Clinical Review
`
`M'wango A. Kashoki, MD, MPH
`N 2|-897
`Medisorb Naltrexone
`
`7.1.3.3 Other significant adverse events
`
`Due to the antagonist effects of naltrexone at the mu opioid receptor, there is a potential risk for
`dysphoria and other mood changes, as well as subsequent suicide in naltrexone-treated patients.
`In a placebo——controlled study in which patients with obesity were randomized to treatment with
`placebo or oral naltrexone 300 mg/day, 19% of patients in the oral naltrexone group developed
`non-serious elevations1n serum transaminases after 3— 8 weeks oftreatment, compared to 0% of
`
`placebo-treated patients.
`
`The Medisorb Naltrexone database suggested that, as a result ofthe route of administration, the
`drug may be associated with injection site reactions. Descriptions of the types of reactions were
`consistent with an inflammatory type response. The safety database also contained information
`regarding elevations in eosinophil count, a case of eosinophilic pneumonia, and reports of rash,
`urticaria, and angioedema. Together, the adverse events were suggestive of an allergic reaction
`following study treatment.
`
`Therefore the safety database was assessed for evidence of increased risk of suicide, elevated
`LFTs, and allergic reactions in Medisorb Naltrexone patients.
`In addition, the specific type and
`severity of injection site reactions were evaluated.
`
`7. 1.3. 3. 1 HEP/l TOTOXICIT)’
`
`Alkermes is ofthe opinion that the risk of hepatocellular injury from Medisorb Naltrexone is
`considerably lower than that of oral naltrexone because:
`0 Whereas the total monthly dose at which hepatotoxicity was observed with oral
`naltrexone (300 mg/day) would be 8400 mg, the total monthly dose of Medisorb
`Naltrexone is 380 mg (which is 22—fold lower than the total monthly dose of oral
`naltrexone).
`0 Administration of Medisorb Naltrexone suspension by gluteal 1M injection avoids first—
`pass hepatic metabolism
`- Medisorb Naltrexone will be dispensed1n single--dose kits and will be administered by a
`health care provider, reducing the risk of patient overdose.
`
`Upon review of the safety data, Alkermes found that in alcohol-dependent subjects treated with
`Medisorb Naltrexone suspension the hepatic safety profile was indistinguishable from that of
`placebo, with assessments ofALT, AST, GGT, and total bilirubin. Also, single——dose 1M
`administration of relatively high doses (141,269,530 and 784 mg) of Medisorb Naltrexone to
`healthy subjects did not result1n hepatotoxicity.
`
`Using the datasets iss-ae_3.xpt and iss-labs.xpt I calculated the number of patients who
`experienced a liver-related adverse event.
`I searched iss—ae_3.xpt for all terms consistent with
`hepatocellular injury, first by system organ class (gastrointestinal disorders, hepatobiliary
`disorders, and investigations), and then by specific terms including: AST, ALT, GGT, ALP,
`
`87
`
`

`

`Clinical Review
`
`Mwango A. Kashoki, MD. MPH
`N 21-897
`Medisorb Naltrexone
`
`SGPT, SGOT, LFT, liver, liver function, liver enzyme, laboratory test, aspartate, alanine,
`bilirubin, alk phos, glutamyltransferase, hepatitis, and jaundice.
`
`I then searched the iss—labs.xpt dataset using the variable “LBNAME,” under which were coded
`tests for AST (SGOT), ALT (SGPT), alanine transaminase, aspartate transaminase, total
`In
`bilirubin, and total bilirubin (mg/dL). These categories were collapsed for ease of analysis.
`addition, I used the variable “LBFLAG” to identify all patients with high values for any of the
`lab tests. This subset of patients was merged with‘the subset from the iss—ae.xpt dataset for
`comparison of events between active— and placebo—treated patients.
`
`7.1.3.3.].1 Reviewer’s Analysis: Patients with Hepatic—Related Adverse Events
`
`7.1 .3.3. 1.1.1 Hepatic-related events — Studies of 4—6 month’s exposure
`
`Based on the iss-aejxpt dataset, 4.8% of all patients (52/1090) in the 4—6 month trials
`experienced adverse events suggestive of hepatocellular injury. Most patients reported more
`than one type of hepatic—related injury. Only 1 of these AEs was considered serious (Subject
`ALK21003-230-024, 190-mg, cholelithiasis) but, based on my review ofthe patient narrative,
`was not related to study treatment.
`.
`
`The frequency of hepatic—related AEs in the combined Medisorb Naltrexone subset was 4.6%
`(37/81 1). This was lower than the frequency in the placebo group (5.6% (12/124)), and was
`comparable to the frequency in the oral naltrexone group (4.6% (3/65)). The risk of hepatic—
`related AEs did not appear to increase with increasing doses of Medisorb Naltrexone: 4.8%
`(10/210) ofthe 190-mg patients, 4.5% (26/576) ofthe 380-mg patients, and 4% (1/25) ofthe
`400-mg patients.
`
`Listed in Table 7.1.3.3.1 .1 (below) are the types ofevents that were hepatic—related.
`
`Liver function test (LFT) abnormalities/increases were the most commonly occurring. Ofthe
`1090 patients with 4—6 months of drug exposure, 67 patients (6.1%) had an elevation of AST,
`ALT, GGT, bilirubin, or alkaline phosphatase. The most frequent type of LFT abnormality was
`an elevation in GGT, however more placebo patients than naltrexone patients reported this AE
`(3.7% of placebo patients vs. 1.6% of Medisorb Naltrexone and 0% of oral naltrexone patients).
`There was no evidence ofa dose response of GGT elevation among the Medisorb Naltrexone
`groups. The greater proportion of placebo patients with high GGT levels compared to Medisorb
`Naltrexone patients likely reflects a higher frequency of drinking in the placebo population than
`in the Medisorb Naltrexone patients.
`
`AST increases were the next most common LFT change, with slightly more patients in the oral
`naltrexone and combined Medisorb Naltrexone groups (~ 1.5% each) experiencing this AE than
`in the placebo group (0.9%). With respect to ALT, patients in the combined 380/400-mg
`Medisorb Naltrexone group and the oral naltrexone groups were more likely to report ALT
`increases (1 .5% and 1.2%, respectively) than placebo patients (0.9%).
`
`88
`
`

`

`Clinical Review
`
`Mwango A. Kashoki, MD, MPH
`N 21-897
`Medisorb Naltrexone
`
`Five patients were reported as having hepatitis (1 case of alcoholic hepatitis, and 3 cases of
`hepatitis C and 1 case of“hepatitis NOS. ”)6. Two cases occurred during ALKZl- 003, and the
`remaining three during ALK21-006. None of the cases was considered serious.
`1 reviewed the
`patient narratives and CRFs to evaluate for a relationship to study treatment. The narratives
`showed that patients experienced elevated LFTs either in the context of increased drinking or
`hepatitis C diagnosis. These factors make it difficult to ascertain whether Medisorb Naltrexone
`is associated with hepatitis, or whether it increases the risk of hepatitis in patients with
`predisposing factors.
`
`REVIEWER COMMENT:
`
`Overall, therefore, the data from the trials of4-6 months’ duration suggest that treatment with
`Medisorb Naltrexone is associated with a slightly increased risk of AST and ALT compared
`to placebo, however this risk is similar to that associated with oral naltrexonc therapy.
`Therefore, Medisorb Naltrexone does not appear to offer a safety advantage— with respect to
`hepatotoxicity— over oral naltrexone. The risk of hepatitis following treatment with
`Medisorb Naltrexone appears to be low.
`
`APPEARS THlS WAY
`0N ORlGlNAL
`
`6 Subject ALK2|006-247-006: alcoholic hepatitis, Subject ALK21006-255-004. hepatitis NOS; Subjects
`- ALK21003- 2l6— 005, ALK21003-216- 012, and ALK2l006- 252- 009: hepatitis C.
`
`89
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`00
`
`
`
`
`
`
`
`
`e\aZo\=e\aZ.x.Ze\oZ¢\=Ze\oZBenzman,3ng3an
`
`ENnzmunz=<”a8:3».as2:.wecanwe:3.22:nzencour—XHZ=20gash—«Z“28:5238.35=<miniall00m
`3.0H0.2_0N.0NN0_0.00E.003.0_5.0002522350»:
`3.0N0.2H2..oN;N.00.0nmodN0:N0conga$34
`3.0N0.2.2o00.0m00.0_0:2088550.12
`
`
`3.0N0.2mg000.0w91Lmmm;C08855Hm<.
`
`
`00.0000.00N_.0_N.0_00.002.0_00.0000.0mOZmacaw:can53.0
`05mm00.0mg000.00mm;m00.008.0_N088550.00524mcosmwflmoi:
`
`
`00.0000.002.00N0_00.00E.0000.0000.0Hmagnum3:02:30:53:08.30:
`
`00.0000.003.0m|_..m0N00.00mm.0N01.0_mN.0mUEESQIIELS0:80:2558050,.umZ
`
`
`00.0000.00.nN.0NN.0_00.00E.0_00.0_E0NmOZmoiv:<usmmrrIL
`
`N.33.00.00mNNmMN.0mmNu.0N0..2.0N000053:?.53A5.
`
`0.000.00N.00..0.0000.00..Lllmv0N$00MEEREAS»..I.
`
`
`
`
` |_|253805
`
`
`
`
`
`
`
`
`
`
`
`'gmlll
`
`sesame
`
`:EEEm3E
`
`08855
`
`mOZSmog
`
`38854.
`
`00322:.0835
`
`.mombmcu
`
`
`
`DMZ
`
`
`
`
`2530:van
`
`
`
`ozosoofi338%a£00820
`
`52.0w
`
`2058.0595
`
`2.233%:
`
`
`
`
`
`
`
`
`moan—um5:9:0IvI05.2:hag—.8033:mauzwmowwsmmm<"m_mbw:<mnu032>om
`
`
`
`
`
`
`
`000-325:2.92.Eofimz.<$5.32
`
`
`
`ocoxobfizDEERE
`
`325m3256
`
`
`
`H.fi.H.m.m.~.hBeak.
`
`
`
`
`
`
`
`

`

`
`
`
`
`3
`
`:NHZ332m
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`235m_:._.mafia:
`
`252$20
`
`
`
`
`
`fix..x.,5
`
`
`
`
`
`E0_00.0000.000.0000.0000.0000.000.0£23305in32.93.,”€239“ng
`
`$20..NEE-bumNseflmE
`
`
`
`00.0000.00~50\0.0000.0000.00SW0\00.0EESNEBFC32.380050\DENSEaEmE
`
`
`00.0.000.00mN.0bmonofiquE©3333an33:83.3:
`fiz:20822:228:52=<
`menz=<8&2;2;we.3».as:338.3.5
`
`
`
`:mSwnzmNuzEmuzczuzoaSuz
`
`
`
`583%cosmfimeo
`
`mEBmE>m
`
`
`
`
`
`vanmswmmEDEOEU
`
`3:682
`
`EEEEEU
`
`3323330.3
`
`H44.
`
`
`
`findsQENH
`
`
`
`
`
`
`
`
`
`Avoszficoov3:555:2:0IvIi=7:523383.00:mo«Zamowmamm<mama—andwmLoBqum
`
`
`
`
`
`
`
`howLNZE:.92.3232.<owcmaz
`
`0:985?Z£8622
`
`325m356
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`

`

`Clinical Review
`
`Mwango A. Kashoki, MD, MPH
`N 21-897
`Medisorb Naltrexone _
`
`. 7.1 3.3112 Hepatic-related events — Studies of > 6-months’s duration
`The iss_ae_3.xpt dataset showed that, among the 572 patients included in this subset of studies,
`44 patients (7.9%) reported 54 A135 suggestive of hepatocellular injury, 4 ofwhich were labeled
`as “serious.” I reviewed the patient narratives and considered only 1 of the “serious” cases to be
`suggestive ofa relationship to Medisorb Naltrexone (subject ALK21003-210—004: acute
`hepatitis)7. This case has previously been discussed in Section 7.1.2.2.4 and the narrative is
`located in the Appendix.
`'
`
`The types and number of hepatic—related AEs in the “> 6-months’ exposure” categories of studies
`are listed in the table below.
`
`/
`
`Table 7.1.3.3.1.1.2: Reviewer’s Analysis: Hepatic—related AEs — studies 0f> 6 months’ duration
`
`Medisorb Valtrexone
`Oral NTX
`
`50 mg
`380—mg
`190- mg
`Adverse Event/Preferred
`
`N = 157
`N = 175
`N = 36
`Term
`System Organ class
`
`%
`%
`%
`N
`N
`5.74
`11
`6.28
`2
`5.56_l
`LLFT abnormalities — Total
`ALT increased
`1.91
`1
`0.57
`0
`0.00
`0.64
`2
`1.14
`0
`0.00
`AST increased
`
`L1_nvestigations
`
` o—mxoz
`
`
`Blood alkaline phosphatase
`NOS increased
`
`0.00
`
`l
`
`0.57
`
`0
`
`0.00
`
`
`0.00
`
`
`
`
`
`
`
`
`2.78
`1
`0.57
`1
`1.91
`3
`Blood bilirubin increased
`GGT increased
`1
`0.64
`1
`0.57
`l
`2.78
`
`Laboratory test abnormal NOS
`0
`0.00 I—1
`0.57
`0
`0.00—l
`LFTs NOS abnormal
`1
`0.64
`4
`2.29
`0
`
`
`I
`
`
`
`.
`
`l-lepatomegaly; Liver palpable
`l-lepatobiliary disorders,
`Investigations
`subcostal
`
`2.78
`Gastrointestinal disorders Abdominal pain NOS;
`10
`6.37
`10
`5.72
`l
`
`abdominal pain upper
`
`2
`
`1.28
`
`1
`
`0.57
`
`O
`
`0.0
`
`Hepatobiliary disorders
`
`0.0
`0
`0.0
`1.28 _l—0
`2
`Cholecystitis acute NOS;
`Cholelithiasis
`
` Hepatobiliary disorders
`"
`1
`I
`0.64—l
`0
`l
`0.00
`0
`0.00
`
`The table shows that LFT abnormalities were the most frequently reported AB in this category of
`studies that was suggestive of hepatocellular injury. Slightly more patients treated with
`Medisorb Naltrexone 380-mg (6.3%) reported LFT abnormalities than did patients in the 190-mg
`
`
`7 The other 3_ SAEs were not considered related to study treatment and were as follows: Subject ALK21003-230-
`018 — abdominal pain; Subject ALK21003-209—031 — Cholecystitis, elevated LFTs; Subject ALK21003-224—015 —
`abdominal pain.
`
`92
`
`

`

`Clinical Review
`
`Mwango A. Kashoki, MD, MPH
`N 21-897
`Medisorb Naltrexone
`
`group (5.7%) and the oral naltrexone group (5.6%). Patients administered Medisorb Naltrexone
`were also more likely to report abdominal pain. The sole case of hepatitis that was likely to be
`treatment related occurred in a patient in the l90—mg arm.
`
`REVIEWER COMMENTS
`
`The information regarding hepatic—related AEs that is available from the open label studies is
`consistent with the information from the blinded/controlled trials: treatment with Medisorb
`Naltrexone, particularly the higher doses, is associated with a slightly increased risk of LFT
`abnormalities. Again, however, no safety advantage ofMedisorb Naltrexone over oral
`naltrexone was demonstrated.
`
`7.1.3.3.].2 Reviewer’s Analysis: Changes in Liver Function Tests
`
`7.1.3.3.].3 LFT changes — Studies of 4—6 month’s duration
`
`This category oftrials includes Study ALK21-002 which, as has been previously noted, was a 4-
`month study in patients with alcohol dependence. The other two trials in the category (ALK21—
`003 and —006) comprise at least 6 months of data. Given the shorter exposure duration in Study -
`002, as well as the relatively few number of Medisorb Naltrexone—treated patients (n = 25), l
`excluded this study from the analysis of the change in LFT values from baseline to Week
`24/Month 6.
`
`Laboratory data were contained in the dataset issglbhzxpt. Using this dataset, I calculated the
`mean LFT values at Baseline and Week 24 (Appendix Table 10.7.a). Increases in LFT values.
`were ofinterest, as these would indicate possible hepatocellular injury.
`
`Only the patients in the oral naltrexone group showed a small increase in the mean AST at Week
`24. The mean AST value for the other treatment groups either remained relatively unchanged or
`was slightly decreased. None ofthe treatment groups showed a numerical increase in the mean
`ALT, GGT, or total bilirubin levels at Week 24.
`
`I also evaluated whether there was a significant difference in the within-group mean LFT values
`from Baseline to Week 24, and ifthere was a difference between treatment groups (active vs.
`placebo) with respect to the mean LFT values at Week 24 (Appendix Table 10.8.b). The
`Medisorb Naltrexone l90—mg group showed statistically significant (i.e. p-value < 0.05)
`decreases in mean AST, ALT, and GGT levels from baseline to Week 24. Similarly, there was a
`statistically significant decrease from baseline for the Medisorb Naltrexone 380-mg and oral
`naltrexone groups with respect to the Week 24 GGT values. There were also statistically
`significant decreases in the 380—m g and placebo groups with respect to the baseline Week 24
`bilirubin values. However, fairly wide confidence intervals were associated with these values,
`making the finding ofa low p—value less relevant.
`
`1 evaluated the between-group dgfferences (i.e. between active and placebo groups) in mean LFT
`values at Week 24 (Appendix Table 10.8.b). The 380-mg and oral naltrexone groups showed
`statistically significantly greater increases in the Week 24 mean GGT values compared to the
`
`93
`
`

`

`Clinical Review
`
`Mwango A. Kashoki. MD, MPH
`N 21-897
`Medisorb Naltrexone
`
`_ placebo group. Again, however, the confidence intervals for these differences were considerably
`wide. The 380—mg group’s mean Week 24 bilirubin value was also slightly larger than the
`placebo group (mean difference = 0.07 mg/dL) and this finding was statistically significant.
`Otherwise, the Week 24 values ofAST and ALT for either of the Medisorb Naltrexone groups,
`while greater numerically, were not statistically different from those ofthe placebo group.
`
`REVIEWER COMMENT:
`
`Although confidence intervals and p—values were calculated, it is important to note that
`the analyses do not take into account that the study was powered on the basis of the-
`primary efficacy endpoint, and not the safety analyses, or that there is an issue of
`multiplicity.
`
`Nevertheless, the data suggest that treatment with Medisorb Naltrexone only slightly
`increases mean LFT values by the end of treatment. The end—of-treatment mean LFT
`values are not much greater than those for the placebo group, suggesting no considerably
`increased risk of hepatocellular injury compared. Also, the mean LFT values suggest no
`difference in hepatic effect of Medisorb Naltrexone versus oral naltrexone.
`
`Finally, I evaluated the Applicant’s shift tables to determine whether more Medisorb Naltrexone
`patients had changes in LFT values from normal to abnormal, or from abnormal to even more
`abnormal (Appendix Tables 107.0 to 1‘). Higher proportions of patients in the Medisorb
`Naltrexone 380-mg group (14%) and the oral naltrexone group (10%) had a shift in ALT value
`from normal to the high limit of normal than did placebo patients (7%) or the 190-mg group
`(6%). Additionally, 3% of patients in the oral naltrexone group had a shift in AST from the high
`limit of normal to 3x the upper limit of normal, compared to 0% in the other groups. Otherwise,
`there no shifts from normal to abnormal in the active groups that were considerably different
`from the shifts observed in the placebo group.
`
`REVIEWER COMMENT:
`
`Overall, therefore, the laboratory data from the 4-6 month category studies show that
`treatment with naltrexone, whether oral or Medisorb naltrexone, can cause slight increases in
`transami-nases, particularly ALT. However, the increases are not considerably greater than
`those observed with placebo treatment. Also, the effects of Medisorb Naltrexone therapy are
`no better or worse than the effects of oral naltrexone therapy.
`
`7.1.3.3.].4 LFT changes — > 6 month studies
`This category of studies comprised ALK21-006, ALK21-003-EXT, and ALKZl-OIO. Since
`study ALK21—006 was ongoing at the time of data cutoff, the number of subjects overall and in
`any given group at the later time points becomes quite small. Therefore the composite data are
`subject to fluctuations due events in individual subjects.
`
`94
`
`

`

`Clinical Review
`
`Mwango A. Kashoki, MD, MPH
`N 2l-897
`Medisorb Naltrexone
`
`At the data cutoff date, the maximum duration of exposure to Medisorb Naltrexone was 30
`months: 1 patient in the 380—mg group, and 3 patients in the l90-mg group. However, Alkermes
`provided information on for the changes in LFTs for only 2 years of study treatment. For
`patients in ALK2l—003EXT and —010, “the 2 year period of observation includes the 6 months of
`exposure in ALK21-003.
`
`Alkermes evaluated the mean LFT values over time (i.e. every 4 weeks), the mean changes in
`LFT values over time, and the proportions of patients who shifted from normal to abnormal (or
`abnormal to more abnormal) over time (Applicant’s SCS, Appendix Tables 2.7.4.43 through
`2.7.4.45, P. 271-375). For subjects in ALK2l—006, baseline considered to be the last pre-dose
`test value. For subjects in ALK2l—003EXT or ALK2l—010, baseline was the initial dose in study
`ALK21—003.
`
`With respect to ALT, the median values at 24 weeks were somewhat higher in subjects in
`ALK21—003EXT or ALK21-010 who has previously been treated with placebo than in subjects
`who had been given active drug (either Medisorb Naltrexone or oral naltrexone). Abnormal
`ALT values were noted among all groups over time, but there was no shift towards a higher
`proportion of abnormal values, either any value above upper limit of normal or over 3xULN,
`with time. Thus, there did not appear to be a trend towards increasing ALT abnormalities with
`continued Medisorb Naltrexone treatment.
`
`At the end of the first 24 weeks oftreatment, mean and median GGT values were lower among
`subjects in the Medisorb Naltrexone 380-mg and oral naltrexone groups, compared with those in
`the placebo and Medisorb Naltrexone l90—mg. As was observed with the ALT values, GGT
`abnormalities were noted among all groups and at all time points, but there was no increase in
`the proportion of abnormal values over time.
`'
`
`There were no considerable increases in AST or bilirubin for any ofthe treatment groups from
`baseline to Week 100.
`
`REVIEWER COMMENT: The data from the studies of > 6 months’ exposure therefore suggest
`that prolonged treatment with Medisorb Naltrexone does not confer an increased risk of
`abnormalities in tests of liver function.
`
`7.1.3.4 Suicide
`
`Patients with a history of substance abuse, including alcohol abuse/dependence, are at risk for
`suicide, whether or not they also have a history of depression.
`In addition, as a mu opioid
`antagonist, naltrexone may cause dysphoria and other mood changes that put patients at risk for
`suicide.
`In previous studies in patients with alcohol and narcotic dependence, depression,
`suicidal ideation, and suicide attempts have been reported in those treated with oral naltrexone.
`Therefore, the NDA data were reviewed to evaluate whether treatment with Medisorb
`Naltrexone was associated with an increased risk of these events.
`
`95
`
`

`

`Clinical Review
`
`Mwango A. Kashoki, MD, MPH
`N 21-897
`Medisorb Naltrexone
`
`7. 1 . 3. 4. 1 APPLICANT ’3 EVA L UA T/0N 0F SUICIDE-REM TED E VENTS
`
`To identify potential suicide cases, Alkermes reviewed all verbatim terms that suggestive of
`suicidal behavior, suicidal ideation, suicide attempt, or completed suicide. The Applicant then
`calculated the proportion of patients in each treatment group who experienced these events, for
`both the 4—6 month studies and the > 6 month studies.
`
`Alkermes found that, among the 1090 subjects who participated in the Medisorb Naltrexone
`istudies with 4 to 6 months of exposure, 16 subjects (1.5%) experienced a suicide—related AE.
`In
`One subject committed suicide (the suicide occurred after completion of study participation).
`addition, 2 subjects attempted suicide, 1 made a suicidal gesture, and 12 (1%) reported suicidal
`ideation.
`
`In study ALK21—003, Alkermes counted 3 suicide-related events in the Medisorb Naltrexone
`group (1% of subjects), and none in the placebo 380—mg groups.
`In the first 6 months of study
`ALK21-006, of the 10 suicide-related events that occurred in the Medisorb Naltrexone 380 group
`(3% of that group), 7 occurred in subjects with alcohol dependence (3% of that group) and 3
`events occurred in subjects with opiate/mixed dependence (3% of that group). Three subjects
`(5%; 2 with alcohol dependence and 1 with opiate/mixed dependence) who received oral
`naltrexone in study ALK21—006 had suicide—related AEs.
`
`,
`
`In studies longer than 6 months, as ofthe data cutoff date, Alkermes identified a total of3
`subjects who reported suicidal ideation: 2 participants in the extension studies to ALK21—003
`(both at the 190 mg dose level), and 1 subject in ALK21-006 (380 mg group) with mixed
`opiate/alcohol dependence. There were no suicide attempts or completed suicides after 6 months
`of treatment.
`’
`
`Alkermes noted that most of the suicide—related events reported during the Medisorb Naltrexone
`clinical program occurred in the context of either depressive symptoms or further substance
`abuse (either alcohol or other drugs).
`
`7. 1. 3. 4.2 REVIEWER ’s E VALUA TION OF SUICIDE-REM TED EVENTS
`
`7.1.3.4.2.1 Suicide-Related Events — Studies of 4 t0 6 months’ exposure
`Using the iss_ae~3.xpt dataset, I attempted to identify all cases consistent with suicide
`(completed suicide, attempted suicide, or suicidal ideation).
`I selected for AEs grouped under
`the SOC terms “psychiatric disorders” and “injury, poisoning, and procedural complications.” 1
`then selected cases using the following HLGTs: injuries; injuries by physical agents; chemical
`injury, overdose and poisoning; psychiatric disorder NEC; psychiatric and behavioral symptoms
`NBC; and suicidal and self-injurious behaviors NBC.
`1 narrowed the cases further by then
`selecting for LLTs that were consistent with suicide e.g. accidental ingestion; deliberate self-
`harm; injury; intentional overdose; and overdose. Patients with an AB of “depression” were
`excluded, unless they also had a suicide-related AB.
`
`96
`
`

`

`Clinical Review
`
`Mwango A. Kashoki, MD, MPH
`N 21-897
`Medisorb Naltrexone
`
`Based on this process, 1 identified 19 patients (19/1090, 1.7%) who had 20 events suggestive of
`suicide in the studies of 4-6 months7 exposure, 8 of whom experienced serious suicide-related
`AEs (see Section 7.1.2.2.3.l).
`
`As shown in Table 7.1.3.4.2.1, suicidal ideation was the most common suicide—related AE,
`followed by suicide attempt. The proportion of patients who reported suicidal ideation and
`suicide attempt was highest in the oral naltrexone group (3.1% and 1.5%, respectively).
`In
`comparison, 1.4% of patients in the Medisorb Naltrexone 380-mg group reported suicidal
`ideation, and 0.9% reported a suicide attempt. Slightly fewer patients (1.0%) in the 190—mg
`group experienced suicidal ideation, and none reported suicide attempts. There were no patients
`in the placebo group who had either suicidal ideation or attempt.
`
`The patient narratives ofthe suicide—related events that 1 considered to possibly drug related are
`provided in the Appendix.
`
`REVIEWER COMMENT:
`
`The table suggests that patients treated with Medisorb Naltrexone 380-m g and oral
`naltrexone are at greater risk of suicidal ideation or suicidal attempts than are patients treated
`with Medisorb Naltrexone 190-mg or placebo. The data also suggest that the risk of these
`events is less for patients in the 380—mg group than for those in the oral naltrexone group.
`
`The patient narratives show that the majority of subjects had a previous history of depression,
`anxiety, or suicide attempts/thoughts and that, with the exception of 3 cases, all events
`occurred more than 30 days since the last dose of study drug. This makes a causal
`relationship between treatment with study drug and suicide-related events less likely.
`Nevertheless, given that all of the cases occurred in patients treated with naltrexone (oral or
`depot formulations), that no placebo patients reported these adverse events, and that an
`association of this AE with naltrexone has been observed in previous trials, a relationship
`between suicide—related ABS and Medisorb Naltrexone cannot be ruled out.
`
`APPEARS T1113 WAY
`0N ORlGlNAL
`
`97
`
`

`

`mo
`
`
`
`o\=
`
`
`
`
`3%.2:...mafia:
`
`.EzE—mo20
`
`
`
`
`
`
`
`
`
`
`
`
`
`82:96Egan.20-0m~-000-§¥1~<“832mq:00-Nmm-000-fimv5<Sufism-m32:96:hmnmw400-wa-000-§¥4<Sufism
`
`Io
`
`
`
`
`
`
`
`
`
`90o8.00Ed08.0@228“Emacasuafiofi
`
`
`
`
`
`
` ENMZ33.“?—
`
`
`
`
`
`0.0000.00mm.0N00.0688%0225
`0.0.000.00mm.0N00.0mmOZ08996
`
`0.0o#2_8.0c8.0N9.37:3383225
`
`
`0.0000.005.0H00.0$85335203-82
`0.00E;0$00.0088-388%quSm053:0me
`
`8.5%02.5.35
`
`
`2:898.332:0%.3£15mI029:;mo358..3;mad.”mi
`
`
`
`
`:onmuEE2080muoEOmmuoEmonmm
`
`
`
`
`bu=<mkuokogom"fiNéfiAEczar—L
`
`chofloaRH:Eionaoflu
`
`r
`
`
`
`
`
`80-3Z52d2.Eofifi.<$532
`
`2855:“Z£80022
`
`325m3256
`
`
`
`
`
`
`
`0.0000.005.0H2:898Ecufloo<UnamEcoflom$5.0:
`
`
`
`
`
`€038:an03:689
`
`
`
`
`
`
`
`
`
`
`

`

`Clinical Review
`
`.
`
`Mwango A. Kashoki. MD, MPH
`N 21-897
`Medisorb Naltrexone
`
`7.1.3.422 Suicide-Related Events — Studies 0f> 6 months’ exposure
`
`Among the patients in this category of studies, I identified 7 who reported 8 suicide—related
`events (7/572, 12%). Three of the events were serious, and included a single event of completed
`suicide (see Sections 7.1 .l and 7.l.2.2.3.l for details). Of note, two ofthe events (suicidal
`ideation and completed suicide) occurred in a patient (subject ALK21003-224-012) who
`discontinued from the 6-month study ALK21—003 and died 73 days after his last dose.
`
`The suicide-related AEs are tabulated in the table below.
`
`I found that suicidal ideation was the
`
`most commonly reported event, occurring with the greatest frequency in patients treated with
`Medisorb Naltrexone 190—mg (2% vs. 0% in the other groups). The oral naltrexone group had
`the highest proportion of suicide attempts (3%) — however this percentage was represented by
`only 1 patient out of 36.
`
`
`Table 7.1.3.422: Reviewer’s Anal sis: Suicide related events —- studies 0f> 6 months’exposure
`
`
`
`
` Suicidal ideation
`Psychiatric disorders
`
`
`Completed suicide
`
`
`
`Overdose NOS
`I
`0
`Selfmutilation
`I
`O
`
`l
`l
`
`'
`
`I 0.26 I
`I 0.26
`I
`
`l
`0
`
`I 2.78
`I_0.00
`
`
`
`
`
`Injury, poisoning and
`procedural complications
`
`I 0.00 I
`I 0.00 I
`
`The narratives of the suicide-related events that I considered to be possibly related to study
`treatment are in the Appendix.
`
`REVIEWER COMMENT:
`
`The data from the studies with longer treatment exposure show that the frequency of suicide—
`related AEs was greater for the Medisorb Naltrexone l90—mg group. However, an
`association ofthese ABS with high-dose Medisorb Naltrexone cannot be ruled out since only
`interim data for two of the long-term studies (ALK21—006 and ALK21—010) were available at
`the time of data cut-off.
`'
`
`7.1.3.5 Allergic reactions
`
`As mentioned in Section 7.1.3.3, the Medisorb Naltrexone database suggested that the drug may
`be associated with injection site reactions (ISRs) ofthe inflammatory type. Also, the safety
`database described elevations in eosinophil count, a case of eosinophilic pneumonia, and reports
`of rash, urticaria, and angioedema. Together, the adverse events were suggestive of an allergic
`reaction following study treatment.
`
`1 evaluated the frequency and. severity oflSRs, pneumonia, eosinophilia, rash, urticaria and
`angioedema to determine whether they occurred in greater frequency among Medisorb
`Naltrexone patients than placebo or oral naltrexone patients. Also of concern was whether
`together, the demonstrated the allergenic potential of Medisorb Naltrexone.
`
`99
`
`

`

`Clinical Review
`
`Mwango A. Kashoki, MD, MPH
`N 21-897
`Medisorb Naltrexone
`
`7. 1.3. 5. 1
`
`INJECTION SITE REACTIONS (ISRs)
`
`To assess injection tolerability, injection sites were inspected by study personnel at least
`monthly. Information on ISRs was collected with other physical examination observations.
`Investigators were to use their own judgment when determining whether or not lSRs were
`clinically significant. Only those reactions deemed clinically significant were considered
`adverse events and were recorded on the Adverse Event sheet of the CRF.
`
`In the NDA submission, information on lSRs was contained in two datasets: the iss—z'srxpt and
`the z'ss.ae.xpt datasets. The iss-isr.xpt dataset comprised information an all lSRs, regardless of
`whether the investigator deemed them clinically significant. The issxaexpt dataset contained
`information on only those IS‘Rs that were severe enough to be recorded on the Adverse Event
`sheet of the CRF.
`'
`
`7.1.3.5.].1 Applicant’s Analysis of ISRs — Studies of 4-6 months’ exposure
`
`The Applicant used the iss—isr.xpt dataset to describe the frequency and severity of lSRs
`following participation in clinical trials. Alkermes found that during studies ALKZl-OOZ,
`ALK21-003, and ALK21-006, a total of Medisorb Naltrexone or placebo 4844 injections were
`administered. Approximately 55% of subjects receiving either placebo or Medisorb Naltrexone
`developed at least one ISR during the course of treatment. The percentage varied from 46% in
`the 2 mL placebo group to 88% in the 400 mg Medisorb Naltrexone group. The most common
`type of ISR observed was tenderness, followed by induration and pain (Table 7.1.3.5.].1). The
`specific type of ISR that occurred most frequently in the Medisorb Naltrexone patients (190— or
`380—mg) than in the placebo patients was injection site induration (25-29% vs. 9%).
`
`Following an analysis ofthe proportion oflSRs after each monthly injection, Alkermes
`concluded that across all treatment groups, the frequency of ISRs declined over the course of
`therapy (Figure 1). Alkermes calculated that, among subjects receiving 380- or 400-mg doses of
`Medisorb naltrexone, the incidence of any ISR decreased from 36% after the first injection to
`1 1% by the end of the study/data collection period.
`
`100
`
`

`

`
`
`
`
`3330mm9.2““Eamm95%2%ME
`
`
`
`
`
`AMQOINwmeamaonfimm4¢v
`
`
`
`X¢me>fiwwflmwnfiwfla
`
`
`
`
`
`wuuwflnwwunmvawawaadmwflwmumannexm‘wa”
`
`
`
`
`
`
`
`wew*mevmeWQQmamagwwmawpvencwan“;wuuwflnsmVuwnaax
`
`
`
`
`
`RMENzE:.02£053“.<09332
`
`
`
`0:32stnhommcuz
`
`325m3256
`
`
`
`2:893.32553hemacawIxi5?38.35no3:32:2aura:€=8=aa<ajwés03$
`
`
`
`
`
`
`
`
`
`nmawmeoUmamaflmmqmmoa-wmmqm.moolmmmammoaafimmaaMawMENmuomflnnmfiaowuumwm@uflwaowwoamcw
`
`
`
`
`
`
`
`
`
`
`
`2:
`
`2%wfifimafia:
`
`EEEmo2o
`
`
`
`
`
`Ahm”mmmAmp“maRom“MamAmwmeAmwVmwnwwVmwflmm«WmAwmmemmmeauwwgwwmuwmacwwmmnaw
`
`
`
`
`
`”mmvwnmhmwUmfirmvm¢Amm.vflw.mmme_mVamhmVmmevnqmnamumugaammafiaseguomflcH
`
`
`
`Am“mmAwvmmAWHwqmflawvmmamVa““wVmAm”mmxwmmmmwmsowaowmam
`
`
`
`Mm“mmn¢wadWuvwmA¢Va”m“m“mvmAmvmqmflmoamauum“maofluommxfi
`
`Awyamnvvmm“mm“mmAw“a“wV5”mmuflusmMmmewwuwflcm
`
`baawfiw”mmWmAmyam«mflvemAmWm“AmvmAm”wAmyam“mguo
`
`
`
`“hmufiemhawywmnmwvwmmHam“mg“Awmvmmw.mmvwm“awvww“mmvwemmmmmam
`
`m:0wuummmmuwmmawwuwh:m
`
`
`
`
`
`a8.muugszskasmgmaa$3.229¢$521?528$
`
`
`
`
`
`mmauwxmuomnasmmo“MWV.02
`
`
`
`Howmm“finmomcammo”mowmmamammowmuumwnumweManasz
`
`
`

`

`
`
`
`
`4..
`
`
`
`
`
`oEloII
`
` 9:03+
`aooémfifi9:ownI-¢-I
`68-556;weownII?I
`
`0587.-..-.
`
`
`
`
`
`
`omMm8«a.m0onaowMom....U+..
`
`
`EmcmmcmwVo
`
`o
`
`cor
`
`N2
`
`o?a?E3;5«Som
`
`oommmmmm
`
`S:v
`
`mm?
`
`KNN:
`
`ON
`
`mtwowmm
`
`m9Km:
`
`x:mm:
`
`com0mm
`
`Rvm
`
`.5852360
`
`3N
`
`oww
`
`mow
`
`Em
`
`mm
`
`”zomE
`
`“29:02
`
`uz9:8v
`
`“288;?39:0mm
`
`”2aooémxév9:8m
`
`
`
`noflnbflfificmS:333%,.5?»25:23..BE5500?:um_mbm:<mznmflaa<Qmun—WE
`
`
`
`
`
`3w;NZTEEn92.EofimM.<09832
`
`
`
`25%:sz£8622
`
`325m3255
`
`

`

`Clinical Review
`
`Mwango A. Kashoki. MD. MPH
`N 21—897
`Medisorb Naltrexone
`
`At the Division’s request, the Applicant calcu