SciELO - Scientific Electronic Library Online

vol.108 issue2Overexpression of Aquaporin 1 on cysts of patients with polycystic liver diseaseImpact of a gluten-free diet on bone mineral density in celiac patients author indexsubject indexarticles search
Home Pagealphabetic serial listing  


Services on Demand




Related links

  • On index processCited by Google
  • Have no similar articlesSimilars in SciELO
  • On index processSimilars in Google


Revista Española de Enfermedades Digestivas

Print version ISSN 1130-0108

Rev. esp. enferm. dig. vol.108 n.2 Madrid Feb. 2016




Idiopathic portal hypertension regarding thiopurine treatment in patients with inflammatory bowel disease

Hipertensión portal idiopática secundaria a tratamiento tiopurínico en pacientes con enfermedad inflamatoria intestinal



Cristina Suárez-Ferrer, Elba Llop-Herrera, Marta Calvo-Moya, María Isabel Vera-Mendoza, Irene González-Partida, Yago González-Lama, Virginia Matallana-Royo, José Luis Calleja-Panero and Luis Abreu-García

Division of Gastroenterology. Hospital Universitario Puerta de Hierro Majadahonda. Madrid, Spain





Introduction: The possibility of developing idiopathic portal hypertension has been described with thiopurine treatment despite compromises the prognosis of these patients, the fact its true prevalence is unknown.
Material and methods: A cross-sectional study was conducted in a cohort of inflammatory bowel disease (IBD) patients followed at our unit, to determine the prevalence of diagnosis of idiopathic portal hypertension (IPH) and its relationship with thiopurine treatment.
Results: At the time of the analysis, 927/1,419 patients were under treatment with thiopurine drugs (65%). A total of 4 patients with IBD type Crohn's disease with idiopathic portal hypertension probably related to the thiopurine treatment were identified (incidence of 4.3 cases per 1,000). Seventy-five percent of patients started with signs or symptoms of portal hypertension. Only one patient was asymptomatic but the diagnosis of IPH because of isolated thrombocytopenia is suspected. However, note that all patients had thrombocytopenia previously. Abdominal ultrasound with fibroscan, hepatic vein catheterization and liver biopsy were performed on all of them as part of the etiology of portal hypertension. In the abdominal ultrasound, indirect portal hypertension data were observed in all patients (as splenomegaly) cirrhosis was also ruled out. The fibroscan data showed significant liver fibrosis (F2-F3).
Conclusion: Idiopathic portal hypertension following thiopurine treatment in IBD patients is a rare occurrence, but it must be borne in mind in the differential diagnosis for early diagnosis, especially in patients undergoing thiopurine treatment over a long period. The presence of thrombocytopenia is often the only predictor of its development in the preclinical stage.

Key words: Idiopathic portal hypertension. Inflammatory bowel disease. Thiopurine treatment.


Introducción: entre los efectos adversos hepáticos secundarios al tratamiento tiopurínico en pacientes con enfermedad inflamatoria intestinal (EII), se ha descrito la posibilidad de desarrollar hipertensión portal idiopática. Esta patología de etiología y prevalencia real inciertas puede comprometer el pronóstico de estos pacientes, por lo que se debe tener un alto grado de sospecha para su diagnóstico precoz.
Material y métodos: se ha llevado a cabo un estudio transversal en una cohorte de pacientes con EII en seguimiento en nuestra unidad para determinar la prevalencia del diagnóstico de HTP idiopática (HTPI) y su relación con el tratamiento tiopurínico.
Resultados: en nuestro centro, en el momento del análisis había 1.419 pacientes en seguimiento por enfermedad inflamatoria intestinal. De estos, 927 pacientes se encuentran bajo tratamiento con fármacos tiopurínicos (o lo han estado durante la evolución de su enfermedad), lo que supone el 65,3% de la población: 689 pacientes con azatioprina (74,3%) y 238 con 6-mercaptopurina (25,7%). En total, se identificaron 4 pacientes con EII tipo enfermedad de Crohn con hipertensión portal idiopática en probable relación con el tratamiento tiopurínico, lo que supuso un 4,3% del total, es decir, una incidencia de 4,3 casos por cada 1.000 pacientes con EII tratados con tiopurínicos. Las características basales de los pacientes se describen en la tabla I. El 75% de los pacientes debutó con signos o síntomas de hipertensión portal: 1 paciente con encefalopatía hepática y 2 pacientes con hemorragia digestiva por varices esofágicas. Solo un paciente se encontraba asintomático, pero se sospechó el diagnóstico de HTP por trombopenia aislada. No obstante, cabe destacar que todos los pacientes presentaban trombopenia previamente aunque no se había sospechado el diagnóstico de HTP a pesar de un exhaustivo estudio. A todos los se realizó ecografía abdominal con fibroscan, cateterismo de venas suprahepáticas, así como biopsia hepática como parte del estudio etiológico de hipertensión portal. En la ecografía abdominal se objetivaron datos indirectos de HTP en todos los pacientes (como esplenomegalia), descartándose asimismo cirrosis hepática. El fibroscan mostraba datos de fibrosis hepática significativa (F2-F3). Además, a todos los pacientes se les realizó una angiorresonancia en la que se descartó trombosis del eje esplenoportal como causa de HTP. Por último, la anatomía patológica de la biopsia hepática descartó la presencia de cirrosis hepática, apoyando el diagnóstico de HTP idiopática (Tabla II).
Conclusiones: la hipertensión portal idiopática secundaria a tratamiento tiopurínico en pacientes con enfermedad inflamatoria intestinal es un fenómeno poco frecuente, pero ha de ser tenido en cuenta en el diagnóstico diferencial para un diagnóstico precoz, principalmente en pacientes con tratamiento tiopurínico de larga evolución. La presencia de trombopenia es a menudo el único factor predictor de su desarrollo en fases preclínicas.

Palabras clave: Hipertensión portal idiopática. Enfermedad inflamatoria intestinal. Tratamiento tiopurínico.



Between the secondary hepatic adverse effects to the thiopurine treatment in patients with inflammatory bowel disease (IBD), there has been described the possibility of developing idiopathic portal hypertension (IPH). This pathology with uncertain etiology can compromise the prognosis of these patients, therefore, it is necessary to have a high degree suspicion for an early diagnosis.

IPH is a rare condition of unknown etiology that covers various syndromes with similar clinical characteristics. To make the diagnosis, one needs to establish unmistakable signs of portal hypertension and to rule out liver cirrhosis and other specific causes of liver disease. The definitive diagnosis is obtained with representative pathology revealing the presence of hepatoportal sclerosis and/or obliterative venopathy (obliteration of small venules portals in fibrous tracts) (1).

One variant of idiopathic portal hypertension is nodular regenerative hyperplasia (NRH). It has been reported in patients treated with thiopurine drugs for IBD. It is a type of dose-dependent injury characterized by damage to the endothelial cells of the sinusoids and hepatic veins, resulting in non-thrombotic occlusion of the blood vessels and the subsequent development of fibrosis and portal hypertension (1-3). Large data sets are not available and only isolated cases have been reported in which thiopurine treatment appears to be the main cause of the syndrome (4). The onset of IPH usually occurs between 3 months and 3 years into the thiopurine treatment. Although the pathogenic mechanism is unknown, it appears to be due to the depletion of glutathione by the use of azathioprine and mercaptopurine (5).

The HNR can produce non-cirrhotic portal hypertension with the onset of esophageal varices, splenomegaly and ascites. These changes can occur even when analysis does not reveal signs of abnormal liver function. This pathology should therefore be suspected in all patients who: a) are receiving thiopurine treatment (or who have received it previously); and b) who exhibit portal hypertension but no other liver disease (6).


Material and methods

A cross-sectional study was conducted in a cohort of IBD patients followed at our unit to determine the prevalence of diagnosis of IPH and its relationship with thiopurine treatment.

For that, the clinical histories of all the patients who are at present in pursuit in our IBD unit have been reviewed. There have been initially selected for the study those patients who were presenting a sign of liver disease (analytical, clinical or radiological alteration) and who were or had been low thiopurine treatment for their IBD.

It was considered to be a significant analytical alteration:

- Pronounced elevation of transaminases, understanding as such an elevation of at least the double of the top profile of the normality.

- Moderate elevation of liver test but remaining stable (for at least one month) or that it requires intervention in some sense for its correction (for example, stopping thiopurine treatment).

- Trombopenia not explained by hematological disorders, whose origin could be liver failure and/or portal hypertension.

We analyzed the baseline characteristics of patients with IPH which included a complete analysis (blood count and biochemistry) to study liver disease (serology, auto-immunity, lysosomal storage diseases protein, immunoglobulins and alpha-1-antitrypsin), and an abdominal ultrasound scan with transient elastography (fibroscan).

A study was also made of portal hypertension including hepatic vein catheterization, gastroscopy for screening of esophageal varices and magnetic resonance angiography (to rule out vascular pathology of the splenoportal axis) and transjugular liver biopsy for definitive diagnosis in all patients.

In the statistical analysis, values are represented as median and range, while categorical data is represented as frequencies (%) and range. Comparisons between groups on continuous variables were performed using Student's t-test or ANOVA as appropriate. Categorical data were compared using the Chi square test. Statistical significance is set at a level of p < 0.05. Statistical analyses were carried out with SPSS 20.0 (SPSS inc., Chicago IL, USA) for Mac.



At the time of analysis, our centre had 1,419 patients being monitored for IBD (ulcerative colitis, Crohn's disease and indeterminate colitis). Of these, 927 patients were under treatment with thiopurine drugs (or have been during the evolution of their disease), representing 65.3% of the population: 689 patients with azathioprine (74.3%) and 238 with 6-mercaptopurine (25.7%).

In the selected sample, 63 patients (6.8% of the patients who had received thiopurines) developed liver disease related to the thiopurine treatment, all other causes of liver disease having been excluded.

Most of them (59 patients, 93.6%) were diagnosed of hepatotoxic of thiopurines, producing in all of them an improvement of the liver test after the decrease of the dose and/or stopping the treatment. Nevertheless, it has to be emphasized that 11 patients were diagnosed of concomitant non alcoholic fatty liver disease.

A total of 4 patients with IBD type Crohn's disease with idiopathic portal hypertension probably related to the thiopurine treatment were identified, representing 4.3% of the total (i.e. an incidence of 4.3 cases per 1,000 IBD patients treated with thiopurine). The baseline characteristics of the patients are described in table I.



Seventy-five percent of patients started with signs or symptoms of portal hypertension: one patient with hepatic encephalopathy and 2 patients with bleeding esophageal varices. Only one patient was asymptomatic but with a suspected diagnosis of IPH as a result of isolated thrombocytopenia. However, one should note that all patients previously had thrombocytopenia although diagnosis of portal hypertension was not suspected despite an exhaustive study.

Abdominal ultrasound with fibroscan, hepatic vein catheterization and liver biopsy were performed on all the patients as part of the etiology of portal hypertension. In the abdominal ultrasound, indirect portal hypertension data were observed in all patients (as splenomegaly) thus also ruling out cirrhosis. The fibroscan data showed significant liver fibrosis (F2-F3). In addition, all patients were subjected to resonance angiography where splenoportal axis thrombosis was ruled out as a cause of IPH. Finally, the pathology of the liver biopsy ruled out the presence of liver cirrhosis, supporting the diagnosis of IPH (Table II).



After the diagnosis of IPH probably stemming from thiopurine treatment, all patients were monitored by the hepatology unit with diagnosis and therapeutic management following clinical practice for patients with portal hypertension. Cases of bleeding varices received endoscopic treatment by ligation of varices with bands and later secondary prophylaxis of gastrointestinal bleeding from esophageal varices with beta-blockers. In addition, treatment with thiopurine drugs was suspended in all cases and it was not re-introduced at any point.



IPH is an uncommon liver disease of unknown etiology that includes a group of disorders with similar clinical characteristics, encompassing a wide spectrum of histological alterations (hence the use of other terms to describe the condition, such as hepatoportal sclerosis, non-cirrhotic portal fibrosis, incomplete septal cirrhosis and nodular regenerative hyperplasia) (7,8). This fact probably reflects different stages of the disease or multiple pathological processes that converge in a syndrome with similar clinical characteristics.

The diagnosis of IPH is a diagnosis by exclusion (i.e., occurring in patients with clinical-radiological findings of portal hypertension without cirrhosis and without vascular changes (with permeability splenoportal axis and hepatic veins) (9-11). The diagnosis is established by pathology, although this ranges from minor changes to the presence of NRH.

The literature shows cases of secondary IPH following treatment with thiopurine drugs in patients with IBD, mostly described as NRH. This and other adverse hepatic effects (such as veno-occlusive disease or peliosis) can arise from between 3 months and 3 years of treatment with thiopurine (12,13).

In our series, we identified 4 patients with unequivocal diagnosis of IPH, with the thiopurine treatment as the only objective causal factor. Nevertheless, although an exhaustive analysis was made to diagnose all possible cases, it has the limitations characteristic of a transversal study.

One should note that in our study all patients had long had IBD Crohn's disease (mean 20 years) and had been treated with thiopurine drugs for a long period (mean 88 months, range 30-120 months). No adverse effects of this nature were found in patients who had not received at least two years of treatment.

IPH, unlike cirrhosis, does not usually lead to abnormal test results in the early stages of the disease. In the case of IPH, liver test results tend to be normal (in the absence of cholestasis or cytolysis) (14). Only the presence of thrombocytopenia may predict the development of IPH (15). In our series, in up to 100% of cases the presence of thrombocytopenia was the first sign of IPH, although as clinical suspicion was low probably due to the low incidence of this disease, the diagnosis was not established until there were other signs and symptoms of portal hypertension. That is, the isolated thrombocytopenia may have been a predictor for the development of portal hypertension. Therefore, in the presence of thrombocytopenia in patients suffering from IBD and treated with thiopurine drugs, exhaustive study of liver disease is needed to discriminate between patients with IPH (even when liver function tests are strictly normal). In fact, in patients with compensated cirrhosis, a figure of platelets lower than 150,000 platelets/mm3 was independently associated with the presence of esophageal varices (16). In our series, all patients had a platelet count of 100,000/mm3 or lower, so if the suspicion of portal hypertension had been high, it could have been a predictor.

Regarding liver function tests, our patients had no major alterations of liver function (GOT average: 50 U/L, average GPT: 34 U/L and average GGT 175 U/L).

This insidious, silent, presentation means that, in most reported cases, the first clinical manifestation is portal hypertension. In our study, 3 patients (75%) were diagnosed following a sign or symptom of portal hypertension, two due to gastric bleeding esophageal varices and the other due to an episode of hepatic encephalopathy. Although these symptoms are the proper ones of the portal hypertension for other causes, it have been reported that variceal bleeding is most frequent, affecting two thirds of the patients (17,18). This explains why the diagnosis of IPH is delayed by the difficulty of diagnosing early pre-clinical stages.

Characteristically, patients with idiopathic portal hypertension show higher values in transient elastography but these are below the threshold described for predicting the presence of clinically significant portal hypertension (18.4 Kpa) (19). In our study, no patient had a normal value in the fibroscan although, as discussed above, it was less than 13 kPa (which corresponds to an F4 stage because all patients exhibited values corresponding to F2-F3).

All patients exhibited portal hypertension when their liver hemodynamic was analyzed, although not all of them showed clinically significant values of portal hypertension. Note that even in patients with clinical manifestations of portal hypertension (such as bleeding gastro-esophageal varices), the portal pressure gradient measured by catheterization of supra-hepatic veins was not clinically significant. In other words, patients with IPH exhibited lower clinical gradients than those estimated for patients with secondary portal hypertension due to cirrhosis, probably because the measurement of hepatic venous pressure gradient translates to sinusoidal pressure and does not allow one to infer the pre-sinusoidal pressure that is characteristic of IPH (20).

The clinical course of patients was similar to that of other patients diagnosed with non-cirrhotic portal hypertension from other causes. Fifty per cent of patients required diuretic therapy due to the presence of ascites. Response to such treatment was good.



We can conclude that IPH following thiopurine treatment in patients with inflammatory bowel disease is a rare occurrence but it must be borne in mind in the early differential diagnosis, particularly in patients undergoing thiopurine treatment over a long period. The presence of thrombocytopenia is often the only predictor of its development at the pre-clinical stage.



1. De Boer NK, Mulder CJ, Van Bodegraven AA. Nodular regenerative hyperplasia and thiopurines: The case for level-dependent toxicity. Liver Transpl 2005;11:1300-1. DOI: 10.1002/lt.20554.         [ Links ]

2. Leon-Montañes R, Trigo-Salado C, Leo-Carnerero E, et al. Treatment with adalimumab in a patient with regenerative nodular hyperplasia secondary to azathioprine. Rev Esp Enferm Dig 2013;105:168-70. DOI: 10.4321/S1130-01082013000300009.         [ Links ]

3. Selsik P, Mary JY, Beaugerie L, et al. Incidence of nodular regenerative hyperplasia in inflammatory bowel disease patients treated with azathioprine. Inflamm Bowel Dis 2011;17:565-72. DOI: 10.1002/ibd.21330.         [ Links ]

4. Vernier-Massouille G, Cosnes J, Lemann M, et al. Nodular regenerative hyperplasia in patients with inflammatory bowel disease treated with azatioprine. Gut 2007;56:1404-9. DOI: 10.1136/gut.2006.114363.         [ Links ]

5. Calabrese E, Hanauer SB. Assessment of non-cirrhotic portal hypertension associated with thiopurine therapy in inflammatory bowel disease. J Crohns Colitis 2011;5:48-53. DOI: 10.1016/j.crohns.2010.08.007.         [ Links ]

6. Arnott ID, Ghosh S. Portal hypertension in the presence of minimal liver damage in Crohn's disease on long-term azathioprine: Possible endothelial cell injury. Eur J Gastroenterol Hepatol 2000;12:569-73. DOI: 10.1097/00042737-200012050-00016.         [ Links ]

7. Leal-Valdivielso C, Naves JE, Mañosa M, et al. Portal hypertension in patients with inflammatory bowel disease. Gastroenterol Hepatol 2010;33:297-302.         [ Links ]

8. Venkatesh PG, Navaneethan U, Shen B. Hepatobiliary disorders and complications of inflammatory bowel disease. J Dig Dis 2011;12:245-56. DOI: 10.1111/j.1751-2980.2011.00511.x.         [ Links ]

9. Forman LM, Lucey MR. Predicting the prognosis of chronic liver disease: An evolution from Child to MELD. Hepatology 2001;33:473-5. DOI: 10.1053/jhep.2001.22481.         [ Links ]

10. Friedrich-Rust M, Ong MF, Martens S, et al. Performance of transient elastography for the non-invasive assessment of liver fibrosis: A meta-analysis. Gastroenterology 2008;134:960-74. DOI: 10.1053/j.gastro.2008.01.034.         [ Links ]

11. Yoshioka K, Kawabe N, Hashimoto S. Transient elastography: Applications and limitations. Hepatol Res 2008; 38:1063-8. DOI: 10.1111/j.1872-034X.2008.00386.x.         [ Links ]

12. Vernier-Massouille G, Cosnes J, Lemann M, et al. Nodular regenerative hyperplasia in patients with inflammatory bowel disease treated with azatioprine. Gut 2007;56:1404-9. DOI: 10.1136/gut.2006.114363.         [ Links ]

13. Navaneethan U, Shen B. Hepatopancreatobiliary manifestations and complications associated with inflammatory bowel disease. Inflamm Bowel Dis 2010;16:1598-619. DOI: 10.1002/ibd.21219.         [ Links ]

14. Schouten N, Garcia Pagan JC, Valla DC, et al. Idiopathic noncirrhotic portal hypertension. Hepatology 2011;54:1071-81. DOI: 10.1002/hep.24422.         [ Links ]

15. Angulo P, Hui JM, Marchesini G, et al. The NAFLD fibrosis score: A noninvasive system that identifies liver fibrosis in patients with NAFLD. Hepatology 2007;45:846-54. DOI: 10.1002/hep.21496.         [ Links ]

16. Zein CO, Lindor KD, Angulo P, et al. Prevalence and predictors of esophageal varices in patients with primary sclerosing cholangitis. Hepatology 2004;39:204-10. DOI: 10.1002/hep.20029.         [ Links ]

17. Khanna R, Sarin SK. Non-cirrhotic portal hypertension - Diagnosis and management. J Hepatol 2014;60:421-41. DOI: 10.1016/j.jhep.2013.08.013.         [ Links ]

18. Sarin SK, Kumar A, Chawla YK, et al. Noncirrhotic portal fibrosis/idiopathic portal hypertension: APASL recommendations for diagnosis and treatment. Hepatology International 2007;1:398-413 DOI: 10.1007/s12072-007-9010-9.         [ Links ]

19. Seijo S, Reverter E, Miquel R, et al. Role of hepatic vein catheterization and transient elastography in the diagnosis of idiopathic portal hypertension. Dig Liver Dis 2012;44:855-60. DOI: 10.1016/j.dld.2012.05.005.         [ Links ]

20. Bissonnette J, Généreux A, Côté J, et al. Hepatic hemodynamics in 24 patients with nodular regenerative hyperplasia and symptomatic portalhypertension. J Gastroenterol Hepatol 2012;27:1336-40. DOI: 10.1111/j.1440-1746.2012.07168.x.         [ Links ]



Cristina Suárez Ferrer.
Division of Gastroenterology.
Hospital Universitario Puerta de Hierro Majadahonda.
C/ Manuel de Falla, 1.
Majadahonda, Madrid. Spain

Received: 05-08-2015
Accepted: 24-09-2015