Our meta-analysis including five eligible studies demonstrated a prevalence of PVT in the NAFLD population of about 8% and, most importantly, that patients with NAFLD and its advanced forms have a pooled 1.34-fold increased risk of prevalent PVT, when compared to advanced liver diseases from other etiologies. At present, available data are not still conclusive on this topic and, first of all, focus on the onset of pre-transplant PVT in LT recipients, thus showing a negative impact of NAFLD etiology sustaining cirrhosis requiring LT on the thrombosis complication. This evidence provids additional support for the assertion that NAFLD and its advanced forms predispone to a prothrombotic state [10,11,12]. Along with cirrhotic complications, another small study including prospectively 94 NAFLD non-cirrhotic patients, showed the occurrence of PVT in 8% of the cohort even without advance liver disease, especially in obese subjects and those with increased leptin levels9. Nevertheless, this study does not evaluate any association with histological features of severity of NAFLD, and no other study includes patients with NAFLD without cirrhosis, so that the question of whether the increased risk for PVT is restricted to patients with NASH or applies to all patients with NAFLD remains largely unsolved so far. If on the one hand, studies on this issue are scarce, on the other hand meta-analysis evidence is even less pronounced. In fact, only one meta-analysis by Li et al. including 22 studies aiming at defining the association of PVT with cirrhosis of different etiologies, quantified the magnitude of the association between metabolic alterations and PVT and highlighted a significant association with T2DM and dyslipidemia. When focusing only on 4 studies (n = 3,385,821 patients) reporting a NAFLD etiology for cirrhosis, the authors showed a significant association between NAFLD and PVT, with a pooled random effects OR 1.61 (95% CI 1.34–1.95) [13]. This risk was slightly higher compared to those found in our analysis. However, this study did not specifically focus on NAFLD, but on metabolic abnormalities and reported a transversal association between PVT and liver disease without prompting the drive of a final causative relationship. In addition, in this analysis, a statistically significant between-study heterogeneity was also observed (I2 = 75%) [14], similar to those found in our study. Another small meta-analysis including 3 retrospective studies with cirrhotic patients from different etiologies, demonstrated an independent association between NAFLD and PVT (random effects OR 2.12, 95% CI 1.45–3.09). However, the solidity of reports was weak as two articles were conference abstract and only one article was an original one [13].

Our meta-analysis has some important limitations that are strictly inherent to the design of the included studies. First, the cross-sectional design of the elibible studies does not allow establishing any temporal and causal association between NAFLD and PVT. Second, although the eligible studies adjusted their results at least for age, sex, smoking, obesity and T2DM, the possibility of residual confounding by some unmeasured (or unknown) factors cannot be ruled out. For example, the majority of the eligible studies reported incomplete adjustments for some important risk factors, such as waist circumference, drug use, procoagulant state. Moreover, in most of the studies, associated oncological diseases were excluded (except for HCC) and probably also prothrombotic states considering that the cohort analysed came from transplant cohorts. However, it is not possible to exclude with certainty any conditions conducive to thrombotic states. However, the NOS quality scale of the eligible studies suggested an overall medium risk of bias. Third, another limitation of the meta-analysis is that the eligible studies used imaging techniques, but none of them used liver biopsy, which is the reference standard for diagnosing and staging liver disease. In addition, almost all studies included Caucasian populations, thus limiting the generalization of results to other ethnicities and pre-transplant population, so selecting a category of patients with advanced liver disease. Fourth, our results should be interpreted with caution, because of the high heterogeneity across the eligible studies. Notably, specific subanalysis by study country and NOS did not substantially reduce the high heterogeneity we observed. In addition, we were unable to perform additional sub-analyses or even meta-regressions (given the relatively small number of the studies included) to further assess the causes of heterogeneity. Therefore, speculatively, we believe that heterogeneity in our meta-analysis may be mainly due to different inclusion criteria of patients, limitation of ultrasound in diagnosing NAFLD and PVT, as well as to different covariates used in the eligible studies.Not with standing these limitations, this is the first uptodate meta-analysis of observational studies available in literature, which specifically focus its attention on the association between NAFLD and PVT. It is beyond the scope of this meta-analysis to discuss in depth the putative underlying mechanisms by which NAFLD and its advanced forms may contribute to the development of PVT. We confirm this association possibly sustaining the higher hypercoagulability state of NAFLD compared to other etiologies. Whether NAFLD is associated with an increased risk of PVT simply as a consequence of metabolic comorbidities which characterize the liver disease, or if NAFLD, especially in its advanced forms, may contribute to its development remains an open question.

In conclusion, our meta-analysis of observational studies reported that patients with NAFLD and its advanced forms have an increased risk of prevalent PVT, when compared to advanced liver diseases from other etiologies. Our results may suggest that clinicians should have particularly attention on the procoagulant state in patients with NAFLD, especially those with advanced forms.

However, further larger studies, possibly prospective, are warranted to confirm this evidence.