Objectives: Neurologic complications are more common in liver transplants than in other solid-organ transplants. One such neurologic complication, peripheral neuropathy, may cause functional limitations for recipients and have a negative effect on posttransplant quality of life. We aimed to examine the risk factors associated with the occurrence of clinical neuropathy after liver transplant and to investigate the frequency of sensory deficits.
Materials and Methods: In this case-control study, we analyzed factors from medical records of 63 recipients who underwent living donor liver transplant during the period from January 2010 to December 2016. A neuropathy symptom score was assigned to identify the patients who had clinical neuropathy (case group) and the patients without clinical neuropathy (control group). Quantitative sensory testing was performed to measure the warm and cold detection thresholds, and the difference between the 2 groups was examined.
Results: Compared with controls, patients with clinical neuropathy were older (61.0 vs 55.4 years; P = .028), had higher rates of diabetes (46.2% vs 16.0%; P = .03), and were taking antiviral agents against hepatitis B (100% vs 62%; P = .006). Patients with neuropathic symptoms had significantly increased frequencies of impairment of warm and cold detection thresholds. In addition, the greater severity of symptoms showed higher detection thresholds of warm (control, 40.7?; mild-to-moderate, 43.8 ?; severe, 46.0 ?; P = .007) and cold (control, 28.8? ; mild-to-moderate, 27.0 ?; severe, 21.8 ? ; P = .008).
Conclusions: Our findings show that older age, diabetes, and treatment with oral antiviral agents against hepatitis B virus were more likely to be associated with the occurrence of clinical neuropathy after liver transplant. Early awareness and careful monitoring are warranted.
Key words : Neurologic complications, Peripheral neuropathy, Quantitative sensory testing
Liver transplant (LT) is the treatment of choice to cure end-stage liver disease. With improvements in medical knowledge and surgical skills, but also because of a shortage of deceased organs, living donor LT (LDLT) has become a vital procedure for many LT programs. In countries where deceased organs are scarce, LDLT has been adopted more widely, particularly in Asia, where some religious and cultural beliefs may foster disfavor toward deceased donation.
Despite improvements in survival rates, postoperative complications may occur after LT, with negative effects on recipient mortality or morbidity. Neurologic complications are more common in LT than in other solid-organ transplants; previous studies have reported neurologic complication rates of 20% to 60%.1-7 There are various possible reasons for neurologic complications in LT recipients, including pretransplant disease status, aspects of surgery, infection, and immunosuppressive drugs.6-10
Peripheral neuropathy is a neurologic complication that may cause functional limitations for patients and thus negatively affect the posttransplant quality of life. Earlier studies have reported that peripheral neuropathy occurs in 17% to 32% of LT recipients.6,9 The most common form of peripheral neuropathy is characterized by symmetric, length-dependent, and sensory-predominant symptoms. These symptoms primarily present as the sensation of numbness, tingling, or burning and typically begin distally in the longest nerves, starting at the feet and ascending gradually to the legs. Hand symptoms often become apparent by the time the leg symptoms have progressed to the knees. Upper limb involvement may never occur.11,12
Quantitative sensory testing (QST), a method of measuring sensory thresholds, is a useful tool to assess and, specifically, to quantify neurologic function. Among various types of QST, thermal threshold testing is widely used. Measurement of the threshold for cold sensation reflects the function of small myelinated Aδ fibers. The threshold for warm sensation reflects the function of unmyelinated C fibers. Elevation of thermal thresholds is correlated with intraepidermal nerve fiber depletion.13 Quantitative sensory testing may detect the peripheral neuropathy associated with chemotherapy14 and may be used to monitor disease progression.15,16
Few studies have explored the risk factors for peripheral neuropathy in LDLT recipients. This study was conducted to find the potential clinical factors associated with neuropathy that occur after LDLT and to examine the frequencies of sensory deficits on the basis of QST.
Materials and Methods
Study design and patients
This was a case-control study of 161 patients who underwent LDLT at Changhua Christian Hospital in Taiwan during the period from January 2010 to December 2016 and were followed up at the hospital’s LT outpatient clinic. The inclusion criteria included recipients of a previous LDLT at least 1 year before the screening visit and who were under immunosuppressive treatment with prolonged-release tacrolimus (Advagraf, Astellas). The exclusion criteria were (1) patients with neurologic disorders or with conditions related to neuropathy (chronic renal failure treated by hemodialysis, poorly controlled diabetes mellitus with hemoglobin A1c >7, history of alcoholism, and treatment with chemotherapy or radiation therapy), (2) patients with histories of neuropathic symptoms or neurologic operations before LDLT, (3) patients with biliary complications, (4) patients with asymmetric symptoms or symptoms with greater severity in the upper extremities versus the lower extremities (to eliminate confounding factors of radiculopathy or entrapment neuropathies), and (5) patients with cognitive impairments. Therefore, a total of 63 recipients were included in the analysis (Figure 1). This study complied with the regulations for living liver donors in Taiwan, who must meet the following conditions: (1) at least 20 years old, or 18 years old with a custodian’s consent; (2) the recipient’s spouse, or within the fifth degree of kinship; and (3) voluntary choice to donate, without coercion. The most common relationships of donor to recipient were children (49 donors), followed by spouses (9 donors), siblings (3 donors), niece (1 donor), and daughter-in-law (1 donor). The study protocol was approved by the hospital’s institutional review board.
Identification and scoring of neuropathic symptoms
A neuropathy symptom score was assigned to each participant to identify the patients with clinical neuropathy (case group) and the patients without clinical neuropathy (control group). In addition, these scores indicated the comparative severities of symptoms. Participants were asked to describe any type of leg discomfort. If the patient described burning, numbness, or tingling in the legs, then a score of 2 was assigned; symptoms of fatigue, cramps, or ache were assigned a score of 1. The presence of symptoms in the feet was assigned a score of 2, the calves was a score of 1, and elsewhere was a score of 0. Nocturnal exacerbation of symptoms was assigned a score of 2; if symptoms were present during both day and night, then a score of 1 was assigned; and 0 was the score for symptoms only during daytime. If ever the symptoms had awakened the patient from sleep, then a score of 1 was added. The patients were asked whether any of the following maneuvers reduced their symptoms: walking (score of 2), standing (score of 1), and sitting or lying down (score of 0). Thus, the highest possible neuropathy symptom score was 9, and degrees of symptom severity were defined as normal (score 0-2), mild (score 3 or 4), moderate (score 5 or 6), and severe (score 7-9). The neuropathy symptom score was determined to have 82% sensitivity and 67% specificity, with nerve conduction studies as reference.17 Positive concordance has been reported among Chinese populations.18
Quantitative sensory testing
Quantitative sensory testing was performed for all participants. Warm and cold thermal detection thresholds were measured with a computerized thermal sensory analyzer (model TSA 2001, Medoc Ltd. Advanced Medical Systems). The method of limits was used, ie, the thermal stimulus applied to the skin was increased or decreased until the participant confirmed the sensation. The detection threshold was defined as the minimum perceptible intensity. A higher warm detection threshold indicated that warmth was first perceived at a temperature higher than the typical threshold of normal perception, and a higher cold detection threshold indicated that coldness was first perceived at a temperature lower than the typical threshold. During the QST, a thermode (30 × 30 mm) with a baseline temperature of 32 °C was placed on the skin at the foot dorsum. From the baseline, the temperature of the thermode was increased (for warm stimuli) or decreased (for cold stimuli) at a rate of 1 °C per second. Participants did not view the computer screen during the test and were instructed to respond immediately (eg, via a computer mouse) when warm/cold sensation was perceived. The detected temperature was recorded in the computer, after which the temperature of the thermode returned to baseline at a rate of 1 °C per second. This exercise was repeated 6 times. The first 2 stimuli served as a rehearsal. The detection threshold for each series was determined as the mean temperature of the final 4 consecutive measurements. Normative data of Taiwanese population were similar to data reported for other populations.15,19-21
The following clinical data were obtained from medical records: basic characteristics, the indication for LDLT, the duration of follow-up after LDLT, the post-LT use of oral antiviral agents for prophylaxis against hepatitis B (HBV) recurrence, and the presence of diabetes mellitus (preexisting diabetes or posttransplant diabetes, which is defined according to the International Consensus Guidelines22 as newly diagnosed diabetes in the posttransplant setting from which transient hyperglycemia after transplant is excluded). Immunosuppressive regimens consisted of prolonged-release tacrolimus with or without either mycophenolate mofetil or everolimus, and all patients were steroid-free when they were enrolled in the study. The collected laboratory data included dose-adjusted blood trough tacrolimus levels, hemoglobin A1c, renal function index (serum creatinine and estimated glomerular filtration rate), liver function index (aspartate aminotransferase and alanine aminotransferase), and blood concentrations of calcium and magnesium.
We used the Fisher exact test for comparisons of categorical variables and the Mann-Whitney U test for continuous variables to assess the differences between cases and controls. A 2-tailed P < .05 was considered statistically significant. To compare values of detection threshold between controls and cases of different severity of symptoms, we used the Kruskal-Wallis test followed by the Dunn post hoc test. All statistical analyses were performed with the SPSS statistical software for Windows (version 16.0).
Comparison of patient characteristics
The characteristics of the participants are shown in Table 1. There were no significant differences between the case and control groups with respect to sex, body mass index, and duration of follow-up after LDLT. Significant differences were found with age, diabetes, and anti-HBV prophylactic treatment. Patients with clinical neuropathy were significantly older (61.0 vs 55.4 years; P = .028), had a higher rate of diabetes (46.2% vs 16.0%; P = .03), and had a higher rate of anti-HBV prophylactic treatment (100% vs 62%; P = .006) compared with the control group (patients without clinical neuropathy). Furthermore, patients with neuropathy symptoms were found to have a significantly higher percentage of preexisting diabetes (38.5% vs 8%; P = .016) compared with the control group. With regard to history of post-LT use of anti-HBV prophylactic agents, telbivudine and entecavir-to-telbivudine (or telbivudine-to-entecavir) treatments were more common in the case group; however, entecavir treatment was more common in the control group (P = .012).
Comparison of quantitative sensory testing results
Quantitative sensory testing data are presented in Table 2. The detection threshold data showed that patients with clinical neuropathy (both mild to moderate and severe) had significantly higher warm detection thresholds than control subjects (control, 40.7 ?; mild to moderate, 43.8 ?; severe, 46.0 ?; Kruskal-Wallis, P = .007; Dunn post hoc, both mild to moderate vs control and severe vs control were P < .05). The results were similar for cold detection thresholds. Patients with severe neuropathy, but not those with mild to moderate symptoms, had significantly higher cold detection thresholds, ie, they required a lower temperature to perceive cold (control, 28.8 ?; mild to moderate, 27.0 ?; severe, 21.8 ?; Kruskal-Wallis, P = .008; Dunn post hoc, severe vs control, P < .05). We also found that patients with clinical neuropathy had significantly higher rates of abnormality for warm detection threshold (P = .042) and cold detection threshold (P = .009) than those in the control group.
Neurologic complications are relatively common after LT. One retrospective study reported that sensory peripheral neuropathy was the most common neurologic complication, with a rate of 32% in a total of 93 LT recipients who had neurologic complications.9 In the present study, we found that older age, diabetes, and anti-HBV agents were associated with clinical neuropathy in LDLT patients. In addition, QST results revealed that patients with clinical neuropathy had greater frequency of impaired thermal sensations than those without clinical neuropathy. Older age and diabetes are well-known risk factors for developing peripheral neuropathy.23 Interestingly, our analyses found that the prevalence of preexisting diabetes was significantly higher in patients with clinical neuropathy than in patients without clinical neuropathy, suggesting that diabetes duration may be considered. This has been supported by a previous study that showed the prevalence of peripheral neuropathy increases as diabetes duration increases.24 In a previous study, multivariate logistic regression analysis showed that diabetes duration was an independent and significant risk factor for peripheral sensory neuropathy (odds ratio, 1.03; 95% CI, 1.01-1.05).25
Patients with HBV-related liver disease account for the majority of LT patients in Asian countries where HBV is the leading cause of chronic hepatitis. For these patients, HBV prophylaxis is essential to prevent HBV reactivation and to improve survival and should be started immediately after surgery. The anti-HBV agents for prophylaxis after LT are oral nucleoside/nucleotide analogues (NA), of which entecavir and tenofovir are preferred because of high potency and low rate of drug resistance.26 Telbivudine is similarly effective in preventing HBV recurrence and also may be associated with improvement of posttransplant renal function.27-30 According to the guidelines from the American Association for the Study of Liver Diseases,26 this prophylactic therapy should be lifelong. Therefore, the safety of NA treatment should be considered.
Peripheral neuropathy is infrequently observed during HBV-related NA treatment. Telbivudine has been reported to be associated with myopathy and neuropathy.31 Interestingly, 2 previous trials showed that telbivudine in LT patients was associated with higher incidence of neuropathy. In de novo HBV patients, about 1.2% of patients treated with telbivudine developed peripheral neuropathy in 4-year follow-up.32 In a 1-year trial, Lee and colleagues29 have reported that 18% of LT recipients treated with telbivudine developed peripheral neuropathy during the trial and post hoc follow-up. Turan and colleagues30 have shown that up to 40% of LT recipients with telbivudine developed polyneuropathy and the 1-year estimated incidence of polyneuropathy and/or myopathy was 28%. Among those patients with peripheral neuropathy, numbness was the typical symptom. It is worth emphasizing that some patients’ neuropathy problems would be resolved after telbivudine was discontinued; however, a small number of patients had persistent symptoms, as shown in the studies by Lee and colleagues29 and Turan and colleagues.30 The association between peripheral neuropathy and entecavir has been infrequently reported for patients with HBV.33,34 In our analysis, the case group had a higher proportion of patients receiving telbivudine and/or entecavir during the post-LT follow-up compared with the control group. However, caution is warranted because a limitation of this study is that we did not know the time sequence of onset of peripheral neuropathy and the prescription of a specific anti-HBV agent. Thus, the causality was not definite. The link between NA treatment and peripheral neuropathy may be associated with the manifestation of mitochondrial toxicity caused by the NA-induced inhibition of human DNA polymerase γ.35
Immunosuppressants also have risks of neurotoxicity that may predispose LT patients to peripheral neuropathy.36 In previous literature, treatment with calcineurin inhibitors such as tacrolimus and cyclosporine has been reported to be an independent risk factor for neurologic complications (odds ratio, 1.7; 95% CI, 1.0-2.8).9 We found no difference in trough level of tacrolimus in our study, in agreement with previous literature.9 In addition, hepatitis C virus-related mixed cryoglobulinemia, a common extrahepatic manifestation, is also associated with development of sensory polyneuropathy. The etiology of this hepatitis C virus-related mixed cryoglobulinemia is associated with deposition of circulating immune complexes into small blood vessels, resulting in purpura, glomerulonephritis, and neuropathy.37
It is probable that peripheral neuropathy is a multifactorial disease, not caused by any single factor. Moreover, factors should be considered as component causes, not as sufficient causes. For example, not all patients with diabetes will develop peripheral neuropathy. Therefore, multiple component causes probably contribute to the development of peripheral neuropathy.23
In our present study QST data showed that patients with clinical neuropathy had significantly higher thresholds and greater frequencies of impairment for warm and cold detection thresholds, implicating the involvement of small fiber neuropathy. Quantitative sensory testing is helpful for clinical diagnosis of peripheral neuropathy in a relatively large subset of patients for whom conventional nerve conduction testing, which focuses primarily on large fiber function, is nondiagnostic.38 Two types of small fibers are known to conduct thermal sensation, ie, unmyelinated C fibers (warm) and lightly myelinated Aδ fibers (cold). Unmyelinated fibers may be affected first during the progress of peripheral neuropathy, so the warm detection threshold is considered specific for evaluation of a small fiber neuropathy.39 The present study also demonstrated consistency between the degree of thermal thresholds and the severity of neuropathy. Thresholds were higher when the severity was greater.
A nerve conduction velocity test was performed on 9 patients in the case group, as indicated by their medical needs. Seven patients had been diagnosed with sensory polyneuropathy (5 without pick-up of sensory potentials in bilateral sural nerves and 2 with decreased amplitude of sensory potentials in bilateral sural nerves, ie, 77.8%). These patients had higher warm and cold detection thresholds (43.3 ? and 26.8 ?) compared with the control group despite a lower abnormality rate of QST outcome (4 of 9 patients, 44.4%). The explanation for this outcome may be that QST is a subjective test and is dependent on patient motivation, alertness, and concentration. Nerve conduction velocity testing and QST are not equally interchangeable, but rather each is complementary to the other. There are some restrictions to QST. For example, abnormalities in the central nervous system may cause symptoms similar to those for peripheral neuropathy. Also, cognitive impairment in participants may result in poor reliability of the QST data.39 However, such patients were excluded in this study.
The main limitation of our study is the small sample size, which may lead to bias. Moreover, the data were from a single medical center, which limits the applicability of the results. Therefore, prospective cohort trials with greater numbers of participants are needed to clarify the risk of peripheral neuropathy in LT recipients.
Older age, diabetes, and oral anti-HBV agents were found to be associated with clinical neuropathy among the LDLT recipients in the present study. Physicians who treat these patients should be aware of the risk for peripheral neuropathy because early recognition and management of neuropathic symptoms may reduce a patient’s risk for morbidity.
Volume : 19
Issue : 7
Pages : 664 - 670
DOI : 10.6002/ect.2020.0392
From the 1Department of Pharmacy, Changhua Christian Hospital; the 2Department of Nursing, HungKung University, Taichung; the 3Department of Neurology, the 4Department of General Surgery, and the 5Transplant Medicine and Surgery Research Centre, Changhua Christian Hospital, Changhua, Taiwan
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Corresponding author: Yao-Li Chen, Department of General Surgery, Changhua Christian Hospital, No. 135, Nanxiao St. Changhua City, Changhua, Taiwan
Phone: +886 04 7238595
Figure 1. Flowchart of Study Enrollment
Table 1. Comparison of Patient Characteristics Between the Case and Control Groups
Table 2. Comparison of Thermal Quantitative Sensory Testing Outcomes Between the Case and Control Groups