High detection rate and risk factors of melanosis coli in end-stage renal disease patients awaiting kidney transplantation: a single-center retrospective study in China

Abstract

Background

Rhubarb-containing Traditional Chinese Medicine (TCM) is widely used to manage renal failure in China. Studies show that anthranoids contained in rhubarb are related to the occurrence of melanosis coli (MC). However, no study on the rate of MC in end-stage renal disease (ESRD) patients has been reported.

Methods

We conducted a retrospective study on 99 ESRD patients who received a colonoscopy as part of pre-transplant work-up between Jan 1, 2017 and July 1, 2020 in a single center in Beijing, China. 200 age- and gender-matched subjects with normal renal function and received a colonoscopy for health screening in the same period were randomly selected as control group. Univariate analysis and multivariate logistic regression were performed in succession to determine the association between ESRD and MC, and also probe the independent MC risk factors in ESRD patients.

Results

Among the 99 ESRD patients, MC was detected in 19, with a detection rate of 19.2% (95% confidence interval [CI]: 11.3%, 27.1%). In contrast, among 200 control subjects, MC was only detected in 1 subject, with an detection rate of 0.5% (95%CI: 0.5%, 1.5%) (ESRD vs. control: odds ratio [OR] = 47.263, 95% CI: 6.222, 358.981; p=0.000). After adjustment of age, body mass index(BMI), albumin, hemoglobin and Charlson comorbidity index, being ESRD was still associated with the presence of MC (OR = 36.251, 95%CI: 3.435, 382.559, p = 0.003). In ESRD patients, both univariate analysis (OR = 4.358, 95% CI: 1.329, 14.286; p = 0.012) and multivariate logistic regression (OR = 4.916, 95% CI: 1.462, 16.525, p = 0.010) found that past use of Rhubarb-containing TCM was the only independent risk factor of MC. However, when the ESRD patients with above use history were excluded, MC rate in ESRD was still higher than non-ESRD group (8.5% vs. 0.5%; OR = 18.512, 95% CI: 2.019, 169.755; p = 0.000).

Conclusion

Our work highlights that MC rate is high and rhubarb-containing TCM use might be a risk factor in Chinese ESRD patients. There is a need for arranging high quality studies to examine the development of MC in ESRD, especially those awaiting transplantation.

Introduction

Chronic kidney disease (CKD) has a high prevalence worldwide, and its prognosis is far from optimal, with renal failure as a frequent outcome, thus shapes a global healthcare challenge. In China, CKD has an overall prevalence of 10.8%, and the prevalence of stages 3 and above CKD is 1.7% ^[1]. Increasing loss of renal function is commonly unavoidable in moderate-to-severe CKD, when the estimated glomerular filtration rate (eGFR) is within 15–59 ml·min^-1·1.73 m^-2. The management measure of CKD in this stage, including diet restriction plus keto acid supplementation ^[2], angiotensin-converting enzyme inhibitor (ACEI) and angiotensin II receptor blocker (ARB) ^[3] etc., is usually somewhat useful but cannot reverse the final fate of CKD. The patients routinely advance into end-stage renal disease (ESRD) and need maintaining replacement therapy though great efforts have been made.

The incurable nature of middle and advanced-stage CKD leaves a place for Traditional Chinese Medicine (TCM) in China. The concept of CKD treatment (≥3 stage) in TCM includes supplementing qi, nourishing and activating blood, removing dampness and turbid toxin, clear the triple energizer and so on ^[4]. Hundreds of TCM herbs and numerous patent TCM medicines are widely used to treat chronic renal failure ^[5]. Rhubarb (in Chinese, Dahuang), prepared from the root of the Rheum palmatum L plant, is one of the most commonly prescribed herbs, and also is one of the main ingredients in majority of these patent TCM medicines ^[5]. In the concept of TCM, rhubarb has the effect of purgation, anti-stasis, astringent and so on ^[6]. Over 20 kinds of anthraquinones exist in rhubarb, including emodin, rhein, and aloe-emodin etc., which have been explored extensively in China ^[11]. In plenty of clinical studies, Rhubarb was found effective in slowing the progression of CKD, abating uremic symptoms, decreasing serum creatinine, raising hemoglobin levels, and adjusting lipid metabolism ^{[5, 8, 9]}.

However, anthranoids contained in rhubarb have been linked to colonic toxicity, including the induction of melanosis coli (MC) ^{[10, 11]}. First described by Cruveilhier, in 1829, and named by Virchow in 1857, MC is a so called benign colonic condition featured by non-specific light to dark brown lipofuscin-like pigments in macrophages of colon’s lamina propria ^{[12, 13]}. Lysosomes use apoptotic cells ingested by marcrophages and then transported into the lamina propria to generate lipofuscin pigment, is thought to be MC’s formation mechanism ^{[14, 15]}. In recent decades, the detection rate of MC has been progressively increased together with the aging of the people, the increasing popularity of constipation and advances in colonoscopic diagnosis ^{[12, 16]}. Factors associated with MC include female, elder, constipation and use of anthranoid laxatives ^{[16, 17]}. The detection rate of MC in China is believed lower than that in Western nations ^[16]. Investigations showed that MC may not be harmless, but a mark of chronic damage of colonic and intestinal mucosa ^[12]. MC is associated with an increased detection rate of adenoma, although the oncogenic effect of MC is still a subject of debate ^{[12, 18]}. A few studies have reported a link between colorectal cancer and MC ^{[19, 20]}, however, other studies have shown no association ^{[21, 22]}.

To our knowledge, no study concerning the detection rate and risk factors of MC in ESRD patients in China has been published till now, though rhubarb or rhubarb-containing TCM is widely used during the treatment of renal failure. During Jan 1, 2017 to July 1, 2020, we performed screening colonoscopy for some ESRD patients awaiting kidney transplantation. This study examined the detection rate of EC in these patients, compared it with that of a group of age- and gender-matched subjects without renal failure, and examined the risk factors of the development of EC in these patients.

Materials And Methods

Study participants

We performed a retrospective study of 99 ESRD patients who underwent colonoscopy as part of pre-transplant work-up in the eighth medical centor of PLA General Hospital, Beijing, China, between Jan 1, 2017 and July 1, 2020. These patients came from different places of China, including Beijing (16 patients), Inner Mongolia (15), Heibei (11), Shanxi (11), Heinan (7), Heilongjiang (6), Tibet (4), Anhui (4), Liaoning (3), and other 13 provinces (22). The inclusion criteria were as follows: predialysis or postdialysis ESRD patients who sought kidney transplantation and underwent screening colonoscopy as part of pre-transplant work-up. The exclusion criteria were as follows: incomplete colonoscopy occurred in the patients, defined as failed to reach the ileocecal junction, due to the inadequate preparation, intolerant pain, tortuosity, stricture, and intro-operative complications. 200 age- and gender-matched subjects with normal renal function and received a colonoscopy for health screening or other minor indicaitons in the same period were randomly selected as control. The following variables were collected from their medical records: age, sex, body mass index (BMI), cause of ESRD, time form serum creatinine elevation to dialysis or colonoscopy, dialysis modality, time on dialysis, blood urea nitrogen, serum creatinine, albumin, hemoglobin, serum parathyroid hormone, serum calcium, serum phosphorus, charlson comorbidity index, past use of rhubarb-containing TCM, detection of MC. The study was reviewed and approved by the ethics committee of the PLA General Hospital. The requirement for informed consent was waived due to the retrospective design of this study.

Definitions

MC was defined as a gross appearance of brown or black discoloration of the colonic mucosa during colonoscopy ^[18]. Pathological evaluation was performed in the cases with other benign and malignant diseases accompanied on colonoscopy.

Statistical analysis

Descriptive statistics were given for all demographic and clinical variables. All available variables were compared between ESRD and control groups, and also compared in ESRD group between patients with and without melanosis coli. After each comparison, all variables with p value of <0.500 were included in a multivariable logistic regression analysis with dichotomized being EC as outcome, for adjusting confounding factors between groups, and for discerning of factors associated with EC in ESRD group respectively. The statistical significance level was set to 0.05. All analyses were performed using Statistical Package for Social Sciences (SPSS) software version 21.

Results

Demographic and clinical characteristics of study population

Demographic and clinical characteristics of study population are shown in Table 1. There were no statistically significant difference in age (43.3±11.0 vs. 44.6±10.9, p=0.339) and distrabutaion of sex (male 77[77.8%] vs. 155[77.5%], p = 0.957) between the two groups. The BMI of ESRD group was significantly lower than that of control group (22.7±3.7 vs. 24.7±3.8, p = 0.000). In ESRD group, the cause of ESRD was glomerulonephritis in 71 cases (71.7%), diabetic nephropathy in 16 cases(16.2%), hypertensive nephropathy in 7 cases(7.1%), polycystic kidney disease in 2 cases (2.0%), and unspecified in 3 cases (3.0%). The median time elapsed from serum creatinine elevation to dialysis or colonoscopy was 36(IRQ12-72) months. Sixty-three ESRD patients (63.6%) were undergoing hemodialysis, 9 (9.1%) were undergoing peritoneal dialysis, and 27 (27.3%) were receiving predialysis care. Their median time on dialysis was 2.4(IRQ0.0-10.8) months. Blood urea nitrogen (25.4±9.8 vs. 4.9±1.4 mmol/L, p = 0.000) and serum creatinine (909.6±282.8 vs. 71.8±13.7 μmol/L, p = 0.000) was significantly higher, while albumin (36.9±5.5 vs. 42.0±8.2 g/L, p = 0.000) and hemoglobin (99.9±19.5 vs. 143.1±15.7 g/L, p = 0.000) was significantly lower in ESRD group vs. control group. The serum parathyroid hormone, calcium and phosphorus was 200.9(IRQ104.1 - 330.0) pg/ml, 2.15±0.21 mmol/L and 2.05±0.55 mmol/L respectively in ESRD group. Charlson comorbidity index was low in both group, but signifcantly higher in ESRD group (1[0 - 2] vs. 0[0 - 1], p = 0.007), while the number of patients with ≥1 comorbidity was similar between the two groups (37[37.4%] vs. 81[40.5%], p = 0.603).

Detection rate of EC in the ESRD patients vs. control

Prevalence of EC in the two groups is also shown in table 1. Among 99 ESRD patients, MC was detected in 19, with a prevalence of 19.2% (95% confidence interval[CI]: 11.3%, 27.1%). In contrast, among 200 control subjects, MC was detected in only 1, with an prevalence of 0.5% (95%CI: 0.5%, 1.5%) (ESRD vs. control: odds ratio[OR] = 47.263, 95% CI: 6.222, 358.981; p=0.000). Logistic regression analysis of variables associated with melanosis coli is shown in table 2. As shown, after adjustment of age, BMI, albumin, hemoglobin and Charlson comorbidity index, being ESRD was still associated with the suffer from MC (Odds ratio[95%CI]: 36.251[3.435, 382.559], p = 0.003).

Factors associated with detection of EC in ESRD patients

In ESRD patients, both univariate analysis (OR = 4.358, 95% CI: 1.329, 14.286; p = 0.012; as shown in table 3) and multivariate logistic regression (OR = 4.916, 95% CI: 1.462, 16.525; p = 0.010; as shown in table 4) found that past use of Rhubarb-containing TCM herbs was the only independent risk factor of MC.

Discussion

As far as we know, this is by far the first article concerning the detection rate and risk factors of EC in ESRD patients. Our patients came from different places of China, which means that they have good representativeness. The major finding of our study is that the detection rate of EC in those ESRD patients reached as high as 19.2%, as compared to 0.5% in the control subjects with normal renal function. After adjusting confounding factors, the detection rate of EC in ESRD group was still higher than that in control. Furthermore, both univariate and multivariate analysis showed that past use of rhubarb-containing TCM medicine was the only risk factor for detection of EC in ESRD patients. AlAmeel T et al ^[23] published their screening colonoscopic results in 169 renal transplant candidates, but no MC was reported. Zheng Y et al ^[24] also reported no MC in 469 screening colonoscopies during kidney transplant evaluation. The difference between their results and ours maybe come from the different population, different treatment background and different focus of attention.

Different detection rates of EC in general population have been reported by different studies in the literature. In autopsy studies, the prevalence of EC is relatively high. In one study, MC was found in 119 of 200 consecutive autopsies, with a prevalence of 59.5% ^[25]. In another international multicentre study ^[26], MC was found in 34% of all autopsies. However, in endoscopic studies, the figure was much lower. Wittoesch JH et al ^[17] reported that 750 cases of MC was found in 91,472 colonoscopies from 1950 to 1954 at the Mayo Clinic, with a prevalence of 0.82%. In another series of patients, he found 137 cases of MC in 12,131 routine proctoscopic examinations from March 1 to September 30, 1955, giving a prevalence of 1.13% ^[17]. Wang S et al ^[16] conducted a multicenter study in China, and detected 6,090 cases of MC in 342,922 colonoscopies between January 2006 and October 2016, obtaining a detection rate of 1.78%. Our study is an endoscopic study, so in control subjects with kidney function normal, we observed a detection rate in a similar level, which was 0.5%. Our detection rate is lower than that reported by Wang S et al ^[16], possibly because our patients are much younger (mean 44.6 vs. 60 years old). Data showed that elder age was related to higher detection rate of EC. In age ≤44 years group of Wang S et al study ^[16], the detection rate of MC was 0.85, which was much similar to our results.

On the other hand, in ESRD group, we found 19 cases of MC in 99 ESRD patients, with a detection rate of 19.2%, which was astonishingly higher than that of control group. After adjusting age, BMI, serum albumin, hemoglobin, charlson comorbidity index, the detection rate in ESRD group was still higher than control, which means that other factors related to ESRD were the cause. In the second part of the study, we examined the risk factors of MC in ESRD patients using univariate and multivariate analyses, and found that only past use of rhubarb-containing TCM medicine was associated with the detection of MC. Rhubarb has long been found as a risk factor of MC. The study results of Badiali D et al ^[27] showed that MC was significantly associated with consumption of anthracene drugs. They found that the prevalence of MC in patients taking anthracene drugs was significantly higher (52.0%) than that without taking anthracene drugs (8.0%) ^[27]. Wittoesch JH et al ^[17] reported that in 750 MC patients, 718 were habitual users of laxatives, while in another series of 137 MC patients, laxatives were used regularly by 132 patients, in which 113 patients used cathartics containing emodin. In an animal study, Chen JY et al ^[14] fed extract of crude rhubarb to guinea pigs for 60 days, and found that rhubarb successfully induced MC with a dosage dependent manner. In another animal study, Cheng Y et al ^[15] fed rhubarb extract to Sprague-Dawley rats for 90 days, and saw MC formation in the rats' colon. They also found that rhein was the metabolite of rhubarb responsible for its toxicity ^[15]. Our study showed that in 19 ESRD and with MC patients, 15 (78.9%) used rhubarb-containing TCM herbs in the past, while in 80 ESRD and without MC patients, 37 (46.3%) used such herbs, which means a significant difference. Our results indicate that in ESRD patients, consumption of rhubarb-containing TCM medicine carries the risk of MC development, just as in general population with normal renal function.

The relationship between EC and colorectal neoplasm has long been a concern. Liu ZH et al ^[12] reported that MC is associated with a higher incidence of colonic non-adenoma polyps, low-grade adenomas, and distal ileal ulcers, which means that MC may not be a harmless pigmentation, but an indication of chronic colonic and intestinal injury. Blackett JW et al ^[28] reported that the presence of MC was associated with increased adenoma detection. Siegers CP ^[19] found in a prospective study of 1095 patients that the incidence of EC was 6.9% for patients with normal endoscopy, 9.8% for patients with adenomas and 18.6% for patients with colorectal carcinomas. A relative risk of 3.04 (1.18, 4.90; 95% confidence interval) was calculated for colorectal cancer as a result of MC from these data ^[19]. Morgenstern L et al ^[20] reported that 17% of MC was accompanied by invasive adenocarcinoma. The data of Biernacka-Wawrzonek D et al ^[29] showed that colon cancer was found in 11.9% of MC patients. Given that ESRD patients undergoing kidney transplantation already have a significantly higher risk of colorectal cancer ^[30], the high prevalence of EC in ESRD patients may further raise this risk.

There exist a number of limitations in this study. Firstly, this is a retrospective study, so selection bias can't be avoided. Secondly, this is a single center study, the sample size is relatively small, and only Chinese population is included. Thirdly, because the data is derived from medical record review, sometimes there is the problem of incoherence, inconsistent in the information provided. For example, bowel habit such as constipation or diarrhea was not described in detail but marked as normal in most of the patients, possibly because it's not directly related to the condition for hospitalization, thus can't be analyzed as an risk factor. Therefore, future prospective, multicenter, multi-national, well-designed studies with large sample size are warranted to assess the detection rate and risk factors of MC in ESRD patients.

In conclusion, there is a significantly higher detection rate of EC in this cohort of ESRD patients than control subjects without renal failure. The detection of EC in ESRD patients is independently related to the past use of rhubarb-containing TCM herbs. We feel that use rhubarb-containing TCM herbs to treat renal failure is the main reason for the occurrence of EC, and therefore recommend take cautions when prescribing those kind of therapies in renal failure patients. Caution of the development of EC should also be placed in the package inserts of rhubarb-containing TCM medicine. Considering that EC is a sign of colonic and intestinal mucosa injury, and the possible association of MC with development of colorectal cancer, the clinical significance of the high detection rate of EC in ESRD patients, and its influence on the long-term prognosis of kidney transplantation, demands further study in the future.

Declarations

Acknowledgements

None

Funding

None

Data Availability Statements

The data underlying this article will be shared on reasonable request to the corresponding author.

Authors' contributions

Zhang W conceived and designed the study and wrote the manuscript. Zhou W, Wang Y, Li Y and Huang W participated in data collection and analysis. All authors were involved in drafting and revising the manuscript, providing intellectual content of critical importance to the work described and approving the final version of the manuscript.

Conflict of interest statement

The authors declare that they have no conflict of interest to disclose. The results presented in this article have not been published previously in whole or part, except in abstract format.

References

1. Zhang L, Wang F, Wang L, Wang W, Liu B, Liu J, et al. Prevalence of chronic kidney disease in China: A cross‑sectional survey. Lancet 2012;379:815‑22. doi: 10.1016/S0140‑6736(12)60033‑6.
2. Koppe L, Cassani de Oliveira M, Fouque D. Ketoacid Analogues Supplementation in Chronic Kidney Disease and Future Perspectives. Nutrients. 2019;11(9):2071. Published 2019 Sep 3. doi:10.3390/nu11092071.
3. Zhang Y, He D, Zhang W, et al. ACE Inhibitor Benefit to Kidney and Cardiovascular Outcomes for Patients with Non-Dialysis Chronic Kidney Disease Stages 3-5: A Network Meta-Analysis of Randomised Clinical Trials. Drugs. 2020;80(8):797-811. doi:10.1007/s40265-020-01290-3.
4. Guo C, Li S, Rao XR. New Goals and Strategies of Chinese Medicine in Prevention and Treatment of Chronic Kidney Disease. Chin J Integr Med. 2019;25(3):163-167. doi:10.1007/s11655-019-3065-z.
5. Li X, Wang H. Chinese herbal medicine in the treatment of chronic kidney disease. Adv Chronic Kidney Dis. 2005;12(3):276-281. doi:10.1016/j.ackd.2005.03.007.
6. China pharmacopoeia committee, 2015. Pharmacopoeia of the People's Republic of China, 26 vol.1, pp. 945 (in Chinese).
7. Li L. End-stage renal disease in China. Kidney Int. 1996;49(1):287-301. doi:10.1038/ki.1996.41.
8. Khan IA, Nasiruddin M, Haque SF, Khan RA. Evaluation of Rhubarb Supplementation in Stages 3 and 4 of Chronic Kidney Disease: A Randomized Clinical Trial. Int J Chronic Dis. 2014;2014:789340. doi:10.1155/2014/789340.
9. Peng A, Gu Y, Lin SY. Herbal treatment for renal diseases. Ann Acad Med Singapore. 2005;34(1):44-51.
10. Freeman HJ. "Melanosis" in the small and large intestine. World J Gastroenterol. 2008;14(27):4296-4299. doi:10.3748/wjg.14.4296.
11. Willems M, van Buuren HR, de Krijger R. Anthranoid self-medication causing rapid development of melanosis coli. Neth J Med. 2003;61(1):22-24.
12. Liu ZH, Foo DCC, Law WL, Chan FSY, Fan JKM, Peng JS. Melanosis coli: Harmless pigmentation? A case-control retrospective study of 657 cases. PLoS One. 2017;12(10):e0186668. Published 2017 Oct 31. doi:10.1371/journal.pone.0186668.
13. Li XΑ, Zhou Y, Zhou SΧ, et al. Histopathology of melanosis coli and determination of its associated genes by comparative analysis of expression microarrays. Mol Med Rep. 2015;12(4):5807-5815. doi:10.3892/mmr.2015.4126.
14. Chen JY, Pan F, Zhang T, Xia J, Li YJ. Experimental study on the molecular mechanism of anthraquinone cathartics in inducing melanosis coli. Chin J Integr Med. 2011;17(7):525-530. doi:10.1007/s11655-011-0786-z.
15. Cheng Y, Zhang H, Qu L, et al. Identification of rhein as the metabolite responsible for toxicity of rhubarb anthraquinones. Food Chem. 2020;331:127363. doi:10.1016/j.foodchem.2020.127363
16. Wang S, Wang Z, Peng L, et al. Gender, age, and concomitant diseases of melanosis coli in China: a multicenter study of 6,090 cases. PeerJ. 2018;6:e4483. Published 2018 Mar 8. doi:10.7717/peerj.4483.
17. WITTOESCH JH, JACKMAN RJ, McDONALD JR. Melanosis coli: general review and a study of 887 cases. Dis Colon Rectum. 1958;1(3):172-180. doi:10.1007/BF02616828.
18. Abu Baker F, Mari A, Feldman D, Suki M, Gal O, Kopelman Y. Melanosis Coli: A Helpful Contrast Effect or a Harmful Pigmentation?. Clin Med Insights Gastroenterol. 2018;11:1179552218817321. Published 2018 Dec 4. doi:10.1177/1179552218817321.
19. Siegers CP, von Hertzberg-Lottin E, Otte M, Schneider B. Anthranoid laxative abuse--a risk for colorectal cancer?. Gut. 1993;34(8):1099-1101. doi:10.1136/gut.34.8.1099.
20. Morgenstern L, Shemen L, Allen W, Amodeo P, Michel SL. Melanosis coli. Changes in appearance when associated with colonic neoplasia. Arch Surg. 1983;118(1):62-64. doi:10.1001/archsurg.1983.01390010048011
21. Nusko G, Schneider B, Ernst H, Wittekind C, Hahn EG. Melanosis coli--a harmless pigmentation or a precancerous condition?. Z Gastroenterol. 1997;35(5):313-318.
22. Nusko G, Schneider B, Schneider I, Wittekind C, Hahn EG. Anthranoid laxative use is not a risk factor for colorectal neoplasia: results of a prospective case control study. Gut. 2000;46(5):651-655. doi:10.1136/gut.46.5.651.
23. AlAmeel T, Bseiso B, AlBugami MM, AlMomen S, Roth LS. Yield of screening colonoscopy in renal transplant candidates. Can J Gastroenterol Hepatol. 2015;29(8):423-426. doi:10.1155/2015/823835.
24. Zheng Y, Chaung KV, Park PJ, et al. The Utility of Screening Colonoscopy During Kidney Transplant Evaluation. Exp Clin Transplant. 2017;15(6):602-608. doi:10.6002/ect.2016.0214.
25. Koskela E, Kulju T, Collan Y. Melanosis coli. Prevalence, distribution, and histologic features in 200 consecutive autopsies at Kuopio University Central Hospital. Dis Colon Rectum. 1989;32(3):235-239. doi:10.1007/BF02554536.
26. Clark JC, Collan Y, Eide TJ, et al. Prevalence of polyps in an autopsy series from areas with varying incidence of large-bowel cancer. Int J Cancer. 1985;36(2):179-186. doi:10.1002/ijc.2910360209.
27. Badiali D, Marcheggiano A, Pallone F, et al. Melanosis of the rectum in patients with chronic constipation. Dis Colon Rectum. 1985;28(4):241-245. doi:10.1007/BF02554044.
28. Blackett JW, Rosenberg R, Mahadev S, Green PHR, Lebwohl B. Adenoma Detection is Increased in the Setting of Melanosis Coli. J Clin Gastroenterol. 2018;52(4):313-318. doi:10.1097/MCG.0000000000000756.
29. Biernacka-Wawrzonek D, Stępka M, Tomaszewska A, et al. Melanosis coli in patients with colon cancer. Prz Gastroenterol. 2017;12(1):22-27. doi:10.5114/pg.2016.64844.
30. Komaki Y, Komaki F, Micic D, Ido A, Sakuraba A. Risk of Colorectal Cancer in Chronic Kidney Disease: A Systematic Review and Meta-Analysis. J Clin Gastroenterol. 2018;52(9):796-804. doi:10.1097/MCG.0000000000000880.