`
`QEN lER EQR DRUG EYALIJATIQN AND RESEARCH
`
`APELIQA! IQN NUMBER: NDA 20845
`
`ED AL
`
`EW
`
`
`
`

`

`
`
`ocr 291999
`
`Primary Medical Review
`of Inhaled Nitric Oxide (I-NO)
`
`NDA 20-845
`
`Food and Drug Administration
`Division of Cardio-Renal Drug Products (HFD--110)
`
`October 29, 1999
`
`By
`Douglas C. Throckmorton, M.D.
`
`
`
`

`

`.--.———---—- ,.,. _.._...,__. M-z-#flu__—.M—M‘
`
`Table of Contents
`
`
`
`
`
`.
`
`
`
`
`
`
`
`
`0.0 Overall Summary
`1.0 Materials Used for Review
`
`
`
`2.0 Background
`
`.0 Review ofCINRGI
`
`3.] Title ofStudy
`'
`
`
`3.2 Sites of Investigation and InvestigatordFinancial Disclosure
`3.3 Background
`‘
`
`3.4 Study Design
`
`3.5 Primary And Secondary ObjectiveIEndpoints
`
`3.6 Number Of Subjects! Randomization
`
`3.7 Inclusion! Exclusion Criteria
`
`3.8 Dosage! Administration?
`
`3.9 Duration! Adjustment of Therapy
`3.10 Safety and Efficacy Endpoint Measured
`
`
`3.11 Statistical Considerations
`
`3.12 Efficacy Outcomes
`
`
`3.12.1 Patient Demographics and Baseline Characteristics
`
`3.12.2 Disposition and Follow-up for Subjects
`
`
`3.12.2a Subject Selection .
`
`3.12.2b Protocol Violationsand Deviations
`
`
`3.12.2: Patient Randomization & Completion
`
`
`3.12.2c Concomitant Therapies Used After Trial Initiation
`
`
`3.12.3 Analysis of Primary Endpoint
`3.12.4 Analysis ofSecondary Endpoints from CINRGI
`
`
`3.12.5 Additional Analyses of CINRGI
`
`3.13 Safety Outcomes
`
`
`3.13.1 Deaths in CINRGI
`3.13.2 Serious Adverse Events in CINRGI‘
`
`
`3.13.3 Adverse Events in CINRGI
`._-
`3.13.4 Discontinuations in CINRGI
`
`
`3.13.5 Lab Adverse Events and Special Studies
`
`
`3.14 Efficacy Summary For CINRGI
`
`
`
`3.15 Safety Summary For CINRGI
`
`3.l6 Overall Summary for CINRGI
`
`4.0 Long-Term Follow-Up for [NO-01! -02 and NINOS
`4.1 NINOS Follow-Up
`
`4.1.] Additional NINOS Analyses: Receipt of ECMO
`4.1.2 NINOS Follow-Up Results
`4.2 [NO-01! -02 Follow-Up
`
`5.0 Integrated Efficacy Summary from NDA Trials
`5.] Demographics
`5.1.1 Population Demographics
`5.1.2 Clinical Demographics
`
`
`5.2 Extent of exposure (dosefduration)
`
`
`5.3 Primary and secondary efficacy endpoints
`
`
`5.4 Success of trials in meeting pro-specified primary endpoints
`
`
`5.4.1 Sub-Group Analyses of the Primary Endpoint
`
`
`5.5 Success of trials in meeting secondary efficacy endpoints:
`
`
`demonstrating a physiological effect of LNG.
`5.6 Success of trials in meeting secondary efficacy endpoints:
`
`
`demonstrating a clinical benefit for I-NO
`
`
`5.7 Clinical effect of I-NO from the secondary data sources
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`NBA 20-845
`Primaiy Medical Review
`10.99
`'
`
`
`2
`
`

`

`
`
`,_ ,_ __.
`
`..—>—- ,
`
`,,,
`
`.__...... .... .————..—--——I-——_____‘__w___
`
`Table of Contents (cont)
`
`
`
`
`
`6.0 Integrated Safety 'Summary from NDA Trials
`
`6.1. General Comments about Adverse Event collection
`
`6.2 Acute pulmonary injury
`
`
`6.3 Chronic Pulmonary Injury
`
`
`6.6. Acute Neurological Injury
`
`
`6.5 Chronic Neurological Injury
`
`
`6.6 Laboratory Abnormalities
`
`
`Increased Methemoglobin and NO; concentrations
`
`
`Eosinophilia
`-
`
`
`Abnormal LFI‘s
`6.7 Other Laboratory Measurements
`
`
`6.8. Miscellaneous Adverse Events
`7.0 Overall Summary of Efficacy and Safety for LNG
`
`
`
`8.0 References
`Appendices
`
`
`9.0 Appendix One: Abbreviations Used
`10.0 Appendix Two: Death Narratives from the CINRGI Trial
`
`
`11.0 Appendix Three: Patient Disposition in the CINRGI Trial
`12.0 Appendix Four: Proposed Label
`
`
`NDA: 20-845“
`Inhaled Nitric Oxide (l-NO)
`NAME or DRUG:
`TRADE NAME:
`[NOmax
`FORMULATION: Gas for inhalation
`RELATED APPLICATION: None
`
`‘
`
`’
`
`DATE OF SUBMISSION: 5.25.99
`_
`DATE RECEIVED BY FDA: 5.26.99
`DATE ASSIGNED T0 CURRENT REVIEWER: 6.17.97(see below)
`DATE REVIEW COMPLETED: 10.29.99
`
`PROPOSED INDICA'I‘ION: Treatment of hypoxic respiratory failure in newborns
`
`SPONSOR/MONITORS: 1N0 Therapeutics, Inc.
`
`A. a I
`
`Douglas C. Throckmorlon, MD.
`Primary Medical Reviewer
`
`NDA 20-845
`Primary Medical Review
`10.99
`-
`
`
`-
`
`-
`
`3
`
`

`

`_________ __ _ ,_,__,__n __
`
`__ W
`
`‘
`-
`____,m.___._._—__-__.___._____
`
`———m—H——H-——wbuuua—
`‘
`
`.
`0.0 Overall Summary of Efficacy and Safety for I-NO
`Inhaled Nitric Oxide (l-NO) has been proposed as a treatment for hypoxia respiratory failure in neonates.
`Clinical support for this indication comes from four clinical trials conducted in this population and submitted as part
`of this NDA, as well as an extensive published literature on the use of l-NO in this setting. Three of these trials were
`previously reviewed as part of an earlier NDA submission by the Sponsor. The fourth trial was completed more
`recently and is reviewed as part of the present document.
`The data from three trials (NINOS, [NO-0]:r -02 and CfNRGI) demonstrate that l-NO administration is
`associated with a significant decrease in the use of extra-corporeal membrane oxygenation (ECMO), an invasive
`method of oxygenating the blood. This effect of I-NO to decrease ECMO use may well be due to the acute effect of [-
`NO to improve oxygenation, rather than due to any other beneficial effect on the course of the disease causing the
`hypoxic respiratory failure. In support of this contention, no beneficial effect of l-NO on mortality or any other
`clinical endpoint was demonstrated by the available data. No effect of-l-NO on mortality (beneficial or adverse) has
`been demonstrated by the data. The effect of l-NO to improve oxygenation is significant, however. and avoidance of
`ECMO is a clinically-desirable outcome. In the absence ofhard clinical benefit (cg, decreased mortality, fewer days
`of hoSpitalization) the safety of I-NO needs to be firmly established prior to allowing its non-investigational use.
`The following safety issues have been raised during one or both of my reviews:
`.
`1) The safety database included small number of subjects, and for most adverse events, the INO-Ol/ -02 was
`the primary source of information. Given the baseline differences between the subjects in the INO-OI/ -02 and the
`other trials, extrapolating between the two populations is also difficult, and open to serious errors of omission due to
`inadequate data. These difficulties have been alleviated to some extent by the addition of 97 additional patients who
`Were exposed to l-NO, bringing the total number of children exposed to I-NO in the NDA database to 375-. The
`difficulties with differences in baseline characteristics are again present in the CINRG] trial, complicating its
`interpretation. innother potentially confounding variable between the CTNRGI trial and the previous trials is the lower
`dose of I-NO administered in ClNRGl (20 ppm reduced to 5 ppm if possible), compared with the NfNOS and lNO-
`01/ -02 trials (20-80 ppm).
`- 2) The available safety database in the original NDA raised several potential safety issues. The most
`troubling of the adverse events, raised in the original medical review, was the possible association of l-NO with acute
`and chronic pulmonary toxicities. This association, like all of the safety data, relied on small numbers of subjects,
`although the association was plausible, given the available data. The addition of the CINRGI trial data, along with
`additional
`long-term follow-up data from NINOS and INO-Ol/ -02 has allayed some of the concerns, especially
`regarding the occurrence of chronic injury. The existing database is inadequate, however, to exclude the occurrence
`‘ of pulmonary toxicity inassociation with the use of I-NO.
`3) There was a definite association of l-NO with the development of methemoglobinemia and elevated NO;
`concentrations, identified in the NTNOS and [NO-Oil -02 trials (especially at the 80 ppm dose). This concern is
`minimized with the use of the lower doses of l-NO in the Cl'NRGl trial (and the proposed dose for the label).
`4) Several other adverse events were also possibly linked to the administration of [-NO based on the data
`available in l997, although the data were insufficient to determine the seriousness of these potential adverse events,
`or to determine their duration or dose-reSponse. The addition of the CINRGI data has resolved some of these safety
`concerns, and no new safety concerns have arisen as a result of the CINRGI trial review. The available data does
`suggest that rapid discontinuation of l-NO'is associated with rebound hypoxia in some patients.
`5) For some adverse events of interest, no data were obtained at all. Most critical of these was the effect of 1-
`NO on coagulation parameters. Other clinical events for which we have either scarce or no clinical data include:
`musculoskeletal injury; non-glomemlar renal injury; effects on the cardiac conduction system, and effects on serum
`electrolytes.
`'
`‘
`'
`'
`‘
`.
`6) The number of patients exposed to I-NO is too small to adequately assess the potential interactions of I-
`NO with disease states, patient demographics and concomitant medications. The potential interaction of l-NO with
`other drugs is of particular importance for drugs commonly used to treat this condition, such as steroids and
`vasodilators (with the exception oftolazoline).
`'
`7) Finally, an issue that cannot be resolved from the database is the potential genotoxicity and
`carcinogenicity of l-NO. The available data on the genotoxicity of l-NO are mixed (see section 4.] in my 1997
`review for details). It
`is true that the duration of exposure to I-NO is limited in these studies, and that [-140 is
`produced (at many-fold lower concentrations) intracellularly. However, the cumulative years of risk for a newborn.
`who receives I-NO is appreciably longer than an adult.
`-
`
`-
`
`I
`i
`'
`
`
`
`
`
`-
`NDA 20-845
`Primary Medical Review
`10.99
`-
`
`-
`
`‘
`
`.
`
`4
`
`

`

`fl m__‘
`
`-..
`
`H_______‘____"____ __
`
`___ MW
`
`-
`
`0.0 Overall Summary of Efficacy and Safety for I-NO (cont)
`The data suggest that I-NO has a dose-dependent, acute effect on oxygenation in newborns with hypoxic
`respiratory failure. This improvement
`in the physiology translates into a reduced use of ECMO, an invasive
`procedure with significant potential morbidity and mortality. There are no data demonstrating a clear beneficial effect
`of I-NO on hard-endpoints (death, days of hospitalization, days of ventilation, incidence of chronic lung disease or
`neurological sequelae). Given that there are other effective therapies and a falling mortality rate from this disease, the
`safety of I-NO becomes more critical to assess.
`Regarding safety, in distinction to the situation in 1997, the available database is more reassuring regarding
`the safety I-NO administration, especially the potential for neurologic injury. Additional data are needed to resolve
`the issue of pulmonary toxicity definitively. No new safety concerns have been identified from the ClNRGl review.
`The safety database is inadequate as regards to certain key adverse events, and insufficient data exist on the
`interaction of I-NO with other medications used in the treatment ofhypoxia respiratory failure.
`In conclusion, a clear clinical benefit of l-NO in this population has not been demonstrated.'A beneficial
`effect of l-NO on the physiology of these severely ill patients (improvement in oxygenation) has been demonstrated.
`This effect allows for a decreased use of an invasive and potentially dangerous procedure (ECMO). While the
`database does not address some safety concerns for l-N0,the available data suggest that the short-term use of l-NO is
`not associated with severe adverse events. Approval of l-NO to improve oxygenation in neonates with hypoxic
`respiratory failure is therefore recommended.
`
`1.0 Materials Utilized in the Review
`.
`.'
`-
`1) NDA 20-845
`2) Medical/Statistical Review of l-NO by DC. 'I'hrockmorton and Walid Nuri, completed 1 1.19.97.
`3) Published literature pertaining to I-NO.
`4) Draft Statistical review by Lu Cui, Ph.D., obtained 10.27.99.
`
`-
`
`.
`
`‘
`
`,
`7
`2.0 Background
`.
`The initial submiSsion of NDA 20-845, for inhaled nitric oxide (l-NO), took place in 1997. In the package
`were data fi'om three substantive trials: NINOS, [N086 and [NO-01! ~02. In addition, a fourth trial (IND-03)
`enrolled 14 patients before being halted. Two of these trials (NINOS and [NO-011' -02) were ultimately considered to
`be pivotal as regards to safety and efficacy-The NINOS trial was conducted by the NIH, while the lNO-Ol/ -02 was a
`combined trial initiated and supported by the sponsor. All of—the trials examined the effects of l-NO in newborn
`infants with hypoxic respiratory failure along with evidence (either clinical or by echocardiogram) of pulmonary
`hypertension. All of the trials were randomized, placebo-controlled, designed to be double-blind. Children with
`congenital diaphragmatic hemiaflung hypoplasia were excluded from these trials.
`'
`NINOS: With regard to efficacy, the NINOS trial was stopped early for perceivedoverwhelming clinical
`efficacy with regard to its primary endpoint (the incidence of ECMO or death). In this trial, I-NO (20-80 ppm)
`reduced the. use of ECMO in patients with hypoxic reSpiratory failure and had an acute effect
`to improve
`oxygenation.
`'
`INC-01! -02: After NINOS stopped early, the INO-OI/ -02 trial had a severe decline in enrollment, and
`ultimately stopped early as well. While this trial did not show a significant effect of I-NO (5-80 ppm) on its primary
`endpoint (death, receipt of ECMO, evidence of neurologic or pulmonary sequelae at 28 days), there was a trend
`towards a reduction in the use of ECMO in the I—NO group as a whole. There was also a dose-dependent effect of 1-
`NO to improve oxygenation. lNO-Dl/ —02 enrolled patients with less severe hypoxia than the other trials in the NDA,
`and was the only trial in the original ‘NDA submission to collect all reported adverse events and measure routine
`laboratories other than methemoglobin and NO; levels.
`
`specified primary endpoint.
`[NO-03: A small trial, IND-03.enrolled 14 patients before similarly being halted. Its contributes to the
`overall safety database.
`.
`.
`
`With regard to safety, concerns were raised regarding the long-term safety of l-NO. Most significant, in the
`final Medical/Statistical review document, dated 11.19.97, the possibility of acute and chronic lung injury following .
`l-NO was raised by this reviewer.
`'
`_
`.
`
`.
`
`'
`
`NDA 20-345
`Primary Medical Review
`
`.
`
`-
`
`5 .
`
`-
`
`I
`
`

`

`__
`
`,,__._. - —--
`
`- -
`
`- ---——————-————-——-——-——-“'__"-‘-*—‘—‘-‘——-—-_—_____
`
`_
`_
`.
`2.0 Background (cont)
`The sponsor ultimately chose to withdraw the NDA application prior to a decision about approvability. At
`thetime of the review, and in subsequent meetings, the weaknesses of the database were discussed with the sponsor,
`and they were encouraged to collect long-term follow-up regarding the clinical course of the infants in the available
`studies (for the NINOS and INO-OI/ -02 trials). The results of this follow-up have been summarized by the sponsor,
`and are presented below.
`In addition, the sponsor obtained the data fi-om a third randomized, placebo-controlled study of l-NO in
`hypoxic respiratory failure (the CINRGI trial). The results of this trial are presented by the sponsor in the present
`submission, and are reviewed below.
`
`.
`2.] Organization of Medical Review
`Because of the nature of the NDA submission, with a large portion of the data previously submitted and
`reviewed, the current document will differ somewhat from the usual format for'an NDA review. In particular, those
`trials 1 have reviewed previously are summarized in the present review, and the reader is referred to the review dated
`11.19.97 for further details. Tables that originally appeared in the previous review will be identified by having two
`numbers in their title: the numbering system used for the present review as well as the number corresponding to their
`location in the I997 review (in parentheses).
`_
`-
`This current review document will initially focus on the CINRG] trial data, followed by a review of the
`long-term data item the NINOS and INO-Ol/ -02 trials. The company has also performed additional analyses from
`the NINOS data addressing some of the concerns expressed by the Agency, and these will be presented in this section
`as well. The third section will focus on integrating the results from the three main clinical trials to assess the overall
`riskfbenefit ratio of the use of l-NO in this neonatal population with respiratory failure. This section will include the
`summary materials for the efficacy and safety data ofthe NINOS, INOSG and lNO-Ol/ -02, drawn largely from my
`original consult, incorporating the relevant data from the CINRGI trial..A review of the relevant published literature
`regarding the safety and efficacy of l-NO will also be included at this point.
`. The conclusions of the Medical Reviewer, including the recommendation regarding the approvability of 1-
`NO, will follow the integrated efficacy and safety summaries. References are to be found at the end of the document,
`followed by the Appendices, which will include the following:
`'
`l) a list of abbreviations used,
`2) narratives for the deaths in the ClNRGl trial, and
`3) the sponsor’s proposed label,
`including comments from the medical, chemistry and pharmacology
`reviewers where appropriate.
`'
`
`'
`
`The reader is referred to the original review of NBA 20-845, dated 11.19.97, for some of the normal aspects
`of an NBA review, including pharmacology, toxicology, carcinogenicity. The reviews of the NINOS, INGSG and
`INO-Ol/ -02 trials are also to be found there. Finally, the previous review contains an extensive review of the
`available animal and human literature published on nitric oxide. '
`'
`The reader is also referred to the Statistical Review of NDA 20-845 by Dr. Cui, submitted along with the
`current Medical Review.
`
`
`
`NBA 20-84 5
`Primary Medical Review
`[0.99
`-
`
`.
`
`6
`
`

`

`3.1 Title of Study
`A comparison of conventional therapy and inhaled nitric oxide in the management of persistent pulmonary
`hypertension of the newborn (Clinical Investigation ofNitric Oxide Research Group Initiative; CINRGI).
`
`3.2 Sites of investigation and Investigators
`The CINRGI investigators and their sites of investigation, along with the number of infants enrolled at each
`center, are summarized below. No financial disclosure information was available at the time of this review, so no
`statement is possible at this time regarding potential conflicts of interest on the part of the investigators.
`
`'Site Name
`# of Enrolled
`
`Su bjects
`
` All Children’s Hospital
`Arnold Palmer Hospital
`
`
`Carolina Medical Center
`Children's Medical Center of Akron
`
`
`Christ Hospital
`
`
`Duke University Hospital
`
`
`Egleston Children's Hospital
`
`
`Georgetown University Medical Center
`
`
`Grady Memorial Hospital
`
`
`- Greenville Memorial Hospital
`
`Medical University ol’South Carolina
`
`
`Ochsner Clinic
`
`
`Richland Medical Center
`
`
`St. Joseph's Hospital and Medical Center
`
`
`University of South Dakota
`
`
`University of Texas
`
`Vanderbilt University
`Wilford Hall Medical Center
`
`Total En rollment'
`
`a. 0fthese, 186 enrolled patients met the entry criteria and were ultimately randomized.
`
`3.3 Background
`
`_
`
`-
`
`--'
`
`-
`
`Initial protocol submitted: 12.95
`This submission was submitted as an amendment to the pilot study protocol (unblinded) which was finalized
`4.94. At that time, 36 patients had been enrolled. The protocol was changed to a randomized, double-blind study at
`,
`this time.
`
`First protocol amendment submitted:
`12.97
`1. Provided further details regarding the blinding of the trial, including the use of shrouds for the gas tanks.
`2. The primary endpoint was changed from the ‘number of infants who needed ECMO‘ to the ‘number of
`infants who received ECMO.’
`
`Second protocol amendment submitted:
`l. A steering committee was established.
`2. Monitoring of NO and NO; levels were added.
`
`6.98
`
`Study Initiation: 4.21.94
`
`EnrollmentTermination: 12.8.98
`
`
`
`NDA 20-345
`Primary Medical Review
`10.99
`-
`
`.
`
`‘
`
`7
`
`

`

`.
`3.4 Study Design
`-
`CINRGI was a multicenter, double-blind, randomized, placebo-controlled trial in infants with evidence of
`persistent pulmonary'hypertension of the newborn (PPHN), but without evidence of structural heart disease. An
`important feature of this trial, compared with the NINOS and INO-Ol/ -02 trials, is that the use of high-frequency
`oscillatory ventilation was encouraged prior to study enrollment for patients with significant parenchymal
`lung
`disease.
`.
`Eligible patients were first categorized according to disease type (e.g., respiratory distress syndrome,
`meconium aspiration syndrome), and then randomized to receive either placebo (N2) or l-NO.
`l-NO was started at 20 ppm and continued for at least 4 hours. At that point, if the P302 was >60 mm Hg
`with a pH $7.55, the dose was decreased to 5 ppm (otherwise the l-NO was continued at 20 ppm for a maximum of 24
`hours). Infants could be continued on l-NO 5 ppm for up to 96 hours ofgas administration or the patient was 7 days or
`age.
`.
`Treatment gas was continued until FiOz was <0.7, the patient had received 96 hours of therapy, or the patient
`was 7 days old. whichever came first. Once the HO; $0.7, weaning attempts were made. Treatment gas could be
`restarted ifthe patient required an Fi0220.80 to support a Pa02260 mm Hg. During the first 24 hours, the gas was re-
`started at 20 ppm. After 24 hours, the gas could only be restarted at 5 ppm. If the patient failed to respond to the
`reinitiation of study gas, they were deemed a treatment failure and the gas was discontinued.
`
`Treatment Failure! Patient Discontinuation
`After initiation of treatment, patients were assessed for the occurrence of treatment failure, defined as:
`]. Sustained hypoxemia:
`a. 0] >40 cm H20 on 3 of S ABGs drawn 30 minutes apart.
`b Pa02<40 mm Hg for 22 hours. or
`c. Pa02 <35 for >l hour.
`“ 2. Systemic hypotension (mean arterial systemic BP <35 mm Hg) unresponsive to medical management.
`3. Inadequate response to treatment gas:
`_
`a. P302<60 mm Hg after 24 hours. or
`b. deterioration in oxygenation status on the initiation of study gas, as evidenced by:
`r‘. a drop of >10 mm Hg PaO; or to a-value <40 mm Hg,
`ii. a drop in oxygen saturation'of>5%, or
`iii. a drop in 8:10; to <88%.
`_
`4. Failure to tolerate weaning from gas
`a. Failure to tolerate a decrease in the study gas‘to 5 ppm after 24 hours at 20 ppm.
`b. Failure to tolerate discontinuation of study gas at the end of 96 hours oftreatment.
`c. Oxygen saturation falls by >5% or to <88%.
`5. Elevated methemoglobin levels (>4%).
`6. Elevated NO; levels (>5 ppm).
`7. Parents withdraw informed consent.
`
`.
`.
`Blinding
`An unblinded investigator monitored the patients for N02 and methemoglobin leVels at all sites, necessitating
`a second, blinded, team who were responsible for all activities and data collection, excluding those data which could
`indicate the therapy (methemoglobin levels, NO; levels,- settings and calibration of the study gas delivery device).
`Every three hours the unblinded respiratory therapist recorded the N02 and NO levels, regardless ofthe treatment gas
`administered to the patients.
`'
`
`Concomitant Medications
`It has been suggested that the effectiveness of I-NO may vary depending on the type of ventilatory support
`given to the patients. To minimize this, the investigators attempted to standardize the delivery of high frequency
`oscillatory ventilation (HFOV) and surfactant. Infants with evidence ofparenchymal lung disease on CXR andlor poor
`lung inflation were started on HFOV prior to entry into the study, and patients with RDS were treated with at least one '
`dose of surfactant.
`
`NDA 20-345
`Primary Medical Review
`l0.99
`-
`
`
`3
`
`.
`
`

`

`3.4 Study Design
`Concomitant Medications (cont!
`All other therapies were allowed by the protocol, although two disease-specific guidelines were given to the
`investigators(NDA vol. 9.6, section 9.4.6):
`in these patients is to
`1. Pulmonary Iglpertensfon with associated parenchyma! lung'disease. The goal
`optimize lung inflation. In some patients, the optimization of lung inflation was expected to decrease pulmonary artery
`pressure.
`In patients with radiographic signs of low lung volume, end-expiratory or mean airway pressure was
`increased in an attempt to recruit collapsed alveoli.
`2. Pulmonary hypertension without significant parenchymol lung disease. in these, patients, the primary goal
`was selective pulmonary vasodilatation. Every attempt was made to avoid lung over-inflation and pressure-induced
`lung injury.
`'
`
`'
`
`3.5 Study Objective! Primary and Secondary Endpoints
`The objective of the CINRGI trial were to assess the safety and efficacy of I-NO added to conventional
`therapy for PPHN, compared with conventional thé'rapy alone. Patients with congenital diaphragmatic hernia (CDH)!
`lung hypoplasia were eligible for enrollment in the trial, but were excluded from the primary analyses.
`I
`
`.
`Primary Endpoint
`1. The number of patients in each treatment group that received ECMO.
`
`“
`
`Secondagy Endpoints
`1. Improvement in arterial oxygenation, measured by arterial-alveolar oxygen ratio (a-A ratio), the-alveolar-
`arterial oxygen gradient (A-aD02), the arterial partial pressure of oxygen (PaOz), and the oxygenation index (0]) in
`the treatment groups.
`'
`2. incidence of the following in the two treatment groups:
`a. Physiologic measures
`i. blood pressure,
`ii. gas exchange,
`iii. methemogiobin levels.
`b. Safety meaSures
`i. discharge home on 02 and/or pulmonary medications,
`ii. neurologic abnormalities
`iii. survival to discharge.
`
`3.6 Number of Subjects] Randomization
`‘
`Central randomization codes were created using blocks of ten patients in each disease stratum (e.g., RDS,
`MAS).-Enrollment books were created for each center including sealed envelopes containing the randomization
`packet. Each envelope was labeled with the site designation, the underlying disease stratum, and a sequential patient
`number. The blinded investigators were to open the envelopes sequentially within each stratum.
`'
`
`3.7 inclusion! ExclusionCriteria
`
`Inclusion Criteria {all must be present}
`1. Estimated gestational age 234 weeks.
`._
`2. Age <96 hours at time of entry.
`3. Severe re5piratory failure defined as OI >25 on Optimized ventilator settings (including HFOV in all
`patients with significant parenchymal disease).
`‘
`4. Postductal arterial line for sampling blood gases.
`
`andJor
`
`At least one ofthe following must also be present:
`5. Echocardiographic evidence of pulmonary hypertension without structural heart disease;
`6. Clinical evidence of pulmonary hypertension with at least one of the following:
`a. differential oxygenation in preductal and postductal areas.
`b. marked clinical lability in oxygenation despite optimized treatment ofthe lung disease.
`1) >2 desaturation events within .12 hours due to PPHN and not vent setting.
`2) need for extreme alkalosis to maintain adequate oxygenation (pH >160).
`
`'NDA 20-845
`Primary Medical Review
`“3.99
`-
`
`
`'
`
`-
`
`9
`
`

`

`_ fi, “0-..... “%._._.____ _;,
`
`,
`
`
`
`"
`
`3.7 Inclusion! Exclusion Criteria (cont)
`Exclusion Criteria [none can be present)
`1. Urgent need for ECMO:
`‘
`a. Refi'actory hypotension (< 35 mm Hg) despite vasopressors and volume support.
`b. Refractory hypoxemia (PaOz <30 mm Hg) despite maximum ventilatory support.
`2. Lethal congenital anomaly.
`3. Significant bleeding diathesis.
`4. Active seizures on anticonvulsants, or a history of prolonged severe asphyxia:
`a. Pa02<10 mm Hg for 1 hour or Pa02Q0 mm Hg for 2 hours.
`b. pH (LO for >1 hour despite resuscitative efforts.
`5. Cyanotic congenital heart disease.
`6. Any other reason that would exclude the use of ECMO.
`
`3.8 Dosage! Administration
`Treatment gas was administered using a dielivety device that diluted the gas (100% N; or NO 800 ppm) 1:20
`delivered to the endotracheal tube. Adjustments to the study gas rates were similar whether the patient received
`placebo or I-NO.
`
`.
`3.9 Duration! Adjustmentof Therapy
`Treatment gas (I-NO or N2) was started at 20 ppm and continued for at least 4 hours. At that point, if the
`Pa02 was >60 mm Hg with a pH $7.55, the dose was decreased to 5 ppm (otherwise the l-NO was continued at 20
`ppm for a maiimum of 24 hours). infants could be continued on l-NO 5 ppm for up to 96 hours of gas administration
`or the patient was 7 days of age.
`Adjustments made to study gas are discussed in section 3.4 above.
`.4
`
`.
`3.10 Safety and Efficacy Endpoints Measured
`The timing of the various tests performed as part of CI‘NRGI is presented in tabular form below.
`
`Table 3.10.1 Timetable for clinical observations and lab measurements in CINRGla.
`_ stuuynay
`
`Discharge
`Day 30!
`Completion
`
`
`
`
`X
`X I
`X
`
`'
`
`_ X
`
`.
`
`X
`
`
`
`illllixx><l
`
`
`
` Oxygenation para meters
`
`Hemodynamics
`
`Ventilation parameters
`
`
`
`
`
`Im—
`
`lemma—
`
`a. Data from NDA vol. 9.6, Table 2.
`
`3.1] Statistical Considerations
`-
`7
`Power
`The sample size was based on an expected rate of ECMO use in the placebo group of 50% compared with a
`30% .ECMO use in the l-NO group. Using an overall alpha of 0.05, with 80% power, along with a orie-to-onc
`randomization of placebo/I-NO group it was determined that 103 patients in each arm would be needed.
`
`_
`Multiplicig
`No adjustments for multiplicity were performed.
`
`~
`
`Interim Analyses
`There were no interim analyses.
`
`NDA 20-845
`Primary Medical Review
`-
`10.99
`
`
`
`10
`
`‘
`
`

`

`3.11 Statistical Considerations (cont)
`.
`_
`, .
`Statistical Analysis
`The primary endpoint of the trial was the number of patients in each treatment group who received ECMO.
`The principal analysis was an intent-to-treat analysis stratified by disease using Cochran-Mantel-Haenszel (CMH)
`method.
`
`The secondary endpoints were analyzed using CMH, Fisher’s 2-tailed exact test, or Student's t test as
`appropriate.
`Because infants with pulmonary hypoplasia were expected to have substantially higher morbidity/mortality
`as well as responses to l—NO, the safety outcomes were analyzed separately and then integrated with the other infants
`in the trial.
`.
`
`_
`3.12 Efficacy Outcomes for the CINRGI Trial
`3.12.1 Subject Demographics & Baseline Characteristics
`The demographic and clinical backgromdfiam for the subjects enrolled in CINRGI are summarized below.
`
`Table 3.12.1.1 Demographics ofCINRGIa.
`
`Demographic
`Placebo
`l-NO
`p Value
`N=89
`N=97
`Gender
`
`
`
`52 (58.4%)
`
`.
`
`44 (45.4%)
`
`.
`
`
`
`,
`
`
`
`
`
`
`
`
`
`
`
`
`Male '
`Race in (V0)
`Caucasian
`Black
`Hispanic
`‘
`Other
`Mean Age Since Birth (hrs 18D)
`Mean Age ($59)”
`Mean Weight, kg (15D)
`Apgar Scores
`
`14:18
`1 Minute
`7:311:222
`5 Minute
`
`a. Data from CINRGl study report, table 7-8.
`b. p Vaiue per sponsor.
`e. Mean age assessed by physical exam at birth.
`
`40 (41.2%)
`44 (49.4%)
`
`43 (44.3%)
`33 (37.1%)
`
`8 (8.2%)
`10 (l 1.2%)
`
`
`2 (2.2%)
`6 (6.2%)
`
`
`29 9:16 5
`30.0302
`
`38.8:2 l
`39 221:1.7
`13:0 6
`
`
`‘
`
`
`
`-
`
`
`
`Table 3.12.1.2 Birth demographics of CINRGla.
`
`
`
`
`
`
`
`Obstetrical complications
`64/89 (71.9%)
`54/97 (55.7%)
`.
`46/96 (47.9%)
`Fetal Distress
`48/89 (53.9%)
`
`
`
`
`ffiékiféifiriétién we}. '§3£§QX§§7£%JET' swam: .
`
`Resuscitation at birth
`9/89 (10.1%)
`8/96 (8.3%)
`
`
`
`
`CPR at birth
`I (1.1%)
`10.0%)
`
`
`
`
`88/95 (92.6%)
`Normal Head U/S
`79/87 (90.8%)
`
`a. Data from ClNRGl study report, table 7-8.
`b. p Value per sponsor.
`'
`c. Mean age assessed by physical exam at birth.
`
`The two treatment groups were well-balanced with regard to the cause of the hypoxic pulmonary failure.
`
`Table'3.12.1.3 Underlying disease leading to hypoxic respiratory failure in CINRGI'.
`Underlying Disease
`.
`Placebo
`l-NO
`'
`N=89
`Nfl?
`35 (39%)
`34 (35%)
`21 (24%)‘
`24 (25%)
`
`8 (9%) -
`8 (8%)
`
`25 (28%)
`31 (32%)
`
`
`
`
`
`
`
`Meeonium aspiration
`Pneumonia/Sepsis
`
`Respiratory Distress Syndrome
`
`Persistent Pulmonary Hypertension
`
`a. Data from Cl'NRGI study report. table 32.
`
`
`
`
`
`NDA 20-345
`Primary Medical Review
`10.99
`.
`
`
`1 1
`
`'
`
`

`

`3.12.1 Subject Demographics & Baseline Characteristics (cont)
`The investigators collected individual information on the severity of the pulmonary disease at entry as well.
`Note-that the control group had significantly more airleak and more pulmonary hemorrhage compared with the l-NO
`group. The severity of the pulmonary injury on CXR was similar in the two groups.
`
`
`
`Table 3.12.1.4 Pulmonary disease at birth in ClNRGla.
`Peri-Natal Demographic
`Placebo
`”“0
`N=89
`N=97
`
`mu-
`fairlsahimdr'pmeii 534 "
`éPtilI-fi'o'n'isi‘yiflehTor’rhagep
`‘1
`o
`‘
`I
`(
`.
`
`
`
`Lung Disease on CXR
`
`
`
`
`
`None
`6 (6.7%)
`4 (14.4%)
`
`
`
`28 (31.5%)
`Mild
`26 (26.8%)
`
`
`
`
`41 (42.3%)
`Moderate
`42 (47.2%)
`
`
`
`
`Severe
`13 (14.6%)
`16 (16.5%)
`
`
`a. Data from ClNRGl study report. table 7-8.
`b. p Value per sponsor.
`'
`
`p Valueb .
`
`
`
`Vasoactive drugs used at time of entry were Well-balanced as to class among the treatment groups, as
`summarized below. There was a trend towards more use of epinephrine in the l-NO group among a small number of
`infants. in data not shown, fl-ie average dose of epinephrine and tolazoline in the small number of infants receiving
`those therapies was higher numerically in the l-NO group.
`7
`
`Table 3.12.1.5