Source: Clinical Gastroenterology and Hepatology 2014;12:1514-1521

Comment

Colonic cleansing with sodium phosphate (SPH) preparations is effective, however, an increased risk of acute kidney injury (AKI) has been reported, secondary to the abrupt increase in blood phosphate levels. In healthy volunteers, it was found that about 50% of the preparation is absorbed in the intestine and 14% is excreted by the kidneys. The retained phosphate could have systemic effects by being deposited in different tissues after binding with calcium. It is not yet clear what consequences this deposit could have, especially in populations considered to be at risk for kidney damage. Most of the published works are inconclusive, or have a low power as a major limitation, secondary to an insufficient number of patients. On the other hand, they do not discriminate between those who undergo studies on an outpatient or inpatient basis, which implies populations that are not always comparable. The FDA advised against the use of FTS in patients with kidney disorders in 2008.

This study is a retrospective cohort, from a secondary database, which listed the reasons for patient care demand towards their health coverage plans in the United States. Patients between 50 and 75 years of age who had undergone surveillance video colonoscopy (CCV) on an outpatient basis from January 2000 to November 2008 were analyzed. The exposure period was considered to be 30 days prior to the CCV, period during which colonic preparation had been prescribed, either with FTS or with polyethylene glycol (PEG). Follow-up began on the day of the CCV and continued for 6 months. Those patients who had presented acute kidney injury, kidney failure of unknown etiology, rhabdomyolysis, on dialysis or in pre-transplant evaluation were excluded. The main event was acute kidney injury (AKI), defined according to the codes of the nomenclator of the 9th revision of the International Classification of Diseases. The covariates considered, in addition to the demographic data of the population, were risk factors for kidney damage, regular medication and that started during the exposure period. Hazard ratios (HR) for the development of AKI were estimated using a Cox proportional hazards regression model. A propensity score for treatment was generated, using all potential confounders, and the analysis was matched to detect inequalities between groups. A sensitivity analysis was performed in which patients operated on before 2005 (no reports of kidney damage due to FTS) and between that year and 2008 (the FDA advised against the use of FTS in patients at risk of AKI) were evaluated separately. ).

A total of 550,696 patients were analyzed, 429,430 had received PEG and 121,266 FTS. There was a slight predominance of women in both groups. In the group that received PEG, the mean age and drug use were higher, as was the prevalence of co-morbidities, such as diabetes, chronic renal failure, hypertension, and cardiovascular disorders. The frequency of hypercalciuria and recent kidney stones was similar in both groups. The mean follow-up was 170.7 days (SD 32 d). A total of 1,595 AKI episodes were recorded, 241 (0.2%) in the FTS-prepared group and 1,354 (0.3%) in the PEG-prepared group. Other events such as kidney failure or dialysis requirement did not show differences between both groups. The crude HR of AKI for the FTS group compared to the PEG group was 0.63 (95% CI 0.55-0.72) and the HR adjusted for comorbidities (CKD, >60 years, diabetics, stones, hypercalciuria, NSAIDs and with thiazide treatment) was 0.86 (95% CI 0.75-0.99). The analysis using the propensity score showed that both groups, FTS and PEG, are comparable. The matched HR was 0.85 (95% CI 0.72-1.01). Likewise, the sensitivity analysis by periods (until 2004 and from 2005 to 2008) did not show significant differences either.

Since previous evidence has not been conclusive regarding the possible risk of complications from FTS, and considering that conducting a clinical trial would be unethical given that the FDA has discouraged its use in this group of patients, this cohort has certain advantages over previously published studies. It analyzes patients exclusively from the outpatient setting and defines the follow-up period. It is innovative, by using a propensity score that makes it possible to homogenize both groups and make them comparable. This is a useful tool in observational studies, since it allows estimating a causal effect of the treatments studied. It also performs an analysis adjusted for numerous risk factors, as well as concurrent medication.

Within limitations, the fact that the exposure factor (FTS or PEG) is determined by the dispensing records in pharmacies and that the detection of patients with ARI has been based on the coding of the social coverage pathologies stands out, which lowers the sensitivity of the data. Like all observational studies, there is the possibility that co-variables that could be directly associated with the ARI event have not been considered. The secondary database used may not be generalizable to other populations outside of the United States.

Based on the aforementioned results, this study did not find an increased risk of AKI associated with the use of FTS, even in the highest-risk subgroups.

Carried out by: Dr. María Julieta Argüero

Source: Gastrointestinal Endoscopy

Comment

Currently, the ASGE (American Society of Gastrointestinal Endoscopy) guidelines consider conventional or cold loop polypectomy as a high-risk procedure in anticoagulated patients, and therefore recommend the suspension of anticoagulation before the procedure is performed.

This recommendation is based on the available evidence that establishes that the risk of post-polypectomy bleeding in anticoagulated patients is approximately 10%. However, the risk of thromboembolic events after discontinuation of anticoagulation is approximately 3%.

In this study published in the journal Gastrointestinal Endoscopy by a Japanese group, the risk of conventional polypectomy versus cold loop polypectomy in anticoagulated patients without suspension of anticoagulation is raised as a question.

This is a prospective randomized study carried out in a single center in Japan, in which patients over 20 years of age who were anticoagulated and who presented polyps smaller than 10 mm in a screening colonoscopy were included for 10 months. Exclusion criteria were: pregnant women, ASA III or IV, obesity (>100 kg), allergy to propofol or insufficient preparation that prevented observation of at least 90% of the colon. The primary outcome was delayed bleeding (bleeding within the first 2 weeks of the procedure that would have required endoscopic intervention) and the secondary outcome was immediate bleeding (oozing bleeding during the procedure for more than 30 seconds that would have required hemostatic clip placement).

The estimated sample calculation was 53 patients in each group; however, an intermediate analysis was performed during the course of the study which showed a high rate of bleeding with conventional polypectomy (40%) and the study was stopped. Finally, 77 patients were prospectively included, of which 7 were excluded because they presented lesions larger than 10 mm. All patients were randomized, leaving 35 patients in the conventional group (CG) and another 35 in the cold loop group (GF). Both groups were similar with respect to demographic characteristics.

In the GF group 78 polyps were resected with an average size of 6.5 mm and in the CG group 81 polyps were resected with an average size of 6.8 mm. Regarding delayed bleeding, a significant difference was observed between both groups, being 14% (5/35) in the GF vs. 0% (0/15) in the GF; P=0.027. Immediate bleeding was 5.7% (2/35) vs. 23% (8/35) respectively; P=0.042. The Odds ratio for post-polypectomy bleeding in the GF was 6.5 (95% CI, 1.9-22.5). Regarding the histopathological study of the polyps, there was no difference in the rate of complete resection between the two groups. A statistically significant difference was observed in the presence of submucosal arteries in the samples analyzed: GC 47% vs GF 32%, P=0.049. Likewise, the presence of injured submucosal arteries was also higher in the GC: 22% vs 39%, P=0.023.
In conclusion, this study shows that cold loop polypectomy seems to be a useful and safe resource in anticoagulated patients, since immediate bleeding could be effectively controlled with hemostatic clips or other methods. As for delayed bleeding, it would appear to be much less than that estimated with conventional polypectomy, making it a safe method that could avoid the suspension of anticoagulation, thus avoiding possible thromboembolic events. A possible explanation for the differences observed between both techniques could be linked to the injury of submucosal arteries, a situation more frequently observed in conventional polypectomy. Although this is a small sample study, future larger and multicenter studies could be important to corroborate these results.

Carried out by: Dr. Ramiro C. González Sueyro

Original article:
Removal of small colorectal polyps in anticoagulated patients: a prospective randomized comparison of cold snare and conventional polypectomy.
Akira Horiuchi, MD,1 Yoshiko Nakayama, MD,1,2 Masashi Kajiyama, MD,1 Naoki Tanaka, MD,1 Kenji Sano, MD,3 David Y. Graham, MD.
Gastrointest Endosc. 2014 Mar;79(3):417-23.