Association between 6-thioguanine nucleotide levels and preventing production of antibodies to infliximab in patients with inflammatory bowel disease.

OBJECTIVE: Combination therapy with infliximab and a thiopurine has been shown to be more effective than monotherapy in patients with inflammatory bowel disease (IBD). The therapeutic efficacy of thiopurines is correlated with 6-thioguanine (6-TGN) levels between 235 and 450 pmol/8×108 erythrocytes. The primary aim of the study was to investigate the association between 6-TGN levels and inhibition prevention of the production of antibodies to infliximab (ATI). DESIGN: We performed a retrospective review of the medical records of patients being treated with infliximab for IBD at University Hospitals Bristol NHS Foundation Trust. Demographic and biochemical data were extracted, alongside thiopurine metabolite levels, trough levels of infliximab and the presence of ATI. χ2 tests were used to investigate the association between 6-TGN levels and prevention of ATI. Logistic regression was used to compare the odds of prevented ATI between those with a 6-TGN level between 235 and 450 pmol/8×108 erythrocytes, those with a 6-TGN level outside of this range, and the baseline group who were on infliximab monotherapy. RESULTS: Data were extracted for 100 patients. Six of 32 patients with a 6-TGN level between 235 and 450 pmol/8×108 erythrocytes developed ATI (18.8%) compared with 14 out of 22 (63.6%) patients with a 6-TGN outside of this range and 32 out of 46 (69.6%) patients on monotherapy (p=0.001). The OR (95% CI) for prevented ATI in those with a 6-TGN between 235 and 450 pmol/8×108 erythrocytes compared with a 6-TGN outside of this range was 7.6 (2.2, 26.3) (p=0.001) and compared with monotherapy was 9.9 (3.3, 29.4) (p=0.001). CONCLUSION: 6-TGN levels between 235 and 450 pmol/8×108 erythrocytes prevented production of ATI. This supports therapeutic drug monitoring to help guide treatment and maximise the beneficial effects of combination therapy for patients with IBD.

Liver steatosis is a risk factor for hepatotoxicity in inflammatory bowel disease patients treated with azathioprine.

BACKGROUND AND AIMS: The literature demonstrates that hepatic steatosis reduces the tolerance of immunosuppression in people with inflammatory bowel disease. It also shows that elevated methylmercaptopurine may be responsible for thiopurine-induced hepatitis. This exploratory study investigates the relationship between hepatic steatosis, methylmercaptopurine and alanine transaminase. METHODS: We performed a retrospective review of patients started on azathioprine treatment at University Hospitals Bristol NHS Foundation Trust between 2014 and 2017. There were 600 patients in total. Ninety-one patients met our inclusion criteria which were at least one ultrasound scan commenting on the appearance of the liver, liver function tests at commencement of azathioprine and liver function tests and a methylmercaptopurine level between 6 and 8 weeks after starting treatment. RESULTS: Of 91 patients included in our study, 32 patients (32%) were identified as having radiological hepatic steatosis on ultrasound imaging and 59 patients had no evidence of steatosis. We found a positive association between methylmercaptopurine levels and change in alanine transaminase in patients with hepatic steatosis (P < 0.001) but not in those with a normal liver on ultrasound imaging. CONCLUSION: We conclude that the higher levels of methylmercaptopurine may be a risk factor for hepatitis in patients with hepatic steatosis but not in those with a normal liver.

Needs-based quality of life in adults dependent on home parenteral nutrition.

BACKGROUND & AIMS: Home parenteral nutrition (HPN) provides life sustaining treatment for people with chronic intestinal failure. Individuals may require HPN for months or years and are dependent on regular intravenous infusions, usually 12-14 h overnight between 1 and 7 days each week. This regime can have adverse impact on the life of people dependent on the treatment. The aim of this study was to establish mean values for the Parenteral Nutrition Impact Questionnaire (PNIQ) and to determine the effect of disease, frequency of infusions per week and patient characteristics on quality of life of patients fed HPN. METHOD: The PNIQ was distributed to patients across nine UK HPN clinics. Data were analysed using linear regression, with PNIQ score as the dependent variable and potential confounders as independent variables. Unadjusted and adjusted models are presented. Higher PNIQ scores reflect poorer quality of life. RESULTS: Completed questionnaires were received from 466 people dependent on HPN. Mean PNIQ score was 11.04 (SD 5.79). A higher PNIQ score (effect size 0.52, CI 0.184 to 0.853) was recorded in those dependent on a higher frequency of HPN infusions per week. Respondents with cancer had a similar mean PNIQ score to those with inflammatory bowel disease (mean 10.82, SD 6.00 versus 11.04, SD 5.91). Those with surgical complications reported a poorer QoL (effect size 3.03, CI 0.642 to 5.418) and those with severe gastro-intestinal dysmotility reported a better QoL (effect size -3.03, CI -5.593 to -0.468), compared to other disease states. CONCLUSIONS: This large cohort study of quality of life in chronic intestinal failure demonstrates that HPN impacts individuals differently depending on their underlying disease. Furthermore, since the number of HPN infusions required per week is inversely related to an individual's needs-based quality of life, therapies that reduce PN burden should lead to an improvement in QoL.

Can Quantab titrator sticks reliably predict urinary sodium?

BACKGROUND & AIMS: Urinary sodium concentration is a commonly used marker for extracellular fluid depletion which is often associated with dehydration. A point of care test for urinary sodium may reduce delays in clinical decision making by offering more timely guidance leading to improved salt and fluid management. We compared laboratory assessed urinary sodium with a potential point of care measure of urinary chloride in a variety of in- and outpatient specialities, to explore its use as an indicator of low urine sodium. METHODS: Urinary chloride concentrations were estimated using a Quantab titrator stick in samples from patients that had been sent for urinary sodium assays. We validated the results of this titrator stick with laboratory-assessed sodium concentrations by deriving correlation coefficients between these methods and using limits of agreement testing. We determined the optimal titrator stick cut-point for identifying low urinary sodium (urinary sodium <20 mmol/L) by maximising the product of the sensitivity and specificity. This level of urinary sodium was used to mirror the British Society of Gastroenterology guidance on short bowel patients Nightingale and Woodward, 2006. RESULTS: We obtained laboratory urinary sodium concentration and Quantab stick chloride measures on 127 samples. Twenty three percent had a urinary sodium below 20 mmol/L so were regarded as biochemically dehydrated. A threshold of <4.3 on the Quantab scale had a positive predictive value for low sodium of 56% (95%CI 40%-71%) and a negative predictive value of 94% (95%CI 87%-98%). CONCLUSIONS: These data suggest that the Quantab stick could be used as a point of care test to aid fluid and salt management decisions in an outpatient setting. Further work to explore the use of the titrator stick in specific patient populations at risk of salt and water depletion is justified.

Communication between primary and secondary care.

Background Up to date patient data is a cornerstone of optimal safety and care, so admission to hospital requires transfer of data held in the community to secondary care. Despite the advent of electronic medical record systems such as Connecting Care and EMIS, the telephone remains the mainstay of communication. Methods A prospective cross-sectional quality improvement project was conducted to assess the time taken in telephone communication between primary and secondary care doctors and determine if access to the electronic shared system, Connecting Care, would improve efficiency. As part of normal junior doctor activity, fifty GPs were contacted between September 2015 and February 2016 to obtain medical data on patients admitted to UH Bristol. Time taken to contact each GP and the duration of the conversation was recorded. One hundred patient records were accessed using Connecting Care between October 2015 and February 2016 and the length of time taken to access information documented. Results Out of 50 phone calls 27 resulted in direct transfer to a GP with time to transfer ranging from 12 seconds to 19 minutes 51 seconds (mean 8 minutes 10 seconds, median 7 minutes 16 seconds). A total of 28 messages were left with the receptionist with 16/28 phone calls being returned and time taken for the call to be returned ranged from 34 minutes to 21 hours 3 minutes (mean 5 hours 50 minutes 4 seconds, median 4 hours). Information was available for 88/100 patient records accessed using Connecting Care with a mean duration to access information required of 1 minute 47 seconds. This was significantly shorter than the mean duration of conversation with GPs (4 minutes 22 seconds), mean total duration of telephone call and mean total duration of time to achieve aim of call with GP practices of 13 minutes 18 seconds and 2 hours 14 minutes 11 seconds respectively. Conclusions This study identifies areas of potential improvement in current methods of communication between primary and secondary care. Direct telephone contact will always have an important role in sharing information. However, access for secondary care doctors to electronic patient records, with patient consent and consideration for confidentiality, would improve efficiency and alleviate time pressures on both busy primary and secondary care doctors. This would have a positive impact on patient care and safety.

In recurrent C. difficile, the CRP response to the primary C. difficile infection predicts whether the same strain or a different strain will cause a second infection.

BACKGROUND: Clostridium difficile is the leading cause of antibiotic-associated diarrhoea and is associated with an increase in morbidity and mortality. There is a wide variance in disease severity with some patients suffering a single, self-limiting episode of diarrhoea while others suffer more intractable problems with recurrent attacks or toxic dilatation. Numerous different C. difficile ribotypes exist, some of which are considered hypervirulent. The magnitude of toxin production alone is not sufficient to explain the varying virulence of these ribotypes, suggesting the involvement of other mechanisms. METHODS: To test the same patient's response to infection with different C. difficile ribotypes, we reviewed 45 patients who suffered two episodes of C. difficile infection and determined by ribotyping and MLVA whether the second episode was due to the same strain or a different strain. RESULTS: Patients harbouring a different strain had significantly higher C-reactive protein (CRP) responses on the first assessed infection (143 mg/L ± 20 vs. 55 ± 9.63, p = 0.0001) and a significantly lower CRP on reinfection (p = 0.048). Same strain patients had a non-significant increase in CRP response on second infection. CONCLUSIONS: This suggests that the inflammatory response to C. difficile is determined by an interaction between host immunobiology, previous exposure and C. difficile strain.

C-reactive protein is a useful predictor of metronidazole treatment failure in mild-to-moderate Clostridium difficile infection.

OBJECTIVE: To assess the ability of C-reactive (CRP) protein, against the other commonly used metrics, to predict metronidazole treatment failure in Clostridium difficile infection. METHODS: We retrospectively reviewed the case notes of 65 patients with C. difficile infection initially treated with metronidazole. Patients were grouped on the basis of outcome: those who responded to metronidazole within 6 days (cut-off as used by previous authors) versus those who required vancomycin. Individual predictor variables were examined between groups (using a t-test, Kruskal-Wallis test, or Fisher's exact test), and the strength of associations was assessed by logistic regression. RESULTS: Of the 65 patients reviewed, 48 (74%) resolved with metronidazole alone. Regression analysis found that (CRP) white cell count and creatinine levels were significantly different across the metronidazole success/failure groups (P<0.01, P=0.01 and P=0.03, respectively). CONCLUSION: (CRP) is a useful predictor of metronidazole treatment failure in mild-to-moderate C. difficile infection.