Novartis showed off its new drug pipeline for the first time in three years Nov. 19, but the R&D day was as much about vindicating the novel research strategy it rolled out in 2002 as it was previewing the work in process.

Six years ago the Swiss pharma launched the Novartis Institute for Biomedical Research in Cambridge, Mass., with Harvard Medical School cardiologist Mark Fishman at the helm and a decidedly more academic than industry bent to its methodology: NIBR would first identify a molecule's pathway and method of action, then pick a disease mediated by that pathway to try the molecule in.

Further, the starter diseases are not the big indications typically sought first by drug developers, but afflictions with small, well-characterized, genetically known patient populations. The idea is to get proof of concept in humans early on, in a Phase I/IIa study that from the get-go includes afflicted patients, then initiate parallel early stage inquiries in other diseases affected by that pathway right away (see "Novartis' Research Experiment," ¹ Also see "Novartis's Research Experiment" - In Vivo, 1 May, 2006.).

The Fishman experiment appears to be working.

Building a "soup to nuts" system in a "big company where you have tremendous resources, but you have to get these groups working together," has taken five or six years, Novartis global development chief Trevor Mundel acknowledged. "But what you can see is where we are now and how well it has actually functioned."

Novartis will have a lot of pipeline activity over the next 18 months, with 24 new filings planned over the next two years, he said.

At its last R&D update, in 2005, Novartis reported 75 projects in the clinic, with 46 new molecular entities, 52 of them in Phase II or beyond - including the ill-fated diabetes drug Galvus (² Also see "Novartis Sticks With Heart Of Pipeline: Growth Is In Cardio-Metabolic Sector" - Pink Sheet, 26 September, 2005., p. 22).

Novartis' candidate portfolio has grown 40 percent since then, Fishman said. While the entire pipeline in 2005 boasted 63 NMEs, today there are 88, and NIBR anticipates reliably producing two a year within the next five years. Even better, half of the new-model POC trials are working, their programs heading into full development, while the other half fail in less-expensive early studies.

Altogether 80 percent of the candidates made the transition from POC to Phase II/III during 2006-2007, a 60 percent improvement over attrition in 2003-2005.

The best way to evaluate the new NIBR model, of course, is to look how it has played out in the pipeline. Perhaps not coincidentally, Novartis' next product submission will be for a monoclonal antibody that exemplifies the strategy of understanding the science, pursuing a small clear target indication in an accelerated clinical program and simultaneously moving ahead on larger follow-up indications.

Next Submission A Graduate Of The New Model

The first BLA filing for Novartis' experimental interleukin-1b blocker ACZ885 (canakinumab) will be completed by the end of 2008, months ahead of schedule, Mundel announced during the update.

Mundel, a translational sciences expert, assumed the global development role following James Shannon's departure earlier in the year as a result of Novartis' pharma division shake-up (³ Also see "Drug Development Chief Departs Beleaguered Novartis" - Pink Sheet, 20 March, 2008.). He was replaced by Timothy Wright, who previously did translational work in inflammatory drug development for Pfizer.

Soon after he took the position, he spotted ACZ885 and pushed for an accelerated timeline. "Looking across the portfolio, this one [program] has such potential for expanding into many indications, starting from the rare indication right through diabetes."

ACZ operates on the cryopyrin inflamasome (NALP3) pathway. In patients with NALP3 gene mutations, IL-1b, normally activated by an inflammatory event, becomes autoactivated and creates autoimmune disease. The same gene can have different mutation sites, causing autoinflammatory diseases with a spectrum of onsets and severity.

Novartis is starting with a rare disease where the risk-benefit argument should be simple. The December filing will seek approval to treat a group of rare autoinflammatory diseases, called Cryopyrin-Associated Periodic Syndromes, that includes Muckle-Wells Syndrome, an "intermediate level" disease on the NALP3 spectrum that has an estimated market of about 1,000 patients.

After a Phase I/IIa trial in Muckle Wells concluded "a short time ago," NIBR expanded the number of subjects, switched from I.V. to subcutaneous injection, did dose ranging within the study, and went right to Phase III on the way to the December application goal, Wright explained.

Novartis is already moving ACZ into other diseases where NALP3 is or may prove a factor. A Phase III program in systemic juvenile idiopathic arthritis (11,000 potential patients) launches in 2009, a Phase II trial in gout (population 3 million) is ongoing, and a Phase III in adult rheumatoid arthritis (3 million with the gene mutation) is planned for 2010.

As Mundel suggested, Novartis also is trying ACZ885 in diabetes and chronic obstructive pulmonary disorder, where Wright predicted it will show utility in a subset of patients.

Novartis' pipeline includes a second autoinflammatory approach, with an earlier-stage antibody, AIN457, that targets the IL-17 pathway. Novartis started out testing the molecule against RA, but on a signal that IL-17 is important in skin inflammation, added a psoriasis trial mid-stream, Wright said. After a single dose, patients experienced rapid, profound response, some sustained out six months, with no safety signals in either study, he said. Currently the antibody is infused, but researchers are switching it to delivery by subcutaneous injection.

Following the Fishman model, NIBR now has AIN457 in a Phase II psoriasis study with parallel indication-expansion studies going in RA, Crohn's/irritable bowel disease, COPD, uveitis and ankylosing spondylitis.

The Future Will Rely On Patient Segmentation

Fishman dismissed the suggestion that NIBR should not pursue therapies in a RA/psoriasis market dominated by Amgen/Wyeth's Enbrel (etanercept), particularly with the advent of etanercept biogenerics, likely under development by Novartis' Sandoz division, a near-term possibility.

Instead, he's advocating the idea that biomarkers will help segment that market and identify still-available opportunities. Fishman projected that within the next few years, predictive biomarkers will identify patients who are high responders to various therapies. "There are many patients who don't respond at all or very well to Enbrel and anti-TNF therapy. One of the things we're exploring is whether those are the same patients who might respond well to ACZ, anti-IL-17, anti-IL-1 compounds," he said.

"We won't be taking ACZ forward in RA without a biomarker. That's the approach we're taking: identify the high responders, and we already know we have very high responders to ACZ there.... The future will rely on patient segmentation."

Novartis' pharma division plans to establish a diagnostics unit in 2009 to support NIBR's companion biomarker strategy and commercialization of the biomarkers (see ⁴ Also see "Novartis Pharma To Add Diagnostics Unit In 2009" - Pink Sheet, 8 December, 2008.).

Parkinson's Side Effect Is First AFQ056 POC

Another of Novartis' new-model programs is built around AFQ056, a glutamate receptor subtype 5 (mGluR5) inhibitor in studies as a therapy for glutamate-mediated movement and pervasive development disorders, Wright said.

AFQ056 achieved POC as a novel treatment for L-dopa-induced dyskinesia, the involuntary twitches suffered by about half of Parkinson's disease patients as a side effect of chronic dopamine therapy. In a one-month, randomized, controlled trial, patients showed improvement in motor control with no deleterious effect on treating the disease.

A parallel study testing AFQ056 as a treatment for the movement disorder component of Fragile X syndrome, a common genetically determined form of mental retardation, will read out shortly, and a third study in an unnamed neurological disorder will start within the next two months, Wright said.

Potential indications for AFQ056, which Novartis expects to submit in the PD-LID indication around 2012, are primary dystonia, tardive dyskinesia and autism spectrum disorder.

NIBR has developed a PET ligand, a labeled molecule to use as a biomarker to indicate the level of GluR5 receptors in the brain and their occupancy, Wright explained. The biomarker will enable investigators to better pick dose and dose range going forward, he said.

- Shirley Haley (s.haley@elsevier.com)

Novartis Has Proof Of Concept For New R&D Model, And NMEs To Show It

More from Archive

More from Pink Sheet