Oki 2014.

Study characteristics
Methods	Secondary citation(s) NA Language of publication English Study design Retrospective, single‐centre study Study centre(s) MD Anderson Cancer Center, Houston, Texas, USA Country USA Median follow‐up time (range) 45 months
Participants	Number of included participants Total: 325 229 participants with PET2 analysed 96 participants with PET3 excluded post‐hoc Inclusion criteria Classic HL Treatment with ABVD Availability of interim PET scan Exclusion criteria Additional treatment (e.g. with brentuximab vedotin or rituximab) except for radiotherapy Consent Not reported Recruitment period January 2001 to May 2011 Age (range, in years) Group I (early‐stage non‐bulky): 32 (median, 18‐77) Group II (stage II bulky): 36 (20‐60) Group III (advanced stage IPS ≤ 2): 30 (19‐79) Group IV (advanced stage IPS ≥ 3): 49 (19‐84) Ethnic group(s) Not reported Stages of disease All stages Comorbidities Not reported Therapy regimen ABVD with or without radiotherapy
Prognostic factor(s)	Prognostic factor(s) Interim PET Definition of prognostic factor(s) Not reported Timing of prognostic factor measurement After 2 or 3 cycles of ABVD Method for measurement (use of specific scale and cut‐off) Deauville 5‐point scoring system Scores of 1‐3 considered negative, scores of 4‐5 considered positive Independent assessment by 3 nuclear medicine physicians Was the same definition and method for measurement used in all participants? No, 10 participants had only PET without CT scan Were prognostic factor(s) assessed blinded for outcome(s), and for each other (if relevant)? Yes
Outcome(s)	Primary outcome(s) and definition(s) Progression‐free survival (PFS), defined as the time from diagnosis to disease progression, relapse or death from any cause Secondary outcome(s) and definition(s) None Timing of outcome measurement At 3 years Was the same definition and method for measurement used in all participants? Yes Was/were outcome(s) assessed blinded for prognostic factor(s), and for each other (if relevant)? No
Missing data	Participants with any missing value? No If yes, how were missing data handled? Not applicable
Analysis	Univariable analysis: Yes Total number of participants included in univariable analysis for each outcome PFS: all Statistical method Kaplan‐Meier survival curves with log‐rank test per subgroup Univariable Cox proportional hazard models How was the prognostic factor treated? Binary Multivariable analysis: No
Risk of bias (QUIPS)	Study participation Low risk Clear description of participants and study characteristics. Study attrition Low risk No loss to follow‐up. Prognostic factor measurement Low risk Adequate measurement and description. Prognostic factor measured the same way for all participants. Outcome: Overall survival Not reported Outcome: Progression‐free survival Outcome measurement Low risk Clear definition. Outcome measured the same way for all participants. 'Other prognostic factors (covariates)' High risk Disease stage not accounted for. Statistical analysis and reporting High risk Exclusion of participants with PET3 during analysis due to lack of prognostic value. Stratification according to disease stage resulted in small sample sizes per subgroup. Outcome: Adverse events Not reported
Notes	Conflict of interest No conflict of interest to disclose for the study. Funding Not reported