(1) It was clarified that certain additional safety criteria must have been met in order to resume either rituximab/placebo or MMF/placebo after treatment interruption, and to receive repeat rituximab/placebo infusions on Day 168 and Day 182. (2) The option to administer rituximab/placebo study drug at a faster infusion rate at Day 168 and Day 182 was removed. (3) Progression of PV, worsening, disease flare, or treatment failure was not to be recorded as an AE or SAE, and was to be captured as efficacy assessment data only. (4) The following sections were added: a section on dosage modification and treatment interruption, a section on the efficacy and safety of MMF in PV, and a section on side effects known to be associated with MMF. (5) Additional criteria for study treatment discontinuation were added, including severe allergic or anaphylactic study treatment-related reaction, NCI CTCAE Grade 4 (life-threatening) event during or within 24 hours of an infusion, and pure red cell aplasia. (6) Text was added to clarify the management of specific AEs, particularly infections and IRRs: - Infections: repeat infusions of rituximab (or matching placebo) were only to be considered after any infection had fully resolved. Additionally, it was clarified that infections should be treated according to the local standard of care instead of symptomatically; - IRRs: it was clarified that subjects who experienced a Grade 4 (life-threatening) event during an infusion were to have their infusion stopped, and that additional infusions should not be given. (7) It was clarified that prolonged peripheral blood B cell depletion was an expected outcome of rituximab treatment and was not to be considered a toxicity. (8) Instructions regarding the initial dose of prednisone or equivalent in the range of 1.0 to 1.5 mg/kg/day was updated to 60 to 120 mg/day from 80 to 120 mg/day.

(1) The PDAI in Appendix 4 was replaced with a corrected version. In the previous version of the protocol (Version 2), the PDAI was reproduced with permission from Clinics of Dermatology (2012); however, in 2015, it was discovered that Clinics of Dermatology had published an erroneous version of the PDAI that omitted the column to evaluate the number of lesions if </=3 for disease activity. The erroneous version of the PDAI was replaced with the validated version, and the protocol was updated to reflect the correct information, including instructions on the scoring method. (2) During an iDMC organizational meeting on 7 Feb 2015, the iDMC advised the Sponsor to consider that clinical definitions of treatment failure may not be the same as the statistical definition of treatment failure (i.e., non-responder). Relevant section of the protocol was therefore amended, per the iDMC recommendation, to align the clinical definitions of treatment failure with “non-responder” classification for the primary analysis. To achieve this, one definition was amended and two definitions were deleted to avoid treatment bias and premature rescue. (3) The non-responder rule was updated to clarify the scenarios in which subjects will be imputed as non-responders. The previous wording referred to the non-responder categorization being applied at the time of withdrawal or treatment failure. This wording was potentially misleading given that the primary endpoint required a 16-week consecutive duration at any time during the 52-week treatment period. Therefore, additional wording was added for clarification purposes and did not deviate from the original intent.

(1) A typographical error stating that urine and serum pregnancy tests were being performed at the sensitivity of 50 mIU/mL was corrected, as pregnancy tests were being performed at the recommended sensitivity of 25 mIU/mL. (2) It was clarified that female subjects who were not postmenopausal must have had two negative pregnancy test results prior to starting study treatment, with a specific time interval between the two pregnancy tests. The timing for the serum pregnancy test during the screening period was changed to clarify that the screening pregnancy test must have been performed 8 to 10 days before the baseline/Day 1 urine pregnancy test (i.e., between Day -8 and Day -10) instead of at Day -1 to Day -28. (3) Contraception requirements were updated to include male patients who had undergone a vasectomy and female partners of male subjects who received MMF. (4) Wording was included that male subjects should advise their female partners to use a highly effective method of contraception during the study and for 12 months after stopping treatment. (5) Footnote “o” in the schedule of assessments was changed to address a typographical error. In the previous version of the protocol (Version 3), the footnote specified that certain blood chemistry values were to be collected at Weeks 1, 15, 24, and 26; however, Week 15 was incorrect, and the timing was changed to Weeks 1, 2, 24, and 26.

(6) It was clarified that, before Week 12, patients who experienced treatment failure required an early withdrawal visit and were to be followed in the SFU period to receive standard-of-care treatment. (7) The inclusion criterion regarding first confirmed diagnosis of PV within the previous 24 months based on histological features of acantholysis via skin or mucosal biopsy was revised to include additional acceptable confirmatory tests for PV diagnosis besides tissue-bound IgG antibodies by direct immunofluorescence on the surface of affected epithelium. The diagnosis of PV was based on histological features of acantholysis via skin or mucosal biopsy and one of the following: tissue-bound IgG antibodies by direct immunofluorescence on the surface of affected epithelium or serological detection of serum Dsg3 autoantibodies against epithelial cell surface either by indirect immunofluorescence microscopy or by enzyme-linked immunosorbent assay. (8) It was clarified that the iDMC would review the safety data of TM subjects.

The protocol was amended to address FDA’s recommendation that data obtained from TM subjects was considered to be exploratory in nature and therefore the primary analysis for establishing efficacy was to be based on the ITT population for subjects who were not recruited via TM. The following changes were made to the protocol to address this recommendation, as follows: (1) An additional minimum of 8 non-TM subjects was to be recruited into this study to maintain sufficient statistical power for the primary analysis, thereby increasing the overall enrollment from approximately 124 to approximately 132 subjects with PV. (2) The Statistical Considerations section was updated, stipulating that all efficacy outcomes would be analyzed using the mITT population and excluding subjects who were enrolled via TM. (3) he primary efficacy outcome measure was revised to reflect the exclusion of TM subjects. (4) A new primary analysis population was added, the mITT population, which included subjects in the ITT population, excluding the 10 subjects enrolled via TM. This population was to be used in the analyses of efficacy outcomes. (5) The analysis population to be used for the secondary efficacy endpoints was changed to the mITT population. (6) Exploratory analyses were updated to include a statement on descriptive statistics for evaluation of the 10 subjects recruited via TM. (7) In addition, the following change has been made: An exploratory objective and exploratory outcome measure to evaluate the proportion of subjects experiencing treatment failure in each treatment arm were added.