The need for gastroscopy in early cirrhotics

Background

Despite recent advances, oesophageal variceal bleeding in cirrhosis is still associated with 10-20% mortality at 6 weeks.¹ Furthermore, in 35-90% of patients rebleeding occurs after spontaneous haemostasis (approximately 40% of rebleeding episodes occurring within the first 5 days).² Once cirrhosis is diagnosed a gastroscopy is recommended to check for oesophageal varices in most cases.³ If oesophageal varices are present, treatment is required to prevent variceal bleeding (primary prophylaxis of variceal haemorrhage).⁴ The recommended treatment is with non-selective beta blockers (NSBB) for patients with small varices and either NSBB or variceal banding if medium to large varices are present.¹ Splenomegaly and consumptive thrombocytopenia are recognized complications of significant portal hypertension secondary to cirrhosis⁵ and potential indices for indirect diagnosis of oesophageal varices.⁶

Aim

Oesophageal varices are a major complication of portal hypertension and occur in 30-70% of cirrhotics.⁷ Is it possible to predict which cirrhotic patients will have oesophageal varices requiring treatment before performing a gastroscopy?

Method

This is a retrospective study looking at all the gastroscopy reports of Child-A cirrhotic patients under the care of a single gastroenterology/hepatology firm performed in 2015 (see table 1 for population demographics).  The diagnosis of cirrhosis was in most cases based on imaging studies as shear wave elastography has only recently been introduced locally. The aetiology of liver cirrhosis included autoimmune liver conditions, alcoholic liver disease, chronic viral hepatitis and non-alcoholic steatohepatitis. Once a patient was recruited their previous endoscopies were also retrieved and matched with their Child-

Pugh score at the time (see table 2). Platelet counts and spleen size as measured on imaging (US/CT/MRI) within three months of endoscopy were also recorded. For the purposes of the study a normal sized spleen is one which is less than 12 cm. During data collection whenever splenomegaly was noted on imaging but no precise measurement was given, a spleen size of 12 cm was allocated. Conversely a normal spleen with no precise measurement was given a value of 11 cm.⁸ The gastroscopy report was used to determine whether varices were present. If present and banding was performed this was also recorded. Once oesophageal banding was performed further endoscopies to control variceal size were excluded from the study since the question of prediction is no longer relevant in these patients. Patients with a previous splenectomy or who were suffering from myeloproliferative conditions (which often effects spleen size and platelet counts) were excluded from the study. Patients with early hepatocellular carcinoma (within the Milan transplant criteria)⁹ were not excluded from the study as it was felt that early cancer would not contribute significantly to portal hypertension. ANOVA statistical analysis was used to measure significant differences between groups regarding platelet count and spleen size.

Results

A total number of 59 patients undergoing 97 gastroscopies (episodes) between 2008 and 2015 were included in the study. The cases were divided into 3 groups; absent varices (AV) group: no varices seen at endoscopy, small varices (SV) group: small varices not requiring banding and banded varices (BV) group: medium–large varices requiring banding at endoscopy. Figure 1 illustrates the number of episodes in each group.

Figure 1

Age Distribution

75.3% (n=73) of episodes were in patients between 50 -69 years of age (Figure 2)

Figure 2

Gender distribution

71.1% (n=69) of episodes studied involved males (figure 3). This gender distribution was similar throughout the three groups.

Figure 3

Spleen size (cm)

89.6% (n=87) of the population studied had a spleen size >12cm as seen in figure 4. Interestingly 23.8% (n=10) of AV group patients had spleens larger than 14 cm. On the other hand, none of the patients in BV group had a normal sized spleen (<12 cm) (table 3).

Figure 4

Statistical analysis using ANOVA showed that spleen size difference between the 3 groups was statistically significant (p= 0.035) (table 4).

Platelet count: Units x10⁹/L

In BV group, 66.7% (n=18) of patients had a platelet count <150 x10⁹/L whereas in AV group 61.9% (n=26) of patients had a platelet count > 150 x10⁹/L (table 5).

Statistical analysis using ANOVA showed that platelet count differences between the 3 groups was statistically significant (p=0.008) (table 6).

Discussion

Several recent studies have explored possible non-invasive predictors of the presence and/or severity of oesophageal varices in cirrhotics including such indices as spleen size, platelet count, spleen stiffness, liver stiffness, liver volume and serum albumin level.^10-11,5 Spleen size and platelet count are indices which are readily available after initial workup of patients with chronic liver disease and were thus selected in this study for further evaluation. Studies have shown that as cirrhosis progresses spleen size tends to increase and platelet count tends to decrease.⁶ The aim of this study was to assess whether one could predict which patients would have oesophageal varices requiring treatment before performing endoscopy and therefore which patients can safely avoid such invasive monitoring.

In this study 66.7% of BV group had platelet counts below 150 x10⁹/L as opposed to 38.1% of AV group patients and this result was statistically significant. There was still however considerable overlap between the 2 groups such that 33.3% (n=9) of BV group patients had platelet counts above 150 x10⁹/L. Platelet counts on their own therefore cannot be used to predict the presence of varices in cirrhotic patients. In the case of spleen size the differences between groups were again statistically different and the overlap between the groups was less marked. A possible inference from the results obtained would be that with spleen size of <12 cm varices requiring banding would not be found. However, only 10.3% (n=10) of the population studied had a spleen size within normal limits, and out of these, 8 episodes had small oesophageal varices (SV group) which would still require treatment with non-selective beta blockers.

In recently published guidelines ² recommendations were made for which cirrhotic patients could safely avoid screening at endoscopy. These guidelines stated that patients with a liver stiffness <20 kPa and with a platelet count >150 x10⁹/L have a very low risk of having varices requiring treatment, and can avoid screening endoscopy. Liver stiffness measurement was not considered in this study as this modality has only recently been introduced in Malta and is subject to numerous variables such as obesity, elevated ALT levels and liver venous or biliary congestion.¹²

The main limitation of our study was that the diagnosis of cirrhosis was based on imaging studies rather than liver biopsies. Liver biopsies are considered invasive and often avoided clinically. In future, similar studies should also include liver stiffness measurements in the diagnosis of cirrhosis. Another limitation of our study was that patients with different aetiologies of cirrhosis were recruited. Considering cirrhotic patients with a single aetiology may produce more robust conclusions. Furthermore, when collecting data on spleen size US, CT and MRI modalities were referenced depending on which imaging study was performed closest to the date of endoscopy. More accurate measurements would have been obtained had all spleen size measurements been calculated using MRI¹³ however since data was collected retrospectively this was not possible. At the start of the study several inclusion and exclusion criteria were laid out before selecting the population to be studied to improve validity of the results obtained. One limitation of the study however was that not all variables that may affect spleen size and platelet count were taken into consideration.

In conclusion, despite the relatively small size of our study the results were statistically significant and it is unlikely that more patients would have resulted in a different outcome. Having said this, the study did not yield any new recommendations to alter existing clinical practice.

References

1. Tripathi D, Stanley A, Hayes P, Patch D, Millson C, Mehrzad H, et al. UK guidelines on the management of variceal haemorrhage in cirrhotic patients. Gut. 2015;64(11):1680-1704.
2. De Franchis R, on behalf of the Baveno VI Faculty. Expanding consensus in portal hypertension. Report of the Baveno VI Consensus Workshop: Stratifying risk and individualizing care for portal hypertension. Department of Biomedical and Clinical Sciences, University of Milan, Gastroenterology Unit, Luigi Sacco University Hospital, Milan, Italy. Journal of Hepatology. 2015;63:743–752.
3. Wani Z, Bhat A, Bhadoria A, Maiwall R, Choudhury A. Gastric varices: Classification, endoscopic and ultrasonographic management. J Res Med Sci. 2015;20(12):1200-1207.
4. Biecker E. Portal hypertension and gastrointestinal bleeding: Diagnosis, prevention and management. World J Gastroenterol. 2013;19(31):5035–5050.
5. Li H, Chen T, Li Z, Zhang X, Li C, Chen X, et al. Albumin and magnetic resonance imaging-liver volume to identify hepatitis B-related cirrhosis and esophageal varices. World J Gastroenterol. 2015 Jan;21(3):988–996.
6. Chen X, Chen T, Zhang X, Li Z, Zeng N, Zhou P, et al. Platelet count combined with right liver volume and spleen volume measured by magnetic resonance imaging for identifying cirrhosis and esophageal varices. World J Gastroenterol. 2015 Sep;21(35):10184–10191.
7. Shin S, Lee J, Yu M, Yoon J, Han J, Choi B, et al. Prediction of Esophageal Varices in Patients with Cirrhosis: Usefulness of Three-dimensional MR Elastography with Echo-planar Imaging Technique. Radiology. 2014;272(1):143-153.
8. Ehimwnma O, Tobechukwu Tagbo M. Determination of normal dimension of the spleen by ultrasound. Niger Med J. 2011;52(93):198-203.
9. De Simone P, Fagiuoli S, Cescon M, De Carkis L, Tisone G, Volpes R, et al. Use of Everolimus in Liver Transplantation: Recommendations From a Working Group. Transplantation. 2016;101(2):239-25.
10. Cherian J, Deepak N, Ponnusamy R, Somasundaram A, Javanthi V. Non-invasive Predictors of Esophageal Varices. Saudi J Gastroenterol. 2011 Jan-Feb;17(1):64–68.
11. Hwi Y, Eun H, Won K, Jae Y, Hyunsik W, Sohee O, et al. The Role of Spleen Stiffness in Determining the Severity and Bleeding Risk of Esophageal Varices in Cirrhotic Patients. Medicine (Baltimore). 2015 Jun; 94(24):e1031.
12. Non-Invasive tests for evaluation of disease severity and prognosis. EASL practice guidelines. Journal of Hepatology. 2015;63:237–264.
13. Taouli B, Ehman R, Reeder S. Advanced MRI Methods for Assessment of Chronic Liver Disease. AJR Am J Roentgenol. 2009 Jul;193(1):14–27.

Figure

Test image