Gastrointestinal pH profiles in patients with inflammatory bowel disease

Authors


Ramadori Leiter der Abteilung Gastroenterologie und Endokrinologie, Medizinische Klinik und Poliklinik, Robert-Koch-Str. 40, D 37075 Göttingen, Germany.

Abstract

Background:

5-Amino salicylic acid preparations are used in therapy for patients with inflammatory bowel diseases. The bioavailability of these drugs depends on their coating.

Aim:

To determine whether intraluminal pH is decreased by the presence of inflammation, thereby altering the release of 5-amino salicylic acid in the intestinal lumen.

Methods:

Intraluminal gastrointestinal pH was measured by means of a radiotelemetry capsule in 12 healthy controls, in 12 patients with Crohn’s disease (five with active disease), and in 11 patients with ulcerative colitis (seven with active disease).

Results:

The median gastric pH values in the patient groups (Crohn’s disease 2.4, range 1.5–4.1; ulcerative colitis 1.95, range 1.55–4.4) were significantly higher than those observed in the controls (1.55, range 0.95–2.6). In the small bowel and colonic segments, all the pH values of Crohn’s disease patients were comparable to those of the controls, as were the pH values in the proximal small intestine and in the left colon in patients with ulcerative colitis. However, the latter group had higher pH values in the terminal ileum, the caecum and the right colon. Patients with active disease had comparable median gastrointestinal pH values to patients in remission.

Conclusions:

The luminal release of 5-amino salicylic acid might not be inhibited by low pH in patients with active inflammatory bowel diseases. This supports a safe disintegration of the slow release mesalazine preparations even in the presence of severe disease.

INTRODUCTION

Inflammatory bowel diseases are chronic diseases of unknown aetiology. Transmural chronic inflammation may involve all intestinal segments in Crohn’s disease. Mucosal inflammation is limited to the colon in ulcerative colitis. The natural course of both variants consists of quiescent phases that are interrupted by flare-ups. Mesalazine (5-amino salicylic acid) and sulphasalazine are effective in relapse prevention1[2]–3 and in active disease.4, 5

Mesalazine acts locally from the intestinal lumen on the inflamed mucosa. Mesalazine is rapidly absorbed by the upper small intestine, hence absorption must be delayed until the substance reach the terminal ileum or the colon. Slow release formulas were developed that prevent mesalazine from early absorption in a pH dependent fashion. Eudragid L-coated mesalazine (Salofalk, Dr Falk Pharma GmbH, Freiburg; Claversal, SmithKline Beecham Pharma GmbH, München) dissolves at an intraluminal pH of 6, whereas Eudragit S-coated mesalazine (Asacolitin, Henning Berlin GmbH, Berlin) has its optimum at pH 7. Both coatings are intended to guarantee the release of mesalazine distal to the terminal ileum.6

The therapeutic value of mesalazine preparations is uncertain in cases of severe intestinal inflammation. One reason for this could lie in insufficient drug release when intestinal pH values are too low, as has been suggested by faecal pH data.7, 8

Some data have shown that the luminal colonic pH is decreased when there is extensive ulcerative colitis.9 However, the data for intestinal milieu changes in ulcerative colitis are sparse and they are not known for Crohn’s disease.

The purpose of this study was to evaluate the intestinal pH in patients with ulcerative colitis and with Crohn’s disease in relation to the disease activity, to define possible situations where the liberation of 5-ASA preparation might be decreased.

PATIENTS AND METHODS

Twelve healthy controls and 11 patients with ulcerative colitis were evaluated, seven of them with active disease (clinical Gomes index = 4) and four in remission (Gomes < 4).10 Eight patients had pancolitis and three had left sided colitis. Their average age was 33.2 ± 10.1 years (mean ± s.d.).

The controls consisted of six healthy men and six healthy women, all Caucasian, with normal diets (no vegetarians or ethnic diets). Their average age was 27.8 ± 4.0 years (mean ± s.d.). None of the controls had a history of previous chronic bowel diseases or bowel operations. All women used birth control hormones to avoid pregnancy during pH-measurements.

Additionally, 15 patients with Crohn’s disease were evaluated, six of them with active disease (CDAI > 150) and nine in remission (CDAI < 150).11 Three patients (one with active disease and two in remission) had to be excluded because their intestinal pH measurements revealed a data loss > 75% (75% loss from possible data recordings in 24 h were accepted), due to a broken antenna. From the remaining 12, six patients had ileocolitis, three colitis and three ileitis. Their average age was 29.0 ± 6.4 years (mean ± s.d.).

Exclusion criteria were ileus, toxic megacolon with pending perforation or peritonitis, intestinal stenosis < 1 cm, pending surgery, planned colonoscopy or barium enema, pH-influencing drugs (cholestyramine, loperamide, antibiotics, etc.), missing written consent, patients younger than 18 years old, pregnancy, special diets (formula or elemental diet) and a diagnosis made less than 3 months earlier.

The patients were not restricted to any diet before or during the study. They were treated by the usual medications in relation to disease activity. Colonoscopies or barium follow-through examinations were performed to exclude patients with relevant bowel stenosis before pH-measurements.

Intraluminal pH was registered using a radiotelemetry capsule (type 7036, Oakfield Instruments, Eynsham, UK) and a portable solid state recording system, described elsewhere in detail.12 In short: after an overnight fast, the previously activated capsule was ingested by both the patients and controls. pH signals were received by the recorder via an antenna and shown on an LCD display. The individuals continued to fast until the capsule emptied from the stomach, which was signalled by a sharp rise in pH. The pH was continuously recorded while individuals identified the capsule by moving a locator over the abdomen indicating the point of highest signal intensity. This position was recorded on a body map divided into four sections over the abdomen (upper right and left, lower right and left quadrants). Subjects noted the quadrant in which the capsule could be localized at 1–2 hourly intervals during the study period. The position of the capsule on arrival in the caecum was always found to be in the right iliac fossa. The capsule than moved round the abdomen from the right to the left side during its passage through the colon.

The capsules were calibrated using pH 4 and pH 9.2 buffer solutions before ingestion and again after recovery.

The study was approved by both local ethic committees. Informed consent after thorough verbal and written information was obtained from each control person and each patient.

Statistical analysis was performed by using Wilcoxon–Mann–Whitney test.

The study was supported by the Deutsche Forschungsgemeinschaft, grant Pr 447/1–1.

RESULTS

Evaluable pH-measurements were obtained in 35 subjects. Four subjects had to repeat the measurements because of high data loss (> 75%), broken antenna and high pill drift in the first measurement cycle. There were a total of four measurements with > 75% data loss recordings that were excluded from statistical analysis. Postcalibration values were within the allowed limit (0.5 units) in 28 cases, one pill was lost with faeces and in another case the antenna was broken, thus postcalibration was not possible.

Results from patients with ulcerative colitis ( Table 1 and Figure 1)

Table 1.  . Median intraluminal gastrointestinal pH values in patients with inflammatory bowel diseases and controls Thumbnail image of
Figure 1.

.  Gastrointestinal pH profile from patients with ulcerative colitis. Medians are depicted. Bold lines and squares indicate patients with active disease, thin lines and squares indicate patients in remission. pH thresholds 6 and 7 are marked. prox = proximal; sb = small bowel; term = terminal.

Patients with active disease had similar median gastrointestinal pH values as patients in remission.

The median pH values in controls ( Table 1; Figure 3) compared to the patients showed significant differences in the stomach (P = 0.02), the terminal ileum (P = 0.002), in the caecum (P < 0.0001), and in the right colon (P = 0.0002).

Figure 3.

. Gastrointestinal pH profile from controls. Medians are depicted. pH thresholds 6 and 7 are marked. prox = proximal; sb = small bowel; term = terminal.

The data were analysed, and the number of patients which had surpassed the pH threshold of 6 or 7 in the most important intestinal segments (the terminal ileum and colon), where the mesalazine must be delivered, were analysed. No patient had values below pH 7 in the terminal ileum (one control subject). Six patients (three in remission and three with active disease) had values between pH 6 and 7 in the caecum (three controls), none were under pH 6 (five controls had pH < 6). Four patients had pH < 7 in the right colon (all 12 controls), none had < 6 (seven controls). All but one patient had pH < 7 in the left colon (as did all the controls), only one patient < 6 (two controls).

The patients’ median pH values in the proximal small intestine were comparable to the controls’ (= 0.93), as were the pH values in the left colon (= 0.24).

A statistical analysis of values obtained from the left colon was not possible because the number of evaluable results was too small.

Taking the results together, no significant luminal acidification was evaluated in cases of active intestinal inflammation.

Results from patients with Crohn’s disease ( Table 1 and Figure 2)

Figure 2.

.  Gastrointestinal pH profile from patients with Crohn’s disease. Medians are depicted. Bold lines and squares indicate patients with active disease, thin lines and squares indicate patients in remission. pH thresholds 6 and 7 are marked. prox = proximal; sb = small bowel; term = terminal.

Again, patients with active disease had similar median gastrointestinal pH values as patients in remission.

The median pH values in controls ( Table 1; Figure 3) compared to the Crohn’s disease patient group showed significantly differences in the stomach (P = 0.0043).

The patients’ median pH values in the proximal small intestine (P = 0.89), in the terminal ileum (P = 0.65), in the caecum (P = 0.41) and in the right colon (P = 0.91) were comparable to the controls.

Four patients (two with active disease, and two in remission) had pH values < 7 in the terminal ileum (one control subject), but only one patient (active disease) had a median pH < 6.8. Five patients had values between pH 6 and 7 in the caecum (three controls), five were under pH 6 (five controls). All patients had pH < 7 in the right colon (all 12 controls), five had pH < 6 (seven controls). All but one patient had pH < 7 in the left colon (as did all the controls), two patients had pH < 6 (two controls).

The median pH value measured in the segments of the left colon was 6.2, range 5.43–7.95. The individuals with active disease had very high data losses and therefore could not be estimated. Data obtained from only the patients in remission showed no differences to the controls (= 0.91).

Overall, no significant luminal acidification was evaluated in cases of active intestinal inflammation.

DISCUSSION

According to the luminal capsule-derived pH measurement of this study, patients with active Crohn’s disease and ulcerative colitis had similar gastrointestinal pH values compared to patients in remission. Therefore, intestinal inflammation does not decrease the pH in the intestinal lumen.

One-third of patients with Crohn’s disease (equal numbers with active disease and in remission) had pH < 7 in the terminal ileum, in comparison to one control subject. However, the medians were close to pH 7. Only one patient had pH 6.2. In the caecum, 41% of patients had values between pH 6 and 7 (25% from the controls), another 41% under pH 6 as the controls.12

Fewer patients with ulcerative colitis than controls had pH values below 6 or 7 in the terminal ileum or colon. In contrast to these results, previous findings suggested a decrease in intraluminal pH in the right colon and in the faeces of patients with ulcerative colitis, particularly in the presence of severe inflammation.7[8]–9, 13 The explanation for the even higher luminal pH values found in this study remains speculative. An underlying systematic error seems unlikely, because the control group pH values fit perfectly to the values from the literature. Another reason could lie in nutritional abnormalities: a low fibre diet is frequently recommended for patients with symptoms,14 and a low fibre intake was found in patients.15 Therefore, the known impairment of luminal short-chain fatty acids (SCFA) in active colitis16, 17 may be further increased, which might lead to less colonic acidification. On the other hand, a low luminal SCFA concentration in the colon is accompanied by an increase in luminal lactate production by changing metabolic pathways from colonic strains in cases of SCFA deficit.18 The resulting low pH could have been antagonised by the higher buffer capacity of the blood and mucus which is present, even in the proximal colonic segments in pancolitis. As most of the patients had pancolitis, this could explain the higher pH level in the right colon.

The slightly elevated pH (0.5–1 log units) in the terminal ileum of ulcerative colitis patients seems of minor relevance considering the limitations of our methods (a pill drift 0.5 pH unit was allowed), and the well-known, large variations in pH measurements.19 The same holds true for results obtained using the gastric pH of patients (see above). More studies are needed to confirm the elevated pH values in the right colon of patients with ulcerative (pan)colitis.

In addition, elevated lactate concentrations were found in the faecal water of patients with Crohn’s colitis, but were less pronounced.7 The lower lactate elevation in Crohn’s colitis might be too small to fully influence the luminal pH in the colon, as shown by our data.

Higher gastric pH values were found in the patient groups. Interestingly, hypochlorhydria was registered in patients after colectomy with ileorectal pouch anastomosis who developed pouchitis.20 The reasons for the apparent gastric hypochlorhydria are not known, but the median difference did not exceed one pH unit and therefore might be neglected.19 Further studies are needed to evaluate the possible pathophysiological consequences of ulcerative colitis and Crohn’s disease per se and the influence of surgical intestinal resection or reconstruction on gastric chlorhydria.

Several technical difficulties were observed during this study. Previous in vitro experiments suggested that artificial low pH values were found when the capsule exceeded the distance to the antenna for optimal data reception, before a ‘low signal’ was registered by the recorder. For this reason, we excluded these patient data from statistical analysis. Identifying an incorrect intraluminal pH in the colon (data loss > 75%) might suggest low pH values. The reason for the unpredictable data from the colon might be that mucus and compact stool components could obstruct the pH-sensitive area of the capsule. Sometimes, no data at all were received from the colon. From this point of view, the conclusions of Fallingborg et al. about the very low colonic pH values in severe ulcerative colitis are doubtful.9

As long as no refined methods are available for measuring colonic pH, intraluminal pH measurements are a matter of scientific investigation, which is not practicable in routine use.

It is concluded from this study that active intestinal inflammation does not decrease the pH in the intestinal lumen of patients with inflammatory bowel disease. pH 6 has been achieved in the terminal ileum in every patient, but pH 7 is missed by one-third of the patients with Crohn’s disease. The luminal release of 5-amino salicylic acid from mesalazine preparations might not be inhibited by active inflammation in those patients.

ACKNOWLEDGEMENTS

Financial support for this study was received from the Deutsche Forschungsgemeinschaft.

Ancillary