Protocol Amendment 1 concerned the clarification of PFS definition and the modification of censoring rules. For patients “who received any new lymphoma-directed therapy (other than rituximab as maintenance) before progression of disease”, the date of censoring was defined as the date of the “last radiologic assessment prior to the start of the new therapy”, instead of the “date of first administration of additional treatment”.

Protocol Amendment 2 concerned mainly: - The change in the study primary objective: OS, initially a secondary objective, was added to PFS as a combined primary objective. Statistical methods including sample size determination were updated taking into account this modification. - The increase in the number of patients to be randomized (from 300 to 350). - The replacement of the stratification factor “prior SCT” by “length of time from initiation of first-line therapy for aggressive NHL until first relapse”. - The addition of safety criteria (study-related AEs and some cardiac events and assessments) during the Follow-up periods. - As regards to the reporting of AEs, it was also specified that cardiac AEs ≥ grade 3 were to be collected until the end of the study and followed until resolution or no further improvement was expected. - Specifications on clinical examination were added. - The modification of inclusion criteria (bone core biopsy was to be obtained within 8 weeks prior to randomization, addition of the necessity to confirm the response to CHOP-R by a second radiographic assessment; removal of the 24-week delay between day 1 of last cycle of CHOP-R or equivalent treatment and subsequent relapse, and update of the laboratory requirements for platelets and absolute neutrophil count). - The addition of primary refractory aggressive NHL as an exclusion criterion. - An update of pixantrone and gemcitabine dose adjustments and delays for hematologic toxicity. - Specifications for low-dose corticosteroids use as concomitant medication. - For the selection of target lesions, measurable lesions in a previously radiated site could not be considered target lesions. - Wording specifications. - Procedures specifications.

Protocol Amendment 3 included: - An update of Safety information in the background section. - The clarification of the use of the terms “NHL” and “DLBCL”. - The modification of inclusion criteria: bone marrow biopsy criteria were clarified, prior CHOP-R was allowed for any type of NHL, patients with transformed follicular lymphoma who may not have received CHOP-R as first-line therapy for aggressive NHL could also be included in the study, definition of measurable disease was adjusted to be consistent with other Sections in the protocol, and it was specified that DLBCL diagnosis was to be confirmed by pathologic review. - The modification of exclusion criteria: definitions of measurable disease and primary refractory disease were updated and clarified, CNS involvement was further detailed, and enrolment of patients who had had certain low-risk cancers commonly found in this population was finally allowed. - Disease Assessment Criteria were extensively clarified. - PET scan requirements were modified. - The prior requirement for central evaluation of echocardiograms was removed, since implementing the central read processes negatively impacted timely site initiation and accrual of patients. There was no clear need for a retrospective central read, but there was considerable negative impact; therefore, central read was deleted. Local echo evaluations were used for safety and treatment decisions. - The timing of IDMC meetings was clarified. - A recommendation was added that rituximab be administered first, and the section regarding drugs administration was reorganized to allow investigators more clearly identifying potential rituximab-related adverse reactions. The window for pixantrone administration was clarified. - The requirement for physical examinations before study drug administration was clarified. - Procedures for dose adjustments and delays were detailed. - Safety reporting was clarified and revised. - Pathology procedures were clarified.

Protocol Amendment 4 aimed to: - Change the primary endpoint of the study to “overall survival” only, following FDA recommendations. PFS was therefore a secondary objective. All the protocol (primary and secondary endpoints, statistical analyses, etc) was modified accordingly. - An interim analysis of OS, to be done when 150 deaths (50%) had occurred, was planned. The final OS analysis was to be performed when 300 deaths had occurred. - Stratification factors were adjusted. - Criteria for eligible patients were modified and clarified to ensure safety and enrolment of the target population. - Bone marrow biopsy requirements were revised. - Troponin sample requirements and processing were clarified. - Dose adjustments and delays for toxicity were detailed. - Follow-up for randomized patients versus not randomized patients was clarified. - The definition of measurable disease was slightly revised. - PET scan requirements were clarified.

Protocol Amendment 5, for North America (NA), concerned: - For the primary objective, eligibility of patients was further detailed: patients were to “have no progression for at least 12 weeks after last dose of a treatment regimen” instead of “were to have had a response to prior therapy”. - In response to a recommendation from the EMA, pixantrone dose was expressed in its base form (instead of its salt form) in the whole document. - It was specified that the study would be conducted in the United States, Canada, Eastern and Western Europe and South America. - Requirements for entering Survival Follow-up period were updated. - The formula to calculate doxorubicin equivalent dose was replaced. - Wording specifications. - Procedures specifications.

Protocol Amendment 5 NNA, for Non-North America (NNA), included: - The rationale for rituximab-pixantrone combination was further detailed (NA + NNA). - It was specified that the study would be conducted in NA, Western Europe and potentially Eastern Europe (NNA). - It was specified that the study would be conducted in NA, Western Europe and Eastern Europe for (NA). - Inclusion criterion on contraceptive methods to be used was updated (NA + NNA). - Possible adverse reactions associated with rituximab were added and concerned cytokine release syndrome, hypersensitivity reactions and anaphylaxis and progressive multifocal leukoencephalopathy (NA + NNA). - Recommendations were given on the use of some concomitant medications (especially those metabolized by CYP1A2) when co-administered with pixantrone (NA + NNA). - In response to EMA, the pixantrone dose was expressed in its base form (instead of its salt form) in the whole document (NA).

Protocol Amendment 6, for North America (NA) included: - The rationale for rituximab-pixantrone combination was further detailed (NA + NNA). - It was specified that the study would be conducted in NA, Western Europe and potentially Eastern Europe (NNA). - It was specified that the study would be conducted in NA, Western Europe and Eastern Europe for (NA). - Inclusion criterion on contraceptive methods to be used was updated (NA + NNA). - Possible adverse reactions associated with rituximab were added and concerned cytokine release syndrome, hypersensitivity reactions and anaphylaxis and progressive multifocal leukoencephalopathy (NA + NNA). - Recommendations were given on the use of some concomitant medications (especially those metabolized by CYP1A2) when co-administered with pixantrone (NA + NNA). - In response to EMA, the pixantrone dose was expressed in its base form (instead of its salt form) in the whole document (NA).

Protocol Amendment 6 NNA, for Non-North America (NNA), concerned mainly: - In response to EMA, the pixantrone dose was expressed in its base form (instead of its salt form) in the whole document. - It was specified that the study would be conducted in NA, Eastern and Western Europe. - Procedures for reporting AEs updated.

Protocol Amendment 7 NNA, for Non-North America (NNA), and aimed to: - Entirely revise the synopsis consistently with the protocol body: several sections were added (No. of study, Name of sponsor, Phase of development, Investigational Drug, Duration of Study and Treatment, Planned No. of Patients, Test Product and Reference therapy, Treatment regimen, Inclusion and Exclusion criteria, Schedule of treatment and assessments, Criteria for evaluation (renamed section title). - Change in EOT window: The EOT visit was to occur at 4 to 7 weeks after last study drug dose, inclusive, or before non-protocol NHL therapy was given, whichever occurred first instead of “the EOT visit occurs at 5 weeks ± 1week after the last study drug administration or before non-protocol NHL therapy is given”. - Gemcitabine dose modifications for hematologic toxicity were completed. - Gemcitabine and pixantrone dose modifications for non-hematologic toxicity were completed. - Concomitant medications were updated: additional recommendations for drug’s photosensitivity. - Use of local laboratories for urgent clinical decisions: Local labs were required, as needed, for time-limited evaluation windows, per protocol, and to support urgent clinical decisions. - Details were added for the description of visits schedules and assessments (for the screening period, the End of treatment visit, and Follow-up periods). - The description of disease assessment was also modified (in case no EOT PET scan could be obtained, for bone marrow biopsy…). - Specification on AEs and AEs reporting were added. - Some minor revisions and wording specifications were implemented.

Protocol Amendment 8, for NA and NNA, unified previous NA and NAA versions. Major changes were: - Primary endpoint (previously OS) was replaced by PFS. - Number of subjects to be randomized was decreased from 350 to 260 patients. Enrolment was to be continued until 195 PFS events occurred, or approximately 260 patients were enrolled, whichever occurred first. Enrolment period was planned approximately 60 months from study initiation. - It was specified that no interim analysis of PFS was planned. - Added exploratory objectives (assessment of DOR and DCR between treatments, proportion of patients who received a SCT after study treatment). - Rituximab given as maintenance therapy was not allowed prior to the EOT visit per protocol window. - Added precisions for patients entering Follow-up periods. - Added definition of EOS (i.e. date when the required OS events have occurred or the study is terminated). - Criteria for efficacy evaluation were further detailed. - Consequently to the change in the objectives and primary criteria, statistical methods were adapted, including the sample size re-estimation. - Updated pixantrone safety in clinical study in the background and rationale. - Precisions were given for patients still on treatment at the end of recruitment (to continue per protocol until EOT and enter the survival follow-up or directly to enter the survival follow-up period). - Added details for completion and discontinuation of treatment. - Added information about pixantrone handling, preparation, administration and storage. - Added recommendations for drug’s for drug’s photosensitivity. - Visit schedule and assessments as well as imaging ad criteria for response, were further detailed. - Updated definition and reporting of AEs. - Added length of time from initiation of therapy for DLBCL or FL grade 3 until first relapse to subgroup analyses. - Added informed consent for collecting survival information. - Added PK study.

Protocol Amendment 9 aimed to: - Updated the total number of patients to be enrolled, after simulation performed by the sponsor, from 260 to 320 patients, to reach the 195 PFS events (per IRC) within a reasonable timeline to meet the study report due date. - Increased enrolment period from 60 to 80 months, to accommodate the increase in the number of planned patients’ enrolment. - Added one additional interim analysis for secondary efficacy endpoint OS, i.e. when 190 OS events (86% of the required number of OS events) have occurred, due to much slower occurrence of OS events than expected. Indeed, based on the updated projection, the time to achieve 220 OS events is year 2022, more than 3 years from the first interim analysis. Adding the 2nd OS interim analysis would provide a chance to stop the trial earlier if the treatment demonstrated superior survival benefit. - The hierarchy order for testing the secondary efficacy endpoints was updated. In view of the importance of OS in these patients and in order to better match the study objectives, there was reorganization of the order of endpoints testing, to put OS ahead of overall response and CR, in the hypothesis testing hierarchy of secondary endpoints. - Clarified treatment blinding. - Added a sub-group analysis. As the current indication is in 3rd and 4th line, a subgroup analysis defined by 0-1 line versus more than 2 lines was added to confirm the efficacy and safety of pixantrone in the current indication and to evaluate them in 2nd line.