Clinical data sources
`
`.
`
`‘
`
`NBA 21 -2 72
`UT-I 5 for pulmonary hypertension
`
`study allowed for the long-term treatment of patients who were treated either with
`active UT-lS or vehicle in studies P01:03, P01: 04 or P01:05. In addition a total of 208
`patients not previously enrolled into clinical studies were treated in an open-labeled
`manner.
`
`4.1.2 Subject enumeration and exposure
`
`As of the cutoff date, exposure was 476 subject-years, with 224 subjects treated for
`more than 1 year. This open-label study comprises the bulk of the exposure to UT—lS.
`The exposure in this study is shown in Figure l.
`
`
`
`
`
`
`
`Figure is a stacked bar chart in which each subject contributes in one of a number of
`states on each day after enrollment. Subjects entering from studies ”1:03, P01:04,
`and P01:05 have dosing information prior to enrollment in P01:06. Data obtained
`fiozn 120—day safety update.
`
`The proportion of subjects who remained alive, in study, and on a non-zero dose of UT-
`15 is shown in Figure 2.
`
`G:\N21272.doc
`
`'
`
`——7——
`
`1 6:09 Friday, March 09, 2001
`
`Last saved
`
`

`

`clinical data sources
`
`p
`
`,
`
`NDA 21-2 72
`UT-1 5 for pulmonary hypertension
`
`For subjects in study P01106 who remained on any non-zero dose, the proportion on
`various doses is shown in Figure 3.
`
`Proportion of subjects remaining on s non-zero dose of “-15 among subjects not
`censored by the reporting cutoff date. This is not I true life table, because subjects
`could go to a zero-dose and subsequently return on treatment.
`
`
`
`
`The demoninstor is the number of subjects on any non-zero dose. Dots from 120-dsy
`safety update.
`
`4.1.3 Demographics
`
`There were few males, few non-Caucasians, and few subjects over age 65. No separate
`analyses were performed in these subgroups.
`
`G:\N21272.doc
`
`——8—
`
`Last saved
`16:09 Friday, March 09, 2001
`
`

`

`vurueui uutu auul LCD
`
`,
`
`JVUA 41 'Z /Z
`
`UT—1 5for pulmonary hypertension
`
`4.2 Secondary source data
`
`_,
`4.2.1 Other studies
`There are no other known studies with UT-lS.
`
`4.2.2 Post-marketing experience
`
`There is no post—marketing experience with UT-15.
`
`4.2.3 Literature
`
`No publications were found that did not correspond with identified studies.
`
`4.3 Adequacy of clinical experience
`
`The development program appears to have been large enough to have reliably detected a
`reasonably sized treatment effect. The population studied contained relatively few males
`and relatively few representatives of racial minorities, but there are no data to suggest
`such groups respond differently to pulmonary hypertension or to treatments for
`pulmonary hypertension.
`
`Long-term exposure in approximately 600 subjects or 500 subject-years is adequate to
`exclude, with 95% confidence, an incidence of unobserved adversity at the rate of about
`one per 150 exposed patients or one per 125 patient-years. This is rather less safety
`data than is frequently available for the evaluation of a new chemical entity.
`
`4.4 Data quality and completeness
`
`Case report forms were provided for all subjects who died or were withdrawn for
`medical reasons. A spot—check comparing values in the CRF with the sponsor's
`electronic data revealed no discrepancies.
`
`APPEARS THlS WAY
`0N ORIGINAL
`
`G:\N21272.doc
`
`-—-9—
`
`16:09 Friday, March 09, 2001
`
`Last saved
`
`

`

`huey. area remew of effectiveness
`
`’
`
`,-
`
`NDA 21-272
`UT—l 5 for pulmonary hypertension
`
`5 Integrated review of effectiveness
`
`Studies P01:04 and P01:05 are the key pivotal studies3. The procedures and
`measurements for these two protocols were identical and the studies were analyzed
`both individually and as a single pooled study. Subjects who enrolled into these studies
`were symptomatic pulmonary hypertension subjects (NYHA Class II-IV), despite
`optimum concurrent therapies. The etiology of the pulmonary hypertension could be
`either primary disease or could be as consequence of either collagen vascular disease or
`left to right congenital shunts.
`
`5.1 Six-minute walk
`
`The primary end point was the change in walking distance from baseline at the end of
`week 12. For the pivotal analyses, missing values for those who discontinued were
`imputed. Those who discontinued either because of death, deterioration or
`transplantation received the worst rank or worst value. Those who discontinued due to
`adverse events had their last rank carried forward or their last metric for walk carried
`forward.
`
`The primary method of analysis was a non-parametric analysis of the pooled studies.
`The database was to be considered demonstrating a benefit for UT~15 if either both
`studies were by themselves significant at the p< 0.049 or if one study was significant
`(p< 0.049) and the pooled studies had a p-value of less than 0.01.
`
`By the sponsor’s own analysis the database would not be considered successful. Neither
`of the studies demonstrated a p—value of < 0.049 (p=0.06 for both studies), although the
`pooled studies demonstrated an overall p-value of < 0.01 (p=0.006 for the pooled
`studies]. The magnitude of the change in median walking distance was small, ranging
`from 2 meters in study P01:04 to 19 meters in study P01:05. The fractional increase in
`walk distance over baseline for UT-15 patients relative to vehicle was between < 1% to a
`6% increase for each or the studies and a pooled increase of approximately 3%.
`
`Not only did the sponsor’s analysis not meet the pre-specified criteria for considering
`the trials a success but also there was an inherent bias in the statistical approach
`employed in the analysis of the study. There was a clear imbalance in the number of
`subjects who discontinued for adverse events. Nearly all such discontinued subjects
`were treated with UT-15 and nearly all those who discontinued did so for infusion site
`pain or infusion site reaction.
`
`There are several consequences that result from this algorithm for imputing data for
`discontinued subjects. First, those who discontinue due to adverse events could never
`be classified as worst outcomes even if they should subsequently die, deteriorate or
`receive a lung transplant. The fraction of subjects who discontinued for adverse events,
`therefore, was shielded from the worst imputed outcome values possible in this study.
`
`Second, nearly all subjects that discontinued in the UT-15 group did so because of
`infusion site pain/ reaction. Since infusion site pain was ubiquitous in the UT-15
`subjects, those who discontinued were possibly suffering from infusion site pain in
`conjunction with a worsening of their pulmonary hypertension. The attribution of cause
`and therefore the imputed value was markedly dependent on this attribution.
`
`Third, the process of imputation presupposes the values at early times are reflective of
`the performance atwthe time of discontinuation. There are clearly subjects whose
`imputed value for walking distance does not reflect their status at the time of
`discontinuation. Subjects who discontinue for pain, whose discontinuation fell within
`
`3 For a flat description, see Section A.4 on page 82.
`
`G:\N21272.doc
`
`——1 0—
`
`Last saved
`16:09 Friday, March 09, 2001
`
`
`
`

`

`uucylutcu ICUIGW UJ CJJCLLLUCILCbb
`
`.-
`
`IVUA 21-2/2
`UT-1 5 for pulmonary hypertension
`
`the time-window of an exercise test and who did not undergo testing were imputed an
`earlier value, which would likely be better than their current status.
`
`Lastly, there was an asymmetry in the need for pain medication that could alter
`vascular dynamics or mitigate some of the disease symptoms particularly those that are
`associated with pain.
`
`In order to deal with the inherent biases due to the unequal rates of discontinuation for
`adverse events the data was analyzed in three additional ways. The first analysis
`included as worst outcomes three UT-lS and two vehicle subjects who died or were
`transplanted during the 100-day window of the study. The resulting p—values of the
`pooled database to 0.02 and that for the individual studies to >0. 1.
`
`The second analysis further includes as worst outcome, those subjects who
`discontinued for adverse events if Flolan® was started within one month of
`
`discontinuation and within the window of the study. There were six additional subjects.
`Two subjects were started on Flolan® either prior to or immediately upon
`discontinuation of UT—lS. Two additional subjects were started within two weeks of
`discontinuation of UT-15 and two within one month of discontinuation of UT-15
`
`therapy. None of these subjects obviously required Flolan® at baseline and the need for
`Flolan® upon discontinuation of UT—lS suggests that the subject’s status had
`deteriorated. The p-values for the pooled and individual studies when treating those
`subjects started on Flolan® within 1 month of discontinuing UT-15 as well as those
`who died or required transplant as worst outcomes, no longer are significant. For the
`pooled data, the p-value was 0.082. For the individual studies the p value was > 0.2.
`
`A third analysis also included all those who were treated with Flolan® during the
`window of the study as worst outcomes. In addition, there was one subject whose
`status at the time of discontinuation appeared to be inconsistent with the imputed
`measurement from week 1. The value for this subject was excluded. The p-values for
`this analysis for the pooled data was >0. 1. The p—values for each of the individual
`studies were >0.2.
`
`The above analyses presume that all subjects who discontinued UT-15 therapy and
`received Flolan® did so because they deteriorated. Some or all of these subjects,
`however, may have been started on Flolan® because no other options were available. An
`alternate analysis, performed by the sponsor imposes a last rank value for all those who
`discontinued prematurely, even if the reason was death, deterioration or need for
`transplantation. This analysis removes one source of the bias against the placebo in
`that no subject received a worst outcome. This analysis is sponsor's analysis # 4 in this
`review. The p-value for the pooled studies was 0.011 and that for the individual studies
`was between 0.07-0.08.
`
`In summary, the study did not succeed by the pre-specified criteria of success. Neither
`study P01:04 nor P01:05 was by itself statistically significant by a method of analysis
`that biases results towards UT-15 treatment. Additional analyses that corrected for the
`asymmetry in adverse events completely eliminate any benefit even for the pooled
`studies.
`
`5.2 Supportive metrics
`
`Since the primary outcome of the study did not succeed by the pre—specified criteria,
`supportive measures of efficacy are more difficult to interpret. Nevertheless, there is a
`suggestion from the supportive information that UT-lS may have some effect on
`symptoms associated with severe pulmonary hypertension. The supportive symptoms
`were collected only among those who completed the study. Those who discontinued for
`any reason did not have any values imputed. In addition, the supportive symptoms
`
`G:\N21272. doc
`
`——1 1——
`
`Last saved
`.
`16:09 Friday, March 09, 2001
`
`
`
`

`

`uuegtuucu I<-Utew UJ <djlvytlbplleba
`
`5
`
`NBA 21 .2 72
`
`UT—I 5for pulmonary hypertension
`
`were administered by the treating physician who might have been aware, based on the
`nature of infusion site reaction the subject’s treatment.
`
`Subjects showed improvement in the composite of sixteen signs and symptoms of
`pulmonary hypertension. The metric that was used was a composite of all these
`symptoms. Subjects were assigned a “+ l” for symptoms present at baseline and absent
`after l2-weeks, and a “-1” for symptoms that went from absent to present. Symptoms
`that were present at baseline and present at end of study, or absent at baseline and
`absent at end of study were assigned a value of '‘0”. The net change for each subject
`was averaged over all those who approximately 1 unit for those treated with UT-15. The
`specific symptoms that were improved or were less frequently worsened in the UT-lS
`group were dizziness, palpitations, orthopnea and chest pain. The most troublesome
`symptoms of pulmonary hypertension, dyspnea and fatigue did not appear to be
`differentially resolve across groups.
`
`A second metric that was prospectively collected as a supportive end-point was the
`dyspnea fatigue index. This metric consists of three components with values ranging
`from 0-4. The three components are I‘magnitude of task”, “magnitude of pace" and
`“functional impairment”. The higher the value, the less symptomatic the subject. There
`was a net increase of approximately 1.4 units in the overall symptom score among those
`treated with UT—lS, approximately equally divided among the three components of this
`metric.
`
`The quality of life metric was the Minnesota Living With Heart Failure questionnaire.
`This questionnaire consists of 21 questions and is divided in to 4 dimensions. This
`questionnaire was validated among subjects with CHF but not among patients with
`pulmonary hypertension, although the specifics of the questionnaire should be broadly
`applicable to patients with pulmonary hypertension. The questionnaire consists of a
`global and three components termed dimensions (i.e. physical, economical and
`emotional dimension). This questionnaire was not apparently administered to all
`subjects. Overall the global QOL did not differ between the two treatments. The physical
`dimension [portion of the questionnaire, however, was statistically favored the UT-15
`group.
`
`Each subject was asked to rank his or her degree of breathlessness after each six—
`minute walk by the Borg-dyspnea scale. This metric ranged from 1—10. The higher
`numbers suggest greater degrees of shortness of breath. The exercise coordinator
`performed this task and consequently is more likely to have been shielded from telltale
`signs suggesting active drug or vehicle use. Both the pooled studies and each of the
`individual studies were highly significant in improvement (p<0.01) of this metric. The
`magnitude was approximately 0.8 units.
`
`5.3 End points related to the natural course of the disease
`
`Despite modest effects on measurements of performance and symptoms, there does not
`appear to be any evidence that UT—15 alters the natural course of pulmonary
`hypertension. Deaths, hospitalizations, cardiovascular/ pulmonary hypertension
`hospitalizations or need for new or increases in medications were no different between
`groups. The need for inotropic or Flolan® support during the 12-week study did not
`differ between the two treatments.
`
`There were a total of 19 subjects who died during the window of the study. Ten of these
`subjects were in the vehicle group and nine in the UT-15 group.
`
`Hospitalizations were equivalent in both groups. There were 40 subjects who were
`hospitalized or had their hospitalizations prolonged among the vehicle group and 38
`among the UT—15 group. Two of those hospitalized among those randomized to vehicle
`were hospitalized after accidentally receiving UT-15. The investigators at the various
`study sites did not adjudicate cause-specific hospitalizations. This reviewer, based on
`G:\N212 72. doc
`Last saved
`
`-—12——-
`
`16:09 Friday, March 09, 2001
`
`

`

`..,..,_. -_yv\4 luvl-vw UJ vapuL/vuvvu
`
`.
`
`lVU/‘l 21-2/2
`
`UTF-1 5 for pulmonary hypertension
`
`the capsular summaries found 22 of those treated with UT-lS and 25 of those treated
`with vehicle had their hospitalizations prolonged or had a de novo hospitalization as a
`consequence of cardiovascular or pulmonary hypertension etiologies.
`
`Subjects who status deteriorates may require new medications or increase in doses of
`ongoing medications. A difference in the need to alter medications may suggest a benefit
`of a given treatment. For the purposes of this assessment the following drug classes
`were considered: loop diuretics, calcium channel blockers, vasodilators (including
`hydralazine, clonidine, nitrates), ACE inhibitors or angiotensin II blockers, oxygen,
`Flolan®, pressors, steroids, digoxin, aldactone or non-loop diuretics. The number of
`subjects who required no change in medication or no increase in medication comparing
`baseline to week-12 were similar in both groups.
`
`There was no difference in the number of subjects who required Flolan® or inotropic
`support. This reviewer counted 12 subjects in the UT-15 group and 10 in the vehicle
`group that required one of these medications.
`
`5.4 Hemodynamics
`Among those who completed the study, there was a modest improvement in
`catheterized hemodynamics. Right atrial pressures, pulmonary artery pressures (mean,
`systolic and diastolic) and pulmonary vascular resistance were decreased. Cardiac
`index, stroke index and mixed venous oxygenation were increased. The effects on
`hemodynamics, though statistically significant were in general small and of uncertain
`consequence. For cardiac index the net change (assuming that the data for those
`measured is consistent with the whole group) there was a net increase of 8%. There was
`an approximately 5% (3 mm Hg) decrease in mean pulmonary artery pressure. There
`was an approximately 18% decrease in pulmonary vascular resistance.
`5.5 Dosing
`Dosing was predicated on improving symptoms of pulmonary hypertension while
`minimizing excessive pharmacologic effect or infusion related adverse events. It is
`therefore not possible to define either the initial, optimal or an appropriate dose range
`of use for UT-15 based on the data from this study or from the database as a whole.
`
`Despite nearly an order of magnitude increase in mean infusion rate, there was minimal
`increase in walking distance among those treated with UT—15. The observed differences
`more reflect a worsening of the distance walked by the vehicle group than by an
`improvement among those taking larger and larger infusions of UT.15. There was no
`randomized withdrawal to ascertain a persistent (or any) benefit of UT-15. In fact among
`the handful of subjects who discontinued UT-15 acutely, no evidence of rebound was
`described. It is therefore unclear if there was any persistent beneficial effect of UT-lS.
`
`5.6 Comparison with Flo lan
`
`Flolan is approved for the treatment of patients with primaryxpulmonary hypertension
`Class III and IV. However, its use is difficult and inconvenient. The infusion of Flolan
`requires the insertion of an indwelling central catheter with the attendant risks of the
`inserting the catheter and the subsequent risk of catheter infection. Flolan has a rapid
`half-life and rapid dissipation of its hemodynamic effects. Any inadvertent interruption
`of the infusion is potentially life threatening. Flolan is chemically labile at room
`temperatures and must be reconstituted every 8 hours or kept at cold temperatures
`during the infusion. UT-15 was developed to avoid these problems and thereby delay
`the time till Flolan treatment becomes infusion.
`
`There is no scientific rationale to concurrently use UT-15 with Flolan. There is also no
`empirical safety or efficacy information on the concurrent use of these drugs. UT-15 is
`intended as treatment of pulmonary hypertension solely to postpone starting Flolan.
`There is no study that randomized patients to Flolan or UT- 15 that demonstrates
`G:\N212 72. doc
`Last saved
`
`-—-13—
`
`1 6:09 Friday, March 09, 2001
`
`
`
`

`

`1. neg. «.uéa . evzew OJ ejjeuweness
`
`,.
`
`NDA 21-272
`UT-l 5for pulmonary hypertension
`
`equivalent outcomes so that there may be unintended negative consequences in the
`delay of Flolan infusion. Comparing the labeling of Flolan to the likely labeling of UT-lS
`the mortality benefit for Flolan does not appear to be uniformly observed with UT-15.
`The current labeling of Flolan states:
`
`‘Suruiual was improved in NYHA functional Class III and IV
`PPH patients treated with FLOLANfor 12 weeks in a
`multicenter, open, randomized, parallel study. At the end of the
`treatment period 8 of 40 patients receiving standard therapy
`alone died, whereas none of the 41 patients receiving FLOLAN
`died (P=0.003). '
`
`In the pivotal UT-lS studies (P01:04 and P01:05) the drug demonstrated no mortality
`benefit. The UT— 15 study population consisted of predominantly (55%) primary
`pulmonary hypertension patients with the vast majority NYHA Class III and this portion
`of the population coincides with the population for which Flolan demonstrated a
`mortality benefit. There were 9 deaths among those randomized to UT-lS and 10
`deaths among those randomized to vehicle during the l2~week study. Five of the 9
`deaths on UT-15 were patients with primary pulmonary hypertension while 8 of the 10
`deaths on vehicle were patients with primary pulmonary hypertension.
`
`Performance benefit on the 6-minute walk for UT~15 patients was small, approximately
`3% of the baseline walk distance. Performance among those with Flolan was
`approximately 35~SO% of baseline walk distance. Admittedly, the basis of comparison is
`across studies with different designs. Nevertheless, the magnitude of effects a does give
`one pause before assuming equivalence between UT-15 and Flolan.
`
`APPEARS THIS WAY
`' 0" ORIGINAL
`
`G:\N2 1 2 72. doc
`
`-—-14—-—
`
`16:09 Friday, March 09, 2001
`
`Last saved
`
`

`

`Huey! uLeu I emcw a] any my
`
`i
`
`NUA 21 -2 72
`UT—1 5 for pulmonary hypertension
`
`6 Integrated review of safety
`
`6. 1 Methodology
`
`6.1.1 Mortality
`
`.
`
`,
`
`Mortality was a primary end point in no studies. Most studies were of duration short
`compared with the lifetime of patients with pulmonary hypertension, so most mortality
`occurred in the 12-week studies P01:04 and P01:05 and the long—term open-label
`follow-on study P01:06.
`
`This review contains narrative summary of all of the deaths developed from the
`sponsor's summaries and the case report forms.
`
`6.1.2 Withdrawals
`
`Case report forms were available for medically related withdrawals. There were also
`narrative summaries of these events.
`
`, 6.1.3 Adverse events
`
`Case report forms provided a mechanism for reporting adverse events and identifying
`their seriousness, severity, and relationship to study drug. The sponsor provided a
`summary narrative and case report forms for serious nonfatal adverse events.
`
`Common adverse events were separately tabulated for studies of normal volunteers,
`placebo-controlled studies in pulmonary hypertension, and chronic studies of open-
`label administration.
`
`6.1.4 Laboratory findings
`
`The principal laboratory data of interest were obtained during placebo-controlled
`studies. The sponsor tabulated changes from baseline in mean laboratory values.
`Electronic datasets provided by the sponsor were used to construct baseline-vs-on-
`treatment plots for data from continuous measurements. Such graphs show the
`relationship between baseline and on—treatment values with a 45-degree line of no effect
`for orientation and, in the margins, box-and-whiskers plots of the distributions in
`various treatment groups.
`
`6.1.5 Vital signs
`
`Vital signs were monitored as a conventional aspect of in-hospital patient care, and they
`were systematically collected for the first 8 hours of randomized treatment in placebo-
`controlled studies P01:04 and P01:05.
`
`6.1.6 ECGs
`
`ECGs were collected at baseline and at the end of active treatment in the placebo-
`controlled studies. These were analyzed much like the laboratory data from these
`studies.
`
`6.2 Results
`
`6.2.1 Exposure
`The sponsor's Integrated Summary of Safety covers the period up to 31 May 2000 for
`deaths and serious adverse events, and up to 4 February 2000 for other safety data.
`
`A summary of studies, exposure to study drug, deaths, and serious adverse events is
`shown in Table 2.
`
`G:\N21272.doc
`
`_
`
`Last saved
`
`-——1 5—
`
`1 6:09 Friday, March 09, 2001
`
`

`

`Integrated review of safety
`
`,
`
`i
`
`NDA 21-272
`UT-I 5for pulmonary hypertension
`
`Table 2. Exposure to UT-154.
`
`-AaysAAA—AAAAA
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`1
`
`Studies of clinical -harmacolo in sub'ects with CHF
`,
`——_————
`IEIil-—_-_- 12
`mmnm
`Studies of oharmacokinetics in normal volunteers
`
`mmnnn '
`_-_—I--_n‘
`mum-u
`mun—_—
`Controlled studies in -ulmon
`h ‘- rtension
`
`mun-um.-
`
`”—m-n
`
`Uncontrolled studies in ulmona
`h pertension
`
`—-n———_
`”mu-u..-
`_——-II-———
` ' Portoulmona h ‘0 rtension
`
`mum-Em—
`Peri - heral vascular disease
`
`A brief description of these studies is given in the following paragraphs.
`
`Study P01:01 was a two-period crossover study in which subjects were titrated to the
`maximum tolerated dose of Flolan and UT-15 and maintained for 90 minutes.
`
`Study P01202 was a parallel study in which subjects received UT-15 10 ng/ kg/ min for
`1.25 h, followed by 5, 10, or 20 ng/kg/min for 2.5 h.
`A
`
`Study P01:03 was a parallel study in which subjects received placebo or UT-15 for 8
`weeks.
`
`Studies P01204 and P01205 were parallel studies in which subjects received placebo or
`UT-15 for 12 weeks.
`
`Study P01:06 is an ongoing open-label study conducted among subjects previously
`enrolled in Studies P01:03, P01:04, or P01:05, and 208 subjects newly enrolled. As of
`the cutoff date, exposure was 476 subject-years, with 224 subjects treated for more
`than 1 year.
`
`Study P01:07 was a two-period crossover study of bioavailability, comparing
`intravenous and subcutaneous administration of UT-IS 15 ng/ kg/ min over 2.5 hours.
`
`Study P01 :08 was a two—period cross-over study of pharmacokinetics, comparing
`placebo with acetaminophen 1000 mg in subjects on sc infusions of UT-15 15
`ng/ kg/ min for 7 days.
`
`Study P01209 was an open-label, forced titration study in which normal subjects
`received ascending doses of UT-IS 2.5, 5, 10, and 15 ng/ kg/ min each for 7 days.
`
`‘ N=total enrollmen: / exposed to UT-JS; With-withdrawals total/medical
`
`5 The study is ongoing. The rmmbers refer to the NDA wtojf date of 1 October 2000.
`
`G:\N21272.doc
`
`——1 6—
`
`-
`
`Last saved
`1 6:09 Ffiday, March 09, 2001
`
`
`
`

`

`nueyl area . emew of safety
`
`‘
`
`,
`
`NBA 21 -272
`UT-1 5for pulmonary hypertension
`
`Study P0121] was an open-label study of the transition to UT- 15 of subjects receiving
`Flolan.
`
`Study P02:01 was an open-label, baseline-controlled study with dosing at 10
`ng/kg/min so for 2.5 h.
`
`Study P76:01 was a single—center, open label study in subjects NYHA III—IV CHF.
`
`6.2.2 Deaths
`
`Two subjects died subsequent to completion of study P01:02.
`
`Study P01202 subject #02005 was a 47 year old, 72-kg, Caucasian
`female with a 4'year history of primary pulmonary hypertension,
`NYHA N at baseline. She completed study with UT-lS without
`- problems except intermittent backache". The following day, she had a
`central venous catheter placed for Flolan administration, but she
`developed electromechanical dissociation and died despite CPR
`efforts, prior to Flolan administration. The investigator and sponsor
`regard the death as not reasonably attributable to UT-lS.
`
`Study P01:02 subject #040047 was a 39 year old female with NYHA
`Ill primary pulmonary hypertension who completed study with UT-15
`without incident. She remained in hospital for insertion of a central
`venous catheter and was discharged on Flolan 4 ng/ kg/ min. She was
`readmitted the following day (2 days after the last dose of UT-IS) with
`syncope and dizziness, seized, had a brady-asystolic arrest, and died
`the same day despite CPR efforts. Death was attributed to
`decompensated cor pulmonale. The investigator and sponsor regard
`the death as not reasonably attributable to UT-15.
`
`In studies P01:04 and P01:05, there were 9 deaths (3.8%) on UT-15 and 10 deaths
`(4.3%) on placebo.
`
`Deaths on UT—lS are summarized in the paragraphs below:
`
`Study P01:04 subject 004017 was a 32 year old, 70-kg female with a
`1-month history of PPH, NYHA III at baseline. She was having
`injection site pain and inject site reaction throughout treatment. She
`developed right heart failure on day 21 (UT-15 4 ng/ kg/ min). She
`was hospitalized and died, 3 days after decreasing the dose of study
`drug. Events were attributed to underlying disease.
`
`Study P01:04 subject 009006 was a 29—year—old, 55—kg female with
`congenital atrial septal defect, NYHA III at baseline. She had a stroke
`(pontine infarct with bilateral loss of vision) on day 2, several hours
`after Swan-Ganz removal. She received UT—15 1 mg] kg/ min for about
`9 hours. Thrombotic stroke was diagnosed by MRI and cerebral
`angiography, and she was treated with thrombolysis, but died 3 days
`after study drug administration. Events were attributed to
`.catheterization and not to study drug.
`
`Study P01:04 subject 010002 was a 40-year—old, 114~kg female with
`mixed connective tissue disease, NYHA IV at baseline. She was
`hospitalized on day 57 of treatment (UT-15 7.5 ng/kg/ min) with right
`
`5 CRF is ambiguous as to whether this subject disoontimted for backache.
`
`7 CRF was not provided.
`
`G:\N2 1 2 72. doc
`
`——1 7—-
`
`1 6:09 Friday, March 09, 2001
`
`Last saved
`
`

`

`Integrated review of safety
`
`,
`
`5
`
`7
`
`NBA 21-2 72
`UT-I 5 for pulmonary hypertension
`
`-
`
`0
`
`0
`
`0
`
`0
`
`0
`
`heart failure, and treated with diuretics, inotropes, and increased UT-
`15 (12 ng/ kg/ min). On day 81 (apparently still on UT-15), she
`developed ventricular tachycardia that did not respond to
`resuscitation.
`
`Study P01:04 subject 023002 was a 15-year-old 40-kg female status
`post repair of ventricular septal defect, NYHA II at baseline. On day
`52 of treatment with UT—lS 4 ng/ kg/ min, she had respiratory arrest
`at home and died in the ER despite resuscitation attempts. Death
`was attributed to her underlying condition.
`
`Study P01:05 subject 004503 was a 36 year old, 49-kg Hispanic
`female with a 3-year history of primary pulmonary hypertension,
`NYHA III at baseline. She discontinued UT-15 5 ng/ kg/ min on day 49
`to undergo medical termination of pregnancy. Her hospital course
`was complicated by low cardiac output, disseminated intravascular
`coagulopathy, oliguria, hypoxemia, and sepsis. She died 7 days after
`discontinuing UT-15. Death was attributed to sepsis.
`
`Study P01:05 subject 051007 was a 28 year old 58-kg Caucasian
`female with recent onset of primary pulmonary hypertension, NYHA II
`at baseline. At her week-12 cardiac catheterization (day 86 on UT-15
`20 ng/ kg/ min), she had vasovagal syncope, followed by complete A—V
`block, electromechanical dissociation, and death.
`
`Study P01:05 subject 054005 was a 20 year old 40-kg Caucasian
`female with congenital left-to-right shunt, NYHA III at baseline. She
`was hospitalized on day 43 of treatment (UT-15 2.5 ng/ kg/ min) for
`worsening hemodynamics and hypoxemia, thought to be pulmonary
`embolus. She was treated with vasodilators and anticoagulation, but
`died 8 days after admission. It is unclear how long previously UT-15
`was stopped. Death was attributed to pulmonary embolus and
`underlying condition and not to study drug.
`
`Study P01:05 subject 055005 was a 32 year old 87-kg female with a
`9-month history of primary pulmonary hypertension, NYHA IV at
`baseline. She was receiving UT-15 1.25 mg] kg/ min up to day 6, when .
`she was hospitalized with recurrent syncope, chest pain, and
`hypotension. Myocardial infarction was suspected, but she developed
`intractable ventricular fibrillation prior to angiography. The
`investigator and sponsor disagree on the possible role of UT-15 in
`these events.
`
`Study P01:05 subject 058001 was a 39 year old, 83-kg Caucasian
`male with a 15-month history of primary pulmonary hypertension,
`NYHA III at baseline. He was receiving UT—lS 1.5 mg] kg/ min on day
`17 when he was admitted to hospital for hemoptysis, hyponatremia,
`and hyperkalemia. Pulmonary embolus was suspected. Renal and
`respiratory function declined and he died on day 7 of admission
`(unclear whether still on UT—lS) with bradycardia followed by cardiac
`arrest. These events were not attributed to UT-15.
`
`Deaths on placebo are summarized in the paragraphs below.
`
`0
`
`Study P01:O4 subject 009012 was a 56 year old 80—kg Caucasian
`male with a 2-year history of primary pulmonary hypertension, NYHA
`IV at baseline. He died on day 9, one day after being admitted to
`
`G:\N2 1 2 72. doc
`
`—1 8—-
`
`1 6:09 Friday, March 09, 2001
`
`Last saved
`
`

`

`Integrated review of safety
`
`_
`
`,
`
`NDA 21-272
`UT~1 5 for pulmonary hypertension
`
`0
`
`0
`
`0
`
`0
`
`0
`
`0
`
`0
`
`0
`
`0
`
`G:\N21272.doc
`
`hospital for right heart failure. Events not considered related to study
`drug.
`
`Study P01204 subject 010001 Was a 65 year old 80—kg male with a 1—
`year history of mixed connective tissue disease, NYHA W at baseline.
`On day 36, he was hospitalized with hematemesis and acute
`respiratory distress. Six days later, he developed bradycardia and
`cardiac arrest from which he could not be resuscitated. Events not
`
`considered related to study drug.
`
`Study P01:04 subject 015003 was a 23 year old 60—kg Caucasian
`female with a 3—year history of systemic lupus erythemetosis and ,
`mixed connective tissue disease, NYHA III at baseline. She was
`hospitalized on day 46 for dyspnea and hypoxemia attributed to
`right-to-left shunting through a patent foramen ovale. She arrested
`during a pericardiocentesis procedure and did not survive. Events
`were not attributed to study drug.
`
`Study P01:04 subject 016003 was a 67 year old 54-kg female