Hypokinetic / Hyperkinetic Sx tail, Decision-making row, Clinical trial Examples of PD (SAD), MAD, Finding (study in the target population), Concomitant Medication, Dose finding / Dose ranging study, Dose selection/finding study

	normal
Motor	When a decision is made to perform a particular action → ↑ DA → ↓ Basal ganglia's normal inhibitory action → ↑ motor activation
Oculo-motor
Associative
Limbic
orbitofrontal

Clinical trial Examples of PD

SAD

design

Output

(1) starting dose (x) increased by an equal amount (x, 2x, 3x, 4x...); (2) doses (x) increased by equal percentage (e.g., by 100%); (3) modified Fibonacci (x, 2x, 3.3x, 5x, 7x, 9x, 12x, and 16x); (4) a variant of the modified Fibonacci scheme where doses are increased by 100% until the first hint of toxicity followed by the modified Fibonacci scheme. Many methods based on concentrations or PK guided dose escalation utilize PK parameters such as AUC or Cmax from the preceding dose group to rationalize the dose increments for escalation (Vaidya and Vaidya, 1981; Graham and Workman, 1992; Krogner and Hirsch, 2002). Doses are escalated to the MTD if appropriate, and AUC or a given PK parameter is monitored. In general, doses are escalated by doubling the dose until 40% of the AUC at the mouse LD10 is reached, and then conventional dose escalation began.

MAD

design	Output
PK data from the single dose FIH study is used to estimate the dosing frequency for the multiple dose study

nding (study in the target population)

design	Output
Critical Pathways to Success in CNS Drug Development; 1st Edition P19. - 50% < MTD - 25% < MTD - 10% < MTD based on HV - 25% < MTD - 50% < MTD	MTD: the dose at which 50% of patients experience severe or multiple AEs probably related to the study drug, ie one patient experiences a SAE - THE DOSE IMMEDIATELY BELOW MD = MTD

ncomitant Medication

DA agonist	before	During protocol	During actual
sandena	Naive (MADB allowed)	Not expected to require DA therapy for at least 52 weeks	Allowed, ~60% started
PMI / Datican / VMAT study (PMI (2.0?))		Not expected to require DA therapy for at least 6 months, but allowed	Allowed, ~51% started during 1y, 79% started during 2y (Simuni, 2018 #198)

ose finding / Dose ranging study

Design

Goal / output

Information on the duration of a PD effect along with PK data obtained in Phase I studies provides a basis for dosage interval or frequency. Dose ranging studies usually include a placebo group, plus a few doses of the test drug — e.g., low dose, medium dose, and high dose. An ideal dose ranging study should cover a wide range of doses from low to high. Typically, the high dose is a dose selected around or below the MTD. MaxED: MaxED, an approach to clinical study design without explicit MTD information, was first proposed by Sheiner et al. (2004) for randomization to support MTD. Although estimates regarding MTD should have been available prior to Phase II, more information will be helpful to re-confirm or to adjust MTD estimates obtained from previous trials. In a dose-response design with placebo, low dose, high dose and several doses in between, the dose range is defined as the range between the lowest and the highest dose. Dose range can be expressed as the ratio of highest dose over lowest dose — as a rule of thumb, in the first dose ranging study, the range should be at least 10-fold. In many cases, when the dose range is too narrow, the dose-response study failed to deliver the necessary information for efficacy or safety, and to work will be needed after these studies. Wong and Lachenbruch (1996) introduce cases using equal dose spacing from low to high doses, that is to divide the distance from placebo to highest dose by the number of active doses, then use that divided distance as the space between two consecutive doses (e.g., 20, 40, 60 mg, respectively). Others may consider some type of log dose spacing, e.g., 1, 3, 10, and 30 mg, respectively, for the design. Hanlett et al. (2002) proposed to use binary dose spacing (BDS) design for dose allocation. If the study includes two test doses and placebo, BDS suggests to pick a mid-point between placebo and MTD, then allocate a dose above the midpoint and another dose below. If the study uses three test doses and placebo, BDS suggests to keep the high dose as the one selected in the two-dose case. Then pick a second midpoint between placebo and the first mid-point, allocate the low dose below the second midpoint, and the medium dose between the two mid-points. statistically optimal experimental design? Optimal design techniques can be used with various statistical models. For a given study objective and a reasonable model, optimization techniques allow one to determine the statistically best set of doses and number of subjects to be used at each dose. These designs help to estimate the parameters of the model; for example, slope and ED50 (the dose response in logistic regression models, and intercept and slope in linear regression models).

Goal / output to estimate the dose response relationships for efficacy.

Dose selection/finding study

design

Goal / output

Although the design of a dose selection study is very similar to a dose ranging study (with placebo or active control, plus a few test doses), the data analysis tends to be hypothesis testing of each test dose against the control. - Statistic stepwise test : testing the most important endpoint first. If this null hypothesis is not rejected, stop. Otherwise, continue to test for the second most important endpoint... vs if two or three endpoints are equally important, then it is possible to combine these endpoints into a single score, and the primary analysis is performed on this composite score. - Testing to see if a specific dose of the study drug is different from placebo - Finding the maximum effective dose - Differentiating efficacy between active doses - Checking to see if there is an increasing dose trend - Demonstrating monotonicity between a particular dose and the active control - There is another angle of sample size estimation: a study is powered to achieve a required amount of precision for an estimated quantity using a confidence interval approach (rather than testing the effect = 0). The quantity could be an accepted age of responses at a given dose — or, more usefully, the dose to give a required range of response.

ple size calculation

CRITICAL PATH INSTITUTE

Uncertain Spans

location	transcription	uncertainty
Concomitant Medication / sandena row	sandena / Naive (MADB allowed)	reads as written; sandena could be a partial of Pasadena; MADB may refer to an MAO-B inhibitor list; preserved verbatim.
Section header	nding (study in the target population)	leading characters Fi- are clipped at the left edge; preserved verbatim.