Does orlistat cause acute kidney injury?

Abstract

Orlistat is an inhibitor of gastric and pancreatic lipase with proven efficacy in the augmentation and maintenance of weight loss. Although its use has been limited by troublesome but benign gastrointestinal side effects, it has more recently been associated with acute kidney injury (AKI). In this review, we summarize orlistat’s benefits and drawbacks and discuss the body of evidence supporting its role as a cause of AKI. Although we cannot yet draw an unequivocal causal link between orlistat and AKI, there is enough evidence to include orlistat exposure in the clinical assessment of patients with AKI.

Introduction

Obesity is fast becoming a worldwide epidemic. Weight management strategies play an important role in mitigating the complications of obesity such as hypertension, dyslipidemia, cardiovascular disease, type II diabetes mellitus and obstructive sleep apnoea [Hubert et al. 1983; Lean et al. 1998; Hu et al. 2001; Davy and Hall, 2004]. Weight reduction achieved through lifestyle and behavioural modifications can be augmented by pharmaceuticals that reduce dietary fat absorption [Li and Cheung, 2011]. Several of these drugs have proven effective, but their use has been limited by undesirable, but largely benign side effects [Li and Cheung, 2011]. Orlistat (Roche, Basel Switzerland) is a gastrointestinal lipase inhibitor that is commonly prescribed for weight reduction. Orlistat is currently available in Canada by prescription but is available without a prescription in Australia, the European Union and the United States. Recent studies have suggested an association between orlistat and acute kidney injury (AKI) [Courtney et al. 2007; Singh et al. 2007; Weir et al. 2011]. In this review we consider the benefits and drawbacks of orlistat and discuss its role as a potential cause of AKI.

Orlistat efficacy

Orlistat, also known as tetrahydrolipstatin is a potent, selective and reversible inhibitor of gastric and pancreatic lipases [Guerciolini, 1997]. The drug acts primarily by reducing fat absorption from the diet via inhibition of triglyceride hydrolysis. Initial human studies demonstrated that orlistat could increase fecal fat excretion by up to 30% [Hauptman et al. 1992].

In the 1990s, several randomized controlled trials (RCTs) examined the efficacy of orlistat on weight reduction and weight loss maintenance. Sjorstrom and colleagues conducted a double-blind placebo-controlled crossover trial that demonstrated that overweight or obese patients achieved a greater reduction in body weight with orlistat (10.2%) compared with placebo (6.1%) after 1 year of treatment combined with a hypocaloric diet [Sjostrom et al. 1998]. Patients were then switched to either orlistat or placebo for a subsequent year and placed on a eucaloric diet. Those who continued orlistat for an additional year regained half as much weight as those switched to placebo. Subsequent RCTs verified the dose–response relationship between orlistat and weight loss in obese patients who were consuming a reduced energy diet [Hauptman et al. 1992; Hill et al. 1999; Davidson et al. 1999; Rossner et al. 2000]. A meta-analysis of 29 studies of orlistat revealed a mean additional weight loss of 2.89 kg per year of orlistat use compared with placebo [Li et al. 2005].

Secondary benefits of orlistat-induced weight loss have also been evaluated. Benefits such as lower plasma total and low-density lipoprotein (LDL) cholesterol concentrations, smaller waist circumference, lower fasting plasma glucose concentrations, and reductions in systolic and diastolic blood pressure have been found among patients treated with orlistat [Hill et al. 1999; Davidson et al. 1999; Rossner et al. 2000]. Orlistat use has also been shown to reduce the risk of developing type II diabetes [Colditz et al. 1995; Must et al. 1999; Hu et al. 2001]. In a 4-year, double-blind trial of 3305 patients with a body mass index (BMI) over 30 kg/m² treated with lifestyle changes, orlistat decreased the risk of new onset type 2 diabetes mellitus by 37% compared with placebo [Torgerson et al. 2004]. Among patients with established diabetes and poor glycemic control, orlistat has been shown to be an effective adjunctive therapy for decreasing fasting plasma glucose concentrations, reducing the percentage of glycated haemoglobin, and decreasing plasma total and LDL cholesterol concentrations [Hollander et al. 1998; Miles et al. 2002; Kelley et al. 2002].

Adverse drug events with orlistat

Despite orlistat’s beneficial effects, several drug interactions and adverse drug events have been noted in trials, drug surveillance studies and case reports. Orlistat can reduce bioavailability of fat-soluble drugs, for example amiodarone, cyclosporine, antiepileptic drugs, antipsychotics and thyroxine [Filippatos et al. 2008]. Orlistat’s most frequently observed adverse reactions result from the abundance of undigested fat in the colon, which results in diarrhoea (relative risk [RR] 3.4), flatulence (RR 3.10), bloating, abdominal pain and dyspepsia (RR 1.48) [Li et al. 2005]. Significant reductions in plasma concentrations of fat-soluble vitamins have also been associated with orlistat use although rarely below the lower limit of normal [Hauptman et al. 1992; Rossner et al. 2000; Torgerson et al. 2004].

More severe adverse drug events have also been described with orlistat use, including liver injury and AKI. Although data are limited to case reports, between 1999 and 2008 the American Food and Drug Administration received 32 reports of serious liver injury in patients using orlistat [DeLancey, 2011]. In 2007, Singh and colleagues provided the initial suggestion of orlistat-induced AKI [Singh et al. 2007]. In a patient with underlying stage III chronic kidney disease (CKD; among other comorbidities), initiation of orlistat treatment coincided with an increase in serum creatinine concentration, an increase in urine oxalate concentration and the presence of calcium oxalate crystals in the lumen of the renal tubules. These findings resolved 1 month after discontinuation of orlistat. Courtney and colleagues reported similar findings in a 55-year old woman with diabetic nephropathy [Courtney et al. 2007]. After 5 months of orlistat therapy, the patient’s estimated glomerular filtration rate (eGFR) declined from a baseline of 66 ml/min/1.73 m² to 12 ml/min/1.73 m² with a documented accumulation of calcium oxalate crystals in the renal parenchyma. In 2007, McLaughlin and Macdougall reported a case series of 33 obese patients with CKD (stages III or IV) taking orlistat for more than 6 months [MacLaughlin and Macdougall, 2007]. In 6 of the 33 patients (18%), eGFR decreased by >10 ml/min/1.73 m² in the 12 months following initiation of orlistat. Following these case reports, a retrospective review of 855 renal biopsies by Karamadoukis and colleagues identified two cases of calcium oxalate crystals in patients with unexplained acute tubular necrosis; both of these patients were taking orlistat at the time of biopsy [Karamadoukis et al. 2009]. These same investigators later identified a 66-year-old man with a history of type 2 diabetes who had been taking orlistat for 3 months. Following hospital admission for AKI, the patient suffered a progressive decline in renal function that eventually necessitated hemodialysis [Karamadoukis et al. 2009]. Most recently, we published a pre–post study using the health administrative databases of Ontario, Canada. We studied 953 patients using orlistat and compared their risk of AKI in the 12 months preceding orlistat initiation to their risk of AKI in the following 12 months [Weir et al. 2011]. This study showed that significantly more patients experienced AKI in the year after orlistat initiation compared with the year prior (18 versus 5, p = 0.01).

Proposed mechanism of orlistat-associated AKI

Orlistat is thought to cause AKI through enteric hyperoxaluria, which can also be seen in patients with fat-malabsorption syndromes. In this mechanism, unabsorbed fat in the small bowel resulting from orlistat use produces calcium soaps, which reduce the availability of free enteric calcium [Ahmed, 2010]. As a result, calcium can no longer bind and sequester oxalate inside the gut and this allows increased intestinal oxalate absorption. This leads to increased renal oxalate excretion. Supersaturation of oxalate in renal tubules is a risk factor for calcium oxalate precipitation, which can result in AKI [Williams and Wandzilak, 1989; Holmes et al. 2001]. This mechanism is supported by the human reports of orlistat-associated AKI and also by the work of Ferraz and colleagues who observed increases in urinary oxalate excretion in rats with orlistat-induced fat malabsorption [Ferraz et al. 2004].

Epidemiology of orlistat-associated AKI

Although the current literature draws associations between orlistat and AKI, it provides little insight into the frequency with which AKI occurs or the factors that may predict it. In our population-based study of 953 patients, we identified 18 individuals (1.9%) admitted to hospital with AKI after they filled prescriptions for orlistat [Weir et al. 2011]. When we consider that 5 individuals (0.5%) had admissions for AKI prior to filling prescriptions for orlistat, this suggests an excess risk of 1.4% over the 1-year follow-up period. However, this estimate is subject to the limitations of our study, most important of which was our inability to identify the aetiology of each patient’s AKI. Even less information is available to identify risk factors for orlistat-associated AKI. Although two out of the three case reports described AKI in patients with pre-existing CKD, CKD is a well-established risk factor for AKI in general, so this observation is unlikely to be specific to orlistat [Hsu et al. 2008]. To assess risk factors for orlistat-associated AKI, we re-examined our data and identified characteristics that differed between those with and without AKI, for which the standardized difference was greater than 10%. Compared with those who did not experience a hospitalization for AKI, patients who did have AKI were slightly older (mean age 60 versus 58 years), were generally in poorer health (mean Charlson comorbidity index 2.38 versus 1.76, mean number of drugs used in the past year 7.5 versus 5.9), and were more likely to have CKD (39% versus 10%), hypertension (89% versus 77%) and heart failure (39% versus 17%). Although this provides an impression of some potentially important risk factors, it is important to note that we assessed a relatively small number of AKI events.

Causality

Despite the associations between orlistat and AKI, establishing causality between drug exposure and adverse outcome is challenging [Macedo et al. 2003; Meyboom, 1998; Karch et al. 1976]. Although innumerable algorithms and probabilistic models have been developed to improve upon the common practice of assigning causality based on expert opinion, no gold standard exists [Agbabiaka et al. 2008]. Therefore, when assessing the relationship between orlistat exposure and AKI, in the absence of a gold-standard methodology, we are left to consider the elements of causal relationships commonly identified in the literature. These elements include an appropriate temporal sequence, a biologically plausible mechanism, the effects of drug challenge, dechallenge and rechallenge, the existence of confirmatory laboratory or pathological data, a consideration of alterative aetiologies and the demonstration of a dose–response relationship [Miremont et al. 1994; Arimone et al. 2005; Begaud et al. 1985; Emanueli and Sacchetti, 1980; Kitaguchi et al. 1983; Danan et al. 1993; Benichou et al. 1993].

In the case of orlistat and AKI, the literature fulfills some but not all of these requirements. Aside from the cross-sectional biopsy study, all studies describe an appropriate temporal relationship between orlistat exposure and AKI outcome. The putative mechanism underlying this relationship is plausible and supported by both animal data and histologic findings in humans. Although there is no documentation of patients being rechallenged with orlistat, the dechallenges reported by the groups of Singh and Courtney showed reversal of the kidney injury. These case reports also provide pathological confirmation of oxalate nephropathy, which is congruent with the proposed pathophysiology.

Although there appears to be strong support for a causal relationship between orlistat and AKI, the body of evidence has some important weaknesses. Case reports offer interesting observations, such as the presence of calcium oxalate crystals on biopsy or the resolution of AKI upon withdrawal of orlistat, but they cannot establish causality nor estimate risk. Observational studies are limited in their ability to establish causality by the influence of bias and confounding variables. The nonrandom allocation of treatment introduces a ‘bias by indication’ where subjects who receive the exposure of interest differ in potentially important ways from those who do not. Confounding variables can also threaten the validity of observational studies, and when these variables are unknown, unmeasured or inaccurately measured, statistical adjustment cannot adequately mitigate their effects. Because these limitations leave the door open for alternative aetiologies of AKI, causality is difficult to establish outside the setting of randomized trials. Unfortunately, trials designed to establish efficacy are not often powered to detect rare outcomes such as AKI [Tsang et al. 2009]. Furthermore, patients with pre-existing CKD, who would likely be at the highest risk for developing AKI, are often excluded from randomized trials [Coca et al. 2006]. The existing trials of orlistat do not provide information on the risk of AKI and given its flagging market share, we are unlikely to see future large-scale trials involving this drug.

Conclusion

Establishing unequivocal causality in the setting of adverse drug reactions is very difficult. Without evidence from randomized trials or multiple observational studies, we cannot confidently label orlistat a cause of AKI. However, we believe there is sufficient evidence to consider exposure to orlistat a potential cause of AKI. In patients with otherwise unexplained AKI, discontinuation of orlistat would be a reasonable course of action and in patients at risk of AKI, such as those with multiple comorbidities or pre-existing CKD, orlistat should be prescribed with caution and close observation.