Placing transdermal selegiline for major depressive disorder into clinical context: Number needed to treat, number needed to harm, and likelihood to be helped or harmed

Journal of Affective Disorders 151 (2013) 409–417

Contents lists available at ScienceDirect

Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad

Review

Placing transdermal selegiline for major depressive disorder into clinical context: Number needed to treat, number needed to harm, and likelihood to be helped or harmed Leslie Citrome a,n, Joseph F. Goldberg b, Kimberly Blanchard Portland c a

New York Medical College, Valhalla, NY, USA Icahn School of Medicine at Mount Sinai, New York, NY, USA c Mylan Specialty L.P., Basking Ridge, NJ, USA b

art ic l e i nf o

a b s t r a c t

Article history: Received 11 April 2013 Received in revised form 13 May 2013 Accepted 14 June 2013 Available online 24 July 2013

Background: This is a quantitative review of existing studies of transdermal selegiline for major depressive disorder. Methods: Data for dichotomous outcomes were extracted from the five 6–8 week studies of transdermal selegiline. Number needed to treat (NNT) vs. placebo was calculated for response and remission using standard definitions. Number needed to harm (NNH) vs. placebo for commonly encountered adverse events (AEs), AEs associated with sexual function, incidence of weight gain ≥5% from baseline, and discontinuation due to an AE, were also calculated. Data was pooled as appropriate and likelihood to be helped or harmed (LHH) ratios contrasting remission with selected tolerability outcomes were determined. Results: When pooling together the two pivotal trials as identified in product labeling, NNT for response was 11 (95% CI 6–109) and for remission, 9 (95% CI 6–28). Pooling all trials, NNH for application site reaction was 7 (95% CI 6–10) and for insomnia, 19 (95% CI 12–41). There were no clinically relevant differences from placebo regarding weight gain or impairment in sexual functioning. NNH for discontinuation due to an AE was 32 (95% CI 19–132). LHH for remission vs. discontinuation from treatment due to an AE was 3.6. Limitations: The studies included were not identical in design. The studies were registrational in nature and thus not necessarily generalizable. Conclusions: NNT for transdermal selegiline for efficacy is similar to that for other antidepressant regimens for which similar analyses have been published. There appear to be no clinically relevant effects of selegiline on weight or sexual functioning. & 2013 Elsevier B.V. All rights reserved.

Keywords: Evidence-based medicine Major depressive disorder Number needed to harm Number needed to treat Selegiline Transdermal

Contents 1. 2. 3.

4. 5. 6.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Pivotal clinical trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Efficacy outcomes: NNT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Safety outcomes: NNH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Discontinuation due to adverse events or lack of efficacy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Consideration of benefit vs. harm: LHH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

n Correspondence to: 11 Medical Park Drive, Suite 106, Pomona, NY 10970, USA. Tel.: +1 845 362 2081; fax: +1 845 362 8745. E-mail addresses: [email protected], [email protected] (L. Citrome).

0165-0327/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jad.2013.06.027

410 410 410 410 411 411 411 411 412 415 416

410

Role of funding source Conflict of interest. . . . Acknowledgment . . . . References . . . . . . . . . .

L. Citrome et al. / Journal of Affective Disorders 151 (2013) 409–417

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

1. Introduction Transdermal selegiline was approved in 2006 by the US Food and Drug Administration (FDA) for the treatment of major depressive disorder (US Food and Drug Administration, 2007). Selegiline itself is a monoamine oxidase inhibitor (MAOI), a class of antidepressant medications that is not currently used extensively because of dietary tyramine restrictions and potential drug-drug interactions. Transdermal selegiline differs from orally administered MAOIs, in that the transdermal formulation can deliver sustained selegiline blood concentrations sufficient to selectively inhibit MAO-A and MAO-B in the brain, producing antidepressant effects, without substantially inhibiting MAO-A in the gastrointestinal tract, thereby reducing the risk of hypertensive crisis with the ingestion of tyramine-rich foods (i.e. the so-called “cheese effect”) (Preskorn, 2006; Nandagopal and DelBello, 2009). Transdermal selegiline is dosed within the range of 6–12 mg/24 h, without the need for dietary precautions at the 6 mg/24 h dose. No cases of hypertensive crisis were reported in clinical trials, even without dietary restrictions. Additional details can be found in several reviews (Nandagopal and DelBello, 2009; Pae et al., 2007; Lee and Chen, 2007; Patkar et al., 2012). This report reviews the evidence-base for transdermal selegiline for the acute treatment of major depressive disorder (MDD) using the metrics of evidence-based medicine, namely number needed to treat (NNT), number needed to harm (NNH), and likelihood to be helped or harmed (LHH), in order to better place this intervention into clinical perspective. Similar work in MDD has been published for vilazodone (Citrome, 2012) and for adjunctive aripiprazole, olanzapine and quetiapine extended release (Citrome, 2010).

2. Methods All clinical trial reports and published papers for the similarly designed short-term double-blind trials of transdermal selegiline in adults with MDD were reviewed (Somerset Pharmaceuticals Inc., 2000a, 2000b, 2000c, 2000d, 2003); Bodkin and Amsterdam, 2002; Feiger et al., 2006; Amsterdam, 2003; Robinson et al., 2007), and reconciled with any revisions that occurred during the FDA’s regulatory approval process (US Food and Drug Administration. 2007). The following categorical data for the different efficacy and tolerability outcomes were extracted for each study arm: response, as defined by a 50% reduction on the Montgomery-Asberg Depression Rating Scale (MADRS) (Montgomery and Asberg, 1979) at last observation carried forward (LOCF) endpoint; remission, as defined by an endpoint rating scale score not exceeding 10 on the MADRS at LOCF endpoint; the proportion of subjects experiencing each of the most commonly encountered spontaneously reported adverse events (AEs)—the specific AEs includes all those that occurred in at least 5% of subjects (across all studies) receiving transdermal selegiline with a rate greater than that observed for placebo as noted in product labeling (Somerset Pharmaceuticals Inc., 2011): headache, diarrhea, insomnia, dry mouth, application site reaction; the proportion of subjects gaining at least 5% of their initial body weight; the proportion of subjects that have at least one AE related to sexual functioning (abnormal ejaculation,

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

416 416 416 416

decreased libido, impotence and anorgasmia in males, and decreased libido in females); the proportion of subjects that discontinued the clinical trial because of an AE; and the proportion of subjects that discontinued the clinical trial because of lack of efficacy. Where the data was not in either the clinical trial reports or published papers, the original locked datasets were queried. Data was pooled across studies as appropriate. Sensitivity analyses were done by excluding the data from the transdermal selegiline 3 mg/24 h study treatment arm from study E114 (Somerset Pharmaceuticals Inc., 2000d), as this dose is not commercially available. For the efficacy outcomes two different pooled populations were examined: the two identified pivotal short-term studies as noted in product labeling (Somerset Pharmaceuticals Inc., 2000a, 2003) were pooled for one analysis and all short-term studies, regardless of whether or not separation from placebo on the primary efficacy outcome was observed, were pooled for the second analysis. NNT and NNH, and their respective 95% CIs, were calculated for transdermal selegiline vs. placebo for the efficacy and tolerability outcomes listed above. The methodology of this technique is described in detail elsewhere (Citrome, 2008, 2009), but essentially NNT and NNH are measures of effect size and indicate how many patients would need to be treated with one agent instead of the comparator in order to encounter one additional outcome of interest. Lower NNTs are evidenced when there are large differences between the interventions in question. For example, a NNT of 2 would be a very large effect size, as a difference can be observed after treating 2 patients with one of the interventions versus the other. A NNT of 50 would mean little difference between the two interventions as it would take treating 50 patients to encounter a difference in outcome. NNH is used when referring to undesirable events. A useful medication is one with a low NNT and a high NNH when comparing it with another intervention. Kraemer and Kupfer (Kraemer and Kupfer, 2006) put forth that an NNT of 2.3, 3.6 and 8.9 correspond to a Cohen's d of 0.8, 0.5 and 0.2, respectively, representing effect sizes that are ‘large’, ‘medium’ and ‘small’. Likelihood to be helped or harmed was subsequently calculated on selected benefits and harms (Citrome, 2010, 2012; Citrome and Kantrowitz, 2008). Subgroup analyses were conducted to examine efficacy outcomes in the following subpopulations in the pivotal studies: subjects with a prior major depressive episode (i.e. recurrent episodes); subjects with a baseline MADRS score430 (indicating severe depression) (Nemeroff, 2007).

3. Results 3.1. Pivotal clinical trials There were 5 short-term double-blind, placebo-controlled, randomized controlled trials of transdermal selegiline in adult subjects with MDD. These are summarized in Table 1. The studies were designed in a similar fashion with all but one (Somerset Pharmaceuticals Inc., 2003) employing a fixed dose design, and all but one (Somerset Pharmaceuticals Inc., 2000a) were 8 weeks in duration. For all of the studies, the key entry criterion regarding severity of illness was a score of at least 20 on the 17-item Hamilton Depression Rating Scale (HAMD-17). All but one

L. Citrome et al. / Journal of Affective Disorders 151 (2013) 409–417

411

Table 1 Double-blind placebo-controlled trials of transdermal selegiline for adults with acute major depressive disorder. Study Length Dates identifier (weeks) conducted

N, randomized Selegiline dose (N)

N, placebo MADRS at baseline Comments

E106a P0052b P9804c E113d E114e

177 265 310f 297 446

88 133 157f 150 146

6 8 8 8 8

1/97–10/97 9/01–8/02 11/98–12/99 8/98–8/99 3/99–4/00

6 mg/24 h (89) 6-9-12 mg/24 h (132) 6 mg/24 h (153)f 6 mg/24 h (147) 3 (151) and 6 mg/24 h (149)

29–30 29 28 27 27–28

Pivotal study. Selegiline superior to placebo. Pivotal study. Selegiline superior to placebo. Selegiline superior to placebo on some outcomes. Selegiline no different from placebo. Selegiline no different from placebo.

MADRS-Montgomery Asberg Depression Rating Scale. a

Somerset Pharmaceuticals, 2000a. Somerset Pharmaceuticals, 2003. c Somerset Pharmaceuticals, 2000b. d Somerset Pharmaceuticals, 2000c. e Somerset Pharmaceuticals, 2000d. f 149 subjects in the selegiline arm and 152 subjects in the placebo arm included in the analysis after removal of 9 subjects from one site due to regulatory matters. b

(Somerset Pharmaceuticals Inc., 2003) study also included a 1-week open-label placebo lead-in period to eliminate placebo responders from being randomized if their symptoms improved by at least 20% on the HAMD-17 or if the score dropped below 20. Two of the studies (Somerset Pharmaceuticals Inc., 2000a, 2003) were considered pivotal by the FDA (US Food and Drug Administration. 2007) and are thus summarized in product labeling (Somerset Pharmaceuticals Inc., 2011). 3.2. Efficacy outcomes: NNT Table 2 provides the response and remission rates, and the respective NNT values, for each of the 5 trials separately and pooled. NNT vs. placebo for response for selegiline across all trials and all doses is 19 (95% CI 10–182), and for remission, NNT was 15 (95% CI 9–37). Results were essentially unchanged when excluding the 3 mg/24 h dose arm. When pooling together the two pivotal trials (Somerset Pharmaceuticals Inc., 2000a, 2003), the NNT for response was 11 (95% CI 6–109), and for remission, NNT was 9 (95% CI 6–28). In the pivotal trials, for the subgroup with baseline MADRS scores 4 30, pooled response rates were 31/84 (36.9%) for selegiline and 23/88 (26.1%) for placebo, for a NNT vs. placebo of 10 (ns); remission rates were 19/84 (22.6%) for selegiline and 12/ 88 (13.6%) for placebo, for a NNT vs. placebo of 12 (ns). In the pivotal trials, for the subgroup with a prior major depressive episode, pooled response rates were 49/147 (33.3%) for selegiline and 26/158 (22.8%) for placebo, for a NNT vs. placebo of 10 (95% CI 5–198); remission rates were 40/147 (27.2%) for selegiline and 21/158 (13.3%) for placebo, for a NNT vs. placebo of 8 (95% CI 5–20). 3.3. Safety outcomes: NNH Table 3 provides the rates for AEs that occurred in at least 5% of subjects (across all studies) receiving transdermal selegiline with a rate greater than that observed for placebo as noted in product labeling: headache, diarrhea, insomnia, dry mouth, application site reaction; the proportion of subjects gaining at least 5% of their initial body weight; the proportion of subjects that has at least one AE related to sexual functioning. For the AEs, those that were statistically significant were for insomnia (12.1% for selegiline vs. 6.7% for placebo, NNH 19 (95% CI 12–41)) and application site reaction (28.5% vs. 13.5%, NNH 7 (95% CI 6–10)). The NNH for selegiline vs. placebo for application site reaction was similar when excluding the 3 mg/24 h dose arm or when limiting the calculations to the pivotal trials only.

Data for weight gain ≥5% from baseline was available for all of the studies except for one (Somerset Pharmaceuticals Inc., 2000a). After pooling, rates for weight gain ≥5% from baseline were higher for placebo than for selegiline. Product labeling (Somerset Pharmaceuticals Inc., 2011) describes 2.1% of 757 patients receiving selegiline gaining ≥5% from baseline vs. 2.4% of 614 patients receiving placebo (NNH
334 (ns)). Moreover, weight loss ≥5% from baseline occurred in 5.0% of patients receiving selegiline vs. 2.8% for placebo, for a NNT of 46 (95% CI 24–582). Rates of AEs related to sexual functioning were low for both selegiline and placebo, with a NNH vs. placebo of 387 (ns). 3.4. Discontinuation due to adverse events or lack of efficacy Overall, about one-quarter of all subjects discontinued for any reason. Table 4 provides the rates for all-cause discontinuation, discontinuation due to an AE, and discontinuation due to lack of efficacy. Regarding the latter, overall rates of discontinuation because of lack of efficacy were low and marginally higher for subjects randomized to placebo (NNT vs. placebo 629 (ns)); when restricting the calculation to only the pivotal trials (Somerset Pharmaceuticals Inc., 2000a, 2003), NNT was 111 (ns). For subjects with a baseline MADRS score 4 30, in the pivotal trials, the proportion of subjects discontinuing because of lack of efficacy was 4/84 (4.8%) for selegiline vs. 8/88 (9.1%) for placebo, yielding a NNT of 24 (ns). Discontinuation due to an AE produced NNH values vs. placebo of 32 (95% CI 19-132) overall. Product labeling (Somerset Pharmaceuticals, 2011) documents slightly lower absolute rates of discontinuation due to an AE (7.1% for selegiline vs. 3.6% for placebo) than what we report here (7.5% for selegiline vs. 4.3% for placebo), however the NNH calculated from product labeling is similar (29, 95% CI 18–80) to what we found. Application site reactions led to discontinuation in 20/817 (2.4%) of subjects receiving selegiline vs. none for placebo, for a NNH vs. placebo of 41 (95% CI 29–72). Insomnia rarely led to discontinuation—4/817 (0.5%) of subjects receiving selegiline and 1/669 (0.1%) of subjects receiving placebo discontinued because of an AE of insomnia (NNH 294 (ns)). 3.5. Consideration of benefit vs. harm: LHH LHH can be used to illustrate trade-offs between benefits and harms in medical decision-making. When contrasting a highly desirable outcome, for example remission, vs. a highly undesirable outcome such as discontinuation from treatment due to an AE, the LHH for selegiline would be 3.6, meaning that selegiline is

3.6 times as likely to lead to a remission vs. a discontinuation due to an AE (see Table 5). Although LHH for remission vs. incidence of an application site reaction is 0.8, meaning that selegiline is 0.8 times as likely to lead to a remission vs. an occurrence of an application site reaction (i.e. 1.25 more likely to lead to an application site reaction than a remission), discontinuation due to an application site reaction was uncommon in the clinical trials. The LHH for remission vs. discontinuation due to application site reaction was 4.6. Similarly LHH for remission vs. incidence of insomnia was 2.1; LHH for remission vs. discontinuation due to insomnia was 32.7. Both weight gain and incidence of sexual dysfunction were low with selegiline, resulting in highly favorable benefit vs. risk determinations when contrasting remission with these adverse outcomes. Fig. 1 further illustrates NNT and NNH with their respective 95% CIs in graphical format.

b

Somerset Pharmaceuticals, 2000a. Somerset Pharmaceuticals, 2003. c Somerset Pharmaceuticals, 2000b. d Somerset Pharmaceuticals, 2000c. e Somerset Pharmaceuticals, 2000d.

4. Discussion

a

CI-confidence interval. NA-not applicable. NNT-number needed to treat. ns-not significant. Response is defined by a 50% reduction on the Montgomery-Asberg Depression Rating Scale (MADRS) at last observation carried forward (LOCF) endpoint; remission is defined by an endpoint rating scale score not exceeding 10 on the MADRS at LOCF endpoint.

136 (20.3%) NA 224 (27.4%) 15 (9 -3 7) 40 (27.4%) NA 10 (11.4%) NA 25 (28.1%) 6 (4–19)

36 (27.3%) 13 (ns)

26 (19.5%) NA

39 (26.2%) 11 (ns)

26 (17.1%) NA

27 (18.4%) −24 (ns)

34 (22.7%) NA

43 (28.5%) 93 (ns)

54 (36.2%) 12 (ns)

817 221 (33.2%) 19 (10-182) 146 54 (37.0%) NA 88 17 (19.3%) NA

N n, response NNT vs. placebo (95% CI) n, remission NNT vs. placebo (95% CI)

89 29 (32.6%) 8 (4–199)

132 50 (37.9%) 15 (ns)

133 41 (30.8%) NA

149 46 (30.9%) 12 (ns)

152 34 (22.4%) NA

147 36 (24.5%) −21 (ns)

150 44 (29.3%) NA

151 54 (35.8%) -82 (ns)

149 60 (40.3%) 31 (ns)

Selegiline Placebo Selegiline 6 mg/24 h Selegiline Placebo 6-12 mg/24 h Placebo Selegiline 6 mg/24 h

Selegiline 6 mg/24 h

Placebo

Selegiline 6 mg/24 h

Placebo

Selegiline 3 mg/24 h

Pooled E114e E113d P9804c P0052b E106a

Table 2 Response, remission, and number needed to treat vs. placebo.

669 190 (28.4%) NA

L. Citrome et al. / Journal of Affective Disorders 151 (2013) 409–417

Placebo

412

Transdermal selegiline appears efficacious and generally tolerable. Although application site reactions are common, other types of AEs such as sedation, weight gain and sexual dysfunction are essentially absent when compared with placebo. We can indirectly compare transdermal selegiline with other antidepressant agents for which NNT and NNH data vs. placebo have been published and find similar values for NNT for response and remission. Vilazodone is a specific serotonin reuptake inhibitor and serotonin 5HT1A receptor partial agonist; two 8-week placebo-controlled randomized clinical trials in outpatients with MDD where vilazodone was titrated to a target dose of 40 mg/d over the first 2 weeks evidenced efficacy for vilazodone as measured by the MADRS (Citrome, 2012). NNT for response vs. placebo was 8 (95% CI 6–16) and for remission was 14 (95% CI 8–55). NNH vs. placebo for discontinuation because of an adverse event (AE) was 27 (95% CI 15–104). The most commonly encountered AEs (incidence≥5% and at least twice the rate of placebo) were diarrhea, nausea, vomiting and insomnia, with NNH values vs. placebo of 6 (95% CI 5–8), 6 (95% CI 5-8), 30 (95% CI 18–82) and 26 (95% CI 16–78), respectively. NNH vs. placebo for any sexual AE was 12 (95% CI 9–18), but systematically collected data using rating scales of sexual function did not reveal treatment associated effects. Vilazodone was not associated with clinically relevant weight change in the short-term trials. Three short-term registration studies of adjunctive aripiprazole, 5 for olanzapine–fluoxetine combination, and 2 for quetiapine extended-release for the treatment of MDD reveal NNT for MADRS response and remission for adjunctive aripiprazole to be 7 (95% CI 5–11) and 8 (95% CI 6–13), respectively, for olanzapine– fluoxetine combination 8 (95% CI 6–17) and 13 (95% CI 8-40), respectively, and for adjunctive quetiapine extended release 300 mg/ d 9 (95% CI 5–24) and 7 (95% CI 5–16), respectively (Citrome, 2010). However, AE profiles for the 3 different adjunctive antipsychotics were more diverse, with adjunctive aripiprazole more strongly associated with akathisia (NNH 6, 95% CI 5–7), adjunctive olanzapine with weight gain (NNH 3, 95% CI 3–3), and adjunctive quetiapine with somnolence (NNH 3, 95% CI 3–4). There have also been published data regarding NNT for the use of duloxetine, fluoxetine and paroxetine in MDD, obtained from studies of 8–9 weeks duration (Cookson et al., 2006). Patients receiving duloxetine 60 mg/d had NNT for HAMD-17 response of 6 (95% CI 4–23) and remission 9 (ns). The NNTs for fluoxetine or paroxetine, 20 mg/day, were 7 (95% CI 5–18) for response and 11 (ns) for remission. NNH for AEs were not reported. Our NNH analysis is consistent with a systematic review of key safety and tolerability data from all clinical trials of transdermal selegiline in patients with MDD accrued during clinical development as reported to the FDA, and which included data from both controlled and uncontrolled clinical trials involving selegiline-treated (n¼2036)

Table 3 Treatment emergent adverse events and number needed to harm vs. placebo. E106a

P0052b

Selegiline 6 mg/24 h

Selegiline 6-12 mg/24 h

Placebo

Selegiline 6 mg/24 h

E113d Placebo

Selegiline 6 mg/24 h

E114e Placebo

Selegiline 3 mg/24 h

Pooled Selegiline 6 mg/24 h

Placebo

Selegiline

Placebo

89 15 (16.9%) −523 (ns) 6 (6.7%) −43 (ns) 6 (6.7%) 95 (ns) 6 (6.7%) −1305 (ns) 31 (34.8%)

88 15(17.0%) NA 8 (9.1%) NA 5 (5.7%) NA 6 (6.8%) NA 15(17.0%)

132 21 (15.9%) -73 (ns) 13 (9.8%) 17 (9-1539) 40 (30.3%) 7 (4–17) 16 (12.1%) 22 (ns) 55 (41.7%)

133 23(17.3%) NA 5 (3.8%) NA 19(14.3%) NA 10 (7.5%) NA 32(24.1%)

149 24 (16.1%) 44 (ns) 8 (5.4%) -83 (ns) 11 (7.4%) 25 (ns) 7 (4.7%) -82 (ns) 49 (32.9%)

152 21 (13.8%) NA 10 (6.6%) NA 5 (3.3%) NA 9 (5.9%) NA 11 (17.2%)

147 33 (22.4%) 15 (ns) 15 (10.2%) 35 (ns) 19 (12.9%) 18 (ns) 14 (9.5%) 29 (ns) 36 (24.5%)

150 23 (15.3%) NA 11 (7.3%) NA 11 (7.3%) NA 9 (6.0%) NA 17 (11.3%)

151 32 (21.2%) 38 (ns) 15 (9.9%) 98 (ns) 8 (5.3%) 54 (ns) 10 (6.6%) 88 (ns) 28 (18.5%)

149 21 (14.1%) -23 (ns) 16 (10.7%) 55 (ns) 15 (10.1%) 16 (9-102) 8 (5.4%) -907 (ns) 34 (22.8%)

146 27 (18.5%) NA 13 (8.9%) NA 5 (3.4%) NA 8 (5.5%) NA 15 (10.3%)

817 146 (17.9%) 64 (ns) 73 (8.9%) 53 (ns) 99 (12.1%) 19 (12-41) 61 (7.5%) 85 (ns) 233 (28.5%)

669 109 (16.3%) NA 47 (7.0%) NA 45 (6.7%) NA 42 (6.3%) NA 90 (13.5%)

6 (4–20) Missing NA 2 (2.2%)

NA Missing NA 0

6 (4–16) 2 (1.5%) 8778 (ns) 1 (0.8%)

NA 2 (1.5%) NA 1 (0.8%)

4 (3-6) 5 (3.4%) 50 (ns) 3 (2.0%)

NA 2 (1.3%) NA 1 (0.7%)

8 (5 -2 2) 1 (0.7%) -51 (ns) 0

NA 4 (2.7%) NA 0

13 (7-285) 3 (2.0%) -1470 (ns) 1 (0.7%)

8 (5 -2 4) 1 (0.7%) -73 (ns) 0

NA 3 (2.1%) NA 2 (1.4%)

7 (6 -1 0) 12 (1.6%) -409 (ns) 7 (0.9%)

NA 11 (1.9%) NA 4 (0.6%)

45 (ns)

NA

17,556 (ns)

NA

74 (ns)

NA

NA

NA

-142 (ns)

-73 (ns)

NA

387 (ns)

NA

AE-adverse event. CI-confidence interval. NA-not applicable. NNH-number needed to harm. ns-not significant. A negative NNH can be interpreted as a NNT.

L. Citrome et al. / Journal of Affective Disorders 151 (2013) 409–417

N n, Headache NNH vs. Placebo (95% CI) n, diarrhea NNH vs. Placebo (95% CI) n, Insomnia NNH vs. placebo (95% CI) n, Dry mouth NNH vs. placebo (95% CI) n, Application site reaction NNH vs. placebo (95% CI) n, Weight gain ≥5%f NNH vs. Placebo (95% CI) n, ≥1 AE related to sexual functioning NNH vs. placebo (95% CI)

Placebo

P9804c

a

Somerset Pharmaceuticals, 2000a. Somerset Pharmaceuticals, 2003. Somerset Pharmaceuticals, 2000b. d Somerset Pharmaceuticals, 2000c. e Somerset Pharmaceuticals, 2000d. f Pooled N selegiline 729, placebo 581. b c

413

414

Table 4 Discontinuation due to an adverse event or to lack of efficacy, NNH and NNT vs. placebo. E106a

P9804c

e

Pooled

Placebo

Selegiline Placebo 6-12 mg/24 h

Selegiline 6 mg/24 h

Placebo

Selegiline 6 mg/24 h

Placebo

Selegiline 3 mg/24 h

Selegiline 6 mg/24 h

Placebo

Selegiline

Placebo

89 10 (11.2%) 4 (4.5%) −85 (ns)

88 15(17.0%) 5 (5.7%) NA

132 32 (24.2%) 9 (6.8%) 22 (ns)

133 27(20.3%) 3 (2.3%) NA

149 41 (27.5%) 11 (7.4%) 48 (ns)

152 41(27.0%) 8 (5.3%) NA

147 42 (28.6%) 12 (8.2%) 21 (ns)

150 35(23.3%) 5 (3.3%) NA

151 42 (27.8%) 8 (5.3%) -552 (ns)

149 40 (26.9%) 17 (11.4%) 17 (ns)

146 37 (25.3) 8 (5.5%) NA

817 207(25.3%) 61 (7.5%) 32 (19-132)

669 155(23.2%) 29 (4.3%) NA

5 (5.6%)

9 (10.2%)

5 (3.8%)

3 (2.3%)

3 (2.0%)

8 (5.3%)

8 (5.4%)

6 (4.0%)

13 (8.6%)

5 (3.4%)

7 (4.8%)

39 (4.8%)

33 (4.9%)

22 (ns)

NA

-66 (ns)

NA

31 (ns)

NA

-70 (ns)

NA

-27 (ns)

70 (ns)

NA

629 (ns)

NA

Somerset Pharmaceuticals, 2000a. Somerset Pharmaceuticals, 2003. c Somerset Pharmaceuticals, 2000b. d Somerset Pharmaceuticals, 2000c. e Somerset Pharmaceuticals, 2000d. b

E114

Selegiline 6 mg/24 h

AE-adverse event. CI-confidence interval. D/C-discontinued. NA-not applicable. NNH-number needed to harm. NNT-number needed to treat. ns-not significant. A negative NNH can be interpreted as a NNT; a negative NNT can be interpreted as a NNH. a

E113d

L. Citrome et al. / Journal of Affective Disorders 151 (2013) 409–417

N n, D/C for any reason n, D/C due to AE NNH vs. placebo (95% CI) n, D/C due to lack of efficacy NNT vs. placebo (95% CI)

P0052b

L. Citrome et al. / Journal of Affective Disorders 151 (2013) 409–417 Table 5 Likelihood to be helped or harmed: remission vs. incidence of adverse events or discontinuation due to adverse events. Outcome to be avoided

NNH of outcome to be NNT for remissionb avoideda

LHH

Discontinuation due to AE Discontinuation due to application site reaction Discontinuation due to insomnia Incidence of application site reaction Incidence of insomnia Incidence of weight gain ≥5% from baseline Incidence of ≥1 AE related to sexual functioning

21 41

9 9

3.6 4.6

294c 7

9 9

32.7 0.8

19 −409c

9 9

2.1 NAd

387c

9

43

AE—adverse event. NA—not applicable. NNH—number needed to harm. NNT—number needed to treat. LHH – likelihood to be helped or harmed. a

From prior tables, all data available for selegiline vs. placebo. From the pivotal trials (Somerset Pharmaceuticals, 2000a, 2003), see text; remission is defined by an endpoint rating scale score not exceeding 10 on the Montgomery Asberg Depression Rating Scale at Last Observation Carried Forward endpoint. c Not statistically significant. d Selegiline had a lower incidence of weight gain ≥5% from baseline than placebo, hence a negative number for NNH and calculation of LHH is moot. b

Fig. 1. Number needed to treat or harm and 95% CIs for selegiline vs. placebo for selected outcomes.

and placebo-treated (n¼668) patients (Robinson and Amsterdam, 2008). The principal side effects of transdermal selegiline therapy were local dermal reactions and insomnia, both of which were doserelated. Side effects often associated with antidepressant treatment, such as sexual dysfunction and excessive weight gain, were uncommon. The authors of that review concluded that transdermal selegiline is safe and well tolerated, with an improved safety margin compared with orally administered MAOIs. The low rates of sexually related AE rates observed with transdermal selegiline were consistent with a metaanalysis of data of the impact of transdermal selegiline 6 mg/24 h on various domains of sexual function using a patient-rated questionnaire administered in the 4 fixed-dosed short-term studies (transdermal selegiline, N ¼389; placebo, N ¼400) (Clayton et al., 2007). The Medex Sexual Dysfunction Subscale was used to assess sexual interest, arousal, and maintenance of interest, orgasm, and satisfaction. Estimates of difference between selegiline and placebo

415

demonstrated a significant improvement in sexual satisfaction and for women there was a significant positive effect on interest, maintaining interest during sex, and satisfaction. Although this report was focused on short-term studies of acute MDD, data is available from a long-term study assessing the safety and efficacy of initial and continuation transdermal selegiline therapy in patients MDD (Amsterdam and Bodkin, 2006). A total of 675 patients were enrolled into the initial 10-week openlabel phase of the study where they received transdermal selegiline 6 mg/24 h. Of these, 322 who responded with a HAMD-17 score of 10 or less were subsequently randomized to double-blind transdermal selegiline 6 mg/24 h treatment or placebo for 52 weeks. At study week 52, significantly fewer subjects randomized to selegiline experienced relapse (25/149 [16.8%]) compared with placebo (50/163 [30.7%]) (P ¼0.0025) (27), for a NNT of 8 (95% CI 5–22). Moreover, subjects receiving selegiline experienced a significantly longer time to relapse compared with those receiving placebo (P ¼0.0048). The incidence for application site reactions for transdermal selegiline in the open-label phase was 28.8%, and was 15.2% vs. 3.7% for placebo in the double-blind phase (27), for a NNH of 9 (95% CI 6–20). There were no clinically relevant concerns regarding sexual dysfunction or weight gain. No cases of hypertensive crisis were reported, despite the lack of requirement for dietary tyramine restrictions. Rare events are usually not detected in registration trials. However, post-marketing surveillance (pharmacovigilance studies) can provide information in this regard (Pae et al., 2012). After FDA approval of transdermal selegiline for MDD, deidentified data were obtained on AEs, regardless of causality, as collected and compiled in the pharmaceutical company's adverse event collection systems/databases after the launch of transdermal selegiline in the United States. All reports of hypertensive crisis, suicide attempts, and overdoses were examined to independently determine relation of the AE to transdermal selegiline. From April 2006 to October 2010, a total of 3155 AEs in 1516 patients were reported (5.2% of the total exposures; N ¼29,141), regardless of causality. The most commonly reported AEs were application site reactions and insomnia. There were 13 self-reports of possible hypertensive events or hypertension, although objective clinical data were not submitted in any of these cases. Thirteen drug–drug interactions (0.04%) were reported.

5. Limitations The short term studies included in our analyses were not identical. As noted earlier, study duration was either 6 or 8 weeks, and not all studies employed a fixed dose design. One study (Somerset Pharmaceuticals, 2000d) included a 3 mg/24 h transdermal selegiline treatment arm, the others did not. All of the studies included a singleblind placebo lead-in to eliminate placebo responders prior to randomization except for one (Somerset Pharmaceuticals, 2003). Except for the first study (Somerset Pharmaceuticals, 2000a), subjects were not advised to follow a tyramine-restricted diet, even while on the higher doses. Regarding the calculations of response and remission, pooling uninformative studies for elucidation of efficacy can be problematic; in the absence of an active control, we cannot definitively distinguish between a negative vs. failed study. The estimates of NNT and NNH calculated from registration trials can be difficult to generalize to patients who would not have met the inclusion and exclusion criteria of these studies. By necessity, patients with significant somatic or psychiatric comorbidities (including substance use disorders) are generally excluded from participating in registration trials, as are patients who are actively suicidal. With regard to discontinuing treatment, persons in clinical trials and the investigators may be

416

L. Citrome et al. / Journal of Affective Disorders 151 (2013) 409–417

incentivized to continue participation; this may render discontinuation rates artificially low. Specific limitations related to the metric of NNT include the importance of also noting the rates that are used to calculate it—a NNT of 10 when it is calculated from rates of 20% vs. 10% as opposed to when it is calculated from rates of 80% vs. 70% are two very different clinical treatment scenarios (Citrome and Ketter, 2013). The actual definitions used to categorize responders and remitters can also impact on the interpretability of NNT; for example a threshold of a MADRS ≤10 to define remission can yield different NNTs than thresholds of ≤8 or ≤12. In addition, the effect of time is important – for example in the previously cited study of duloxetine, NNT for remission decreased steadily from a NNT of 79 after 1 week to a NNT of 6 after 9 weeks (Cookson et al., 2006). Although NNT measures successful outcomes in ‘patient units’ compared with commonly reported continuous metrics (such as changes in the MADRS total score), some precision is lost and NNT would not be suitable as a primary outcome measure when reporting a clinical trial. There remains a need for reporting effect sizes for continuous measures, such as Cohen's d and/or area under the curve (Kraemer and Kupfer, 2006). Limitations regarding NNH include the care that must be exercised when selecting outcomes for the calculation of this metric—the harms must be clinically relevant and applicable to the individual patient being treated. When contrasting benefits and harms using LHH, the trade-offs must be consistent with clinical common sense regarding importance to the clinician and the patient, and whether the effects are temporary or permanent (Citrome and Kantrowitz, 2008). For example, in the case of vilazodone, the risk of transient nausea, although common, may not be necessarily prohibitive considering the increased likelihood of achieving response, unless discontinuation because of nausea is relevant for that particular patient (Citrome, 2012).

6. Conclusions Transdermal selegiline is efficacious and tolerable, with NNT for response and remission similar to that for other antidepressant regimens for which similar analyses have been published. The most common AEs attributable to selegiline are application site reaction and insomnia. Discontinuation from the clinical trials because of application site reaction is uncommon and discontinuation because of insomnia is rare. There appear to be no clinically relevant effects of selegiline on weight or sexual functioning. Further characterization of the ideal candidate for transdermal selegiline therapy would be of interest.

Role of funding source In the past 36 months, L. Citrome, was a consultant for, has received honoraria from, or has conducted clinical research supported by the following: Alexza, Alkermes, AstraZeneca, Avanir, Bristol-Myers Squibb, Eli Lilly, Envivo, Forest, Genentech, Janssen, Lundbeck, Merck, Mylan, Novartis, Noven, Otsuka, Pfizer, Shire, Sunovion and Valeant. J.F. Goldberg was a consultant for, has received honoraria from, or has conducted clinical research supported by the following: Astra Zeneca, Eli Lilly, Mylan, Novartis, and Sunovion. K. Portland is a full-time employee of Mylan.

Conflict of interest This study was supported by Mylan Specialty L.P., Basking Ridge, NJ. L. Citrome designed and undertook the analysis and wrote the first draft of the manuscript. All authors contributed to and approved the final manuscript.

Acknowledgment We thank Chao Wang, PhD, statistician for Mylan Specialty L.P., for extracting the requested data from the study databases.

References Amsterdam, J.D., 2003. A double-blind, placebo-controlled trial of the safety and efficacy of selegiline transdermal system without dietary restrictions in patients with major depressive disorder. Journal of Clinical Psychiatry 64, 208–214. Amsterdam, J.D., Bodkin, J.A., 2006. Selegiline transdermal system in the prevention of relapse of major depressive disorder: a 52-week, double-blind, placebosubstitution, parallel-group clinical trial. Journal of Clinical Psychopharmacology 26, 579–586. Bodkin, J.A., Amsterdam, J.D., 2002. Transdermal selegiline in major depression: a double-blind, placebo-controlled, parallel-group study in outpatients. American Journal of Psychiatry 159, 1869–1875. Citrome, L., 2008. Compelling or irrelevant? Using number needed to treat can help decide. Acta Psychiatrica Scandinavica 117, 412–419. Citrome, L., 2009. Quantifying risk: the role of absolute and relative measures in interpreting risk of adverse reactions from product labels of antipsychotic medications. Current Drug Safety 4, 229–237. Citrome, L., 2010. Adjunctive aripiprazole, olanzapine, or quetiapine for major depressive disorder: an analysis of number needed to treat, number needed to harm, and likelihood to be helped or harmed. Postgraduate Medicine 122 (4), 39–48. Citrome, L., 2012. Vilazodone for major depressive disorder: a systematic review of the efficacy and safety profile for this newly approved antidepressant—what is the number needed to treat, number needed to harm and likelihood to be helped or harmed? International Journal of Clinical Practice 66, 356–368. Citrome, L., Kantrowitz, J., 2008. Antipsychotics for the treatment of schizophrenia: likelihood to be helped or harmed, understanding proximal and distal benefits and risks. Expert Review of Neurotherapeutics 8, 1079–1091. Citrome, L., Ketter, T.A., 2013. When does a difference make a difference? Interpretation of number needed to treat, number needed to harm, and likelihood to be helped or harmed. International Journal of Clinical Practice 67, 407–411. Clayton, A.H., Campbell, B.J., Favit, A., Yang, Y., Moonsammy, G., Piontek, C.M., Amsterdam, J.D., 2007. Symptoms of sexual dysfunction in patients treated for major depressive disorder: a meta-analysis comparing selegiline transdermal system and placebo using a patient-rated scale. Journal of Clinical Psychiatry 68, 1860–1866. Cookson, J., Gilaberte, I., Desaiah, D., Kajdasz, D.K., 2006. Treatment benefits of duloxetine in major depressive disorder as assessed by number needed to treat. International Clinical Psychopharmacology 21, 267–273. Feiger, A.D., Rickels, K., Rynn, M.A., Zimbroff, D.L., Robinson, D.S., 2006. Selegiline transdermal system for the treatment of major depressive disorder: an 8-week, double-blind, placebo-controlled, flexible-dose titration trial. Journal of Clinical Psychiatry 67, 1354–1361. Kraemer, H.C., Kupfer, D.J., 2006. Size of treatment effects and their importance to clinical research and practice. Biological Psychiatry 59, 990–996. Lee, K.C., Chen, J.J., 2007. Transdermal selegiline for the treatment of major depressive disorder. Neuropsychiatric Disease and Treatment 3, 527–537. Montgomery, S.A., Asberg, M., 1979. A new depression scale designed to be sensitive to change. British Journal of Psychiatry 134, 382–389. Nandagopal, J.J., DelBello, M.P., 2009. Selegiline transdermal system: a novel treatment option for major depressive disorder. Expert Opinion on Pharmacotherapy 10, 1665–1673. Nemeroff, C.B., 2007. The burden of severe depression: a review of diagnostic challenges and treatment alternatives. Journal of Psychiatric Research 41, 189–206. Pae, C.U., Bodkin, J.A., Portland, K.B., Thase, M.E., Patkar, A.A., 2012. Safety of selegiline transdermal system in clinical practice: analysis of adverse events from postmarketing exposures. Journal of Clinical Psychiatry 73, 661–668. Pae, C.U., Lim, H.K., Han, C., Neena, A., Lee, C., Patkar, A.A., 2007. Selegiline transdermal system: current awareness and promise. Progress in Neuropsychopharmacology and Biological Psychiatry 31, 1153–1163. Patkar, A.A., Portland, K.B., Pae, C.U., 2012. Selegiline transdermal system in major depressive disorder. Neuropsychiatry 2, 125–134. Preskorn, S.H., 2006. Why the transdermal delivery of selegiline (6 mg/24 h) obviates the need for a dietary restriction on tyramine. Journal of Psychiatric Practice 12, 168–172. Robinson, D.S., Gilmor, M.L., Yang, Y., Moonswammy, G., Azzaro, A.J., Oren, D.A., Campbell, B.J., 2007. Treatment effects of selegiline transdermal system on symptoms of major depressive disorder: a meta-analysis of short-term, placebo-controlled, efficacy trials. Psychopharmacology Bulletin 40, 15–28. Robinson, D.S., Amsterdam, J.D., 2008. The selegiline transdermal system in major depressive disorder: a systematic review of safety and tolerability. Journal of Affective Disorders 105, 15–23. Somerset Pharmaceuticals, Inc, 2000a. Clinical Study Report. A double-blind, placebo-controlled, parallel assessment of the safety and efficacy of the selegiline transdermal system in patients with major depression. Protocol S9303-E106-96B. Somerset Pharmaceuticals, Inc., 2000b. Clinical Study Report. A double-blind, placebo-controlled, parallel assessment of the safety and efficacy of the selegiline transdermal system in patients with major depression. Protocol S9303-P9804. Somerset Pharmaceuticals, Inc., 2000c. Clinical Study Report. A double-blind, placebo-controlled, parallel assessment of the safety and efficacy of the

L. Citrome et al. / Journal of Affective Disorders 151 (2013) 409–417

selegiline transdermal system in patients with major depression. Protocol S9303-E113-98B. Somerset Pharmaceuticals, Inc., 2000d. Clinical Study Report. A double-blind, placebo-controlled, parallel assessment of the safety and efficacy of two doses of the selegiline transdermal system (10 mg and 20 mg) in patients with major depression. Protocol S9303-E114. Somerset Pharmaceuticals, Inc., 2003. Clinical Study Report. A Phase III flexible dose titration study of the safety and efficacy of the selegiline transdermal system in patients with major depression. Protocol S9303-P0052.

417

Somerset Pharmaceuticals, Inc., 2011. EMSAM (selegiline transdermal system). Product label. Revised August 2011. Available at: 〈http://www.emsam.com/ assets/emsam/pdfs/pi.pdf〉. Accessed 1 February 2013. US Food and Drug Administration, 2007. Emsam (Selegiline) Transdermal System Drug Approval Package. Available at: 〈http://www.accessdata.fda.gov/drugsatf da_docs/nda/2006/021336s000_021708s000TOC.cfm〉. (Accessed 31.01.13).