Your Vioxx Concerns - Are Valdecoxib and Naproxen Dangerous?

Your Vioxx lawyer provides you the complete transcript of the February 16^th, 2005 joint meeting of the FDA's Arthritis Advisory Committee and the Drug Safety and Risk Management Advisory Committee. This section discusses Valdecoxib and Naproxen. We have formatted the complete transcript of the three day conference for easy of navigation to provide you with the best possible Vioxx information. To contact a Vioxx Lawyer, click here for a FREE case evaluation.

FDA Presentation: COX-2 CV Safety: Valdecoxib-Naproxen

DR. WITTER: What I am going to try and do is bring back some of the discussion I had earlier and specifically try and set some of this into some kind of a context. I was the primary reviewer for parecoxib. I was not the primary reviewer for valdecoxib so I have had to rely on reviewing reviews for the information I have here.

In terms of valdecoxib, the NDA came in on January 15 of 2001; 60 studies and I have listed them here again. We like to focus on the arthritis studies. There were 10 of those. There was a long-term exposure included which I will talk about briefly. I would just note again, as we have been discussing, that there has not been conducted a long-term outcomes type study. So that we are complete here, the original approval for valdecoxib did not contain a sulfonamide warning. That was addressed by subsequent label changes and "Dear Healthcare Provider" letters.

To give you a sense of comparison then from earlier studies, the patient-years are described here for OA and RA. You can see that the numbers are smaller than what we were describing earlier for example with celecoxib.

Turning to the deaths in the NDA database, there were 22 deaths and 17 of thee occurred during the double-blind studies, 4 in the CBG trial. So I will discuss that when I talk about parecoxib. Two of those were cardiovascular related. There were 8 deaths in patients receiving valdecoxib. Half of those were cardiovascular related. There were 3 in patients receiving NSAIDs; 2 were cardiovascular related. There were 2 non-cardiovascular related deaths that occurred in the cancer pain trial. During the open-label studies there were 5 deaths, 3 were cardiovascular related.

So, taking that information and looking at the number of deaths and patient-years and trying to give you some sense of comparison between my prior presentation, you can see that the highest mortality rate is in the group of valdecoxib plus the CABG patients at 4.7 percent. Recall that it was 3.7 percent; it was the highest from the prior discussion. If we exclude the 2 cases in CBG we come down to a rate of 3.5 percent. In the open-label studies the rate calculates out to 1.4 percent.

There were a couple of analyses that were conducted, special analyses that are listed here to look at the NDA for valdecoxib. This was to address the rate of serious thromboembolic cardiovascular events. They were in two patient populations, one that was described as high risk and the other was at risk. So, the high risk patients were those that had a history of angina, myocardial infarction, coronary-artery disease and cerebrovascular accident, while the at risk patients were described as those patients who had hypertension, hyperlipidemia or smoking.

The endpoints as defined by the NDA at that point for this special analysis were MI, myocardial ischemia, unstable angina, cardiac arrest, sudden death, CVA, TIA, pulmonary embolism, venous thrombosis, embolism in general, peripheral gangrene and peripheral ischemia.

Looking then at the high risk group and looking at cardiovascular safety in this group, you can see that there are small numbers of patients that fit into this category in particular when looking at the placebo arm here. When you look at the incidence rates of the events per 100 patient-years, you can see that there doesn't appear to be a consistent dose effect across the various doses. Valdecoxib doesn't appear to be any different than the NSAID comparators. The result here, looking at placebo, certainly appears to be a spurious result based upon the small number of patients and the event rates there. Looking at the at risk patients, there are more patients in this category. It gives us more patient-years to look at. The number of events is small. Again, calculating the incidence rates and the events per 100 patient-years, once again there doesn't appear to be any strong dose-response correlation here between the increasing doses of valdecoxib but they don't appear to be any different or any greater than what was seen in the comparator group.

As I mentioned, there was a study that was conducted at the urging of the agency to give us a better idea of the long-term cardiovascular events. This was study 047. This was a 6-month study that was conducted in patients with OA and RA. It was basically naproxen 500 BID against two doses of valdecoxib, 20 mg and 40 mg BID. I have listed here the percentage of patients who completed the 26-week trial, 43 percent naproxen and about 50 percent in both of the valdecoxib arms.

I would like to draw your attention to worsened blood pressure. There was a statistically significant, at p less than 0.05, increase in worsened blood pressure in the 40 mg BID group compared to naproxen. In general when you look at this data there was a dose trend against valdecoxib for all the events, with the exception here of palpitations. It was comparable across all the groups.

Turning then quickly to parecoxib, as we heard it is an intravenous/intramuscular formulation. One of the questions is why would we want to develop or anybody want to develop something like this? So, what I have done here in trying to help answer that question is the label that was in the toradol label--this was Table 3. What this table represents is a postmarketing study of 10,000 patients non-randomized, looking at the issues of incidence of clinically serious GI bleeding after 5 days on increasing doses of toradol, ranging from, in this case, less than or equal to 60 mg up to greater than 120 mg. There are two age categories here, less than 65 and greater than or equal to 65 years. The patients are broken out into those either without or with a history of perforations, ulcers or bleeds. As you can see and I have highlighted here, and that was one of the points of having this included in the labeling, is that as one increases the dose you increase the number of events. A quarter of the patients in fact had these serious GI bleeds. There also is an increase as you go through the categories of increasing event rate with age. So, I think this is part of the answer to the question as to why one we want to develop an intravenous or parenteral formulation of a COX-2 inhibitor.

Just to review quickly, parecoxib has a half-life of about 15-30 minutes. It breaks down into valdecoxib. What this does, and this is what we were concerned about, this allows exposure to different patient populations that have differing risk factors. The trials, however, were intended to address the issues in analgesia and we have some analgesic experts on the panel here. For example, the concept of multimodal analgesia is very much in the popular press these days. It is established that COX-2 has a role in all forms of pain, but there are also studies that looked at parenteral analgesia and opioid sparing and certain of these studies were conducted in concert with valdecoxib which you have heard about, the CABG trials, and I will just briefly review those too.

The original NDA for parecoxib was submitted on September 11 or 2000, 36 studies. They had a variety of studies, as I listed here. Just drawing your attention to the post-surgical analgesia trials, there were 8. There were 4 preoperative or preemptive analgesic trials, and there were 2 studies looking at opioid staring. The CAB 035 was one of those. The long-term safety is what I have already described in the valdecoxib 047.

This was CABG-I. The first CABG 035 as we know, was 2:1 randomization in terms of parecoxib to the placebo. I just want to point out the placebo in this case really refers to standard of care so this is patient controlled analgesia and opioids. The study was broken up into two phases, as we heard. The first 3 days was the IV/IM formulation and then when patients were able to take medicines by mouth they were transitioned into the valdecoxib, same dose, 40 mg twice a day or every 12 hours up until 14 days. Aspirin was a requirement for the study at less than or equal to 325 mg. Patients were studied to 30 days for events, which I will point out in a second.

This was a first of its kind study. This was a study to address the concerns that we had in the agency about this particular drug going into a high risk population. There were a lot of concerns. We were certainly aware of the various hypotheses and issues that are out there with COX-2s. So, we challenged the sponsor to come up with a study in a high risk population. This was the agreed to design of the CABG study but, as I alluded to here, it was a complex study not only because of the patients but because of the procedures and the co-medications.

So, to help address this there were blinded committees that were established to verify that the adverse events met established criteria to help figure out dates and attribution, and this is what was called then CRAEs, clinically relevant adverse events. As has been pointed out just earlier, there were no active controls in this CABG or the other CABG trial, and the discussion we had, which is what you are having, is would that have been an appropriate control in the first place given the risk factors associated with toradol for example?

The exposure, just to give you a sense, was more than 7 days. The bulk of the patients achieved that endpoint. To give you a sense of what the CRAEs were, they were defined as deaths, cardiovascular events, pericarditis, congestive heart failure, renal failure/dysfunction, TIA event, major non-GI bleeding requiring transfusion and infection which required institution of antibiotics and pulmonary complications.

What I would like to do is just briefly talk about some of these and give you a sense of the adjudication and what was actually looked at, pointing out once more that events were followed up to 30 days post last dose of study drug.

Looking at myocardial infarction in terms of a CRAE, to qualify into that definition you had to have two of the following four criteria as I have listed on this slide. For example, chest pain that was typical, not relieved by rest or nitrates; you had enzyme elevation as I have listed here, CK-MBs, LDHs, troponin levels. You had new wall motion abnormalities or you had EKG changes looking at ST-T and Q waves, as I have indicated on this slide. So, you had to meet two of the four criteria to be qualified as having an MI.

Turning to the events then and to some extent repeating the results you have seen but just to go over it again, there were the two groups, placebo and the parecoxib 40 mg BID group. I have listed here the intravenous for the first three days and the entire study. Looking at any event, you have a statistically significant, at p less than 0.12 by Fisher's exact test--the number of CRAEs in the entire study as compared to the placebo arm. When you look essentially at all of the adverse events as defined as CRAEs, just going down the list here, most of these are against parecoxib and valdecoxib. I would just draw your attention to some interesting ones. The MI, for example, there was only 1 event that fit the CRAE definition in both of these. On the other hand, there were 9 events that fit the CVA definition in the parecoxib/valdecoxib group.

Looking at the issue of MI adjudication, I just want to make this point, that there were 13 possible MIs. There were 11 that were in fact sent to the committee. These were 9 events in parecoxib and 2 in the placebo. Of these events, only 2 MIs survived the adjudication process so there was 1 that was listed for parecoxib and 1 placebo, which is what I just described in the prior slide. I note here that one of the rejected events was in the parecoxib group which resulted in death, probable MI of the patient.

What this brings up is the difficulty that we had on both sides trying to, you know, adjudicate these events relating to the timing of the drug. As I have suggested before, this was a complex setting. There wasn't a lot of experience in looking at this. So, that was a factor. Nonetheless, these results factored into my recommendation that this drug not be approved. It also was not approved because there was essentially limited information in terms of efficacy. It was essentially single dose information.

So, there was discussion that ensued with the sponsor in terms of thinking through these events and understanding a way forward. I am sorry, let me just describe the deaths for a second in parecoxib. There were 4, as I said before. There was one in a 58 year-old male who died of a duodenal ulcer. There was one in a 69 year-old female who died on day 19 of a probable MI. There was one in a 56 year-old male who died of septicemia, pneumonia. There was also one in a 62 year-old male who died of an infarct in the left cerebellum.

Given what I just said before, the issue was that perhaps the dosing was too high in that study. There was consideration that adjudication of events on the day of surgery and giving the drug on the day of surgery was not a good idea so that dosing for parecoxib would be delayed until the day after surgery. Then there was an attempt to try to get a handle on whether these events were occurring during the intravenous phase or during the oral phase, or both. So, that was part of the explanation for the repeat study, CABG-II, 071.

This then also had the CRAE definition, again studied for 30 days looking for events. This was a larger study. The groups this time are fairly balanced in terms of the numbers so you have placebo/placebo here; placebo for the first three days; valdecoxib to finish out the study; and then parecoxib/valdecoxib.

When you look at any of the CRAE events in either group that contains the COX-2 agents, there is a statistically significant difference compared to the placebo arm. When you look at all of the events of the CRAEs, for the most part they trend against the COX-2 selective agent, with the notable exception of DVTs. There was one in the placebo group and none in the other group.

I should just comment because there was a comment about it before, in CABG-I as well as this there were issues of wound healing and wound complication which was, to some extent, an unexplained finding. I should also just go back and remind everyone that there was an issue of hypotension that we noted, particularly in CABG-I for which still to this day there isn't, to my mind, a good explanation for.

The deaths in 071 included in the placebo group an intestinal perforation. The placebo/valdecoxib group included cardiac arrest, pneumonia and cardiac failure. The parecoxib/valdecoxib group included cardiac arrest, pulmonary embolism, myocardial infarction and ventricular fibrillation.

The question then ensued would the concern about what had happened in 035 in the at risk population extend to other patients, so there was study 069 that was designed which was meant to look at more general surgery with basically the same doses that we had seen in 071, in the second CABG trial. So, there was a 40 mg loading dose followed by 20 mg BID. These were more general surgical patients which included a mixture of orthopedic, GI, GYN, thoracic and a small amount of others.

Looking at the CRAEs in this study, and I have just then listed here for the entire study. Again, the groups are exactly balanced in terms of the numbers. When you look this time at the number of events the results look different in the sense that there tends to be more of these events in the placebo arm than the parecoxib/valdecoxib arm. With the exception of looking at MI, cardiac arrest and cardiac death, there are more events ion the parecoxib/valdecoxib arm than there are in the placebo arm.

This trial was included, as was indicated, in the current label for valdecoxib, as is study 071 which I didn't mention.

The deaths in 069 for the placebo included a cardiac failure, carcinoma, a mesenteric vein thrombosis and a cardiac arrest. In the parecoxib/valdecoxib group it included GI hemorrhage, MI and pulmonary embolism.

I will skip these slides and just make the following point, that as we think through safety for NSAIDs and COX-2s, what we have been hearing is, you know, think about the data that we have but I think we need to worry about the data that we don't have. As others have said and I am just reinforcing it here, the absence of evidence is not evidence of absence. So, there is a lot there that we still need to know. Thank you.

DR. WOOD: Great! Let's move straight on to the next presentation, and that is our last presentation for tonight. Then we will have the questions after that, if anyone is up to it still; hopefully not.

Weitz & Luxenberg is no longer accepting Vioxx cases.

see also:

COX-2 Safety-Celecoxib Vioxx Side Effects Lawyer - Vioxx/Celebrex FDA Transcript - FDA: COX-2-Celecoxib
Celecoxib Side Effects - The complete transcript of the FDA meeting.

Call to Order The Meeting Begins - Vioxx, Celebrex & Bextra FDA - Call to Order
Call to Order - The start of the FDA meeting discussing Vioxx safety

FDA Transcript - Day 1 Conference for Safety - Vioxx, Celebrex & Bextra FDA Transcript
Read here for the complete transcript of the FDA meeting for Vioxx