The history of antisense drug development is filled with disappointments. Researchers have had as rough a time coming up with compounds that work effectively in humans as others did with monoclonal antibodies. Both are tales of faded early promise and wilderness years spent fixing what went wrong. But monoclonal-based therapeutics have worked out: Nine are FDA-approved and nearly 100 more are in the clinical pipeline. Today, there's every reason to believe that antisense therapeutics too will come into their own: Already, 10 drugs are in the clinic and one has been blessed by the FDA. If the toughest years of antisense development now seem to be ending, the enduring antisense companies may also be headed for major success, especially in this post-genomics era, for which their tools seem tailor-made.
Antisense therapeutics must once have seemed a wonderful path to biotech glory -- it was all so straightforward. Design an oligonucleotide to bind a gene’s mRNA, inhibit its translation, choke off its protein, and thereby exert a therapeutic effect; all you need is to know is the gene’s sequence. Then quick into the clinic for testing. The side effects probably won’t amount to much. Invite the FDA to make its nod, don’t forget to be modest when the acclaim rolls in, and truck the cash home in silver-plated wheelbarrows.
A lovely daydream. But write it down, whip it out, strike it rich has nothing to do with what actually happened. A more than twenty-year, elusive chase after the idea of a lightening-quick drug discovery method has trekked across some hard ground. It has taken unflagging doggedness to make antisense drug technology work. Churchillian phrases are not at all out of place to describe the resilience and fortitude antisense development demands, surveying obstacles with “stern and tranquil gaze,” ready when necessary to “nerve the arm of strife.” These pioneers suffer no doubts that antisense will be ascendant. Its finest hours lay ahead.
“The question isn’t any longer will antisense deliver a drug,” says Frank Bennett with confidence: “It’s what will be the next antisense drug.” The question he doesn’t answer -- because no one yet knows for certain -- is what the impact of antisense drugs on medicine will be.
Bennett is VP of antisense research at Carlsbad CA-based Isis Pharmaceuticals Inc. -- which has the unique honor of bringing to market the first antisense drug. That compound, Vitravene (fomivirsen), was approved by the FDA in 1998; it's a 21-base long oligonucleotide (GCGTTTGCTCTTCTTCTTGCG) for treating cytomegalovirus (CMV)-induced retinitis. If one drug that inhibits gene function by binding to its mRNA and blocking protein translation can be made, certainly others will follow.
But Vitravene is also the only antisense drug to win FDA approval, the solitary commercial achievement of an idea reaching back to 1978, when Stephenson and Zamecnik showed an antisense oligo inhibited replication of Rous sarcoma virus in vitro. (Proc. Natl. Acad. Sci. USA. 1978 Jan; 75(1):285-8.)
Isis is proud of Vitravene, and should be; it combats opportunistic CMV infections of the eye that destroy the retina and reduce AIDS patients to blindness. The problem with Vitravene outside the antisense community is that it’s a drug for a small disease, and lacks the financial significance (past 12-months sales: $157,000, according to IMS Health) that's often needed to get the attention of either analysts or potential partners.
Image courtesy Isis Pharmaceuticals Inc.
Vitravene was always destined for a small market, now even smaller thanks to the success of HIV protease inhibitors in delaying AIDS onset. Until a blockbuster antisense drug come along that makes stock prices pop like a champagne cork, in some eyes antisense will continue to look like just another brilliant idea whose fizz fell flat.
Monoclonal antibodies, the other great drug idea from the 1970s, also suffered from a glitter-to-ashes stigma, and for a very long time. The earliest clinical forays were disasters, and monoclonal makers spent years painstakingly figuring out how to make them work. (See the Signals article, "Companies Load Up On Magic Bullets," for more details.) “But they were persistent. Everybody forgets how long it took,” says Sudhir Agrawal, chief scientific officer of Hybridon Inc., of Cambridge MA, one of the most prominent companies dedicated to antisense.
Antisense drug makers identify readily with the monoclonal antibody story, seeing themselves in a parallel saga of struggle giving way at last to success. Now they are focused on drug targets with major market potential, understand what to do about side effects, and believe to the marrow that the wilderness years of antisense, of disappointing clinical trials, desiccated wallets and doubting peers, finally are drawing to a close.
Probably no one who has kept up with antisense developments disagrees with Isis' Bennett that “antisense is the most direct technology to capitalize on genomics.” Antisense companies believe they can come up with drug candidates faster than anyone else, filling their pipelines from sequence data ready for the plucking. They also have the advantage of working with compounds (nucleic acids) that exhibit pharmacological uniformity -- a development and formulation luxury unknown for proteins or small organic molecules -- giving them a head start in moving drugs from the lab to the clinic.
The antisense companies are ready for the race.
The three largest antisense companies -- Genta Inc., Hybridon and Isis -- were formed in the late 1980s, nearly a decade after the field’s first entrant, AVI BioPharma Inc., set up headquarters in a Corvallis, OR garage under the name Antivirals Inc. That these four have just one FDA-approved drug among them speaks volumes about the challenges they've encountered along the drug-development path, not to mention the time lost in following false trails. “Antisense is very simple, very elegant,” says Hybridon's Agrawal. “But in the last 10 years we found that there is a hidden message in the DNA.” Unraveling the reasons for the body’s responses to antisense drugs has made them safer and more predictable to use.
The safety problems of monoclonal antibodies, originally made from mouse antibodies, were tamed by making them more human: The closer their amino acid residues resembled human residues, the more acceptable they became to the body. As a result, many monoclonals today are of fully human sequences.
Antisense development went in the opposite direction, surmounting problems by making the molecules less human. The phosphodiester backbone of natural DNA was one of very earliest things to go; natural backbone oligos didn’t withstand bloodstream nucleases long enough to be useful. All antisense drugs now have some other sort of backbone for better stability. Phosphorothioates, although slated for retirement, are still the trade’s workhorse backbones.
Other early problems concerned the human immune response. Bolus injections of antisense drugs activated the immune system’s complement cascade and caused unacceptable blood pressure effects. That was solved by switching to two-hour infusions.
More elusive was the CG problem. Hybridon experienced it while testing an early drug candidate against human papilloma virus (HPV). Cell culture and lab animal experiments showed dose-dependent repression of HPV. But the delight of seeing a true antisense effect turned to consternation when the drug also inhibited a second strain of HPV whose target mRNA mismatched the oligo drug by five bases -- ruling out the explanation that the drug inhibited the second virus by blocking its translation.
What emerged was that a CG (cytosine-guanine) dinucleotide in the oligo had indirectly stimulated an immune reaction against HPV. CG occurs commonly in bacterial and viral DNA, but far less frequently in human DNA than random association predicts. The human immune system exploits this distinction by treating CG dinucleotides as an early warning signal of infection, releasing immune cytokines and activating TH1 T-helper cells. Hybridon’s gotten around this problem by using mixed backbone oligonucleotides, consisting of RNA phosphorothioate nucleotides for CGs and DNA nucleotides for the other bases. Hybridon researchers have found that RNA-based CG doesn’t provoke the immune response. Other companies have their own methods of dealing with the CG problem.
Antisense companies gradually learned how to design around CG and other structural motifs causing side effects, producing antisense drugs that square with the requirements for efficacy: stability, efficient cell penetration, specificity for the target mRNA, and high affinity of hybridization. And they wisely selected drug targets that accommodated the limitations of a nascent technology. A key decision, taken early by Isis with Vitravene, was to favor medicines that could be locally applied. Vitravene is injected into the eye about once a month. “A systemic disease where you treat long term would have put a lot of pressure on manufacturing,” Bennett recalls, “so we looked at more local therapies where we would not have to manufacture large quantities of drugs.” The decision also let Isis put off for another day dealing with the regulatory agencies over systemic exposure issues.
Local delivery has become an antisense industry theme. For instance, privately held EpiGenesis Pharmaceuticals Inc. was founded in 1995 to develop therapeutics for lung diseases, building on the discovery that inhaled antisense compounds were effective in the lungs of animals. That finding was made by the Cranbury, NJ company's chairman and CEO Jonathan Nyce, while he was a professor in the University of North Carolina system.
“We like to think the inhaled route is essentially topical administration,” says Jim Mannion, EpiGenesis' president. Antisense drug delivery benefits from the lungs’ huge surface area, “approximately that of a tennis court,” and a naturally occurring surfactant that helps rapid distribution. The company’s drug programs are in asthma, chronic obstructive pulmonary disease, allergic rhinitis and lung cancer, markets altogether valued by Mannion’s estimate at 20 billion dollars.
EpiGenesis' flagship product, EPI-2010, is designed to control overexpression in the lung of the A1 adenosine receptor, a key event in initiating asthma. Mannion hopes EPI-2010’s week-long duration of action will do something even greater than relieving symptoms -- it could prevent symptoms before they occur. He’s optimistic this will especially benefit kids, where weekly rituals -- “mom giving the medicine on Sunday afternoon” -- would prevent childhood asthma.
Image courtesy AVI BioPharma Inc.
AVI BioPharma's plan is to use local delivery of an antisense drug to prevent arterial restenosis after balloon angioplasty. When a cardiologist uses a balloon catheter to enlarge the lumen of a narrowed artery, the expanding balloon damages the arterial wall and triggers a wound healing response -- rapid growth and division of cells in the damaged area. In up to 40 percent of cases, the healing fails to stop when repair is finished, and continued cell growth results in restenosis, in which the artery becomes as clogged as before, even closed completely.
AVI’s Resten-NG, based on its NeuGene technology, inhibits c-myc gene expression in the balloon-damaged area through catheter release of the drug at the same time and place as the angioplasty. Result: Wound healing stops in its tracks before really getting started; animal experiments show restenosis is prevented. A single dose suffices, lasting the 6 to 12 hours needed to prevent wound healing, according to the company’s president and COO, Alan Timmins. AVI believes follow-up doses won’t be needed. Millions of balloon angioplasties each year mean Resten-NG may have a very big market. Timmins expects Resten-NG’s Phase II trials will finish by the end of 2001.
Hossein and Kasra Ghanbari, father and son, are CEO and COO of 12-man, 1999 startup Panacea Pharmaceuticals Inc., of Rockville, MD. They too take advantage of catheter delivery, theirs inserted through the skull, circumventing the blood-brain barrier and allowing concentrated delivery of antisense drug PAN-346 to kill residual cancer cells left after brain tumor surgery.
PAN-346 inhibits aspartyl (asparaginyl)-B hydroxylase (AAH), which localizes in the invasive periphery of brain tumors and plays a role in cell motility and invasiveness; overexpressed, AAH inhibits proteins that trigger programmed cell death. The key to Panacea's approach is that complete AAH antisense inhibition isn’t required for therapeutic effect: “As low as 40 percent inhibition would change the cells back to normal,” says Hossein Ghanbari. So great is the need for improved therapy -- glioblastoma multiforme and astrocytoma are among the deadliest cancers -- that Panacea is in discussions with several foundations (including the Canadian Brain Tumor Consortium) concerning financial support so Phase I trials can start as soon as possible, probably in 2001.
Monitoring AAH levels in spinal fluid will keep dosage on course, increasing PAN-346 when needed, weaning patients if therapy succeeds. The Ghanbaris hope a course of PAN-346 over a period of weeks would eliminate the tumor cells remaining after surgery. There’s no doubt about their optimism. “A yellow brick road for antisense,” Kasra Ghanbari calls it.
“For most cancers, it’s really hard to point to one protein and spend millions and millions to develop the right tool and do clinical testing that shows knocking it out has this incredible prolonging of survival of a human being.” Acknowledging that, Ray Warrell, CEO of Berkeley Heights, NJ-based Genta, is convinced that Genasense, an antisense inhibitor developed to knock out the Bcl-2 gene, is a precisely right, landmark tool in cancer therapy.
Genasense departs from the local delivery strategy by being administered systemically. Genta dedicated years of effort to showing systemic administration works with an antisense drug, and Warrell is proud to say his company was the first to succeed. But by targeting multiple diseases, Genasense exemplifies another strategy: Several antisense pipelines have at least one drug candidate intended for testing on multiple diseases. And the most popular arena for getting several tries at smashing the ball into the net is cancer.
“I don’t think it’s an understatement to say that the goodness of Bcl-2 as a target was not appreciated by us or anybody else at the time it was discovered,” Warrell says. But now it is known that a primary reason cancer cells refuse to die -- even after the horrible damage inflicted on them by chemotherapy and radiotherapy -- is because of overexpression of Bcl-2. High protein concentrations of Bcl-2 prevent apoptosis (programmed cell death), which normally seals the fate of miscreant and faltering cells.
Bcl-2 overexpression is widespread among cancers. Used in combination with standard chemotherapeutics, Genasense is now in Phase III trials for advanced melanoma. In 2001 trials will begin for acute myelocytic leukemia, chronic lymphocytic leukemia and multiple myeloma. Farther ahead may come Genasense trials for prostate and breast cancer.
Other companies are striking diverse cancer targets with drugs that, like Genasense, are largely add-on drugs, meant for use in conjunction with other drugs or therapies. Numerous cancers overexpress AAH, the target gene of Panacea’s PAN-346. The Ghanbaris hope eventually to apply it to breast, colorectal and other cancers. GEM231, Hybridon’s most advanced cancer compound, targets another overexpressed cancer gene, a regulatory subunit of protein kinase A that regulates cell proliferation in most human cancers; its Phase II trials in combination with Taxol and Taxotere have started. Hybridon’s preclinical pipeline contains a multi-cancer inhibitor of the nuclear excision repair protein, an enzyme induced for DNA repair in response to the chemotherapy drug cisplatin.
AVI BioPharma’s CYP3A4 inhibitor project envisions an altogether different set of applications for an antisense drug, affecting many diseases for which drug treatments already exist. CYP3A4 is one of a set of liver enzymes that metabolize drugs. Its substrates include caffeine, Viagra, nicotine -- in fact, Timmins says, at least half of all FDA-approved drugs. Antisense inhibition of CYP3A4 would slow metabolism, allowing its drug substrates to last longer in the body. Many medicines might then accomplish therapeutic effects in smaller, potentially safer doses. Timmins even sees a CYP3A4 inhibitor saving drugs that otherwise would wash out of clinical trials because of toxicity. Slowing CYP3A4 would likewise slow elimination of many compounds ingested from food and drink, an issue the company will need to address carefully. But for toxic yet potentially lifesaving compounds, CYP3A4 inhibition might well be worth dietary restrictions. The project is preclinical.
Antisense company researchers have increasingly been able to exploit the technology's advantage of speed in discovering drug targets and moving drugs into the clinic. As the largest antisense company, with 11 drugs in the pipeline, Isis has been in best position to use the technology to discover drug targets. Bennett explains: “The pharmaceutical companies right now are trying to sift through 100,000 genes to identify the ones that make the best targets for drug discovery efforts. We can take a sequence and within a week have an inhibitor for testing in a cell culture model.”
After finding an oligo with an interesting in vitro phenotype, Isis needs only a couple of weeks more to be ready to study its effects in animal models. No technology is faster for target validation, Bennett says: Antisense is certainly far faster than making knockout mice, and useful with any species. By automating antisense inhibition assays, Isis has built a target validation service called Genetrove that has several pharmaceutical company customers.
Isis uses the same screening to fill its own pipeline. “Historically,” says Bennett, “we would look at a pathway and try to identify which gene would make the best drug target. Now, we don’t have to make a decision. We make antisense inhibitors for every pathway gene and find out which one gives us the most robust effect.” Isis applied antisense screening to 60 targets related to diabetes and found three worth pursuing for drug development. Already the company has inactivated 2,000 genes and expects in three years to have done 10,000.
And, of course, the same oligo that takes only a week to make and a few more to validate as a target might also be a good drug. “Development of antisense is much quicker than development of small molecule drugs,” Bennett says. “We’re in a much better position to capitalize on these discoveries.”
“Small molecules take too long,” seconds Hybridon's Agrawal. “You have to express the protein and screen combinatorial libraries to come up with a molecule. But you don’t know how to manufacture it, how it inhibits the protein, or know the pharmacokinetics.” Here nucleic acids have strong advantages: “You know how to manufacture, you know how to do quality control, what the pharmacokinetics will be -- because it doesn’t change with sequence -- and you know what the safety profile will be. You don’t need to formulate them differently.” And he adds that unlike small molecules, where different manufacturing processes mean different facilities, antisense drugs can be produced in the same plant, since the machinery stays the same. “That makes it very easy for the drug development team.”
Antisense drug developers have also overcome the drawback of administration by injection. Both Hybridon and Isis have shown that antisense compounds can be given in tablets. Isis has set its sights on an oral solid formulation of ISIS 104838, an antisense treatment for rheumatoid arthritis and Crohn's disease. Its target: TNF-alpha, turf of several other drugs, including Immunex Corp.’s Enbrel, with $576 million in sales in the last 12 months.
If antisense inhibition of TNF-alpha bears out Bennett’s prediction of a better side effect profile than Enbrel’s, an oral formulation could give Immunex something to worry about -- Enbrel works by protein antagonism, and requires injection for effect.
Formulation improvements are being complemented by next-generation antisense chemistries, improving upon phosphorothioates in particular, for enhanced therapeutics and economics.
Genta, for instance, has taken the tremendously expensive step of using chirally pure materials in its next-generation compounds. Warrell says the cost is made worthwhile by greater affinity for target mRNA and vastly improved half-life (2-4 hours for phosphorothioates, in excess of 300 hours for the new compounds). This means lower doses, benefiting Genta with lower manufacturing costs and patients with improved safety.
AVI BioPharma’s Resten-NG for balloon angioplasty uses the company’s NeuGene antisense chemistry. DNA and phosphorothioate oligos have charged backbones; Resten-NG’s backbone is neutral. Timmins explains that a neutral backbone makes the drug less likely to stick to unintended targets by charge interactions, and more selective for its true target. Resten-NG binds at the mRNA AUG start site for translation, an invariant design feature of NeuGene-based drugs, and has its inhibitory effect entirely by blocking translation. (Phosphorothioates are not always designed to bind at the AUG start, and can both physically block translation and induce destruction of mRNA through an enzyme called RNase H.)
Farther afield in antisense chemistry -- and in the passion it stirs -- is PNA (peptide nucleic acid), which Pantheco A/S will apply to develop antisense-based antibiotics. Anker Lundemose, CEO of the Copenhagen, Denmark firm, knows no better words to convey the aficionado’s admiration of PNA than this quote from Michael Egholm, one of its inventors: “If God had been a chemist, he would have invented PNA, not DNA.”
Pantheco’s founding is a return home for PNA, which was invented at the University of Copenhagen in the early 90s. Isis jumped right on this invention, and licensed the exclusive rights for developing PNA-based drugs in 1992. Then, two years ago, Danish investors persuaded Isis to sublicense to Pantheco the right to develop PNA-antisense antibiotics. Last September, the two companies expanded their agreement to include treatments for diabetes and cardiovascular diseases.
The helical structure of PNA looks much like RNA and DNA. However, in PNA the bases are attached not to a phosphate-sugar backbone but to the backbone of an ordinary peptide. A PNA antisense oligo readily blocks mRNA translation. There is no RNAse H effect.
PNA/DNA duplex. Image courtesy Pantheco A/S
Maybe He just thought medicinal chemistry a bore, but the marvel The Maker missed is that modification of the PNA backbone imparts unusual new properties to a nucleic acid. “You can vary the PNA backbone with all sorts of modifications which will ensure an optimal drug profile,” says Lundemose. “That could be stability on the bench, half-life in animals, metabolism, volume of distribution, or targeting to an organ like the lungs. You have a very good chance of coming up with the right profile, simply because you can change the backbone chemistry.”
Ordinary PNA barely penetrates bacteria. But with appropriate backbone modifications, antisense PNA goes right in and suppresses its target. Lundemose won’t disclose the targets of Pantheco's PNA-based antibiotics, but hopes the first compound will start clinical trials by 2002.
Arousing less rapture, but still remarkable, are ribozymes: Might they be able to accomplish the same therapeutic ends as antisense oligos, but with milder side effects? Alene Holzman, vice president of corporate development at Ribozyme Pharmaceuticals Inc. (RPI), located in Boulder CO, believes the ribozymes in her company’s pipeline will do just that.
RPI began in 1992 with a mission to develop Tom Cech's Nobel Prize winning ribozyme discoveries into the basis for a new class of therapeutics. A ribozyme is made of RNA, approximately 35 nucleotides long, and can be designed to bind a target mRNA just like a DNA-based oligo. Upon binding, enzymatic ribozyme activity cleaves the mRNA, destroying its capacity for translation. A ribozyme binds its target with slightly less affinity than an antisense oligo, but Holzman points to that as an advantage: It means a ribozyme is less likely to bind the wrong target. Nor are ribozymes associated with complement activation, as phosphorothioates are. Holzman predicts ribozymes’ safety profile will exceed those of many antisense oligos.
RPI's lead drug candidate, Angiozyme, is an anticancer ribozyme that works by halting the growth of tumor-associated blood vessels. Targeted against the VEGF (vascular endothelial growth factor-2) receptor, Angiozyme will be moving into Phase II trials in breast, lung, colorectal, melanoma and renal cancers.
Herzyme, another of RPI's anticancer ribozymes, will counteract overexpression of the Her-2 gene in some forms of breast and ovarian cancer. Herceptin, Genentech Inc.’s monoclonal antibody against the Her-2 protein for the same indication, had sales of $262 million in the past twelve months. RPI thinks Herzyme will have a better safety profile than Herceptin (Herceptin's side effects include cardiac toxicity and severe hypersensitivity reactions).
No one talked about, much less though about, genomics when antisense got started. But with their target-hunting blitzkriegs and the uniform pharmaceutical properties of nucleic acids smoothing their paths toward clinical trials, the antisense companies could become some of the post-genomics era’s biggest beneficiaries. And the challenges to Enbrel and Herceptin serve warning that the old reliables already on pharmacy shelves won’t be immune from antisense attack.
“My own view of antisense is that it is where monoclonal antibodies were six or seven years ago,” says Genta's Warrell. “They were the magic silver bullets of the early 80s -- and then you couldn’t give them away. Antisense is exactly the same. Lots of promise, and then it’s down the drain. And right at this particular time you’re looking at a resurgence of interest.”
“The day will come when you have a gene that has been identified for a specific disease and you can design a drug within a week,” says Hybridon's Agrawal. And if he is right that there are some billion-dollar drugs in these pipelines? “You will see more and more groups shifting to antisense and the applications will expand. Just like what happened with monoclonals.”
So if the genomics companies are the young lions in the grab for genome gold, the antisense companies -- if they prove by deed their claim for speed -- may come to be enviously regarded as the cheetahs in the contest. | |
originally published 01/05/2001
| |  |