Thomas Edison failed countless times before finally inventing the incandescent light bulb in 1879. But those failures, Edison argued, were simply lessons learned - necessary steps that led him to success.

Edison's path is a snapshot of history that today's scientists might occasionally need to refer to, especially if working in antisense. The whole business of drug development can be trial and error, of course, but the concept of antisense has been around for 25 years and still has not reached fruition. Researchers, though, aren't waving the white flag.

Yes, there have been failures: Isis Pharmaceutical Inc.'s Alicaforsen for one. In December, the drug failed in two Phase III trials in Crohn's disease. That same week, Berkeley Heights, N.J.-based Genta Inc.'s antisense drug Genasense failed in advanced multiple myeloma. But neither company has given up on their drugs, with Carlsbad, Calif.-based Isis recently completing a Phase II of Alicaforsen - the last of its first-generation antisense products - to treat patients with ulcerative colitis, and Genta recently filing a new drug application for its drug to treat chronic lymphocytic leukemia. If approved, Genasense could be the light bulb for the field, although Isis has gotten one antisense drug to the market: Vitravene, which gained FDA clearance in 1998 to treat AIDS-related cytomegalovirus-induced retinitis, a smaller indication.

Considering that those two products were not the first to disappoint, there isn't a lot of confidence around the science. Another Isis compound, Affinitak, also failed in a Phase III for non-small-cell lung cancer two years ago.

"I don't think it's well known what exactly might have caused those failures," said Reni Benjamin, an analyst with New York-based Rodman & Renshaw, who covers Hybridon Inc., Genta and AVI Biopharma, a company working on third-generation antisense products. "It could be the target. It could be maybe the antisense didn't stay in the body long enough. Maybe they had to improve their pharmacodynamic profile. Some people feel that antisense therapeutics will never work, but that needs to be shown in the clinic. I think that enough activity has occurred in clinical trials that it at least warrants further investigation."

There are fewer than 20 antisense drugs in development, and most are second- and third-generation products. Isis has announced its move into the second-generation candidates, which use alkyl modifications instead of sulfur modifications, and appear to be 10 times more potent. But Cambridge, Mass.-based Hybridon (now called Idera Pharmaceuticals Inc.) has said it will pursue antisense only with partners, preferring to use in-house resources for the new area of Toll-like receptors (TLRs).

Sailen Barik, a professor of biochemistry and molecular biology at the University of Southern Alabama in Mobile, said that part of the issue with antisense has to do with its specificity, or lack thereof, and its stability.

Since antisense technology emerged, researchers have worked to make its candidates more stable "without sacrificing its function or activity," Barik said. But those changes "also have made it a little more non-specific and more toxic."

Then, the focus changed with the introduction of RNAi in the late 1990s.

"The discovery of RNAi had nothing to do with antisense. It had a very different story," Barik said. "But in the end, people saw it could be used as an antisense."

Like antisense, RNAi can shut off disease pathways at the nucleic acid level, but it also might offer improvements in toxicity, stability and potency. Final results from the first RNAi strategies in clinical trials could be available by the end of this year, Barik said.

Antisense first emerged in 1978 to much enthusiasm. It gave scientists hope they could find drugs that would selectively turn off diseases by interfering with a gene's ability to express a protein, thereby blocking a disease pathway.

In developing antisense drugs, scientists target and interfere with messenger RNA (mRNA), an essential component in building proteins that lead to disease, thus silencing gene expression. But after millions of dollars being spent and little to show for it, particularly in the fields of cancer and infectious diseases, interest has waned from companies like Idera, as well as investors.

Idera's decision to focus on Toll-like receptors might have as much to do with that technology's potential as it does with the disappointments of antisense drugs.

"One of the reasons why [Idera changed its focus] is there is a little bit of a negative factor or connotation surrounding antisense now," Benjamin said. Idera may have strategically decided "to switch gears because [TLRs] are a very interesting space," he said. "Investors are interested. Companies like big pharma are interested in it."

Antisense drugs are single-stranded RNA molecules that hybridize with the mRNA to block protein translation, while short interfering RNA (siRNA) consists of double strands of RNA that induce destruction of the mRNA of the targeted gene via a complex known as RNA-induced silencing complexes, or RISCs.

TLRs, on the other hand, appear on the surfaces of antigen-presenting cells and help regulate immune responses to bacterial, viral and fungal infections. The first one was cloned in 1998. They have garnered a great deal of interest from companies such as Basel, Switzerland-based Novartis AG, which signed a $570 million deal in June with Anadys Pharmaceuticals Inc., of San Diego.

While TLRs represent new research, antisense now falls under the umbrella of RNA therapeutics, as does RNAi, siRNA, microRNA and RNA-based aptamers. Second-generation antisense is a brother to first-generation antisense, and RNAi is a cousin, said Mark Monane, an analyst with New York-based Needham & Co. Inc., which covers Isis.

"What's fair to say is the focus on antisense has been consumed under the banner of RNA medicines," he said.

Earlier this week, two major players in RNA therapeutics, Isis and Cambridge, Mass.-based Alynlam Pharmaceuticals Inc., signed a co-exclusive license agreement with Stanford University to discover and develop products to treat hepatitis C virus infection by inhibiting liver-specific microRNA. And last week, Novartis acquired almost 20 percent of Alnylam as part of a potential $700 million RNAi agreement.

While RNA technologies might not be traditional antisense, they are built on the same principles.

"It's a pretty exciting field," Barik said, adding that company laboratories are working to find more stable and more specific formulations to use RNA therapies "at a lower dosage to minimize the side effects and toxicity."

"It's getting there," he said. "It's more promising than antisense at this point."

In November, San Francisco-based Sirna Therapeutics Inc. started the world's first clinical study of a chemically optimized siRNA, Sirna-027, to treat patients with wet age-related macular degeneration. Interim results presented in May showed that Sirna-027 was safe and well tolerated, with no systemic or local adverse events, and all patients tested experienced visual acuity stabilization, while almost half had improvement in their vision.

But some up and coming antisense drugs also have shown positive results, at least in early trials. In the summer, Portland, Ore.-based AVI BioPharma said studies conducted at the Scripps Research Institute in La Jolla, Calif., showed its Neugene third-generation antisense compounds were able to inhibit infection by the severe acute respiratory syndrome coronavirus. The company also submitted an investigational new drug application to the FDA for AVI-4065 to treat hepatitis C virus. And its lead antisense compound, Resten-NG, is expected to enter Phase III after Phase II data showed statistically significant efficacy in preventing restenosis.

There is still promise in the field, analysts say.

"It's not unlike antibodies. Everybody loves them now," Monane said. "But the first-generation antibodies were murine antibodies. They were non-human."

The murine antibodies evolved into humanized antibodies, which led to that field's light bulb, Herceptin. South San Francisco-based Genentech Inc.'s drug became the first humanized antibody approved in 1998 to treat HER2+ metastatic breast cancer.

It would be unwise to misconstrue a strategic move by Isis to focus on RNAi as the end to all antisense therapies.

"It's the next step in evolution," Monane said. "And I think that's what we're seeing in antisense and RNA medicines. So I don't think it's fair to say that antisense is dead. RNA medicine is the umbrella now in thinking about these kinds of therapies."