Trial objective: adjunctive tavapadon in levodopa-treated adults with motor fluctuations

“To evaluate the efficacy, safety, and tolerability of tavapadon as adjunctive therapy to oral levodopa in adults with PD experiencing motor fluctuations.” (opens the source at this quote in a new tab)
“The Flexible-Dose, Adjunctive Therapy Trial in Adults With Parkinson Disease With Motor Fluctuations (TEMPO-3) was designed to evaluate the efficacy, safety, and tolerability of flexible-dose tavapadon (5-15 mg once daily) as an adjunctive therapy for PD in levodopa-treated adults experiencing off-time fluctuations.” (opens the source at this quote in a new tab)

Research question stated by the investigators

“Can adjunctive tavapadon—an oral, once-daily, selective dopamine (D) D1/D5 agonist—improve motor control for people with Parkinson disease (PD) experiencing motor fluctuations while minimizing risk of adverse events?” (opens the source at this quote in a new tab)

Registry statement of purpose

“The purpose of this study is to assess the effect of tavapadon on the change from baseline in total daily hours of "on" time without troublesome dyskinesia in L-Dopa-treated participants with Parkinson's Disease (PD) who are experiencing motor fluctuations.” (opens the source at this quote in a new tab)

Protocol rationale for the trial

“By targeting the D1R subtype, tavapadon aims to improve motor control symptoms while minimizing adverse events (AEs) that may be mechanistically linked to D2/D3 dopamine receptor agonists (notably, dose-limiting hypotension, impulse control disorders, sleep disorders, and, potentially, some forms of hallucinations).” (opens the source at this quote in a new tab)
“This prospective, Phase 3, multicenter, multinational, randomized, double-blind, placebo-controlled, parallel-group, 27-week trial will evaluate the efficacy, safety, tolerability, and pharmacokinetics (PK) of flexible doses of tavapadon (5 mg to 15 mg once daily [QD]) as adjunctive therapy for PD in levodopa (L-Dopa)-treated adults with motor fluctuations to further define and confirm the magnitude of efficacy and risk/benefit profile of this target dose range.” (opens the source at this quote in a new tab)

Summary: sites, countries, and enrollment period

The primary publication reports 148 sites in 14 countries and a trial period of September 2020 to February 2024. The registry record for NCT04542499 lists 146 locations in the same 14 countries: the United States (54), Poland (13), Germany (11), Spain (11), Italy (10), Ukraine (8), Bulgaria (7), France (6), Israel (6), Australia (5), Czechia (5), Hungary (4), Serbia (4), and Canada (2). The registry gives an actual start date of September 23, 2020, an actual primary completion date of January 29, 2024, and an actual study completion date of February 15, 2024. The protocol (CVL-751-PD-003) names Cerevel Therapeutics, LLC as sponsor; the registry now lists AbbVie.

Number of sites, countries, and study period

“This was a phase 3, double-blind, placebo-controlled randomized clinical trial conducted between September 2020 and February 2024 with a 4-week safety follow-up. Participants were recruited from 148 sites across 14 countries.” (opens the source at this quote in a new tab)
“The trial was conducted from September 2020 to February 2024. Participation lasted approximately 35 weeks, including a 4-week maximum screening period, a 27-week treatment period, and a 4-week safety follow-up period (Figure 1A).” (opens the source at this quote in a new tab)

Randomized, double-blind, placebo-controlled, parallel-group design

“The TEMPO-3 trial was a prospective, phase 3, multicenter, double-blind, placebo-controlled, parallel-group, 27-week randomized clinical trial conducted at 148 clinical sites across 14 countries (the trial protocol and statistical analysis plan are available in Supplement 1 and Supplement 2, respectively).” (opens the source at this quote in a new tab)
“The trial was conducted in accordance with the Declaration of Helsinki, International Council for Harmonisation Good Clinical Practice guidelines, and was approved by independent ethics committees before initiation; participants or their representatives provided written informed consent before screening and enrollment. This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guidelines.” (opens the source at this quote in a new tab)

Registered official title and masking

“A Phase 3, Double-Blind, Randomized, Placebo-Controlled, Parallel-Group, Flexible-Dose, 27-Week Trial to Evaluate the Efficacy, Safety, and Tolerability of Tavapadon as Adjunctive Therapy for Parkinson's Disease in Levodopa-Treated Adults With Motor Fluctuations (TEMPO-3 Trial)” (opens the source at this quote in a new tab)

Maintenance of blinding

“Participants, investigators, and all blinded sponsor personnel remained blinded until final database lock.” (opens the source at this quote in a new tab)
“All participants took 3 tablets of tavapadon, placebo, or a combination of tavapadon and placebo once daily during the treatment period; all tablets and packaging were identical in appearance for blinded dose adjustment.” (opens the source at this quote in a new tab)

Scientific rationale for the design stated in the protocol

“The randomized, double-blind, placebo-controlled, parallel-group trial design is widely accepted as one that minimizes the risk of bias and that is appropriate for evaluating the effects of a trial treatment in indications in which use of a placebo is ethical (US FDA, 2001).” (opens the source at this quote in a new tab)

Visit schedule, diary collection, and entry into the open-label extension

“Assessments will be conducted in the clinic at the end of Weeks 2, 5, 8, 11, 14, 18, 22, 26, and 27. Subjects will complete the Hauser diary on 2 consecutive days within the 7 days before each scheduled postbaseline clinic visit.” (opens the source at this quote in a new tab)
“Subjects who complete through Week 27 of the trial may have the opportunity to enter an open-label extension trial (Protocol CVL-751-PD-004).” (opens the source at this quote in a new tab)

Design described in the prescribing information

“The efficacy of JUVMO for the treatment of PD as adjunctive therapy to levodopa was established in a 27-week, double-blind, randomized, placebo-controlled, parallel-group, flexible-dose trial, Study 3 (NCT04542499).” (opens the prescribing information at this quote)

Place of the trial in the phase 3 program

“The TEMPO clinical development program is evaluating the efficacy, safety and tolerability of tavapadon across a broad Parkinson’s population, including two monotherapy Phase 3 trials (TEMPO-1 and TEMPO-2) and one adjunctive Phase 3 trial (TEMPO-3). Cerevel is also conducting a fourth, open-label extension (OLE) trial (TEMPO-4) to assess the long-term safety and tolerability of tavapadon.” (opens the source at this quote in a new tab)

Key inclusion criteria reported in the primary publication

“Eligible participants included adults aged 40 to 80 years with a diagnosis of PD consistent with the UK Parkinson Disease Society Brain Bank criteria, a modified Hoehn & Yahr score of 2, 2.5, or 3 in the on-state (ie, good mobility, smoother movement, fewer motor/nonmotor symptoms), a minimum of 2.5 hours of daily off-time on 2 consecutive days during the screening period (ie, return/worsening of motor/nonmotor symptoms often experienced between medication doses), a minimum stable dose of 400 mg of levodopa per day divided at least 3 or 4 times daily, and a good response to levodopa per investigator judgment.” (opens the source at this quote in a new tab)
“Participants were required to demonstrate an acceptable ability to complete the Hauser diary” (opens the source at this quote in a new tab)
“on concordance testing during the screening period, and to maintain the same levodopa dose and frequency for the trial duration.” (opens the source at this quote in a new tab)

Key exclusion criteria and permitted concomitant medications

“Exclusion criteria included atypical or secondary parkinsonian syndrome, previous or planned surgical intervention for PD, Montreal Cognitive Assessment score less than 26, history of or current diagnosis of a clinically significant impulse control disorder, history of psychosis or hallucinations for 12 months or less, or current use (≤14 days of baseline) of DAs.” (opens the source at this quote in a new tab)
“Previous and concurrent use of catechol-O-methyl transferase inhibitors, monoamine oxidase inhibitor B inhibitors, amantadine, istradefylline, or anticholinergic drugs was permitted if initiated more than 90 days before the baseline visit; dosage changes for concurrent PD medications were not permitted.” (opens the source at this quote in a new tab)

Levodopa regimen and diary requirements as registered

“Participants who return a completed self-reported home diary for motor function status (Hauser diary) during the screening period (after diary training and concordance testing has occurred), with recordings for 2 consecutive days (ie, 2 consecutive 24-hour periods) showing at least 2 and half hours of "off" time on each of the 2 days.” (opens the source at this quote in a new tab)
“Participants who are on a stable dose of L-Dopa for at least 4 weeks prior to screening and are taking a minimum total daily dose of 400 milligram (mg) divided in at least 4 doses per day of standard carbidopa/levodopa or divided in at least 3 doses per day of extended-release carbidopa/levodopa capsules. The carbidopa/levodopa dose and frequency must be maintained for the duration of the trial.” (opens the source at this quote in a new tab)

Registered exclusion criteria: psychiatric, cardiovascular, and drug interaction conditions

“Participants with a history of psychosis or hallucinations within the previous 12 months.” (opens the source at this quote in a new tab)
“Participants who are currently receiving moderate or strong CYP3A4 inducers or CYP3A4 inhibitors(except for topical administration).” (opens the source at this quote in a new tab)
“Participants with clinically significant orthostatic hypotension (eg, syncope).” (opens the source at this quote in a new tab)
“Participants with moderate or severe renal impairment (creatinine clearance as estimated by Cockcroft-Gault formula <30 mL/min or on dialysis).” (opens the source at this quote in a new tab)

Randomization, flexible dosing, and fixed levodopa dose

“Participants were randomized 1:1 to receive flexible-dose tavapadon (5-15 mg once daily) or placebo as adjunct to levodopa, using a computer-generated randomization scheme.” (opens the source at this quote in a new tab)
“Tavapadon was gradually titrated to the maximum tolerated dose (up to 15 mg once daily) over the first 14 weeks of the trial during the dose titration and dose adjustment phases (eTable 1 in Supplement 3). Further adjustments during the maintenance phase were not permitted. For both treatment groups, the levodopa dose was fixed, without dose adjustments for the duration of the trial.” (opens the source at this quote in a new tab)

Tavapadon and placebo dosing schedule

Trial day	Titration step	Blinded treatment assignment
Blinded dose titration phase	“Days 1-4”	“0.25 mg of tavapadon or placebo QD”
Blinded dose titration phase	“Days 5-8”	“0.5 mg of tavapadon or placebo QD”
Blinded dose titration phase	“Days 9-12”	“0.75 mg of tavapadon or placebo QD”
Blinded dose titration phase	“Days 13-16”	“1 mg of tavapadon or placebo QD”
Blinded dose titration phase	“Days 17-20”	“1.5 mg of tavapadon or placebo QD”
Blinded dose titration phase	“Days 21-24”	“2.25 mg of tavapadon or placebo QD”
Blinded dose titration phase	“Days 25-40”	“3 mg of tavapadon or placebo QD”
Blinded dose titration phase	“Days 41-56”	“5 mg of tavapadon or placebo QD”
Blinded dose adjustment phase	“Days 57-72”	“10 mg of tavapadon or placebo QD”
Blinded dose adjustment phase	“Days 73-104”	“15 mg of tavapadon or placebo QD”
Blinded maintenance phase	“Days 105-189”	“Maximum tolerated dose of tavapadon (5-15 mg) or placebo QD”

Maintenance doses achieved

“Participants received the highest tolerated dose level that was achieved during the dose adjustment phase (5 mg, 22 participants [8.8%]; 10 mg, 32 participants [12.7%]; 15 mg, 169 participants [67.3%]). No adjustment in the tavapadon dose was allowed during the maintenance phase.” (opens the source at this quote in a new tab)

Dose reduction and rechallenge rules

“A reduction in dose (from 15 mg QD to 10 mg QD or from 10 mg QD to 5 mg QD) will be allowed based on individual subject tolerability. Subjects who require a dose reduction may be rechallenged with the higher dose, at the discretion of the investigator, to address symptomatic needs.” (opens the source at this quote in a new tab)
“Subjects will receive the highest tolerated dose level that is achieved during the Dose Adjustment Phase for the duration of the subsequent 13-week Maintenance Phase. Adjustments in the dose of tavapadon will not be allowed during the Maintenance Phase. Subjects who cannot tolerate their maintenance dose will be discontinued from the trial.” (opens the source at this quote in a new tab)

Justification for the 5 mg to 15 mg dose range

“These data suggest that doses <5 mg QD are unlikely to deliver clinically meaningful benefit over placebo in motor symptom control to most patients with PD.” (opens the source at this quote in a new tab)
“The safety and pharmacokinetic data from this Phase 1 trial suggest that doses >15 mg QD are unlikely to have an optimal risk/benefit profile for patients with PD.” (opens the source at this quote in a new tab)
“Taken together, these findings suggest that an optimal dose range of tavapadon for the treatment of PD may reasonably be between 5 and 15 mg QD for the individual patient, based on symptomatic needs and tolerability thresholds.” (opens the source at this quote in a new tab)

Dosing described in the prescribing information

“Patients in Study 3 were on a fixed dose of levodopa and were titrated to JUVMO 15 mg once daily unless prevented by intolerance, resulting in maintenance doses between 5 mg and 15 mg (n=251), compared to placebo (n=254).” (opens the prescribing information at this quote)

Primary, key secondary, and other secondary endpoints

“The primary end point was the change from baseline to week 26 in total daily on time without troublesome dyskinesia (referred to as good-on-time) based on the 2-day average of the Hauser diary.” (opens the source at this quote in a new tab)
“The key secondary end point was change from baseline to week 26 in total daily off-time. Other secondary end points included change from baseline in good-on-time and off-time at all time points, change from baseline in Movement Disorder Society-Unified Parkinson Disorder Rating Scale (MDS-UPDRS) parts I, II, and III individual and combined scores (week 26 and all time points), Parkinson Disease Questionnaire (PDQ) 39 score,” (opens the source at this quote in a new tab)
“and the EuroQol 5-Dimension 5-Level (EQ-5D-5L) index and visual analog scale (VAS) scores.” (opens the source at this quote in a new tab)
“MDS-UDPRS assessments were conducted approximately 2 to 3 hours after the last dose of levodopa, and participant scores could have been collected during on or off status.” (opens the source at this quote in a new tab)

Exploratory, post hoc, and safety end points

“Change from baseline in MDS-UDPRS parts II+III scores was a predefined exploratory end point. Additionally, a post hoc analysis evaluated changes from baseline to week 26 (the end of treatment) in additional motor states (eg, on-time with nontroublesome dyskinesia, on-time without dyskinesia) across the normal waking day as recorded in the Hauser diary.” (opens the source at this quote in a new tab)
“Safety end points included AEs, clinical laboratory values, vital signs (including orthostatic values), and electrocardiograms. Columbia-Suicide Severity Rating Scale (C-SSRS),” (opens the source at this quote in a new tab)
“Epworth Sleepiness Scale (ESS),” (opens the source at this quote in a new tab)
“and Questionnaire for Impulsive-Compulsive Disorders in Parkinson Disease-Rating Scale (QUIP-RS)” (opens the source at this quote in a new tab)
“scores were also assessed at each visit.” (opens the source at this quote in a new tab)

Registered primary outcome measure and diary categories

“Change From Baseline in the Total "On" Time Without Troublesome Dyskinesia Based on the 2-day Average of the Self-completed Home Diary for Motor Function Status (Hauser Diary)” (opens the source at this quote in a new tab)
“The Hauser diary asks participants to rate their mobility for each 30-minute period and to record their status for the majority of the period in 1 of 5 categories as: "on" time without dyskinesia, "on" time with nontroublesome dyskinesia, "on" time with troublesome dyskinesia, "off" time, or asleep.” (opens the source at this quote in a new tab)

Registered MDS-UPDRS secondary outcome and score ranges

“The MDS-UPDRS rating tool was used to follow longitudinal course of Parkinson's Disease. It was made up of 4 parts: Part 1: Non-motor aspects of experiences of daily living (13 items. Score range: 0-52); Part 2: Motor aspects of experiences of daily living (13 items. Score range: 0-52); Part 3: Motor examination (18 items. Score range: 0-132); Part 4: Motor complications (6 items. Score range: 0-24. Part 4 was not collected in this trial). Each item has 0-4 rating on scale from 0 (normal) to 4 (severe). Higher values represent a worse outcome.” (opens the source at this quote in a new tab)

Protocol rationale for the Hauser diary and the clinically important difference in off time

“The minimum clinically important difference (smallest change that is meaningful to patients) for a reduction in “off” time (key secondary efficacy endpoint) has been reported to be 1.0 hour (Hauser et al, 2011; Hauser et al, 2014).” (opens the source at this quote in a new tab)
“The MDS-UPDRS is a comprehensive assessment that is designed to monitor the burden and extent of PD across the longitudinal disease course and provides a clinical endpoint in therapy trials (Goetz et al, 2008).” (opens the source at this quote in a new tab)

Diary recording procedure in the protocol

“Subjects will record diary entries at home at 30-minute intervals for 24 hours on 2 consecutive days within the 7 days before randomization at the Baseline Visit and at 30-minute intervals for 24 hours on 2 consecutive days within the 7 days before each postbaseline clinic visit.” (opens the source at this quote in a new tab)

Phase 2 effect size that informed the trial

“In a Phase 2 trial of tavapadon in L-Dopa-treated subjects with PD who were experiencing motor fluctuations (Trial B7601003), analysis of the change from baseline in average daily “on” time without troublesome dyskinesia showed that the magnitude of benefit of tavapadon 15 mg QD, relative to placebo, was approximately 1 hour at Week 10.” (opens the source at this quote in a new tab)

Development of the diary: dyskinesia categories and patient-defined good and bad time

“In both studies, patients concurrently completed a test and a reference diary.” (opens the source at this quote in a new tab)
“In Study II, we evaluated a new home diary designed to separate dyskinesia that had a negative impact on patient-defined functional status from dyskinesia that did not. There were 816 evaluable time periods from 17 patients; 84.9% of off time and 89.9% of on time with troublesome dyskinesia was considered bad time while 85.5% of on time without dyskinesia and 93.8% of on time with nontroublesome dyskinesia was considered good time.” (opens the source at this quote in a new tab)
“In evaluating the efficacy of an intervention, assessment of change in off time and change in on time with troublesome dyskinesia provides a more accurate reflection of clinical response than change in off time alone.” (opens the source at this quote in a new tab)

Validation of good on time as a trial outcome

“In the current study, PD patients with motor fluctuations and dyskinesia (present more than 25% of the awake day and at least moderately disabling) completed daily diaries on 3 consecutive days in each of 2 consecutive weeks.” (opens the source at this quote in a new tab)
“Three hundred two patients from 10 countries participated.” (opens the source at this quote in a new tab)
“Mean age was 65.5 years (range, 37–89 years), and 48% of patients were women.” (opens the source at this quote in a new tab)
“They were instructed to place a tick mark on a daily diary card every 30 minutes reflecting their predominant status over the prior half-hour period.” (opens the source at this quote in a new tab)
“Seventy-six percent of the missing or duplicate entries occurred after Day 3.” (opens the source at this quote in a new tab)
“Coefficients of reliability as calculated by Cronbach's alpha were as follows: 2 days, r = 0.806; 3 days, r = 0.868; 4 days, r = 0.918; 5 days, r = 0.934; 6 days, r = 0.946.” (opens the source at this quote in a new tab)
“Mean percentages of the awake day (± SD) for each of the diary categories for all patients over the 6 days were as follows: off, 24.7 ± 15.2; on without dyskinesia, 27.4 ± 19.8; on with nontroublesome dyskinesia, 26.7 ± 14.9; and on with troublesome dyskinesia, 21.2 ± 16.0. Mean percent of the awake day spent in good on time (ONG%) was 54.1 ± 19.1.” (opens the source at this quote in a new tab)
“Mean ONG% was observed to be very stable in time (repeated measurement ANOVA, P = 0.9939).” (opens the source at this quote in a new tab)
“The mean correlation for any pair of days was r = 0.74 (SD = 0.06, range = 0.65,0.85). Between any 2 consecutive days, the mean correlation was r = 0.79 (SD = 0.04, range = 0.71,0.82).” (opens the source at this quote in a new tab)
“Under these conditions, the point estimate ICC was 0.715, and the 95% one-sided lower limit was found to be 0.710. The standard error of measurement (SEM) was calculated to be 10.75%.” (opens the source at this quote in a new tab)
“Sample size calculations ranged from N = 42 (4 days of diaries at baseline and endpoint, alpha = 0.05, beta = 0.2, one tail) to N = 99 (2 days of diaries at baseline and endpoint, α = 0.05, β = 0.1, two tails).” (opens the source at this quote in a new tab)
“Of note, the VAS response to the question “How much of the day today did you experience a good response?” more strongly correlated with ONG% (0.41) than ON% (0.24).” (opens the source at this quote in a new tab)
“The maximum error (at the 95% level) as determined by the standard error of measurement was 15.2% for 2 days of diaries, 12.4% for 3 days, and 10.8% for 4 days.” (opens the source at this quote in a new tab)
“Test-retest reliability was good, and reliability increased with increasing number of diary days but compliance diminished beyond 3 days. Good on time (ONG = on time without dyskinesia or with nontroublesome dyskinesia) most strongly correlated with patients' perceived duration of a good response through the day and is an important outcome variable.” (opens the source at this quote in a new tab)
“We anticipate that most clinical trials will use 2 to 4 days of diaries for each assessment point. Having more diary days provides greater reliability but increases cost and effort and potentially reduces compliance. This diary provides good test–retest reliability at 2 days and increasing reliability with increasing number of days. However, most of the diary errors occurred after 3 days. Perhaps the most important determinant in selecting the number of diary days is the anticipated efficacy of the intervention, as more diary days allows detection of smaller changes.” (opens the source at this quote in a new tab)
“Although patients received training regarding the definitions of the various diary designations, the study did not include external validation to assess the accuracy of patient responses.” (opens the source at this quote in a new tab)

Minimal important difference in good on time by number of diary days

Diary days	Minimal important difference in good on time, % of the awake day
1	“21.5”
2	“15.2”
3	“12.4”
4	“10.8”
5	“9.6”
6	“8.8”

Minimal clinically important change in off time: rasagiline trial data

“PRESTO included 472 levodopa-treated PD subjects with motor fluctuations who were randomized to rasagiline, 0.5 or 1.0 mg/day, or matching placebo and followed for 6 months.” (opens the source at this quote in a new tab)
“The PRESTO study utilized home diaries to capture subject-rated ‘‘on,’’ ‘‘off,’’ and asleep time every half hour for 3 days prior to the baseline, week 6, week 14, and week 26 visits.” (opens the source at this quote in a new tab)
“Subject and investigator CGI-I were determined at week 26; both were used for this analysis, however, we considered subject-rated CGI-I to be the better measure of clinically important change.” (opens the source at this quote in a new tab)
“The Spearman correlation coefficient between CGI-I and change in ‘‘off’’ time at week 26 for advanced PD subjects in the PRESTO study was 0.46 (all P < .001).” (opens the source at this quote in a new tab)
“Based on subject-rated CGI-I, mean (SD) changes in ‘‘off’’ time for subjects in an active treatment group of the PRESTO study were −1.9 (2.2) hours for those rated minimally improved and −0.5 (2.2) hours for those rated minimally worse. Corresponding values for subjects in the placebo group were −1.9 (2.7) and 0.3 (2.3) hours.” (opens the source at this quote in a new tab)
“The optimal cutoff for change in ‘‘off’’ time to distinguish rasagiline-treated subjects rated minimally improved versus unchanged was −1.2 hours; the optimal cutoff for placebo subjects was −1.8 hours.” (opens the source at this quote in a new tab)
“The optimal cutoff to distinguish rasagiline-treated subjects rated minimally worse versus unchanged was 0.2 hours; the optimal cutoff for placebo subjects was −0.7 hours.” (opens the source at this quote in a new tab)
“In general, values for no change, minimally improved, and minimally worse were very similar to those derived using subject-rated CGI-I.” (opens the source at this quote in a new tab)
“In addition, we found an MCIC for reduction in "off" time of 1.0 hours as defined by mean reduction in "off" time in active treated subjects self-rated as minimally improved on CGI-I minus mean reduction in "off" time in placebo-treated subjects self-rated as unchanged (1.9-0.9 hours).” (opens the source at this quote in a new tab)
“As defined by mean change in subjects rated minimally improved on CGI-I, we found the MCIC for reduction in ‘‘off’’ time in patients with motor fluctuations to be 1.9 hours in both actively and placebo-treated subjects. However, we also note that there was a mean reduction in ‘‘off’’ time of 0.9 hours in placebo-treated subjects and 0.7 hours in actively treated subjects self-rated as unchanged on the CGI-I. This suggests that there may be a greater placebo effect on diary-recorded ‘‘off’’ hours than on CGI-I (in both placebo and active treatment groups).” (opens the source at this quote in a new tab)
“Similarly, we found that a reduction in ‘‘off’’ time of 1.2 hours best distinguished active treatment subjects rated minimally improved from those rated unchanged.” (opens the source at this quote in a new tab)
“By definition, MCICs derived from ROC cutoffs are smaller than those derived using mean scores. Whether one approach is superior to the other to MCIC is unknown.” (opens the source at this quote in a new tab)
“We hypothesize that many methodological factors can influence determination of the MCIC, and a range of values is likely to emerge from multiple studies.” (opens the source at this quote in a new tab)
“There are multiple limitations of our analyses. There could potentially be bias in the external anchor, in this case, the CGI-I.” (opens the source at this quote in a new tab)
“In our study, there was moderately good correlation between change in ‘‘off’’ time and subject-rated CGI-I scores, but the correlations between change in UPDRS scores and investigator-rated CGI-I scores, although acceptable, were not as strong.” (opens the source at this quote in a new tab)

Change in off time by subject-rated global impression, and receiver operating characteristic cutoffs: rasagiline trial data

Treatment group and subject-rated global impression of improvement	Sample size	Change in off time, hours, mean (SD)	ROC curve cutoff, hours	Sensitivity	Specificity	Area under curve
Rasagiline, very much improved	“26”	“−3.3 (2.2)”	“−2.2”	“0.73”	“0.71”	“0.8”
Rasagiline, much improved	“54”	“−3.5 (2.6)”	“−2.5”	“0.65”	“0.78”	“0.79”
Rasagiline, minimally improved	“69”	“−1.9 (2.2)”	“−1.2”	“0.71”	“0.57”	“0.65”
Rasagiline, no change	“63”	“−0.7 (2.2)”	Not applicable	Not applicable	Not applicable	Not applicable
Rasagiline, minimally worse	“48”	“−0.5 (2.2)”	“0.2”	“0.42”	“0.68”	“0.54”
Rasagiline, much worse	“12”	“−0.6 (2.6)”	“0.3”	“0.58”	“0.7”	“0.52”
Placebo, much improved	“16”	“−1.9 (1.9)”	“−1.7”	“0.75”	“0.66”	“0.67”
Placebo, minimally improved	“27”	“−1.9 (2.7)”	“−1.8”	“0.56”	“0.66”	“0.65”
Placebo, no change	“44”	“−0.9 (2.5)”	Not applicable	Not applicable	Not applicable	Not applicable
Placebo, minimally worse	“43”	“0.3 (2.3)”	“−0.7”	“0.74”	“0.52”	“0.63”
Placebo, much worse	“8”	“1.7 (4.3)”	“0.2”	“0.88”	“0.64”	“0.72”

Minimal clinically important difference in off time: pramipexole extended-release trial data

“In the APD trial, we calculated the MCID for OFF time to be −1.0 hours in the pramipexole ER group and −1.3 hours in the pramipexole IR group.” (opens the source at this quote in a new tab)
“In surveying the literature to date, studies suggest that the lower limit of MCID for total UPDRS (I + II + III) is −3.8 and for change on OFF time is −1.0 hour” (opens the source at this quote in a new tab)

Minimal clinically important differences for MDS-UPDRS Parts I, II, and III

“Nine hundred eighty-five paired investigations of 365 patients were reviewed, implementing three different techniques simultaneously.” (opens the source at this quote in a new tab)
“Any improvement greater than 2.64 points or any worsening more than 2.45 points on MDS-UPDRS Part I represent a minimal, yet clinically meaningful change.” (opens the source at this quote in a new tab)
“In reference to Part II, the smallest changes considered clinically relevant were 3.05 and 2.51 points for improvement and deterioration, respectively.” (opens the source at this quote in a new tab)

“Therefore, 728 paired follow-up visits of 260 patients were utilized for the analyses.” (opens the source at this quote in a new tab)
“The MCID estimates for MDS-UPDRS ME were asymmetric: -3.25 points for detecting minimal, but clinically pertinent, improvement and 4.63 points for observing minimal, but clinically pertinent, worsening.” (opens the source at this quote in a new tab)

“Based on our calculations, the MCID thresholds for the MDS-UPDRS I+II+III are –6.7 and +5.2 points for identifying clinically relevant improvement and worsening.” (opens the source at this quote in a new tab)
“Any improvement greater than 4.9 points or any worsening more than 4.2 points on MDS-UPDRS II+III represent a minimal, yet clinically meaningful, change.” (opens the source at this quote in a new tab)

Investigators' interpretation of the observed off-time reduction against the clinically important difference

“Improvements in good-on-time are associated with better clinical outcomes, improved quality of life, and reduced economic burden.” (opens the source at this quote in a new tab)
“The LSM decrease in off-time from baseline to week 26 observed in the tavapadon group (−1.88 hours) is comparable with improvements in off-time seen with other adjunctive PD therapies, including D2/D3 agonists, and surpasses the previously reported minimum clinically important difference for a reduction in off-time of approximately 1.0 hour.” (opens the source at this quote in a new tab)

Planned sample size, power, and interim sample size increase

“A sample size of 184 participants per group provided greater than or equal to 90% power to detect a 1-hour treatment difference in the primary outcome measure at a 2-sided .05 significance level, with an assumed SD of 2.5 hours. Because sample size assumptions were based on a preliminary trial with limited sample size and a different dosing schedule, an interim analysis to assess the adequacy of the sample size assumption was performed by an independent external review committee, and the target trial sample size subsequently increased from 368 to 500 participants to maintain statistical power at 90%.” (opens the source at this quote in a new tab)

Protocol sample size assumptions

“A sample size of 184 subjects per group (total, 368 subjects) will provide at least 90% power to detect a change in the primary outcome measure (change from baseline in “on” time without troublesome dyskinesia) of 1 hour with a standard deviation of 2.5, a 2-sided alpha level = 0.049, and assuming a 27% dropout rate.” (opens the source at this quote in a new tab)
“This analysis will be conducted by an independent Interim Analysis Review Committee (IARC), which may make a recommendation to increase the overall sample size up to a maximum of 528 subjects.” (opens the source at this quote in a new tab)

Participants screened, randomized, and analyzed

“Of 842 screened individuals, 507 (mean [SD] age, 64.9 [8.5] years; 186 female [37%]; 321 male [63%]) were randomized to receive tavapadon (n = 252) or placebo (n = 255) in addition to levodopa (randomization by site shown in eTable 2 in Supplement 3); 505 received at least 1 dose of study treatment and were included in the safety analysis, and 487 met criteria for the mITT population (Figure 1B).” (opens the source at this quote in a new tab)

Number screened stated in the abstract

“After screening 824 adults with PD, 507 participants (mean [SD] age, 64.9 [8.5] years; 321 male [63%]; mean [SD] disease duration, 6.7 [4.5] years; mean [SD] baseline daily off-time, 5.5 [2.4] hours) were enrolled and received tavapadon (n = 252) or placebo (n = 255).” (opens the source at this quote in a new tab)

Participants analyzed for the primary and key secondary end points

Measure	Tavapadon	Placebo
Participants in the primary and key secondary analysis (prescribing information, Table 10)	“(N = 242)”	“(N = 245)”

Participants with week 26 data in the registry results

Measure	Placebo	Tavapadon
Overall number of participants analyzed, week 26 on time without troublesome dyskinesia	“201”	“153”
Least squares mean (SE) change from baseline, hours	“0.619(0.188)”	“1.721(0.207)”

Modified intention-to-treat and safety analysis sets and the analysis model

“Efficacy analyses were conducted using a mixed model for repeated measures (MMRM) approach based on a modified intention-to-treat (mITT) population (ie, randomized participants who received ≥1 dose of the assigned treatment and had both baseline and ≥1 postbaseline Hauser diary entries). For the primary and key secondary analyses, MMRM included fixed effects of treatment, visits, and interaction between treatment and visit with baseline total daily on-time or off-time (for the primary and key secondary analysis, respectively) included as a covariate and participant included as a random effect. Statistical comparisons for other secondary end points were not corrected for multiplicity; therefore, P values should be interpreted as nominal. The safety analysis set included all participants who received 1 or more doses of tavapadon or placebo.” (opens the source at this quote in a new tab)

Analysis populations defined in the protocol

Population	Description	Analysis
ITT	“All randomized subjects”	“Demographic and Baseline Characteristics”
FAS	“All randomized subjects who receive at least 1 dose of IMP. This will be the safety analysis set.”	“Safety analysis”
mITT	“All randomized subjects who receive at least 1 dose of IMP and have both a baseline and at least 1 postbaseline Hauser diary entry”	“Primary analysis set for efficacy”
Endpoint completers	“All subjects in the mITT population who complete the Hauser diary at endpoint”	“Sensitivity analysis for efficacy”
PK analysis set	“All randomized subjects who receive at least 1 dose of IMP and have at least 1 measurable tavapadon (CVL-751) concentration”	“PK analysis”

Hierarchical testing and alpha adjustment for the interim analysis

“The hypothesis testing will be done in hierarchical order from the primary endpoint onward.” (opens the source at this quote in a new tab)
“hypothesis will be tested at an α level of 0.049 (2-sided) to ensure that the overall Type I error rate is below 0.05. Subsequent testing in the secondary endpoints will be conducted at an α level of 0.05 (2-sided).” (opens the source at this quote in a new tab)

Handling of intercurrent events and missing data

“A hypothetical strategy will be used to address the intercurrent events of discontinuation or use of prohibited medications with the data post intercurrent events treated as censored.” (opens the source at this quote in a new tab)
“Sensitivity analyses for the primary and secondary endpoints will be conducted with different assumptions on missing mechanism. Sensitivity analysis will also be conducted on the endpoint completer set with MMRM.” (opens the source at this quote in a new tab)

Missing-at-random assumption in the statistical analysis plan

“The missing values are assumed to be missing at random for the main analysis, including missing due to missed visits or early termination due to COVID-19 control measures.” (opens the source at this quote in a new tab)
“A pattern mixture model (PMM) approach will be used to address potentially MNAR patterns.” (opens the source at this quote in a new tab)

Discontinuation by treatment group

“A total of 142 participants (28.0%) discontinued the trial, including 93 (36.9%) in the tavapadon group and 49 (19.2%) in the placebo group. Discontinuation reasons included both AEs and non-AE related reasons, such as withdrawal by participant and site terminations due to the war in Ukraine (Figure 1B).” (opens the source at this quote in a new tab)

Participant flow and reasons for not completing the trial

Milestone or reason	Placebo	Tavapadon
Started	“255”	“252”
Completed	“206”	“159”
Not completed	“49”	“93”
Adverse event	“23”	“43”
Lack of efficacy	“5”	“2”
Lost to follow-up	“3”	“4”
Failure to meet continuation criteria	“0”	“1”
Treatment with prohibited concomitant medications	“1”	“0”
Non-compliance with study drug	“0”	“2”
Physician decision	“0”	“1”
Site terminated by sponsor	“3”	“5”
Withdrawal by subject	“12”	“33”
Other	“2”	“2”

Exposure and follow-up duration

“All-cause mortality were reported from enrollment to the end of study, median time on follow up (median time subjects were followed) was 190 and 189 days for Placebo and Tavapadon, respectively. Treatment-emergent adverse events and serious adverse events were collected from first dose of study drug until 4 weeks after the last dose of study drug; mean duration on study drug was 169.6 and 151.7 days for Placebo and Tavapadon, respectively.” (opens the source at this quote in a new tab)

Timing of discontinuations due to adverse events

“AEs led to treatment discontinuation in 17.1% of participants in the tavapadon group compared with 9.1% in placebo; discontinuations due to AEs occurred most commonly during the titration phase (placebo, 17 [7%]; tavapadon, 28 [11%]).” (opens the source at this quote in a new tab)

“Of the 43 patients receiving JUVMO who discontinued because of an adverse reaction, 38 (88%) did so during the titration and adjustment phases. The most common adverse reactions leading to discontinuation included nausea, dyskinesia, visual hallucination, and dizziness.” (opens the prescribing information at this quote)

Demographic and disease characteristics of the randomized population

“Demographics and baseline disease characteristics were similar between groups and reflective of a population with PD and motor fluctuations (Table 1); the mean (SD) duration since diagnosis was 6.7 (4.5) years, and participants experienced a mean (SD) of 5.52 (2.37) hours of daily off-time and 10.02 (2.60) hours of daily on-time without troublesome dyskinesia at baseline. Participants self-reported the following races and ethnicities: 2 American Indian or Alaska Native (0.4%), 4 Asian (0.8%), 6 Black (1.2%), 491 White (96.8%), and 4 other (0.8%).” (opens the source at this quote in a new tab)

Baseline demographics and disease characteristics by treatment group

Characteristic	Placebo QD (n = 255)	Tavapadon 5-15 mg QD (n = 252)	Overall (N = 507)
Age, mean (SD) [range], y	“64.1 (8.5) [43-80]”	“65.6 (8.4) [43-80]”	“64.9 (8.5) [43-80]”
Age <65 y, No. (%)	“119 (46.7)”	“104 (41.3)”	“223 (44.0)”
Age ≥65 y, No. (%)	“136 (53.3)”	“148 (58.7)”	“284 (56.0)”
Female, No. (%)	“85 (33.3)”	“101 (40.1)”	“186 (36.7)”
Male, No. (%)	“170 (66.7)”	“151 (59.9)”	“321 (63.3)”
American Indian or Alaska Native, No. (%)	“2 (0.8)”	“0”	“2 (0.4)”
Asian, No. (%)	“2 (0.8)”	“2 (0.8)”	“4 (0.8)”
Black, No. (%)	“4 (1.6)”	“2 (0.8)”	“6 (1.2)”
White, No. (%)	“245 (96.1)”	“246 (97.6)”	“491 (96.8)”
Other race, No. (%)	“2 (0.8)”	“2 (0.8)”	“4 (0.8)”
Weight, mean (SD), kg	“79.9 (15.7)”	“79.5 (17.2)”	“79.7 (16.5)”
BMI, mean (SD) [range]	“27.0 (4.4) [16.6-40.2]”	“27.4 (5.1) [16.6-42.4]”	“27.2 (4.7) [16.2-46.4]”
Time since initial diagnosis, mean (SD), y	“6.4 (4.6)”	“6.9 (4.5)”	“6.7 (4.5)”
Daily on-time without troublesome dyskinesia, mean (SD), h	“10.1 (2.6)”	“9.9 (2.6)”	“10.0 (2.6)”
Daily off-time, mean (SD), h	“5.4 (2.5)”	“5.6 (2.2)”	“5.5 (2.4)”
MoCA total score, mean (SD)	“28.0 (1.3)”	“28.0 (1.3)”	“28.0 (1.3)”
Modified H&Y stage 2, No. (%)	“146 (57.3)”	“135 (53.6)”	“281 (55.4)”
Modified H&Y stage 2.5, No. (%)	“50 (19.6)”	“63 (25.0)”	“113 (22.3)”
Modified H&Y stage 3, No. (%)	“59 (23.1)”	“54 (21.4)”	“113 (22.3)”
MDS-UPDRS part I score, mean (SD)	“7.5 (4.9)”	“8.0 (5.1)”	“7.7 (5.0)”
MDS-UPDRS part II score, mean (SD)	“12.5 (7.1)”	“13.3 (6.5)”	“12.9 (6.8)”
MDS-UPDRS part III score, mean (SD)	“32.4 (14.3)”	“32.6 (14.3)”	“32.5 (14.3)”
MDS-UPDRS part II + III score, mean (SD)	“44.9 (18.9)”	“45.9 (18.2)”	“45.4 (18.5)”
Adjunctive PD medication added to levodopa: yes, No. (%)	“108 (42.4)”	“115 (45.6)”	“223 (44.0)”
Amantadine, No. (%)	“40 (15.7)”	“49 (19.4)”	“89 (17.6)”
COMT inhibitors, No. (%)	“19 (7.5)”	“17 (6.7)”	“36 (7.1)”
Istradefylline, No. (%)	“2 (0.8)”	“1 (0.4)”	“3 (0.6)”
MAO-B inhibitors, No. (%)	“77 (30.2)”	“76 (30.2)”	“153 (30.2)”
Adjunctive PD medication added to levodopa: no, No. (%)	“147 (57.6)”	“137 (54.4)”	“284 (56.0)”
Daily levodopa dose, mean (SD), mg	“817.6 (644.0)”	“811.9 (451.4)”	“814.8 (556.1)”

Timing of baseline MDS-UPDRS assessments

“Assessments conducted approximately 2 to 3 hours after last dose of levodopa and collected during on or off status.” (opens the source at this quote in a new tab)

Baseline characteristics stated in the prescribing information

“Patients had a mean age of 64.9 years (range: 43 to 80 years), 63% were male, 97% were White, and 5% were Hispanic. The mean duration of PD at baseline was 6.7 years (range: 0.1 to 30.9 years). The mean total ON time without troublesome dyskinesia at baseline was 10.0 hours and was similar between treatment groups.” (opens the prescribing information at this quote)

Primary end point: change in good-on-time at week 26

“The study met its primary end point, with the tavapadon group demonstrating a statistically significant increase of 1.1 hours in total daily good-on-time from baseline to week 26 compared with the placebo group (least-squares mean [LSM] increase of 1.70 hours with tavapadon vs 0.60 hours with placebo; LSM treatment difference, 1.1; 95% CI, 0.6-1.7; P < .001; Cohen d = 0.40). The increase in daily good-on-time from baseline was observed starting at week 2, and separation between tavapadon and placebo was observed as early as week 8 and sustained through week 26 and end of treatment (week 27) (Figure 2A).” (opens the source at this quote in a new tab)

Key secondary end point: change in off-time at week 26

“The tavapadon group experienced a statistically significant reduction from baseline to week 26 in the key secondary end point of off-time compared with the placebo group (LSM decrease of 1.88 vs 0.93 hours, respectively; LSM treatment difference, −0.94 hours; 95% CI, −1.48 to −0.41; P < .001; Cohen d = 0.35). Reductions in off-time with tavapadon relative to placebo were observed starting at week 8 and continued to end of treatment (Figure 2B). Predefined sensitivity analyses were consistent with the primary analysis for both primary and key secondary end points (data not shown).” (opens the source at this quote in a new tab)

Change from baseline to week 26 in primary and key secondary measures

Measure	Juvmo (N = 242)	Placebo (N = 245)
On time without troublesome dyskinesia, baseline mean (SD), hours	“9.9 (2.58)”	“10.2 (2.58)”
On time without troublesome dyskinesia, week 26 mean (SD) change from baseline, as printed	“1.7 (0.21)”	“0.6 (0.19)”
Off time, baseline mean (SD), hours	“5.6 (2.26)”	“5.4 (2.36)”
Off time, week 26 mean (SD) change from baseline, as printed	“-1.9 (0.20)”	“-0.9 (0.18)”

Treatment differences for the primary and key secondary measures in the prescribing information

Measure	Juvmo vs placebo
On time without troublesome dyskinesia, LS mean (SE) difference, as printed	“1.1 (0.6, 1.7)”
On time without troublesome dyskinesia, p-value	“<0.0001”
Off time, LS mean (SE) difference, as printed	“-0.9 (-1.5, -0.4)”
Off time, p-value	“0.0006”

MDS-UPDRS parts II and III and the combined score

“Treatment with tavapadon also resulted in an improvement from baseline in MDS-UPDRS part II scores at week 26 compared with placebo (LSM decrease, 1.4 vs 0.1 points, respectively; LSM treatment difference, −1.2 points; 95% CI, −2.2 to −0.2; nominal P = .02), with significant reductions observed starting at week 11 and continuing to end of treatment (eFigure in Supplement 3). Reductions in MDS-UPDRS part III scores at week 26 were also significantly greater in participants treated with tavapadon vs placebo (LSM decrease, 7.0 vs 4.6 points, respectively; LSM treatment difference, −2.4 points; 95% CI, −4.3 to −0.5; nominal P = .01) (eFigure in Supplement 3).” (opens the source at this quote in a new tab)
“The exploratory analysis of change from baseline in MDS-UPDRS part II+III combined scores demonstrated a significant reduction in the tavapadon group vs the placebo group at week 26 (LSM decrease, 8.4 vs 4.7 points; LSM treatment difference, −3.7; 95% CI, −6.3 to −1.1; nominal P = .005) (eFigure in Supplement 3).” (opens the source at this quote in a new tab)

Pooled good on-time estimate across two adjunctive trials in a meta-analysis

“B7601003 at its 15 mg dose reported + 1.04 h at Week 10 (95% CI − 0.53 to 2.61), an estimate in the same direction and of almost identical magnitude but not statistically significant in a trial terminated at half enrolment.” (opens the source at this quote in a new tab)
“The pooled estimate was + 1.09 h per day (95% CI 0.57 to 1.61; p < 0.001) with no heterogeneity whatsoever (I² = 0.0%, Q(1) = 0.01, p = 0.944)” (opens the source at this quote in a new tab)

TEMPO-3 good on-time confidence interval that differs between the text and figure 6 of a meta-analysis

“Fernandez et al. [12] reported a gain of + 1.10 h per day with tavapadon versus placebo (95% CI 0.60 to 1.70; p < 0.001)” (opens the source at this quote in a new tab)
“Fernandez et al. 2026 242 245 89.1% 1.10 [0.55; 1.65]” (opens the source at this quote in a new tab)

Post hoc analysis: motor states across the waking day

“In a post hoc analysis, a comparison of changes from baseline in motor states across the waking day revealed the improvements in total daily good-on-time at week 26 were primarily driven by the increase in total daily on-time without dyskinesia (Figure 3).” (opens the source at this quote in a new tab)
“Analysis of normal waking day was based on normalized waking hours of 16 hours per day for each individual patient.” (opens the source at this quote in a new tab)

Post hoc motor state results presented at a congress

“Participants receiving adjunctive tavapadon (n=252) demonstrated a significant increase of 0.7 hours in ON time without dyskinesia during the normal waking day versus placebo (n=255; 1.6 hours vs 0.9 hours, respectively; P=0.0426).” (opens the source at this quote in a new tab)
“Change from baseline in OFF time during the normal waking day also showed a significant reduction of 0.8 hours with tavapadon vs placebo (1.1 hours vs 1.9 hours, respectively; P=0.0022). Participants receiving tavapadon experienced a marginal decrease from baseline in ON time with troublesome dyskinesia versus placebo (−0.008 hours vs 0.1 hours; P=0.3397), and an increase in ON time with nontroublesome dyskinesia versus placebo (0.3 hours vs 0.05 hours; P=0.1815).” (opens the source at this quote in a new tab)

Results first presented at a congress

“Adults aged 40–80 years (N=507) were enrolled. Treatment with tavapadon (5–15 mg once daily) significantly increased total daily ON time without troublesome dyskinesia by 1.1 hours relative to placebo (1.7 hours vs 0.6 hours, respectively; P<0.0001). A significant reduction versus placebo in change from baseline in daily OFF time was also observed.” (opens the source at this quote in a new tab)

Topline results announced by the sponsor

“The trial met its primary endpoint – patients treated with tavapadon adjunctive to LD experienced a clinically meaningful and statistically significant increase of 1.1 hours in total “on” time without troublesome dyskinesia compared to those treated with LD and placebo (1.7 hours vs. 0.6 hours, p <0.0001). A statistically significant reduction in “off” time, the key secondary endpoint, was also observed for the tavapadon treatment arm.” (opens the source at this quote in a new tab)

PDQ-39, EQ-5D-5L, and MDS-UPDRS part I

“There were no notable differences in MDS-UPDRS part I, PDQ-39, or EQ-5D-5L Index and VAS scores (data not shown).” (opens the source at this quote in a new tab)
“Finally, participants treated with tavapadon showed nominally significantly greater reductions in MDS-UPDRS parts II and III scores, corresponding to improvements in activities of daily living and investigators’ evaluations of motor function, respectively. No significant differences between groups were observed in the PDQ-39, EQ-5D-5L index, and VAS scores.” (opens the source at this quote in a new tab)

Patient-reported daytime sleepiness on the Epworth Sleepiness Scale

“In TEMPO-3, LSM decrease in ESS total score from baseline was 0.9 with tavapadon and 0.7 points with placebo (LSM difference from placebo [95% CI], −0.2 [−0.8, 0.4]; P=0.53). Item-level scores related to sleepiness on the Movement Disorder Society-Unified Parkinson’s Disorder Rating Scale and Parkinson’s Disease Questionnaire-39, including frequency of unexpectedly falling asleep during the day, did not differ from placebo.” (opens the source at this quote in a new tab)

Summary of adverse events in the safety population

Safety outcome	Placebo QD (n = 254)	Tavapadon 5-15 mg QD (n = 251)
AEs, No. (%)	“140 (55.1)”	“180 (71.7)”
Mild	“65 (25.6)”	“80 (31.9)”
Moderate	“64 (25.2)”	“82 (32.7)”
Severe	“11 (4.3)”	“18 (7.2)”
AE related to study drug	“54 (21.3)”	“105 (41.8)”
AEs with onset in titration phase, No./No. (%)	“83/254 (32.7)”	“100/251 (39.8)”
AEs with onset in adjustment phase, No./No. (%)	“54/237 (22.8)”	“100/214 (46.7)”
AEs with onset in maintenance phase, No./No. (%)	“65/220 (29.5)”	“71/188 (37.8)”
Serious AEs	“14 (5.5)”	“17 (6.8)”
AE leading to discontinuation	“23 (9.1)”	“43 (17.1)”
Death	“0”	“1 (0.4)”
Nausea	“11 (4.3)”	“36 (14.3)”
Dyskinesia	“4 (1.6)”	“25 (10.0)”
Dizziness	“8 (3.1)”	“19 (7.6)”
Headache	“7 (2.8)”	“17 (6.8)”
Fall	“13 (5.1)”	“16 (6.4)”
Orthostatic hypotension	“3 (1.2)”	“15 (6.0)”
Hallucination, visual	“3 (1.2)”	“14 (5.6)”
COVID-19	“9 (3.5)”	“13 (5.2)”
Somnolence	“11 (4.3)”	“13 (5.2)”

Overall adverse events, serious adverse events, and death

“In the safety population, 180 participants (71.1%) in the tavapadon group and 140 participants (55.1%) in the placebo group experienced at least 1 AE (Table 2). Most AEs were nonserious (93.2%) and mild to moderate in severity. Rates of serious AEs (Table 2) were similar between groups. Serious AEs reported in 2 or more participants in the tavapadon group were fall (n = 3) and abdominal pain (n = 2). Severe AEs observed in 2 or more participants in the tavapadon group were dyskinesia (3 [1.2%]), abdominal pain (3 [1.2%]), dizziness (2 [0.8%]), and nausea (2 [0.8%]). One AE of death due to metastatic pancreatic carcinoma reported in the tavapadon group was considered unrelated to study treatment.” (opens the source at this quote in a new tab)

Adverse events by treatment phase

“For the tavapadon group, the percentage of participants experiencing AEs was higher during the dose titration (39.8%) and adjustment (46.7%) phases compared with the maintenance (37.8%) phase. The most common AEs reported in the tavapadon group (≥5% incidence) were nausea (14.3%), dyskinesia (10.0%), dizziness (7.6%), headache (6.8%), falls (6.4%), and orthostatic hypotension (6.0%) (Table 2). Rates of nausea, dyskinesia, dizziness, and headache were higher during the titration phase and decreased with continued exposure. Rates of visual hallucinations and orthostatic hypotension were higher during the dose adjustment and maintenance phases compared with the titration phase.” (opens the source at this quote in a new tab)

Somnolence, fatigue, and daytime sleepiness

“Incidence rates of somnolence (tavapadon, 5.2%; placebo, 4.3%) and fatigue (tavapadon, 4.0%; placebo, 4.3%) were similar between treatment groups. Mean change from baseline to the last visit in ESS total score and subscores were also similar in the tavapadon and placebo groups, with a numerically lower percentage of participants reporting a score greater than or equal to 10 (sleepy) in the tavapadon groups vs the placebo group.” (opens the source at this quote in a new tab)

Impulse control disorders and suicidality

“Overall, 9 participants (tavapadon, 6 [2.4%]; placebo, 3 [1.2%]) experienced AEs suggestive of ICDs. Of the 6 participants in the tavapadon group, all events were nonserious and mild to moderate in severity; 1 event led to tavapadon discontinuation. All events but 1 were resolved by the end of the trial. Mean changes from baseline to the last visit in QUIP-RS total scores were similar between the tavapadon and placebo groups (data not shown). There was no evidence of increased suicidality with tavapadon based on the C-SSRS (data not shown).” (opens the source at this quote in a new tab)

Blood pressure and orthostatic hypotension

“Blood pressure decreases in the tavapadon group were observed. At the end of the study (week 27), a mean (SD) decrease of 8.6 (16.3) mm Hg was observed in standing systolic blood pressure (vs 0.8 [15.2] mm Hg with placebo) and a mean (SD) decrease of 6.0 (10.8) mm Hg was observed in standing diastolic blood pressure (vs 1.8 [9.7] mm Hg with placebo), with similar changes observed for supine blood pressures. The frequency of measured orthostatic hypotension was comparable between treatment groups (eTable 3 in Supplement 3), but the incidence of orthostatic hypotension reported as an AE was higher in the tavapadon group (6.0%) compared with the placebo group (1.2%) (Table 2). There were no clinically relevant safety findings in clinical laboratory findings, eye examinations, or cardiovascular assessments.” (opens the source at this quote in a new tab)
“Of tavapadon-treated participants experiencing falls, only 1 was receiving treatment for orthostatic hypotension.” (opens the source at this quote in a new tab)

Measured orthostatic hypotension

Measure	Placebo QD (N = 254)	Tavapadon 5-15 mg QD (N = 251)
Systolic blood pressure ≥20 mm Hg decrease upon standing compared to supine, n (%)	“51 (20.1)”	“45 (17.9)”
Diastolic blood pressure ≥10 mm Hg decrease upon standing compared to supine, n (%)	“57 (22.4)”	“64 (25.5)”

Adverse reactions in Study 3 in the prescribing information

Adverse reaction	Juvmo (N = 251), %	Placebo (N = 254), %
Nausea	“14”	“4”
Dyskinesia	“10”	“2”
Dizziness	“8”	“3”
Headache	“7”	“3”
Hallucination	“7”	“1”
Orthostatic hypotension	“6”	“1”
Hypotension	“4”	“<1”
Abdominal pain	“3”	“<1”
Asthenia	“3”	“0”

Warnings and precautions: rates observed in Study 3

“In Study 3 in patients taking JUVMO as adjunctive therapy to fixed dose levodopa [see Clinical Studies (14.2)], orthostatic hypotension was reported in 8% of patients treated with JUVMO compared to 2% of patients on placebo. Hypotension was reported in 4% of patients treated with JUVMO compared to 0.4% on placebo.” (opens the prescribing information at this quote)
“In Study 3, impulse control disorders were reported in 2% of patients treated with JUVMO compared to 1% of patients on placebo.” (opens the prescribing information at this quote)
“In Study 3, when added as an adjunct to a fixed dose of levodopa, hallucination was reported in 7% of patients treated with JUVMO compared to 1% of patients on placebo.” (opens the prescribing information at this quote)
“In Study 3, among patients 65 years of age and older, hallucination was reported in 9% of patients treated with JUVMO compared to 1% of patients on placebo.” (opens the prescribing information at this quote)
“In Study 3, 10.0% of patients treated with JUVMO reported dyskinesia compared to 2% on placebo. Consider decreasing the dosage of JUVMO or other dopaminergic medications if dyskinesia occurs.” (opens the prescribing information at this quote)

Serious adverse events and deaths in the registry results

Measure	Placebo	Tavapadon
All-cause mortality, affected / at risk (%)	“0/255 (0.00%)”	“1/252 (0.40%)”
Serious adverse events, affected / at risk (%)	“14/255 (5.49%)”	“17/252 (6.75%)”

Investigators' interpretation of the safety findings

“The safety profile of tavapadon was favorable and consistent with earlier-phase tavapadon trials” (opens the source at this quote in a new tab)
“most AEs were nonserious, mild or moderate in severity, and allowed participants to remain on treatment. Rates of discontinuation due to AEs were higher for tavapadon vs placebo. Common dopaminergic AEs (eg, nausea, dyskinesia, dizziness) reported in participants treated with tavapadon were expected.” (opens the source at this quote in a new tab)
“In this 27-week trial, there were no notable differences between tavapadon and placebo in the incidence of ICDs or the QUIP-RS total score.” (opens the source at this quote in a new tab)

Summary: limitations of TEMPO-3

TEMPO-3 compared tavapadon with placebo for 27 weeks and included no active comparator, so it provides no direct comparison with D2/D3 dopamine agonists or other adjunctive therapies; the investigators note that adverse events such as impulse control disorders may emerge with longer treatment. Discontinuation was unequal between groups: 93 of 252 participants (36.9%) receiving tavapadon and 49 of 255 (19.2%) receiving placebo discontinued, including 43 and 23 for adverse events and 33 and 12 by participant withdrawal. The registry results report week 26 primary end point data for 153 tavapadon and 201 placebo participants, compared with 242 and 245 in the modified intention-to-treat population, and the primary analysis treated data after discontinuation as censored under a hypothetical strategy with a missing-at-random assumption. The target sample size was increased from 368 to 500 participants after an interim analysis, and the primary hypothesis was tested at a 2-sided alpha of 0.049. Comparisons for secondary end points other than off-time were not corrected for multiplicity, so MDS-UPDRS results carry nominal P values, and MDS-UPDRS assessments were performed 2 to 3 hours after the last levodopa dose in either the on or off state. The levodopa dose and dosing frequency were held fixed for the whole trial. The population was 96.8% White, excluded people with a Montreal Cognitive Assessment score below 26, a clinically significant impulse control disorder, or psychosis or hallucinations within 12 months, and excluded users of moderate or strong CYP3A4 inhibitors or inducers. The trial ran during the COVID-19 pandemic and the war in Ukraine, which led to site terminations. The prescribing information reports orthostatic hypotension in 8% versus 2% in its warnings section and 6% versus 1% in its adverse reaction table, and the publication reports 6.0% versus 1.2%. The primary publication gives two screening counts (842 in the results and 824 in the abstract) and two percentages for any adverse event with tavapadon (71.1% in the text and 71.7% in the table).

Limitations stated by the investigators

“The 27-week treatment duration was comparable with other phase 3 trials investigating therapies for PD with motor fluctuations” (opens the source at this quote in a new tab)
“however, as people with PD in real-world settings may receive treatment for years,” (opens the source at this quote in a new tab)
“research is needed to support the long-term safety and efficacy of tavapadon. This study was not designed to compare tavapadon with D2/D3 DAs or other adjunctive therapies for PD. The study population was predominantly White (96.8%), and further study of tavapadon in underrepresented populations may be needed. This trial was held during the height of the COVID-19 pandemic, which may have contributed to the longer-than-usual times to complete enrollment, and the discontinuation rate.” (opens the source at this quote in a new tab)
“As some AEs associated with dopaminergic treatment, such as ICDs, may emerge with prolonged treatment, additional data from the long-term extension trial” (opens the source at this quote in a new tab)
“of tavapadon may further the understanding about the safety of tavapadon adjunctive to levodopa in people with PD and motor fluctuations.” (opens the source at this quote in a new tab)

Sample size assumptions based on a trial with a different dosing schedule

“Given that the assumptions for the sample size were based on a preliminary study using a different dosing schedule, it is unknown if these assumptions are appropriate for this study as designed.” (opens the source at this quote in a new tab)

Absence of active comparator and trials longer than 27 weeks across the evidence base

“Its proposed neuropsychiatric advantage over D2/D3 agonists remains untested: no trial used an active comparator or exceeded 27 weeks.” (opens the source at this quote in a new tab)

Power of TEMPO-3 to detect a difference in impulse control disorders

“Impulse control disorder (RR 2.02, 0.51 to 8.00; nine events, one trial) and somnolence (1.20, 0.37 to 3.92) were rated at very low certainty.” (opens the source at this quote in a new tab)
“The point estimate is above unity and the interval is compatible with anything from a halving to an eight-fold increase in risk; with nine events in total, this trial had approximately 18% power to detect a doubling of the observed placebo rate. No inference about impulse control disorder risk, in either direction, is supported by these data.” (opens the source at this quote in a new tab)
“For impulse control disorder the evidence was rated down one level for indirectness, because the 27-week maximum exposure of the included trials is short relative to the multi-year latency over which these disorders typically emerge, and two levels for imprecision, because nine events across a single trial yield a confidence interval spanning a halving to an eight-fold increase in risk.” (opens the source at this quote in a new tab)