Amendment 1: • Added crossover to treatment with selinexor and dexamethasone (SdX) as an option for subjects in the Vd arm after PD was confirmed by the IRC if they had significant tolerability issues with bortezomib (e.g., higher than Grade 2 PN or Grade 2 or higher PN with pain). • Changed the third key secondary efficacy objective/endpoint from DOR to OS. • Revised the OS1 and time-to-next-treatment secondary objectives to add SdX and added a new exploratory objective (i.e., to assess disease response to SdX treatment) and endpoint (i.e., IMWG response criteria for subjects treated with SdX) for SdX to assess response for subjects who crossed over to SdX. Also clarified that for OS1, subjects on the Vd arm who crossed over were censored at the date of crossover. • Revised the definition of the time to response to “the duration of the time from randomization to the first documented response (>= PR) per IMWG response criteria”. • Updated the IMWG response criteria for myeloma to align with the most recent IMWG criteria (Kumar 2016). The definition for minimal residual disease was changed from “minor” to “minimal” response to align with the IMWG Consensus Criteria. • The process for crossover was modified to prevent premature crossover. •Revised exclusion 12 to clarify that subjects treated with an investigational anticancer therapy within 2 weeks before C1D1 were specifically excluded from the study. • Clarified that symptom-directed physical examinations were only to be performed if clinically indicated • Clarified that clinical plasmacytoma assessments are to be performed if clinically indicated at MM Disease Assessment Visits and at Durability of Response and Survival Follow-up Visits. Also corrected the window for detection of plasmacytomas at baseline by physical examination/palpation from “within 45 days” to “within 28 days” before C1D1. • Clarified that a skeletal survey was required at the End of Treatment Visit.

Amendment 2: • Added details for the Interactive Response Technology system that was to used to perform treatment randomization. • Added details for continuation of the study treatment for subjects if the study was terminated early to comply with International Council for Harmonisation Good Clinical Practice E6. • Clarified that double-barrier contraception methods were considered effective but not highly effective to align with the recommendations of the Clinical Trial Facilitation Group. Also clarified that sexual partners who were surgically sterilized were not exempt from the contraception requirements unless they were “permanently” surgically sterilized. • Added the requirement for pregnancy testing (serum human chorionic gonadotropin or urine) for females of childbearing potential before dosing on Day 1 of Cycles >= 2 to align with the recommendations of the Clinical Trial Facilitation Group. • For the PFS primary efficacy endpoint, changed the analysis to the stratified log-rank test and stratified Cox model (previously in Version 1.0 of the protocol). Also, specified that the stratified log-rank test was to be used for the secondary analyses of OS, DOR, and OS1, and that the exploratory analysis of the treatment discontinuation rate was to be performed using the stratified log-rank test. • For the ORR efficacy analysis, specified that subjects missing MM disease assessments after C1D1 were to be imputed as non-responders. • Changed the timing of the secondary analyses from “after significance is reached for PFS” to “at the time of ORR analysis” and specified that “statistical significance of the secondary endpoints will not be claimed until the ORR and PFS have reached significance.” • Changed the Hochberg procedure for testing the secondary endpoints to a hierarchical testing procedure.

Amendment 3: • Changed ORR from a primary endpoint to a key secondary endpoint to address concerns expressed by the Agencies regarding including ORR as a primary endpoint (i.e., an analysis of ORR could jeopardize the integrity of the study for the ultimate assessment of PFS). • Revised the definition of “IRC-confirmed PD” and renamed the term as “IRC PD confirmation.” • Added a description of 8 tumor lysis syndrome cases reported across all of selinexor development as of May 2018. • Removed the split of the alpha level between PFS (0.02) and ORR (0.005) and added the assumed exponential dropout rate of 0.65%. • Revised the total number of PFS events required for the final analysis from 284 to 267, for the IA for sample size re-estimation from 85 to 81, and for the IA for futility or superiority from 213 to 201. • Removed the secondary OS1 objective/endpoint. • Removed the secondary objective/endpoint comparing ORR, PFS, and DOR for subjects with 1 versus >1 prior anti-MM regimen and added it as a subgroup exploratory analysis. • Revised the basis for the determination of the Revised International Staging System stage used in stratification of randomization from “at original MM diagnosis” to “at study entry, based on screening results.” • Added an exception to the requirement that subjects were to either remain on the study treatment until PD was confirmed by the IRC or until the subject discontinued the study treatment, completed the End of Treatment Visit, and was followed for survival. The exception only applied to subjects in the Vd arm who had to terminate bortezomib prior to IRC confirmed PD due to significant toxicities. • Revised inclusion criterion 12 for contraception requirements and guidance for pregnancy and breastfeeding. • Clarified the wording for how selinexor should be administered and removed the need to take selinexor with food.