The liver safety profile of Novartis' Prexige appears to be one factor behind FDA's request for additional studies on the COX-2 inhibitor prior to approval in osteoarthritis and acute pain.

FDA issued a "not approvable" letter for Prexige (lumiracoxib) Sept. 22, requesting final data from Novartis' long-term TARGET study; additional data supporting the dose and dose interval used in acute pain models; and an additional hip osteoarthritis study.

Novartis described FDA's requests as an attempt to better characterize the overall risk-benefit ratio for Prexige.

"When you look at the response we got from the FDA, it's not addressing a specific liver enzyme elevation problem, it's an overall risk-benefit efficacy/safety assessment," Global Pharma Development Head Joerg Reinhardt, PhD, said on a Sept. 24 analysts call.

The potential for liver toxicity - especially in the Prexige 400 mg dose - is addressed in labeling approved in the U.K. on Sept. 15.

"Elevations of alanine aminotransferase (ALT) and/or aspartate aminotransferase (AST), greater than three times the upper limit of normal have been reported in placebo/active-controlled clinical studies in approximately 1.2% of patients in clinical trials up to one year with lumiracoxib 100 mg and 200 mg daily."

Marked elevations greater than eight times the upper limit of normal "have been observed in 0.0% with 100 mg once daily or twice daily and 0.19% of patients at 200 mg once daily," ¹ U.K. labeling says. "Chronic use with 400 mg daily dose was associated with more frequent and more marked elevations in ALT/AST."

U.K. labeling also notes that "rare cases of hepatitis have been reported" in the ongoing longer-term TARGET study among patients administered lumiracoxib 400 mg once daily. Patients who developed hepatitis were unblinded from the study.

"The basis for this statement in the label was an incidence of 0.07% of jaundice with Prexige in the TARGET study versus 0.05% of jaundice with ibuprofen and 0.02% of jaundice with naproxen," Reinhardt told analysts. "All of these cases were reversed and normalized."

Novartis noted that labeling for the three marketed COX-2 inhibitors - Merck's Vioxx (rofecoxib) and Pfizer's Celebrex (celecoxib) and Bextra (valdecoxib) - contains a class precaution on hepatic adverse events.

Elevations in ALT and AST (less than 3x ULN) have been reported in approximately 1% of patients in clinical trials with NSAIDs, COX-2 inhibitor labeling says. "Rare cases of severe hepatic reactions," including jaundice and fatal fulminant hepatitis, have been reported with use of Vioxx and Celebrex.

With the additional data requests, Prexige approval could be delayed until the 2005-2006 timeframe.

FDA's user fee deadline was Sept. 19, but the two-day government shutdown caused by Hurricane Isabel pushed the user fee deadline to Sept. 23.

"We have a number of studies ongoing that we believe may address some of the concerns that the FDA has raised," Reinhardt said. "We're going to have to discuss with the FDA what their requirements in terms of these individual trials are in the next two to four weeks."

How soon Novartis can respond to FDA's "not approvable" letter depends on whether the agency requires completion of the hip osteoarthritis study prior to approval. TARGET could be submitted to FDA in the third quarter of 2004, but an additional OA trial would not be submitted until the second half of 2005.

If Novartis has to conduct an "osteoarthritis trial in the hip indication, it will take not only the three months of treatment, but also we would probably have to provide 12 months extension data of that trial, which does determine then the overall timeline," Reinhart said.

"And in that case, as we said, we believe that we will not be able to submit the full data package before the second half 2005," he continued. "If that is not needed, we believe we can submit earlier."

Novartis expressed some optimism that existing data from studies in hand and knee osteoarthritis (200 mg) and osteoarthritis of the hip (400 mg) could address FDA's concerns. Given the available data in OA, Novartis does not "fully understand" the agency's request, Reinhardt told analysts.

"We will discuss with FDA in the coming weeks exactly what their requirements are, and then we will be able to give you an exact timing as to when it will be possible to submit the full data package," he said. "However, we believe that it will not be possible to launch Prexige in the U.S. before 2005."

Novartis also believes it has a "pretty comprehensive package" of acute pain studies that could help address FDA's data request, Reinhardt added. Three additional studies - for use in tension headaches, dental pain and orthopedic surgery - will be completed soon, he noted.

"We have several aspects that we will have to discuss with the FDA. One is how many models we actually need to get a pretty broad claim in acute pain, and within these models, what exactly do you have to show in terms of onset, in terms of duration, in terms of re-dose, in terms of duration to rescue medication?"

An acute pain approval could be complicated by the increased evidence of liver toxicity at the higher doses; Novartis is seeking a 400 mg dose for the indication. The company is proposing a 200 mg dose for osteoarthritis.

Prexige is the third "second-generation" COX-2 inhibitor to suffer an approval delay, Pfizer received a "not approvable" letter in July 2001 for parecoxib ( Dynastat in Europe) and plans to resubmit the NDA for acute pain in 2004.

Likewise, Merck withdrew the NDA for Vioxx follow-on Arcoxia (etoricoxib) in March 2002 and plans to refile in the second half of the year with longer-term cardiovascular safety data (² 'The Pink Sheet' March 18, 2002, In Brief).

Assuming a first-cycle review, Arcoxia could now be on the market at least a year ahead of Prexige.

Market share for the three available COX-2 inhibitors has stabilized. During the Banc of America conference Sept. 17, Merck CEO Ray Gilmartin characterized the market as being in "equilibrium," both in terms of coxib share versus traditional NSAIDs, and the COX-2s' shares versus one another.

Completion of the TARGET study is on schedule; Novartis submitted an interim analysis of the primary endpoint (GI perforation, obstruction and bleeds) to FDA in August. "The last patient is scheduled to complete in December of this year, and with the last patient follow-up in January, the full report will be available in July," Reinhardt said.

Novartis began the 18,000-patient Therapeutic COX-189 Arthritis Research & GI Event Trial in December 2001 (³ Also see "Novartis COX-189 TARGET Study Will Assess CV Safety Of COX-2 Inhibitor" - Pink Sheet, 12 November, 2001., p. 32). Patients received either 400 mg once-daily, ibuprofen 800 mg t.i.g. or naproxen 500 mg twice daily for 12 months.

If the results are positive, TARGET could shift attention away from Prexige's liver safety issues by offering definitive evidence of a cardiovascular and gastrointestinal advantage over non-selective NSAIDs - something that so far has eluded Merck and Pfizer.

While Merck's long-term VIGOR study demonstrated a GI benefit for Vioxx, it also showed an increase in thrombotic events. In contrast, Pfizer's CLASS study verified Celebrex' CV safety but did not confirm a GI benefit (⁴ [A#00630070000], p. 3).

Both Merck and Pfizer are conducting additional long-term studies. Pfizer's MATRICS studies will further investigate Celebrex' gastrointestinal safety profile (⁵ [A#00650250007], p. 10). Merck is conducting cardiovascular safety studies with Vioxx and Arcoxia (⁶ Also see "Merck Arcoxia 6,000-Patient Trial Will Study GI, Cardiac Events" - Pink Sheet, 17 June, 2002., p. 15).

Prexige's CV and GI safety profiles appear favorable in U.K. labeling. "In clinical trials with lumiracoxib at daily doses of 200 mg and 400 mg, the relative risk (95% CI) compared to placebo of myocardial infarction, cardiovascular death or stroke were 1.04 (0.36, 3.05) for the overall population (lumiracoxib n=5,431) and 1.19 (0.27, 5.29) for patients at high CV risk (lumiracoxib n=2,627)," labeling says.

"These were reviewed carefully by the U.K....and the feeling here was that the numbers 1.04 and 1.19 were not different. And they did come to the conclusion that there was no increased risk for Prexige in the [high-risk] subpopulation," Novartis Clinical R&D Head James Shannon told analysts.

The incidence of gastrointestinal perforation, ulcers and bleeds in patients treated with lumiracoxib 200 mg (1.6/100 patient years) was eight times lower than in patients treated with NSAIDs diclofenac, naproxen and ibuprofen (13.7/100 patient-years), U.K. labeling says.

"A relatively higher incidence of all the GI-related events (PUBs, decrease in hemoglobin and hematocrit, symptomatic ulcers) was observed with 400 mg compared to 200 mg especially over long-term treatment. However the frequency of these events with 400 mg lumiracoxib is still lower" than NSAIDs.

Novartis Prexige Is “Not Approvable;” COX-2 Launch At Least Two Years Away

More from Archive

More from Pink Sheet