The protocol was updated to add the EudraCT number to the cover page.

The protocol was amended for the following reasons: • To provide additional detail in the study objective and endpoints. • To exclude patients unable to give consent by removing the option for a legally acceptable representative to consent. • Revise the entry criteria to: o Increase the required platelet count for study entry to ≥75,000/μL o Require females of childbearing potential and males who have partners of childbearing potential to use 2 effective contraceptive methods and to specify which methods are considered effective o Exclude subjects with major surgery less than 30 days prior to first dose of study drug , live vaccines within 1 month of first dose of study drug, current severe immunodeficiency, history of recurring or chronic infections or with underlying conditions which may further predispose patients to serious infection, or concomitant use of bleomycin. • Require study discontinuation in the case of pregnancy. o Require pretreatment assessment for hepatitis B. o Require electrocardiogram assessment at each cycle. o Require pregnancy testing (for women of childbearing potential only) at each cycle and require that pregnancy testing be by serum β-hCG • Clearly specify the requirement for physical examinations and require physical examination to include neurological examination at all time points where physical examination is required, including where it is required as part of the lymphoma assessment. • To clarify that sampling for brentuximab and MMAE exposure was only required for subjects on the treatment arm (brentuximab vedotin) and to extend this sampling to all combination treatment cycles (up to Cycle 6). • To limit ATA sampling to the treatment arm (brentuximab vedotin). • Clarify that safety reporting is required in all regions of study conducted as required by local and international standards. • To update the reference to the NCCN guideline for the prevention and treatment of cancer-related infections.

The protocol was amended to • Clarify that the objectives of the study were to investigate the effects of study treatment both in patients with DLCBL and those with follicular NHL grade 3b • Allow a second dose reduction of bendamustine (to 50 mg/m2) • Recommend prophylactic growth factor support and clarified that transfusions and intrathecal prophylactic treatment for cerebral/meningeal disease are permitted • Correct an error regarding timing of posttreatment assessments: The timing of all posttreatment assessments is relative to Cycle 1 Day 1 (not the end of treatment) • Add an additional safety assessment visit during the second week of Cycles 1 and 2 • To change the definition of PFS and OS to begin at randomization (rather than date of first treatment) and to include subsequent treatment for lymphoma (other than post-treatment consolidative radiotherapy, post-treatment chemotherapy for the purpose of mobilizing peripheral blood stem cells, and consolidative autologous or allogeneic SCT) as an event in the progression analyses to reflect current statistical planning • Remove the requirement that biopsy samples be obtained within than 1 year before screening • Clarify that tumor samples are only to be collected from archival samples; no tumor biopsies are required to be performed for this study • Make minor editorial and formatting changes throughout the protocol

The protocol was amended to • Reflect that Teva Pharmaceuticals is discontinuing distribution of the bendamustine hydrochloride liquid formulation TREANDA, which was being used at all US sites in this study. Investigators may continue to dispense TREANDA until its expiration date, as described in previous versions of the protocol and in the pharmacy manual. When acquiring new supplies, sites are to acquire bendamustine hydrochloride lyophilized powder. • Allow the assessment of CD30 for eligibility assessment to be performed either by local or central laboratory • Clarify that a new biopsy may be required for subjects with relapsed disease if one has not been obtained since relapse • To correct an inconsistency between the schedule of follow-up assessments and the text description of follow-up events • Make minor editorial and formatting changes throughout the protocol