Spotlight on Impactful Research: Long‐Term Calcineurin Inhibitor Therapy and Brain Function in Patients After Liver Transplantation

Short abstract

http://aasldpubs.onlinelibrary.wiley.com/hub/journal/10.1002/(ISSN)2046-2484/video/15-4-reading-cheung a video presentation of this article

- http://aasldpubs.onlinelibrary.wiley.com/hub/journal/10.1002/(ISSN)2046-2484/video/15-4-interview-cheung an interview with the author

https://www.wileyhealthlearning.com/Activity/7088571/disclaimerspopup.aspx questions and earn CME

Abbreviations

CKD

chronic kidney disease

CNI

calcineurin inhibitor

MRI

magnetic resonance imaging

mTOR

mammalian target of rapamycin

RBANS

Repeatable Battery for the Assessment of Neuropsychological Status

VWCN

ventricular width at the level of the caudate nucleus

WMH

white matter hyperintensity

Calcineurin inhibitors (CNIs), tacrolimus and cyclosporine, remain the cornerstone of immunosuppression regimens after liver transplantation, although variations may be taken in patients with neurological complications, renal dysfunction, metabolic complications, infections, or increased risk for malignancy, including recurrence of hepatocellular carcinoma or development of de novo malignancies. CNI reduction or elimination alongside use of nonnephrotoxic immunosuppression (mycophenolic acid or mammalian target of rapamycin [mTOR] inhibitors) is a strategy often recommended to achieve long‐term improvement in renal function1 with potential for additional benefits to prevent the other complications noted earlier. Independent risk factors for CNI‐related neurotoxicity include hepatic encephalopathy prior to transplant, pretransplant sodium levels, and older donor age.2 Although the neurological complications related to CNI therapy have been described in the literature, this study by Pflugrad et al.3 sheds further light on the long‐term effects of CNIs on brain function and structure.

This single‐center observational study sought to find a dose‐dependent effect of CNIs on brain function measured by the RBANS (Repeatable Battery for the Assessment of Neuropsychological Status) test and brain structure quantified by magnetic resonance imaging (MRI) findings of hyperintensities and evidence of brain atrophy. The low‐dose versus standard‐dose CNI groups were split based on tacrolimus or cyclosporine trough levels less than or greater than 5 and 50 μg/L, respectively. The medication regimen for the CNI‐free group included various combinations of mycophenolic acid, azathioprine, mTOR inhibitors, and corticosteroids in patients who had previously been treated with CNI therapy for 4.2 ± 3.8 years but had not received any CNI therapy for 8 ± 2.8 years prior to the study. The median time since transplant for these cohorts was 10 years. The control group was a cohort of healthy patients who did not have a liver transplant or CNI exposure that was matched for age, sex, and education level.

The low‐dose CNI group scored the lowest in nearly all of the individual domains of the RBANS, followed by standard‐dose CNI, CNI‐free, and then finally the control group. Patients receiving low‐dose (P = 0.01) and standard‐dose (P = 0.04) CNIs had significantly worse scores in visuospatial/constructional ability compared with the control group. Even though both CNI treatment groups scored worse on the total RBANS score, only the low‐dose CNI group was statistically different (P = 0.01) from the control group. This interesting observation of the lower total RBANS score including the individual index scores in the low‐dose CNI group compared with the standard‐dose CNI group may suggest ongoing susceptibility to the neurotoxic side effects of the CNIs despite lowering the dose in those with impaired renal function, whereas patients who were later CNI‐free did not demonstrate these long‐term effects. Is there a correlation between the degree of nephrotoxicity and neurotoxicity and how much is reversible in each case? The answer to this question will shed light on whether we should be advocating for CNI‐free regimens in all patients with renal dysfunction. The current practice for most transplant hepatologists includes lowering the CNI dose in patients with renal dysfunction to prevent further injury while still ensuring prevention of organ rejection. As the organ allocation rules shift to limit simultaneous liver‐kidney transplants, there are increasing numbers of patients with chronic kidney disease (CKD), particularly with the growing number of patients with nonalcoholic steatohepatitis who have additional CKD risk factors both before and after transplantation.

A negative correlation was found between RBANS sum score and the total number of years on standard‐dose CNIs (r = −0.4, P = 0.04), total tacrolimus dose (r = −0.4, P = 0.04), and tacrolimus mean trough (r = −0.5, P = 0.01). However, similar results were not observed with cyclosporine therapy. Because there appears to be a difference between tacrolimus and cyclosporine, it would be interesting to see this data further delineated between the two, including total cumulative doses for each and how they correlate with the endpoints. In particular, what happens to the patients who were initially treated with tacrolimus but later switched to cyclosporine, or vice versa? Use of tacrolimus, rather than cyclosporine, has improved graft and patient survival in the first year after transplant with fewer episodes of rejection,4 but tacrolimus is associated with greater neurotoxicity, as supported by this study. Thus, application of these data to clinical practice may suggest the use of tacrolimus for the first year when patients are relatively more immunogenic and subsequently switching the patient to cyclosporine or discontinuation of CNIs for appropriately selected patients to circumvent the long‐term effects on the neurological system.

Patients treated with CNIs had significantly more parietal (P = 0.03) and temporal (P = 0.01) white matter hyperintensities (WMHs) and brain atrophy as defined by a broader ventricular width at the level of the caudate nucleus (VWCN). Positive correlation was found between total WMH and mean trough level for both tacrolimus (r = 0.4, P = 0.03) and cyclosporine (r = 0.04, P = 0.005). Differences in parietal WMH and VWCN were found in patients with diabetes and hypercholesterolemia, respectively. Of note, the presence of hepatic encephalopathy, posttransplant encephalopathy, and stage 3 CKD did not have an impact on the measurements of brain structure. VWCN was an independent prognostic factor of RBANS score.

There are a few limitations worth mentioning. The patient population in this study is likely not representative of those seen in a typical liver transplant center. Nonalcoholic steatohepatitis and alcohol are now the leading indications for transplantation,5 but those diagnoses included only a small portion of the patients in this study. The RBANS is a tool that was originally developed to detect dementia. Although its utility as a diagnostic tool in mild cognitive impairment has been hampered by poor sensitivity, its high specificity makes it a reasonable tool for exclusion.6 Lastly, more longitudinal data would be helpful to further understand these patients, that is, obtaining the RBANS and MRI data prior to transplant and at additional time points after transplant. The control group in this study was not patients who received liver transplant, so direct comparison is not possible because the patients were not subjected to the potential neurotoxicities from the decompensated cirrhotic state. Data on patients prior to transplant and at multiple times after transplant would further delineate the neurotoxicities that are attributed to the immunosuppressive agents and whether some may be reversible. For instance, in the CNI‐free group, these patients were subjected to CNI therapy immediately after transplant and then eventually had removal of CNIs. Comparing the neurological data on these patients while they were receiving CNIs and then later when CNI‐free will assist in determination of reversibility potential.

Further research is needed to make better recommendations regarding immunosuppression regimens as they relate to minimization of neurotoxicities. Screening for early evidence of neurotoxicity will likely soon become a routine part of posttransplant care and may be done as proposed by the authors of this study. Although the observed decrease in brain volume early after transplantation may simply be a result of resolving cerebral edema from the cirrhotic state, particularly in the setting of hepatic encephalopathy, later and more marked brain atrophy alongside an increase in WMH is more likely related to the cardiometabolic side effects of immunosuppression agents.7 Notably, however, the other immunosuppression classes, including mTOR inhibitors and corticosteroids, also have cardiometabolic side effects,8 so switching to these medications from CNIs likely will not fully circumvent the volume loss that is observed.

As more patients are able to achieve operational tolerance after liver transplantation,9 this will be an additional and unique subset of patients to determine the long‐term neurological effects simply from the cirrhotic state prior to transplant and to truly delineate the effects of the immunosuppressive medications. Until operational tolerance becomes the norm and while we remain dependent on immunosuppressive agents to maintain longevity of graft organs, there remains an ongoing need to develop therapies that have less toxicity than CNIs while having equal or greater effectiveness in prevention of organ rejection.