“The little guys are going to have to judgewhere their equities are best spent. There’sno question in my mind that it’s incrediblyimportant that the industry as a wholework with the FDA. If the outcome is goodfor bigger companies, it will be good for lit-tle companies.”

One development that may help smallerbiotech firms hoping to develop medicinesand diagnostics based on pharmacoge-nomic data is the potential creation of anInterdisciplinary PharmacogenomicReview Group that would evaluate data thatdoes not have a regulatory effect. The FDAwants access to research information so thatits reviewers can bring themselves up tospeed on the new technologies and theirimplications. “We need product developersand researchers to share their results withus so we can incorporate it into what we’redoing,” says McClellan. By familiarizingthemselves with such data as soon as possi-ble, FDA reviewers should be prepared toproperly evaluate similar information whenproducts based on pharmacogenomic datastart making their way through the clinic.

Woodcock says FDA considers the pro-posal a blueprint for other emerging tech-nologies. “It could also be applied toproteomics by extrapolation. Really, theseissues (of what do with data from emergingfields of research) apply to all of the tech-niques in the development and regulationof drugs.”

Jim Kling, Washington, DC, USA

N E W S

590 VOLUME 21 NUMBER 6 JUNE 2003 NATURE BIOTECHNOLOGY

US FDA contemplates collection of pharmacogenomic dataOn April 9, FDA’s science advisory commit-tee met to discuss a draft proposal forincorporating pharmacogenomic data intothe regulatory process. At the meeting,Janet Woodcock, director of FDA’s Centerfor Drug Evaluation & Research (CDER;Rockville, MD), presented a draft proposalthat met with general approval from indus-try representatives.

Pharmacogenomic data attempts todefine relationships between a patient’sgenomic profile and a response to a givendrug. Some firms hope to use such infor-mation to develop drugs that are tailored toa patient’s genetic profile, but most compa-nies currently focus on toxicology studiesthat examine a patient’s ability to metabo-lize the drug being tested. For example, ofthe top 27 drugs cited in reports of adversereactions, at least 59% are metabolized byan enzyme that has a poor-metabolizingvariant. While it is not conclusive that theseprotein variants are responsible for theadverse reactions, “[these data suggest that]polymorphism matters when it comes todrug safety,” says Lawrence Lesko, areviewer with CDER.

But the FDA recognizes that pharma-cogenomics is still in its infancy, and verylittle clinical significance can be read intomost studies. “It’s unclear how this infor-mation fits into the regulatory processother than to raise a potential red flag,”admits FDA commissioner MarkMcClellan, and that is what worries indus-try. “The amount of data that would be sub-mitted [for pharmacogenomic studies] is inthe tens of thousands of data points, whichis the potential subject for a great deal ofpost hoc analysis and data mining,” saysBrian Spear, director of pharmacogenomicsat Abbott Laboratories (Abbott Park, IL,USA). “If you look hard enough, you canprobably find some association with some-thing that would give you pause.”

At the April 9 meeting, Woodcock madeclear that there are no plans to requirepharmacogenomics testing, but when it isdone, the agency intends to require submis-sion of some data but not others. The pro-posal suggests that submitted data beseparated into two categories: data with aregulatory impact, and data with no regula-tory impact. Specifically, the agency wouldrequire access to data used in safety/efficacyevaluation, such as prescreening patientsfor a phase I clinical trial, or data that goesinto dosing calculations. Even animal

model pharmacogenomics might beincluded if, for example, it explains why atoxic response might be species-dependent.“In general, results intended to influencethe course of the clinical developmentprocess will be considered part of the safetyand efficacy evaluation,” says Woodcock.The agency does not plan to consider dataused solely for research purposes. The FDAhopes to have new guidelines in place in thenext 16 months and is currently planning aworkshop for the fall.

Industry representatives were happy withwhat they heard. “I think most of us believethat if this is data that safety decisions arebeing made around, then you have to pres-ent the data… I’m quite pleased with theproposal,” said Harold Davis, vice presidentof preclinical safety assessment at Amgen(Thousand Oaks, CA, USA).

While big pharma sees pharmacogenomicsas a toxicology exercise, other, smallerbiotechs like Millennium Pharmaceuticals(Cambridge, MA, USA) and GenaissancePharmaceuticals (New Haven, CT, USA) listtreatments for genetically defined popula-tions as a drug discovery priority. Still, few (ifany) representatives from small biotechsshowed up at the meeting, and none returnedcalls asking for comment.

Such disinterest could be because the dis-cussion is still in the early stages, saysGillian Woollett, vice president for scienceand regulatory affairs at the BiotechnologyIndustry Organization (Washington, DC).

The Indian agbiotech industry suffered mul-tiple blows at the hands of the government inlate April, calling into question whethergenetically modified (GM) crops have afuture in the country. Not only has the entryof additional GM crops into the Indian mar-ketplace been delayed, but the marketing ofMahyco Monsanto Biotech’s (MMB;Mumbai) approved GM cotton beyond 2004is also up in the air.

On April 26, a parliamentary committeecalled for an independent review, the resultsof which are expected by the end of August,of last year’s decision by the GeneticEngineering Approval Committee (GEAC;New Dehli) to give MMB a three-yearlicense to market three insect-resistant GM

cotton hybrids (containing the gene forBacillus thuringiensis toxin; Bt) in six centraland southern states of India (Nat.Biotechnol. 20, 415, 2002). A day earlier,GEAC refused to grant permission to MMBto sell its Bt cotton to farmers in northernIndia, citing sensitivity to curl leaf virusspread by white flies that are rampant inthat region. Also on April 25, GEAC calledfor more field trials and biosafety tests fromProAgro (Gurgaon) for its GM mustard,rejecting the firm’s commercial applicationfor a second time since 2001.

All of these setbacks have been influencedby conflicting reports over the performanceof Bt cotton in the country in 2002, the firstyear of sowing. MMB claims that Bt cotton

India dawdles over Bt-cotton

©20

03 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reb

iote

chn

olo

gy