Target population for trial

Ureteric colic describes the pain felt when a stone formed in the urine collection part of the kidney (usually resulting from the aggregation of calcium-based crystals) passes down the ureter, the urine drainage tube connecting the kidney to the bladder (Figure 1). The pain is typically severe and recurrent; each episode usually lasts for an hour or two and is interspersed with periods without pain. Female sufferers consider it to be more intense than the pain experienced during childbirth.¹ Repeated episodes have a detrimental influence on perception of quality of life.² Pain is usually felt in the abdomen but can go down into the groin and scrotum in males, and labia in females. Some people also get increased frequency and urgency of urination. The pain is likely to relate to direct contact of the stone with the epithelial cells lining the ureter and sustained contraction (spasm) of the smooth muscle surrounding the ureter in response to the obstruction of urine flow.³ The severity of the pain leads to secondary symptoms such as nausea, vomiting and fever. Pain episodes generally continue until the stone passes into the bladder from where it can be freely voided by urination. Kidney stone disease is highly prevalent in most countries, where it affects 5% to 10% of the population,⁴ and, as there is no effective disease-modifying treatment, sufferers may have repeated episodes following the first bout of colic, with an approximate lifetime risk of recurrence of 50%.^5,6 The cause of kidney stone formation is multifactorial with genetic, environmental and dietary influences all playing a part. In epidemiological terms it is more common in people aged 15–60 years, in men and in those of Caucasian race, and there is a higher incidence during the summer months.⁷

FIGURE 1

Anatomy of urinary tract showing definition of ureteric segments. Reproduced, with permission, from Medindia. URL: www.medindia.net/news/michigan-hospitals-lead-the-way-in-preventing-common-and-costly-urinary-tract-infections-116433-1.htm (accessed July (more...)

Ureteric colic is one of the most common reasons for people to seek emergency health care, with an annual incidence of around 30 out of 100,000 in high-resource parts of the world.⁸ In the UK, 83,000 people required emergency care for ureteric colic in 2009,⁹ and NHS England health episode statistics data for 2012–13 showed 31,000 hospital admissions with a 1-day median stay and a NHS tariff cost of £19.3M.¹⁰ In the USA, there were 600,000 emergency room visits in 2000 at an estimated health-care cost of US$490M.⁷ In both countries, the incidence of ureteric colic increased by over 50% during the previous decade.^7,9

Clinical assessment

The diagnosis of ureteric colic is usually straightforward from the characteristic history of pain, the lack of abdominal tenderness and, to some extent, the finding of non-visible haematuria on reagent strip testing of urine.⁸ At this point the patient is given effective pain relief in the form of opiates or non-steroidal anti-inflammatory drugs (NSAIDs), alone or in combination.^11,12 Urine is then checked for infection by microscopy and culture, and renal function is estimated by measurement of serum creatinine. Once the patient is comfortable, the presence of a stone in the ureter can be reliably confirmed by imaging using computerised tomography scanning of the kidneys, ureters and bladder (CT KUB) without the use of intravascular contrast agents. Further analysis of the CT KUB images will localise the stone to the upper, mid or lower ureter (see Figure 1) and will estimate stone size defined by the maximum linear dimension in millimetres. Less commonly, patients may have evidence of severe urinary and bloodstream infection (urosepsis) as a result of the stone obstructing drainage from the affected kidney; they may have stones in both ureters; or they may have a single kidney, all of which cause marked impairment of renal function. Each of these situations will require resuscitation and urgent intervention and, therefore, such patients are not the focus of this trial.

Interventions for ureteric colic

Guidance on patient management options with summarised relative benefits and harms has been formulated under a joint initiative between the European Association of Urology and the American Urological Association, and form the basis of clinical management for people with kidney and ureteric stones, particularly across Europe and North America.¹³ The options for treatment fall under three headings – symptom management, MET and active treatment – and these are discussed below.

Outcomes of interest

Cohort studies and observations from placebo or standard therapy groups of RCTs suggest that about 50–85% of people with ureteric colic will pass their stone unaided (spontaneous stone passage) within 4 weeks of diagnosis. Speedier passage of the stone would tend to reduce overall pain burden and thereby lessen the impact of any pain on the patient’s lifestyle, in terms of time off work and interference in day-to-day activities. From a clinical perspective, confirming stone passage is difficult. Patients are sometimes encouraged to sieve their urine (most stones are the size of a match head), but adherence is doubtful and small fragments can easily be missed. Simple imaging by plain single abdominal radiography or by ultrasound to confirm the absence of a stone has low diagnostic accuracy,¹⁷ whereas further definitive imaging by repeated CT KUB gives levels of radiation exposure that are generally considered to be unacceptable for this predominantly young patient group.¹⁸ Clinical definition of stone passage is, therefore, the absence of pain or other relevant symptoms or signs. For the 30% of patients who continue to experience symptoms, further intervention (sometimes with MET if not previously used, but more usually with active stone removal by ESWL or ureteroscopy) will be required following repeat imaging. The urgency of any active intervention will depend on individual patient circumstances and resource availability in the particular health-care setting. There is an increased likelihood that active treatment will be required with larger stones (conventionally described as > 5 mm) and with stones located in the mid or upper ureter at the time of first presentation (Table 1). The need for further treatment is an important and routinely measurable outcome for both patients and providers of health care, as active intervention is associated with harm to patients and increased health-care costs.¹³

TABLE 1

Approximate probability of spontaneous stone passage according to stone characteristics

Background

Given that care for most patients with ureteric stones is delivered with the expectation of spontaneous stone passage, several strategies aimed at reducing pain, hastening stone passage and increasing the rate of stone passage have been proposed and trialled. Such strategies are termed MET. Agents that initially appeared to be useful but then failed to show efficacy in more robustly designed studies included diuretics and administration of high fluid load to increase urinary hydrostatic pressure above the stone;²⁰ steroidal anti-inflammatory drugs and NSAIDs to reduce ureteric oedema and inflammation around the stone;^21,22 and antimuscarinic drugs to inhibit ureteric muscular contraction.²³ The two drug classes that appear to show efficacy in repeated small-scale RCTs and subsequent meta-analysis are calcium channel antagonists and alpha-adrenoreceptor antagonists.²⁴

Experiments using animal and human tissue models demonstrate that ureteric smooth muscle contraction can be stimulated by activation of adrenergic receptors, particularly the alpha-1D subtype.^25–28 Blockade of alpha-1 receptors by specific pharmacological antagonists (alpha-blockers) such as doxazosin,²⁹ terazosin,³⁰ alfuzosin³¹ and, most typically, tamsulosin^32,33 results in ureteric smooth muscle relaxation. It was hypothesised that this would translate into clinical benefit for people with ureteric colic. Smooth muscle contraction is stimulated in part by the influx of calcium ions into the smooth muscle cell through specific channels in the cell membrane, which are opened and closed by changes in the degree of electrical polarisation. These channels can be blocked by specific pharmacological antagonists, such as nifedipine, resulting in less calcium influx and reduced smooth muscle contraction. Experimental work demonstrating this effect in vitro encouraged translation to the clinical care of people with ureteric colic.^34–38

Pharmacological characteristics of putative agents

Tamsulosin is a readily available alpha-blocker which is licensed by the European Medicines Agency and the US Food and Drug Administration at a dose of 400 µg in the form of a modified-release (MR) oral tablet to be taken daily for relief of lower urinary tract symptoms in men.^39,40 It is generally well tolerated but is associated with common (0.1–1%) risks of dizziness and retrograde ejaculation in men. The calcium channel blocker nifedipine is available in the form of MR tablets or capsules in varying doses, ranging from 20 mg to 60 mg, with 30 mg being most often used.⁴⁰ The drug is licensed for the treatment of hypertension, Raynaud’s phenomenon and prophylaxis of angina, although it has been largely superseded for these indications by more specific and effective calcium channel antagonists. The common side effects (0.1–1%) include headache, dizziness, flushing, constipation and oedema of the lower limbs.

Evidence review for efficacy of tamsulosin and nifedipine as medical expulsive therapy

Multiple RCTs have been carried out testing the efficacy of both alpha-blockers (typically tamsulosin 400 µg) and calcium channel blockers (typically nifedipine 30 mg) compared with placebo, standard care, which may include other interventions, and each other. The medications are prescribed for use either up until the time of spontaneous stone passage or for up to one per month without stone passage. The conduct, quality and results of these trials have been examined by a number of systematic reviews and the results combined in meta-analyses, the findings of which concerning the main outcomes of interest will now be summarised.

Stone clearance

Comparison of the rate of spontaneous stone passage between MET and control is the primary outcome for the great majority of RCTs included in the meta-analyses. The absolute rates of stone passage are likely to vary according to trial eligibility criteria, such as type of diagnostic imaging used, stone location, stone size, the time point at which the outcome is censored and the protocol used to decide whether or not the stone has passed. Variation in these trial features is illustrated from 22 studies included in a Cochrane review¹⁴ that compared tamsulosin 400 µg with control (Table 2).

TABLE 2

Variation in key trial design characteristics

Despite these inconsistencies in trial design, the available meta-analyses all demonstrate an apparent beneficial effect of both tamsulosin and nifedipine as agents to increase the proportion of patients with ureteric colic who spontaneously pass their stone within a reasonable time frame (Table 3).

TABLE 3

Summarised stone expulsion rates after tamsulosin or nifedipine therapy from meta-analysis

Regarding the effect of stone size, the meta-analysis by Seitz et al.¹⁵ reported a relative risk (RR) in favour of tamsulosin versus control of 1.25 [95% confidence interval (CI) 1.12 to 1.40] for stones < 5 mm, and 1.62 (95% CI 1.50 to 1.74) for stones ≥ 5 mm. For nifedipine against control, the RR for stones < 5 mm was 1.49 (95% CI 1.17 to 1.88) and for stones ≥ 5 mm was 1.49 (95% CI 1.31 to 1.69). Similarly, Campschroer et al.¹⁴ reported absolute rates of stone clearance for alpha-blocker against control for stones ≤ 5 mm of 83% versus 56%, with a RR of 1.4 (95% CI 1.2 to 1.7), and for stones > 5 mm of 67% versus 39%, with a RR of 1.7 (95% CI 1.3 to 2.1). Summarised results for the effect of stone location on clearance rates for tamsulosin against standard therapy were reported by Fan et al.,⁴³ with a RR of 1.55 for lower ureteral stones (95% CI 1.43 to 1.68) and a RR of 1.28 for upper ureteral stones (95% CI 1.04 to 1.57). Similarly, Campschroer et al.¹⁴ reported absolute rates of stone clearance for alpha-blockers against control for stones in the lower (distal) ureter of 79% versus 55% (RR 1.4, 95% CI 1.2 to 1.6) and for stones in the mid or upper ureter of 39% versus 27% (RR 1.5, 95% CI 0.9 to 2.4). The results of these meta-analyses should be interpreted with caution given the uncertainties associated with estimating effect size in subgroups of the overall trial population. Overall, it does appear that these drugs demonstrate efficacy for passage of larger stones and for stones in the upper section of the ureter that are considered to be less likely to pass without active intervention.⁸

Time to stone passage

Shorter duration of stone episode is likely to be associated with less pain and less social inconvenience, such as time off work, which may be of benefit to patients. Most RCTs examined time to stone passage as a secondary outcome, although the degree to which this was reported varied, thus making meta-analysis difficult. However, the direction of effect was consistent in favouring tamsulosin or nifedipine (Table 4). It should also be noted that censoring of exact time of passage is reliant on precise patient report or timing of follow-up imaging leading to reporting inaccuracy.

TABLE 4

Summarised results for time to stone expulsion after tamsulosin or nifedipine therapy from meta-analysis

Pain episodes/use of analgesia

Pain is likely to be the main reason for continued ill health in people with ureteric stones and it drives both transient quality-of-life detriment and the need for further intervention. However, it is difficult to measure in an ambulatory setting, particularly for conditions such as ureteric colic which are characterised by episodic pain of varying severity and frequency. The RCTs and subsequent meta-analyses reported this outcome, but the differences in definition make measurement of comparative effect uncertain (Table 5).

TABLE 5

Summarised results for number of pain episodes after tamsulosin or nifedipine therapy from meta-analyses

Need for hospitalisation

The final main outcome of interest is the proportion of patients that require further active management, which will mainly consist of stone removal using ESWL or ureteroscopy. Within the RCTs and subsequent meta-analyses this is mainly reported as the rate of further hospital admissions for treatment of a ureteric stone (Table 6).

TABLE 6

Summarised hospitalisation rates after tamsulosin or nifedipine therapy from meta-analysis

Cost-effectiveness

A cost-minimisation study has been performed which assessed the potential health economic benefit of MET compared with observation alone.⁴⁴ A decision analytical model was used to predict comparative costs and resource use for a MET strategy using tamsulosin based on cost data obtained from the USA and four European countries, and efficacy data from existing meta-analyses. The costs of adverse events and other treatment-related complications were not included and a cost–utility analysis was not performed. The study found that use of tamsulosin for MET might lead to a saving of US$1132 over observation, with this conclusion being unchanged by sensitivity analyses. The cost-effectiveness of MET for different health-care systems remains unknown.

Summary

All seven meta-analyses using different selection criteria and reporting protocols appear to suggest that treatment with either tamsulosin or nifedipine at the time of presentation with acute ureteric colic increases the likelihood of eventual spontaneous stone passage, hastens the time to stone passage and reduces the risk of unwanted consequences such as pain and the need for an intervention to remove the stone. This leads the authors of these reviews to conclude that the balance of evidence supports routine use of these therapies, although with caveats regarding individual trial quality and trial size, with all but one trial having fewer than 100 participants in each group (Table 7). One large trial compared the use of tamsulosin (400 µg) against nifedipine (30 mg) as MET for stones sized 4–7 mm located in the very distal ureter (ureter course within the bladder wall) across 10 centres in China.⁴⁵ The trial randomised 3189 patients, and, at 4 weeks, the stone expulsion rate was 96% for tamsulosin and 74% for nifedipine with no further details given, although the Cochrane review¹⁴ gave a RR of 1.3 (95% CI 1.2 to 1.4) and found the trial to have low risk of bias. However, limited details given in the trial report makes quality assessment difficult; there was no sample size calculation to justify the large sample and diagnosis was by any imaging method, although follow-up and primary outcome assessment was by weekly CT KUB until stone passage or up to 4 weeks. Mean time to stone passage was given as 78.35 hours in the tamsulosin group and 137.93 hours in the nifedipine group. This duration of stone episode appears much shorter than that recorded in previous trials and may relate to the inclusion criteria and recruitment policies for this particular study. Additionally, the Jadad quality score was low (one) and the trial was supported by a manufacturer of tamsulosin.

TABLE 7

Summary of conclusions of meta-analyses

The possibility that the conclusion from published meta-analyses regarding the benefits of MET is incorrect has to be borne in mind, as up to one-third of meta-analyses that show positive outcomes of a therapy are later altered by the inclusion of results from single, large, multicentre, robust, well-designed RCTs.⁴⁶ In the case of MET trials it is noted that use of less diagnostically accurate methods of imaging for participant inclusion prior to the widespread availability of CT KUB may lead to selection bias, because either smaller or radiolucent stones are missed, or people without a stone are included. Similarly, older, less accurate methods of imaging were widely used to decide if the stone had passed. This, together with lack of blinding in non-placebo-controlled studies, could lead to ascertainment bias in favour of the intervention group.