An FDA advisory panel voted overwhelmingly — 19-1 — that a proposed Risk Evaluation and Mitigation Strategy (REMS) program would not ensure a favorable benefit-risk balance for tanezumab, a nerve growth factor inhibitor that drugmaker Pfizer hopes to market for osteoarthritis (OA) pain.
It’s yet another setback in a long and rocky road for the agent, for which Pfizer has struggled to show definitely that it is both safe and effective, with the former proving most problematic.
On Wednesday and Thursday, the agent underwent review in a joint meeting of the FDA’s Arthritis Advisory Committee and the Drug Safety and Risk Management Advisory Committee. If approved, the drug would be given by subcutaneous injection every 8 weeks.
Not a Game-Changer
“I’m sorry to say I voted no. I was hoping that we would see a drug that had a greater benefit, where the benefit would outweigh the risks and unfortunately I don’t think that’s the case,” said panel member Edward V. Cheng, MD, of the University of Minnesota Medical School in Minneapolis.
Evidence from tanezumab’s clinical trials indicated reductions in pain, but a few patients receiving the drug showed unexpectedly rapid OA progression — sometimes affecting previously non-arthritic joints. In a recent study comparing tanezumab with nonsteroidal anti-inflammatory drugs (NSAIDs), there were 33 reported cases of joint damage in presumably unaffected joints, with 31 being in patients on tanezumab and only two in patients on an NSAID.
Pfizer’s proposed REMS included certification of the healthcare setting and pharmacy, patient education, and x-rays of the knees and hips at baseline and yearly thereafter. The company also proposed that treatment should be stopped after two doses if patients don’t respond.
Committee members offered various reasons for their near-unanimous negative vote (the sole “yes” vote was given by a patient advocate). Because OA is a chronic condition, the assumption was that treatment would be required for lengthy periods, and Pfizer’s available follow-up data were only for one year. What might occur with longer term use is not known.
There also has been little information on which patients might be most at risk for rapidly progressing OA, other than that the risks were elevated two- to threefold higher among patients who also were taking NSAIDs. Moreover, with real-world use, unlike the more strictly controlled setting of a clinical trial, there could be unforeseen risks, such as off-label use of the drug.
Understanding of the biology of tanezumab also is limited. “We don’t know enough about this molecule,” said David Pisetsky, MD, PhD, of Duke University Medical Center in Durham, N.C.
A hypothesis offered by Pfizer for the progressive OA risk was that the reduced pain may have led patients to increase their activity, putting new mechanical loading on the joints, but panel members said this needed proof.
In addition, the trial data showed only modest pain reduction that weren’t superior to NSAIDs in the head-to-head trial.
“I concur with the sponsor’s contention that there’s an unmet need for efficacious treatments for osteoarthritis, and I really hear the plea from patients with chronic arthritis pain looking for relief. Unfortunately, this drug tanezumab does not fulfill this need and it has similar clinical efficacy and is clinically comparable to existing therapies. It’s really no better than taking aspirin or ibuprofen and does not prevent or delay total joint replacement,” Cheng said.
“This is not a game changer for OA,” he added.
Tanezumab is an immunoglobulin G type 2 monoclonal antibody that binds to nerve growth factor, which is upregulated in states of inflammation and injury. It represents a novel approach to pain control, with a mechanism of action different from NSAIDs and opioids. It also does not appear to have the cardiovascular and gastrointestinal adverse events associated with NSAIDs, nor opioids’ potential for addiction or overdose.
However, the history of tanezumab has been plagued with safety difficulties of its own. The development program began in 2004, and there have now been a total of 41 clinical studies that included almost 18,000 patients with pain conditions such as OA and chronic low back pain. The initial studies evaluated this agent administered intravenously in doses up to 10 mg, but in some of those studies there were cases of rapidly progressing OA that necessitated total joint replacement, and in 2010 and again in 2012, the FDA placed partial clinical holds on studies of nerve growth factor antibodies. The trials were resumed in 2015 using lower doses administered subcutaneously, and with efforts to minimize the potential joint hazards.
Post-2015 Data
Three large studies were completed under the risk mitigation program, forming the basis of Pfizer’s presentation.
The first (study 1056), included 696 patients whose mean age was 61. Led by Thomas J. Schnitzer, MD, PhD, of Northwestern University in Chicago and published in JAMA in 2019, the 16-week study evaluated patients receiving one subcutaneous injection of 2.5 mg of tanezumab on day 1 and either another 2.5-mg injection or a 5-mg injection at week 8, or placebo at both time points.
In the 2.5-mg group, scores on the Western Ontario and McMaster Universities (WOMAC) Osteoarthritis Index declined somewhat more with the drug versus placebo.
Rapidly progressive OA was observed in 2.2% of patients in the 2.5-mg group, in 0.4% of the 2.5/5-mg group, and in none of the placebo group. The incidence of total joint replacement in the three groups was 3.5%, 6.9%, and 1.7%, respectively.
The second study (1057) randomized 849 patients to 2.5 mg or 5 mg tanezumab or placebo every 8 weeks, with the primary endpoints being change from baseline at week 24 in WOMAC pain and physical function scores and patient global assessment of OA. Statistically significant improvements were seen on all three co-primary endpoints for the 5-mg dose, according to lead investigator Francis Berenbaum, MD, PhD, of the Sorbonne in Paris, but only on the pain and physical function scores for the 2.5-mg group. Rapidly progressive OA was reported in 1.4% of the 2.5-mg group and in 2.8% of the 5-mg group.
In studies 1056 and 1057, the treatment effect sizes for WOMAC pain were 0.60 and 0.46, respectively, which is considered a “modest” effect size compared with placebo.
Study 1058 was the head-to-head trial comparing the two tanezumab doses with NSAIDs over 56 weeks.
Patients initially received open-label NSAID therapy with prescription strength naproxen, celecoxib, or diclofenac and those who failed to achieve adequate pain relief were then randomized to tanezumab 2.5 or 5 mg or to continue on their NSAID.
“Hypothetically, if tanezumab is effective in NSAID non-responders, the pain should change minimally in patients who remain on NSAID and should decrease in patients switched to tanezumab. However, that is not what was observed,” FDA staff remarked in a briefing document. In the three groups, the average baseline pain score was approximately 7 out of 10, and all three showed an initial rapid decline in pain and then a long plateau at about 2.5 to 3 points below baseline.
The 2.5-mg dose did not meet any of the three co-primary endpoints at week 16, though the 5-mg dose showed significant differences versus NSAIDs for the WOMAC pain and physical function scores.
Safety was evaluated on a composite endpoint of rapidly progressive type 1 (decrease in joint space of 2 mm or more in one year) or type 2 (loss or destruction of bone), primary osteonecrosis, subchondral insufficiency fracture, or pathologic fracture. In a report presented at the 2019 annual meeting of the American College of Rheumatology, this endpoint was observed at rates of 37.4 per 1,000 patient-years for tanezumab 2.5 mg (P=0.002 vs NSAIDs), 71.5 per 1,000 for 5 mg (P<0.001 vs NSAIDs), and 14.8 per 1,000 for NSAIDs. The majority of these events (83.2%) were classified as rapidly progressive OA type 1.
In its briefing document, the FDA explained that in these studies, the risk for developing a composite joint adverse outcome compared with NSAIDs was 2.4 excess events (95% CI 1-3.8) per 100 patient-years for tanezumab 2.5 mg and 5.8 (95% CI 4-7.6) for tanezumab 5 mg. The risk of needing a total joint replacement also was elevated in the tanezumab groups in the placebo-controlled studies, with hazard ratios of approximately 2 for the 2.5-mg dose and above 3 for higher doses.
Pfizer had sought approval for the 2.5-mg dose.