Delay in diagnosis of colorectal cancer (CRC) is associated with worse outcomes. Although studies have shown the incidence of CRC missed at endoscopy to be 2.5%-7.7% (Morris et al., 2014), there are additional non-endoscopic factors that may lead to delays. This study aimed to identify all factors leading to a delayed diagnosis, including endoscopic "misses".
Methods
All patients diagnosed with CRC at Kings College Hospital, London between 2011 – 2018 were included. We identified patients seen in an outpatient clinic or underwent endoscopy within 36 months preceding diagnosis. 'Delayed' cancers were grouped into 'clinical factors' and 'technical factors'. 'Clinical factors' included the subset of post-colonoscopy colorectal cancers (PCCRC). The Joint Advisory Group on GI endoscopy (JAG) have defined PCCRC as being cancers diagnosed within 36 months of an endoscopy.
Results
797 cases of CRC were diagnosed in the study period and 60(7.5%) were seen in the preceding 36 months. 46 patients (5.8%) were determined to have a delayed diagnosis, of which 24(52.2%) were diagnosed within 1 year of initial investigation. 38 delayed diagnoses were due to clinical factors: PCCRC (n=23), incomplete endoscopy (n=2), inadequate investigation (n=7), missed on rigid sigmoidoscopy (n=1) and incomplete bowel preparation (n=5) with an average delay of 172 days. 8 delays were caused by technical factors: Incomplete follow up (n=4), delayed investigation (n=2), histology not reviewed (n=1), missed on CT (n=1). The rate of missed cancer in the 2WW, Bowel Cancer Screening and Routine referral pathways was 4.8% (n=14), 5.1% (n=8), 7.3% (n=28) respectively. The incidence of missed cancer in the right colon was significantly higher (p=0.068, 95% CI 0.9–.57). Further interrogation showed the highest incidence in the hepatic flexure (10.5%), splenic flexure (9.4%), caecum (7.5%), and anal canal (6.5%). 91.3% of PCCRCs versus 47.8% of other missed cancers identified another pathology at the initial endoscopy which was documented on the report (p=0.003, 95% Confidence interval 2.1–0.6). A delay in diagnosis was not associated with more advanced TNM staging or K-ras mutations.
Conclusions
Our rate of missed cancers is equivalent to that published in the literature, with the majority of missed cancers due to PCCRC. Delays in diagnosis are related to avoidable factors, such as improving the quality of bowel prep or ensuring further investigations in patients with incomplete endoscopies. Endoscopists should also be aware of the increased miss rate in specific locations. Non-cancer pathology identified at endoscopy is also associated with missed cancer, so endoscopists should be careful to maintain a careful examination for concomitant cancers.
Vedolizumab was recently granted NICE approval for moderate-to-severe Crohn’s disease (CD) and ulcerative colitis (UC). Novel pathways agreed by our CCG meant that patients at Guy’s & St. Thomas’ and King’s College Hospitals had early access to vedolizumab.
Methods
Records of patients commencing vedolizumab between Nov 2014–15 were screened. Those completing at least 14 weeks of treatment were included. Clinical activity was assessed using Harvey-Bradshaw Index (HBI) or Simple Clinical Colitis Activity Index (SCCAI) at baseline, 14 and 30 weeks. Response: HBI/SCCAI reduction ≥3. Remission: HBI < 5 or SCCAI <3. Continuous data are summarised as medians (range). Pre- and post-induction values were compared using Wilcoxon signed-rank test.
Results
60 patients (CD: 32, 53%, UC 25, 42%, IBD-U 3, 5%) commenced vedolizumab (m:f 29:31, age: 39 (18–74), follow-up: 5 months (1–13)). 19 were excluded from our analysis (3 IBD-U, 5 stomas, 11 treated for <14 weeks). Clinical data from the remaining 41 was analysed. Of 32 patients with active disease at baseline, 17 (53%) responded and 11 (34%) achieved remission by week 14. The response and remission rates for CD were 8/15 (53%) and 6/15 (40%). In UC they were 9/17 (53%) and 5/17 (29%). Response and remission rates in anti-TNF experienced patients were 12/26 (46%) and 6/36 (35%) compared to 5/6 (83%) and 5/6 (83%) in anti-TNF naïve patients, respectively. 7/11 (64%) with active disease at baseline who completed 30 weeks of treatment responded and achieved remission. Faecal calprotectin fell significantly (pre-induction: 1076 (90–4800), post-induction: 478 (10–3184), p = 0.029 for n = 14) and CRP remained stable (pre-induction: 4 (1–70), post-induction: 4 (1–72), p = 0.28 for n = 40). Rates of steroid use at each time point: 19/41 (46%) at baseline, 11/41 (21%) at week 14 and 3/15 (20%) at week 30. Surgery was required in 4/41 (10%, CD:3 and UC:1).
Conclusion
Our experience mirrors a previously reported real-world cohort1 and demonstrates similar efficacy to the GEMINI trials. This data demonstrates a meaningful reduction in clinical and biochemical disease activity as well as a steroid-sparing effect in patients with complex and previously refractory disease. We did not see a significant difference in efficacy between patients with UC and CD.
Reference
1 Christensen B. et al. Post-marketing experience of vedolizumab for IBD: The University of Chicago experience. ECCO; Barcelona, 2015.
Acute upper gastrointestinal bleeding (UGIB) is a common medical emergency that has a 10% mortality rate,1 requiring specialist input and management.2 We conducted a retrospective review last year which showed that the mean length of stay (days) was shorter in the GI group: 5.5 ± 5.7 vs 15.7 ± 20.8 (p = 0.02).3 We conducted a prospective analysis to assess if the above results held true.
Methods
A prospective review of case-notes (Electronic patient record-EPR) was conducted for all patients admitted to Kings College hospital with suspected UGIB between January and September 2013. Patients were divided as to whether they came immediately under the care of Gastroenterologists (GI) or general physicians (non-GI) after initial evaluation in the Acute Admission Unit. Patients were assigned on the basis of bed availability in a ward-based system. Statistical comparisons were made as appropriate with two tailed t-test or chi- squared test.
Results
138 patient episodes were reviewed of which 63 and 75 were treated by GI and Non-GI physicians. The two groups were broadly similar in their baseline characteristics. Mean length of stay (days) was significantly shorter in the GI group: 6.6 ± 5.6Vs 10.66 ± 11.3 (p = 0.006). Other comparators are shown in the table.
Conclusion
The length of stay of patients with UGIB is significantly shorter when receiving specialist care. In line with previous reports,4 we found that the incidence of UGIB was higher in males. Patients managed by GI physicians received less blood transfusion compared to the Non-GI physicians. The time to endoscopy was significantly shorter when receiving specialist care. Mortality rates in both groups compared favourably to the national average.
References
CG141 Scope for improvement: A toolkit for a safer Upper Gastrointestinal Bleeding (UGIB) service. www.bsg.org.uk Venkatachalapathy SV, Grasso N, Hayee B et al., Specialist care of in-patients with non-variceal upper gastrointestinal bleeding is associated with a dramatically shorter length of stay. Gut 2013;62:A10 doi:10.1136 Lanas A, García-Rodríguez LA, Polo-Tomás M et al., Am J Gastroenterol 2009;104:1633-41
Introduction There is still uncertainty about the role of faecal calprotectin (FCAL) as a marker of mucosal healing in inflammatory bowel diseases. Available studies have provided evidence on the prognostic role of FCAL in predicting relapses during a short period of time but there is a lack of data on associating FCAL levels over time to hard clinical endpoints. In this study we test the hypothesis that FCAL monitoring identifies Crohn’s disease (CD) patients with persistent intestinal inflammation requiring surgery. Method From a large IBD cohort of patients currently under follow up at King’s College Hospital, we identified all the CD patients who were diagnosed locally and had serial FCAL in the context of their routine care (at routine appointments and during flare-ups). Utilising prospectively kept electronic patient records we identified 20 patients, meeting these monitoring criteria, who required a bowel resection for CD≥12 months after diagnosis (cases) and matched them in a 1:2 ratio with controls based on disease duration. Flares were identified based on the physician assessment and endoscopic or radiological findings. Continuous variables are summarised as medians followed by interquartile range. The Fisher exact test was used to compare categorical variables, the Mann-Whitney test for continuous variables and the ROC curve for diagnostic analysis. Results Median time to surgery was 9.5 years (8, 11) [control group follow up: 8 (7, 10), p=0.28]. Right hemicolectomy was the commonest procedure (14, 70%) followed by panproctocolectomy (2, 10%), small bowel resection (2, 10%) and stricturoplasty (2, 10%). The two groups did not differ in clinical characteristics or time to initiation of a biologic or immunosuppressants. Flares and hospitalisations were more common in the cases group [ 20 (100%) vs. 23 (56%), p=0.0005 and 19 (95%) vs. 19 (48%) p=0.0002, respectively]. At the time of diagnosis, there was a numerical difference in median FCAL between groups [cases: 652 ug/g (168, 1020) vs. controls: 304 ug/g (120, 750), p=0.2]. One year after the diagnosis the median FCAL (FCAL1) was higher in the cases group [ 549 ug/g (152, 1115) vs. 68 ug/g (26, 184), p= Conclusion FCAL is a marker of mucosal healing in CD. Frequent monitoring identifies patients with clinically meaningful levels of intestinal inflammation associated with flares, hospitalisations and surgery. Disclosure of Interest None Declared
Background: Faecal calprotectin (FCAL) is a useful test for monitoring of inflammatory bowel disease (IBD) activity. However, providing a stool sample in person to the hospital laboratory is anecdotally unpopular. A new FCAL kit (IBDoc™, Bühlmann) enables self-testing using a proprietary collection tube, camera smartphone and app. The aims of this study were to assess patients' adherence to and experience of using IBDoc™; to compare the assay to the standard laboratory test; and to determine if IBDoc™ can be used to predict a flare of disease within a four month period. Methods: After focussed training, participants were asked to use IBDoc™ once a month for four months and provide a standard stool sample to be tested with standard ELISA (Bühlmann). The following questionnaires were applied before and after testing: GAD-7 (anxiety), PHQ-9 (depression), IBD-control-8, Multi-dimensional Health Locus of control (MHLC) and Cognitive Behavioural Responses to Symptoms (CBSRQ). Patients were also asked to record their experiences and preferences for testing on a proprietary questionnaire. Electronic patient records and endoscopy and histopathology reports were retrospectively reviewed for patients who had FCAL results by both methods at one time point. A faecal calprotectin of >100 μg/g was defined as a positive result. Results: Overall, 54 patients (Crohn's: 23, UC: 31, mean age 36.0) were enrolled. Participants completed a median of 3 tests during the study with 19/54 (35%) completing all four set time points and 17/54 (32%) returning no samples. There was no difference in any of the questionnaire scores between compliant and non-compliant patients. Overall, 85% of respondents stated a preference for IBDoc™ of which 74% would want this to be in the context of prompt contact from the hospital team in the event of a positive result. There was moderate correlation of FCAL results between the two methods (r=0.77, p<0.0001). At least one paired laboratory FCAL and IBDoc™ result was available for 37 patients, of which 30 were in remission at the time of the test. To predict a flare within four months, the IBDoc™ FCAL had a sensitivity of 89%, specificity of 33% and NPV of 87.5%, compared with 78%, 57% and 86% respectively for the laboratory test. Conclusions: There was reasonable uptake and adherence to a demanding testing regimen with 85% of respondents preferring the IBDoc™ test over other methods. The home testing kit results show only moderate correlation to laboratory results. A negative FCAL (<100 μg/g) by either method is a useful test to exclude a flare within four months, but positive results should be interpreted with caution and repeat testing would be advisable prior to treatment escalation.
The early identification of primary non-response to anti-TNFα therapy facilitates the timely management of patients with inflammatory bowel disease (IBD). A recent, pilot study to detect prognostic markers of early response to anti-TNFα therapy identified the two genes coding for the calprotectin subunits (S100A8, S100A9) to be among the most highly expressed gene transcripts in non-responders. This study tests the hypothesis that measurements of faecal calprotectin (FCAL) pre- and post- anti-TNFα induction can predict primary non-response in both Crohn’s disease (CD) and ulcerative colitis (UC)
Methods
Retrospective study of 32 CD and 18 UC patients treated over a two-year period. Outcomes were assessed using Harvey-Bradshaw Index (HBI) or Simple Clinical Colitis Activity Index (SCCAI) (response: drop in HBI/SCCAI >3, remission: HBI < 5, SCCAI <3, steroid free) at 6 months. ΔFCAL was calculated as (FCALpost induction – FCALpre induction)*100/FCAL pre induction.
Results
At 6 months, 23 (72%) CD and 10 (56%) UC patients had responded. In remission were 17 (53%) and 5 (28%) respectively. Comparing non-responders to combined response and remission groups, the area under the curve of ΔFCAL to predict outcomes at 6 months was 0.97 for CD and 0.96 for UC. Using ROC analysis, a decrease of 70% returned a sensitivity and specificity of 99% and 96%, respectively (likelihood ratio, LR = 23) in CD. For UC a a decrease of 70% had a sensitivity and specificity of 88% and 86%, respectively (LR = 6).
Conclusion
A drop in FCAL < 70% after induction predicts primary non-response to anti-TNFα in both CD and UC.
Colonoscopy remains the gold standard procedure for screening and polyp detection, with adenoma detection rate (ADR) being a widely-accepted key performance indicator (KPI). It has long been recognised that even experienced endoscopists incur an appreciable ‘miss-rate’ and a number of novel devices have been marketed to assist this aspect of practice. The Endorings™ device is a simple soft silicone, single-use device consisting of a series of rings arranged around a central tubular core. As the colonoscope is inserted the rings fold backward to allow intubation and flare on withdrawal to flatten colonic folds and aid inspection.
Methods
Prospective data was collected during screening colonoscopy (performed by two accredited colonoscopists) when the Endorings™ device was used and compared the KPI results to outcomes from the previous few months, for the same two colonoscopists) when the device was not in use (ie. historical controls).
Results
The ADR without Endorings™ (n = 85) was 49.4% with a per-procedure detection rate (ppr) of 0.97. With Endorings (n = 66) 66.7% (p = 0.0006) with ppr of 1.625. This represents a 35% increase in ADR and a 68% increase in the number of polyps detected at any given procedure. There were no significant differences in completion rates, withdrawal time, use of sedation or comfort scores. The device was removed in 5/66 procedures due to interference with intubation (in the presence of either an angulated sigmoid or diverticulosis). No complications were recorded.
Conclusion
Use of the Endorings™ device was associated with a significant increase in ADR. Qualitatively, the three-ring design was felt to interfere with normal intubation such that insertion technique had to be modified. An updated design iteration with two rings in slightly different positions along the central tube, has been produced and appears to offer a significant advantage in this regard. Furthermore, the central tube can be pushed further along the distal end of the colonoscope to allow the terminal ileum to be intubated with the device in place. The Endorings™ may offer an advantage in screening colonoscopy and, in this cohort, further prospective investigation is warranted.
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)