Sage Journals: Discover world-class research

Abstract

Background: Monoamine oxidase type B (MAO-B) inhibitors exhibit neuroprotective effects in preclinical models of PD but clinical trials have failed to convincingly demonstrate disease modifying benefits in PD patients.

Objective: To perform a secondary analysis of NET-PD LS1 to determine if longer duration of MAO-B inhibitor exposure was associated with less clinical decline.

Methods: The primary outcome measure was the Global Outcome (GO), comprised of 5 measures: change from baseline in the Schwab and England (ADL) scale, the 39-item Parkinson’s Disease Questionnaire (PDQ-39), the UPDRS Ambulatory Capacity Scale, the Symbol Digit Modalities Test, and the most recent Modified Rankin Scale. A linear mixed model was used to explore the association between the cumulative duration of MAO-B inhibitor exposure and the GO, adjusting for necessary factors and confounders. Associations between MAO-B inhibitor exposure and each of the five GO components were then studied individually.

Results: 1616 participants comprised the analytic sample. Mean observation was 4.1 (SD = 1.4) years, and 784 (48.5%) participants received an MAO-B inhibitor. The regression coefficient of cumulative duration of MAO-B inhibitor exposure (in years) on the GO was – 0.0064 (SE = 0.002, p = 0.001). Significant associations between duration of MAO-B inhibitor exposure and less progression were observed for ADL (p < 0.001), Ambulatory Capacity (p < 0.001), and the Rankin (p = 0.002).

Conclusions: Our analysis identified a significant association between longer duration of MAO-B inhibitor exposure and less clinical decline. These findings support the possibility that MAO-B inhibitors slow clinical disease progression and suggest that a definitive prospective trial should be considered.

Keywords

Parkinson’s disease MAO-B inhibitor rasagiline selegiline treatment disease modification

INTRODUCTION

There has long been interest in whether monoamine oxidase type B (MAO-B) inhibitors slow progression of Parkinson’s disease (PD) and improve long-term outcomes. Indeed, the MAO-B inhibitors selegiline, rasagiline, and safinamide have all demonstrated neuroprotective effects in PD models [1 –3]. However, clinical trials of PD patients have failed to convincingly prove disease modifying effects.

To date, the two largest, prospective, randomized clinical trials to attempt to evaluate whether MAO-B inhibitors provide a disease modifying effect were the DATATOP (Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism) [4] and ADAGIO (Attenuation of Disease Progression with Azilect Given Once-daily) [5] studies. In the DATATOP study [4], the risk of previously untreated patients progressing in disability to the point of requiring levodopa was reduced by 57% for subjects randomized to blinded treatment with selegiline 10 mg/day compared to placebo (p < 10^–10). However, this assessment of disease progression was confounded by the detection of a small but significant symptomatic benefit (∼1.72 total UPDRS points) at 1 month that was unknown at the time the trial was designed. In the ADAGIO study [5], early vs. delayed treatment with rasagiline 1 mg per day provided benefits that were consistent with a disease-modifying effect in early, otherwise untreated patients, but early vs. delayed treatment with rasagiline 2 mg per day did not. A few smaller, long-term trials [6, 7] in patients treated with levodopa suggested that MAO-B inhibitors provide increasing benefit with longer duration of exposure, potentially consistent with slowing of disease progression. However, relatively small sample sizes, high withdrawal rates, and emphasis on motor outcomes limited the strength of their conclusions.

We undertook a secondary analysis of theNET-PD Long-term Study-1 (LS1) [8] to determine if longer duration of MAO-B inhibitor exposure was associated with less clinical decline in that trial, where all participants were required to be on dopaminergic therapy at trial entry and therapy could be adjusted as appropriate over time. NET-PD LS1 provided a favorable opportunity to performsuch an analysis because it was one of the largest and longest PD clinical trials ever conducted and it included a framework to assess clinical disease progression.

METHODS

We performed a secondary analysis of NET-PD LS1 [8]. NET-PD LS1 was a multicenter, double-blind, placebo-controlled trial of 1741 participants with early PD to determine whether creatine monohydrate 10 mg/day is more effective than placebo in slowing long-term clinical decline [9]. Participants eligible for enrollment were within 5 years of diagnosis and had taken levodopa or a dopamine agonist for at least 90 days but not longer than 2 years. Participants continued their PD medications upon entering the study, and during the course of the trial PD medications were adjusted as appropriate. Study visits were scheduled at 3, 12, 24, 36, 48, 60, 72 and 84 months after baseline or until the end of the trial. The objective of the study was to compare clinical decline between treatment groups using a global statistical test (GST) that provided a multidimensional assessment of disease progression [8 –10]. Enrollment took place from March 2007 to May 2010 and follow-up for each participant was defined as the time from baseline until loss to follow-up, death, or July 17, 2013, when the trial was terminated early for futility [8].

Outcome measure

The primary outcome measure for our analysis was the Global Outcome (GO) [11], a standardized version of the GST, which is comprised of 5 measures: 1) change from baseline in the Schwab and England Activities of Daily Living (ADL) scale [12]; 2) change from baseline in the PD 39-item Questionnaire (PDQ-39) [13]; 3) change from baseline in the Unified PD Rating Scale (UPDRS) Ambulatory Capacity (AC) (sum of items 13, 14, 15, 29, and 30) [14]; 4) change from baseline in the Symbol Digit Modalities (SDM) Test for cognitive function [15]; and 5) the most recent measurement of the Modified Rankin scale (mRS) for global disability [16].

For each yearly visit (months 12, 24, 36, 48, 60, and 72) the five measures were ranked (with higher ranks representing worse clinical decline) and the sum of the five ranks was calculated for each participant [8, 9]. The GO is defined as the standardized form of the rank-sum, where GO = summed-rank/(5n) and n represents the number of participants with the five measures assessed at that visit [11]. Thus, at each yearly visit, the GO provided a score for each participant that represented their multi-domain clinical progression relative to other participants and allowed us to test the association between the exposure of interest and overall clinical progression. The GO has a bell-shaped distribution with an overall mean of 0.5 and ranges within 0 (best possible) to 1 (worst possible).

Analytic sample

All NET-PD LS1 participants who completed the baseline evaluation and at least one yearly evaluation that included all five measures of the GO were included in the analysis.

Definition of duration of MAO-B inhibitor exposure

At every study visit, NET-PD LS1 participants reported all medications and dates taken since their previous visit. For each participant we determined whether they received an MAO-B inhibitor during the NET-PD LS1 trial and their cumulative duration of MAO-B inhibitor exposure (in years) between baseline and each annual visit (total number of days from baseline on an MAO-B inhibitor divided by 365.25). Two alternative definitions of duration of MAO-B inhibitor exposure were used for sensitivity analyses. These were: (1) cumulative percentage of time in the trial that a participant received an MAO-B inhibitor between baseline and each annual visit (total number of days on an MAO-B inhibitor divided by the number of days in the trial between baseline and the yearly visit multiplied by 100%), and (2) percentage of time that a participant received an MAO-B inhibitor since the previous annual visit (total number of days on an MAO-B inhibitor since last visit divided by the number of days in the trial between the current yearly visit and the previous yearly visit multipliedby 100%).

Statistical analyses

A linear mixed model was used to explore the association between the cumulative duration of MAO-B inhibitor exposure (in years) and the GO, adjusting for covariates. This model allowed us to utilize repeated measures of the MAO-B inhibitor cumulative duration and the GO at all yearly visits, under the relatively weak missing data assumption of missing at random. In the model, we included subject-specific and site-specific random intercepts to account for within-subject and within-site correlations. We forced several additional factors into the linear mixed model. These included the randomization assignment in NET-PD LS1 (creatine vs. placebo), length of time between baseline and each annual visit (years), total levodopa equivalent daily dosage (LEDD) [17] at baseline, and the change in LEDD from baseline to each visit. We also included baseline age and its interaction with visit time (years), because these two factors were identified as important adjustors in other NET-PD LS1 analyses [18] and additional analysis confirmed the necessity of including them here. Other baseline covariates that were considered as confounders for adjustment of the exposure and outcome relationship included gender, race and ethnicity (non-Hispanic White vs. other), education level, marital status, insurance status, Beck Depression Inventory (BDI) total score, UPDRS Motor + ADL score, UPDRS Mentation, Behavior and Mood subscore, intellectual impairment (UPDRS item 1) score, total functional capacity (TFC) score, total Scales for Outcomes of Parkinson’s disease–cognition (SCOPA-COG) score, total European Quality of Life-5 Dimensions (EQ5D) score, time since PD symptom onset, and time since PD diagnosis. We first identified confounders, i.e., those factors that were associated with both cumulative duration of MAO-B inhibitor exposure and the GO with p < 0.25 in a univariable analysis. We used the variance inflation factor (VIF) to examine multicollinearity. If multicollinearity was not detected, the identified factors were included for regression adjustment, together with the factors that were forced into the model. In the exploratory analysis stage, we compared this confounder adjustment approach to the propensity score approach [19] in analyzing categorized MAO-B inhibitor duration using only the subset of participants with the observed fifth year outcomes and obtained similar results (data not reported). Due to the lack of standard methods and software for propensity score matching/weighting with continuous, repeatedly measured exposures, we adopted the regression adjustment approach.

If a significant association between cumulative duration of MAO-B inhibitor exposure (in years) and the GO was identified after adjusting for other covariates, each component of the GO would then be evaluated individually, adjusting for covariates. For the change in ADL, PDQ-39, AC, and SDM, a linear mixed model with subject-specific and site-specific random intercepts that included the cumulative duration of MAO-B inhibitor exposure and the same set of adjusting covariates was used. We dichotomized the Modified Rankin outcome (0–1 vs. 2–5) [20], where a Rankin score of 0-1 represents no significant disability. A logit link was used in the generalized linear mixed model that includes subject-specific and site-specific random intercepts.

To provide context and aid interpretation of results for the components of the GO, the estimated regression coefficient of the cumulative duration of MAO-B inhibitor exposure among all participants was compared with the estimated regression coefficient of time among participants who never received an MAO-B inhibitor. For each of the five individual measures, a generalized linear mixed model was estimated among participants who never received an MAO-B inhibitor between baseline and the last yearly visit, adjusting for visit time (in years) and treatment assignment (creatine vs. placebo) as a fixed effect and including person-specific, and site-specific random effects. The estimated coefficient of time reflects the average yearly change of the GO component among participants who did not receive an MAO-B inhibitor. We also implemented an alternative approach to estimating average yearly changes. Details of this alternative approach are provided in the supplemental materials, Section S1.

An additional linear mixed model with subject-specific and site-specific random intercepts was estimated to explore the association between the cumulative duration of MAO-B inhibitor exposure (in years) and change in total LEDD (in mg), adjusting for the same set of covariates except for change in LEDD itself.

RESULTS

Analytic sample

Of the 1741 participants in NET-PD LS1, 1616 completed baseline and at least one yearly evaluation that included all five measures of the GO; their baseline characteristics are presented in Table 1. At baseline, mean age (SD) was 61.7 (9.4) years, 1039 (64%) were male, mean time from diagnosis was 1.6 (1.1) years, mean total UPDRS score was 25.9 (11.2) and mean LEDD was 379.2 (243.4) mg (Table 1). Mean time from baseline to last yearly visit with a GO was 4.1 (1.4) years and mean time from diagnosis to last yearly visit with a GO was 5.7 (1.8) years.

Cumulative duration of MAO-B inhibitor exposure

Of the 1616 participants in our analytic sample, 784 (48.5%) received an MAO-B inhibitor between baseline and their last yearly visit with a GO. Of these, 586 (74.7%) received rasagiline (Azilect) only, 135 (17.2%) received selegiline (Eldepryl, Zelapar, EMSAM) only, and 61 (7.8%) received both rasagiline (Azilect) and selegiline (Eldepryl, Zelapar) (Table 2). The cumulative duration of MAO-B inhibitor exposure by visit year is presented in Table 3, and the overall cumulative duration and the cumulative percentage duration of MAO-B inhibitor exposure at the last yearly visit is depicted in Supplemental Figure 1. As seen in Table 3, of the 709 participants completing a Year 5 visit, 357 (50%) received an MAO-B inhibitor between baseline and the Year 5 visit; the median cumulative duration of exposure for these participants was 3.48 years. For all participants who received an MAO-B inhibitor during the study, median cumulative duration of exposure at last yearly visit that included a GO was2.9 years.

Confounder selection and adjustments

We retained the following baseline factors according to the p < 0.25 criteria: education, BDI total score, UPDRS Motor + ADL score, UPDRS Mentation, Behavior and Mood subscore, intellectual impairment score (UPDRS item 1), TFC score, total EQ5D score, total SCOPA-COG score, and length of time since PD diagnosis. The VIFs for the identified factors, the duration of MAO-B inhibitor exposure, and the factors that were forced into the model (visit time, treatment assignment, baseline LEDD, change in LEDD, age and age by time interaction) were less than 5. Therefore, multicollinearity was not detected, and we included all as covariates in the regression model.

Primary Analysis –Association between cumulative duration of MAO-B inhibitor exposure and the GO

The regression coefficient of cumulative duration of MAO-B inhibitor exposure (in years) on the GO was –0.0064 (SE = 0.002, p = 0.001). Since higher GO scores indicate worse clinical decline, this result indicates a significant association between longer duration of MAO-B inhibitor exposure and less clinical decline. This result does not change meaningfully if we do not force treatment assignment (creatine vs placebo), baseline LEDD, and change in LEDD into the model.

Association between cumulative duration of MAO-B inhibitor exposure and individual components of the GO

The regression coefficients of cumulative duration of MAO-B inhibitor exposure were significant for change in ADL (p < 0.001), change in AC (p < 0.001), and the modified Rankin (p = 0.002) (Table 4). These results indicate a significant association between longer duration of MAO-B inhibitor exposure and less clinical decline for these three outcomes. A one year increase in MAO-B inhibitor use was associated with benefit equivalent in magnitude to 17% of the annual decline in ADL (Schwab and England), and 24% of the annual decline in ambulatory capacity (AC) observed in participants not receiving an MAO-B inhibitor (Table 4). On average, a one year increase in MAO-B inhibitor use was associated with a benefit equivalent in magnitude to 19% of the annual increase in the probability of having a Rankin score ≥2 (Table 4). The alternative approach analysis led to very similar results (Table 4). We did not observe a significant association between increasing duration of MAO-B inhibitor exposure and less decline on the PDQ-39 or SDM test for cognitive function.

Sensitivity analyses

Results of the two sensitivity analyses based on alternative definitions of duration of MAO-B inhibitor exposure were consistent with the primary analysis. Specifically, using either of the alternative definitions of duration of MAO-B inhibitor exposure, the coefficient of duration of MAO-B inhibitor exposure was significantly associated with the GO, and longer duration of MAO-B Inhibitor exposure was significantly associated with less clinical decline. For the individual components of the GO, both sensitivity analyses identified significant associations between duration of MAO-B inhibitor exposure and change in ADL, change in AC, and the Rankin outcome but significant associations with the PDQ-39 and SDM were not observed. Details of the sensitivity analyses are provided in the supplemental materials, Section S2.

Association between the cumulative duration of MAO-B inhibitor exposure (in years) and change in LEDD

A one year increase in MAO-B inhibitor cumulative duration of exposure was associated with a –15.6 mg difference in change in total LEDD, adjusting for covariates (p < .001). The magnitude of this estimated association is equivalent to 18% of the annual increase in total LEDD among participants who did not receive an MAO-B inhibitor (Table 4).

DISCUSSION

In this exploratory analysis of NET-PD LS1, increasing duration of MAO-B inhibitor exposure was significantly associated with less clinical decline as assessed by the Global Outcome (GO) measure. This was a robust finding and sensitivity analyses demonstrated that this result was insensitive to the definition of duration of MAO-B inhibitor exposure used. The model adjusted for potential confounders and results were similar whether or not we forced treatment assignment (creatine vs placebo), baseline LEDD, and change in LEDD into the model. In addition, the direction of the association was the same (increasing duration of exposure associated with less clinical decline) for four of the five components of the GO and the association was significant for three of the five components. Increasing duration of MAO-B inhibitor exposure was significantly associated with less progression on the modified Schwab and England activities of daily living (ADL) scale, the Ambulatory Capacity (AC) scale, and the modified Rankin scale for global disability. Additional analyses indicated that a one year increase in cumulative duration of MAO-B inhibitor exposure was associated with benefit equivalent in magnitude to 17% of the annual decline in ADL (Schwab and England), 24% of the annual decline in ambulatory capacity (AC), and 19% of the annual decline in global disability (Rankin) observed in participants not receiving an MAO-B inhibitor. In addition to less clinical decline, increasing duration of MAO-B inhibitor exposure was significantly associated with less increase in levodopa equivalent daily dosage (LEDD).

Our study has several important limitations. The key limitation is that we performed a post-hoc analysis and this was not a randomized, blinded, prospective trial with participants randomized to treatment with an MAO-B inhibitor vs placebo. Although we sought to identify confounders and adjust for them, it remains possible that there are confounders of which we are unaware. MAO-B inhibitor exposure may also be an indicator of participants receiving better care, including such interventions as physical therapy, exercise, or counseling. It could also be that participants receiving MAO-B inhibitors were more engaged in their disease management and took a more active role in their health care. In addition, MAO-B inhibitor use was not blinded in the NET-PD LS-1 trial. It is possible that treating physicians used MAO-B inhibitors more commonly in participants who were doing well or stopped MAO-B inhibitors in participants who were doing poorly. We also note that we relied on patient reports of medication usage at annual visits and cannot exclude the possibility of recall bias, underreporting, or other inaccuracies. We also acknowledge that the association between MAO-B inhibitor use and less clinical decline may or may not be long-lasting. The length of follow-up and the cumulative duration of MAO-B inhibitor use varied across participants, and some of the participants only used it for a short time. Therefore, it is possible that part of the detected association is attributable to a short-term benefit. Nevertheless, since the mean follow-up time was 4.1 years, we believe that the detected association is not dominated by a short-term benefit of MAO-B inhibitor use.

Strengths of our study include the fact that we analyzed data from NET-PD LS1, one of the largest and longest, prospective trials in PD to date. Participants were treated as deemed appropriate by their physicians, similar to “naturalistic” studies. In addition, the GO provides a multidimensional assessment of disease progression [8 –11] and was chosen for use in NET-PD LS1 (in the form of the GST) as an appropriate clinical measure of disease progression.

Why we did not observe any trend between increasing duration of MAO-B inhibitor exposure and less decline on the Symbol Digit Modalities Test (SDM) is a matter of speculation. Although the SDM is frequently used in clinical trials as a measure of cognitive abilities and provides a measure of attention, it does not provide information about other individual cognitive domains or global cognition. Thus, we cannot exclude the possibility of an association between duration of MAO-B inhibitor exposure and slowing of other aspects of cognitive decline. In addition, our analytic sample was relatively young, well educated, early in their disease, and cognitive impairment and progression was generally limited. To the extent that cognitive decline did occur, it is possible that it was in part unrelated to progression of PD but rather due to co-morbid vascular or Alzheimer’s pathology in some participants. Lastly, it is possible that symptomatic treatment with cognitive enhancers (cholinesterase inhibitors, memantine) mitigated our ability to detect an association; cognitive enhancers were used by 10.2% (85/832) of participants who never received an MAO-B inhibitor and by 6.4% (50/784) of participants who did receive an MAO-B inhibitor.

Our overall findings are consistent with several prospective, randomized, placebo controlled trials. Larsen et al. [7] conducted a trial in which 163 early PD patients were randomized to blinded treatment with selegiline or placebo and treated with levodopa that was adjusted as clinically indicated. After 3 months, mean total UPDRS scores were 3.3 units more improved in the selegiline group than in the placebo group while both groups were receiving ∼350 mg levodopa/day. At 5 years, in the selegiline group, mean total UPDRS scores were 11.5 units more improved (p < 0.01) and mean daily levodopa dose was 16% lower (p < 0.05) compared to the placebo group. Similarly, Pålhagen et al. [6] randomized 157 early, untreated PD patients to blinded treatment with selegiline 10 mg/day or placebo, to which levodopa could be added and adjusted as necessary. After 12 months of combination therapy, in the selegiline group, mean total UPDRS scores were 2.9 units more improved and mean daily levodopa doses were 4.5% lower. After 5 years of combination therapy, in the selegiline group, mean total UPDRS scores were 9.9 units more improved (p = 0.07), UPDRS ADL scores were significantly better (p < 0.05), and mean daily levodopa doses were 16% lower (p = 0.0002) compared to the placebo group. In each of these trials, increasing benefit for UPDRS scores was observed with increasing duration of MAO-B inhibitor exposure compared to placebo even as the placebo groups received higher daily levodopa dosages. Similar to our analytic sample, both of these trials involved patients who were treated with dopaminergic medication, raising the possibility that that slowing of clinical decline associated with an MAO-B inhibitor may be most robust in patients treated with dopaminergic therapy, although this is speculative.

Our results should be interpreted cautiously given the limitations of our methodology. Nonetheless, our findings are consistent with the hypothesis that MAO-B inhibitors slow PD clinical decline. A long-term simple study similar to NET-PD LS1 with participants randomized to an MAO-B inhibitor vs placebo and treated with dopaminergic therapy that can be adjusted as clinically indicated should be considered to definitively evaluate the potential long-term benefits of MAO-B inhibitors.

CONFLICT OF INTEREST

Dr. Hauser reports grants from NIH NINDS, during the conduct of the study; personal fees from Teva Pharmaceuticals, personal fees from USB Biosciences, personal fees from AbbVie, personal fees from Eli Lilly, personal fees from Novartis, personal fees from Biotie Therapies, personal fees from Lundbeck, personal fees from Pfizer, personal fees from Allergan Neuroscience, personal fees from Neurocrine Biosciences, personal fees from Chelsea Therapeutics, personal fees from Auspex, personal fees from Acadia Pharmaceuticals, personal fees from Michael J Fox Foundation, personal fees from GLG, personal fees from AstraZeneca, personal fees from Acorda Therapeutics, personal fees from Impax Pharmaceuticals, personal fees from Cynapsus Therapeutics, personal fees from Cowan Group, personal fees from Guidepoint Global, grants from National Parkinson Foundation, other from University of South Florida, personal fees from USWorldMeds, personal fees from Neuropore, personal fees from Prexton, outside the submitted work.

Dr. Li reports grants from NIH NINDS, during the conduct of the study.

Dr. Pérez reports grants from NIH NINDS, during the conduct and analysis of the study.

Ms. Ren reports grants from NIH NINDS, during the conduct of the study;.

Dr. Weintraub has nothing to disclose.

Dr. Elm reports grants from NINDS, during the conduct of the study; personal fees from TEVA, outside the submitted work.

Dr. Goudreau reports grants from NIH NINDS, during the conduct of the study. Participated in Teva Speaker’s Bureau and advisory panel discussions.

Dr. Morgan reports grants from NIH NINDS, during the conduct of the study; grants and personal fees from Impax, grants and personal fees from Acorda, grants and personal fees from Teva, grants and personal fees from Abbvie, grants and personal fees from Acadia, personal fees from UCB, grants and personal fees from Cynapsus, grants and personal fees from Lundbeck, grants and personal fees from National Parkinson Foundation, grants from Biotie, grants from CHDI, grants from Parkinson Study Group, outside the submitted work.

Dr. Fang reports grants from NIH NINDS, during the conduct of the study.

Dr. Aminoff reports grants from NIH NINDS, during the conduct of the study.

Dr. Christine reports grants from NIH NINDS, during the conduct of the study.

Dr. Dhall reports grants from NIH NINDS, during the conduct of the study; personal fees from Teva Pharmaceuticals, personal fees from Acadia Pharmaceuticals, personal fees from UCB Pharma, personal fees from Lundbeck Pharmaceuticals, personal fees from Impax Pharmaceuticals, outside the submitted work.

Dr. Umeh reports grants from NIH NINDS, during the conduct of the study.

Dr. Boyd reports grants from NIH NINDS, during the conduct of the study.

Dr. Stover reports grants from NIH NINDS, during the conduct of the study; other from Serina Therapeutics, other from Biotie Therapeutics, other from the Michael J Fox Foundation, personal fees from Rush Medical Center, personal fees from US World Meds outside the submitted work.

Dr. Leehey reports grants from NIH NINDS,during the conduct of the study; other from Colorado Department of Public Health and Environment, other from Adamas Pharmaceuticals, other fromUS World Meds LLC, other from Pharma Two B, from US World Meds LLC, outside the submitted work.

Dr. Zweig reports grants from NIH NINDS, during the conduct of the study;

Dr. Nicholas reports grants from the NIH NINDS during the conduct of the study; other from the Fox foundation, other from Adamas, other from Cynapsus; personal fees from UCB, outside the submitted work.

Dr. Bodis-Wollner reports grants from NIH NINDS, during the conduct of the study.

Dr. Willis reports grants from NIH NINDS, during the conduct of the study.

Dr. Kieburtz reports grants from NIH NINDS, during the conduct of the study; personal fees from Acorda, personal fees from AstraZeneca, personal fees from Biotie, personal fees from CHDI, personal fees from Clintrex, personal fees from Cynapsus, personal fees from Intec, personal fees from Lundbeck, personal fees from Medivation, personal fees from Orion, personal fees from Otsuka, personal fees from Pharma2B, personal fees from Roche/Genentech, personal fees from Synagile, grants from Teva, personal fees from Upsher-Smith, personal fees from US WorldsMeds, personal fees from Vaccinex, grants from Michael J Fox Foundation, personal fees from The United States –National Institutes of Health (NIH, NINDS), personal fees from Astellas Pharma, personal fees from BioMarin Pharmaceutica, personal fees from Britannia, personal fees from Clearpoint Strategy Group, personal fees from Corium International, personal fees from Forward Pharma, personal fees from Genzyme, personal fees from INC Research, personal fees from Melior Discovery, personal fees from Neuroderm, personal fees from Neurmedix, personal fees from Pfizer, personal fees from Prana Biotechnology, personal fees from Prothena/Neotope/Elan Pharmaceutical, personal fees from Raptor Pharmaceuticals, personal fees from Remedy Pharmaceuticals, personal fees from Sage Bionetworks, personal fees from Sanofi, personal fees from Serina, personal fees from Sunovion, personal fees from Titan, personal fees from Vertex Pharmaceuticals, personal fees from Weston Brain Institute, outside the submitted work;.

Dr. Tilley reports grants from NIH NINDS, during the conduct of the study.

Footnotes

ACKNOWLEDGMENTS

This analysis and the study on which it is based (NET-PD LS1) were sponsored by NIH NINDS.

Appendix

References

Magyar

, Pálfi

, Jenei

& Szöko

(2006) Deprenyl: From chemical synthesis to neuroprotection. J Neural Transm Suppl, 71, 143–156.

Weinreb

, Amit

, Riederer

, Youdim

& Mandel

(2011) Neuroprotective profile of the multitarget drug rasagiline in Parkinson’s disease. Int Rev Neurobiol, 100, 127–149.

Sadeghian

, Mullali

, Pocock

, Piers

, Roach

& Smith

(2016) Neuroprotection by safinamide in the 6-hydroxydopamine model of Parkinson’s disease. Neuropathol Appl Neurobiol, 42, 423–435.

The Parkinson Study Group (1989) Effect of deprenyl on the progression of disability in early Parkinson’s disease. N Engl J Med, 321, 1364–1371.

Olanow

, Rascol

, Hauser

, Feigin

, Jankovic

, Lang

, Langston

, Melamed

, Poewe

, Stocchi

, Tolosa

& ADAGIO

Study Investigators

(2009) A double-blind, delayed-start trial of rasagiline in Parkinson’s disease. N Engl J Med, 361, 1268–1278.

Pålhagen

, Heinonen

, Hägglund

, Kaugesaar

, Mäki-Ikola

, Palm

& Swedish Parkinson Study Group (2006) Selegiline slows the progression of the symptoms of Parkinson disease. Neurology, 66, 1200–1206.

Larsen

, Boas

& Erdal

(1999) Does selegiline modify the progression of early Parkinson’s disease? Results from a five-year study. Eur J Neurol, 6, 539–547.

Writing Group for the NINDS Exploratory Trials in Parkinson Disease (NET-PD) Investigators, Kieburtz

, Tilley

, Elm

, Babcock

, Hauser

, Ross

, Augustine

, Aminoff

, Bodis-Wollner

, Boyd

, Cambi

, Chou

, Christine

, Cines

, Dahodwala

, Derwent

, Dewey

RB Jr

, Hawthorne

, Houghton

, Kamp

, Leehey

, Lew

, Liang

, Luo

, Mari

, Morgan

, Parashos

, Pérez

, Petrovitch

, Rajan

, Reichwein

, Roth

, Schneider

, Shannon

, Simon

, Simuni

, Singer

, Sudarsky

, Tanner

, Umeh

, Williams

& Wills

(2015) Effect of creatine monohydrate on clinical progression in patients with Parkinson disease: A randomized clinical trial. JAMA, 313, 584–593.

Elm

& NINDS NET-PD Investigators (2012) Design innovations and baseline findings in a long-term Parkinson’s trial: The National Institute of Neurological Disorders and Stroke Exploratory Trials in Parkinson’s Disease Long-Term Study-1. Mov Disord, 27, 1513–1521.

10.

Huang

, Goetz

, Woolson

, Tilley

, Kerr

, Palesch

, Elm

, Ravina

, Bergmann

, Kieburtz

& Parkinson Study Group (2009) Using global statistical tests in long-term Parkinson’s disease clinical trials. Mov Disord, 24, 1732–1739.

11.

Bega

, Kim

, Zhang

, Elm

, Schneider

, Hauser

, Fraser

, Simuni

& on behalf of the NET-PD LS1 Investigators (2015) Predictors of functional decline in early Parkinson’s disease: NET-PD LS1 Cohort. J Parkinsons Dis, 5, 773–782.

12.

Schwab

& England

(1969) Projection technique for evaluating surgery in Parkinson’s disease. In Third Symposium on Parkinson’s Disease, Gillingham

, Donaldson

, eds. E & S Livingstone, Edinburgh Scotland, pp. 152–157.

13.

Peto

, Jenkinson

, Fitzpatrick

& Greenhall

(1995) The development and validation of a short measure of functioning and well being for individuals with Parkinson’s disease. Qual Life Res, 4, 241–248.

14.

Fahn

, Elton

& UPDRS Development Committee (1987) Unified Parkinson’s Disease Rating Scale. In Recent Developments in Parkinson’s Disease, Marsden

, Calne

, Goldstein

, eds. Macmillan Healthcare Information, Florham Park, NJ, pp. 153–163.

15.

Smith

(2002) Symbol Digit Modalities Test Manual, Western Psychological Services, Los Angeles, CA.

16.

van Swieten

, Koudstaal

, Visser

, Schouten

& van Gijn

(1988) Interobserver agreement for the assessment of handicap in stroke patients. Stroke, 19, 604–607.

17.

Tomlinson

, Stowe

, Patel

, Rick

, Gray

& Clarke

(2010) Systematic review of levodopa dose equivalency reporting in Parkinson’s disease. Mov Disord, 25, 2649–2653.

18.

Wills

, Pérez

, Wang

, Su

, Morgan

, Rajan

, Leehey

, Pontone

, Chou

, Umeh

, Mari

, Boyd

& for the NINDS Exploratory Trials in Parkinson Disease (NET-PD) Investigators (2016) Association between change in body mass index, Unified Parkinson Disease Rating Scale scores, and survival among persons with Parkinson disease. Secondary analysis of longitudinal data from NINDS Exploratory Trials in Parkinson Disease Long-term Study 1. JAMA Neurol, 73, 321–328.

19.

Guo

& Fraser

(2014) Propensity Score Analysis: Statistical Methods and Applications, Second Edition. SAGE publications, pp. 448.

20.

The National Institute of Neurological Disorders and Stroke (NINDS) rt-PA Stroke Study Group (1997) A systems approach to immediate evaluation and management of hyperacute stroke. Experience at eight centers and implications for community practice and patient care. Stroke, 28, 1530–1540.

Longer Duration of MAO-B Inhibitor Exposure is Associated with Less Clinical Decline in Parkinson’s Disease: An Analysis of NET-PD LS1

Abstract

Keywords

INTRODUCTION

METHODS

Outcome measure

Analytic sample

Definition of duration of MAO-B inhibitor exposure

Statistical analyses

RESULTS

Analytic sample

Cumulative duration of MAO-B inhibitor exposure

Confounder selection and adjustments

Primary Analysis –Association between cumulative duration of MAO-B inhibitor exposure and the GO

Association between cumulative duration of MAO-B inhibitor exposure and individual components of the GO

Sensitivity analyses

Association between the cumulative duration of MAO-B inhibitor exposure (in years) and change in LEDD

DISCUSSION

CONFLICT OF INTEREST

Footnotes

ACKNOWLEDGMENTS

Appendix

References