We hypothesized that subjects treated with angiotensin blockers would have decreased levels of fibrosis as measured by noninvasive indices when compared to subjects not exposed to these medications. Conversely, we found that angiotensin blockade in a cohort of coinfected subjects did not attenuate the progression of liver fibrosis. In fact, there was a statistically significant correlation of worsening fibrosis on the Forns index for subjects who had taken ACE-Is/ARBs for three years compared to subjects who had not been exposed to these medications. While not statistically significant, the trend in the ACE-I/ARB group was progressively worse in all groups at one year and continued to worsen when going back three years. This finding is in contrast to prior data that portrayed an anti-fibrotic effect of angiotensin inhibition. To our knowledge, though, this is the only study that has looked at angiotensin inhibition in subjects with coinfection.
There are a few plausible explanations for our conflicting results. First, we did not distinguish between subjects who were taking an ARB or an ACE-I. A recent study in bile-duct-ligated mice suggests that ARBs may be more effective in suppressing hepatic fibrosis compared to ACE-Is [57
]. A future study may show different results depending on the method of angiotensin suppression. In addition, there was no standardization of dose of ACE-I or ARB. It is possible that high doses of these medications could lower blood pressure enough to impair liver perfusion causing worsening of fibrosis scores. Doses of these medications should be standardized in future research.
Another possible explanation is that subjects taking an ACE-I or ARB had more unmeasured comorbidities than subjects not on these medications. The ACE-I or ARB may have been added for HIV-associated or diabetic nephropathy, hypertension, or heart failure. It is possible that these subjects appeared to have elevated fibrosis scores because they were sicker than the group that did not require these medications. On multivariate analysis, though, ACE-I/ARB use was independently associated with an elevated Forns score, after controlling for co-morbidities such as diabetes. In addition, older subjects and subjects of black race had significantly higher Forns scores, which is consistent with data from prior studies that have shown worsening disease and poorer treatment responses in these groups [58
]. Subjects in the control group were younger and were significantly less likely to be black when compared to the ACE-I/ARB group, which could have created a healthier control group.
There also could be a deleterious interaction between HIV positivity and angiotensin blockade or between ART for HIV and ACE-Is/ARBs. On multivariate analysis, lower levels of HIV viral load were associated with elevated Forns scores. It is possible that subjects who had lower HIV viral loads were taking ART and had medication-induced liver toxicity either from the combination of ACE-Is/ARBs and ART or from ART alone. This explanation could be examined by comparing fibrosis indices in a similar cohort of subjects with HCV monoinfection who have been treated with an ACE-I or ARB.
A recent study in CHC monoinfection relates the possibility that the effects of angiotensin inhibition may occur earlier in liver fibrosis and may be missed in persons with advanced liver disease, often seen in coinfection [60
]. This study used data from the Hepatitis C Long-term Treatment against Cirrhosis (HALT-C) Trial [61
] to evaluate the effects of continuous ACE-I/ARB use for 3.5 years on liver fibrosis progression as determined by serial liver biopsies [60
]. Researchers were unable to demonstrate a benefit for angiotensin blockade in attenuating progression of liver fibrosis: 33.3% of subjects on an ACE-I/ARB had a 2-point increase in fibrosis, compared to 32.5% and 25.7% of subjects on other antihypertensives or on no medications, respectively (P
= 0.21). The authors comment that the majority of their subjects had significant fibrosis at baseline, which is similar to our coinfected cohort. If angiotensin inhibitors exert maximal antifibrogenic effects early in the fibrosis process and have decreased activity at later stages of fibrosis, the affects of these medications may be missed in a cohort with advanced liver disease. This may especially be true in coinfection as the progression of liver injury is often more rapid than in HCV monoinfection; thus, it would be prudent to determine the anti-fibrotic nature of these medications soon after diagnosis of HCV, before the development of liver fibrosis.
One reason for the limited number of human studies in liver fibrosis is the need for invasive measurements, such as liver biopsy, to record the progression of liver disease. Moreover, liver biopsy has been shown to be a less than ideal gold standard for comparison. Recent research has focused on noninvasive indices of liver fibrosis, using common laboratory values to estimate the severity of liver disease. Three of these indices, the Forns score, the APRI, and the FIB-4, were used and validated in the present study to document liver fibrosis in subjects with coinfection. We found a correlation between the presence or lack of significant fibrosis as determined by radiological and biopsy data and level of fibrosis as determined by the use of these noninvasive indices in coinfected subjects. The PPVs were too low to be considered clinically useful, but the NPVs for all of the indices were high enough to make them clinically relevant in ruling out significant fibrosis. This is important as these indices may be used to identify persons at low risk for significant fibrosis and possibly decrease the need for liver biopsies in this group. At more advanced levels of immunosuppression from HIV it seems possible that some of these markers would be susceptible to error, such as from HIV-associated thrombocytopenia. The present study validated these indices despite a range of CD4 counts and HIV viral loads. The best use of these indices in a coinfected population would be to rule out significant fibrosis and avoid biopsy, possibly in a patient who is delaying treatment and has no other signs of hepatic inflammation.
The current study has limitations that prevent definitive conclusions on the effects of angiotensin inhibition. Due to the retrospective nature of the study, causation cannot be inferred. Prospective and ideally randomized controlled trials are needed to determine if these medications may have hepatic toxicity alone, or in combination with ART. Also, fibrosis scores became higher overtime, as cumulative exposure to an ACE-I or ARB was increasing. There were a small number of subjects on ACE-Is/ARBs consistently over the years, though, which may have limited our power to detect a significant difference between the groups earlier than three years. To verify this trend, further research should follow a larger number of subjects over a longer period of time, including the time preceding significant liver fibrosis. Last, we did not have data on other mechanisms of fibrosis, such as non-alcoholic fatty liver disease (NAFLD). Although we collected data on body mass index and diabetes, we did not always have a biopsy-proven diagnosis of NAFLD, which may result in higher fibrosis scores despite being on an ACE-I or ARB.