Hospital Pharmaceutical Assessment Final Summary Discussion Document No 6 (text extract)

This is the text extract for Hospital Pharmaceutical Assessment Final Summary Discussion Document No 6, browse documents here.

Hospital Pharmaceutical Assessment Final Summary Discussion Document No. 6

Celecoxib (Celebrex) and rofecoxib (Vioxx) - selective COX-2 inhibitors - for osteoarthritis and rheumatoid arthritis

Why produce this discussion document?

Summary

⇒ Celecoxib and rofecoxib are anti-inflammatory, analgesic and anti-pyretic agents that act via selective inhibition of cyclooxygenase-2 (COX2). They are indicated for the treatment of pain and inflammation in osteoarthritis and rheumatoid arthritis. ⇒ Two large pivotal trials have been conducted to assess the effectiveness of celecoxib (CLASS) and rofecoxib (VIGOR) compared with conventional nonsteriodal anti-inflammatory drugs (NSAIDs) in reducing gastrointestional (GI) complications. ⇒ COX-2s are no more effective than conventional NSAIDs in reducing pain, but may be associated with a lower rate of GI complications. However, there is evidence that this benefit is reduced with the coadministration of aspirin. ⇒ There is an increasing amount of evidence to suggest that some COX-2 selective inhibitors may have the potential to increase the risk of cardiovascular events compared to conventional NSAIDs. ⇒ The cost/QALY of celecoxib compared with conventional NSAIDS is estimated to be at least $60,000-$220,000 (depending on population risk of GI ulcer), wi th no overall benefit seen in celecoxib compared with diclofenac, or rofecoxib compared with naproxen.

At the November 2002 meeting of PHARMAC’s Hospital Pharmaceutical Advisory Committee (HPAC), members considered that it would be useful for PHARMAC to make available to District Health Boards (DHBs) some of it’s economic analyses, for pharmaceuticals used as much in hospitals as the community. One of the most commonly requested analyses by HPAC and DHB hospitals has been D evelop consensus for celecoxib R educe duplication (Celebrex®) and U ndertake analysis rofecoxib (Vioxx ®).

G enerate discussion

The focus of this S hare information summary document is therefore to highlight for DHBs the costeffectiveness, effectiveness and safety of the selective COX-2 inhibitors, celecoxib and rofecoxib, compared with conventional nonsteroidal anti-inflammatory drugs (NSAIDs), for the treatment of rheumatoid arthritis and osteoarthritis. The methodology used to construct this summary document is outlined in Appendix 1. In April 2003, PHARMAC distributed a draft version of this COX-2 inhibitor discussion document for comment. This final version has incorporated or addressed all significant issues that were raised during consultation.

⇒ COX-2 inhibitors provide little (if any) additional clinical benefit over traditional NSAIDS and at significantly higher cost. The net cost of funding celecoxib would be approximately $30 million per year.

INFORMATION FOR DHB USE ONLY

Introduction NSAIDs are used widely to treat pain and inflammation. There has been a large amount of interest in the new selective cyclooxygenase-2 (COX-2) inhibiting NSAIDs since their introduction. There is a perception that these drugs are less toxic than conventional NSAIDs. But are these drugs really safer than currently funded NSAIDs? Do they reduce the risk of serious upper gastrointestinal complications as initially claimed? Are the costs of COX-2s justified in terms of their tolerability and safety? Are they costeffective? Why doesn’t PHARMAC fund COX-2s for use in the community? Should hospitals fund them? This document attempts to answer these questions and explains PHARMAC’s stance regarding the funding of COX-2s. What are COX-2 Inhibitors? Celecoxib and rofecoxib are new members of the COX-2 inhibitor class of agents1 . These agents act to inhibit prostaglandin synthesis, primarily by selectively i hibiting the COX-2 enzyme. COX-2 n is induced in response to inflammatory stimuli, leading to the synthesis and accumulation of inflammatory prostanoids, in particular prostaglandin E2. This in turn causes inflammation with oedema and pain. Celecoxib and rofecoxib therefore act as antiinflammatory, analgesic and anti-pyretic agents by blocking the production of inflammatory prostanoids via COX-2 inhibition. Celecoxib and rofecoxib are indicated for the treatment of pain and inflammation in osteoarthritis (a chronic degenerative joint disease) and rheumatoid arthritis (a systemic inflammatory disease affecting joints and many other organs), and for acute pain and primary dysmenorrhea.

What other treatments are available? Paracetamol is first-line therapy for patients with mild to moderate symptoms of osteoarthritis (OA) and rheumatoid arthritis (RA) due to its effectiveness in relieving mild to moderate pain, and its safety profile [4,106]. Conventional NSAIDS (e.g. diclofenac and naproxen) are used in the short-term to control acute inflammatory episodes, and in cases where paracetamol is inadequate. Currently 14 different brands of NSAIDs are subsidised in New Zealand. For patients at risk of a gastrointestinal haemorrhage or gastroduodenal ulcer, a proton pump inhibitor (PPI) such as omeprazole can be added to treatment [15,106]. How much is the Cox-2 market worth? COX-2 selective inhibitors are among the most heavily advertised drugs in the United States and are significant revenue earners for the pharmaceutical companies. Merck Sharp and Dohme (MSD) spent US$161 million (NZ$347 million) advertising rofecoxib in the United States in 2000. This exceeded the advertising budgets for Pepsi and Budweiser beer [101]. In the same year the sales of rofecoxib quadrupled to US$1.5 billion (NZ$3.2 billion). In 2001 sales increased to US$2.6 billion (NZ$5.6 billion) [2]. Pfizer sells celecoxib, which had $3.1 billion (NZ$6.7 billion) in sales in 2001. Billions of dollars are being spent on COX-2 selective inhibitors worldwide. Is that investment justified? We examine the answer to this question in this discussion document. How effective are COX-2s?

COX-2s are no more effective than conventional NSAIDs in reducing pain and improving physical and global functions in both OA and RA patients.

Another COX-2 inhibitor assessed by PHARMAC is meloxicam. Information and PTAC’s assessesment of meloxicam is available in Appendix 2. INFORMATION FOR DHB USE ONLY

Celecoxib The Celecoxib Long-term Arthritis Safety Study (CLASS) was an amalgamation of two double blind RCTs for celecoxib (800 mg/day) involving 8,059 patients, one comparing with ibuprofen (2400 mg/day) over 12 months, the other to diclofenac (150 mg/day) over 15 months. These trials, which were funded by Pharmacia, found

that there was no difference in the effectiveness of celecoxib compared with ibuprofen or diclofenac, in terms of the reduction in pain [38]. Other clinical trials assessing the effectiveness of celecoxib compared with conventional NSAIDs in patients with OA [21,22,40] and RA [41,42] have reported similar results. A Cochrane review also concluded that celecoxib controls the symptoms of RA to a similar degree to that of naproxen, diclofenac, and ibuprofen [20]. Rofecoxib The Vioxx Gastro-selective Outcomes Research (VIGOR) study was a prospective, randomized, double-blind, one year study, evaluating the efficacy of rofecoxib (50 mg per/day) compared with naproxen (1000 mg per/day) in 8,076 patients with RA. The study found rofecoxib and naproxen to be similar in efficacy, with similar rates of discontinuation [35]. Other clinical trials comparing rofecoxib with conventional NSAIDS in patients with OA and RA over a period of 6 -54 weeks have reported similar results [31,33,34,36,37,65,100]. A Cochrane review also showed that rofecoxib demonstrates similar efficacy as naproxen for RA [64]. Celecoxib versus Rofecoxib We have identified three RCTs that compared the efficacy of celecoxib and rofecoxib. Geba et al. assessed the relative efficacy of rofecoxib (12.5 and 50 mg), celecoxib (200 mg) and acetaminophen (paracetamol) (4000 mg) in 382 patients with OA over a period of 6 weeks. This study, which was sponsored by MSD, found that patients administered rofecoxib (25 mg) had more effective relief of persistent pain when compared to patients’ administered celecoxib or acetaminophen, with a similar safety profile [45]. Gibofsky et al. assessed the relative efficacy of celecoxib (200 mg) and rofecoxib (25 mg) in 475 patients with OA of the knee over a period of 6 weeks [104]. This study found celecoxib and rofecoxib demonstrated similar efficacy, and were significantly better compared with placebo [104]. Bianchi and Broggini compared the efficacy of celecoxib (200 mg), rofecoxib (25 mg) and nimesulide (100 mg) in 30 patients with OA of the knee for 7 days [105]. The study found that numesulide was significantly more effective in providing symptomatic relief compared with celecoxib and rofecoxib [105].

INFORMATION FOR DHB USE ONLY

The results of these trials do need to be interpreted with caution. It is not clear whether the doses used in these trials are comparable. It has been suggested that celecoxib may have been more effective if the dose was 400 mg [46,47]. Also, the duration of the trials are only between one to six weeks, and the results may change over a longer period of time. Do COX-2s reduce the risk of serious gastrointestinal complications?

There is varying evidence regarding the effectiveness of COX-2s in reducing gastrointestinal (GI) complications. The two dominant trials, CLASS (6-months) and VIGOR (9-months), reported reductions in ulcer complications with COX-2s compared with conventional NSAIDS. In the VIGOR trial this difference was statistically significant. In the CLASS trial, the incidence of symptomatic ulcers and ulcer complications combined was significantly less for patients administered celecoxib. However, the full 12-15 month CLASS data found no difference in ulcer complications between celecoxib and existing NSAIDs overall. The lower GI risks associated with rofecoxib may have been overstated as the VIGOR trial did not allow concurrent use of low-dose aspirin.

Initially it looked as if COX-2s would provide a similar level of pain relief as NSAIDs, but with a reduced risk of gastrointestinal (GI) bleeding; a result that many received eagerly. Short-term RCTs showed fewer cumulative gastroduodenal erosions and endoscopic ulcers with the COX-2 inhibitors (9-15%) than with conventional NSAIDs (41-46%) [6,7]. Further studies, however, began to question the integrity of the original results. New information suggests that previous claims over the GI safety of COX-2s may have been overstated. Celecoxib The CLASS study results were published in the Journal of the American Medical Association (JAMA), but only the first six months of data were reported from two trials of longer duration. The 6 -month trial results reported a statistically significant reduction with celecoxib in ulcer complications and symptomatic ulcers combined compared with conventional NSAIDS (8% and 1.2% respectively). However, the reduction in

ulcer complications alone for patients administered celecoxib compared with existing NSAIDS was not found to be statistically significant (0.3% and 0.6%, respectively). Further, the full 12-15 month results reported no significant difference in ulcer complications between celecoxib and existing NSAIDs (0.5% and 0.6%, respectively). Almost all the ulcer complications that had occurred during the second half of the trials were in the celecoxib group [62,63]. Why the difference between short and long-term data? The authors of the CLASS study attempted to explain the differences in the short and long-term data by arguing that it was caused by the high dropout rate of patients being administered conventional NSAIDs resulting in a lower rate of GI events in these patients [61]. However, as Peter Juni and colleagues pointed out, the number of dropouts increased gradually over the study period, with no sudden increase after six months, and there was little difference in withdrawals between treatment groups [17]. A concern is that the published data from the CLASS trial have been widely distributed and believed. About 30,000 prints were bought from the publisher, which coincided with their increase in sales of celecoxib [17]. Former JAMA editorin-chief, George Lundberg, reportedly stated that “for a group of researchers to send incomplete information to a journal for consideration, while knowing that a more complete set will be reviewed by an authority figure like the FDA would seem very strange. That is, unless the timesensitive marketing advantage of a drug with blockbuster sales potential was so compelling that the manufacturer was willing to take that chance to gain an early mass sales advantage” [18]. In fact, the selective reporting of the 6-month CLASS results was subject to an almost unprecedented editorial in the BMJ, heavily critical of JAMA’s reporting of the CLASS trial results [17]. Juni et al commented on how the authors failed to justify the post hoc changes in trial design, outcomes, and analysis, and provided an unconvincing explanation for reporting only on the six-month follow-up. They stated that “publishing and distributing overoptimistic short term data using post hoc changes to the protocol, while omitting disappointing long-term data of two trials, which involved large numbers of volunteers, is misleading”.

INFORMATION FOR DHB USE ONLY

Concurrent use of aspirin:

Low-dose aspirin might offset any GI benefit of COX-2 inhibitors.

In CLASS, those concurrently taking low-dose aspirin (approximately 21% of patients) did not show any reduction in GI complications with celecoxib compared with conventional NSAIDS (further information in Appendix 3). Other Evidence Deeks et al. undertook an industry-funded systematic review and meta-analysis of RCTs comparing celecoxib with conventional NSAIDs or placebo. The authors reported equivalent efficacy between celecoxib and conventional NSAIDS, but significantly greater tolerability in terms of withdrawals from the trials as a result of GI complications including symptomatic ulcers, perforation, and haemorrhage [28]. However, the Deeks meta -analysis has proved controversial [72], mainly as it did not use the 1215 month CLASS data that were reported to the FDA [29, 30]. PHARMAC staff went on to apply the full CLASS data to the Deeks analysis, and found that these made a substantial difference to the GI safety profile of celecoxib [29]. Further information on PHARMAC’s analysis, and key findings of CLASS and other clinical trials on celecoxib, is included in Appendix 3. In a prominent RCT, published after Deeks metaanalysis, Chan et al looked at the use of COX-2s in patients with a history of recent ulcer bleeding. They found the risk of recurrent bleeding was high regardless of treatment (4.9% of those using celecoxib and 6.4% using diclofenac-plusomeprazole had recurrent bleeding within six months). The authors suggested that neither regimen could completely protect patients at high risk from recurrent ulcer complications [24]. An accompanying editorial stated that the results were unexpected, with neither regimen providing a good or even acceptable level of protection from recurrent ulcer bleeding. “It is clear that our current strategy for the prevention of NSAIDrelated ulcers and ulcer complications requires reexamination and revision”[70]. Rofecoxib The VIGOR study demonstrated that rofecoxib was associated with a lower incidence of complicated ulcers compared with naproxen

(0.4% compared with 0.9%, respectively). The risk of a complicated GI event was estimated to be 1.37 per patient per year with naproxen and 0.59 with rofecoxib (p=0.005) [35]. These results need to be interpreted with caution as they may not be replicated in postmarket use for the following reasons: 1. patients in the VIGOR trial were not able to be administered aspirin concurrently, hence the benefit of rofecoxib may be overstated (as aspirin was shown to offset GI benefit of cox2 inhibitors in the CLASS trial); 2. the comparator used in the VIGOR trial was naproxen, and some argue that diclofenac and ibuprofen are associated with a lower risk of gastrointestional complications, and hence are the more relevant comparator [17]; 3. the difference in GI events emerged only after one month of treatment, indicating that rofecoxib may not reduce GI risk during short-term treatment [106]; 4. patients in poor general health were ineligible for the VIGOR trial, hence results cannot be extrapolated to their situation [106]. Laine et al. performed a post hoc analysis on serious lower and upper GI events in the VIGOR tria l. They found that the risk of serious lower GI events was 54% lower for patients administered rofecoxib compared with naproxen [43]. The number needed to treat with rofecoxib instead of naproxen to prevent one GI event was 10-12 in the highest risk patients (prior event, age >or=75 years, or severe rheumatoid arthritis), 17-33 in patients with other risk factors, and 42-106 in low-risk patients [44]. It should be noted that patients who were administered NSAIDs in the CLASS and VIGOR trials were maintained on the maximum dose of the NSAIDs, rather than having their dose reduced to the minimum effective dose, as is done in clinical practice. Patients therefore had an elevated risk of GI complications [4]. Appendix 3 describes GI effects in two other rofecoxib trials.

Is there an increased risk of cardiac or renal events associated with COX-2s?

The CLASS trial found that there was no significant difference between the treatment groups in the incidence of major cardiovascular events. However, the VIGOR trial reported a fourfold increase of myocardial infarction in patients administered rofecoxib.

There is growing international concern regarding links between COX-2 inhibitors and an increased risk of serious adverse cardiac events. This is a subject of ongoing debate [49,51,52,53,54,55,67]. Small, short-term trials have reached different conclusions regarding the effect of COX-2 inhibitors on cardiac events. One study found that the prevalence of cardiac events was similar in patients prescribed celecoxib and rofecoxib [54], another found that prevalence was lower in patients administered celecoxib compared with rofecoxib [49], and a recent trial found that COX2s potentially may benefit those with cardiovascular disease [55]. Mukherjee et al. performed a meta-analysis to investigate the effect of COX-2s on cardiovascular events. The review found that there was a potential increase in cardiovascular events with the use of COX-2 inhibitors [51]. Data from the FDA Adverse Event Reporting System (AERS) indicates that rofecoxib and celecoxib are associated with rates of renal failure similar to conventional NSAIDs. However, this has been debated in the literature [56,57,58,59,60]. Celecoxib In the CLASS trial, in which 21 percent of patients took aspirin, there was no significant difference between the treatment groups in the incidence of major cardiovascular events. White et al. analysed data from the CLASS trial to determine the incidences of serious cardiovascular thromboembolic events for all patients as well as the subgroup of patients who were not administered aspirin concurrently. They found that the incidence of serious cardiovascular thromboembolic events was not significantly different for all patients administered celecoxib or conventional NSAIDs, and for the subgroups of patients not taking aspirin [23].

INFORMATION FOR DHB USE ONLY

Rofecoxib In the VIGOR study, patients administered rofecoxib had a four-fold increase in myocardial infarctions compared with patients on the conventional NSAID naproxen (0.1% vs 0.4%, respectively). Serious thrombotic events were twice as high in the rofecoxib group (RR 2.38), although these data were not reported in the VIGOR publication. This increase in thrombotic cardiovascular events was found to be significantly higher in all patients administered rofecoxib (including patients who were not needing aspirin). Why the difference in results? The fact that the VIGOR trial found a statistically significant difference in major cardiovascular events, and the CLASS trial did not, may have been due to differences in study patients between the trials (the CLASS trial had a higher proportion of patients with OA, and patients with RA have an increased risk of thrombotic events); the use of aspirin by some patients in the CLASS trial; or the nature of the comparator NSAID used in the trials (naproxen – the comparator NSAID in the VIGOR trial - may be protective against cardiovascular events)[26]. Interaction with Warfarin, Ace Inhibitors and Diuretics

The interaction between COX-2 inhibitors and warfarin can cause serious haemorrhagic events. The interaction between COX-2 inhibitors, ace inhibitors and diuretics can cause renal failure and impairment at a similar rate to conventional NSAIDs.

So how safe are COX-2 inhibitors overall?

The incidence of overall serious adverse events in the two key COX-2 trials (CLASS and VIGOR) was higher for COX-2 inhibitors than conventional NSAIDs. The absolute risk increase was 1 -1.5% and the number needed to treat to cause one harmful event was between 67 and 100.

The view that COX-2s are ‘safe’ to use in those patients for whom conventional NSAIDs w ould normally be prescribed may not be well founded. A recent POEM (Patient-Oriented Evidence that Matters) in the BMJ, summarising a recent detailed analysis of safety effects in CLASS and VIGOR [67], stated “The FDA’s data show no benefit in terms of gastrointestional mortality with COX-2 inhibitors, a trend toward greater all-cause mortality, and a significant increase in serious adverse events…COX 2 inhibitors have serious potential harms that have been minimised in reports of research sponsored by drug companies” [66]. Patients who began treatment on celecoxib and rofecoxib in New Zealand between December 2000 and October 2003 are currently being monitored by the Intensive Medicines Monitoring Programme (IMMP). In June 2002, 13 case reports were received of acute psychiatric events for patients administered COX-2 inhibitors [109]. In April 2003, 17 cases of hepatotoxicity associated with COX-2 inhibitors were reported, including 6 cases of significant liver injury [107]. In November 2003, 7 cases of visual disturbance with COX-2 inhibitors were reported by the IMMP in the BMJ, including a case of temporary blindness and a case of a visual field defect [108]. Recovery from these events occurred soon after discontinuing medication. It should be noted that these adverse effects have also been reported with conventional NSAIDs. Celecoxib The 6-month data from the CLASS trial reported in JAMA showed that significantly more patients who were administered conventional NSAIDs had nausea, constipation, and hypertension, and significantly more patients who were administered celecoxib had skin reactions including rash and urticaria [38]. However, this report excluded some key safety data.

The interaction between COX-2 inhibitors and warfarin can cause serious haemorrhagic events by increasing the international normalised ratio (INR) by approximately 10% [10,26,48,110]. This was raised as a concern by New Zealand’s Medicines Adverse Reactions Committee (MARC) [10]. In June 2002, the Centre for Adverse Reactions Monitoring (CARM) released an article outlining the dangers of combining ace inhibitors, NSAIDs and diuretics, due to the effect the combination may have on renal failure and impairment [8].

INFORMATION FOR DHB USE ONLY

The 12-15 month CLASS results reported on the FDA website found that there was no significant difference in total serious adverse events2 or mortality for patients administered celecoxib compared with other NSAIDs (rate of serious adverse events was 6.8% and 5.8% respectively, mortality rate was 0.4% and 0.43% respectively). However, in the FDA results there was an increase in other serious events 3 for patients administered celecoxib compared with other NSAIDs (5.8% versus 4.8%, respectively), representing an absolute risk increase of 1% and the number needed to treat to cause one harmful event of 100. 22.4% of celecoxib patients and 24.6% of patients on other NSAIDs withdrew due to adverse events. Deeks’ meta -analysis (which used the 6-month CLASS data) found that withdrawals due to all adverse events were not significantly increased with celecoxib compared with placebo or conventional NSAIDs, although withdrawals due to gastrointesional events were lower from patients administered celecoxib compared with conventional NSAIDs. Rofecoxib As with CLASS, only the favourable results of the VIGOR trial were submitted for publication. The article only included discontinuations due to GI events rather than all discontinuations, which favoured rofecoxib. The total incidence of adverse events was significantly higher for patients administered rofecoxib compared with naproxen (9.3% vs 7.8%, respectively, p=0.013), causing an absolute risk increase of 1.5% and a number needed to treat to cause one harmful event of 67 [95,96]. In the FDA results, the number of patients that needed to be withdrawn due to hypertensionrelated adverse events was significantly higher in the rofecoxib group compared with naproxen (1.0% vs 0.2%, respectively, p<0.001) [97]. Even though rofecoxib reduced the incidence of complicated ulcers, the absolute risk reduction was 0.52, where the absolute risk increase from other serious adverse events was 1.5 (mainly due

to an increase in serious cardiovascular adverse events), resulting in an unattractive risk to benefit ratio [96]. What was PTAC’s assessment celecoxib and rofecoxib? of

The Pharmacology and Therapeutics Advisory Committee (PTAC) considered applications for celecoxib and rofecoxib between 1999 and 2001. Further details of PTAC’s assessment can be found in Appendix 4. Celecoxib PTAC considered that celecoxib appears to be similar in effectiveness and safety as other NSAIDs, but may have a better GI safety profile. However, the committee considered that the place of therapy and safety profile of celecoxib still needed to be fully elucidated from post-marketing experience. The committee considered celecoxib to be expensive and agreed that the additional expenditure over NSAIDs would not be justified considering the modest decrease in serious GI complications. In addition, patients at high risk of adverse events, including gastro-intestinal ones, would still have to be considered at high risk if treated with celecoxib and that, for this reason, PTAC could not recommend targeting to any specific sub-groups of patients. Rofecoxib PTAC considered that the efficacy of rofecoxib was comparable with other conventional NSAIDs. The committee considered that rofecoxib had a superior GI safety profile when compared with conventional NSAIDs. However, as with celecoxib, it considered that the true safety prof ile, particularly in elderly patients at high risk, would only become apparent after considerable post-marketing experience. The committee considered that concomitant use of low dose aspirin (to provide an anti-platelet effect) is likely to be routine in high-risk cardiac patients and this will reduce the GI safety advantages of rofecoxib. Similarly, an increase in cardiovascular risk associated with rofecoxib in patients not on aspirin will also offset the safety advantages. The committee considered that patients with a high risk of developing upper GI perforations, ulcers and bleeding (PUBs), who are prescribed rofecoxib, might still be prescribed a Proton Pump Inhibitor (PPI).

Total serious adverse events include death, hospitalisation, or extension of hospitalisation, and any life-threatening event or disability (including complicated ulcers). 3 Other serious adverse events exclude death and complicated ulcer. INFORMATION FOR DHB USE ONLY

The committee considered that the cost offsets in terms of reduction in the use of PPIs in patients with gastrotoxicity induced by conventional NSAIDs might not necessarily be high after they switched to rofecoxib. The committee noted that despite the lack of comparative data there was good evidence suggesting that a combination o f PPI with a conventional NSAID protects against further PUBs, and could be used as an alternative approach. The committee considered that patients at high risk of adverse events, including gastro-intestinal ones, would still be at risk if treated with rofecoxib and that, for this reason, it could not recommend targeting to any specific sub-groups of patients. The committee also considered the price of rofecoxib was disproportionately high compared to its potential benefits. Has an economic analysis been done in New Zealand? In early 2002 PHARMAC staff undertook a preliminary economic analysis for celecoxib that utilised the 6 -month CLASS data. This analysis was referred to in the draft version of this discussion document, and showed that the incremental cost-utility of celecoxib compared with conventional NSAIDs was approximately $502,000 per quality-adjusted life year (QALY). A detailed analysis on celecoxib and rofecoxib for the symptomatic relief of OA and RA in patients not receiving aspirin was completed for PHARMAC in December 2003 [119]. This analysis was based on a Markov model developed by the Canadian Coordinating Health Technology Agency (CCOHTA) to assess the costeffectiveness of celecoxib and rofecoxib over a period of 5 years, using data from the VIGOR and CLASS4 trials [68]. The results of the analysis indicated that the cost/QALY of celecoxib compared with conventional NSAIDs is approximately $60,000-$220,000 (depending on population risk of GI ulcer), with no overall benefit seen in celecoxib compared with diclofenac, or rofecoxib compared with naproxen. Note that this report contains only a summary of the full analysis. For a copy of the full analysis, please contact the hospital pharmaceutical analyst at PHARMAC.

Clinical Inputs in Model The detailed analysis included a systematic review of clinical trials on celecoxib and rofecoxib that showed that the two classes of anti-inflammatories have equivalent efficacy [98], so only the potential for harm from upper GI and MI events were modelled. The analysis estimated the incremental costs and benefits of rofecoxib versus naproxen, and celecoxib versus diclofenac, and celecoxib versus ibuprofen, in two separate patient groups: 1. all patients with RA or OA; 2. patients at high risk of an upper GI ulcer. The mode l was based on the following key assumptions:

Table 1: Key Assumptions used in Model Effectiveness of COX-2s Loss of benefit of COX-2s with aspirin GI Events OA and RA patients have the same probability of upper GI (UGI) events and MI. UGI events occur at a constant rate over time. No significant difference in UGI events and MI for low or high doses of COX-2s [90]. No significant difference in the probability of a UGI event for patients receiving a non-NSAID analgesic (e.g. paracetamol) or a COX-2 inhibitor [68,114,115]. All symptomatic ulcers are treated for 3 months with PPIs and 100% are cured. Mortality rates among patients with complicated UGI events are independent of the model of therapy (medical or surgical), hence an overall mortality rate was used. The COX-2 doses that are modelled are those that are indicated as maintenance therapy in RA and OA [117,118]. These are lower than the doses used in VIGOR and CLASS. The analysis assumes that the safety profiles of COX-2s are identical at high and usual doses, although this may be an optimistic assumption. MI Events All patients who have an MI and survive are hospitalised. NSAIDs are assumed not to be cardio-protective. Other Non-compliance with analgesic use was not modelled (the systematic review found no significant difference in withdrawal).

The analysis was based on the 12- and 15-month CLASS data. INFORMATION FOR DHB USE ONLY

The detailed analysis incorporated the following events:

Table 2: Event Rates used in Model Variable GI Events Proportion at high risk of UGI events Probability of symptomatic and complicated ulcer in individuals with a history of ulcer Probability of hospitalisation with a complicated ulcer Probability of surgery if hospitalised Probability of retrying NSAIDs after GI bleed Case fatality following complicated UGI event Effectiveness of PPIs in ulcer prophylaxis when receiving conventional NSAID therapy MI Events Case fatality following MI Values 8% (0-100%) 3.3% (2.5-4%) 63% 8.5% 10% (5-15%) 11.7% (low value 4.85%) 40% Ref. 35, 38 35, 38

The utility of acute MI was derived from the New Zealand disability weight for treated acute MI [89], and was estimated to be 0.745 . Costs The following costs were included in the analysis:

Table 3: Costs used in Model Cost (NZ$) Pharmaceuticals (cost per day) Rofecoxib 25 mg od Celecoxib 200 mg bd Naproxen 500 mg bd 6 Diclofenac 75 mg bd 6 Ibuprofen 600mg tds 6 Paracetamol 1 g qds 6 Omeprazole 20 mg od 6 Hospitalisation Costs Fatal MI 7 Non-fatal MI 7 Inpatient medical treatment for complicated ulcer 7 Inpatient surgical treatment for complicated ulcer 7 Outpatient Treatment Specialist Gastroenterologist visit, initial consultation, excl GST 8 Outpatient gastroscopy, excl. GST 9 GP visit including patient co-payment, excl. GST 10 $1.20 ($1-$2.30) $1.20 ($1-$2.30) $0.212 $0.135 $0.1863 $0.0784 $0.827 $2,848 $4,188 $2,258 $4, 016

82 68 Expert advice 83, 111, 112 82, 84

Probability of a fatal MI occurring outside hospital Post MI mortality at age 60

35-56% (age/gender dependent) 62-85% (age/gender dependent) (1/34)exp(0. 055*years post MI) NZ mortality data 1995/96

$178 $622 $33

87, 113

Other Events Probability of death from any cause

Quality of Life The baseline quality of life for patients with RA and OA was estimated for the detailed analysis as 0.688 (where 1=perfect health, and 0=death) [68]. The effect of ulcer complications on quality of life was estimated using results reported in a COX-2 economic analysis [88], which were derived from a study of quality of life in peptic ulcer disease. In this study a symptomatic ulcer had a utility of 0.88 (representing severe dyspepsia), a complicated ulcer treated medically in hospital had a utility of 0.49 for 4 days treatment (utility of 0.86 for one 3 -month cycle), and a complicated ulcer requiring surgery had a utility of 0.46 (utility of 0.83 for 3-month cycle).

Cost of Cox-2s The current cost of rofecoxib and celecoxib to a consumer is approximately $2.01 for rofecoxib 25 mg daily and $2.20 for celecoxib 200 mg daily. It is likely that the price would be significantly lower if subsidised therefore an arbitrary lower price was chosen as the base case. Other non-pharmaceutical health sector costs The cost of a symptomatic ulcer included the cost of a GP visit and two endoscopies. The cost of an outpatie nt complicated ulcer included the cost of a GP visit, a specialist visit and two endoscopies [116]. The cost of managing coronary artery disease (CAD) was estimated as $200 per month [90].

INFORMATION FOR DHB USE ONLY

Weighted average of acute MI (0.605) for approximately 6 days (median length of hospitalisation), and 0.75 for the remainder of the 3-month cycle. 6 April 2003 Pharmaceutical Schedule 7 2000/01 NZ public hospital weighted DRGs 8 Private gastroenterology clinic, Auckland 2003 9 Private hospital, Auckland 2003 10 Statistics New Zealand 9

So are COX-2 inhibitors cost-effective?

Rofecoxib was found to be associated with higher total costs and fewer benefits compared with naproxen, and celecoxib was found to be associated with higher costs and fewer benefits compared with diclofenac. In patients at average risk of a GI event, the incremental cost-utility ratio (ICUR) of celecoxib compared to ibuprofen is over $200,000 per QALY. For patients at high risk of a GI event, the ICUR is over $60,000 per QALY. This suggests that COX-2s do not represent good value for money compared with other pharmaceuticals that could be funded.

For individuals at high risk of MI, the adverse cardiovascular effects could outweigh any GIrelated benefits, especially as patients are likely to be on concurrent low dose aspirin. While the use of low dose aspirin may reduce any adverse cardiovascular effects, it will also reduce any GIrelated benefits of COX-2 inhibitors. In summary, celecoxib and rofecoxib are not cost effective compared with NSAIDs for the symptomatic treatment of RA and OA. At a cost per QALY of over $60,000, COX-2s are relatively poor value for money compared with other pharmaceuticals that could be funded (refer to Appendix 5 for the cost/QALY of other pharmaceuticals). As the risk of GI bleeds and MI events increases with age, the results of the analysis on celecoxib compared with ibuprofen were tested to determine how sensitive they were to changes in the age of patients. Varying the age of patients between 45 years and 75 years, caused the cost per QALY to decrease from $90,000 to $40,000. Because the incidence of UGI events is higher than MI events, COX-2s provide additional benefits for the older age group, despite the potential for higher MI rates. However, if aspirin is withheld, the MI rates may be higher, and hence the cost/QALY would increase. The lowest cost per QALY ($33,000) was obtained in a 2-way sensitivity analysis for celecoxib compared with ibuprofen in elderly patients with a high risk of GI events but less severe arthritis. The detailed analysis confirms the results of PHARMAC’s earlier preliminary analysis, and is consistent with findings internationally. For instance, the Canadian Coordinating Office for Health Technology assessment (CCOHTA) found that that COX-2s are relatively cost-ineffective when used for patients with average risk of GI events, and more cost-effective for patients with a history o UGI events [68]. Note the CCOHTA f analysis did include concomitant increases in MI risk with age. Similarly poor cost-effectiveness was seen in other North American analyses [94]. A recent report in ‘The American Journal of Managed Care’ suggests that the cost-effectiveness of COX2 inhibitors may have previously been overestimated [99]. Previous analyses assumed that the use of gastroprotective agents was lower for patients administered COX-2 inhibitors,

Cost effectiveness is frequently measured in terms of the incremental cost-utility ratio (ICUR). This is defined as the difference in net costs divided by the difference in net benefits. This is where benefits are measured in terms of quality-adjusted life years (QALYs) that result from substituting one treatment for another. Net costs include nonpharmaceutical costs borne by other parts of the public health sector.

Table 4: Cost/QALY Results Pharmaceutical Comparator Cost/QALY Result Average Risk of GI Event Rofecoxib Naproxen Higher costs and fewer benefits Celecoxib Diclofenac Higher costs and fewer benefits Celecoxib Ibuprofen $220,000 High Risk of GI Event Rofecoxib Naproxen Higher costs and fewer benefits Celecoxib Diclofenac Higher costs and fewer benefits Celecoxib Ibuprofen $68,000

In predominately average risk population, rofecoxib was dominated by naproxen and celecoxib was dominated by diclofenac (i.e. the COX-2s have higher total costs and provide fewer benefits in terms of number of QALYs). Celecoxib compared with ibuprofen had an incremental cost per QALY gained of over $200,000, and as such is not considered to be cost effective. For patients at high risk of a GI event, rofecoxib was dominated by naproxen and celecoxib was dominated by diclofenac. The incremental cost per QALY was approximately $68,000 for celecoxib compared to ibuprofen, indicating that celecoxib is not cost-effective for high-risk populations.

INFORMATION FOR DHB USE ONLY

however analysis of medical claims data has indicated that this is not the case. Rather, the use of gastroprotective agents was actually higher amongst patients administered COX-2 inhibitors compared with those administered conventional NSAIDs. This resulted in an increase in the cost/QALY by more than $80,000 [99]. “But Australia funds them…….” The rapid uptake of celecoxib and rofecoxib following their listing on the Pharmaceutical Benefits Scheme (PBS) in Australia has been unprecedented by any other pharmaceutical. 800,000 prescriptions were written in the first 30 days following PBS listing of celecoxib in August 2000, and 1.5 million prescriptions were written by the end of the first four months [3,101]. Recent research indicates that physicians are using COX 2 inhibitors indiscriminately. In Australia, 59% of patients prescribed a COX-2 inhibitor for the first time had not received a prescription for a conventional NSAID in the previous 12 months [5]. In Canada, 75% of patients prescribed a COX-2 inhibitor had not had a prescription filled for an NSAID in the 120 days before a COX 2 inhibitor was first prescribed, and 20% had not received an NSAID in five years [71]. When celecoxib was initially listed on the PBS in Australia in August 2000, the projected cost was in the order of $40m in the first year. Instead it cost over $170m in the first year, and has continued to increase. COX-2s now cost Australians AUS$218.6m a year (NZ$256.3m) [5]. COX-2s were identified as a major reason for the health budget blowout in Australia in 2002 and are now under review. What effect would funding of these drugs have on the pharmaceutical budget? PHARMAC’s analysis has indicated that the cost of funding COX-2 inhibitors is likely to be in excess of $30 million per year. PHARMAC understands that the private market is in the region of $10 million per year. This estimate was obtained before the pharmaceutical suppliers began direct-to-consumer advertising (DTCA) of these products, and when celecoxib and rofecoxib were the only COX-2 inhibitors

INFORMATION FOR DHB USE ONLY

marketed in New Zealand (valdecoxib and etoricoxib are now available in New Zealand and DTCA is being used extensively). It is therefore likely that the current private for COX-2 inhibitors is in excess of $10 million per year. The prevalence of OA and RA in New Zealand is at least 130,000. Considering an annual cost of $400 per patient, even if only 50 percent of the patient population accessed Cox II inhibitors, the market would increase to $26 million. Given the likely impact of intensive marketing, this figure could easily be higher. If the uptake of COX-2 inhibitors in New Zealand was similar to Australia, the gross expenditure could be as high as NZ$41 million per year (after adjusting for population variances). Accounting for savings from standard NSAIDs, the net cost of Cox II inhibitors in this case would be up to $35 million per year. Note that Australia restricts DTCA, hence uptake in New Zealand may be higher. In New Zealand, such large spending would need to come from DHBs’ budgets, and would be at the expense of other health funding priorities. Will COX-2s ever be funded in the community? Given the minimal benefits associated with COX2 inhibitors compared with NSAIDs, the potential for increased risk of cardiovascular events, and the high cost of the drugs, PHARMAC has some reservations about the COX-2 inhibitors. PHARMAC has decided that on balance we cannot justify subsidising the COX-2 inhibitors at the current price because to do so would compromise our ability to fund other needed medicines and for DHBs to fund other healthcare. If funding were available, other medicines would represent better value for money and have a higher priority. Therefore, at its October 2003 Meeting, the PHARMAC Board resolved to decline the listing of celecoxib, rofecoxib and meloxicam on the Pharmaceutical Schedule. Our analyses have indicated that currently the COX-2 inhibitors do not represent an appropriate use of taxpayers’ funds. However, PHARMAC will continue to examine the evidence as it becomes available.

Should hospitals fund them? This discussion document has highlighted the concerns PHARMAC has regarding COX-2 inhibitors and explains our stance for not funding them. We hope that by sharing this information DHBs will question their reasons for funding (or not funding) COX-2s and discuss the benefits, safety and costs associated with such decisions. The evidence on COX-2 inhibitors shows that COX-2 inhibitors provide little (if any) additional clinical benefit over conventional NSAIDs. COX2s are not exempt from the severe GI effects associated with NSAIDs. Assuming equivalent efficacy, the risk to benefit ratio of COX-2 inhibitors depends on the cumulative effects of GI and cardiovascular side-effects, for which the data remain controversial. At present, it is difficult to give an accurate account between the relief of pain and improved function on the one hand, and the likelihood of serious adverse effects on the other [16]. PHARMAC's analyses indicate that the costeffectiveness of celecoxib and rofecoxib is poor overall compared with conventional NSAIDs, at $60,000-$220,000 per QALY. Conclusion “Pandora’s box has been opened. We have developed a false sense of security about the protective value of proton-pump inhibitors and COX-2 inhibitors, and now it is clear that the problem needs to be re-examined completely” [70]. To date there is no evidence available suggesting that the COX-2 inhibitors work better than older medicines, such as the conventional NSAIDs. There is mounting evidence suggesting that the GI profile of the COX-2 inhibitors is not substantially better than that of older medicines such as the conventional NSAIDs. While the relative risk of serious GI complications (perforation, obstruction or bleeding) may be reduced with COX-2 inhibitors, in real terms the absolute risk of such complications is so low that this reduction provides only a very small advantage over other NSAIDs. There is also an increasing amount of evidence that suggests that some COX-2 inhibitors may

INFORMATION FOR DHB USE ONLY

have a potential for increased risk of cardiovascular events compared with conventional NSAIDs. MARC has also raised similar concerns. This discussion document has outlined the effectiveness, safety, and cost-effectiveness associated with COX-2 inhibitors, and explains PHARMAC’s stance for not funding them.

References

1. Mathers C, Vos T, Stevenson C, The Burden of Disease and Injury in Australia, Australian Institute of Health and Welfare, 1999 Peterson M, F.D.A. takes mixed stance on a leading arthritis drug, New York Times, 06-08-2002 http://www.nps.org.au/ National Prescribing Service Newsletter, Osteoarthritis – have COX-2s changed its management?, Issue 18, 2001 Prescribing Practice Review 16, Cox-2 selective NSAIDs, National Prescribing Service Limited, 2001 Goldstein JL, Correa P, et al., Reduced incidence of gastroduodenal ulcers with celecoxib, a novel cyclooxygenase2 inhibitor, compared to naproxen in patients with arthritis, Am J Gastroenterol 2001, 96: 1019-1027 Laine L, Harper S, et al., A randomised trial comparing the effect of rofecoxib, a cyclooxygenase 2-specific inhibitor, with that of ibuprofen on the gastroduodenal mucosa of patients with osteoarthritis, Gastroenterology 1999; 117: 776-783 Savage R, A dangerous trio, Prescriber Update 2002, 23(2): 20 Topham-Kindley L, Cox 2 drugs not devoid of common side effects, Doctor September 2002

22. Kivitz AJ, Moskowitz RW, et al., Comparative efficacy and safety of celecoxib and naproxen in the treatment of osteoarthritis of the hip, J Int Med Res 2001, 29(6): 467-79 23. White WB, Faich G, et al., Comparison of thromboemolic events in patients treated with celecoxib, a cyclooxygenase-2 specific inhibitor, versus ibuprofen or diclofenac, Am J Cardiol 2002, 15(4): 425-30 24. Chan FK, Hung LC, Suen BY, Wu JC, Lee KC, et al. Celecoxib versus diclofenac and omeprazole in reducing the risk of recurrent ulcer bleeding in patients with arthritis, NEJM 2002;347: 2104-2110. 25. Goldstein JL, Silverstein FE, et al., Reduced risk of upper gastrointestional ulcer complications with celecoxib, a novel COX-2 inhibitor, Am J of Gastro 2000, 95(7): 1681-1690 26. Fitzgerald GA, Patrono C, The coxibs, selective inhibitors of cyclooxygenase-2, NEJM 2001, 345(6): 433-442 27. Therapeutics letter, Selective COX-2 inhibitors: Are they safer NSAIDs?, Therapeutics Initiative, January/February 2001 28. Deeks JJ, Lesley AS, Bradley MD, Efficacy, tolerability, and upper gastrointestinal safety of celecoxib for treatment of osteoarthritis and rheumatoid arthritis: systematic review of randomised controlled trials, BMJ 2002, 325: 619-27 29. Metcalfe S, Dougherty S, McNee W, Celebrex’s relative gastrointestional safety is overstated, BMJ 2003, 326: 334-5. http://bmj.com/cgi/content/full/326/7384/334 30. Juni P, Sterchi R, Dieppe P, Problems compromise review’s validity, BMJ 2003, 326: 334. http://bmj.com/cgi/content/full/326/7384/334 31. Day R, Morrison B, et al., A randomised trial of the efficacy and tolerability of the COX-2 inhibitor rofecoxib vs ibuprofen in patients with osteoarthritis. Rofecoxib/Ibuprofen Comparator Study Group, Arch Intern Med 2000, 160(12): 1781-7 32. Hawkey C, Laine L, et al., Comparison of the effect of reofecoxib (a cyclooxygenase 2 inhibitor), ibuprofen, and placebo on the gastroduodenal mucosa of patients with osteoarthritis: a randomised, double-blind, placebo-controlled trial. The Rofecoxib Endoscopy Multinational Study Group, Arthritis Rheum 2000, 43(2): 370-7 33. Saag K, van der Heijde D, Rofecoxib, a new cyclooxygenase 2 inhibitor, shows sustained efficacy, comparable with other nonsteroidal anti-inflammatory drugs – A 6-week and a 1-year trial in patients with osteoarthritis, Arch Fam Med 2000, 9: 1124-1134 34. Schnitzer TJ, Truitt K, The safety profile, tolerability, and effective dose range of rofecoxib in the treatment of rheumatoid arthritis. Phase II Rofecoxib Rheumatoid Arthritis Study Group, Clin Ther 1999, 21(10): 1688-702 35. Bombardier C, Laine L, et al., Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheum atoid arthritis, VIGOR Study Group, NEJM 2000, 343: 1520-8 36. Cannon GW, Caldwell JR, Rofecoxib, a specific inhibitor of cyclooxygenase 2, with clinical efficacy comparable with that of diclofenac sodium: results of a one-year, randomised, clinical trial in patients with osteoarthritis of the knee and hip. Rofecoxib Phase III Protocol 035 Study Group, Arthritis Rheum 2000, 43(5): 978-87 37. Acevedo E, Castaneda O, et al., Tolerability profiles of rofecoxib (Vioxx) and Arthrotec. A comparison of six weeks treatment in patients with osteoarthritis, Scan J Rheumatol 2001, 30(1): 19-24

2. 3. 4. 5. 6.

8. 9.

10. Medsafe, Adverse reaction reporting and IMMP, Adverse reactions of current concern, March 2003, www.medsafe.govt.nz/profs/adverse/cc.htm 11. Coulter D, Acute psychiatric events with COX-2 inhibitors, Prescriber Update 2002, 23(2): 21 12. Savage R, Can patients stomach COX-2 inhibitors, Prescriiber Update 2002; 23(2): 18-19 13. National Institute for Clinical Excellence, Technology Appraisal Guideance – No. 27, Guidance on the use of cyclo oxygenase (Cox) II selective inhibitors, celecoxib, rofecoxib, meloxicam, and etodolac for osteoarthritis and rheumatoid arthritis, July 2001. http://www.nice.org.uk/pdf/coxiifullguidance.pdf 14. FDA panel finds no safety benefit for Celebrex, Scrip 2001, 2616: 19 15. Treatment of rheumatoid arthritis – Unknown long-term effects, Prescrire 2001, 10(52): 55-61 16. Jones R, Efficacy and safety of COX 2 inhibitors – New data are encouraging but the risk:benefit ratio remains unclear, BMJ 2002, 325: 607-8 17. Juni P, Rutjes AWS, Dieppe PA, Are selective COX 2 inhibitors superior to traditional non steroidal antiinflammatory drugs?, BMJ 2002, 324: 1287-8 18. Gottlieb S, Researchers deny any attempt to mislead the public over JAMA article on arthritis drug, BMJ 2001, 323: 301 19. Mamdani M, Juurlink DN, Anderson GM, Little is known about COX 2 inhibitors, Letters to the editor, BMJ 2002, 325: 162 20. Garner S, Fidan D, et al., Celecoxib for rheumatoid arthritis (Cochrane Review), The Cochrane Library, Issue 1, 2003 21. McKenna F, Borenstein D, et al., Celecoxib versus diclofenac in the management of osteoarthritis of the knee, Scan J Rheumatol 2001, 30(1): 11-8

INFORMATION FOR DHB USE ONLY

38. Silverstein FE, Faich G, et al., Gastrointestional toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis – The CLASS study: A randomized controlled trial, JAMA 2000, 284(10): 1247-1255 39. Lanas A, Clinical Experience with cyclooxygenase-2 inhibitors, Rheumatology 2002, 41(suppl 1): 16-22 40. Bensen WG, Fiechtner JJ, et al., Treatment of osteoporosis with celecoxib, a cyclooxygenase-2 inhibitor: a randomise controlled trial, Mayo Clin Proc 1999, 74(11): 1095-105 41. Simon LS, Weaver AL, et al., Anti-inflammatory and upper gastrointestinal effects of celecoxib in rheumatoid arthritis – a randomised controlled trial, JAMA 1999, 282(20): 1921-1928 42. Emery P, Zeidler H, et al., Celecoxib versus diclofenac in longterm management of rheumatoid arthritis: randomised doubleblind comparison, Lancet 1999, 354(9196): 2106-11 43. Laine L, Connors LG, et al., Serious lower gastrointestinal clinical events with nonselective NSAID or coxib use, Gastroenterology 2003, 142(2): 288-92 44. Laine L, Bombardier C, et al., Stratifying the risk of NSAIDrelated gastrointestinal clinical events: results of a double-blind outcomes study in patients with rheumatoid arthritis, Gastroenterology 2002, 123(4): 1006-12 45. Geba GP, Weaver AL, et al., Efficacy of rofecoxib, celecoxib, and acteaminophen in osteoarthritis of the knee – a randomised trial, JAMA 2002, 287(1): 64-71 46. Henderson P, Rofecoxib vs celecoxib vs acetaminophen for treatment of osteoarthritis, JAMA 2002, 287 (14), 1799 47. Saseen J, Which is most effective for osteoarthritis of the knee: rofecoxib, celecoxib, or acetaminophen?, J Fam Pract 2002, 51(4) 48. Schwartz JI, Bugianesi KJ, et al., The effect of rofecoxib on the pharmacodynamics and pharmcokinetics of warfarin, Clin Pharmacol Ther 2000, 68(6): 626-36 49. Whelton A, Fort JG, et al., Cyclooxygenase-2 specific inhibitors and cardiorenal function: a randomised, controlled trial of celecoxib and rofecoxib in older hypertensive osteoarthritis patients, Am J Ther 2001, 8(2): 85-95 50. Ahmad SR, Kortepeter C, et al., Renal failure associated with the use of celecoxib and rofecoxib, Drug Saf 2000, 25(7): 53744 51. Mukherjee D, Nissen SE, Topol EJ, Risk of cardiovascular events associated with selective COX-2 inhibitors, JAMA 2001, 286(8): 954-959 52. Konstam MA, Weir MR, Current perspective on the cardiovascular effects of coxibs, Cleve Clin J Med 2002, 69(Supp 1): S147-52 53. Bing RJ, Cyclooxygenase-2 inhibitors: is there an association with coronary or renal events?, Curr Atheroscler Rep 2003, 5(2): 114-7 54. Harley C, Wagner S, The prevalence of cardiorenal risk factors in patients prescribed nonsteroidal anti-inflammatory drugs: Data from managed care, Clin Ther 2003, 25(1): 139-49 55. Chenevard R, Hurlimann D, et al., Selective COX-2 inhibition improves endothelial function in coronary artery disease, Circulation 2003, 107(3): 405-9 56. Niccoli L, Bellino S, Cantini F, Renal tolerability of three commonly employed non-steroidal anti-inflammatory drugs in elderly patients with osteoarthritis, Clin Exp Rheumatol 2002, 20(2): 201-7

57. Zhao SZ, Reynolds MW, et al., A comparison of renal-related adverse drug reactions between rofecoxib and celecoxib, based on the World Health Organization/ Uppsala Monit oring Centre safety database, Clin Ther 2001, 23(9): 1478-91 58. Eras J, Perazella MA, NSAIDs and the kidney revisited: are selective cyclooxygenase-2 inhibitors safe?, Am J Med Sci 2001, 321(3): 181-90 59. Whelton A, Maurath CJ, et al., Renal safety and tolerability of celecoxib, a novel cyclooxygenase-2 inhibitor, Am J Ther 2000, 7(3): 159-75 60. Brater DC, Harris C, et al., Renal effects of COX-2 selective inhibitors, Am J Nephrol 2001, 21(1): 1-15 61. Silverstein F, Simon L, Faich G, Reporting of 6-month vs 12month data in a clinical trial of celecoxib. In reply. JAMA 2001, 286(19): 2399 62. Witter J. Celebrex capsules (celecoxib). NDA 20-998/S-009. Medical officer review. US Department of Health and Human Services food and Drug Administration, 2000. www.fda.gov/ohrms/dockets/ac/01/briefing/3677b1_03_med.p df. 63. Lu HL. Celebrex capsules (celecoxib). NDA 20-998/S-009. Statistical reviewer briefing document for the advisory committee. US Department of Health and Human Services food and Drug Administration, 2000. www.fda.gov/ohrms/dockets/ac/01/briefing/3677b1_04_stats.d oc. 64. Garner S, Fidan D, et al., Rofecoxib for rheumatoid arthritis (Cochrane Review), The Cochrane Library, Issue 1, 2003 65. Geusens PP, Truitt K, et al., A placebo and active comparatorcontrolled trial of rofecoxib for the treatment of rheumatoid arthritis, Scand J Rheumatol. 2002, 31(4): 230-8. 66. POEM – Harms outweigh benefits of COX 2 for many patients, BMJ 2003: 326:0. http://bmj.com/cgi/content/full/326/7389/0/f. 67. Wright JM, The double-edged sword of COX-2 selective NSAIDs, CMAJ 2002, 167(10): 1131-7, http://www.cmaj.ca/cgi/content/full/167/10/1131?ijkey=/dBFX YG0o7bUU 68. Maetzel A, Krahn M, Naglie G, The cost -effectiveness of celecoxib and rofecoxib in patients with osteoarthritis or rheumatoid arthritis, Canadian Coordinating Office for Health Technology Assessment 2001, Technology Report No. 23. http://www.ccohta.ca/ 69. Rapid responses to: Deeks JJ, Lesley AS, Bradley MD, Efficacy, tolerability, and upper gastrointestinal safety of celecoxib for treatment of osteoarthritis and rheumatoid arthritis: systematic review of randomised controlled trials, BMJ 2002, 325: 619-27. http://bmj.com/cgi/eletters/325/7365/619 70. Graham DY. NSAIDs, Helicobacter pylori, and Pandora's Box. N Engl J Med. 2002;347:2162-4. 71. Mamdani M, Rochon P, Laupacis A, Anderson G. Initial patterns of use of COX-2 inhibitors by elderly patients in Ontario: findings and implications. eCMAJ 2002 167: 11251126. http://www.cmaj.ca/cgi/content/full/167/10/1125 72. 19 electronic rapid responses accepted by the BMJ, at http://bmj.com/cgi/eletters/325/7365/619 73. NZ WIES 8A cost weights 2001/02 (ANDRG prices), weighted by 1995/96 DRG volumes 74. Australian Refined Diagnosis Related Groups, Version 4.1, Definitions Manual, Commonwealth Department of Health and Aged Care, Canberra. (1998)

INFORMATION FOR DHB USE ONLY

75. Furst DE, Kolba KS, Fleischmann R, et al., Dose response and safety study of meloxicam up to 22.5 mg daily in rheumatoid arthritis: a 12 week multicenter, double blind, dose response study versus placebo and diclofenac, J Rheumatol 2002, 29(3): 436-46 76. Wojtulewski JA, Schattenkirchner M, Barcelo P, et al., A six month double-blind trial to compare the efficacy and safety of meloxicam 7.5 mg daily and naproxen 750 mg daily in patients with rheumatoid arthritis, Br J Rheumatol. 1996, 35(suppl 1): 22-8 77. Yocum D, Fleischmann R, Dalgin P, et al., Safety and efficacy of meloxicam in the treatment of osteoarthritis: a 12-week, double-blind, multiple-dose, placebo-controlled trial. The Meloxicam Osteoarthritis Investigators. Arch Intern Med 2000, 160(19): 2947-54 78. Linden B, Distel M, Bluhmki E, A double-blind study to compare the efficacy and safety of meloxicam 15 mg with piroxicam 20 mg in patients with osteoarthritis of the hip, Br J Rheumatol. 1996, 35 (suppl 1): 35-8 79. Goei The HS, Lund B, Distel MR, Bluhmki E, A double-blind, randomised trial to compare meloxicam 15 mg with dicofenac 100 mg in the treatment of osteoarthritis of the knee, Osteoarthritis Cartilage 1997, 5(4): 283-8 80. Dequeker J, Hawkey C, et al., Improvement in gastrointestional tolerability of the selective cyclooxygenase (COX)-2 inhibitor, meloxicam, compared with piroxicam: results of the Safety and Efficacy Large-scale Evaluation of COX-inhibiting Therapies (SELECT) trial in osteoarthritis, Br J Rheumatol 1998, 37(9): 946-51 81. Hawkey C, Kahan A, et al., Gastrointestinal tolerability of meloxicam compared to diclofenac in osteoarthritis patients. International MELLISSA Study Group. Meloxicam Large-scale International Study Safety Assessment, Br J Rheumatol 1998, 37(9): 937-45 82. Silverstein FE, Graham DY, Senior J, Davies H, Struthers B, Bittman R. Misoprostol reduces serious gastrointestinal complications in patients with rheumatoid arthritis receiving nonsteroidal anti-inflammatory drugs. A randomized, doubleblind, placebo -controlled trial. Annals of internal medicine 1995;123(4):241-9. 83. Katschinski B, Logan R, Davies J, Faulkner G, Pearson J, Langman M. Prognostic factors in upper gastrointestinal bleeding. Digestive diseases and sciences 1994;39(4):706-12. 84. Rostom A, Dube C, Wells G, Tugwell P, Welch V, Jolicoeur E, et al. Prevention of NSAID-induced gastroduodenal ulcers (Cochrane Review). Issue 2. Oxford: Update Software: The Cochrane Library; 2003. 85. Norris R, United Kingdom Heart Attack Study C ollaborative Group. Fatality outside hospital from acute coronary events in three British health districts, 1994-5. British Medical Journal 1998;316:1065-70. 86. Statistics New Zealand. New Zealand Life Tables (1995-97) Reference Report. Available at http://www.stats.govt.nz/domino/external/web/prod_serv.nsf/ht mldocs/New+Zealand+Life+Tables+(1995-97)++Reference+Report . 87. Mark D, Hlatky M, Califf R, Naylor C, Lee K, Armstrong P, et al. Cost effectiveness of thrombolytic therapy with tissue plasminogen activator as compared with streptokinase for acute myocardial infarction. New England Journal of Medicine 1995;332:1481-24. 88. Spiegel B, Targownik L, Dulai G, Gralnek I. The cost effectiveness of cyclooxygenase-2 selective inhibitors in the management of chronic arthritis. Annals of internal medicine 2003;138(10):795-806.

89. Tobias M, the New Zealand Burden of Disease Study team. The burden of disease and injury in New Zealand. Public health intelligence occasional bulletin 1. Wellington: Ministry of Health; 2001. 90. CCOHTA 2002. The cost -effectiveness of celecoxib and rofecoxib in patients with osteoarthritis or rheumatoid arthritis. 91. Hernandez-Diaz S, Rodriguez LA. Association between nonsteroidal anti-inflammatory drugs and upper gastrointestinal tract bleeding/perforation: an overview of epidemiologic studies published in the 1990s. Arch Intern Med 2000;160(14):2093-9. 92. Singh G, Rosen RD. NSAID induced gastrointestinal complications: the ARAMIS perspective--1997. Arthritis, Rheumatism, and Aging Medical Information System. J Rheumatol Suppl 1998;51:8 -16. 93. Wilson PW, D'Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation 1998 May 12;97(18):1837-47. 94. Spiegel B, Targownik L, Dulai G, Gralnek I. The cost effectiveness of cyclooxygenase-2 selective inhibitors in the management of chronic arthritis. Annals of internal medicine 2003;138(10):795-806 95. US Food and Drug Administration. NDA 21-042, s007, Vioxx Gastrointestional Safety 2001. http://www.fda.gov/ohrms/dockets/ac/01/briefing/3677b2_03_ med.doc 96. Therapeutics letter – COX-2 inhibitors update: Do journal publications tell the full story?, Therapeutics Initiative 2001-02 97. US Food and Drug Administration. NDA 21-042, s007, Statistical Reviewer Briefing Document for the Advisory Committee 2000. 98. Rafter N, Jackson, R, A review of the effectiveness of Cox-2 inhibitors compared with NSAIDs, 2003 99. Verification of a Decision Analytic Model Assumption Using Real-World Practice Data: Implications for the Cost Effectiveness of Cyclo-ozygenase 2 Inhibitors (COX-2s), American Journal of Managed Care 2003, 9(12): 785-794 100. Lisse JR, Perlman M, Johansson G, Shoemaker JR, Schechtman J, Skalky CS, Dixon ME, Polis AB, Mollen AJ, Geba GP; ADVANTAGE Study Group. Gastrointestinal tolerability and effectiveness of rofecoxib versus naproxen in the treatment of osteoarthritis: a randomized, controlled trial. Ann Intern Med. 2003 Oct 7;139(7):539-46. 101. Dowden JS. Coax, COX and cola. Med J Aust. 2003 Oct 20;179(8):397-8. 102. Cutts C, LaCaze A, Tett S. A clinical audit of the prescribing of celecoxib and rofecoxib in Australian rural general practice. Br J Clin Pharmacol. 2002 Nov;54(5):522-7. 103. Kerr SJ, Mant A, Horn FE, McGeechan K, Sayer GP. Lessons from early large-scale adoption of celecoxib and rofecoxib by Australian general practitioners. Med J Aust. 2003 Oct 20;179(8):403-7. 104. Gibofsky A, Williams GW, McKenna F, Fort JG. Comparing the efficacy of cyclooxygenase 2-specific inhibitors in treating osteoarthritis: appropriate trial design considerations and results of a randomized, placebo-controlled trial. Arthritis Rheum. 2003 Nov;48(11):3102-11.

INFORMATION FOR DHB USE ONLY

105. Bianchi M, Broggini M. A randomised, double-blind, clinical trial comparing the efficacy of nimesulide, celecoxib and rofecoxib in osteoarthritis of the knee. Drugs. 2003;63 Suppl 1:37-46. 106. [No authors listed] Rofecoxib in rheumatoid arthritis: new indication. No better that other NSAIDS. Prescrire Int. 2003 Jun;12(65):89-90. 107. Coulter D, COX-2 inhibitors and hapatotoxicity, Prescriber Update 2003; 24(1):2 www.medsafe.govt.nz/profs/puarticles/cox2liver.htm 108. Coulter DM, Clark DW, Savage RL. Celecoxib, rofecoxib, and acute temporary visual impairment. BMJ. 2003 Nov 22;327(7425):1214-5. 109. Coulter D, Acute psychiatric reactions with COX-2 inhibitors, Prescriber Update 2002,; 34(2):21 http://medsafe.govt.nz/profs/puarticles/cox2psych.htm 110. Medsafe Editorial Team, Interact ion between COX-2 inhibitors and warfarin, Prescriber update 22:16-18, http://medsafe.govt.nz/profs/puarticles/cox2warf.htm 111. Branicki F, Coleman S, Fok P, Pritchett C, Fan S-t, Lai E, et al. Bleeding peptic ulcer: a prospective evaluation of risk factors for rebleeding and mortality. World Journal of Surgery 1990;14:262-70. 112. Zimmerman J, Siguencia J, Tsvang E, Beeri R, Arnon R. Predictors of mortality in patients admitted to hospital for acute upper gastrointestinal hemorrhage. Scandinavian Journal of Gastroenterology 1995;30:327-31. 113. The GUSTO Investigators. An international randomized trial comparing four thrombolytic strategies for acute myocardial infarction. New England Journal of Medicine 1993;329(10):673-82. 114. Langman MJ, Jensen DM, Watson DJ, Harper SE, Zhao PL, Quan H, et al. Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs. Journal of the American Medical Association 1999;282(20):1929-1933. 115. Watson DJ, Harper SE, Zhao PL, Quan H, Bolognese JA, Simon TJ. Gastrointestinal tolerability of the selective cyclooxygenase-2 (COX-2) inhibitor rofecoxib compared with nonselective COX-1 and COX-2 inhibitors in osteoarthritis. Archives of Internal Medicine 2000;160(19):2998-3003. 116. New Zealand Guidelines Group. The management of dyspepsia. Best practice evidence-based guideline. Consultation draft. April 2003. http://www.nzgg.org.nz/development/wip.cfm#Dyspepsia 117. Celebrex data sheet. http://medsafe.govt.nz. 118. Vioxx data sheet. http://medsafe.govt.nz. 119. Rafter N, Milne R, Jackson R. Technology Assessment Report No. 55 – Listing rofecoxib and celecoxib in the Pharmaceutical Schedule. 2003

Appendix 1 - Methodology

For this review, we searched PubMed, the Cochrane database, the TRIP database (http://www.tripdatabase.com/ ), and various internet sites11 . Sites were searched between 19 March and 7 April 2003. We used information from randomised control trials (RCTs), review articles, meta-analyses, guidelines, and economic analyses that were identified in the searches. We also used reference lists of sourced articles to obtain further papers not identified in the above searches; pharmaceutical industry submissions and data supplied to PTAC through PHARMAC. We then undertook a narrative synthesis of the above systematically derived data. This summary has been reviewed both within PHARMAC and by PTAC members.

Appendix 2 – Meloxicam

Meloxicam (Mobic) is a COX-2 inhibitor that is indicated for the symptomatic treatment of painful osteoarthritis (arthrosis, degenerative joint disease) and symptomatic treatment of rheumatoid arthritis. Clinical trials have shown that meloxicam has comparable therapeutic efficacy to standard dosages of other NSAIDs such as piroxicam, diclofenac and naproxen and caused fewer gastrointestinal adverse events than those NSAIDs [75,76,77,78,79,80,81]. PTAC considered an application to list meloxicam on the Pharmaceutical Schedule in February 1999 for the treatment of osteoarthritis and rheumatoid arthritis. Further information was considered by PTAC in February 2001. The minutes of the meeting are summarised below: The committee considered that meloxicam appears to have similar efficacy to other NSAIDs with minor advantages in terms of the incidence of serious gastrointestinal effects. It noted that at a dose of 7.5 mg/day the absolute risk of serious gastro-intestinal side effects was slightly less than diclofenac 100 mg/day and appears similar to other COX-2 inhibitors. The committee considered that the risk of serious gastro-intestinal events in high-risk patients remained significant with meloxicam and that it could not be assumed that meloxicam may be safely used in patients with risk factors. It considered that patients at higher risk of adverse events, especially gastro-intestinal ones, would still be at higher risk if treated with meloxicam.

Other sites searched: Medsafe, CCOHTA, NICE, FDA, Australia Prescriber. Sites were searche d between 19 March and 7 April 2003. Search terms were: celecoxib Limits: Randomized Controlled Trial; celecoxib AND cost*; celecoxib AND rofecoxib; rofecoxib Limits: Randomized Controlled Trial; rofecoxib AND cost*; celecoxib AND economic; rofecoxib AND economic; rofecoxib AND cardiovascular; celecoxib AND cardiovascular; rofecoxib AND warfarin; celecoxib AND warfarin etc.

INFORMATION FOR DHB USE ONLY

Appendix 3 – Further Information on GI events

CLASS Trial The CLASS study results were published in JAMA, but only the first six months of data were reported from two trials of longer duration. One of the trials was a 15month trial comparing celecoxib with ibuprofen and the other was a 12-month trial comparing celecoxib with diclofenac. The primary endpoint of the CLASS study, which was published in JAMA, was upper GI ulcer complications. The rate of symptomatic ulcers during the first six months of treatment was also assessed. The trials reported a statistically significant reduction in ulcer complications and symptomatic ulcers compared with conventional NSAIDS [38]. However, the 12-15 month data show that the risk of complicated ulcers was not significantly different between celecoxib and existing NSAIDs (0.5% and 0.6%, respectively). Almost all the ulcer complications that had occurred during the second half of the trials were in the celecoxib group [62, 63]. The six-month data reported a relative risk reduction (RRR) of 39% in serious upper GI events and endoscopic ulcers for patients administered celecoxib [38]. However, the 1215 month data showed no statistically significant reduction in relative risk. The CLASS study also found that those concurrently taking low-dose aspirin (approximately 21% of patients) did not show any reduction in GI complications with celecoxib compared with conventional NSAIDS. The upper GI complication rates for patients who were also administered low dose aspirin was 2.0% for patients administered celecoxib and 2.1% for patients administered ibuprofen/diclofenac. Non-aspirin users who were administered celecoxib had a statistically significant 42% RRR, where aspirin users who were administered celecoxib showed no risk reduction (RR 1.02 (0.59 1.74)). The difference between the subgroups’ RRRs over the 12-15 months was statistically significant (p 0.03). Therefore, low-dose aspirin may offset any benefit of COX-2 inhibitors. Other Evidence Goldstein et al. conducted a pooled analysis of 14 RCTs (11,008 patients) and a separate analysis on an open-label trial (5,155 patients) on celecoxib compared with conventional NSAIDs for patients with RA or OA. In the RCTs, no upper GI events occurred in the placebo group, 2 occurred with celecoxib (annualised incidence 0.2%) and 9 occurred with conventional NSAIDs (1.68%, p=0.002). Nine cases occurred in the open-label study (annualised incidence 0.18%). The study concluded that the incidence of upper GI ulcer complications was 8-fold lower with celecoxib compared with conventional NSAIDS, and a similar rate to placebo [25].

INFORMATION FOR DHB USE ONLY

Chan et al. assessed the effectiveness of celecoxib compared with diclofenac combined with omeprazole in 287 patients with RA and OA who presented with ulcer bleeding. The study found there to be no significant difference in the two treatments’ effect on recurrent GI bleeding [24]. Other trials have reported either significantly lower [41,42] or comparable [40] rates of GI complaints with celecoxib [41, 42] compared with conventional NSAIDs. Deeks Meta-Analysis Deeks et al. report the finding of a systematic review and meta-analysis of RCTs comparing the efficacy, tolerability and upper gastrointestinal safety of celecoxib compared with conventional NSAIDs or placebo. They identified nine trials, involving 15,172 patients with OA or RA. CLASS contributed to over half of the patients analysed. The authors reported equivalent efficacy between celecoxib and conventional NSAIDS, but significantly greater tolerability in terms of withdrawals from the trials as a result of gastrointestinal complications including symptomatic ulcers, perforation, and haemorrhage. Deeks et al. reported that the RRR in serious GI events for patients administered celecoxib was 46% [28]. However, this meta-analysis has proved controversial, and a number of correspondents pointed out several errors (in total 19 rapid responses were accepted by the BMJ, see http://bmj.com/cgi/eletters/325/7365/619). Questions were raised about the appropriateness of pooling of data from the two 6 -month CLASS trials. Another criticism was the use of short-term trial results (only the CLASS trial was longer than 26 weeks) when patients generally take NSAIDs for many years [30]. The meta-analysis therefore does not provide us with any information on the long-term safety of celecoxib. Perhaps most importantly, a number of authors pointed out that the meta-analysis did not use the 12-15 month CLASS data that were reported to the FDA [29, 30]. PHARMAC staff went on to apply the full CLASS data to the Deeks analysis, and found that these made a substantial difference to the GI safety profile of celecoxib [29]. Combining the 12-month CLASS data with the seven trials in Deeks analysis, and then adjusting for the much longer exposure time experienced in CLASS, the overall RRR in serious GI events decreased to 32% (from 46% reported by Deeks). These results suggest that although celecoxib still causes statistically significant reductions in GI adverse events overall, reductions were appreciably less than suggested [29] (see Figure 1 below).

Figure 1. Withdrawals because of adverse GI effects in celecoxib vs. NSAID RCTs

event rate 1.5% 2.0%

Rofecoxib VIGOR Trial The VIGOR study demonstrated that rofecoxib was associated with a lower incidence of complicated ulcers compared to naproxen (0.4% compared with 0.9%, respectively) [35]. Laine et al. performed a post hoc analysis on serious lower and upper gastrointestinal events in the VIGOR trial. They found that serious lower GI events were 54% lower for patients administered rofecoxib compared with naproxen [43]. The number needed to treat with rofecoxib instead of naproxen to prevent one GI event was 10-12 in the highest risk patients (prior event, age >or=75 years, or severe rheumatoid arthritis), 17-33 in patients with other risk factors, and 42-106 in low-risk patients [44]. Other Trials Hawkey et al. evaluated the effect of rofecoxib, ibuprofen, and placebo on the gastroduodenal mucosa of 775 patients with OA over a period of 24 weeks. The study found that ulcers were significantly less common in patients treated with rofecoxib [32]. Laine et al. reached similar conclusions in the trial of 742 OA patients randomised to received rofecoxib or isoprofen [7].

0.0%

0.5%

1.0%

2.5%

3.0%

3.5%

control treatment

CLASS 12/15-month

RCTs in Deeks + CLASS 12/15-mnth

Deeks et al also reported no statistically significant difference between low-dose aspirin and non-aspirin use for both endoscopic ulcers and for CLASS. PHARMAC however found that including the 12-15 month CLASS data gave a non-significant 28% RRR for aspirin use, compared with a 72% RRR for nonaspirin use – a statistically significant difference between RRRs. Adjusting for CLASS’ longer exposure gave again a non-significant 4% RRR for aspirin users, versus 52% for non-aspirin use, p <0.01). Hence PHARMAC disagreed with any implication that celecoxib’s benefits extend equally to aspirin users [29] (see Figure 2 below). Figure 2. Aspirin use in celecoxib vs. NSAID RCTs, for GDU + withdrawals from GI events: RCTs in Deeks + CLASS 12/15-month, exposure/variance-weighted

0.0%

CLASS 12/15-month, aspirin

1.0%

2.0%

event rate 3.0% 4.0%

5.0%

6.0%

7.0%

CLASS 12/15-month, non-aspirin

control treatment

RCTs in Deeks + CLASS 12/15-mnth, aspirin

RCTs in Deeks + CLASS 12/15-mnth, non-aspirin

Further information on PHARMAC’s analysis (and Deeks et al. response) can be found on PHARMAC’s website: www.pharmac.govt.nz/economic_analysis.asp.

INFORMATION FOR DHB USE ONLY

Appendix 4 – PTAC minutes

Included below are the amalgamated and summarised PTAC minutes on celecoxib and rofecoxib from the February 2000 to November 2001 PTAC meetings. Celecoxib PTAC considered an application for celecoxib in November 1999. Further information on celecoxib was considered in February 2000, May 2000, August 2000, and May 2001. The committee considered that celecoxib appears to be similar in effectiveness and safety as other NSAIDs, but may have a better gastrointestinal safety profile. However, the latter issue has not been fully elucidated. The committee noted there have been numerous reports suggesting an interaction between warfarin and celecoxib. The committee noted that an increased risk of bleeding for some patients on therapy with celecoxib and warfarin would preclude the funding of celecoxib for this subgroup on safety grounds. As with other NSAIDs, the post-marketing surveillance (PMS) data suggest that the elderly, diabetics, patients with prior renal disease, and concomitant ACE inhibitor/diuretic use are risk factors for acute renal failure. The committee noted that dyspepsia is a common side effect of celecoxib, which may lead to significant concomitant PPI use. The committee also noted that some patients would be on treatment with aspirin for coronary or cerebrovascular disease, and therefore already on treatment with PPI, which would make the benefit of using celecoxib instead of a NSAID questionable and the cost of treatment significantly greater. PTAC considered celecoxib to be extremely expensive and agreed that the additional expenditure over NSAIDs would not be justified considering the modest decrease in serious gastrointestinal complications. In addition, patients at high risk of adverse events, including gastro-intestinal ones, would still have to be considered at high risk if treated with celecoxib and that, for this reason, PTAC could not recommend targeting to any specific sub-groups of patients. The committee considered that the place in therapy and safety profile of celecoxib still needs to be fully elucidated from post-marketing experience. The committee, therefore, considered that the maximum benefit of celecoxib would only be achieved if the drug was made available to all patients requiring treatment with NSAIDs. The committee therefore supported the listing of celecoxib alongside NSAIDs and considered that this should be given a moderate priority.

Rofecoxib PTAC considered an application in November 2000 to list rofecoxib 12.5 mg and 25 mg tablets on the Pharmaceutical Schedule. Further information was considered by PTAC in November 2001. PTAC noted that rofecoxib is a selective COX-2 inhibitor that does not have a sulphonamide structure. The committee considered that the efficacy of rofecoxib was comparable with other conventional NSAIDs. The committee considered that rofecoxib had a superior gastrointestinal (GI) safety profile when compared with conventional NSAIDs. However, as with celecoxib, it considered that the true safety profile, particularly in elderly patients at high risk, would only become apparent after considerable postmarketing experience. The committee noted the new evidence relating to the potential prothrombotic effects of rofecoxib shown in the VIGOR study. The committee considered that this could have been due to a positive effect of naproxen rather than a negative effect of rofecoxib. The committee considered that concomitant use of low dose aspirin (to provide an anti-platelet effect) is likely to be routine in high risk cardiac patients and this will reduce the GI safety advantages of rofecoxib. Similarly, an increase in cardiovascular risk associated with rofecoxib in patients not on aspirin will also offset the safety advantages. The committee considered that patients with a high risk of developing upper GI perforations, ulcers and bleeding (PUBs), who are prescribed Vioxx, might still be prescribed a PPI. The committee considered that the cost offsets in terms of reduction in the use of PPIs in patients with gastrotoxicity induced by conventional NSAIDs might not necessarily be high after they switched to rofecoxib. The committee noted that despite the lack of comparative data there was good evidence suggesting that a combination of PPI with a conventional NSAID protects against further PUBs, and could be used as an alternative approach. The committee considered that patients at high risk of adverse events, including gastro-intestinal ones, would still be at risk if treated with rofecoxib and that, for this reason, it could not recommend targeting to any specific sub-groups of patients. The committee also considered the price of rofecoxib was disproportionately high compared to its potential benefits. The committee considered there to be no clinical reason not to list rofecoxib on the Pharmaceutical Schedule should an acceptable commercial arrangement be reached between the supplier and PHARMAC.

INFORMATION FOR DHB USE ONLY

Appendix 5 – Other Cost/QALYs

Table 5: Comparison of Cost/QALY with other Economic Evaluations for Hospital Pharmaceuticals

Pharmaceutical Infliximab*

Year 2002

Indication(s) Assessed Crohns disease

Zoledronic acid*

2003

Linezolid* Drotrecogin alfa (activated)* Naltrexone Venlafaxine Etanercept Rituximab

2003 2003

Hypercalcemia of malignancy (HCM); Bone metastases in breast cancer and multiple myeloma; Osteoporosis Methicillin-resistant staphylococcus aureus infections Severe sepsis

Marginal $/QALY Single dose: $53,000 Retreatment: $118,000 Maintenance: $382,000 HCM: $20,000-$40,000 Bone metastases: $800,000

Additional cost of $1,300 High risk of death: $35,000 All patients: $97,000 Low risk of death: less effective and more costly $3,000 $10,000-$23,000 (depending on response to treatment) $36,000 - $62,000 (depending on length of use of cyclosporin) $33,000 - $46,000

2003 2003 2003 2003

Alcohol addiction Refractory depression Rheumatoid arthritis CD20 positive diffuse large B-cell NHL (stage II and above) in combination with CHOP (cyclophosphamide, vincristine, doxorubicin and prednisone) chemotherapy Last-line treatment for neuropathic pain Blastic and accelerated phases of Chronic Myeloid Leukemia (CML); Chronic phase after failure of interferon Schizophrenia Venous thromboembolism Renal transplantation

Gabapentin* Imatinib mesylate*#

2002 2002

$3,000-$5,000 Blastic phase: $35,000-$51,000 Accelerated phase: $100,000 Chronic phase: $110,000

Quetiapine* Enoxaparin Tacrolimus*

$20,000 Cost-saving Renal Rescue: $2,500 Primary: $3,000-$22,000 (depending on dose) * Analyses available for DHBs to download from HPAD: www.pharmac.govt.nz/hpad/ # The decision by PHARMAC to fund imatinib mesylate (Glivec) was not based on its cost-effectiveness alone. Rather, PHARMAC was able to negotiate cost-savings from a bundle of pharmaceuticals that made the funding of Glivec relatively cost-neutral.

2001 2001 1998

INFORMATION FOR DHB USE ONLY

Metadata

Title

Hospital Pharmaceutical Assessment Final Summary Discussion Document No 6

Abstract

Celecoxib (Celebrex) and rofecoxib (Vioxx) - selective COX-2 inhibitors - for osteoarthritis and rheumatoid arthritis.

Page 1

Note

This text has been extracted from the source PDF document.

Also available as plain text.

Please contact webmaster to discuss alternative format options.