All adult patients, who underwent LT, who were on IR-Tac-based IS regimen for at least 6 months, with stable liver function test (LFT) and serum creatinine levels <2.0 mg/dl were enrolled in this study.

Patients were excluded in case of pregnancy, breastfeeding, malignancy, severe systemic infection requiring any therapy that could modify Tac pharmacokinetics, or the use of any other investigational drugs.

In general, the IS therapeutic protocol of our centre requires that corticosteroids are not used unless the patient has autoimmune pathologies. In patients with stable liver function, Tac monotherapy is usually achieved 1 year after the transplantation (22).

The conversion from IR-Tac to PR-Tac started as soon as the new formulation was available in our hospital. All patients enrolled in the study were switched to PR-Tac, individually, during a 2 months period, and they were followed-up for at least 10 years. The starting dose of PR-Tac was exactly the same as the dose of IR-Tac taken by the patient at the time of conversion (1: 1).

Tac levels were measured in our central laboratory using a high-performance liquid chromatography-mass-spectrometry procedure (23). Patients were closely monitored during the study and Tac doses were adjusted to maintain adequate blood levels to maintain normal liver function and preventing rejection.

At follow-up, physical examination and measurement of vital signs were performed, contingent adverse events were noted, and laboratory test results were checked. Arterial hypertension was defined as systolic blood pressure >140 and/or diastolic >90 mm Hg at two subsequent visits or when antihypertensive treatment was prescribed. Diabetes mellitus was defined as fasting glucose >126 mg/dl at two subsequent visits or when hypoglycemic treatment was used. Dyslipidemia was defined as cholesterolemia >220 mg/dl and/or triglyceridemia >200 mg/dl at two subsequent visits or when using hypolipidemic treatment.

LFTs including aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyl-transferase and bilirubin were performed at each clinic. Elevated transaminases, defined as twice the upper limit of our laboratory cut-off (AST >68U/L; ALT >110U/L), triggered closer surveillance and a liver biopsy when LFT abnormalities persisted (24). Graft loss was defined as retransplantation or death. Renal function was assessed using the glomerular filtration rate (eGFR, Modification of Diet in Renal Disease formula, MDRD) formula.

Adherence was assessed using the ‘‘Basel Assessment of Adherence Scale to Immunosuppressives” (25) (BAASIS) questionnaire. This tool consists of a four-item validated questionnaire and a Visual Analog Scale (VAS). The first part addresses adherence including timing, missed dose and a “drug holiday” defined as >24 h interval between two consecutive doses. The VAS is a 100-point score, where patients report adherence in the previous 4 weeks from 0 to 100 (drug therapy never/always taken as prescribed) thereby assessing adherence as a continuous variable. The BAASIS form was completed by patients once pre-conversion to assess adherence to IR-Tac formulation and again at one- and ten-years following conversion to assess adherence to PR-Tac formulation and tolerability over time.

A de novo questionnaire (unpublished data) was developed to address those aspects of the IS regimen that influence patient care and the general perception of good or poor QoL in patients taking PR-Tac versus IR-Tac. The questionnaire was designed to be short, simple and easy to understand, to ensure a high completion rate with minimal missing data. The questionnaire, filled in anonymously by patients, collected demographic information (age, gender, and marital and employment status), and included three additional questions. Demographics were collected to facilitate the interpretation of the data at the end of the study. The first question was completed in the pre-conversion phase and queried the possible difficulty of taking multiple daily doses of drugs using a binary response option (YES/NO). The response was followed by a four-point Likert scale measuring the degree of difficulty, with two positive (very, quite) and two negative (little, very little) quantitative responses. At 12 and 120 months, questions two and three were administered. The second question assessed the possible satisfaction of the new drug regimen using a binary response option (YES/NO). The third question evaluated the perception of an improvement in QoL following the intake of the single-dose drug with a four-point Likert scale measuring the degree of improvement with two positive (very, quite) and two negative (little, very little) quantitative responses.

A positive response to the first question assumed dissatisfaction with taking multiple daily drugs, which was confirmed by positive responses on the Likert scale. Positive responses to questions 2 and 3 indicated satisfaction with the new therapeutic regimen and increased perception of QoL.

Data were presented as means (standard deviation), medians (interquartile range; IQR), or frequencies (percentage) as appropriate. For adherence data, categorical variables collected during follow-up were compared to baseline values using Fisher’s exact-test, while continuous data were compared to baseline values using the paired Student’s t-test. To simultaneously subject-wise as well as time-related changes, and possible interactions between them, all other variables were analyzed using multivariate linear mixed models modelling timepoints as a repeated within-subject factor and employing an unstructured estimate of the covariance matrix. As opposed to general linear models, mixed models have the advantage of being able to account for heterogeneous distances between timepoints, missing data as well as unequal variances and covariances. To account for possible confounding due to inter-patient variability, all models included gender, categorized disease etiology, time between therapy inception and conversion, and Tac blood levels (primary endpoint only) as covariates of interest.

Whenever a statistically significant (p < 0.05) overall effect of time was found, pairwise comparisons between timepoints were performed and corrected for multiple comparisons across pairs of timepoints using the Dunn–Šidák procedure.

Written informed consent was obtained from each patient prior to enrolment, without any patient refusing to participate in the study.

This study was conducted in accordance with the Declaration of Helsinki and approved by an Independent Ethics Committee prior to implementation.

At enrolment, 74 patients (69%) were on IR-Tac monotherapy. Median Tac daily dose was 2.0 (IQR, 1.5–3.0) mg and similar values were reported within the first 12-month after conversion. Eight-six out of 107 (81.9%) patients continued with the same Tac dosage after 1 year and the dosage was decreased in six patients (5.7%), increased in 11 patients (10.5%) and two (1.9%) patients were withdrawn from PR-Tac: the first patient due to frequent episodes of hypertension, diarrhea and vertigo and reconverted to IR-Tac; the second one due to de novo intestinal adenocarcinoma and subsequently switched to mTOR-inhibitor monotherapy.

By the end of the follow-up, 91 (85%) patients were still on the PR-Tac IS regimen: 56 patients (52.3%) were maintained on the same dosage as baseline; 24 patients (22.4%) had their dosage decreased and 11 patients (10.3%) increased. Six (5.6%) were converted to a different IS drug between 12th and 120th month. Three patients (2.8%) initiated single drug treatment with mycophenolate mofetil due to blood hypertension at 28, 56 and 68 months respectively. Three patients (2.8%) were converted to mTOR-inhibitor due to HCC recurrence, breast cancer and colon adenocarcinoma at 58, 89 and 112 months, respectively. A total of eight (7.4%) patients died due to a cardiovascular accident (n = 4) or malignancy (n = 4; lung cancer (n = 2) and esophagus cancer (n = 2)) (Table 2).

The median serum Tac blood levels were 5.80 (IQR, 4.1–7.1) ng/ml and 3.80 (IQR, 3.1–4.5) ng/ml respectively (p < 0.0001) at baseline and 1 month after conversion, respectively. When the LFTs remained stable no dose adjustments were considered. In patients without dose adjustments, the median Tac blood levels remained 2.0 mg [IQR, 1.5–3.0] We found a significant effect of the timepoint factor (p < 0.0001) on Tac blood levels, which in post-hoc comparison appeared to be associated with the following differences: baseline vs. 1, 6, 60 and 120 months (1 m < baseline: p < 0.0001; 6 m < baseline: p = 0.023; 60 m < baseline: p = 0.002, 120 m < baseline: p = 0.001); 1-month vs. 6, 60 and 120 months (1 m < 6 m: p = 0.023; 1 m < 60 m: p = 0.003; 1 m > 120 m: p = 0.001); 6 months vs. 12, 60 and 120 months (6 m < 12 m: p = 0.01, 6 m > 60: p = 0.001, 6 m > 120 m: p = 0.001) (Figure 1).

No patient experienced clinical or biopsy-proven acute rejection (BPAR) after conversion. The 10-year survival was 92.6%.

There was no statistically significant effect for the timepoint factor on liver function. Even the comparison of glucose levels and cholesterol and triglycerides values between the pre and post conversion periods was not statistically significant. We found a significant effect of the timepoint factor (p < 0.0001) on eGFR (Figure 2), which in post-hoc comparison appeared to be associated with the following differences: baseline<120 months (p < 0.0001); baseline>1 month (p < 0.0001); baseline>6 months (p = 0.049); month 1<6, 12, 60 and 120 (all p < 0.001).

The BAASIS questionnaire addressed different aspects of adherence, with the aim of identifying those areas where adherence has significantly increased. At baseline, 84 (78.5%) patients reported forgetting to take at least one drug dose in the previous 4 weeks, whereas this dropped to 24 (22.4%) and 27 (25.2%) patients at 12 and 120 months, respectively (p < 0.0001). Sixty-six patients (61.7%) declared at baseline that they had possibly missed two consecutive drug doses, which dropped to 12 (11.4%; p < 0.0001) and 22 (20.5%; p < 0.0001) patients at 12 and 120 months, respectively. Sixty-three patients (58.9%) did not respect the therapeutic intake time at baseline but this decreased to 15 (14%; p < 0.0001) and 25 (23.3%; p < 0.0001) patients at 12 and 120 months, respectively. Five patients (4.7%) admitted to taking lower drug dosages than medically prescribed, which dropped to three (2.8%; p = 0.72) and no (0%, p = 0.06) patients at 12 and 120 months, respectively.

Median VAS ratings of patient adherence were 90 (IQR,75–100) at baseline and were significantly higher at 12 months (97 [IQR, 85–100]; p = 0.0009) and 120 months (95 [IQR, 87–100]; p = 0.0008) (Table 3).

The three separate questions regarding QoL were completed by all participants at baseline. At baseline 66 (61.7%) patients indicated that they experienced difficulty with taking multiple doses of immunosuppressants daily, which was similar after 12 months (n = 69; 65.7%) and 120 months (n = 61; 67%). Filter questions showed that 40 (60.6%), 49 (71.0%) and 45 (73.8%) of patients found it very difficult to take more than one type of drug at baseline, 12 months and 120 months, respectively. Twenty-six (39.4%), 20 (72.4%) and 48 (78.7%) patients found it very difficult to take one or more doses of the same drug at baseline, 12 months and 120 months, respectively. Thirty (45.4%), 67 (97.1%) and 60 (98.4%) patients found it very difficult to take drugs at different times at baseline, 12 months and 120 months, respectively.

The second item addressed the degree of satisfaction with the PR-Tac therapeutic regimen. The data showed that 98 (93.3%) and 79 (86.8%) patients were very satisfied at 12 and 120 months, respectively. The third item asked if there had been an improvement in QoL after conversion. Eighty (76.2%) and 75 (82.4%) patients confirmed that their QoL had improved at 12 and 120 months respectively, with the filter question showing that 82.5% and 86.8% felt that QoL had very much improved at 12 and 120 months, respectively (Table 4).

The median C/D ratio at baseline was 2.67 (IQR, 2.7–4.0). After one and 12 months, the ratio decreased to 1.87 (IQR, 1.9–2.6) and 2.03 (IQR, 2.0–2.6), respectively and remained stable during the follow up (2.03 [IQR, 2.0–2.9]: p < 0.000001) (Figure 3).

We observed a significant effect of the timepoint factor (p < 0.0001) associated with the following differences: 12 months vs. baseline, 1, 6, 60, 120 months (12 m < baseline: p = 0.0001, 12 m > 1 m: p = 0.001, 12 m > 6 m: p = 0.007, 12 m < 60 m: p = 0.001, 12 m > 120 m: p = 0.053), 6 months vs. baseline, 1, 12, 60, 120 months (6 m < baseline, 6 m > 1 m, 6 m < 12 m, 6 m < 60 m, 6 m < 120 m; all p = 0.0001), 1 month vs. baseline, 6, 12, 60, 120 months (1 m < baseline, 1 m < 6 m, 1 m < 12 m, 1 m < 60 m, 1 m < 120 m; all p = 0.0001), 120 months vs. baseline, 1, 6, 12, 60 months (120 m < baseline, 120 m > 1 m, 120 m > 6 m, 120 m < 12 m, 120 m < 60 m; all p = 0.0001). In addition, we compared 10 consecutive pre-conversion timepoints to 10 consecutive post-conversion measurements to evaluate the change of Tac blood levels and dose in a long-term observation: the mean C/D ratio was significantly higher pre-conversion compared to post conversion (3.29 [IQR, 2.7–4] versus 2.58[IQR, 2.3–2.9]: p = 0.008), while the coefficient of variation of the C/D ratio was significantly lower pre-conversion compared to post-conversion (2.12 versus 1.19: p = 0.003).