Drug development: Holding out for reinforcements

Signs of emerging drug resistance are turning the hunt for new malaria treatments into a race against the clock.

The first signs of trouble were subtle. Several patients living near the Thai-Cambodia border had received first-line malaria treatment of artemisinin-combination therapy (ACT), but failed to clear the Plasmodium parasites from their blood. Their doctors were flummoxed. “There were these sentinel sites reporting reduced efficacy of ACTs,” recalls Arjen Dondorp, deputy director of the Mahidol-Oxford Tropical Research Unit in Bangkok. Dondorp and his colleagues suspected drug resistance. “But it was always uncertain, and could potentially be ascribed to the drugs not being taken properly or under-dosed.” In 2006, they started to investigate the likelihood of ACT resistance more thoroughly.

Cambodia is cracking down on counterfeit malaria drugs and has outlawed artemisinin monotherapies. Credit: ©WHO/SONNY KRISHNAN

The stakes are high. The emergence of drug resistance has already rendered once-effective malaria treatments — chloroquine and sulfadoxine-pyrimethamine — less reliable. Today, ACTs are the weapon of choice against malaria, and the possibility of losing them has the research community scrambling to understand the situation and to develop new drugs as reinforcements or even replacements.

Cracks in the armour

Artemisinin is a molecule extracted from the sweet wormwood plant Artemisia annua. In its natural form, it is degraded in a matter of hours within the body. Treatment requires a week-long course — a demanding regimen that can fail if patients do not complete it.

Clinical researchers overcame the limitations of natural artemisinin by developing more robust derivatives, such as artesunate or artemether, and coupling them with partner drugs such as mefloquine or lumefantrine. The World Health Organization (WHO) recommends five such combinations for distribution to endemic malaria regions. They have, in general, proven successful, ridding most patients of the immature blood-stage parasites after only 3 days.

Resistance has only recently become the biggest concern with ACT. For years, the primary issue was uncertainty with supply. Production depended on A. annua agriculture, and drug prices were prone to dramatic fluctuations. “The average price for 1 kilogram of high-quality artemisinin spiked to US$1,000 in 2005 due to a shortage, but dropped to around US$195 in 2007 due to overproduction,” says Tue Nguyen, vice president for research and preclinical development at OneWorld Health, a non-profit organization in South San Francisco, California. “The current price is around US$450/kg and increasing.” Nguyen's organization is leading efforts to counter this volatility, partnering with Amyris — a synthetic-biology company in Emeryville, California — and French pharmaceutical giant Sanofi to produce bulk quantities of artemisinic acid, an artemisinin precursor, using genetically-engineered yeast. In 2012, Nguyen anticipates making 9,000 kg of artemisinin — enough for millions of doses — at an initial price of US$400/kg; in 2013 they expect to quadruple production. Stabilizing supply of artemisinin should help reduce some of the volatility in the cost of ACTs and smooth the availability of drugs.

Nevertheless, a stable supply will be of little comfort to patients in Southeast Asia facing potential treatment failure. Several independent investigations by Dondorp and Harald Noedl of the Medical University of Vienna have confirmed the emergence of ACT resistance in the region¹, and the problem appears to be spreading. “There was a dramatic slowing of parasite clearance,” says Dondorp. “ACTs are starting to fail in western Cambodia, but what may be even worse is that this phenotype has also arrived in western Thailand, at the Thai–Myanmar border.” In some communities, a standard ACT course no longer stops malaria in up to half of the patient population. This resistance is specific to artemisinin, but it could eventually lead to resistance to partner drugs too.

Southeast Asia has incubated treatment-resistant malaria in the past (see The numbers Game, page S14). With both chloroquine and sulfadoxine-pyrimethamine, resistance spread to the Indian subcontinent and eventually Africa. With ACT, “So far, we have not seen any signs of resistance outside of Southeast Asia,” says Pascal Ringwald, coordinator of the Drug Resistance and Containment Unit within the WHO's Global Malaria Programme, which is closely monitoring reports of potential malaria resistance.

Unfortunately, the basis of this resistance remains unclear. “We have screened 185 different P. falciparum isolates — mostly from Southeast Asia — and they are all extremely sensitive to the artemisinins in vitro,” says Xinzhuan Su, chief of the Malaria Functional Genomics section at the National Institute of Allergy and Infectious Diseases in Bethesda, Maryland. Nevertheless, many of these strains show resistance to artemisinin in malaria patients. A study published in April 2012 identified a segment of the P. falciparum genome that appears to contribute to reduced parasite clearance in patients from Cambodia and Thailand, although further analysis will be needed to identify the relevant genetic variants².

David Fidock, a microbiologist at Columbia University, New York, suspects that the resistance seen in Plasmodium is only a transient survival mechanism, and that genetic mutations that confer robust artemisinin resistance without weakening the parasite's fitness are yet to emerge. “My personal sense is that it's very difficult for the parasite to gain stable resistance to artemisinin,” says Fidock. “Lab studies indicate that the initial gain of drug tolerance is readily lost by the parasite after the removal of drug pressure.” Furthermore, many scientists think that artemisinin works in a nonspecific manner: as Plasmodium feeds on haemo- globin, iron is released that activates artemisinin, causing extensive chemical damage to the parasite (see 'One parasite — many hiding places').“There's probably not a specific target on the parasite that the drug is attacking that could mutate to prevent inhibition,” says Fidock.

Even without a clear cause, the problem is undeniable, and the WHO has produced the Global Plan for Artemisinin Resistance Containment (GPARC), now being implemented across Southeast Asia. GPARC pairs surveillance with improved clinical practice: catching malaria early and ensuring complete elimination of Plasmodium. “We're focusing on the success of diagnostics and treatment,” says Ringwald, “and using very effective first-line drugs, mosquito control, lots of advocacy and research into new operational strategies to kill off reservoirs of the parasite.

New drugs would also help. One of the most promising is a compound known as OZ439, which differs structurally from the artemisinin drugs, but retains the endoperoxide chemical group that is crucial to their success. OZ439 outperforms artemisinin in an important regard. “It has a half-life that has never been seen for an artemisinin derivative,” says biochemist Sergio Wittlin of the Swiss Tropical and Public Health Institute (Swiss TPH), who helped discover the compound. “Natural products have half-lives in the range of 1 hour, but this has a half-life of 20 hours in orally dosed rats, which also held true in a phase I trial.” The drug is currently in phase II trials, and hopes are high that OZ439's prolonged existence in the bloodstream, paired with its novel structure, will make it effective where ACTs are failing.

Not so neglected

The past decade has seen a surge in investment in antimalarial drug discovery. According to a report by the Bill & Melinda Gates Foundation's Global Funding of Innovation for Neglected Diseases (G-FINDER) programme, 2010 global spending on malaria research and development (R&D) totalled US$547.2 million, with drug research accounting for the largest component (42%). Nearly a quarter of this R&D investment came from the private sector, an important contribution to what was previously considered a neglected tropical disease.

Alongside the Gates Foundation as a driver of this funding boom is the Medicines for Malaria Venture (MMV). With pharmaceutical and biotechnology partners, MMV coordinates distribution of public-sector and philanthropic funds to ensure that money can be rapidly allocated to scientists making progress in malaria drug development at any institution, anywhere in the world. “MMV is so knowledgeable and focused and professional in its efforts — it's like a virtual drug discovery organization,” says Nick Cammack, who heads medicinal development for GlaxoSmithKline (GSK) in Tres Cantos, Spain. The Tres Cantos campus is devoted to tropical-disease research and a nexus for public–private sector collaboration. “We fund research on a 50:50 basis, with GSK and MMV each contributing half, and we have joint objectives to find new molecules for clinical development,” he says.

As well as investment and expertise, drug companies engaged in the drive for anti- malarial research are also contributing libraries of chemicals — resources once kept under lock and key. Wittlin's Swiss TPH colleague Matthias Rottmann is part of a multi-institutional collaboration with the Novartis Institute for Tropical Diseases that has worked with one such library. “It was almost 2 million products, basically all the chemicals Novartis has,” says Rottmann. In 2010, GSK surprised many observers by making public the chemical structures and screening data of more than 13,000 molecules with apparent antimalarial activity, identified from its library of roughly 2 million compounds³. MMV has since worked with GSK, Novartis and St Jude Children's Research Hospital in Memphis, Tennessee, to compile a 'greatest hits' collection called the Malaria Box, a freely available sampler of 400 chemicals with demonstrated activity against P. falciparum. According to MMV's chief scientific officer Tim Wells, this curated collection is proving popular. “It became available in the last week of December 2011, and it has already 'sold out' its first batch,” he says.

Compounds that show promise in these screens are only starting points. “I saw one article that described our findings as '13,000 new malaria drugs', which is an exaggeration to say the least,” says GSK's Cammack. Each promising lead will need significant optimization work, he adds. Accordingly, although scientists are free to use the Malaria Box as they like, Wells anticipates that most will seek further assistance through MMV or private sector partners to shepherd promising molecules into clinical testing.

MMV is assisting in early stage development of dozens of potential drug candidates. Among the most promising is the synthetic molecule NITD609, one of several spiroindolones isolated in 2010 from a Novartis library⁴. “They were fairly potent in cellular assays right away, and they were attractive with regard to how long they stay in the bloodstream and their chemical tractability,” says cell biologist Elizabeth Winzeler at The Scripps Research Institute in La Jolla, California, and a corresponding author on the study. NITD609 is now poised for a phase II trial in Thailand. “It has a very good safety profile, and looks very promising,” says Rottmann, who has also worked on NITD609. Unfortunately, patient recruitment has stalled owing to events such as catastrophic flooding — problems that have also dogged trials of OZ439. “Now the malaria season is past its peak, researchers at the trial sites will have to wait until next season to finish the phase II trial,” said Wittlin, in January 2012. The onset of the rains in June should enable these trials to get back on track.

Seeking the ultimate weapon

Although the early data for both OZ439 and NITD609 are encouraging, their efficacy has been demonstrated only against blood-stage parasites. Yet Plasmodium has a complex, multi-stage lifecycle — especially P. vivax, which remains an especially tricky target (see 'Danger below the radar'). Any therapeutic strategy intended to eradicate malaria must wipe out gametocytes — the stage of the parasite life- cycle that moves from humans to mosquitoes. The ultimate goal for researchers is to design treatments that render a person inhospitable to parasitic infection after one dose: a strategy termed 'single-encounter radical cure and prophylaxis'. Efforts are underway to pin point vulnerabilities in non-blood-stage parasites, and although no new drugs have unambiguously passed this test, Fidock's team has obtained surprising data on a very old drug — methylene blue⁵. “This was the first synthetic compound ever used in humans — in 1891 — and it has very potent activity against both early and late-stage gametocytes, including the ability to block transmission to mosquitoes,” says Fidock. Methylene blue fell into disuse because of its disconcerting tendency to turn urine green and the whites of the eyes blue, but these might be acceptable side effects should the compound prove valuable for parasite eradication.boxed-text

The motivation and resources are clearly present for antimalarial drug discovery and development, but time is a looming problem. “When you talk about next-generation drugs, launch dates are projected for 2018 or 2019,” says MMV's Wells, “but people are worried about what's going to happen next year.” For the time being, surveillance and containment strategies are the best hope. “It all comes back to whether emerging ACT resistance might soon leave us with essentially no effective treatments for a period of some years,” says Fidock, “and that's an unresolved question right now.”

Box 1: Danger below the radar: Targeting the 'other' Plasmodium species

When people talk about malaria, they're generally talking about Plasmodium falciparum, the predominant species in Africa and the cause of hundreds of thousands of deaths each year. Plasmodium vivax, more of a problem in Asia and South America, tends to get overlooked, largely because of its reputation for being less lethal.

Epidemiologist J. Kevin Baird, of the Eijkman-Oxford Clinical Research Unit in Jakarta, Indonesia, sees this as a serious mistake. “You can still read in current medical textbooks that P. vivax very rarely kills,” says Baird. “But that's just not so, and there's very good evidence that it kills patients in the same variety of ways as P. falciparum.” He cites recent unpublished data from his team indicating that P. vivax was the sole parasite found in 28% of malaria deaths in one Indonesian hospital.

Plasmodium vivax's mild reputation is only part of the problem — it's also hard to study. It infects only blood-cell precursors known as reticulocytes, which are difficult to obtain and vanish quickly in culture. “Vivax can only be cultured on a short-term basis,” says Matthias Rottmann, a biochemist at the Swiss Tropical and Public Health Institute. “A reticulocyte ages very fast: after a few days it's not a reticulocyte any more — it's an erythrocyte.” This means drug studies must be performed at sites where P. vivax is readily available, and makes genetic manipulation of the parasite virtually impossible.

Furthermore, it has a nasty tendency to hide. P. vivax enters a dormant stage — a hypnozoite — in the liver, lingering for months before re-awakening. “We know nothing about hypnozoite metabolism,” says Elizabeth Winzeler, a cell biologist at The Scripps Research Institute in La Jolla, California. “We don't know what targets would be useful, and we have no way of validating them.

Antimalarial drugs that kill blood parasites leave hypnozoites unscathed. A drug called primaquine eliminates P. vivax hypnozoites, but can be profoundly toxic for people with a mutated form of the G6PD gene. In a cruel irony, such mutations are widespread in malarial regions — and detecting them requires clinical resources that can be scarce in the developing world. “Even here in Jakarta, I would have a hard time finding somebody who could do a G6PD test,” says Baird. A related drug, tafenoquine, now in phase II trials, allows a sharply reduced dosing regimen, lowering the risk of harm for G6PD-deficient patients. Nevertheless, toxicity may still prove an issue, in which case effective use would require improved G6PD diagnostics or better-designed dosing strategies.

Lack of access to good experimental models is another limiting factor. Winzeler's team developed an assay to screen compounds against liver cells infected with a rodent-specific strain of malarial parasite⁶. This yielded at least one lead, although it has not yet been tested against human hypnozoites. Otherwise, the best models available are rhesus macaques — far from ideal, as monkey models are logistically complicated, expensive and ill-suited to large-scale drug discovery.

Fortunately, this once-neglected problem is getting a lot more attention. “The Gates Foundation put quite a lot of money in two years ago, and there are five or six programmes looking at how to improve cell culture,” says Tim Wells, chief scientific officer at Medicines for Malaria Venture. But Baird believes that priority should go to bolstering research efforts in regions where the crisis is most acute. “As difficult and expensive as it might be,” he says, “we have to invest in research sites near where malaria patients live.” ME