Tremor Induced by Cyclosporine, Tacrolimus, Sirolimus, or Everolimus: A Review of the Literature

We calculated the prevalence rates for tremor induced by each of the drugs as a range and for studies that had a sample size greater than 100 we calculated mean ± standard deviation.

We found 33 studies reporting patients developing cyclosporine-induced tremor [8‐40] (Table 1). The geographical distribution of data on cyclosporine-induced tremor was as follows: 13 studies from the Americas, 14 from the European Region, two from the Eastern Mediterranean Region, and one from the South-East Asian Region, while one study did not specify the region.

Data were extracted from mostly prospectively conducted studies. Although the prevalence of cyclosporine-induced tremors in all studies taken together appeared to be wide (range 0.1–71.4%), the assessment of prevalence in larger cohorts with n > 100 (Fig. 2A) was determined to be mean ± standard deviation 17.1 ± 6.3% (189/1109 patients; five studies). Most studies did not report the time frame of tremor development (n = 18). A total of 11 studies reported an early-onset tremor, while one study reported late-onset chronic development of tremors. Three studies reported patients developing both early-onset and late-onset tremors. One study with a large number of kidney transplant patients (n = 402) found 20% of participants developed tremor between 1 and 3 months, 12% between 4 and 6 months, and 10% between 7 and 12 months [17].

Most studies reported the risk of tremor in the context of kidney (n = 9) and liver (n = 7) transplantation, with one study including patients with both kidney and liver transplantation (n = 1). Tremor was also reported in the context of inflammatory bowel disease (n = 5), autoimmune/connective tissue diseases (n = 4), bone marrow transplantation (n = 4), autoimmune disorder of the eye (n = 2), and chronic disease of the spinal cord (n = 1). The risk of cyclosporine-induced tremor was more likely in patients with kidney (28 ± 16.1%) and bone marrow transplants (40 ± 14.7%) compared with liver transplants (9 ± 7.7%). When combining data for nontransplant indications, a prevalence of 21.5 ± 9.7% was noted (Fig. 2B).

Most studies reported their target or mean for monitoring the serum and plasma concentrations of cyclosporine (Table 1). However, data on longitudinal monitoring of drug concentrations and, more importantly, the correlation of blood concentrations with tremor intensity were lacking. One observational study with pediatric renal transplant recipients found that the area under the curve correlated positively with the intensity of tremors [12].

Some studies reported co-occurring neurological side effects such as headaches (n = 14), paresthesia/hypoesthesia/hyperesthesia (n = 14), seizures (n = 8), and insomnia/sleep disorders (n = 5). The more serious neurotoxic adverse effects such as encephalopathy (n = 2) and coma (n = 2) were reported by a small number of studies. Occasionally, studies reported co-occurring ataxia (n = 1) and parkinsonism (n = 1).

Treatments for tremors varied, and ranged from no treatment specified (n = 21), tapering to eventual discontinuation of the drug (n = 6), change to a lower dosage (n = 5), or supplementation with magnesium (n = 2). None of the studies reported data on potential risk factors and patient demographics for developing tremors.

Studies reporting tacrolimus-induced tremor listed in Table 2 [4, 5, 9, 13‐15, 20, 25, 27, 37, 41‐73] mainly comprised kidney (n = 18) and liver transplantation (n = 17), with the rest including bowel-related diseases (inflammatory bowel disease, n = 7), heart transplant (n = 2), rheumatoid arthritis (n = 1), and bone marrow transplant (n = 1). Reports of tremor was geographically distributed as follows: 20 studies from the Americas, 13 from Europe, eight from the Western-Pacific Region, and two from the Eastern-Mediterranean Region.

Although many studies involved a prospective design, a wide range (0.1–71.4%) for tremor was reported. Data extracted from larger cohorts (n > 100) revealed a frequency of 21.5 ± 8.4% (490/2276; 12 studies) for tremor (Fig. 2A). In one of the largest studies by Yocum et al. (n > 800), the prevalence of tremor was observed to be lower (9.1%) [73]. The risk of tremor in nontransplant indications was about 11.3 ± 9.7% (Fig. 2B). While most studies did not report the time frame, some specified the tremor as early onset (n = 10) and some reported as late onset (n = 4). Three studies reported patients with tremors in both settings. Most studies reported the risk of tremor in the context of kidney (n = 18) and liver (n = 17) transplantation, with one study including patients with both kidney and liver transplantation (n = 1). Tremor was also reported in the context of inflammatory bowel disease (n = 7) and autoimmune/connective tissue diseases (n = 1). The risk of tacrolimus-induced tremor was higher in patients with kidney transplants (30.1 ± 19.7%) and bone marrow transplants (41.9 ± 1.7%) compared with patients with liver transplants (11.5 ± 9.8%).

Similar to cyclosporine, the majority of the studies employed plasma monitoring of drug concentrations for therapeutic dosing (Table 2); however, longitudinal monitoring and, more importantly, data on whether blood concentrations correlated with tremor intensity were lacking. A few studies compared the risk of tremor with twice-daily/immediate-release versus once-daily/prolonged or extended-release tacrolimus [42, 52, 63, 64, 68]. The incidence of tremor with the twice-daily formulation had a range of 1.7–34.5% compared with the once-daily formulation range of 2–27.5% (Fig. 3). In one study, tremor was less likely to develop when converted to a once-daily extended-release formulation that was prepared with Meltdose technology (improved tacrolimus bioavailability) [64]. Similarly, in another study, conversion to extended-release tacrolimus led to a reduction in tremor intensity [58]. However, several other studies did not observe a statistically significant difference in the incidence of tremors between twice-daily immediate-release and once-daily extended-release formulations [42, 52, 63]. Some studies reported co-occurring neurological side effects such as headaches (n = 16), seizures (n = 12), paresthesia/hypoesthesia/hyperesthesia (n = 9), and insomnia/sleep disorders (n = 8). A small number of studies reported more serious neurotoxic effects, such as encephalopathy (n = 3), visual disturbances (n = 2), and coma (n = 5). Occasionally studies reported co-occurring ataxia (n = 1) and parkinsonism (n = 1).

None of the studies presented data on patient demographics and potential risk factors for tremors. Many studies also did not specify whether a treatment for tremor was employed (n = 20). Of the studies that described the various strategies for mitigating tremor, we found some changed to a lower dosage (employed by n = 11 studies), one study switched patients to another immunosuppressant drug such as cyclosporine, some studies used longer acting tacrolimus (n = 5), tapered tacrolimus to eventual discontinuation (n = 3), and supplemented with magnesium (n = 1).

While multi-drug therapy for immunosuppression means that these drugs were frequently co-administered with CNIs, a few studies have reported the occurrence of tremor in the context of sirolimus alone, without the presence of a CNI (n = 5) [14, 26, 74‐76] (Table 3). The geographical distribution of reports on sirolimus-induced tremor was as follows: three studies from the Americas, one from Europe, and one from the Western Pacific regions. These studies primarily included patients with a kidney transplant; one study included patients with heart and liver transplants. When combining data from all studies, tremors were seen in 22 out of 402 patients, with an average of 12.9% (range 0.48–44%). In the post-marketing surveillance from Korea (n = 209, largest), the prevalence of tremor was 0.48%. While there was no comment on the timeframe of tremor development (n = 3), tremor was reported as early onset by one study and late onset by one study. Two studies reported specific trough concentrations, and one described trough concentrations as within the therapeutic range. No treatment was provided for tremors (n = 3), and the tremors were described as self-limiting (n = 2). One study reported headaches, insomnia, and fatigue and another study reported depression, polyneuropathy, transient ischemic attacks, seizures, and stroke to co-occur with tremor.

Tremor related to the use of everolimus was reported in a recent study in the setting of a kidney transplant. Unlike sirolimus, everolimus was administered in conjunction with low-dose CNI therapy. The tremor prevalence was noted to be about 9.7% among 1014 patients across 42 countries [77]. Dose adjustment with everolimus was observed to reduce the intensity of tremor.

In the early 1960s, azathioprine and prednisolone were the main immunosuppressive agents used after solid organ transplant surgery [3]. However, these drugs were associated with a 50% rate of acute rejection leading to reduced allograft survival [78]. Cyclosporine was introduced in 1982 and the drug with potent immunosuppression properties worked as a CNI. With availability of this drug, the incidence of acute rejection after solid organ transplant surgery significantly dropped to 25% [79]. Furthermore, there were substantial improvements in the rates of allograft survival [80] but 20–39% of patients were observed to develop tremor [81]. The neurotoxicity risk with cyclosporine in bone marrow transplant recipients was found to be as low (4.2–28.8%) [82, 83]. A few years later in 1987, tacrolimus, another CNI agent, was introduced for immunosuppression purposes [84]. Compared with cyclosporine, tacrolimus was 100 times more potent, thus leading to further lowering of the incidence of acute rejection and improvement of graft survival rates. Tacrolimus became an integral part of induction and maintenance immunosuppression therapy. Currently, more than 90% of solid organ transplant recipients are discharged with a CNI‐based immunosuppressive regimen and the majority of them receive tacrolimus instead of cyclosporine [3, 85, 86]. Some studies compared the safety and efficacy profiles of the two drugs. In a phase III US multicenter trial, at the 1-year follow-up after transplant surgery, there was a significantly low incidence of acute rejection in the tacrolimus group (30.7%) versus the cyclosporine group (46.4%). At 5 years, the rate of patients with serum creatinine levels >150 μg/L was lower in the tacrolimus group (40.4%) versus the cyclosporine group (62%), but tremor and paresthesia were more common in the tacrolimus group (54.1%) versus the cyclosporine group (33.8%) [27]. In one study, over 70% of patients had resolution of neurological side effects arising from tacrolimus therapy when switched to cyclosporine [87]. In another study with prospectively collected single-center data, the prevalence rates of tremor were even higher with cyclosporine (62%) and tacrolimus (82%) [14]. Webster and associates, in a meta-analysis of 4102 patients, found significant reductions in graft loss in tacrolimus-treated recipients compared with cyclosporine, but more tremor, headache, diarrhea, dyspepsia, and vomiting [88]. In contrast to these data, our literature review found that tremor prevalence rates with cyclosporine (17%) and tacrolimus (21.5%) were lower overall and shared a similar likelihood of developing tremors.

Cyclosporine (rapamycin) forms a complex with cyclophilin, whereas sirolimus/rapamycin bind to a family of immunophilins called FK506 (same as tacrolimus). While acting on the FK binding protein-12, sirolimus, which interferes with the cellular signal transduction pathways, does not inhibit calcineurin [89]. Sirolimus has gained attention as an immunosuppressive therapy in organ transplantation mainly owing to its unique mechanism of action, reduced risk of organ toxicity, and its ability to synergize with other immunosuppressant drugs without overlapping toxicity [6]. One study directly compared sirolimus with cyclosporine to find a lower risk of tremor (10% vs 46%) [26].

Everolimus is a second-generation rapamycin derivative with functions of mammalian target of rapamycin inhibition similar to sirolimus but with a higher potency [7, 90]. The two drugs exhibit significant differences in their pharmacokinetic, pharmacodynamic, and toxicodynamic properties, resulting in distinct clinical profiles. In one large study involving kidney transplant patients who were randomized to receive either everolimus with lower than typical doses of CNI or mycophenolic acid with standard doses of CNI, the investigators found that nearly 9.7% of patients developed tremor when exposed to everolimus compared with 13.5% with standard exposure CNIs [77]; these findings suggest that the dose is an important contributor. When combining data from all sources, sirolimus and everolimus indeed appear to have a lower propensity to induce tremors than cyclosporine and tacrolimus agents (Fig. 2).

As the therapeutic window is narrow, CNI agents require individual dose titration and mandate serum concentration strict monitoring to achieve a correct balance between maximizing efficacy and minimizing dose-related toxicity. Transplant surgeries in the past found low tacrolimus concentrations to correlate with rejections, whereas high concentrations were associated with nephrotoxicity [91, 92]. Some studies have suggested the following risk factors including elevated blood concentrations of cyclosporine or tacrolimus [93], administration of drugs through the intravenous route, co-administration of drugs that inhibit cyclosporine and tacrolimus metabolism (e.g., high-dose methylprednisolone), and the presence of hypomagnesemia leading to the development of tremor [53, 94]. In our data analysis, tremors seemed to be more frequent in kidney transplant patients compared with liver transplant patients. Unfortunately, the prevalence of data on drug-induced tremor in bone marrow transplants was inadequate, and the heart transplant data were not individually presented (combined with kidney transplant data). In our analysis, most studies did not specifically find a correlation between tremor intensity and drug concentrations. While most studies tended to monitor the trough concentrations, one study in the literature showed that the incidence of adverse reactions with cyclosporine was not different whether the area under the curve or trough concentration was monitored [95]. Few studies mentioned whether the tremor was early onset during the induction phase (drug concentrations relatively high) or the chronic maintenance phase when lower doses were employed. Finally, the studies included in our review did not ascertain whether a higher propensity of drug-induced tremors accompanied older age, male sex, and polypharmacy [1, 80].

Several clinical and nonclinical studies have shown that the pharmacokinetics of twice-daily and once-daily tacrolimus are significantly different [42]. Once-daily extended-release formulation has about a 30% greater relative bioavailability, about a 30% lower peak-to-trough fluctuation, and a consistently lower daily dose than prolonged-release tacrolimus [63, 96‐98]. In one pooled analysis of over 800 kidney transplant recipients, a longer-acting once-daily formulation was found to be as effective as twice-daily tacrolimus [99]. The STRATO clinical trial found a lower risk of tremor with the once-daily extended-release formulation compared with twice-daily tacrolimus formulations [58]. In one study examining the real-world experience with conversion, a pre-conversion incidence of tremor (20.8%) was found to be substantially lower post-conversion (11.8%, p < 0.0001) [64]. In another study, among patients who were switched from twice-daily to once-daily formulations because of tremors, 88% reported a significant amelioration in symptoms [50]. In contrast to these data, some studies did not find a significant difference between the two formulations in the prevalence of tremors [42, 63]. In our assessment of combined data, we found that the tremor developing with a twice-daily tacrolimus formulation (17%) was comparable to a once-daily formulation (18%). Thus, further research could potentially resolve the current inconsistencies in findings.

While the exact mechanism underlying the immunosuppressant-induced tremor is unknown, we speculate that several factors contribute to the pathophysiology. The intracellular binding proteins for both cyclosporine (cyclophilin) and tacrolimus (FK binding protein-12 or FKBP-12) [100, 101] are enriched in the central nervous system [102]. Cyclosporine and tacrolimus via calcineurin inhibition likely modulates the activity of both excitatory NMDA and inhibitory GABA receptors [80]. The pathophysiological mechanism for tremor oscillations is robustly linked to a dysfunction of these receptors [103]. Given the symptoms of ataxia co-occurring in some patients [14, 104], there is no doubt that the cerebellum is involved; a critical substrate also for the pathogenesis of tremor. It could be speculated that the use of cyclosporine and tacrolimus lowers the threshold for tremor generation for patients predisposed to developing these symptoms (a two-hit hypothesis). Many clinicians find that the tremors improve with a lowering of the dosage or discontinuation of the CNI drugs; however, some patients continue to exhibit bothersome tremors despite these measures, suggesting that the drugs could induce permanent damage in the tremor circuitry. The lower prevalence or risk of tremors with non-CNI drugs is intriguing, either related to the paucity of data or mechanistically, the drugs do not affect the tremor circuitries, but this will need further work.

There are no standard guidelines for treating tremor resulting from the use of cyclosporine, tacrolimus, sirolimus, and everolimus. In our review of the data, most studies that reported the prevalence of tremor did not address the potential treatment modalities. Some studies implemented the discontinuation or reduction of the dosage to achieve remission or amelioration. Some studies reported that control of magnesium levels could be a potential strategy to lower the possibility of developing tremor [24, 35]. Some patients with immunosuppressant drug-induced tremor may respond to treatments proven to be of benefit for other tremor disorders such as propranolol and primidone [105, 106]. These potential modalities will need systematic studies in large well-characterized cohort of immunosuppressant drug-induced tremors.

We acknowledge that there are pitfalls in interpreting the data extracted from the literature. Pooled analyses using statistical tests could not be conducted as the included studies did not use the same study design, populations were not homogeneous, and the reporting methods used by most studies were quite heterogeneous. Many studies did not report the onset of tremors and whether there was a definitive temporal relationship between drug initiation and tremor. Most studies did not list the specific method for clinical or electrophysiological characterization. Although kidney and bone marrow transplant studies reportedly had a higher prevalence rate of tremor, we acknowledge that there could be a publication bias. Moreover, some data could not be analyzed as diverse etiologies were combined in a single dataset, such as kidney and heart transplant data. Finally, specific risk factors applicable to the emergence of tremors were not addressed.