Long-term treatment of gout involves taking urate-lowering therapy (ULT), most commonly with allopurinol or febuxostat. National and international guidelines recommend starting ULT at low dose and increasing gradually over a number of months to lower the serum urate below a therapeutic biochemical target level (treat-to-target) [1–4]. New crystal formation is therefore prevented and existing crystals dissolve gradually so that flares cease, tophi reduce in size, and long-term joint damage is prevented, effectively ‘curing’ gout [5]. One barrier to successful ULT is the tendency for flares to increase in frequency after initiation of ULT or dose increases. As well as causing significant pain and disability, such ULT-initiated flares may be perceived by people with gout and practitioners as exacerbating the severity of gout, leading to ULT being stopped [6]. Clinical guidelines therefore recommend that colchicine, non-steroidal anti-inflammatory drugs (NSAIDs), or corticosteroids can be co-prescribed when initiating ULT in order to prevent ULT-induced flares [1–4].

Colchicine prophylaxis is undoubtedly effective. Randomised trials demonstrate that co-prescribing colchicine prophylaxis when initiating ULT reduces the incidence of ULT-induced flares [7]. One small randomised trial (43 participants) undertaken in people initiating allopurinol found that 33% of those randomised to colchicine 0.5 mg twice daily had a flare within the 6-month follow-up period compared with 77% of those receiving placebo [8]. In a more recent larger randomised trial (200 participants), placebo was not non-inferior to colchicine 0.5 mg daily for mean number of flares per month (0.61 vs. 0.35) when initiating ‘start low, go slow’ treat-to-target allopurinol [9]. The Febuxostat Outcome Research Towards Urate Lowering in the Next Decade-1 (FORTUNE-1) trial randomised 241 adults with gout to febuxostat step-wise dose increase from 10 mg to 40 mg daily only, febuxostat fixed dose 40 mg daily plus colchicine prophylaxis, and febuxostat fixed dose 40 mg daily only, finding that fewer people experienced flares over 12 weeks with febuxostat fixed dose plus colchicine 0.5 mg daily (19%) than with febuxostat fixed dose alone (36%) [10]. There is limited evidence of the effectiveness of prophylactic NSAIDs or glucocorticoids for flare prophylaxis, other than for azapropazone, an NSAID no longer widely used in clinical practice [11].

Colchicine is inexpensive in many countries including the UK and Australia, however, a recent increase in the cost of colchicine in the USA led to a reduction in colchicine initiations and an increase in costs to people with gout [12]. A subsequent 2-arm decision tree cost-effectiveness analysis modelled from both US and Australian inputs found that colchicine was very cost-effective in preventing gout flares when initiating allopurinol compared with placebo, particularly in an Australian setting where colchicine was considerably cheaper [13]. Twice daily colchicine prophylaxis resulted in a cost of $1,276 and 0.49 quality-adjusted life-years (QALYs) compared with $516 and QALY 0.47 for placebo, equating to an incremental cost-effectiveness ratio of $34,004 per QALY gained.

Cardiovascular disease is prevalent in people with gout. In one primary care-based cross-sectional study of 1,079 people with gout, the prevalences of obesity, hypertension, hyperlipidaemia, diabetes mellitus, and chronic kidney disease (CKD) were 35%, 40%, 10%, 16%, and 25% respectively, with 15% of participants having coronary heart disease [14]. There has therefore been great interest recently in the possible cardiovascular benefits of colchicine prophylaxis [15], following large randomised trials that have shown cardioprotective effects of colchicine in people with chronic coronary heart disease or a recent myocardial infarction (MI) [16, 17]. However, to date, there has not been a randomised trial of cardioprotective effects of colchicine in people with gout. A systematic review and meta-analysis of four observational studies including 10,026 participants with gout found that colchicine was associated with a 29% lower risk of MI, although this was not statistically significant (risk ratio 0.71, 95% confidence interval 0.36–1.39), highlighting the need for further robust evidence [18]. The duration of colchicine treatment in the cardiovascular disease trials was often longer than is used for gout flare prophylaxis in clinical practice and since the participants had a history of chronic coronary heart disease or a recent MI, they are likely to have been at higher risk of cardiovascular events than people with gout, many of whom have cardiovascular risk factors but not necessarily established cardiovascular disease. It is therefore unclear whether the potential cardiovascular benefits of colchicine apply to the way it is currently used for gout flare prophylaxis in practice.

Given the clear evidence of effectiveness of colchicine prophylaxis, should all people with gout initiating ULT receive anti-inflammatory prophylaxis? Uptake of prophylaxis appears to be variable and dependent on the clinical setting. Prophylaxis is commonly co-prescribed when ULT is initiated by rheumatologists in the UK (94% of initiations), France (72%), Greece (41%), and by physicians in the USA (90%) [19–21], in contrast to only 25% and 7% of primary care initiations in the UK and Australia respectively [22, 23]. Furthermore, in a large UK primary care-based randomised trial which compared nurse-led treat-to-target ULT with usual primary care for gout, only 1% of participants in the intervention arm opted for prophylaxis after being offered it when initiating ULT [5]. Although the proportion of participants having a flare during the first year of follow-up was greater in the intervention arm (54.0% vs. 39.8%), this did not prevent excellent adherence to ULT, the vast majority of participants achieving target serum urate levels, or significant clinical improvements in flare rates, tophi, and quality of life at 2 years. The clinical severity of gout and risk of ULT-induced flares are likely to be lower in primary care populations, potentially explaining discrepancies between primary and secondary care prescribing and suggesting that good outcomes can be achieved from ULT without prophylaxis in certain patients. There have been few studies comparing the characteristics of people who receive prophylaxis with those who do not. In a cross-sectional analysis from the secondary care-based Consortium of Rheumatology Researchers of North America (CORRONA) gout registry, people taking prophylaxis had shorter gout duration, more flares, greater healthcare use in the preceding year, greater disease activity and pain, and were more likely to have tophi than those not receiving prophylaxis [24].

National and international clinical guidelines make varying recommendations regarding how strongly they recommend anti-inflammatory prophylaxis. The 2012 American College of Rheumatology gout guideline and its 2020 update strongly recommended co-prescribing colchicine, NSAID, or prednisolone/prednisone prophylaxis when initiating ULT [1, 25]. Other guideline development groups have made less stringent recommendations concerning prophylaxis in their recent guideline updates than in earlier versions and highlighted that the patient perspective is important and that the pros and cons of prophylaxis should be discussed with people with gout initiating ULT. Whereas the first British Society for Rheumatology gout guideline recommended in 2007 that colchicine (or an NSAID or coxib if colchicine is not tolerated) should be co-prescribed when initiating ULT [26], the 2017 update advocated less strongly that it can be considered [3]. Both the 2016 European Alliance of Associations for Rheumatology (EULAR) and 2022 UK National Institute for Health and Care Excellence (NICE) guidelines recommend discussing prophylaxis with people starting ULT [2, 4]. Whilst recommending colchicine prophylaxis (or NSAID if colchicine is not tolerated) for the first 6 months of ULT, the 2016 EULAR recommendations state that prophylaxis against flares should be fully explained and discussed with the patient [2]. The 2022 NICE guideline advises discussing with the person the benefits and risks of taking medicines (colchicine, NSAID, or low-dose corticosteroid) to prevent gout flares when starting or titrating ULT [4].

There are multiple reasons why people with gout and clinicians may be reluctant to use anti-inflammatory prophylaxis (Table 1). Harm from prophylaxis is likely to be an important consideration. Comorbidities are common in people with gout [14] and the presence of comorbidities, particularly CKD, is a contraindication to NSAIDs and is thought to increase the risk of colchicine toxicity. Gastrointestinal symptoms, particularly diarrhoea, are the most common side-effects of colchicine. In a systematic review of 35 randomised trials of colchicine in the treatment of 18 different conditions, diarrhoea affected 17.9% of participants in the colchicine arms [27]. However, only five trials were undertaken in people with gout, two of which were trials of short courses of colchicine to treat flares rather than longer courses for prophylaxis. In the aforementioned small randomised trial of colchicine prophylaxis by Borstad et al. [8] the incidence of diarrhoea over the 6-month follow-up was 38% in the colchicine arm period compared with 4.5% in the placebo arm. The potential for more serious haematological, neuropathic, or myopathic side-effects, particularly when co-administered with CYP3A4 inhibitors, is a commonly cited concern but the systematic review by Stewart et al. [27] found these to occur infrequently in clinical trials. A more recent large cohort study undertaken in the UK Clinical Practice Research Datalink found that bone marrow suppression, neuropathy, and myalgia occurred more commonly in 13,945 people initiating allopurinol with colchicine than in 13,945 matched individuals initiating without prophylaxis, but absolute incidence rates were uncommon (incidence for each in those receiving colchicine < 120/10,000 person-years) [28]. There were no occurrences of rhabdomyolysis and no consistent associations between adverse events and use of other medications or CKD [29]. In a parallel cohort study, acute kidney injury (incidence 161/10,000 person-years), angina (467/10,000 person-years), MI (157/10,000 person-years), and peptic ulcer disease (82/10,000 person-years) were more common in 25,980 people initiating allopurinol with NSAID prophylaxis than in 25,980 individuals initiating without [28].

Patient and clinician perceptions of a person’s individual risk of ULT-induced flare are also likely to be a key factor determining use of prophylaxis. A large secondary care cohort study of 8,828 people with gout initiating ULT found that 68% had at least one flare within 12 months of initiating ULT [30]. Colchicine, NSAIDs, and corticosteroids were taken by more than 80% of participants at baseline and during the 12-month follow-up period, although the duration of treatment and whether this was to treat a flare or for prophylaxis was not stated. In the placebo arms of two randomised trials of colchicine prophylaxis by Borstad et al. [8] and Stamp et al. [31] 77% and 73% of participants respectively had a flare within six months of initiating allopurinol suggesting that approximately one-quarter of people initiating ULT without prophylaxis in secondary care do not appear to flare in the first six months following ULT initiation and arguably do not require prophylaxis. Furthermore, whereas the risk of ULT-induced flares is high when initiating ULT at historical fixed doses (e.g., allopurinol 300 mg daily), contemporary practice of starting at low dose and increasing the dose gradually leads to fewer flares [7, 10]. There have been few studies of other risk factors for ULT-induced flares. Secondary care-based cohort studies and an analysis nested in the aforementioned non-inferiority trial by Stamp et al. [9] have suggested that older age, having a high serum urate or C-reactive protein level, more comorbidities, having a flare in the month preceding ULT initiation, higher allopurinol starting dose, taking diuretics, and not taking prophylaxis were associated with greater risk of ULT-induced flares, but risk factors investigated did not include the presence of tophi, which might also be expected to influence likelihood of a ULT-induced flare [30–32]. Only one study investigated whether prior flare frequency increased the risk of a ULT-induced flare, finding that people having at least two flares in the 6 months prior to ULT initiation were no more likely to experience a ULT-induced flare than those having only one flare in the prior 6 months [31]. Identifying clinical characteristics that influence an individual’s risk of ULT-induced flares may allow use of prophylaxis to be tailored and personalised [7].

Whilst recent guideline recommendations suggest that the perspectives of people with gout about prophylaxis should inform treatment decisions, little is known about what the perspectives of patients and prescribers are and what influences uptake of prophylaxis. Qualitative research studies highlight that people with gout are aware that ULT can precipitate flares and that prophylaxis is available to prevent these [6, 33, 34]. One study from New Zealand of patients’ perceptions and experiences of taking colchicine highlighted their concerns about polypharmacy and potential side-effects that can affect adherence [34]. Our clinical experience is that people with gout may be concerned that taking prophylaxis requires them to take additional tablets, which can be compounded by the need for gastroprotection if taking an NSAID or corticosteroids and/or bone protection if taking corticosteroids. Conversely, people with gout may find prophylaxis particularly acceptable if they have had frequent or severe flares or bad experiences of ULT-induced flares previously. However, existing studies have not explored the factors which influence people with gout to opt for prophylaxis, their information needs to support this decision, or their views on the relative merits of prophylaxis versus the ‘pill in the pocket’ approach, the main alternative to prophylaxis, which involves having medication to treat a flare available to be taken as soon as the flare starts [2]. It seems likely that some people with gout may prefer to manage ULT-induced flares as they arise rather than take regular prophylaxis, depending on their previous experience of flares (frequency, severity, and duration) and ULT (if taken previously), their perception of their individual flare risk, and their concerns about polypharmacy and adverse events.

In conclusion, ULT-induced flares can be a significant clinical problem, which contributes substantially to suboptimal adherence to ULT and poor outcomes in people with gout. There is robust evidence of the effectiveness and safety of colchicine to prevent flares when initiating ULT but limited evidence for NSAIDs or corticosteroids. However, not all people with gout experience a flare after initiating ULT, even without prophylaxis, and some may not wish to take flare prophylaxis, preferring to manage flares as they arise (Table 1). We advocate a personalised approach between people with gout and clinicians to reach shared treatment decisions when discussing the possibility of co-prescribing prophylaxis, which explores their views about prophylaxis, makes them fully aware of its benefits and risks (including those of different therapeutic options for prophylaxis), the alternatives to it (the ‘pill in the pocket’ approach), and what they should do in the event of a flare (including not stopping ULT), and provides them with a prescription for flare treatment that they can take as soon as a flare occurs. This approach not only provides a safety net for patients should a flare occur but also empowers them to make decisions about their care.