Product Development
Doing more with less
How Ovid, Clementia found endpoints in two untested rare diseases
In rare diseases where natural history data are scarce, Phase II studies may have to perform dual duty to both seek efficacy signals and test out new endpoints. Ovid Therapeutics Inc. and Clementia Pharmaceuticals Inc. are taking very different approaches to doing so in some of the first clinical trials ever conducted in their respective diseases.
Endpoint development is challenging in rare diseases for more reasons than low numbers of patients. These diseases may affect multiple body systems and have different impacts on patients over the course of their lives, making it hard to find meaningful and sensitive measures of improvement. In addition, the natural history may not be well understood, which can make adapting and interpreting conventional endpoints difficult.
Ovid designed its first clinical trial in Angelman syndrome to test not only an intervention, but also a slate of potential outcomes. The Phase II STARS trial tested the GABA A receptor δ (GABRD) agonist gaboxadol (OV101) vs. placebo in 88 individuals with Angelman Syndrome.
On Aug. 6, Ovid reported that gaboxadol met the primary endpoint of safety, and the lower of two doses tested met the first prespecified efficacy endpoint of Clinical Global Impressions-Improvement (CGI-I).
But Ovid ended the week down $3.36 (35%) on concerns over the inconsistency of the data: the higher dose didn’t meet the endpoint, and neither dose significantly improved secondary behavioral, sleep or motor function endpoints.
“There’s greater acceptance of data integration and different data sources.”
Chairman and CEO Jeremy Levin said gaboxadol’s mechanism of action explains the dose-response behavior, and investors didn’t appreciate that the trial was intended to study the endpoints themselves, not necessarily to improve all of them (see “Sidebar: (Over)shooting STARS?”).
Sidebar: (Over)shooting STARS?
“We didn’t anticipate statistically significant results. We thought we would get something that would help direct us for our next trial,” said Levin.
Ovid hasn’t yet reported on all the outcomes or discussed its path forward, but Levin said the study showed the CGI-I can capture meaningful improvements across a broad swath of the patient population.
Clementia made its case to FDA for a novel endpoint in fibrodysplasia ossificans progressiva (FOP) by combining data from interventional Phase II studies of palovarotene with patient community surveys and a natural history study. Palovarotene is a retinoic acid receptor γ (RARG) agonist.
The primary endpoint of the Phase III MOVE study is yearly change in new bone volume as shown by CT imaging. MOVE is supplementing that endpoint with additional outcomes that include the FOP-Physical Function Questionnaire (FOP-PFQ) and the disease-agnostic Patient-Reported Outcomes Measurement Information System (PROMIS) global health scale.
The two biotechs’ different approaches to identifying endpoints were determined by the characteristics of the diseases they hope to treat. Ovid’s approach, which hinged on including a large placebo group in STARS, may turn out to save it the expense of conducting a natural history study.
That approach wouldn’t have been feasible in FOP, where patients who are untreated risk irreversible and eventually fatal declines.
“It’s important to distinguish between those disorders where the individuals die in childhood and those where the individual can and does live a normal lifespan,” said Levin.
Proof by placebo
Levin said Ovid’s trial was the first industry-sponsored study in Angelman syndrome, and no disease-specific outcomes had been developed or tested. Furthermore, lack of natural history data kept the company from creating disease-specific measures or adapting existing outcomes to the Angelman population prior to the trial.
Angelman syndrome is a neurodevelopmental disorder present from birth that is characterized by a spectrum of neurological features including impaired communication, hyperexcitability, motor dysfunction, sleep disturbance and seizures. It affects 16,000-27,000 patients in the U.S.
To find suitable endpoints, Ovid spent a year reviewing roughly 200 potential outcomes that had been used in registrational trials in similar indications. The company then consulted Angelman caregivers and physicians to winnow the list to 17 scales deemed most appropriate to assess Angelman symptoms.
Ovid also asked for help prioritizing the outcomes based on their importance to caregivers and the likelihood they could show measurable changes in a clinical trial. In STARS, secondary efficacy outcomes were analyzed hierarchically starting with CGI-I, which physicians and caregivers rated most important.
The study took about two and a half years to complete.
Prespecifying the order in which to assess the endpoints gave reassurance that the statistically significant improvement in CGI-I was reliable.
STARS’s 1:1:1 randomization also increased confidence that the improvement was real, because it resulted in a larger placebo group than is typical in a rare disease trial. Without natural history data to draw on, it would have been difficult to tell whether a small placebo group’s performance was typical of the broader population.
“The only way to determine that without cherry-picking results is to do what we did: conduct a double-blinded, placebo-controlled trial.”
“CGI-I captures the totality of the disorder in a robust fashion. The only way to determine that without cherry-picking results is to do what we did: conduct a double-blinded, placebo-controlled trial,” Levin said.
Levin said STARS also showed the other secondary outcomes should be adapted to better suit Angelman syndrome before they can be used in subsequent studies.
These included a caregiver-recorded sleep diary, the Aberrant Behavior Checklist (ABC) and the Anxiety, Depression and Mood Scale (ADAMS) assessed by caregivers, and the physician-rated Modified Performance Oriented Mobility Assessment (MPOMA) score, which captures aspects of mobility such as balance and gait.
Levin said putting caregivers in charge of recording sleep didn’t work because older Angelman patients may sleep alone, and caregivers wouldn’t disturb them to see if they were asleep.
The trial may have been less likely to record changes in MPOMA scores, he said, because scores were determined by independent physicians who watched films of patients, and these physicians were not familiar with Angelman syndrome patients and the types of movements they normally make.
Few physicians who treat Angelman syndrome have experience with clinical testing tools, according to Levin. Thus Ovid couldn’t survey them before the trial to compare how they would have rated patients on MPOMA.
The ADAMS and ABC assess symptoms or behaviors that the majority of Angelman patients don’t have. Ovid included them to determine their usefulness in patients who do have those symptoms.
Ovid has yet to analyze other secondary endpoints and additional data collected in STARS, like actigraphy to assess sleep, treadmill testing to assess gait, metabolomic studies and EEGs, which might yield more insights on endpoints.
The company plans to report data from STARS at the American Academy of Child and Adolescent Psychiatry annual meeting in October, and at additional scientific meetings this year. It plans to begin an open label extension study, ELARA, next quarter.
Levin said Ovid’s next step is to discuss the approval pathway and endpoints with FDA; he declined to say whether Ovid will seek approval based on the results of STARS.
Piecing together
Like Ovid in Angelman, Clementia is the first company to conduct a clinical trial in patients with FOP, CEO Clarissa Desjardins told BioCentury.
To compensate for scant data on the condition and lack of clinical precedents, Clementia started with outreach in the patient community to get more information about the disease and ideas for how to measure it.
FOP is a progressive musculoskeletal disorder estimated to affect 3,500 people worldwide. Episodes of heterotopic ossification (HO) -- or abnormal bone formation -- gradually replace skeletal muscle and soft tissue to form a second skeleton. Ossification episodes are preceded by a flare-up that includes symptoms such as pain, swelling or stiffness. The accumulating bone increasingly restricts movement, leading to physical disability and early death.
Starting about five years ago, the company adapted existing mobility questionnaires to create the FOP-Physical Function Questionnaire (FOP-PFQ) and validated it through surveys of the patient community and physician KOLs.
In July 2014, Clementia began a Phase II study in which 40 FOP patients experiencing flare-up symptoms were randomized 3:1 to receive one of two doses of palovarotene or placebo for six weeks, followed by six weeks of follow up.
These patients were subsequently enrolled into Part A of an open-label extension study that tested different palovarotene doses to treat flare-ups, and 41 additional patients were enrolled into Part B, which tested chronic palovarotene dosing that was temporarily stepped up after a flare-up.
“We were able to combine data from the placebo and natural history study because of this objective, quantifiable endpoint.”
In parallel with the Phase II program, Clementia began a three-year, 114-patient natural history study to collect data on disease progression and outcomes that could serve as potential FOP endpoints. The study started in December 2014.
The Phase II and natural history studies assessed patients using the FOP-PFQ; individual functional endpoints of mobility, pain or swelling; and HO imaging by whole body CT scans.
In the placebo-controlled phase of the Phase II trial, palovarotene resulted in non-statistically significant improvements in mean new HO volume and incidence of new HO. The compound did not show an effect on functional outcomes.
In 33 evaluable patients in Part B, mean volume of new HO at 12 months for patients receiving palovarotene was 28% lower than age-matched patients from the natural history study. It was 65% lower when excluding one patient who had evidence of multiple flares but did not subsequently receive a higher dose of palovarotene.
The company has not yet reported functional outcome data from Part B, but Desjardins said the natural history data suggest that the change from expected new HO volume seen in the trial corresponds to clinically meaningful preservation of mobility.
Desjardins said the company’s work to correlate functional outcomes important to patients with an objective imaging endpoint helped convince regulators to accept an external control group in its Phase III program.
“We were able to combine data from the placebo and natural history study because of this objective, quantifiable endpoint,” she said.
The Phase III MOVE trial is testing chronic treatment with daily 5 mg palovarotene. The dose will increase to 20 mg per day for four weeks to treat flare-ups, followed by 10 mg daily for eight weeks, or until an investigator deems the flare-up resolved.
Untreated patients from the natural history study will serve as a control; patients in the natural history study who experience an eligible flare-up can elect to enroll in the MOVE trial rather than continue.
The primary endpoint of MOVE is annual change in new HO volume. The study is slated to run through 2020; interim readouts are planned for 2Q19 and 4Q19.
Tapping the source
National Organization for Rare Disorders (NORD) Director of Research Programs Vanessa Boulanger suggested another way to test the sensitivity and meaningfulness of new endpoints for rare diseases before entering the clinic: registries.
“In a registry, you typically have pretty broad representation of people living with a certain condition -- there are different ages and disease effects represented, expression of symptoms and perhaps different longevity since the diagnosis,” she said.
And where there are registries, there may also be a history of baseline information that can help interpret responses in a clinical trial, allowing patients to serve as their own control, and potentially enabling enrollment of fewer patients who receive placebos in trials.
“If you have data on me from five years ago or two years ago, you can look back at my data to see how I’m reporting later on,” she said.
NORD’s IAMRARE platform allows patients and organizations to contribute and store data, such as medical information and responses to questionnaires, to serve as the basis for registries or natural history studies. Companies can partner with NORD to create disease-specific registries or, if available, access natural history data to inform study design and complement data in from clinical programs.
Boulanger also recommended that companies developing new endpoints look for additional sources of data, including real-world data.
“The way we’re generating and collecting and analyzing medical and health data has changed drastically over the course of the past 10 years: there’s greater acceptance of data integration and different data sources, and I think that’s reflected by the FDA as well,” she said.
Companies and Institutions Mentioned
American Academy of Child and Adolescent Psychiatry, Washington, D.C.
Clementia Pharmaceuticals Inc. (NASDAQ:CMTA), Montreal, Quebec
National Organization for Rare Disorders (NORD), Danbury, Conn.
Ovid Therapeutics Inc. (NASDAQ:OVID), New York, N.Y.
U.S. Food and Drug Administration (FDA), Silver Spring, Md.