Aim: This was to describe the different types of lesions that can be found in patients with liver cirrhosis during upper gastrointestinal endoscopy. Patients and Methods: Consecutive patients with liver cirrhosis who gave their consent to participate in the study were recruited. The diagnosis of liver cirrhosis was made by clinical and radiological features and a total of 56 patients were recruited. After taking informed consent, upper gastrointestinal endoscopy was performed on all the patients. Results: The patients consisted of 43 (76.8%) males and 13 (23.2%) females. Mean age was 45.5 13.8 years (range: 20-77 years). Analysis showed that 21 (37.5%) patients were less than 40 years of age, 16 (28.6%) were between 40 years and 49 years, and 19 (33.9%) were 50 years and above. Esophageal varices was diagnosed in 54 (96.4%) patients. The esophageal mucosal findings observed were erosions (7.1%), ulcer (1.8%), candidiasis (8.9%) while one (1.8%) patient had both esophageal erosions and candidiasis. Gastric varices were observed in 11 (19.6%) patients and the analysis showed that the most frequent form was isolated gastric varices type 1 (IGV1) was recorded in six (10.7%) patients. Portal hypertensive gastropathy (PHG) was seen in 45 (80.4%) patients. Other gastric mucosal lesions were erosions (12.5%) and ulcers (10.7%). Duodenal varices were not seen in any of the patients. Other lesions seen were duodenitis, duodenal erosions, and duodenal ulcers in one (1.8%) patient, one (1.8%) patient, and four (7.1%) patients, respectively. Conclusion: This study has revealed different pattern of lesions that can be seen in patients with liver cirrhosis apart from varices.
Keywords: Endoscopy, liver cirrhosis, esophageal candidiasis, ulcers, varices
Liver cirrhosis is a clinicopathologic condition that is characterized by hepatic fibrosis, nodular regeneration, and distortion of hepatic architecture.  It presents with an array of clinical manifestations as well as complications with a high mortality and is common worldwide. In the United Kingdom, about 30,000 individuals live with liver cirrhosis and annually, another 7,000 new cases are diagnosed.  In the year 2000, liver cirrhosis accounted for 1.2% of all deaths in the United States.  In a study in Nigeria, it was found to account for 20.4% of all admissions to the medical wards over a 6-year period. 
The distortion of hepatic architecture seen in liver cirrhosis disrupts the hepatic blood flow and functions. The major complications seen in liver cirrhosis are portal hypertension, ascites, coagulopathy, hepatic encephalopathy, and hepatocellular carcinoma. 
Upper gastrointestinal endoscopy (UGIE) has been found to be of specific use in patients with chronic liver disease, especially when there is gastrointestinal hemorrhage,  which is one of the most important complications of liver cirrhosis and is mostly variceal. However, there are other causes such as gastritis and gastric ulcers and duodenal ulcers, which may be unrelated to the underlying cirrhosis. 
The aim of this study was to describe the different types of lesions that can be found in patients with liver cirrhosis during upper gastrointestinal endoscopy.
This descriptive study was conducted at the endoscopy unit of a tertiary health facility. Consecutive patients with liver cirrhosis who gave their consent to participate in the study were recruited.
The diagnosis of liver cirrhosis was made by clinical and radiological features, and a total of 56 patients were recruited. After taking informed consent, upper gastrointestinal endoscopy was performed on all the patients. The oropharynx was sprayed with 2% xylocaine and the patients were placed in the left lateral position, and a mouth gag was then placed between the incisor teeth. The gastroscope (Olympus, UK, exera III, GIF-HQ190) was then introduced under direct vision into the oropharynx, the esophagus, stomach, and duodenum. The general rules for study of portal hypertension (2 nd edition) as proposed by the Japan Society for portal hypertension were used to describe the varices. 
Esophageal varices were described based on form (F), location (L), color (C), presence of red color signs (RC), bleeding signs, and mucosal findings.
In terms of location, locus superior (Ls) is for varices located in the upper esophagus, locus medialis (Lm) for those in the middle, and locus inferior (Li) for the varices in the lower part of the esophagus.
In relation to the form, F0 stands for varices with no varicose appearance, F1 for straight, small caliber varices, F2 for moderately enlarged beady varices, and F3 for markedly enlarged, nodular, or tumor-shaped varices.
In terms of color, the varices are classified as either white, when they look whitish like large folds of the esophageal mucosa or blue, when they are bluish white or cyanotic and distended by blood.
The red color signs are classified into red wale markings (RWMs), which are dilated venules on the mucosal surface, cherry-red spots (CRSs), which are red spots on the mucosal surface, and hematocystic spots (HCSs), which are large, round, crimson-red projections like blood blisters.
Bleeding signs are classified into those present during bleeding, which are gushing, spurting, or oozing, and those that occur after hemostasis, which are either red or white plug. Mucosal findings were recorded as either erosions, ulcers, scars, or any other findings.
Gastric varices were described as gastroesophageal varices type 1 (GOV 1) if esophageal varices extended into the stomach along the lesser curve and GOV 2 if the extension was along the greater curve. Isolated gastric varices type 1 (IGV 1) were those located in the fundus while IGV 2 were those located in other parts of the stomach or duodenal bulb. Gastric mucosal findings were recorded as portal hypertensive gastropathy (PHG), erosions, ulcers, or any other mucosal findings.
Duodenal varices were recorded as present if found in the second or distal portions of the duodenum, and the mucosal findings were recorded as erosions, ulcers, or any other findings.
A total of 56 patients comprising 43 (76.8%) males and 13 (23.2%) females participated in the study. The mean age was 45.5 ± 13.8 years with a range of 20-77 years. Analysis of the age groups showed that 21 (37.5%) patients were less than 40 years of age, 16 (28.6%) were between 40 years and 49 years, and 19 (33.9%) were 50 years and above. The most common symptom presented by the patients was abdominal swelling in 46 (82.1%) of them. This was followed by leg swelling in 30 (53.6%) patients. Clinical examination revealed that 31 (55.4%) patients had ascites, 23 (41.1%) had hepatomegaly, and 19 (33.9%) had prominent anterior abdominal wall veins. Stigmata of chronic liver disease observed in the patients were Dupuytren’s contracture in 21 (37.5%) patients, palmar erythema in 12 (21.4%) patients, and sparse axillary hair in 12 (21.4%) patients.
The upper gastrointestinal endoscopy showed that 54 (96.4%) patients had esophageal varices.
Location of esophageal varices
Analysis of the esophageal varices revealed that 33 (61.1%) patients had their varices located in the lower esophagus, 6 (11.1%) in the middle, and 3 (5.6%) in the upper esophagus. Multiple locations of the varices were also observed as follows: Middle and lower portions in 10 (18.5%) patients, upper and middle portions in 1 (1.9%) patient, and in the whole length of the esophagus in 1 (1.9%) patient [Table 1].
Columns of the esophageal varices
In terms of the number of columns, the most frequently observed was three columns, which were seen in 20 (35.7%) patients followed by a single column observed in 15 (26.8%) patients [Table 1].
Color of the esophageal varices
Analysis of the color of the varices showed that blue varices were observed in 36 (64.3%) patients while white varices were seen in 14 (25%) patients [Table 1].
Forms of the esophageal varices
The most frequent form of the varices was f 0 , which was observed in 20 (37%) patients followed by f 1 in 15 (27.8%) patients. A combination of various forms of the varices was observed in 12 (22.2%) patients with the most frequent combination being f 2 and f 3 seen in 6 (11.1%) patients [Table 1].
Red color signs
The most common red color sign observed was RWM seen in 15 (26.8%) patients, HCS was seen in 1 (1.8%) patient while CRS was not observed in any of the patients [Table 1].
Esophageal mucosal findings
Esophageal mucosal findings observed in the patients were erosions in four (7.1%) patients, ulcer in one (1.8%) patient, candidiasis in five (8.9%) patients while one (1.8%) patient had both esophageal erosions and candidiasis [Figure 1].
Gastric varices were observed in 11 (19.6%) patients and the analysis showed that the most frequent form was IGV1 was recorded in six (10.7%) patients while four (7.1%) patients had GOV1, and a combination of GOV1 and IGV1 was seen in only one (1.8%) patient. GOV2 and IGV2 were not observed at all in our patients [Figure 2].
Gastric mucosal findings
The most frequent lesion was PHG seen in 45 (80.4%) patients. Other mucosal lesions were gastric erosions in seven (12.5%) patients and gastric ulcers in six (10.7%) patients [Figure 3].
Duodenal varices were not seen in any of the patients. Other lesions seen were duodenitis, duodenal erosions, and duodenal ulcers in one (1.8%) patient, one (1.8%) patient, and four (7.1%) patients, respectively [Figure 4].
In this study, the prevalence of esophageal varices was 96.1%. This figure is much higher than the 53.6% prevalence reported by Ajayi et al. in Ekiti, Nigeria, which is in the same geographic zone where this study was conducted. Although the sample sizes were the same, the marked difference in the prevalence of esophageal varices could be explained by the fact that only f1-f3 varices were included in their study, whereas our study included f0 varices in addition and this group constituted the highest prevalence (37%).
The prevalence of esophageal varices in this study is also higher than the 75% reported by Achinge et al. in Jos, Plateau State, Nigeria. Although, the sample is higher in this particular study, the difference observed in the prevalence could be due to the fact that only newly diagnosed patients with liver cirrhosis were recruited in their study, whereas our study included previously diagnosed patients in addition, and this could have accounted for the higher prevalence.
This prevalence is also much higher than that reported in other parts of the world. In a study by Giannini et al. in Italy, the prevalence of esophageal varices was 61%. This study was retrospective compared to our own, which was prospective but their sample size was higher. The lower prevalence in their study could be explained by the stringent inclusion criteria employed because patients with active gastrointestinal bleeding at admission, those who had previously undergone one form of treatment for esophageal varices, and those who were on primary prophylaxis for variceal bleeding were excluded from their study.
Also, Ullah et al. and Sarangapani et al. reported a lower prevalence of 65% and 72.6%, respectively. Again in these particular studies, many patients with cirrhosis who were presumed to have esophageal varices were excluded from the studies, and this could explain the lower prevalences recorded.
The prevalence of esophageal erosion of 7.1% seen in our study is higher than that reported by Schechter et al. (2%). In this particular study, the sample size is higher than in our study. Although a lower prevalence of esophageal erosion was observed at endoscopy in that study, typical symptoms of gastroesophageal reflux disease was observed in 53% of their patients and abnormal gastroesophageal reflux was demonstrated in 37% of these patients using 24-h pH ambulatory recording.
It has been observed that cirrhotic patients with esophageal varices are prone to developing esophageal motor disorders, a delay in esophageal clearance time, and abnormal gastroesophageal reflux compared to those without esophageal varices. ,,, In our study, almost all the patients had esophageal varices and this could explain the high prevalence of esophageal erosions observed. The presence of ascites in the majority of our patients also could have predisposed to erosive esophagitis. It has been observed that ascites through increase in intragastric and intraabdominal pressure could predispose to gastroesophageal reflux with resultant esophageal erosions. , In a study by Nebel et al.,  lower esophageal sphincter pressure (LESP) in patients with cirrhosis was not different from that of control subjects but was found to be significantly higher than the LESP recorded in patients with cirrhosis and ascites. It was also observed that the LESP significantly increased after the drainage of ascitic fluid.
Another mechanism that aids gastroesophageal reflux in cirrhosis has to do with plasma vasoactive peptide and neurotensin, which are known to reduce LESP, and these have been found to be elevated in patients with liver cirrhosis. , Li et al. reported a higher prevalence (43%) of erosive esophagitis in their patients with liver cirrhosis. In this particular study, the sample is much higher than in our study. Apart from this, a majority of their patients had medium and large varices and erosive esophagitis was found in 43% and 54% of them, respectively, in contrast to our study where the majority of the patients had small varices. Another observation in their study was the increased prevalence of erosive esophagitis with increasing age such that 63.1% of those with this abnormality were older than 60 years. However, in our study a majority of the patients were younger than 50 years of age.
Esophageal candidiasis was observed in 8.9% of our patients and this is much higher than the 0.8% recorded by Ou et al. in a retrospective study. Although their sample was higher, the category of patients studied might have contributed to the wide difference in the prevalence. While they studied non-human immunodeficiency virus (HIV)-infected subjects among whom 3,017 had liver cirrhosis, we prospectively studied only patients with liver cirrhosis. However, it is possible that some of our patients had underlying HIV or other immunosuppressive conditions, which could have been responsible for the high prevalence. However, none of our patients was screened for these conditions. Although it is known that esophageal candidiasis is common in immunocompromised states such as HIV infection or as a result of iatrogenic immunosuppression, , liver cirrhosis has also been found to lower immunity and as such predispose to bacterial and fungal infections. Reasons for this include reduced antibody production, increased production of immunosuppressive cytokines, reduced leukocyte function, reduced opsonization, and disturbances of complement. ,,
In this study, the prevalence of gastric varices was 19.6%, which is slightly higher than that reported by Mumtaz et al.  (15%). While their study included patients with other causes of portal hypertension, this present study recruited only patients with liver cirrhosis as the cause of portal hypertension. Another observation is that 86% of their patients with gastric varices had hepatitis C virus (HCV) infection as the etiology of the liver disease, whereas only 1.8% of our patients had HCV infection. These two factors might explain the difference in the prevalence of gastric varices. In our study, IGV1 was the most frequent type observed compared to theirs where GOV1 was the most frequent. Also, GOV2 and IGV2 that were not observed in our study were recorded by Mumtaz et al.  In a study by Sarin et al.,  IGV2 were reported to predominantly arise after esophageal variceal ligation. This may explain the absence of IGV2 in our patients since none of them had variceal band ligation prior to the study. In the study by Schechter et al.,  the prevalence of gastric varices was 8%, which is lower than in the present study. This may be explained by the fact that most of their patients had less severe clinical presentation compared to our patients. In another study by Sarin et al.,  the prevalence of gastric varices is the same as in this present study. However, GOV-1 was the most common type they observed in their patients.
Gastroduodenal ulcer was seen in 17.8% of our patients. This finding is of importance in the sense that this may be a source of upper gastrointestinal bleeding (UGIB) in some of these patients. This is paramount so as not to attribute every case of UGIB in patients with liver cirrhosis to varices alone. In a study by Gado et al.,  peptic ulcer was found in 60% of their patients with liver cirrhosis who had acute nonvariceal UGIB and was the commonest cause of bleeding in this category of patients. Also, in their patients with acute variceal UGIB, peptic ulcer was diagnosed in 17% of them. Their study recruited patients with liver cirrhosis and acute UGIB and our study recruited nonbleeding patients. This might explain the higher prevalence of peptic ulcer recorded in the former.
However, the prevalence of peptic ulcer in this present study is comparable to that reported in other previous studies. ,,, The presence of peptic ulcer in cirrhotic patients may be associated with some ulcerogenic factors that are specific to patients with liver cirrhosis. Among the proposed factors are hypergastrinemia, decreased gastric prostaglandin E2 levels, and the observed portosystemic shunting in liver cirrhosis, which may prevent ulcerogenic factors from being cleared by the liver. ,, Also, it has been observed that the presence of portal hypertension and PHG may predispose the gastric mucosa and duodenal mucosa to damage by the ulcerogenic factors as well as impair the ability of these to repair damage. , In this present study, PHG was observed in 80.4% of our patients, and this may explain the high prevalence of gastroduodenal ulcer observed.
Duodenal varices were not seen in any of our patients. These lesions are said to be rare in patients with liver cirrhosis, occurring in 0.4% of the patients.  However, a prevalence of 40% had been recorded in patients with hepatic portal hypertension undergoing angiography. 
Upper gastrointestinal endoscopy has revealed a different pattern of lesions in patients with liver cirrhosis, apart from varices. This underscores the importance of this procedure in the diagnosis of these additional lesions so as to prevent the complications that can arise from these lesions if not diagnosed and treated appropriately.
We thank all the nurses and other supportive members of the staff of the endoscopy suites of the University College Hospital, Ibadan, Oyo State, Nigeria for their support during the study.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
Source of Support: None, Conflict of Interest: None
[Figure 1], [Figure 2], [Figure 3], [Figure 4]