17 Oct 2014

Thanks to an enormous effort to collect and comingle data from 17 ALS clinical studies, researchers can now access thousands of patient records to understand the disease’s natural progression, hunt for biomarkers, and plan better therapeutic trials. Project organizers from the nonprofit Prize4Life in Cambridge, Massachusetts, and the Neurological Clinical Research Institute at Massachusetts General Hospital (MGH) in Boston, describe the Pooled Resource Open-Access ALS Clinical Trials database (PRO-ACT) in the October 8 Neurology online. In this paper, first authors Nazem Atassi and James Berry of MGH also offer a few examples of how the PRO-ACT data can be mined. They confirmed suspicions, raised in much smaller studies, that higher levels of uric acid and creatinine in the blood, as well as being heavy, predict a slower course of ALS.

“It is really exciting to ponder what is buried in there and what we can discover," said John Ravits of the University of California, San Diego, who was not involved in the project. Since PRO-ACT’s official opening in 2012 (see Dec 2012 conference story), more than 370 users from 42 different countries have requested free access to the database, said study author Neta Zach of Prize4Life. Alberto Ascherio of the Harvard School of Public Health in Boston, who is not involved with developing the database, said the large number of records will allow users to explore questions that individual studies cannot. “You can look at biomarkers and disease progression in unprecedented ways,” he said.

Major Merger
Typically, data from trials of unsuccessful drugs languish in file drawers or company vaults. Senior author Melanie Leitner, formerly of Prize4Life and now at Biogen Idec in Cambridge, Massachusetts, and colleagues have brought that valuable data out into the open. The project, a collaboration with the Northeast ALS Consortium, was funded by the ALS Therapy Alliance. (Prize4Life collaborates with Alzforum on AlsGene and funds this reporter’s position.)

Diane Stephenson of the Critical Path Institute in Tucson, Arizona, has organized a similar effort for Alzheimer’s disease called C-Path Online Data Repository (see Dec 2010 conference story). Stephenson was impressed that the organizers obtained information not only on people in the placebo arm, but also the treatment arms. “Typically, companies do not want to share [that data],” she said. In querying PRO-ACT, scientists might even discover that some of those drugs worked for a subset of people with ALS, she speculated.

Getting access was but the first step, however. Study author Alexander Sherman of MGH led the effort to put the data into one unified format. Each study reported data in its own way. Some used metric units, some U.S. standard. Studies labeled gender differently, as M and F or 1 and 0. The authors had to make all those data entries match. Standardization is an underappreciated, but essential, part of big-data studies, Stephenson said.

PRO-ACT currently contains the records from 8,635 people who participated in 16 Phase 2 or 3 trials and one observational study. Nearly 5,000 of those records list the ALS-Functional Rating Scale (ALS-FRS) scores, a standard measure for disease progression. More than 3,000 include date of death, because the person died while the trial was ongoing. More than 7,000 records indicate whether the patient took riluzole, the only treatment approved for ALS. Other data points include blood tests, lung capacity, and family and medical histories.

The organizers are further curating PRO-ACT, Zach said. In November, data on adverse events and concomitant medications will be added. That will help researchers understand what other conditions people had besides ALS. For example, Zach speculated, scientists might find a link between ALS and immune disease. In 2015, PRO-ACT will expand by a few trials, including studies of ceftriaxone, minocycline, and dexpramipexole (Cudkowicz et al., 2014; Gordon et al., 2007; Rudnicki et al., 2013).

First Findings
In their Neurology report, the PRO-ACT scientists offer a taste of what PRO-ACT can do. In the past, scientists have linked uric acid, creatinine, and body mass index with ALS progression, but those small studies, with tens to hundreds of subjects, were tentative (see May 2011 news story; Jul 2009 news story on Paganoni et al., 2012; Ikeda et al., 2012). Using PRO-ACT, Leitner and colleagues confirmed that those measures indeed affected disease course in the larger sample. “It is nice to see that the use of this database validated those markers,” said Richard Bedlack of the Duke ALS Clinic in North Carolina, who was not involved in the project.

Uric acid, or urate, is an antioxidant that reportedly protects against Parkinson’s (Ascherio et al., 2009). “This [PRO-ACT] study, while not definitive or conclusive, is a welcome addition that suggests there is something going on [with uric acid] in ALS,” said Ascherio. Sabrina Paganoni at MGH and colleagues are planning to study the safety of the urate precursor inosine for the treatment of ALS. Ascherio cautioned against the unsupervised use of inosine because it can cause kidney stones and gout.

The developers predict that PRO-ACT will help researchers design better trials. With PRO-ACT it should be possible to simulate ALS trials the way Alzheimer’s researchers can with CODR, said Stephenson (see Jul 2013 news story). For example, clinicians could use PRO-ACT to predict disease course and determine how many subjects they require to see a certain drug effect. “It will help accurately power trials,” Stephenson said.

Clinical trial organizers could also use PRO-ACT to refine how they select their subjects. In one publication with PRO-ACT data, researchers sorted cases into fast and slow progressors (Gomeni and Fava, 2014). This could help exclude patients whose disease course is so slow that they would not be expected to benefit from treatment over the short time of a typical drug study, said Zach.

Researchers could incorporate PRO-ACT data directly into trials by mining it for historical controls, Atassi suggested. Since ALS is rare, it can be difficult to recruit enough people for trials, and no one wants to end up in the placebo arm (see Oct 2011 news story). One option would be to give every participant in a new trial the experimental drug, and use PRO-ACT cases as the placebo control. Because the database contains so much information, researchers could select controls who otherwise match their subjects. For example, they might want to make sure the controls have similar uric acid levels as do people in the treatment arm. Bedlack thought this could speed up Phase 1 and 2 trials to identify drugs that should move on to Phase 3 studies. Those would still have to be double-blinded and placebo-controlled to satisfy regulatory requirements.

Bedlack added that PRO-ACT data have helped him investigate off-label treatments on behalf of the networking site ALSUntangled. Many people with ALS self-medicate, and some are interested in vitamin D and propofol. Comparing people who used those treatments to others who did not in PRO-ACT, Bedlack found no difference in disease course (ALSUntangled Group, 2013; ALSUntangled Group, 2014).

The study authors pointed out a major limitation of the database: The data reflect people who participated in clinical trials, not people with ALS as a whole (see Chiò et al., 2011). As such, it may miss key demographics. For example, PRO-ACT data are skewed toward a younger population because trials often have a maximum age limit.

Could PRO-ACT help answer the question that most worries many people with ALS: “How much longer do I have to live?” Atassi doubts it, but Prize4Life hopes it will. The nonprofit has awarded $50,000 in prizes to database users who developed algorithms to predict survival (see Nov 2012 news story). One, Torsten Hothorn at the University of Zurich, has linked declining ALS-FRS scores to progression and published his calculations (Hothorn and Jung, 2014). A publication on the contest results is forthcoming, Zach added.

PRO-ACT comes at a perfect time, said Stephenson. This summer, ALS research received a huge influx from donations as part of the ALS ice bucket challenge (see Sep 2014 news story). Now that both data and the funds are available, Stephenson sees a bright future. “Hopefully, [PRO-ACT] will foster a lot more interest in drug development,” she said.—Amber Dance

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

Paper Citations

Cudkowicz ME, Titus S, Kearney M, Yu H, Sherman A, Schoenfeld D, Hayden D, Shui A, Brooks B, Conwit R, Felsenstein D, Greenblatt DJ, Keroack M, Kissel JT, Miller R, Rosenfeld J, Rothstein JD, Simpson E, Tolkoff-Rubin N, Zinman L, Shefner JM, Ceftriaxone Study Investigators. Safety and efficacy of ceftriaxone for amyotrophic lateral sclerosis: a multi-stage, randomised, double-blind, placebo-controlled trial. Lancet Neurol. 2014 Nov;13(11):1083-91. Epub 2014 Oct 5 PubMed.
Rudnicki SA, Berry JD, Ingersoll E, Archibald D, Cudkowicz ME, Kerr DA, Dong Y. Dexpramipexole effects on functional decline and survival in subjects with amyotrophic lateral sclerosis in a Phase II study: subgroup analysis of demographic and clinical characteristics. Amyotroph Lateral Scler Frontotemporal Degener. 2013 Jan;14(1):44-51. Epub 2012 Sep 17 PubMed.
Paganoni S, Zhang M, Quiroz Zárate A, Jaffa M, Yu H, Cudkowicz ME, Wills AM. Uric acid levels predict survival in men with amyotrophic lateral sclerosis. J Neurol. 2012 Sep;259(9):1923-8. Epub 2012 Feb 10 PubMed.
Ikeda K, Hirayama T, Takazawa T, Kawabe K, Iwasaki Y. Relationships between disease progression and serum levels of lipid, urate, creatinine and ferritin in Japanese patients with amyotrophic lateral sclerosis: a cross-sectional study. Intern Med. 2012;51(12):1501-8. PubMed.
Ascherio A, LeWitt PA, Xu K, Eberly S, Watts A, Matson WR, Marras C, Kieburtz K, Rudolph A, Bogdanov MB, Schwid SR, Tennis M, Tanner CM, Beal MF, Lang AE, Oakes D, Fahn S, Shoulson I, Schwarzschild MA, Parkinson Study Group DATATOP Investigators. Urate as a predictor of the rate of clinical decline in Parkinson disease. Arch Neurol. 2009 Dec;66(12):1460-8. PubMed.
Gomeni R, Fava M, Pooled Resource Open-Access ALS Clinical Trials Consortium. Amyotrophic lateral sclerosis disease progression model. Amyotroph Lateral Scler Frontotemporal Degener. 2014 Mar;15(1-2):119-29. Epub 2013 Sep 26 PubMed.
ALSUntangled Group. ALSUntangled No. 22: Propofol. Amyotroph Lateral Scler Frontotemporal Degener. 2013 Dec;14(7-8):640-2. Epub 2013 Aug 23 PubMed.
ALSUntangled Group. ALSUntangled No. 24: Vitamin D. Amyotroph Lateral Scler Frontotemporal Degener. 2014 Jun;15(3-4):318-20. Epub 2014 Apr 2 PubMed.
Chiò A, Canosa A, Gallo S, Cammarosano S, Moglia C, Fuda G, Calvo A, Mora G, . ALS clinical trials: do enrolled patients accurately represent the ALS population?. Neurology. 2011 Oct 11;77(15):1432-7. PubMed.
Hothorn T, Jung HH. RandomForest4Life: a Random Forest for predicting ALS disease progression. Amyotroph Lateral Scler Frontotemporal Degener. 2014 Sep;15(5-6):444-52. PubMed.

More Data, More Progress: ALS Researchers Debut Shared Database

Quick Links

Tools

Comments

Make a Comment

References

News Citations

Paper Citations

External Citations

Further Reading

Papers

News

Primary Papers

Annotate