Safety and efficacy of fostamatinib in rheumatoid arthritis patients with an inadequate response to methotrexate in phase II OSKIRA-ASIA-1 and OSKIRA-ASIA-1X study.

OBJECTIVE: The primary objectives of two phase II studies of fostamatinib were to evaluate efficacy (OSKIRA-Asia-1: NCT01569074) and long-term safety/tolerability (OSKIRA-Asia-1X: NCT01640054) in patients from Asia with active RA despite MTX treatment. METHODS: OSKIRA-Asia-1 was a 12-week, multicentre, double-blind, placebo-controlled, parallel-group study. Patients were randomized to receive one of four fostamatinib doses (groups A-D; n = 31, 33, 33, 33) or placebo (group E; n = 33). OSKIRA-Asia-1X was a long-term extension study (100 mg fostamatinib qd) of patients who completed OSKIRA-Asia-1. RA signs and symptoms were measured by ACR response criteria and DAS based on a 28-joint count. Physical function status was assessed with the HAQ-Disability Index. Safety findings were monitored. RESULTS: In OSKIRA-Asia-1, fostamatinib revealed numerical improvements in ACR 20% response (ACR20) at week 12 in group A (100 mg bid) and group B (100 mg bid, then 150 mg qd) vs placebo. Statistically significant improvements in ACR20 and ACR50 at week 8 and in ACR70 at week 12, and statistically significant achievement in low disease activity (defined as DAS based on a 28-joint count ≤3.2 based on C-reactive protein) occurred in groups A and B. Improvement in physical function was numerically higher in group A. The most common adverse events were hypertension, diarrhoea and neutropenia. In OSKIRA-Asia-1X, the most common adverse events were nasopharyngitis, hypertension, RA and neutropenia. CONCLUSION: Fostamatinib achieved both statistically and clinically significant improvements in RA signs and symptoms. The safety and tolerability of fostamatinib (plus MTX) were consistent with previous studies. TRIAL REGISTRATION: OSKIRA-Asia-1 trial registration: https://clinicaltrials.gov, NCT01569074; OSKIRA-Asia-1X trial registration: https://clinicaltrials.gov, NCT01640054.

Effects of Fostamatinib on the Pharmacokinetics of the CYP2C8 Substrate Pioglitazone: Results From In Vitro and Phase 1 Clinical Studies.

Fostamatinib is a prodrug that undergoes gastrointestinal tract dephosphorylation to form the active metabolite, R406. Here we report its cytochrome P450-inducing potential. In vitro, R406 3 and 10 µM induced CYP2C8 to levels representing 53% and 75%, respectively, of the level achieved by the positive control, rifampicin. Induction of other enzymes was minor. The effect of fostamatinib (100 mg twice daily) on the pharmacokinetics of a single oral 30-mg dose of the CYP2C8 substrate pioglitazone and its metabolite, hydroxy pioglitazone, was then investigated (open-label, nonrandomized, 2-period phase I study [n = 15]). Coadministration of fostamatinib and pioglitazone (vs pioglitazone alone) was associated with lower mean maximum plasma concentration values for pioglitazone (geometric least-squares mean ratio, 82.8; 90% confidence interval, 64.2-106.8) and hydroxy pioglitazone (90.9; 78.6-105.1), an increase in pioglitazone AUC (117.8; 108.4-128.0), a decrease in hydroxy pioglitazone AUC(0-t) (89.7; 78.9-101.9), and an increase in pioglitazone geometric mean t1/2λz (9.4-12.8 hours). No tolerability concerns were identified upon coadministration. These data suggest that although clinical significance has not been formally evaluated, fostamatinib is unlikely to have a clinically significant effect on the pharmacokinetics of pioglitazone (which may be extrapolated to other CYP2C8 substrates). However, vigilance is advised should these agents be prescribed together.

Effects of CYP3A4 Inhibitors Ketoconazole and Verapamil and the CYP3A4 Inducer Rifampicin on the Pharmacokinetic Parameters of Fostamatinib: Results from In Vitro and Phase I Clinical Studies.

BACKGROUND: Fostamatinib (R788) is a spleen tyrosine kinase (SYK) inhibitor. The active metabolite of fostamatinib, R406, is primarily metabolized by CYP3A4. OBJECTIVES: The aim of this study was to characterize hepatic microsomal metabolism of R406 and confirm the role of CYP3A4 in R406 metabolism, determining whether co-administration of CYP3A4 inhibitors (ketoconazole, verapamil) or inducers (rifampicin) affects R406 pharmacokinetics. METHODS: R406 stability was determined using human hepatic microsomes. The CYP450 isoforms responsible for R406 metabolism in humans were identified using expressed CYP450 isoforms and specific chemical inhibitors. The ketoconazole interaction study (double-blind, randomized, placebo-controlled, two-period crossover) involved fostamatinib administration (single 80-mg dose), alone and with ketoconazole (200 mg twice daily). The verapamil and rifampicin interaction studies (open-label, two-period, fixed-sequence) involved fostamatinib administration (single 150-mg dose), alone and with immediate-release verapamil (80 mg three times daily) or rifampicin (600 mg once daily). Standard pharmacokinetic parameters were calculated in all studies. RESULTS/DISCUSSION: Hepatic microsomes showed time-dependent loss of R406 and formation of para-O-demethylated R406. Microsomal metabolism of R406 was markedly inhibited by CYP3A4 inhibitors and, in the expressed CYP450 studies, the rate of R406 disappearance was greatest with CYP3A4. In the clinical studies, co-administration of ketoconazole caused a 2-fold (CI 1.77-2.30) increase in R406 exposure. Verapamil increased R406 exposure (39% increase, CI 8-80), whereas rifampicin co-administration decreased exposure by 75% (CI 68-81). Fostamatinib was well tolerated. CONCLUSION: The oxidative metabolism of R406 is predominantly catalyzed by CYP3A4. In clinical studies, exposure to R406 is affected by concomitant administration of CYP3A4 inducers/inhibitors. These findings should be taken into account when considering co-prescription of fostamatinib with such agents.

Pharmacokinetic Properties of Fostamatinib in Patients With Renal or Hepatic Impairment: Results From 2 Phase I Clinical Studies.

PURPOSE: Phase III trials of fostamatinib, an oral spleen tyrosine kinase inhibitor, in the treatment of rheumatoid arthritis have been completed. Herein, we report the effects of renal and hepatic impairment on the pharmacokinetic (PK) properties of the active metabolite of fostamatinib, R406, in plasma, and on the urinary excretion of R406 and its metabolite N-glucuronide. METHODS: Two Phase I, single-center, open-label clinical trials determined the PK properties and tolerability of fostamatinib in subjects with normal or impaired renal or hepatic function. Twenty-four subjects in the study in renal impairment (8 per group: normal renal function, moderate renal dysfunction, or end-stage renal disease [ESRD]), and 32 subjects in the study in hepatic impairment (8 per group: normal hepatic function or mild, moderate, or severe hepatic impairment) received a single 150-mg dose of fostamatinib. Patients with ESRD in the study in renal impairment participated in 2 treatment periods separated by a ≥1-week washout. In these patients, fostamatinib was administered after dialysis or 2 hours before dialysis. FINDINGS: Geometric mean R406 Cmax and AUC values were less in the combined renally impaired group than in the group with normal renal function; Tmax was similar across groups. However, renal impairment had no apparent effect considered clinically relevant on unbound R406. In patients with ESRD, R406 exposure was less when fostamatinib was administered after compared with before dialysis. Urinary excretion of R406 N-glucuronide was decreased with increasing severity of renal impairment. Renal elimination of R406 was negligible in all groups. Varying degrees of hepatic impairment had no consistent effects on the PK properties of R406. R406 Cmax values were 10% to 15% less in all hepatically impaired groups than in the group with normal hepatic function. AUC and Tmax values were similar between the groups with normal and severely impaired hepatic function; in the groups with mild or moderate hepatic impairment, AUC was less and Tmax was greater. The geometric mean percentage of unbound R406 ranged from 0.64% to 1.95% and was greatest in the group with severe hepatic impairment. The urinary excretion of R406 was minimal. The amount of R406 N-glucuronide excreted in urine was greater in severely hepatically impaired patients. Fostamatinib 150 mg was generally well tolerated. IMPLICATIONS: In these patients, renal or hepatic impairment did not affect exposure to the active metabolite of fostamatinib, R406, to a clinically relevant extent. ClinicalTrials.gov identifiers: NCT01245790 (renal) and NCT01222455 (hepatic).

Effects of Fostamatinib on the Pharmacokinetics of Digoxin (a P-Glycoprotein Substrate): Results From in Vitro and Phase I Clinical Studies.

PURPOSE: Fostamatinib, a spleen tyrosine kinase inhibitor and prodrug of the active metabolite R406, is being developed as an anti-inflammatory drug for several indications for which polypharmacy is likely. Digoxin, indicated for congestive cardiac failure, may be used for certain supraventricular dysrhythmias. The studies reported herein examined whether fostamatinib and R406 are inhibitors of P-glycoprotein (P-gp) in vitro and evaluated the effect of fostamatinib on the pharmacokinetic parameters of digoxin to understand drug-drug interaction (DDI) potential in the clinic. METHODS: Inhibition of P-gp-mediated digoxin transport by fostamatinib and R406 was determined across Caco-2 cell monolayers. Apparent permeability of digoxin was determined and used to calculate efflux ratios and percentage inhibition. Half maximal inhibitory concentrations (IC50) and theoretical gastrointestinal concentration [I2] (dose in moles per 250 mL) were calculated to gauge clinical DDI potential. In a subsequent Phase I study, the plasma concentration-time profiles and resulting pharmacokinetic parameters were examined across 2 treatment periods: (1) oral digoxin loading dose of 0.25 mg BID on day 1 and 0.25 mg once daily on days 2 to 8, and (2) oral digoxin 0.25 mg once daily and oral fostamatinib 100 mg BID on days 9 to 15. FINDINGS: Fostamatinib (but not R406) was determined to be a P-gp inhibitor in vitro (IC50 = 3.2 µM). On the basis of a theoretical gastrointestinal concentration (I2)/IC50 ratio of 216 ([I2] = 691 µM), predictions indicated the potential for absorption-based DDI in vivo through inhibition of intestinal P-gp. In the clinical study, when digoxin was co-administered with fostamatinib, digoxin levels were higher before dosing and throughout the dosing interval, and an increase in exposure to digoxin was observed. Co-administration led to a 1.70-fold increase in digoxin maximum plasma concentration at steady state (Cmax,ss) versus digoxin administration alone (2.18 vs 1.32 ng/mL). Median digoxin time of Cmax was earlier when digoxin was co-administered with fostamatinib (1.00 vs 1.48 hours). The digoxin AUC during the dosing interval at steady state was increased 1.37-fold with co-administration. No severe or serious adverse events or deaths were reported. IMPLICATIONS: Fostamatinib was confirmed to be a P-gp inhibitor in vitro and in vivo, and a DDI with digoxin was apparent. Co-administration of digoxin and fostamatinib was generally well tolerated. However, continued review of digoxin response and dose is advisable should these agents be prescribed concomitantly. ClinicalTrials.gov identifier: NCT01355354.

Effects of fostamatinib, an oral spleen tyrosine kinase inhibitor, in rheumatoid arthritis patients with an inadequate response to methotrexate: results from a phase III, multicenter, randomized, double-blind, placebo-controlled, parallel-group study.

OBJECTIVE: This phase III, 52-week study compared fostamatinib with placebo (for 24 weeks) in patients with active rheumatoid arthritis (RA) and an inadequate response to methotrexate (MTX) therapy. METHODS: Patients taking MTX were randomized (1:1:1) to receive fostamatinib 100 mg twice daily for 52 weeks (group A), fostamatinib 100 mg twice daily for 4 weeks and then 150 mg once daily (group B), or placebo for 24 weeks and then fostamatinib 100 mg twice daily (group C). At week 24, the co-primary end points were change from baseline in the American College of Rheumatology 20% (ACR20) improvement response rates and change in the modified total Sharp/van der Heijde score of radiographic damage (SHS). RESULTS: In this study, 918 patients were randomized and received ≥1 dose of study drug (fostamatinib or placebo); the demographic and baseline clinical characteristics were well balanced. Following treatment with both fostamatinib regimens, a statistically significant difference in the ACR20 improvement response was achieved at week 24 as compared with that in patients receiving placebo (49.0% [group A] and 44.4%, [group B] versus 34.2%; P < 0.001 and P = 0.006, respectively), but there was no statistically significant difference in the SHS between either fostamatinib group and placebo (P = 0.25 and P = 0.17, respectively). The most common adverse events in patients in groups A, B, and C were hypertension (15.8%, 15.1%, and 3.9%, respectively) and diarrhea (13.9%, 15.1%, and 3.9%, respectively). Elevated blood pressure (≥140/90 mm Hg) occurred at ≥1 visit in 44.2%, 41.6%, and 19.3% of patients in each respective group. CONCLUSION: With the use of either fostamatinib regimen in patients with RA, statistically significant, but not clinically significant, improvements in the ACR20 improvement response over placebo were achieved at 24 weeks, whereas a significant difference in the SHS was not seen. The overall level of response to treatment with fostamatinib was lower than had been observed in the phase II program, but similar adverse events were reported.

Substance misuse and psychiatric illness: prospective observational study using the general practice research database.

OBJECTIVES: To quantify the relation between substance misuse and psychiatric illness in the UK general practice population in terms of (a) the relative risk of developing one condition given prior exposure to the other and (b) the proportion of cases of one condition attributable to exposure to the other. DESIGN: Population based prospective observational study using the general practice research database (GPRD) between 1993 and 1998. The 230 GP practices represent 3.1% of the population. SETTING: England and Wales. PARTICIPANTS: 1.4 million registered patients of whom 3969 had both substance misuse and psychiatric diagnoses between 1993 and 1998. MAIN OUTCOME MEASURES: Relative risk (RR) for subsequent psychiatric illness among participants exposed to substance misuse and RR for subsequent substance misuse among participants exposed to psychiatric illness. Population attributable risk (PAR) of psychiatric illness attributable to substance misuse and of substance misuse attributable to psychiatric illness. RESULTS: The baseline prevalence of psychiatric illness over the study period was 15% and 0.3% for substance abuse. RR for psychiatric illness for substance misusers compared with non-substance misusers was 1.54 (95% CI 1.47 to 1.62). RR for substance misuse among psychiatric compared with non-psychiatric cases was 2.09 (95% CI 1.99 to 2.22). PAR for psychiatric illness attributable to substance misuse was 0.2%. PAR for substance misuse attributable to psychiatric illness was 14.2%. DISCUSSION: Only a comparatively small proportion of psychiatric illness seems possibly attributable to substance use whereas a more substantial proportion of substance use seems possibly attributable to psychiatric illness. This study does not support the hypotheses that comorbidity between substance misuse and psychiatric illness is primarily the result of substance misuse or that increasing comorbidity is largely attributable to increasing substance misuse.

Psychological, sleep, lifestyle, and comorbid associations with headache.

OBJECTIVE: To investigate the associations of headache occurrence, severity, and frequency with psychological, sleep, and lifestyle characteristics, and comorbid conditions. BACKGROUND: Whilst associations for individual headache types, particularly migraine, have been investigated, possible associations between headache of all types and general health characteristics have not been explored. METHODS: Cross-sectional postal survey in an adult general population sample registered at five general practices in North Staffordshire, UK. RESULTS: Headache occurrence was associated with anxiety (odds ratio 4.09, 95% confidence interval 3.0, 5.6) and sleep problems (moderate sleep problems OR 3.60, 95% CI 2.5, 5.0), and the strength of the associations increased with higher levels of anxiety and sleep problem. Whilst depression showed an association with headache, this was not seen in respondents with depression in the absence of anxiety. Headache occurrence was also associated with comorbid pain anywhere in the body (OR 2.12, 95% CI 1.7, 2.6), with the strongest associations being for the neck and upper body areas compared with other areas. There was no overall link with alcohol or caffeine consumption. Increasing severity and frequency of headache resulted in stronger associations, and there were strong associations between the occurrence, severity, and frequency of headache, and both sleep problems and psychological distress. CONCLUSION: Poor sleep and anxiety appear to make a substantial contribution to the impact of headache on sufferers' lives in the general population.

One-year follow-up of headache in an adult general population.

OBJECTIVE: To investigate variation in headache occurrence and characteristics over 1 year. BACKGROUND: Headache is a common condition which can affect the work, home, and social lives of sufferers, yet surprisingly little is known about how headache changes over time. METHODS: Postal survey to a random general population sample of 5000 adults aged 18 years plus, with follow-up survey to all baseline responders at 1 year and a subsample of 500 being surveyed at 3-monthly intervals between the baseline and 1-year surveys. RESULTS: A total of 1589 (74% response) responded to the 1-year follow-up and 282 of the subsample responded to all five surveys at 3-monthly intervals. Among 1-year respondents with recent headache at baseline (defined as occurring during the previous 3 months), nearly all (94%) also reported headache during the follow-up year. One-third of respondents without recent headache at baseline reported a new episode of headache during the follow-up year. Most (85%) respondents with recent headache at both baseline and 1-year follow-up reported a variation in at least one headache characteristic. These findings were replicated in the sample completing the 3-monthly surveys. Although most of this subgroup reported their headache occurrence status was unchanged during each 3-month period, only a few (3%) respondents with headache in each period reported no variation at all in headache characteristics during the study. CONCLUSIONS: While prevalence of recent headache was stable over time for individuals, there was considerable variation in headache characteristics.

A method to determine if consenters to population surveys are representative of the target study population.

BACKGROUND: Searching medical records of study non-responders to investigate selection bias is no longer acceptable. We explore an alternative by comparing consultation rates in survey responders who consented to medical record review, with anonymized consultation rates for the total practice populations. METHODS: Anonymized aggregated consultation rates for the year following a population-based survey were calculated for headache and a number of other conditions (chosen to reflect a mixture of chronic and episodic conditions). These rates were compared across two groups of adults: (i) responders to the survey who consented to medical record review and (ii) a 'population group' created from records of the general practices participating in the survey to represent all patients aged 18 years and over at the mid-point in the study year. The consultation rates for the conditions were compared across the two groups using direct standardization. RESULTS: Adjusted consultation rates were similar but generally higher in the responders. CONCLUSIONS: This alternative method applied here offers one potential approach to determine whether study respondents are representative of the population from which they were sampled with respect to general practice consultations.

What has happened to antimicrobial usage in primary care in the United Kingdom since the SMAC report? - description of trends in antimicrobial usage using the General Practice Research Database.

BACKGROUND: The aim of the study was to assess antibiotic prescribing within the United Kingdom for three of the Standing Medical Advisory Committee recommendations 'four things which could be done'. METHODS: We conducted a retrospective survey of morbidity and antibiotic prescribing data between 1993 and 2001 using the national General Practice Research Database. Antibiotic prescribing was linked to diagnoses of cough/cold and sore throat; length of antibiotic course for uncomplicated cystitis. RESULTS: The rate of antibiotic prescribing for cough/cold declined between 1993 (43.7 per 1000 patient years at risk) and 1999 (23.5 per 1000 patient years at risk) and has since increased slightly (to 30.5 per 1000 patient years at risk in 2001). Antibiotic prescribing for sore throat declined between 1995 (80.6 per 1000 patient years at risk) and 1999 and has since remained static (42.1 per 1000 patient years at risk in 2001). Trimethoprim was the most commonly used antibiotic for episodes of uncomplicated cystitis and the prescription of 3 day (or less) courses has increased from 16.4 per cent in 1998 to 41.5 per cent in 2001. CONCLUSIONS: For the SMAC recommendation to limit prescribing for uncomplicated cystitis to 3 days in otherwise fit women there has been demonstrable impact since the publication of the SMAC report. For two recommendations (no prescribing of antibiotics for simple coughs and colds; no prescribing of antibiotics for viral sore throats) the impact has been less clear against the background of a general reduction in antimicrobial prescribing.

Prevalence of comorbid psychiatric illness and substance misuse in primary care in England and Wales.

STUDY OBJECTIVE: To estimate the annual period prevalence of co-occurring psychiatric illness and substance misuse among patients in primary care. DESIGN: Analysis of the general practice research database. SETTING: England and Wales, 1993-1998. PARTICIPANTS: Registered patients at 230 general practices representing 3.1% of the population. A comorbid case was defined as one with both a psychiatric diagnosis and substance misuse diagnosis (not including alcohol or tobacco) within a calendar year. A potentially chronic comorbid case was one that met this definition and, in addition, was treated in subsequent years for either a psychiatric condition or substance misuse. MAIN RESULTS: The annual period prevalence of comorbidity increased from 50/100 000 patient years of exposure (PYE) to 80/100 000 PYE, an increase of 62% during the study period. Rates of comorbid psychoses, comorbid schizophrenia, and comorbid paranoia increased by 147%, 128%, and 144%. The average age of comorbid cases decreased from 38 years to 34 years. Over 80% of comorbid cases were newly diagnosed in each study year, although many are treated in subsequent years for either psychiatric illness or substance misuse. CONCLUSIONS: This study provides data on the nature and extent of comorbidity in primary care in England and Wales. As the comorbidity rate is increasing by about 10% each year, and as comorbid cases are becoming younger, it is probable that the comorbidity rate will have increased beyond the study end point.

Naratriptan for the treatment of acute migraine: meta-analysis of randomised controlled trials.

OBJECTIVE: To evaluate the comparative efficacy and tolerability of naratriptan in the treatment of acute attacks of migraine. DESIGN: Meta-analysis of randomised controlled trials using a random effects model. SUBJECTS: A total of 4499 patients suffering from moderate or severe attacks of acute migraine reported in ten trials. MAIN OUTCOME MEASURES: Response rate ratios for headache relief, pain-free response and sustained relief (4-24 hours). Adverse events were estimated with the rate ratio (RR), risk difference and number needed to harm. RESULTS: Pooled RRs relative to placebo for pain-free response at 2 and 4 hours for naratriptan 2.5 mg were 2.52 (95% CI: 1.78-3.57) and 2.58 (1.99-3.35). Naratriptan 2.5 mg was more effective than naratriptan 1 mg; the corresponding RRs for pain-free response at 2 and 4 hours were 1.54 (95% CI: 1.28-1.86) and 1.35 (1.20-1.51). In contrast, naratriptan 2.5 mg was less effective in pain-free response than either rizatriptan 10 mg at 4 hours (RR: 0.68; 95% CI: 0.55-0.85) or sumatriptan 100 mg at 4 hours (RR: 0.79; 95% CI: 0.67-0.93). However, significantly fewer patients experienced adverse effects with naratriptan 2.5 mg than with rizatriptan 10 mg (RR: 0.73; 95% CI: 0.56-0.97) or sumatriptan 100 mg (RR: 0.68; 95% CI: 0.55-0.86). CONCLUSIONS: Naratriptan is an effective and well-tolerated treatment for acute attacks of migraine. Head-to-head comparisons suggest that naratriptan 2.5 mg is significantly more effective than the 1 mg dose. Rizatriptan 10 mg and sumatriptan 100 mg were superior to naratriptan in terms of headache relief, while zolmitriptan 2.5 mg seemed to have comparable efficacy. Randomised controlled trials have shown that at licensed doses (1 and 2.5 mg), naratriptan is associated with a lower incidence of adverse effects than rizatriptan, sumatriptan and zolmitriptan. The incidence rates of adverse effects were similar to placebo.

Safety profile of the triptans.

The triptans are 5-HT(1B/1D) agonists used as migraine and cluster-specific agents. Seven are in commercial use worldwide; in order of release these are sumatriptan, zolmitriptan, rizatriptan, naratriptan, almotriptan, frovatriptan and eletriptan. Sumatriptan has been in clinical use since 1991, and although postmarketing studies have stimulated much debate of triptan strengths and weaknesses, their overall safety profile appears excellent. The most serious adverse events are cardiovascular, due to coronary artery narrowing as a consequence of coronary artery 5-HT(1B) receptor activity. Triptans are contraindicated in patients with vascular disease. Other events are even more rare, and include the potential for drug-drug interactions, based on metabolic elimination pathways. Serotonin syndrome has been a concern, but one large prospective study failed to document a single case, and reports are sporadic and not clearly causative.

Cost-effectiveness analysis of stratified versus stepped care strategies for acute treatment of migraine: The Disability in Strategies for Care (DISC) Study.

BACKGROUND: The Disability in Strategies for Care (DISC) study was the first large randomised controlled trial to compare alternative treatment strategies in the acute treatment of migraine. With 835 patients in its intention-to-treat efficacy analysis, DISC compared a stratified care strategy, where initial therapy was based on clinical need as determined by the Migraine Disability Assessment Scale (MIDAS) and two stepped care strategies (across attacks and within attacks), where first-line therapy with a simple combination analgesic was escalated, if response had been inadequate, to zolmitriptan, a migraine-specific therapy. OBJECTIVE: To report on the cost effectiveness of these three strategies from a societal perspective. STUDY DESIGN AND METHODS: A cost-effectiveness analysis was undertaken using data from the DISC study, and including both health service and productivity costs. Data were collected prospectively on drug usage (main therapy and rescue medication); resource use associated with adverse events was estimated by a clinician blinded to treatment strategy. Health service resource use was costed using UK unit costs (1999 to 2000 values). Data were collected using diary cards on the amount of time patients lost from work, and on reduced effectiveness at work, due to a migraine attack. This facilitated an estimate of the productivity costs associated with the treatment strategies. To assess cost effectiveness, the differences in costs between the strategies were related to the two primary outcome measures in the trial: headache response 2 hours after initial therapy and disability-adjusted time during the first 4 hours after initial therapy. RESULTS: Although the mean health service cost was higher in the stratified care group (mean over 6 attacks of pound 28.25 versus pound 11.74 and pound 23.15 in the stepped care across attacks group and within attacks group, respectively), mean productivity costs over 6 attacks were lower in the stratified group (pound 112.22 versus pound 144.70 and pound 127.53). The total mean cost over six attacks was, therefore, lowest in the stratified care group (pound 138.95 compared with pound 157.19 in the stepped care across attacks group and pound 148.53 in the stepped care within attacks group), although these differences did not reach statistical significance. In terms of headache response, stratified care was statistically significantly more effective than both forms of stepped care. Using disability-adjusted time, stratified care was statistically significantly more effective than stepped care across attacks, but not against stepped care within attacks. CONCLUSION: Given its lower mean costs and higher mean effectiveness, a stratified care strategy, which included zolmitriptan, was the dominant strategy and was unequivocally more cost effective from a societal perspective than either stepped care strategy. When the uncertainty around these means was considered, stratified care had the highest probability of being cost effective.