All quiet on the regulatory front, mostly
FDA defines combination product as one “comprised of two or more regulated components” – drug, biologic, or device (including diagnostics) – that are distinct. Thus, a lidocaine patch qualifies (device plus drug) but an acetaminophen plus codeine tablet (two drugs) does not.
Notwithstanding FDA’s “final rule” on combination products, promulgated in 2005, regulations have not significantly changed since FDA instituted the combination product office in 2002.

FDA’s Office of Combination Products (OCP) assigns regulatory oversight for combinations through a formal “Request for Designation,” to one of FDA’s three major Offices (CBER, CDER, CDRH) overseeing biologics, drugs, and devices, respectively. According to former OCP director Mark Kramer, the Agency applies regulation from all three Offices, as appropriate, for every combination product.

The Request for Designation is a sponsor’s best opportunity to make a case for their preferred status. “If a sponsor can provide adequate documentation and justification for a particular designation, they have a fair chance of at least having input into that decision,” notes Rosina Robinson, principal consultant at Medical Device Consultants (MDC; North Attleboro, MA), which advises device and diagnostics companies.

“If similar products already exist, or are under review, we could assign the applicant to the center with the most expertise in that area,” notes Kristina Lauritsen, Ph.D., a scientific reviewer at OCP. Barring that, FDA considers be the most significant safety and efficacy factors. Products that still fail categorization, usually by virtue of multiple modes of action, are subjected to an assignment algorithm.

FDA could not provide information on types of combinations submitted to and approved by OCP. Nor does the agency formally track incoming Requests for Designation by FDA Center. But data from OCP’s website for October 1, 2006 through September 31, 2007 suggest that requests for designation overwhelmingly involve devices plus a drug and/or a biologic. Of 36 RFDs submitted during this time period all but one fell into that category. Two additional submissions were for a “biologic and tissue.” According to FDA’s data, regulatory review times are somewhat shorter for combination products than for non-combinations submitted to the same FDA office.

One of the initial steps when an application hits FDA is to define its “primary mode of action”—basically, will the product be treated as a drug or as a device? While the most popular combinations—drug-releasing patches and stents—are generally approved as devices (the drugs tend to be generic), PMAs are not written in stone for entire product classes. Tomorrow’s stents may hold innovative drugs specifically developed for the device, or multiple drugs, or a drug-biological combination, which could tip the PMA balance for this product class towards CDER or CBER. Advances in medical materials could at some later date push this class of product back into the device category.

Future combination products may raise PMA questions that border on the philosophic. For example, REVA Medical (San Diego) and Rutgers University have co-developed a bio-erodible drug-eluting stent that dissolves after several months. It is not inconceivable that the label for the eventual product, which entered human testing in 2007, could stipulate drug co-therapy that continues long after the stent disintegrates. Which FDA office oversees approval of a combination in which the device is innovative but temporary, while the drug is a generic but treatment with it lasts years or a lifetime?

The Combination Products Coalition (Washington, DC), an independent organization, has proposed its own guidance and assignment algorithm, available at the organization’s website (combinationproducts.com). Although she had not heard of CPC (which did not respond to interview requests), Lauritsen said her office was open to any suggestions.

The consensus view is that FDA has come a long way towards providing a fair, open regulatory path for combinations. “One reason it’s easier than ten years ago is the agency has done a good job of publishing their decisions. Companies now have a much better idea of the regulatory environment in which these products are approved,” says Rosina Robinson of MDC.

MARCO BAFAN, NERAC

Despite progress, there are some indications that FDA Offices do not always cooperate as smoothly together, and with OCP, as they could. According to Marco Bafan, an analyst with research advisory firm Nerac (San Diego), ambiguity persists in oversight disputes whose resolutions carry immense economic implications for sponsors. Although reimbursement as a percentage of cost of goods is higher for drugs than for devices, companies prefer a device designation because it represents the clearest, best-defined route to approval through a 510(k) (essentially equivalent) or supplemental approval route.

Issues can arise post-marketing when a product with a device PMA is associated with drug-related adverse events (or vice versa). “Clearly CDRH, which holds oversight, lacks expertise to handle that type of event,” Bafan notes, particularly when safety concerns may initiate label changes.

From his dealing with clients, Bafan has picked up on friction between the CDER/CBER and CDRH, particularly on matters of drug safety. Meanwhile CDRH has reportedly had issues with CBER and CDER not communicating details on drug/biologic components completely, or in a timely manner.

No slam-dunk
Metabolic diseases have become a fertile area for developing combinations of drugs or biologics and devices. But these ventures are not without risk, even for sure-bet markets like diabetes.

Approved in January 2006, Pfizer’s Exubera inhaled powdered insulin combination product was hailed as the first advance in diabetes care since the hypodermic needle.
Consisting of innovative formulation and device, Exubera had all the ingredients of a blockbuster. Yet Pfizer sold just $12 million of the product in the nine months post-approval. By late 2007 Pfizer had withdrawn Exubera and subsequently took a related $2.8 billion charge. Co-developer Nektar Therapeutics learned about Exubera’s demise from a Pfizer press release.

There is no shortage of theories as to why Exubera failed. Lungs are less-than-ideal entry points for systemic drugs. Add to this the long success of injected insulin and the fact that most diabetics are not insulin-dependent, and the prospect for inhaled insulin seems less rosy. By contrast drug-device combinations that treat respiratory diseases, which affect close to 50 million Americans, target the affected organs while limiting systemic exposure.

Warren Levy, CEO of peptide therapeutics firm Unigene (Fairfield, NJ), says Exubera’s failure resulted in part from its cumbersome, unwieldy device that did not allow fine-tuning the dose, an important issue with diabetics. “Physicians were also wary of the dosage form, which consisted of insulin-coated particles whose inhalation caused a temporary decrease in lung function in some patients.”

Complexity is a common problem with self-administered drug-device combinations. A 2007 study in the journal Respiratory Medicine found that just 38% of asthmatics who did not receive formal instruction used their inhalers correctly. Of those receiving training only 48% operated the inhaler properly.

Pfizer’s debacle left three major players and one small biotech in the once-promising development arena for inhalable insulin. Then, in January 2007 Novo Nordisk pulled out. CEO Lars Sorensen stated in a press release that his firm’s AERx inhaled insulin was “unlikely to offer significant clinical or convenience benefits” over Novo’s own pen injectors.
The next-to-last size 14-EEE shoe dropped in mid-March, when Eli Lilly terminated its Phase III AIR insulin product, co-developed with Alkermes. In an interview in Chemical & Engineering News Lilly CEO Sidney Taurel cited uncertainty in gaining approval and capturing significant market share.

That leaves two companies standing: small-cap Generex Biotechnology (Toronto, ON), which is developing its Oral-Lyn inhaled insulin, and MannKind (Valencia, CA) with Technosphere Insulin. Both products are just entering Phase III testing. MannKind, whose share price took a dive after the Lilly announcement, is perhaps staking its future on inhaled insulin. As you read this, the company is putting the finishing touches on a 250,000-sq-ft renovation/ construction project, in Danbury, CT, for a plant it hopes will manufacture Technosphere.

Beating the “me-too” rap
The largest class of combination products to date comprises simple devices with off-patent drugs, prompting this observation of Steve Richter of Microtest: “The lion’s share of approved combination products are quite ordinary.”

Combination products rarely involve New Molecular Entities, but that will change as more biologics assume the form of combination products. “For now the device is usually the novelty,” observes Dale Robinson of PA Consulting. As the culture of combination products matures, however, drug developers may look to devices as the “commodity” component of an innovative therapeutic. “Some new drugs may only be available through a unique delivery system,” Robinson says.

STEVE RICHTER, PHD, MICROTEST

While small-molecule/device combinations will probably dominate for the foreseeable future, the greatest opportunities will arise from innovative products approved under drug and biologic designations, or those involving tissue engineering or regenerative medicine.
“We’re waiting for the new biotech piece to this puzzle,” says Microtest’s Richter, “particularly with respect to drug-biologic combinations.”

Combinations involving cell- or tissue-based treatments would likely include one or more drugs or biologics, and perhaps a device as well. But as Richter points out cells and tissues represent a regulatory gray area, “There are significant regulatory gaps for them that FDA needs to address.”

One buzzword to emerge from combination products is “convergence,” which has come to mean the combination of two or more distinct scientific disciplines in one medical product. In a 2006 paper in Nature Biotechnology, Robert Langer of MIT and Scher Shmulewitz at Medgenesis Partners (Tel Aviv, Israel) describe a future in which information technology, microprocessors, smart materials, and microelectromechanical devices dominate the high-value end of the combination product marketplace.

Diagnostic products already employ some of these technologies. From there, integration with drugs does not seem too much of a stretch. The nearest-term promise lies in nanoparticles, most likely of the erodible kind, for drug delivery. Nanoparticle-drug/biologic combinations might also renew interest in therapies that combine a therapeutic and a targeting component, for example a drug plus a magnetic particle, or a drug-particle-antibody combinations. Drug discovery conferences were awash with antibody-drug combinations during the late 1980s and throughout the 1990s, but enthusiasm waned with emerging safety and efficacy issues.

Microelectronic and micromechanical sensors are now inexpensive and reliable enough to implant in the human body, to control delivery of dosage-sensitive drugs and biologics for oncology, cardiovascular, and metabolic disorders. Recognizing this, IBM and Hewlett Packard have invested in healthcare nanotech, as have Dow Chemical, Medtronic, and Alza on “smart” drug delivery systems. MicroCHIPS (Bedford, MA), which is affiliated with Langer’s group at MIT, has a diabetes program based on wireless sensing and actuation of embedded chips. MicroCHIPS has received equity investments from device giants Boston Scientific and Medtronic, and on the pharmaceutical side from Novartis.

An artificial pancreas based on advanced materials and microsensors would very quickly make everyone forget about inhaled insulin. PC