Rx for Lilly To Save Its Insulin Business

The secret is out: Eli Lilly and Company (Lilly) has an insulin problem. That was made abundantly clear by the company's January 5, 2006 announcement that it would halt construction of a new insulin plant in Virginia because they now anticipate being able to meet North American demand from existing factories. While demand for the company's insulin is still growing, its just not growing as fast the company forecast it to grow when plans for the new factory were announced in 2002.

According to U.S. insulin sales volume data compiled by IMS Health Inc., a leading global source for pharmaceutical market intelligence, Lilly claimed as much as 85% of the U.S. insulin market back in 1995, and as recently as 2000, the company still commanded an impressive 82% of the market. But during the past 6 years, Lilly has really squandered its near-monopoly of the American insulin market by ignoring the needs of the market and deploying insulin profits, research and marketing dollars into to other, unrelated therapeutic areas (including treatments for such tragic ailments as erectile dysfunction and urinary incontinence). In 2006, Lilly's share of the U.S. insulin market had plunged to just 43%. As I reported previously, in September 2005, Lilly ceded its title as the market share leader for U.S. insulin sales to rival Novo Nordisk, the first time since the discovery of insulin. This is no small problem; insulin still accounts for just under one-third (~30%) of the company's total worldwide sales, and Humalog remains one of the company's biggest "blockbuster" products, with worldwide sales of more than $1 billion per year -- only Zyprexa (for schizophrenia) and Gemzar (an expensive cancer treatment) sell more. Considering the fact that Humalog is now a decade old, that's still pretty good!


This is a new position for Lilly, and one the company's not accustomed to being in. Although Lilly still produces nearly 25% of the world's insulin, since the U.S. accounts for so much of the world market, the company's share of the global market is also shrinking. While Humalog still holds more than 50% of the U.S. market for rapid-acting insulins, it faces increased competition from Novo Nordisk and Sanofi Aventis and has had difficulty increasing its share of that product category. Analysts have predicted that insulin sales could potentially more than double in the next four years. But Lilly lost the giant United Healthcare account at the beginning of 2006, and the company doesn't even compete in the long-acting insulin analog market, forcing doctors and patients to buy these from competitors even if they wanted to buy exclusively from Lilly.

The Indianapolis Star reported, "Lilly's problems in the insulin market grew in part out of a tough choice the company made seven years ago."

Facing a cash crunch when its blockbuster antidepressant Prozac lost its U.S. patent protection in 2001, Lilly knowingly chose to take money from its insulin marketing budget at a time when insulin sales were poised to take off, and the company dramatically cut back its insulin sales force five years ago.

Ron W. Hoven, Lilly's former brand leader for diabetes care, told the paper from Lilly's hometown that "The company was forced to make some very difficult trade-off decisions."

When that occurred, doctors suddenly noticed that Lilly's sales reps stopped handing out free insulin samples, which had become a leading marketing tactic.

"You couldn't find a Lilly insulin in your fridge," said Dr. Luigi Meneghini, director of the diabetes clinic at the Diabetes Research Institute in Miami.

While this may partially explain the decline, those were marketing decisions, and it takes much more than marketing to compete in the pharmaceutical business. For example, it does not explain why Lilly put all of its research resources into other therapeutic areas. Today, the company cannot even serve diabetes patients' full physiological needs, as Lilly does not offer a long-acting insulin analog, and so far, hasn't announced anything in development on this front, either. Not one new insulin is in Phase II or III stage development according to Lilly's most recent annual report. Kelly Close reported from the 19th World Diabetes Congress sponsored by the IDF in Capetown back in December that Lilly never mentioned a long-acting analog was even in development at any of the company's symposiums.

Although the Indianapolis Star claims that Lilly Labs tried to develop its own long-acting analog, they also report that the company was "unable to develop a comparable product to Lantus." Frankly, I am not convinced Lilly put much effort (or money) into doing so. After all, this is the company who invented Humalog. I still believe Humalog may in some ways be superior to all of the other rapid-acting analogs on the market (although because its not in my drug formulary, I now use Sanofi's Apidra and I'm pretty pleased with it). One of the reasons I believe Humalog may be superior is because Humalog is the only insulin analog that contains all of the same amino acids that are found in regular human insulin. The same cannot be said for Novolog or Apidra, both of which have introduced new amino acids into the insulin structure which do not exist in human insulin (or for that matter, in insulin found in any other species). Humalog's structural simplicity (two amino acids, lysine and proline, have had their positions on the beta chain of the insulin molecule reversed) is also admirable from a chemical engineering standpoint. This suggests Lilly could probably develop a long-acting analog if they really tried and allocated a sufficient budget towards it. But it is rumored that Lilly executives were very skeptical about the need and demand for a product like Lantus, which has since gone on to become the world's best selling insulin analog.

David Kliff, the publisher of Diabetes Investor, a Chicago newsletter for investors in diabetes-related companies, thinks that Lilly was "caught flat-footed" by the success of Lantus. "Maybe they were a little bit arrogant about it," he said, remembering a diabetes trade show he attended where employees at the Lilly booth scoffed at the potential for Lantus to become a big seller.

Lilly's insulin business wasn't always in such dire straits. Lilly was the first company to commercialize the insulin derived from animal pancreases way back in 1923. In 1982, Lilly brought the first human insulin of recombinant DNA origin to market (although truth be told, Genentech developed the technique used to produce Humulin back in 1978 and sold it to Lilly). As recently as 1996, Lilly was the first company anywhere on earth to introduce an insulin analog, and at the time it was introduced, Humalog was billed by the company as the first new insulin to be introduced in over 14 years.

Today, the company now needs to play "catch-up" with Novo Nordisk and Sanofi-Aventis. But drug development takes years (if not decades), and had Lilly invested anything in this market during the past decade, they might at least have something in their drug pipeline by now, but since they didn't, their fastest option is to pour money into marketing, another area they did not invest in. Lilly announced plans to do just that by increasing its U.S. insulin sales force by 40% last year, and an additional 50% this year. But that is more about distributing its products, its not about meeting market needs, and the consensus is that Lilly was forced to do this in order to compete with Novo's plans to add another 800 salespeople (ultimately bringing Novo's total salesforce in the U.S. to some 1,900 nationwide).

But Lilly has proven that its marketing and product development teams can accomplish great things if they try. As I reported last June, this year, Lilly will launch a slick new insulin pen device called the Humapen Memoir, which has a memory that stores information on the last 16 doses, and is truly an innovative product (although I told the company they need to launch one that can dose in 1/2 units, as the core type 1 market needs to be able to deliver much more precise dosages than the type 2 market given their sensitivity to minor changes in dosage). But whether that is sufficient to turn Lilly's insulin business around seems questionable. They need something else to differentiate their insulin products from all the "me-too" analogs now on the market.

Differentiation from Lilly's Labs

So what could Lilly do to turn this business around? Sure, they could partner with another drug company to perhaps expand their product line, and they may need to in order to complete their insulin product portfolio. But as as the current leader learned (and I reported in my last posting), its tough to topple Lantus. Lilly's partnership with Amylin Pharmaceuticals has been an outstanding success indicating the company has a track record of successful partnerships, although Lilly was quite unprepared for the success of both Byetta and Symlin. But in my opinion, Lilly really need look no further than their own research laboratories to give their products a differentiating edge over rivals.

Nearly a decade ago (back in July 1997), Lilly helped fund a joint study with Washington University at St. Louis and the National Institutes of Health, and the researchers discovered that in diabetic rats, treatment with C-peptide reversed vascular and nerve damaged blood vessels and helped to repair them. In the rats treated with C-Peptide, the nerve cells worked normally and vessels almost completely stopped leaking.

C-Peptide is a by-product of the production of insulin, and is present in non-diabetic people but is scarce or completely absent in people with type 1 diabetes, and for many with type 2 diabetes, it is also in relatively short supply.

Although some researchers back in the 1970's began asking if patients with type 1 diabetes might be suffering from a lack of C-peptide, a few trials suggested that C-Peptide did nothing. Most notably, back in 1993, Julio Santiago, M.D., professor of medicine and pediatrics at the Washington University School of Medicine injected human patients with type 1 diabetes with low doses of C-Peptide -- just enough to match normal levels -- but saw no effects. The conclusion was that the protein was merely a byproduct of insulin production and served no physiological purpose. However, some researchers questioned whether this approach may have overlooked the fact that patients had been without C-Peptide for many years.

Trying a different approach, the team at Washington University tried injecting diabetic rats with much larger doses of synthetic human C-peptide, exceeding the levels of C-peptide that rats produce naturally. The results were dramatic: nerve cells worked normally and vessels almost completely stopped leaking. Because relatively large doses were needed to achieve the effect, researchers suspected that C-peptide therapy could also help type 2 diabetes patients who already have normal levels of the protein.

The 1997 finding suggested that the long-accepted view of how proteins affect cell function was far from the whole story. The study was described in the July 25, 1997 issue of the journal Science.

"Since this protein is so effective at preventing and reversing vascular leakage, it brings up the possibility that it could prevent cardiovascular disease in both types of diabetes," said Dr. Joseph R. Williamson, M.D., professor of pathology at the Washington University School of Medicine and senior researcher of the 1997 study.

Swedish researchers have subsequently validated and replicated this research, and recently, took it a step further: testing it on humans. In the January 2007 issue of the journal Diabetes Care, researchers at the highly-regarded Karolinska Institute in Stockholm concluded that C-peptide treatment for 6 months improves sensory nerve function in early-stage type 1 diabetic neuropathy. Of course, in the time that Lilly has let lapse since their initial discovery, the Swedes have since patented the manufacture of "recombinant insulin C-peptide". I guess the saying "you snooze, you lose" is applicable here.

The resulting company is a firm called Creative Peptides, a development stage biopharmaceutical company, engaged in the discovery and development of biopharmaceutical drugs to treat long-term complications of Type 1 diabetes. There is no reason to believe Lilly could not partner with them. Creative Peptides' lead drug (synthetic C-Peptide made from rDNA origin) development project is currently in phase II clinical trials.

It is my understanding that in the manufacture of synthetic human insulin, the C-peptide chain must actually be removed from the pro-insulin to become what we know as synthetic insulin. In effect, the manufacturers are removing the beneficial C-Peptide before they sell it to us, and I cannot help but wonder if U.S. patent law would prohibit Lilly from selling insulin with the C-Peptide chain too? This is a part scientific, part legal question, but I am certain that someone at Lilly should try to answer it. The bigger question is will they, or are they more interested in accelerating their inhaled insulin that the entire pharmaceutical industry has already concluded will be a blockbuster (in spite of disappointing sales for Exubera so far)?

***** EDITOR'S NOTE: For what its worth, Lilly's President and Chief Operating Officer, John C. Lechleiter, wrote a letter to the Indianapolis Star in response to the article on the company's decision to pull the plug on the Virginia insulin plant entitled "Lilly battles pandemic of diabetes". Although I am not convinced by it, readers may wish to be aware of his letter. *****