- Top of page
Background Many oral 5-aminosalicylic acid (5-ASA) formulations are designed to maximize 5-ASA release in the colon where it acts topically on the colonic mucosa. Delayed-release formulations and azo-prodrugs minimize 5-ASA absorption in the upper gastrointestinal (GI) tract.
Aims To review methods for assessing 5-ASA release and colonic distribution from oral formulations, and the potential use of this information for guiding clinical decisions.
Methods PubMed and recent conference abstracts were searched for articles describing techniques used to assess 5-ASA release from ulcerative colitis (UC) therapies.
Results In-vitro GI models, although unable to simulate more complex aspects of GI physiology, can provide useful data on 5-ASA release kinetics and bioaccessibility. Gamma-scintigraphy is useful for investigating GI disintegration of different formulations, but may not accurately reflect 5-ASA distribution. Plasma pharmacokinetic studies provide data on systemic exposure, but not on colonic distribution or mucosal uptake. Mucosal biopsies provide direct evidence of colonic distribution and may predict clinical efficacy, but must be interpreted cautiously because of considerable inter-subject variability and other confounding factors.
Conclusion While assessment of 5-ASA release is important, limitations of individual measurement techniques mean that randomized clinical studies in UC patients remain the best guide for dosing and treatment regimen decisions.
- Top of page
Ulcerative colitis (UC) is a chronic, idiopathic inflammatory disorder affecting the colon and rectum, characterized by an unpredictable course of relapse and remission.1 At first presentation of the disease, approximately 40% of adult patients have UC that is limited to the rectum, known as proctitis.2 Left-sided or distal colitis, defined as UC extending proximally in the colon but not beyond the splenic flexure (approximately 60 cm from the anal verge), occurs in approximately 40% of adult patients presenting with UC. Less than 20% of adult patients at presentation have extensive UC, which extends proximally to the splenic flexure.2
The cardinal symptom of UC is diarrhoea containing blood and mucus. However, many patients with UC also develop other symptoms and abnormalities including abdominal pain, tenesmus, anaemia and weight loss.1, 3, 4 Less common symptoms can include arthritis/arthralgias, cutaneous conditions, renal disorders, osteoporosis, and inflammation of the eye, liver and biliary system.3, 5
The recommended first-line therapy for the treatment of active symptoms, induction of remission and maintenance of remission in patients with mild-to-moderate UC is the anti-inflammatory agent 5-aminosalicylic acid (5-ASA; mesalazine; Figure 1).1, 6 A number of potential targets for 5-ASA action have been proposed; among these is the peroxisome proliferator-activated receptor-γ (PPAR-γ), which is known to be involved in UC inflammation.7 Indeed, 5-ASA can act as a synthetic agonist of PPAR-γ.8 However, additional mechanisms of action, independent of PPAR-γ activation, have also been proposed. These include the inhibition of: prostaglandin synthesis (via inhibition of cyclo-oxygenase); chemotactic leukotriene synthesis (via inhibition of lipoxygenase);9 interleukin-1 (IL-1) synthesis;10, 11 and nuclear factor-kappa B activation by tumour necrosis factor alpha12 and IL-1.13 5-ASA may also act as a biological antioxidant by scavenging oxygen free radicals.14
As 5-ASA is believed to exert a direct effect on the colonic mucosa through a variety of anti-inflammatory mechanisms,15 direct application of this agent to the colonic mucosa is required. There are a number of strategies for achieving this delivery to the target tissue. Rectal administration of gels, foams and enemas containing 5-ASA is effective for administering the active drug directly to the rectum, sigmoid or left colon. Indeed, recent studies in patients with extensive UC have shown that combination of both oral and rectal therapies may maximize 5-ASA concentration throughout the colon and is superior to oral therapy alone.16 However, patients often dislike rectal formulations because of difficulty with this mode of administration and problems with discomfort, retention and leakage.17–19 In contrast, the variety of currently available oral 5-ASA formulations are more acceptable to patients.
Oral administration of 5-ASA presents a different challenge as the majority of oral 5-ASA, when taken in an unformulated fashion, is rapidly absorbed in the small intestine, leaving little or no 5-ASA to treat the colon (Figure 2). Sulfasalazine was the first 5-ASA to be used for the treatment and maintenance of symptoms of UC. The azo-bonded sulfapyridine molecule protects the active drug until bacterial azoreductase cleavage in the colon. However, the sulfasalazine molecule is associated with allergic reactions and a number of dose-dependent adverse effects.20 As a result, formulations based on the active moiety, 5-ASA, have been developed.1, 3, 21 Two main methods have been employed by manufacturers of 5-ASA-based UC treatments to prevent upper-gastrointestinal (GI) 5-ASA release.15, 17 The first is to link 5-ASA (as with sulfasalazine, via a di-azo bond) to another compound to form a non-absorbable prodrug. This molecule can then be cleaved in the colon by the bacterial enzyme azoreductase. Agents that utilize this strategy are balsalazide and olsalazine (in which the 5-ASA molecule is coupled to a benzoic acid derivative or another 5-ASA molecule). The second strategy is to coat the formulation in a gastro-resistant coating that prevents 5-ASA release until luminal conditions approach pH 7 (normally in the terminal ileum). This allows a bolus of 5-ASA to be released in the terminal ileum and the proximal colon. Such formulations include Asacol delayed-release mesalazine tablets (Procter & Gamble Pharmaceuticals, Cincinnati, OH, USA); Salofalk tablets (Axcan Pharma, Mont St Hilaire, QC, Canada) and Salofalk Granustix (Axcan Pharma).
Figure 2. Proposed metabolic pathway of 5-ASA after oral administration. The shaded area (large intestine) indicates the site of topical action. Unformulated 5-ASA is absorbed rapidly from the small intestine, and many current formulations are designed to delay release of 5-ASA until the terminal ileum or proximal colon. 5-ASA, 5-aminosalicylic acid; N-AC-5-ASA, N-acetyl-5-ASA.
Download figure to PowerPoint
Historically, there have been two drawbacks common to these oral 5-ASA-based UC therapies. Firstly, azo-bonded and delayed (bolus)-release formulations may not deliver therapeutically effective doses of 5-ASA to the left side of the colon (a site affected in all patients with UC). Indeed, clinical studies have shown that mucosal 5-ASA concentrations using azo-bonded or bolus-release formulations are typically highest in the right-sided colon, whereas in the rectum the concentration of 5-ASA is much lower.22, 23 Secondly, it has thus far been necessary to dose these formulations multiple times daily. This has been considered essential to ensure that therapeutically effective 5-ASA doses are maintained in the colon, and indeed, formulations dosed in this way have been shown to be efficacious for the treatment of UC in clinical studies.1, 6 However, patient compliance with these dosing schedules has been demonstrated to be poor in clinical practice, leading to reduced drug efficacy and thus poorer disease control (i.e. an increased number of UC flares).24–27
Recently, new technologies have been used to address the problem of bolus release of 5-ASA. MMX Multi Matrix System technology (Shire Pharmaceuticals Inc., Wayne, PA, USA), utilizes lipophilic and hydrophilic excipients to allow prolonged release of 5-ASA throughout the colon, following degradation of a gastro-resistant coating. This delivery system allows once-daily administration of high-concentration tablets (1.2 g 5-ASA per tablet).28, 29 Another formulation that avoids bolus release is Pentasa (Shire Pharmaceuticals Inc.; trademark licensed from Ferring A/S, Copenhagen, Denmark) which comprises 5-ASA containing microspheres enclosed within a moisture-sensitive, ethylcellulose, semi-permeable membrane. This allows pH-independent release of the active drug. Unlike other formulations, Pentasa starts releasing 5-ASA in the duodenum and continues throughout the entire GI tract.
Micropellet release systems are becoming more widespread. Provided as individual sachets containing a single dose, these formulations utilize granules to effect a delayed and sustained release of 5-ASA with similar delivery properties and systemic exposure to tablets.30, 31 Salofalk 500 mg and 1 g (Axcan Pharma and Falk Pharma, Freiburg, Germany) and Pentasa 1 g (Ferring A/S) are currently available in some European markets. Phase III trials are continuing in the US and a 1.5 g, once daily encapsulated mesalazine micropellet formulation (Salofalk Granustix) is also in clinical development (Salix Pharmaceuticals Inc., Morrisville, NC, USA).
Although randomized clinical trials have demonstrated that various formulations of 5-ASA are effective in either treating symptoms or inducing remission in UC,1, 6, 28, 29 it has been difficult to elucidate how each formulation or delivery system releases 5-ASA in the GI tract. This is further complicated by the variability in transit times and pH conditions that exist in UC patients, which can have a large impact on how 5-ASA is released from formulations and taken up by the colonic mucosa.32 While data exist pertaining to the predicted release mechanism of the various 5-ASA formulations, these data are not precise, which is in part because of limitations in the techniques available to perform these measures. Nevertheless, over the past 20 years, numerous studies have been performed that have added to our understanding of the release of 5-ASA from different formulations using a variety of different methods. These include: traditional pharmacokinetic (PK) investigations;33 simulated GI tract systems;34, 35 insertion of intestinal multilumen tubes for aspiration and marker perfusion;36 in-vitro dissolution tests;37, 38 gamma scintigraphy30, 31, 39–44 and tissue biopsy.22, 45 This review will compare the major techniques that are currently used in the study of 5-ASA release and will discuss what can be concluded from each technique in terms of formulation release characteristics, and 5-ASA bioaccessibility and efficacy.
Simulated GI models
Simulated GI models are a relatively recent development and are used to study the fate of ingested products in a dynamic bioenvironment that closely mimics the physiological conditions in the lumen of the human adult GI tract.46, 47 These model systems have frequently been used for investigating the bioaccessibility and site of action of nutritional supplements.48–50 While such systems cannot assess efficacy, they do provide a standardized environment in which to investigate the dissolution characteristics of 5-ASA formulations in a way that is not possible using other techniques.
A simulated GI model [the GI simulated system (GISS)] has been used to assess the release characteristics of several commercially available 5-ASA products for the treatment of UC.34 This system consists of four vessels, which are used to simulate the stomach, jejunum, ileum and proximal colon, respectively. Although transit time, pH, osmolarity and agitation can be varied and controlled with the GISS, it lacks the simulation of peristalsis and the removal of digestion/disintegration products. Despite these limitations, however, some interesting results have been observed, which provide information regarding the release characteristics of delayed-release 5-ASA formulations. All three of the delayed-release products assessed (Asacol tablet, Salofalk tablet, and Salofalk Granustix in sachet) began releasing 5-ASA in the proximal and mid parts of the simulated small intestine. The two tablet formulations released most of their 5-ASA prior to entry to the simulated colon, but by the end of the experiment had cumulatively released almost the entire administered dose of 5-ASA. In contrast, Salofalk Granustix had a much slower release profile, with 5-ASA being steadily released in the simulated colon.34 However, 5-ASA was released so slowly from the sachet formulation that less than half of the total dose was released by the end of the experiment.34 Therefore, it is a distinct possibility that while a patient may consume apparently similar doses of active drug, from various formulations, the actual concentration of 5-ASA delivered and where it is released may differ substantially, thus potentially affecting clinical efficacy.
Another GI model, the TNO GastroIntestinal Model (TIM system; TNO Quality of Life, Zeist, The Netherlands), is available for studying drug release in a simulated GI tract.35 This system consists of two computer-controlled, multicompartmental systems that simulate peristaltic movements and introduce gastric/biliary/pancreatic secretions in a controlled environment. The first system (TIM-1) simulates the stomach and small intestine and the second (TIM-2) simulates the colon.35 Recently, release kinetics of 5-ASA from MMX mesalazine tablets (Mezavant; Lialda) were assessed using the TIM system. The release kinetics of 5-ASA were investigated under standardized fed and fasted conditions with automated sampling via dialysis at various sections of the system. Less than 1% of the 5-ASA was released from the tablet in the simulated stomach and small intestine (prior to introduction into the simulated colon). Under fasted conditions, 78% of the 5-ASA was released into the simulated colon and under fed conditions, 68.5% was released. Notably, substantial quantities of 5-ASA were released during the 8–18-h sampling period [49.6 mg/h (fasted) and 40.7 mg/h (fed)]. Importantly, and in contrast to other techniques such as plasma PK analysis, this particular system allows a spatial-time analysis to determine where in the simulated colon 5-ASA was released. Indeed, the authors reported that under both simulated fasted- and fed-state conditions, cumulative 5-ASA recovery in the colon dialysate was sigmoidal in nature. These data support the notion that MMX technology prolongs release of 5-ASA from MMX mesalazine in a way that may allow greater distribution of 5-ASA throughout the colon.
Ultimately, while the usefulness of in-vitro systems is limited by the fact that they are not able to completely replicate human gut physiology (including the complex immunological and inflammatory factors that are present in patients with UC), and provide no information on the clinical effectiveness of 5-ASA therapies, the results from these two in-vitro studies clearly demonstrate that simulated GI models have a role to play in providing information regarding 5-ASA release profiles. Indeed, such model systems allow reproducible results to be obtained rapidly and in a non-invasive fashion. Perhaps the greatest advantage of in-vitro systems over the other methods of assessing release of 5-ASA from oral UC therapies is that they allow different formulations to be compared in a standardized environment.
Gamma-scintigraphy was originally developed as a technique for detecting gamma-ray emitting molecules localized in specific structures and organs of the body. The incorporation of gamma-emitters into drug formulations allows gamma-scintigraphy to be used to visualize the distribution or accumulation of drug within the body.40, 41, 51–54 Gamma-scintigraphy requires that a drug formulation be labelled with a gamma-radiation-emitting tracer. This can be achieved either by directly incorporating a gamma-emitting compound into the dosage form, or by neutron activation of a formulation containing a nonradioactive tracer. Because gamma-emitters are rare amongst the constituent elements of the majority of drugs (e.g. hydrogen, carbon, nitrogen, oxygen, phosphorous and sulphur), alternative strategies employing other elements are required for gamma-scintigraphy. For the evaluation of the transit and disintegration of complex formulations, such as enteric-coated tablets or pellets, labelling can be performed by the addition of a nonradioactive tracer (e.g. samarium-152 oxide) that is not absorbed from the GI tract. Samarium-152 oxide can then be converted into a gamma-emitter (samarium-153) by neutron activation of the formulated product.31, 51, 55, 56 This approach enables the fate of the tracer to be followed once released from the formulation. However, a limitation of the technique is that the drug and tracer are separate and their disposition after release may be different.
A number of studies have utilized gamma-scintigraphy to assess the release characteristics of oral 5-ASA formulations.30, 39, 40, 42, 53, 57, 58 In these studies, gamma-scintigraphy was used to visualize intact gastro-resistant tablets travelling through the upper-GI tract and their subsequent dissolution, together with progression of the radioactive tracer molecule through the colon at more advanced timepoints.30, 39, 40, 57 However, while scintigraphic images gave a partial impression of release, they could not quantify the amount of 5-ASA released, which might have occurred at a different rate from that of the tracer. Ultimately, while being a useful tool, gamma-scintigraphy provides no accurate information on the subsequent distribution and fate of 5-ASA once it is released from the formulation.
Often, for a drug with a site of action in rapid equilibrium with plasma, estimating therapeutic effects is helped by an understanding of the drug’s distribution within the body (PK). It may also be necessary to understand how the effects of the drug at the site of action [pharmacodynamics (PD)] relate to their concentration in the plasma. For some drugs, plasma PK and PD can be used in clinical practice to guide manipulation of dose and dosage regimens to optimize therapies within tolerable limits. However, as 5-ASA acts topically in the treatment of UC, systemic exposure to 5-ASA is not necessarily related to therapeutic efficacy. Moreover, because only low levels of 5-ASA are released into the plasma via the colonic mucosa, it is difficult to relate mucosal concentrations of 5-ASA or how 5-ASA is distributed along the length of the colon to the plasma concentration.
Poor uptake of 5-ASA from the colonic mucosa into the bloodstream is caused partly by the hydrophilic nature of 5-ASA, which means that only a very small proportion of the 5-ASA in the colon will diffuse into the plasma through lipid membranes.59 Furthermore, paracellular absorption is reduced at the tight junctions of the colonic mucosa, further reducing the extent of systemic absorption, particularly for delayed-release and azo-bonded formulations compared with formulations that release more 5-ASA into the small intestine. Also, 5-ASA and its metabolite [N-acetyl-5-ASA (N-Ac-5-ASA)] are secreted back into the colonic lumen following uptake into the colonic mucosa,60 reducing further the amount of 5-ASA that will progress to the plasma. These observations are supported by clinical data, which show that serum concentrations of 5-ASA after colonic instillation are only one-tenth of those after jejunal instillation.61 Moreover, following administration of a single dose of either a delayed-release or an azo-bonded formulation, only approximately 20% of the 5-ASA is taken up systemically (assessed by analysis of 5-ASA and its main metabolite N-Ac-5-ASA in urine).62
Although there are limits to the utility of plasma PK and PD assessments for 5-ASA in the treatment of UC, they have more relevance in the context of avoiding toxicity as systemic exposure of active drug and its metabolites drive the occurrence of adverse events outside the GI tract. PK and PD assessments can often act as a bridge between nonclinical and clinical safety evaluation. Indeed, they have been used successfully in supportive safety analyses in dose-ranging studies.61, 63 In the case of 5-ASA, it is important that such PK analyses be determined following multiple doses, as steady-state 5-ASA plasma concentrations are several-fold higher than following administration of a single dose.64
The results of plasma PK analyses of delayed-release formulations of 5-ASA have shown high inter-patient variability,65 mainly because of the low bioavailability of 5-ASA. As most PK studies use only small sample sizes, making definitive conclusions is difficult. Nevertheless, in a number of studies, 5-ASA formulations have been compared on the basis of plasma PK data.62, 64–68 While similar 5-ASA plasma PK profiles may potentially predict similar safety profiles, conclusions regarding efficacy, dosing schedules or colonic release characteristics of 5-ASA should be considered with caution. Indeed, as it is the local effect of mesalazine that determines efficacy, not systemic concentration, only well-controlled clinical trials can determine efficacious doses and/or dose regimens.
As plasma PK is not directly related to the availability of 5-ASA at the colonic mucosa, bioequivalence studies, which can be used for many drugs to validate potential therapeutic equivalence, are not appropriate for delayed-release 5-ASA formulations. Indeed, the FDA have specifically recommended that bioequivalence ‘where the drug substance produces its effects by local action in the GI tract….can be achieved using bioequivalence studies with clinical efficacy and safety endpoints and/or suitably designed and validated in-vitro studies if the latter studies are either reflective of important clinical effects or are more sensitive to changes in product performance compared to a clinical study’.69 In addition, the European regulatory authorities (Committee of Proprietary Medical Products) have stated that for ‘locally acting products PK bioequivalence generally is not a suitable way to show therapeutic equivalence, since plasma levels are not relevant for local efficacy, although they may play a role with regard to safety’.70 This said, several generic balsalazide formulations have been approved on the basis of bioequivalence to the original product Colazide (Salix Pharmaceuticals, Inc., Morrisville, NC, USA). It is assumed that the fact that these were all formulations of an identical prodrug was pivotal to this decision. Otherwise, no two 5-ASA formulations and specifically no two formulations delivering the drug by different technologies are currently considered to be bioequivalent.
Mucosal tissue concentrations
The concentration of 5-ASA in mucosal tissues can be assessed from biopsy samples. Although useful, this technique is fraught with difficulties and is prone to large inter- and intra-patient variability. Because of its invasive nature (requiring endoscopy with mucosal biopsy), direct assessment of mucosal concentration is not always feasible in randomized studies. Further, indirect techniques are also described below. Studies that have been performed using mucosal biopsy have shown that high mucosal 5-ASA concentrations are associated with improved mucosal healing in patients with UC.45, 71, 72 In a study of 21 patients with UC taking oral 5-ASA therapy, mucosal concentrations of 5-ASA were found to be significantly higher in patients with no or only mild mucosal damage on endoscopy compared with those having moderate or severe mucosal damage. While this may reflect greater mucosal 5-ASA uptake, it may also reflect the decreased epithelial cell turnover that is seen as healing occurs. It has also been shown that levels of soluble IL-2 receptor, a pro-inflammatory marker, were lower in patients with relatively high mucosal 5-ASA concentrations.71 The association between high mucosal 5-ASA levels and low disease activity was also shown when assessing 5-ASA concentrations from rectal biopsies in a study of 29 patients taking oral sulfasalazine, or delayed-release 5-ASA, with or without rectally administered 5-ASA.45 The link between mucosal 5-ASA levels and disease activity was strengthened further by the results from a study of 18 patients with UC deemed at high risk of relapse, despite existing 5-ASA therapy. Following increases in the total daily 5-ASA dose (using both oral and rectal formulations), significant increases in mucosal 5-ASA concentrations were observed relative to baseline. Significant reductions in the number of relapses were also seen during the period of the study.72
Differences in tablet delivery systems, as well as in intestinal behaviour and colonic segmental transit time, may lead to differences in drug availability at the level of the colonic mucosa as assessed by tissue biopsy.22, 73, 74 For example, De Vos et al.22 reported that in 61 patients with irritable bowel syndrome in whom GI transit had been accelerated by the administration of a promotility agent (metoclopramide), higher mucosal concentrations were achieved after administration of slow release 5-ASA preparations than after azo-bound drugs. The authors also drew attention to the fact that an administered dose of 5-ASA may show high interindividual variability in mucosal concentrations.22
Increasing the oral dose of some 5-ASA formulations may not necessarily lead to a corresponding increase in mucosal concentrations at the sites of inflammation, irrespective of increasing plasma concentrations.75 These observations could be related to the release of 5-ASA from most bolus-release formulations in the terminal ileum or proximal colon. Indeed, studies with a range of azo-bonded and bolus-release formulations have shown that mucosal 5-ASA concentrations were typically highest in the right-sided colon, whereas in the rectum (a site affected in the majority of patients), the concentration of 5-ASA was much lower.22, 23 The uptake of majority of 5-ASA in the terminal ileum or proximal colon may explain why increased oral dosing with some formulations only leads to a limited dose response in 5-ASA levels in the rectum, assessed by rectal mucosal biopsies,22, 75 despite a progressive increase in serum and urine 5-ASA concentrations.22, 75 Conversely, there are data to suggest that some formulations may be able to achieve a greater mucosal concentration following higher oral doses. For example, in a recent pilot study, higher 5-ASA mucosal concentrations were observed in the sigmoid colon and rectum in patients with UC following treatment with MMX mesalazine 4.8 g/day given once daily than following treatment with MMX mesalazine 1.2 or 2.4 g/day given once daily.63 In a biopsy study of patients taking either a mean dose of 6.75 mg/day balsalazide (containing 2.4 g of 5-ASA) vs. patients taking a mean dose of 3.74 g/day Asacol, patients in the balsalazide group demonstrated similar or higher mucosal concentrations of 5-ASA.76 Taken together, all these findings suggest that the relationship between dose and mucosal concentrations may be formulation-dependent and independent of relationships between dose and plasma concentrations.
Interindividual variability in mucosal 5-ASA concentrations may contribute to the variability in clinical effectiveness that is often observed with oral 5-ASA therapy. In-vivo studies have shown that the same oral dose does not always exert the same therapeutic effect and that increasing the oral dose does not always provide additional therapeutic benefit to all patients.77–80 The source of the interindividual variability is unclear, but may involve differences in drug disposition in the colonic lumen and mucosa. Factors such as local pH conditions, transit rates and activities of the relevant enzymes and transporters responsible for colonic mucosal 5-ASA metabolism and apical secretion back into the colonic lumen may all be relevant factors. Clearly, lower tissue 5-ASA concentrations during maintenance therapy may predispose patients to relapse. Simply increasing the oral dose, however, may not be sufficient to maintain or improve remission rates.
Some investigations have attempted to estimate the concentration of 5-ASA in mucosal tissues from the faecal concentrations of 5-ASA and its major metabolite N-Ac-5-ASA. Patients are required to swallow dialysis bags which are subsequently excreted with the faeces. Analysis, of such dialysis bags has provided some useful information about the PK of the aminosalicylates.81
However, collection and analysis of faeces and dialysis bags is unpleasant for patients and technicians. Furthermore, the stability of compounds released from the mucosa in stored faeces (or dialysis fluid) is uncertain. Techniques that examine faecal concentrations therefore seem unlikely to achieve widespread application in clinical practice or trials.
Similarly, in-vivo rectal dialysis utilizes a cylindrical dialysis membrane that is inserted into the rectum in contact with the rectal mucosa. Compounds that are released from the mucosa move across the dialysis membrane into the dialysis fluid. As previously, their concentration is measured after removal of the dialysis bag. This in-vivo technique has been used to study the effects of 5-ASA82 and insertion of the dialysis bag does not appear to alter mucosal function unless the procedure is repeated frequently within a short period.83 However, if equilibrium dialysis is to be performed, the procedure can take ≥4 h.84
While dialysis presents a potentially less invasive option than mucosal biopsy, the technique is hampered by the possibility of degradation or faecal contamination. Furthermore, large molecular weight compounds may have increased difficulty in entering the dialysis bag, thus providing an inaccurate representation of the mucosal concentrations. Importantly, rectal dialysis is not representative of the mucosal concentrations in locations other than the rectum. In contrast, faecal dialysis cannot accurately represent any one segment of the mucosae and indeed may be prone to interference from the latest delayed-release 5-ASA formulations.