CENTER FOR DRUG EVALUATION AND
`
`
`
` RESEARCH
`
`
`
`
` APPLICATION NUMBER:
`
`
`
` 208276Orig1s000
`
`
`
`
` STATISTICAL REVIEW(S)
`
`
`
`
`
`
`
`
`

`

` U.S. Department of Health and Human Services
`
`Food and Drug Administration
`Center for Drug Evaluation and Research
`Office of Translational Sciences
`
`Office of Biostatistics
`
`S T A T I S T I C A L R E V I E W A N D E VA L U A T I O N
`
`CLINICAL STUDIES
`
`
`NDA: Supplement:
`Drug Name:
`Indication(s):
`Applicant:
`Date(s):
`Review Priority:
`
`208276
`Remodulin Implantable System (RIS)
`Pulmonary Arterial Hypertension
`United Therapeutics Corporation
`December 16, 2015
`Standard (Safety Review)
`
`DBI/OB/OTS/CDER
`Biometrics Division:
`Fanhui Kong, PhD
`Statistical Reviewer:
`Concurring Reviewers: Hsien Ming Hung, PhD
`
`Medical Division:
`Clinical Team:
`CDTL
`Project Manager:
`
`Division of Cardiovascular and Renal Products
`Patricia Beaston, MD
`Shari Targum, MD
`Wayne Amchin,
`
`Keywords: pulmonary arterial hypertension (PAH), Remodulin (treprostinil) injection,
`Remodulin implantable system (RIS), implantable intravascular catheter, safety profile,
`
`catheter-related complications, single arm study.
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

`Table of Contents
`
`
`1
` EXECUTIVE SUMMARY .................................................................................................................................3
`
`
`
`2
`
`
`
`3
`
`
`
`4
`
`
`5
`
`
`INTRODUCTION ...............................................................................................................................................4
`
`
`
`OVERVIEW......................................................................................................................................................4
`
`2.1
`
`
`DATA SOURCES ..............................................................................................................................................5
`
`2.2
`
`
`STATISTICAL EVALUATION ........................................................................................................................5
`
`
`
`DATA AND ANALYSIS QUALITY......................................................................................................................5
`
`3.1
`
`
`
`EVALUATION OF EFFICACY.............................................................................................................................6
`
`3.2
`
`
`
`EVALUATION OF SAFETY ................................................................................................................................7
`
`3.3
`
`
`
`3.3.1
`Study Design and Endpoints ..................................................................................................................7
`
`
`
`
`3.3.2
`Statistical Methodologies.......................................................................................................................9
`
`
`
`3.3.3
`Patient Disposition, Demographic and Baseline Characteristics........................................................10
`
`
`
`3.3.4
`Safety Results and Conclusions ...........................................................................................................11
`
`
`
`FINDINGS IN SPECIAL/SUBGROUP POPULATIONS .............................................................................15
`
`
`
`SUMMARY AND CONCLUSIONS ................................................................................................................15
`
`
`
`STATISTICAL ISSUES .....................................................................................................................................15
`
`5.1
`
`
`
`COLLECTIVE EVIDENCE ................................................................................................................................15
`
`5.2
`
`
`
`CONCLUSIONS AND RECOMMENDATIONS .....................................................................................................16
`
`5.3
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`2
`
`

`

`1 EXECUTIVE SUMMARY
`
`In this NDA the Sponsor submitted a pivotal study DelIVery to assess the safety profile of the
`Model 10642 Implantable Intravascular Catheter for the delivery of Remodulin® (treprostinil)
`injection in the treatment of patients with pulmonary arterial hypertension (PAH) who met the
`
`
`approved Remodulin Injection indication, using the approved formulation, and approved
`intravenous route of administration.
`
`The DelIVery for PAH study is a multicenter, prospective, single arm, non-randomized open
`
`label investigational clinical trial. The purpose of the trial is to evaluate the safety profile of the
`
`
`Model 10642 Implantable Intravascular Catheter, a component of the PAH Implantable
`Vasodilator Therapy (PIVoT) system, to deliver Remodulin Injection, for the treatment of PAH.
`The study population will include patients currently treated with the approved intravenous (IV)
`
`infusion route of delivery of Remodulin Injection for PAH.
`
`The study was conducted at up to 10 sites in the U.S. A total of 64 subjects were enrolled and 60
`were implanted of which 9 died during study participation. Given PAH is a severe and fatal
`disease, the serious AE rate and survival rate were consistent to those of the general PAH patient
`population. A total of 7 primary endpoint complications were observed during the accumulation
`
`of 71,920 patient days, resulting in a total of 0.10 catheter-related complications per 1000 patient
`days and the one-sided upper 97.5% confidence bound of 0.20 which is below 2.5 per 1000
`
`patient days, the rate identified from the external Central Venous Catheter literature. This result
`
`
`seems to provide adequate evidence to support the safety of the Model 10642 investigational
`Implantable Intravascular Catheter for the delivery of Remodulin injection for the treatment of
`
`
`
`the patients with PAH.
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`3
`
`

`

`INTRODUCTION
`
`Pulmonary arterial hypertension is an incurable disease with median survival approaching 7
`
` years. Although prostacyclins are accepted as one of the most efficacious regimens, there is
`underutilization of the therapy. Current options allow patients continuous parenteral prostanoid
`therapy via an external infusion pump, either with an indwelling central venous catheter or by
`
`
`
`subcutaneous injection. This method creates fear and angst for some PAH patients. In addition,
`
`indwelling central venous catheters are associated with the risk of blood stream infections and
`sepsis, which can be fatal. Patients receiving subcutaneous injections often experience infusion
`site pain.
`
`Medtronic, Inc. sponsored a multicenter, prospective, single arm, non-randomized open label
`investigational clinical trial, named “DelIVery” for PAH. The purpose of this clinical trial was to
`
`evaluate the safety profile of the Model 10642 Implantable Intravascular Catheter, a component
`of the PAH Implantable Vasodilator Therapy (PIVoT) system. The PIVoT system also includes a
`market approved implantable drug delivery pump with programmer (SynchroMed II), and a
`sutureless connector to connect the investigational catheter to the SynchroMedII pump. This
`fully implanted system was used to deliver the currently marketed pharmaceutical product
`treprostinil (Remodulin). Because Remodulin therapy is the only parenteral with a 4-hour half-
`life and proven long-term stability, it is the only prostacyclin candidate for the PIVoT system.
`
`The sponsor tested the hypothesis that the investigational implantable intravascular catheter is
`safe when used in the PIVoT system to deliver treprostinil by demonstrating that the incidence
`
`rate of catheter-related complications is less than catheter-related complications using external
`
`
`
`systems. This drug-device combination was developed to change the current prostacyclin
`treatment by removing the need to prepare medication multiple times a week and wear an
`
`external pump. It minimizes the risk of bloodstream infections and generates impactful
`improvements in PAH patients’ lives.
`
`1.1 Overview
`
`This study focuses on the safety of delivery of Remodulin Injection in the treatment of patients
`with PAH who meet the approved Remodulin Injection indication, using the approved
`formulation, and approved intravenous route of administration. The data generated by the study
`
`are intended to provide adequate safety information necessary to support the FDA approval of a
`marketing application for the Model 10642 catheter and the labeling updates for the SynchroMed
`II system, as well as an NDA-supplement from United Therapeutics to support the updates to
`
`Remodulin Injection labeling.
`
`This open-label, uncontrolled study was designed to evaluate the safety of Model 10642
`Implantable Intravascular Catheter when used with the Medtronic SynchroMed II Implantable
`
`Infusion System to deliver Remodulin compared with historical literature. For ancillary
`
`
`endpoints, data collected during the follow-up visits were analyzed by comparing with the
`
`subjects’ Baseline values (receiving Remodulin via an external infusion pump).
`
`4
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

` The original protocol was developed on October 14, 2010. The study design was conditionally
`
`
` approved by FDA Center for Devices and Radiological Health (CDRH) under Investigational
` Device Exemption (IDE) G100017 for the DelIVery for PAH clinical protocol Version 4 on
`
`January 21, 2011, limiting the study to five subjects and one U.S. institution. The protocol was
`amended once to allow enrollment of up to 70 subjects at 10 sites. Additionally, six IDE-
`
`supplements resulted in changes to the conduct of the study.
`
`This NDA was first submitted on January 26, 2015 and the submission was refused to file on
`February 24, 2015 because the application was not sufficiently complete to permit a substantive
`review.
`
`The revised NDA was submitted on December 16, 2015 which included additional analyses of
`the clinical study data set presented in the DelIVery for PAH, premarket application (PMA)
`
`Clinical Report for data cutoff point of July 25, 2014. On April 15, 2016, a revision was
`submitted with the analysis results being updated based on additional data collected by the new
`data cutoff date of January 08, 2016. This date will be referred to as the data cutoff date in this
`
`review. New data sets and programs were provided. This review is conducted based on the
`newest submission.
`
`1.2 Data Sources
`
`The sponsor’s electronic data sources were stored in the directories
`of \\CDSESUB1\evsprod\NDA208276\0007\ and \\CDSESUB1\evsprod\NDA208276\0010\ of
`the Center’s electronic document room of the Agency. Data sources include all raw data sets in
`
`SDTM format, analysis data sets in ADAM format, individual data listings, SAS programs for
`deriving the data sets and analysis results, the study reports, protocol amendments, statistical
`
`analysis plan, literature referenced, etc.
`
`
`
`2 STATISTICAL EVALUATION
`
`2.1 Data and Analysis Quality
`
`
`The sponsor provides high quality data sets along with the programs to produce the analysis data
`sets and safety results which allow the statistical reviewer to confirm the results.
`
`Data collected include demographics, medical history, medications, adverse events, exit/death
`information, deviations, device malfunctions, serum treprostinil levels, fluoroscopic/radiograph
`images of catheter location, pump programming and fill data, New York Heart Association
`(NYHA) classification, physical assessment, six minute walk test, quality of life (QoL)
`questionnaires, and subject and caregiver involvement assessment.
`
`5
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

`According to Clinical Study Report (CSR), data were collected using an electronic data
`
`
`management system for clinical trials. Electronic case report form (eCRF) data were stored in a
`
`secure, password-protected database that was backed up nightly. Access to the database was
`restricted to trained and delegated personnel. Data were reviewed using programmed and manual
`data checks. Data queries were made available to sites for resolution. Study management reports
`
`were generated to monitor data quality and study progress.
`
`Investigators were required to ensure accuracy, completeness, and timeliness of the data reported
`to Medtronic. Electronic case report forms were maintained and signed electronically within the
`
`electronic data capture system during the trial by authorized study site personnel. The data for
`
`this report were stored as a frozen data set and will be retained indefinitely. At the end of the
`study, the data will be frozen and will be retained indefinitely by Medtronic.
`
`2.2 Evaluation of Efficacy
`
`The primary objective of this study was to determine the safety of the investigational catheter.
`
`An ancillary objective of the study was to determine the percent change of six-minute walk
`
`
`distance from baseline to six weeks post-implant and to characterize changes in quality of life
`
`(QoL). There was no primary efficacy endpoint in this study.
`
`The percent change in six-minute walk distance from baseline to 6 weeks post-implant in meters
`was calculated. The mean percentage increase in six-minute walk test values in the first six
`
`weeks post implant was 0.2% (95% confidence interval: -4.9% to 5.2%).
`
`The CAMPHOR (Cambridge Pulmonary Hypertension Outcome Review) QoL questionnaire
`was given at baseline, 6-week, three-month, six-month and twelve-month follow-up visits.
`The CAMPHOR has three scales: symptoms, activities, and QoL, which are calculated
`
`separately. The changes in all three scale scores from baseline to six months post-implant were
`calculated for each subject. For categorization purposes, a decrease in the score of at least two
`points indicates better quality of life. An increase of at least two points indicates worse quality of
`life.
`
`Mean changes of the scores from CAMPHOR between baselines and six month follow-ups with
`95% confidence intervals as well as percentages of the subjects falling into each of three levels
`are presented (better, no change, worse) (Table 3.1). Data from all successfully implanted
`subjects with CAMPHOR scores at baseline and six months are included in this analysis.
`Changes in the three scores were minimal, with the average scores being within two of baseline,
`
`
`and many subjects scoring no change.
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`6
`
`

`

`Table 3.1 CAMPHOR Baseline to Month 6 Results
`of the Implanted Subjects
`
`Symptom Scale Activity Scale
`
`QoL Scale
`
`-0.37 ± 3.51
`(-1.32, 0.54)
`
`1.03 ± 3.10
`(0.20, 1.85)
`
`Mean Score Change ± SD
`95% Confidence Interval
`Change
`18 (32%)
`10 (18%)
`18 (32%)
`Better
`29 (51%)
`26 (46%)
`25 (44%)
`No Change
`10 (18%)
`21 (37 %)
`14 (25%)
`Worse
`
` Source: Table 11-3 in Summary of Drug-Device Safety Update: IDE G100017.
`
`-0.51 ± 3.07
`(-1.32, 0.30)
`
` These results are expected since the treatment therapy being provided in this study was the same
`
`
` therapy received at baseline; only the drug delivery method of using an internal versus external
`drug delivery system was changed.
`
`2.3 Evaluation of Safety
`
`2.3.1 Study Design and Endpoints
`
` It’s not clear when the original Clinical Investigational Plan (CIP) was developed. The final
`
`
`version of the CIP (Version 5) was finished on February 2, 2011. In this CIP, the study design,
`primary objective, analysis endpoints and the detailed statistical analysis methods were
`developed. The first enrollment was on June 14, 2011.
`
`Study Design
`
` Study G100017 is an open-label, uncontrolled study designed to evaluate the safety of Model
`
`
` 10642 Implantable Intravascular Catheter when used with the Medtronic SynchroMed II
` Implantable Infusion System to deliver Remodulin compared with historical literature in the
`
`
`
` PAH patient population. Prior to enrollment, subjects were required to have been receiving IV
`
`Remodulin, and to have had a stable dose (no change in dose) for at least four weeks.
`Additionally, subjects were not to have been initiated on a new oral PAH therapy within two
`months prior to enrollment.
`
`Implant of the system occurred within two weeks of the baseline visit. A successful implant
`
`included implant of the SynchroMed II Implantable Infusion System and the Implantable
`
`Intravascular Catheter, and completion of the prime bolus procedure. After the prime bolus
`
`procedure was completed, the pump delivered drug at the subject’s prescribed therapeutic rate,
`and the subject was transitioned off the external delivery system.
`
`7
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

` Implanted subjects were seen at scheduled follow-up visits; one week, six weeks, three months,
`
`
`
` six months, twelve months, and then every six months thereafter. Subjects were also seen at
` unscheduled follow-up visits as needed, primarily for pump refills, Remodulin drug dose change,
`
`
`and adverse events (AEs). Clinical data were also collected at system modifications, upon
`notification of adverse events or device malfunctions, deviations, exits, and in cases of death or
`pregnancy. Subjects will continue to be followed until marketing approval is granted from the
`
`US FDA or official study closure, whichever occurs first.
`
`Rationale for Study Design
`
`Remodulin Injection is a well-established and effective therapy for PAH. It is currently delivered
`
`
`
`subcutaneously, and intravenously (IV). Since the new delivery system delivers Remodulin
`Injection to the same place, effectiveness does not need to be tested. The new component in the
`system is the Model 10642 Implantable Intravascular Catheter. The remainder of the system,
`including the SynchroMed II Implantable Infusion System has been used in the market for many
`
`
`years. A single arm study was chosen because subjects could not be blinded and the primary
`
`safety objective is to assess the safety of the catheter, for which there is no reasonable control
`
`group.
`
`Sample Size
`
`A total of 64 subjects were enrolled (consented) in the study at 10 US sites, and 60 subjects were
`successfully implanted with the system.
`
`A total of 22,000 patient days of follow-up were estimated using a simulation program with the
`assumption of the rate of catheter-related complications followed a Poisson distribution with a
`
`
`mean rate 1.5 events per 1000 patient days to reach a power of 90% to detect the safety of the
`investigational catheter if the exact upper 97.5% confidence bound was below 2.5 per 1000
`patient days. Implanted subjects were expected to have an average of 440 days of follow-up. A
`
`total of 50 implanted subjects were adequate to achieve 22,000 patient days. The maximum
`enrollment sample size was 70 subjects. Analyses were performed when subjects completed six
`
`months of follow-up and 22,000 patient days were accumulated.
`
`The study was expected to enroll up to 70 subjects to ensure at least 50 subjects to be
`successfully implanted.
`
`Primary Endpoints:
`
`The primary endpoint was related to the safety of the investigational catheter. The primary
`
`
`endpoint is catheter-related complications per 1000 patient days. A complication was defined as
`any adverse event that resulted in death, involved any termination of significant device function,
`
`or required an invasive intervention. Catheter-related complications included catheter-related
`systemic bloodstream infections, site infections, and complications from catheter thrombosis,
`mechanical dysfunction, and catheter dislocation. Procedure-related pneumothorax
`
`complications were also counted toward the primary endpoint.
`
`8
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

`According to the sponsor, the Adverse Event Advisory Committee (AEAC) adjudicated all the
`adverse events according to prespecified rules as catheter-related complications or just
`observations. If an event was classified as a complication, even its relatedness to the catheter was
`
`
`unknown, it was still counted as a catheter-related complication. Additionally, pneumothorax
`
`was adjudicated by the AEAC as procedure-related, and classified as a complication, was
`counted toward the primary endpoint. The number of patient days contributed by each subject
`
`
`was the latest known date of follow-up. The number of patient days for subjects who had the
`system explanted was the removal date minus the implant date.
`
`2.3.2 Statistical Methodologies
`
`According to sponsor, the statistical analysis plan (SAP) was developed based on the revised
`study protocol on January 26, 2012, before data were analyzed. According to SAP, data from all
`
`
`the centers that participated in this protocol were combined for analysis. Once all active subjects
`had completed the six month follow-up and there were at least 22,000 days of follow-up among
`
`implanted subjects, a PMA report would be prepared.
`
`To demonstrate that the Model 10642 Implantable Intravascular Catheter was safe when used
`with the Medtronic SynchroMed II Implantable Infusion System to deliver Remodulin, the
`
`following hypotheses were developed.
`
`Hypothesis:
`x HO: Rate of catheter-related complications • 2.5 per 1000 patient days
`x HA: Rate of catheter-related complications < 2.5 per 1000 patient days
`
`The Model 10642 Implantable Intravascular Catheter is considered to be safe if the one-sided
`
`upper 97.5% confidence bound of the rate of catheter-related complications among all the
`
`implanted subjects is less than 2.5 per 1000 patient days. A one-sample exact test for the Poisson
`rate was used to obtain the 97.5% one-sided upper confidence bound of the catheter-related
`complications.
`
`The catheter-related complication rate of 2.5 per 1000 patient days in the PAH population was a
`rough estimate from a series of published studies. The estimated rate of central venous catheter
`(CVC) systemic infections for bloodstream infections goes from 0.43 (Dicknson et al., 2009) to
`1.13 (Kallen, et al., 2008) per 1000 patient days. The estimated site infection rate is from 0.26
`(Oudiz, et al., 2004) to 0.87 (Akagi, et al., 2007) per 1000 patient days. Finally the combined
`estimated rate of complications from catheter thrombosis, mechanical dysfunction, and catheter
`dislocation in the general CVC population is from 0.36 (Smith, et al., 1998; Bozzetti, et al.
`2002) to 0.51 (Moureau, et al., 2002) per 1000 patient days Adding the three higher estimated
`rates together gives a roughly estimated total rate of catheter-related complication of up to 2.5
`per 1000 personal days.
`
`9
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

`2.3.3 Patient Disposition, Demographic and Baseline Characteristics
`
`Disposition
`
`A total of 64 subjects were enrolled (consented) in the study at 10 US sites, and 60 subjects were
`successfully implanted with the system. Four subjects exited the study prior to implant (two
`developed external line infections, one clinically worsened, and one was too small to accept the
`
`implantable pump bulk and weight). Nine subjects had died and one subject had discontinued
`therapy due to AEs.
`
`Figure 3.1 Subject Disposition
`
`Source: Figure 10-1 in Summary of Drug-Device Safety Update: IDE G100017
`
`Demographic and Other Baseline Characteristics
`
`
`The demographics of the subjects in the DelIVery for PAH clinical study enrolled into this
`clinical study reflected those of the overall PAH patient population. Of the 60 implanted
`subjects, 80% were female. The mean/median age was 50.1/52.0 years with the range from 24 to
`
`74. Additionally, 13 (22%) of the 60 implanted subjects were Asian (3%), Hispanic (13%) or
`African American (5%), and 47 (78%) were Caucasian.
`
`10
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

`Table 3.2 Key Demographic/Baseline Characteristics
`of the Implanted Subjects
`
`Gender
`
`Age
`
`Race
`
`Demographic/Baseline
`Characteristic
`Female
`Male
`Mean ± SD
`Median
`25th – 75th Percentile
`Minimum - Maximum
`Asian
`African American
`Hispanic
`Caucasian
`
`Implanted Subjects
`
`48 (80%)
`12 (20%)
`50.1 ± 13.5
`52.0
`38 - 61
`24 - 74
`2 (3%)
`3 (5%)
`8 (13%)
`47 (78%)
`
`
`
` Source: Table 11-1 in Summary of Drug-Device Safety Update: IDE G100017
`
`Protocol Violations/Deviations
`
` Overall, there were 141 study deviations among the 64 enrolled subjects. The majority of the
`
`deviations (107) were due to one or more assessments not completed or not completed per
`protocol.
`
`
`
`
`
`As of the data cutoff date, there were 20 instances in 6 subjects in which the refill interval
`
`exceeded 12 weeks, the limit specified by the study procedures. In addition, three individual site
`
`personnel participated in the study prior to being approved; one subject did not complete a
`
`pregnancy test at Baseline; one subject was enrolled into this study within 30 days of exiting a
`
`drug post-market surveillance study; and one initially missed a check box on the consent form
`but completed it at a later date. No protocol deviations (e.g., longer than 12 week refill interval,
`or implant not per protocol) directly impacted the study results including the frequency of overall
`AEs and 6MWD results.
`
`2.3.4 Safety Results and Conclusions
`
`
`According to the sponsor, all new and/or worsening AEs experienced by the subjects, including
`deaths, were collected throughout the study duration starting at subject enrollment and reported
`
`to the Sponsor on an AE electronic Case Report Form (eCRF). Documented pre-existing
`
`conditions were not considered AEs unless there was a change in the nature or increase in
`severity of the condition.
`
`11
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

`As of the data cutoff date for this report, the 60 implanted subjects averaged 3.3 years of follow
`up for a total of 197.0 patient-years. Fifty-one subjects completed their three year visit. The
`range of individual ongoing subject participation was 87 to 1,625 days. The mean Remodulin
`dose at Baseline was 72.2 ± 29.5 ng/kg/min with a range of 22 to 160 ng/kg/min. At the most
`recent study visit prior to this data cutoff, the mean Remodulin dose was 79.8 ± 26.2 ng/kg/min
`with a range of 26.9 to 142.0 ng/kg/min.
`
`Brief Summary of Adverse Events
`
`At the time of the data cutoff for this report, 23 pre-implant AEs were reported in 17 subjects and
`
`1,030 treatment-emergent (including 2 during implant and 1,028 post-implant) AEs had been
`
`reported in 60 subjects. Twenty AEs (in 17 subjects) were considered “unavoidable” that
`
`included AEs related to the implant procedure (such as anesthesia-related nausea/vomiting
`
`within the first 24 hours and pocket site/incisional pain that had an onset and resolution within 72
`
`hours of the implant procedure).
`
`The most common post-implant AEs (including unavoidable AEs) were implant site pain, upper
`respiratory tract infection, worsening PAH, dyspnea, injection site reaction, headache, nausea,
`hypotension, fatigue, injection site pain, and dizziness. Of the1,028 post-implant AEs, 294 in 59
`subjects were considered related to Remodulin, one or more system components and/or a study
`
`procedure.
`
`A total of 146 post-implant SAEs occurred in 45 of the 60 implanted subjects. Twenty-three post
`implant SAEs in 16 subjects were considered related to Remodulin or one or more system
`components or study procedures. Nine subjects died during this analysis period resulting in a
`survival probability of 95% at one year after enrollment into the study and 90% two years after
`
`
`
`enrollment. None of these deaths were adjudicated to be related to Remodulin, a study
`
`
`procedure, or any component of the system. One subject exited the study after the system was
`explanted due to AEs, and one subject had Remodulin replaced with saline in the pump in
`
`preparation for lung transplant.
`
`Primary Analysis of Primary Safety Endpoint
`
`As of the data cutoff date, six catheter-related complications (all requiring invasive system
`
`
`modifications) and one procedure-related pneumothorax complication had occurred in this study.
`One subject had three of the six catheter-related complications and ultimately had the implanted
`
`system explanted as a result of sepsis, pump pocket infection and catheter incision site infection
`after her third system modification.
`
`According to the pre-defined performance goal, the safety objective on the primary endpoint
`
`would be met if the one-sided upper 97.5% confidence bound of the rate of catheter-related
`complications was less than 2.5 per 1000 patient days. A total of 7 primary endpoint
`
`complications were observed during 71,920 patient days, resulting in a total of 0.10 catheter-
`related complications per 1,000 patient days (one sided upper 97.5% confidence bound of 0.20).
`12
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

`These 7 complications included 3 catheter dislocations, 2 punctured catheters, one pneumothorax
`
`and one venous stenosis.
`
`Statistical Comments:
`
`1. The statistical method is considered to be acceptable given the study design. However,
`considering that this was an open label study and there was no control arm, the
`reliability of the statistical result could be a concern.
`2. Due to the change of the data cutoff date, the efficacy results as well as safety results
`need to be updated although this does not change the safety conclusion of the study.
`
`Adverse Events other than the Primary Endpoints
`
`
`The following adverse events results are extracted from the Summary of Drug-Device Safety
`
`
`report submitted by the sponsor.
`
`Of the 1,028 post-implant AEs that occurred during the study, 294 events were considered
`related to Remodulin. As of the data cutoff date, 1,798 pump refills had been conducted with
`subjects having an average of 45 days between refills. Eighty-five AEs related to the refill
`
`
`process occurred in 34 subjects.
`
`A total of 164 post-implant AEs (including unavoidable AEs) were considered to be related to
`
`the implant procedure or a system modification. A system modification was required when the
`implanted catheter and / or pump required invasive modification. Of the 164 AEs related to
`implant or system modification, 22 events in 13 subjects were also considered to be related to
`Remodulin.
`
`Fourteen AEs in 11 subjects were related to the SynchroMed II pump. The events of implant site
`
`extravasation (2 events), device dislocation (flipped pump) and muscle spasms were also related
`to the implant procedure. Five AEs in three subjects (including dyspnea, renal failure, deep vein
`thrombosis, and two cases of thrombophlebitis superficial) were considered to be related to the
`implant procedure and potentially associated with a thromboembolic event.
`
`
`
`A total of 180 infection AEs occurred in 54 (90%) subjects during the study. These include all
`AEs under the MedDRA SOC of Infections and Infestations, plus all reports of general fevers,
`post operation fevers, phlebitis, diarrhea, and other events treated with antibiotics or associated
`
`
`with positive cultures noted in the diagnostic testing. Of the 180 infections, 176 in 52 (87%)
`subjects occurred post implant and 4 occurred prior to the implant. Eight infection-related AEs in
`5 subjects were considered related to the original implant procedure or a system modification.
`The remaining 172 were considered unrelated to any other study procedure, a system component,
`
`or Remodulin.
`
`The implant process has the potential to increase the risk of thrombotic / embolic events. A total
`of 25 AEs in 18 (30.0%) subjects fit into this category. Of these, five events in three (5.0%)
`
`
`13
`
`Reference ID: 3955858
`Reference ID: 4409052
`
`

`

`subjects were adjudicated as being related to the implant procedure. The remaining 20 events in
`
`this category were adjudicated to be unrelated to any other study procedure, a system component,
`
`or Remodulin.
`
`To examine the potential impact of the implant process on arrhythmias, an analysis of
`
`arrhythmias present both before and after implant was conducted. Twenty-two subjects (36.7%)
`had a history of at least one reported arrhythmia at baseline. Twelve subjects (20%) had a history
`
`of atrial arrhythmia, two (3.3%) had a history of ventricular arrhythmia, and four (6.7%) had a
`history of AV junctional arrhythmia and blocks. Additionally, one subject (1.7%) had a history
`
`of both atrial and ventricular arrhythmia, two (3.3%) had a history of atrial arrhythmia and AV
`
`junctional arrhythmia and blocks, and one (1.7%) had a history of all three classes of arrhythmias
`at Baseline.
`
`There were 7 system mo