials, we aimed to investigate the role of consolidative RT in patients with bulky advanced HL who experience metabolic CR after chemotherapy.

Methods and Materials

HD0607 and HD0801 were specifically selected for a combined analysis due to their similar study design, patient population, inclusion/exclusion criteria, treatment, and endpoint of interest definitions. A systematic review was performed to evaluate for other contributing trials and yielded no additional findings (Appendix E1 and Fig. E1). Both trials enrolled adults with advanced HL (stage IIB-IV) who had at least 1 site of bulky disease (largest dimension ≥5 cm) measured on transverse sections of the computed tomography (CT) imaging and achieved metabolic CR by PET/CT after 2- and 6-cycles of ABVD. The Consolidated Standards of Reporting Trials diagram for the combined analysis is displayed in Fig. E2.

RT procedures were overall consistent between studies, with sites of initial bulk specified to be treated with approximately 30 Gy using conventional fractionation. RT detail is provided in Table E1. The primary endpoint of this combined analysis was PFS, defined in both studies as time from registration/randomization to the date of progression/relapse or death from any cause.

Reconstructed patient data were extracted from the published figures using a robust, 2-stage, iterative extraction process based on the Kaplan-Meier estimation method (Appendix E1).¹⁰ This established methodology is used to analyze time-to-event data when specific patient-level data are either not available or when specific data points are not consistently reported across multiple studies.^10,11 Reconstructed outcome estimates and Kaplan-Meier plots for each trial are demonstrated in Table E2 and Figs. E3 and E4.

Descriptive analyses were performed on categorical data where appropriate to summarize the combined cohort. Survival outcomes in the combined analysis were estimated using the Kaplan-Meier method. The log-rank test was used to test for differences between groups. Statistical analyses were performed using R version 4.2.

Results

Baseline patient and disease characteristics of both trials that have been previously reported are summarized in Table E3. Of all 445 bulky sites (patients on HD0706 could have >1 bulky site), 73% were in the mediastinum and 27% were located at sites other than the mediastinum. The maximum size of the bulky lesion(s) was evenly distributed across the 3 size categories (5-7 cm, 7-10 cm, 10+ cm).

In the intent-to-treat analysis, data from all randomized patients on both trials (n = 412) were used to estimate PFS (Fig. 1). The 5-year PFS rates in the no RT versus RT groups were 90.1% (95% CI, 86.0-94.3) versus 90.1% (86.0-94.5), respectively (P = .81).

Figure 1.

Progression-free survival in the combined intent-to-treat analysis.

A total of 373 patients were included in the per-protocol analysis: 191 in the no RT group and 182 in the RT group (Fig. 2). Five-year PFS rates in the no RT versus RT groups were 90.9% (95% CI, 86.9-95.1) versus 92.9% (89.1-96.9), respectively (P = .31).

Figure 2.

Progression-free survival in the combined per-protocol analysis.

Both studies reported the size of largest bulky disease for all 412 patients randomized but only reported PFS outcomes based on size using the per-protocol cohort (n = 373). Combined 5-year PFS estimates between treatment groups stratified by size of largest bulky disease are summarized in Table E4. There were no PFS differences between the no RT versus RT groups by size category, nor was there an association between size and PFS in patients who did not receive RT (Fig. E5; P = .59).

Discussion

In this unique combined analysis of 2 randomized trials, we failed to demonstrate an association between consolidative RT and PFS in patients with bulky advanced HL. These findings are particularly important considering the controversy surrounding the role of RT for this patient population, where current data are limited, and its routine use is difficult to justify. We performed a systematic review to rule out the existence of other randomized trials reporting outcomes analogous to those included in this combined analysis and found no additional studies. Though both studies produced high-quality randomized evidence, each is individually underpowered, and one was not powered by a defined statistical design “mainly because of a lack of published comparators.”^8p2 Each individual trial on its own would not be expected to provide adequate statistical power to detect a small improvement in PFS. However, trials that enroll the same population and randomize patients to similar treatment groups offer an opportunity for combined analysis.¹²

The evidence supporting the use of RT in HL is heterogeneous and difficult to interpret in the context of modern therapy. Its use has been associated with improved PFS in prior non-ABVD-based trials without PET response imaging.^1,13 Considering the current area of treatment for advanced HL with ABVD guided by PET/CT imaging, it is unclear whether the results of this trial with older chemotherapy and without functional imaging remain valid. Our findings are more consistent with Alliance/CALGB 50801, a phase 2 trial of PET-adaptive therapy for patients with bulky, early stage HL who were treated with 6 cycles of ABVD alone without RT in the setting of metabolic CR.¹⁴ The RATHL trial did not allow consolidative RT for advanced HL patients in metabolic CR and reported similar rates of 3-year PFS.¹⁵ RATHL was identified but not included for combined analysis because of the lack of PET/CT imaging after completion of chemotherapy. A combined analysis of patterns of failure in HD0607 and HD0801 was not possible given inconsistency in reporting, but failure within sites of initial bulk was rare. These data suggest that local control at the bulky site(s) would only reduce the risk of progression by a small amount; an adequately powered trial to clarify this benefit may not be feasible. Even a small difference in PFS may not provide clinical benefit with modern targeted salvage systemic therapies.^16,17

This study is limited by the lack of individual patient-level data not amenable to reconstruction, precluding sensitivity analyses of predictors of recurrence and limiting our ability to control for factors such as International Prognostic Score, site of bulk, or RT dose. Thus, these findings should be interpreted with caution.

Conclusion

In conclusion, this combined analysis of 2 randomized trials was unable to detect a PFS benefit to consolidative RT in patients with advanced HL who present with initial sites of bulky disease and have achieved metabolic CR after 2 and 6 cycles of ABVD.

Disclosures

The authors have no conflict of interest to disclose. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.