Involved Field Radiotherapy Versus No Further Treatment in Patients with Newly Diagnosed Stage 1A or 2A Hodgkin Lymphoma and a ‘Negative’ PET Scan after 3 Cycles ABVD. Survival and Cause of Death after a Median of 16 Years Follow-up in the UK RAPID Trial

Background

In early-stage Hodgkin lymphoma (HL) abbreviated chemotherapy (ACT) followed by involved field radiotherapy (IFRT) is commonly employed but results of the RAPID trial (New England Journal of Medicine, 2015) showed that most patients achieving PET ‘negative’ status are probably cured by ACT alone. Patients taking part in RAPID received 3 cycles ABVD followed by a centrally-reviewed PET scan; those with a ‘negative’ PET scan (score of 1 or 2 on a 5-point scale) were randomised 1:1 to IFRT or no further treatment (NFT), whilst those with a ‘positive’ PET scan had a 4th cycle ABVD and IFRT. Here we report long-term survival and cause of death, obtained using data linkage approaches, for patients participating in the RAPID trial.

Methods

602 patients (53.3% male, 46.7% female; median age 34 years) with newly diagnosed, stage1A or 2A HL and no mediastinal bulk, were recruited to RAPID from across England, Wales, Scotland and Northern Ireland between 2003 and 2010. Following 3 cycles ABVD, 571 pts had a PET scan of which 426 were classified as PET ‘negative’ (score 1, n=301; score 2, n=125) and 420 were randomised to receive IFRT (n=209) or NFT (n=211).

Long-term survival and cause of death data for RAPID patients recruited from England and Wales were obtained through linkage to national death registration data via the National Health Service (NHS) Digital Data Access Request Service (DARS) using NHS number and date of birth. Disease progression data are not available via national registries, and consistent trial mandated follow-up ceased after publication of the primary results in 2015 when the median follow-up was 60 months and most progressions or relapses will have already been observed. Patients recruited from Scotland and Northern Ireland were not included in the DARS linkage exercise due to required approvals, however their available trial data are included in the analysis.

Results

Of the total 602 RAPID patients, 535 were from England and Wales and 459 of these were matched by DARS. Of the 420 randomised PET ‘negative’ patients, 374 were from England and Wales and 319 were matched by DARS. Age, sex, stage, number of nodal sites, EORTC/GHSG prognostic groups and PET scores were similar in DARS matched and not matched patients in the PET ‘negative’ group, furthermore, randomised arms were balanced. The DARS matched data were combined with the available trial data of patients who could not be matched, including those from Scotland and Northern Ireland, to give a median follow-up of 16.0 years (inter-quartile range: 13.7-17.7) for the entire RAPID population on which the following survival and cause of death data are based.

From the randomised PET ‘negative’ group (n=420), there have been 11 deaths in the IFRT arm (respiratory, n=4; cardiac, n=2; non-Hodgkin lymphoma (NHL), n=2; carcinoma, n=1; HL, n=1; stroke, n=1) and 15 deaths in the NFT arm (NHL, n=4; carcinoma, n=3; cardiac, n=3; respiratory, n=2; hepatic failure, n=1; stroke, n=1; suicide, n=1). The 15-year overall survival rates are 94.2% (95% CI: 89.7-96.8) and 92.3% (95% CI: 87.0-95.5) respectively, with a hazard ratio (IFRT vs. NFT) of 0.71 (95% CI: 0.33-1.55), p=0.39. Of the 145 PET ‘positive’ patients after 3 cycles ABVD, 20 have died (HL, n=7; respiratory, n=6; infection, n=2; carcinoma, n=1; cardiac, n=1; dementia, n=1; haemophagocytic syndrome, n=1; NHL, n=1).

Conclusions

Utilising UK national death registration data for patients recruited to the RAPID trial, there is no evidence, after 16.0 years median follow-up, of a difference in overall survival between patients in the IFRT arm and the NFT arm achieving PET ‘negative’ status following 3 cycles of ABVD. It is of note that in the PET ‘negative’ group there has been one death from HL (IFRT arm). Commonest causes of death, distributed across both randomised arms, were NHL (n=6), respiratory (n=6), cardiac (n=5) and carcinoma (n=4). In the PET ‘positive’ group there were 7 deaths from HL highlighting the known adverse prognostic significance of a positive PET scan after chemotherapy.

These results show that in newly diagnosed patients with stage 1A or 2A HL and no mediastinal bulk, who have a ‘negative’ PET scan following 3 cycles of ABVD, consolidation radiotherapy has no impact on long-term survival. Further work to investigate the possible effects of RT and ABVD on the incidence of non-fatal second cancers and cardiac disease, through data linkage to relevant national registries, is underway.