Visual evoked potentials as a method for the prospective assessment of tacrolimus neurotoxicity in patients after kidney transplantation

Patient recruitment and laboratory measurements were carried out during a routine visit in the outpatient transplant clinic at the Department of Nephrology, Transplantation and Internal Medicine of the Medical University of Silesia in Katowice. Patients who had recently undergone transplantation and had stable renal excretory function qualified for the study, which was done in accordance with the Declaration of Helsinki. The Bioethics Committee of the Medical University of Silesia granted permission for the study protocol (KNW/0022/KB1/93/13).

The inclusion criteria were: age between 20 and 60 years, written informed consent, visual acuity ≥ 0.5 (decimal) on the Snellen chart, and treatment with one pharmacologic form of tacrolimus from the transplantation procedure to the time of the present study. The exclusion criteria included: absence or withdrawal of consent at any stage of the study, photosensitive epilepsy or other central nervous system disease (stroke condition, previous neurosurgical intervention, or major head injury), pre- or posttransplant diabetes, history of optic nerve (n. II) and/or retina diseases, cataract worsening visual acuity, nystagmus, photophobia or other unpleasant sensations caused by flashes of light, and treatment with other potentially neurotoxic drugs.

Both VEP examinations were done in 35 out of the 43 patients recruited into the study. Table 1 shows the clinical characteristics of study group during the baseline and follow-up examination. All study patients were treated with orally given tacrolimus during the whole period after kidney transplantation. At transplantation, all patients received the twice-daily tacrolimus formulation (Prograf). During the observation period of the study, 15 patients were converted from the twice-daily to the once-daily tacrolimus formulation (Advagraf) within a few months after the baseline examination. The main study outcomes were further analyzed separately for the different tacrolimus formulations.

The rest of the immunosuppressive regimen consisted of mycophenolate mofetil (orally, 500 mg BID) or sodium (orally, 360 mg BID) and steroids (prednisone, 7.5 mg during the baseline and 5 mg during the follow-up examination).

During the routine laboratory test, additional blood samples were withdrawn for the determination of several biochemical parameters and for the calculation of tacrolimus exposure by using the area under the curve (AUC). The AUC for tacrolimus was calculated with the formula:

AUC = 13.17 + (5.43 × “Tc0”) + (4.72 × “Tc3”).

where Tc0 is the tacrolimus blood level before the morning dose (trough level) and Tc3 is the tacrolimus blood level measured 3 h after the morning dose ingestion [13]. In the follow-up examination, only the tacrolimus trough level was measured.

Kidney graft function was measured by using an estimated glomerular filtration rate (eGFR) calculated according to the Modification of Diet in Renal Disease (MDRD) formula.

The historic control group, which consisted of patients who met the medical criteria of the present study but had no history of chronic kidney disease, was also included. In all control subjects, as the prospective study group, VEP measurements were done twice in the same laboratory. The control group included 15 subjects (27 eyes) with normal kidney excretory function, for whom the results of 2 consecutive VEP measurements were available. The median age was 52 (interquartile range, IQR, 46–66) years and the time between VEP examinations was 24 (19–31) months.

The laboratory measurements included the assessment of plasma concentrations of high-sensitivity C-reactive protein (hsCRP), IL-6, and brain derived neurotrophic factor (BDNF), as well as asymmetric dimethylarginine (ADMA) and endothelin-1 (ET-1). Table 2 provides detailed information on commercially available ELISA kits. All parameters were assessed twice, concomitantly with the baseline and follow-up VEP examinations.

General ophthalmologic examinations were done at the nearby Professor K. Gibiński University Clinical Centre of the Medical University of Silesia in Katowice. The best-corrected vision acuity was measured by using the Snellen chart. Patients with a defect above 3 dioptres were excluded. The first study examination (I) was carried out after the patient qualification, and the second examination (II) was done after a mean period of 25 (range, 21–28) months. Both examinations were carried out according to the International Society for Clinical Electrophysiology of Vision (ISCEV) standards, with minor variations in luminance and recording filters. Table 3 shows the detailed technical data on the VEP measurements; all study variables were measures twice. Visual evoked potentials (VEP) were tested by using the Reti-Port electrophysiologic device from Roland Consult (Germany), in accordance with the standards of the ISCEV [14]. The patient’s scalp was glued with gold cup electrodes by using abrasive paste (Nuprep Skin Prep Gel, Weaver and Company) and conductive paste (Ten20 Conductive Paste, Weaver and Company), according to the international system 10/20: the active electrode located at the Oz point (above the occipital area) in relation to the electrode at the Fz point (above the frontal area). The ground electrode was glued at point Cz (above the parietal area). A black-and-white checkerboard pattern with alternating phase change (pattern VEP; PVEP) was used to elicit a visual cortical response. Each eye of the patient was stimulated with a small (0.25 degree) and a large (1 degree) angular standard. For flash stimulation, (flash VEP; FVEP), standard flashes with a frequency of 1.4 Hz were used in the Ganzfeld stimulator. The amplitude and latency of the P100 PVEP wave (Fig. 1a) and P2 FVEP wave (Fig. 1b) were analysed. The amplitude was measured from the top of the negative peak preceding the positive (P) wave (N75-P100 and N2-P2), and the latency from the beginning of the stimulation to the occurrence of the highest value of the marked wave.

Each VEP test was preceded by checking of the visual acuity of the distance in the patient’s correction with glasses using Snellen charts and an intraocular pressure (IOP) test using a Goldmann applanation tonometer (CSO, Italy), and then finished with anterior and posterior examination of the eyeball in a slit lamp (Haag-Streit) by an indirect method using a Super Field lens (Volk, US) to monitor possible pathologic changes in the eye that may affect the result of the VEP test or the result of neurotoxicity (primarily changes in the optic disc, such as pallor and blurring or raising boundaries, etc.). Before the fundus examination, short-acting (4–6 h) mydriatics were given: parasympatholytic (1% Tropicamidum; Polfa) or sympathomimetic (10% Neosynephrin-POS; URSAPHARM) drops. If the patient did not consent to mydriasis, the fundus was checked through the narrow pupil. Refusal of consent for pupil dilation did not affect the patient's further participation in the study and did not change the patient’s treatment.

The results of VEP measurements were scored according to the reference values (ISCEV) and expressed as percentages of the study subjects, data for the right and left eyes were combined and then individually analyzed. The primary outcomes of the present study were the relative changes in VEP parameters from the baseline values. Additionally, explanatory analyses were carried out on the variability of VEP measurements, including age, eGFR, tacrolimus level, and laboratory parameters.

Data were collected on a Microsoft Excel 2007 spreadsheet and then analyzed by using the STATISTICA 13.3 software package for Windows (TIBCO Software Inc., Palo Alto, CA, USA) and MedCalc version 19.2.1 (MedCalc Software, Mariakerke, Belgium). The distribution normality was analyzed for all variables with the use of the Shapiro–Wilk test. All study data (except the mean visual acuity) were presented as median values with an interquartile range. Statistical calculations were done by using the Mann–Whitney U test and the Wilcoxon signed-rank test for variables with a non-normal distribution. The Spearman’s correlation coefficient R was used to assess the level of relationship between different variables. At the follow-up examination, further analyses were done separately for the two different tacrolimus formulations: twice-daily and once-daily tablets. Stepwise multiple regression analyses were performed for the P100 wave (1°) and (15^’) latencies and the corresponding amplitudes as dependent variables, including age, eGFR, and log tacrolimus trough level as potential independent variables. Analogic regression analyses were done for the P2 wave latency and amplitude. The second model was used for the P100 (15’) latency and amplitude and P2 amplitude as dependent variables, including age, eGFR, log tacrolimus trough level, and log BDNF plasma level. The level of statistical significance in all analyses was set at p < 0.05.

The mean visual acuities on a decimal scale were 0.974 (95%CI, 0.952–0.996) and 0.920 (0.920–0.872) for the right and left eye, respectively, in the first examination, and 0.969 (0.947–0.991) and 0.912 (0.864–0.959) in the follow-up examination. No statistical differences were observed in the comparative test for visual acuity in both study examinations. In individual cases (5.8%), visual acuity decreased over a two-year observation period by a maximum of one line on the Snellen chart.

There were no statistically significant differences between the IOP measurements in the first and in the follow-up examinations (Table 1). In the assessment of the anterior eye segment structures, pathologic changes were observed in only one patient in the entire study cohort. During the follow-up examination, posterior subcapsular cataract was detected in both eyes, with long-term chronic kidney disease and pretransplant steroid therapy considered to be the most likely cause. The image of the eye fundus for the whole group did not differ from the norm. During the two-year follow-up, the patients did not complain of any eye symptoms.

As expected, the tacrolimus daily dose and trough levels were significantly lower during the follow-up study examination than in the early post-transplant period (Table 1). Notably, there were no significant differences in tacrolimus dosing and trough levels between the patients treated with the two different tacrolimus formulations (data not shown).

At baseline, the median differences in PVEP P100 latencies between a given patient’s eyes were: 2.3 (IQR, 0.6–5.3) ms (1° checks), 3.0 (1.0–5.2) ms (15’ checks); for amplitudes, median interocular differences were 1.7 (1.0–2.6) μV (1°), and 1.8 (0.9–3.0) μV (15’), with maximum intrapatient differences of 15.3 and 17.6 ms, and of 9.8 and 10.7 μV, respectively. In FVEP, the median baseline differences were 5.2 (IQR, 1.9–7.6) ms and 1.8 (1.3–3.2) μV, respectively.

In the PVEP examinations, in the study group, the follow-up latency (1°) was significantly longer than in the baseline post-transplant examination (Table 4, Fig. 2a). Additionally, a significant decrease of amplitude (1°) was observed during the same period (Fig. 2b). In contrast, there were no significant changes for the 15’ checks in P100 latency or amplitude values during the study period (Table 4).

In the control group, no significant differences were observed between the baseline and follow-up PVEP measurements [median P100 latency (1°): 107.2 (IQR, 104.5–111.6) vs. 105.1 (102.2–107.4) ms; p = 0.16; median latency (15’): 109.8 (105.1–122.1) vs. 117.8 (112.7–123.0) ms; p = 0.40; median amplitude (15’): 8.1 (4.6–11.3) vs. 9.3 (5.3–17.2) μV; p = 0.20], except for amplitude (1°), for which the median value increased over time [7.5 (5.5–10.6) vs. 8.5 (5.9–9.4) μV; p < 0.05]. Thus, the latency (1°) and amplitude (1°) percentage changes in the control group were −1.9 and 13.5%, respectively. Similar longitudinal values of latency and amplitude were also found in FVEP examinations, both in the study and control groups. There were also no significant differences in the analyzed VEP parameters between the study and control groups, both at baseline and in the follow-up period.

At the first examination, P100 latencies were positively correlated between the check sizes (1° versus 15’) (R = 0.294; p = 0.015) but P100 was not correllated with P2 latency (R = 0.218; p = 0.072). P100 latency for the 1° checks was negatively correlated with (1°) and (15’) amplitudes (R = −0.487; p < 0.001 and R = −0.271; p = 0.025, respectively) and with P2 amplitude (R = −0.469; p < 0.001). There was a weak correlation between tacrolimus trough level and P2 wave latency (R = 0.241; p < 0.05). No other correlations were observed between either tacrolimus trough level or tacrolimus AUC and the other VEP parameters measured shortly after kidney transplantation. Notably, almost all VEP parameters (except P100 amplitude (15’) and P2 latency) showed a baseline association with recipient age, and all examined latencies, but not amplitudes, were much more pronounced in the follow-up examination (Table 5). However, in the stepwise multiple regression analysis, only eGFR (r_partial = 0.248; p < 0.05), but not age or log tacrolimus trough level, independently influenced P2 latency.

In the follow-up examination, there were only weak associations of P100 latencies between the check sizes (R = 0.276; p = 0.021) and between P100 for the 15' checks and P2 latency (R = 0.346; p < 0.01), whereas the correlations of (1°) amplitude with (15’) amplitude (R = 0.710); p < 0.001) and P2 wave amplitude (R = 0.512; p < 0.001) were much more pronounced. Notably, P2 wave amplitude, but not latency, correlated with tacrolimus trough level (R = −0.450; p < 0.001 and R = 0.183, p = 0.13, respectively) (Fig. 3a). Interestingly, when separate analyses were done for the two tacrolimus formulations, both the above associations were observed only in the twice-daily formulation subgroup (for latency R = 0.393; p = 0.013, and for amplitude R = −0.375; p = 0.019), whereas in the once-daily formulation subgroup, only amplitude (R = −0.480; p = 0.007), but not latency (p = 0.43), correlated with tacrolimus level. In the whole study group, the stepwise multiple regression analyses showed that P2 wave latency was independently influenced only by log tacrolimus trough level (r_partial = 0.472; p < 0.001), whereas P2 wave amplitude was affected by log tacrolimus trough level (r_partial = −0.420; p < 0.001) and age (r_partial = −0.404; p < 0.001).

In the analysing of inflammatory markers, IL-6 values significantly correlated only with P100 (15’) latency in the first (R = 0.300; p = 0.02), but not in the follow-up, examination.

The significant baseline association between BDNF plasma level and patient age (R = 0.457; p < 0.001) was not found in the follow-up examination (p = 0.98). Early after transplantation, there was a positive correlation of BDNF with P100 (15’) latency (R = 0.305; p = 0.017) (Fig. 3b) and a negative correlation with P2 wave amplitude (R = −0.304; p = 0.017) (Fig. 3c). In the whole study group, there was no significant association between BDNF and any VEP parameter in the follow-up measurement. However, there were significant positive correlations of BDNF with P100 (15’) latency (R = 0.499; p = 0.005) and P2 latency (R = 0.409; p = 0.025) only in patients treated with the once-daily, but not the twice-daily, tacrolimus formulation. Stepwise multiple regression analysis showed that log BDNF (r_partial =−0.383; p = 0.03) and eGFR (r_partial =−0.371; p = 0.004) were independently associated with P100 (15’) amplitude. Age and log tacrolimus trough level were not included in the final model.

In the follow-up examination, P100 (1°) latency and (1°) amplitude were associated with ET-1 levels (R = 0.299; p = 0.013 and R = −0.263; p = 0.03, respectively), but not with tacrolimus blood trough level.

In both examinations, there were no significant associations of VEP parameters with ADMA.