598 VOLUME 26 NUMBER 6 JUNE 2008 NATURE BIOTECHNOLOGY

FDA scrutinizes human stem cell therapies

With several companies and dozens of aca-demic researchers pushing to turn stem cells into therapeutics, officials at the US Food and Drug Administration (FDA) are striv-ing to stay a step ahead. Yet judging from an advisory meeting held in April, anticipating the regulatory challenges associated with cell therapies will entail a delicate balancing act. On the one hand, such cell therapies may offer new therapeutic avenues for intractable diseases and injuries; on the other, they pose difficult-to-appraise risks, such as cancer and immunogenicity.

Along with those ‘deal-killer’ risks loom other, more pragmatic questions about harnessing and delivering human embry-onic stem (hES) cells for specific diseases and anatomic sites being treated; about lot production, purity and other quality con-trol measures; and about appropriate ani-mal models, clinical trial designs and other safeguards in testing and using such cells. Although committee members briefly con-sidered requiring that suicide genes be incor-porated into hES cells as a failsafe measure for protecting patients, they concluded that

such schemes seem too technically compli-cated to enable implementation and might “cause more problems than [they] would solve.” Even so, there was wide agreement that hES recipients should be followed on a long-term basis for signs of cancer or other “unanticipated events.” Those conducting clinical trials with hES cells need to be vigi-lant for signs and symptoms that might arise, however unexpectedly, from the procedure. No one wants the cell therapy field to repeat the setback that gene therapy experienced following the death of Jesse Gelsinger, who died after receiving a single injection of an adenoviral vector. Indeed, only a month after the advisory panel emphasized the need for strict safety measures, on May 14, Geron, of Menlo Park, California received notice from the FDA that approval for its application to begin a clinical trial using hES cells for spi-nal cord injuries had been delayed. Thomas Okarma, Geron’s president and chief execu-tive officer, was “disappointed with the news,” particularly as the company has been in discussions with the FDA for the past four years. He had hoped that “the breadth

and depth of the submission, some 21,000 pages, predicated on those discussions with the agency” would be enough for to the trial to go ahead.

Although no specific product has entered the clinic, Geron is not the only company developing cell-based products headed for human trials. Front-runners include prod-ucts aimed at treating a rare retinal-pigment deficiency that leads to blindness, insulin-deficient diabetes and Parkinson’s disease and other degenerative conditions involving the central nervous system (CNS). San Diego, California–based Novocell is looking to direct hES cells into glucose-responsive insu-lin-secreting cells aimed at treating diabetes; Mytogen, of Charlestown, Massachusetts, a wholly owned subsidiary of Advanced Cell Technologies, is focused on developing reti-nal pigment epithelial cells from hES cells to treat patients with macular degeneration.

Except perhaps in the case of insulin-pro-ducing cells (dysfunctional pancreatic beta cells are known to underlie diabetes), a major challenge that developers of hES candidate products face is in defining their products. Will it be a single cell type at a particular differentiated state or some mix of cells at different states of development? Further complicating this question, many features of such cells are likely to be in flux as they are grown, stored, shipped, delivered and, ultimately, introduced into patients—with some of those changes plausibly affecting safety and efficacy in clinical use. Among many issues, the cells’ chromosomal makeup—karyotype—is high on the list of concerns. In culture, many ES cells contain normal chromosomal sets and structures, but a smaller proportion have either miss-ing or extra chromosomes, or abnormali-ties in chromosomal size and shape, a state called aneuploidy. The frequency and type of aneuploidy vary widely, and when cells are expanded to boost their numbers, there are opportunities for these abnormalities to develop. Because aneuploidy is often associ-ated with cancer, there are questions about threshold percentages of aneuploid cells that might be acceptable.

Retinal pigment cells provide a concrete example. Jonathan Dinsmore, senior vice president and general manager for Advanced Cell Technology and Mytogen, says that 95–99% of his company’s hES cell–derived retinal cells are “terminally differentiated,” as indicated by the expression of some 15 sepa-rate markers. But because treating one eye

Vigilance for chromosomal abnormalities will be high as human embryonic stem cells move into the clinic.

Stev

e G

schm

eiss

ner /

Pho

to R

esea

rche

rs, I

nc

NEWS©

2008

Nat

ure

Pub

lishi

ng G

roup

ht

tp://

ww

w.n

atur

e.co

m/n

atur

ebio

tech

nolo

gy

NATURE BIOTECHNOLOGY VOLUME 26 NUMBER 6 JUNE 2008 599

will probably take 1.5 million cells, there will always be numerous cells that are not of the cell type intended to deliver to the patient.

The pluripotency of ES cells is not only their virtue, but also their vice. As Kenneth Chien of Massachusetts General Hospital and Harvard Medical School in Charlestown, Massachusetts, who served a consultant to the FDA Cellular, Tissue and Gene Therapies Advisory Committee when it met last April, admits, hES cells are “probably the most complex human therapeutic imaginable.” ES cells can produce teratomas, an accumula-tion of many different cell types resulting in a benign form of tumor. But teratomas can grow and fill confined anatomical spaces, such as the central nervous system and spinal cord, which could prove disastrous. In addition, teratomas have some tendency to lose their differ-entiated status to develop into frankly malignant teratocarcinomas.

Developing a standard screening for tera-tomas, or teratocarcinomas, is a “key barrier” to overcome, says Chien. Committee mem-ber Savio Woo of the Mount Sinai School of Medicine in New York agrees that every com-mercial sponsor will need to demonstrate to the FDA what number of cells is safe for each such product. Rigorous criteria will be needed “before jumping into patients.” Many committee members concurred that it would be unacceptable if cancers derived from hES cells developed in patients involved in pio-neering clinical trials. Another unanticipated outcome could result from hES cells entering a site and either not doing what they are sup-posed to or increasing susceptibility to the very disease they are intended to treat.

Yet, Ole Isaacson of McLean Hospital of Harvard Medical School, in Belmont, Massachusetts, believes this complexity might work in the opposite way, proving beneficial to patients. Because hES cells are subject to “feedback control,” they are unlike “conven-tional pharmacologic agents” and thus may behave properly when situated in a particular anatomic site, he suggests. To make matters even more complicated, another committee member, Doris Taylor of the University of Minnesota in Minneapolis, felt that acces-sory cells within the hES donor mix—which are distinguishable from the ‘main’ cells and somehow aid them—could prove critical to the overall potency of hES cell batches, adding to the difficulty of defining product purity.

Indeed, there was wide agreement among the FDA committee that setting cell purity standards too high could also “backfire.” Within a batch of hES cells administered to a patient, there may be a minority of accessory cells that may be necessary for poorly under-stood cell-cell signaling—possibly to provide feedback, for instance to insulin-produc-ing cells to either boost or shut down their activity. Nonetheless, defining the degree of tolerable differentiation for cells, accept-able thresholds for their “heterogeneity” (in regard to their states of development), and other criteria for releasing cell lots will be needed. Requirements will doubtless be tai-lored to reflect the specific hES cells prod-ucts for specific clinical applications and

particular kinds of patients.

Such uncertainties need to be addressed with extensive stud-ies in animals before hES cells can ven-

ture into the clinical arena. Generally, the aim is to “mimic the human setting as well as possible,” says Jane Lebkowski, senior vice president of regenerative medicine at Geron, referring in this case to hES cells being developed to treat spinal-cord injuries that are being tested in rodents. Yet, even here, complications set in, according to Melissa Carpenter, former vice president of R&D at Novocell. Such testing of hES cells in rodents requires giving the animals immunosup-pressive agents so that they can tolerate hES xenografts, or using ‘nude’ mice with geneti-cally impaired immune systems that accept transplanted cells—deviating from the anticipated clinical protocols, which would involve transfers of cells between individu-als of the same species. One exception is the hES retinal pigment cells being developed to treat macular degeneration, because the eye is “privileged,” meaning that such cells are not subject to rejection, according to ACT’s Dinsmore.

Dosing issues might be best addressed in nonhuman primates, or other species big-ger than mice and rats, to gain a better idea of how many cells to use. But because of immune-suppression complexities and the spiraling costs involved in moving away from mice into larger animals, the panel steered away from insisting on this option. Despite the immunogenicity caveats, safety and effi-cacy are needed for the lead-up to any clini-cal trial, and hES cells are no exception.

Jeffrey L Fox, Washington DC

hES cells are “probably the most complex human therapeutic imaginable.”

Canada charts biologics pathHealth Canada has posted a draft guidance for the approval of biogenerics, which it has termed subsequent entry biologics (SEBs). According to the government agency’s guidelines, manufacturers can include a slimmed-down clinical package in their submission as long as they can show similarity between their product and a previously approved biologic, even one not approved in Canada. Manufacturers can rely in part on publicly available information, but must show—through side-by-side comparisons—that the quality, safety and efficacy of the SEB is comparable to those of the original biologic. Andrew Storey, vice president of Cangene, a biopharmaceutical company based in Winnipeg, Manitoba, welcomed the guidance but has reservations. “You want doctors to be able to prescribe it like a generic drug, but [Health Canada] is suggesting that additional studies will be required. That’s not the case for generic drugs and that shouldn’t be the case for [subsequent entry] biologics.” Health Canada will convene a stakeholder consultation in early June. In the US, California has recently introduced a bill that paves the way for biogenerics legislation. ‘The Pathway for Biosimilars Act’ proposes a minimum of 12 years’ exclusivity, plus another 2 years for a medically significant innovation. The European Agency for the Evaluation of Medicinal Agents has already implemented a regulatory pathway for ‘biosimilars’ and is reviewing a raft of new products. —Hannah Hoag

FDA gets personalNew guidelines published by the US Food and Drug Administration (FDA) bring the prospect of personalized medicine a giant step closer. The document, released in April, aims to ensure that consistent definitions of key terms in pharmacogenomics are applied across the US, Europe and Japan, the three regions covered by the International Conference on Harmonization, a project gathering together regulatory authorities and pharma experts. “If we don’t have a commonly agreed upon understanding of the terms that we are using, we cannot harmonize on issues for which these terms are critical,” says Felix Frueh of the Center for Drug Evaluation and Research at the FDA, in Silver Spring, Maryland, who helped compile the report. “Adherence to this terminology will facilitate and streamline drug applications and submissions,” he says. The document defines a ‘genomic biomarker’ and the difference between ‘pharmacogenomics’ and ‘pharmacogenetics’. It also aims to unify the way that biological samples and their associated genomic data are coded. The guidelines came just as ParagonDX of Morrisville, North Carolina, received FDA clearance for its warfarin-sensitivity kit. At the same time, the National Institutes of Health has put out a call for researchers’ input into current needs in pharmacogenomics research. The initiative’s objective is to “highlight opportunities, reveal gaps, and aid in identifying specific, achievable goals that will advance the field.” —Henry Nicholls

IN brief

NEWS©

2008

Nat

ure

Pub

lishi

ng G

roup

ht

tp://

ww

w.n

atur

e.co

m/n

atur

ebio

tech

nolo

gy