No abstract

SUMMARY BackgroundDiarrhoea is a common presentation throughout the course of HIV disease. AimTo review the literature relating to aetiology, investigation and management of diarrhoea in the HIV-infected adult. MethodsThe PubMed database was searched using major subject headings 'AIDS' or 'HIV' and 'diarrhoea' or 'intestinal parasite'. The search was limited to adults and to studies with >10 patients. ResultsDiarrhoea affects 40-80% of HIV-infected adults untreated with antiretroviral therapy (ART). First-line investigation is by stool microbiology. Reported yield varies with geography and methodology. Molecular and immunological methods and special stains have improved diagnostic yield. Endoscopy is diagnostic in 30-70% of cases of pathogen-negative diarrhoea and evidence supports flexible sigmoidoscopy as a first line screening procedure (80-95% sensitive for CMV colitis), followed by colonoscopy and terminal ileoscopy. Radiology is useful to assess severity, distribution, complications and to diagnose HIV-related malignancies. Side effects and compliance with ART are important considerations in assessment. There is a good evidence base for many specific therapies, but optimal treatment of cryptosporidiosis is unclear and only limited data support symptomatic treatments. ConclusionsThe immunological response to HIV infection and antiretroviral therapy remains incompletely understood. Antiretroviral therapy regimens need to be optimised to suppress HIV while minimising side effects. Effective agents for management of cryptosporidiosis are lacking. There is an urgent need for enhanced regional diagnostic facilities in countries with a high prevalence of HIV. The ongoing roll-out of antiretroviral therapy in low-resource settings will continue to change the aetiology and management of this problem, necessitating ongoing surveillance and study.

No abstract

MRI assessment of the circumferential resection margin in rectal cancer is associated with high accuracy. Neoadjuvant chemoradiotherapy has a detrimental effect on this accuracy, although accuracy remains high. In the presence of persistently predicted positive margins, complete resection remains achievable but may necessitate a more radical approach to resection.

Background Autoimmune pancreatitis (AIP) is an uncommon form of chronic pancreatitis. Whilst being corticosteroid responsive, AIP often masquerades radiologically as pancreatic neoplasia. Our aim is to appraise demographic, radiological and histological features in our cohort in order to differentiate AIP from pancreatic malignancy. Methods Clinical, biochemical, histological and radiological details of all AIP patients 1997-2016 were analysed. The initial imaging was re-reviewed according to international guidelines by three blinded independent radiologists to evaluate features associated with autoimmune pancreatitis and pancreatic cancer. Results There were a total of 45 patients: 25 in type 1 (55.5%), 14 type 2 (31.1%) and 6 AIP otherwise not specified (13.3%). The median (IQR) age was 57 (51-70) years. Thirty patients (66.6%) were male. Twenty-six patients (57.8%) had resection for suspected malignancy and one for symptomatic chronic pancreatitis. Three had histologically proven malignancy with concurrent AIP. Two patients died from recurrent pancreatic cancer following resection. Multidisciplinary team review based on radiology and clinical history dictated management. Resected patients (vs. non-resected group) were older (64 vs. 53, p = 0.003) and more frequently had co-existing autoimmune pathologies (22.2 vs. 55.6%, p = 0.022). Resected patients also presented with less classical radiological features of AIP, which are halo sign (0/25 vs. 3/17, p = 0.029) and loss of pancreatic clefts (18/25 vs. 17/17, p = 0.017). There were no differences in demographic features other than age. Conclusion Despite international guidelines for diagnosing AIP, differentiation from pancreatic cancer remains challenging. Resection remains an important treatment option in suspected cancer or where conservative treatment fails.