Abstract
Introduction
Over the span of the last few decades, calcineurin inhibitors, such as tacrolimus or cyclosporine, have served as a regimen backbone in the prevention of graft-versus-host disease (GVHD) in allogeneic stem cell transplant recipients. Initial continuous intravenous administrations have been favored due to the consistent, steady-state degree of immunosuppression, particularly prior to engraftment, as opposed to pharmacokinetic peaks and troughs of intermittent dosing. While some centers opt to initiate patients on oral formulations, concern for absorption secondary to mucositis, intestinal GVHD, or other factors may necessitate intravenous dosing. Our institution utilizes intravenous tacrolimus as the calcineurin inhibitor of choice for initial GVHD prevention, however scant literature exists regarding alternative non-enteral strategies to tacrolimus continuous intravenous (CIV) administration.
There are limited data on the practicality of adopting an intermittent intravenous infusion (IIV) tacrolimus approach.1,2 Prior reports cited caution against IIV tacrolimus use due to increased rates of nephrotoxic and neurologic complications as was observed in solid organ transplant recipients although dosing was generally higher than 0.03 mg/kg/day in that population. 1 Additionally, some authors place preference for initial use of intravenous tacrolimus but note that the continuous infusion method is a disadvantage for its use. 1
To date, few published studies report on the utility of IIV tacrolimus, and none compare tacrolimus IIV to CIV. These studies either validated their current tacrolimus IIV practices or concluded IIV was a feasible alternative to CIV (Table 1).
Literature review of intermittent intravenous tacrolimus use in allogeneic stem cell transplant.
47.1% did not have complete follow-up.
Serum creatinine ≥ 2 mg/dL.
Acute kidney injury defined as a increase in serum creatinine ≥ 0.3 mg/dL in a 48-h period.
Follow-up timeframe not specified.
NR: not reported; BID: twice daily; aGVHD: acute graft-versus-host disease.
Although endpoint definitions were heterogeneous, in general, nephrotoxicity incidence ranged from 10 to 42% and severe neurotoxicity 0–9.8%. Acute GVHD (aGVHD) grades II–IV were less than 35% and relapse rates were 3.3–8.7% by day +100–180 depending on study methodology.3–6 Chronic GVHD (cGVHD) was seldomly reported due to variation in follow up durations but occurred in up to 25% of cases according to Skeens et al. 5
At the advent of our center's allogeneic stem cell transplant program, CIV tacrolimus (0.03 mg/kg/day based on ideal body weight [IBW]) was our preferred administration method with a general goal range of 5–15 ng/mL. Tacrolimus monitoring was driven by the pharmacy team in collaboration with the transplanting physician. Whole blood levels were drawn at least three times weekly during the IV tacrolimus period but could be monitored more frequently per clinician discretion for reasons such as renal dysfunction or a change in drug interactions.
We anecdotally noted that this CIV approach was logistically tedious for both patients and staff and increased safety risks with infusion pump programming among other complications. As noted by Hashmi and colleagues, allogeneic stem cell transplant patients are susceptible to requiring multiple intravenous infusions for a variety of reasons; the continuous use of a patient's lumen may not be ideal in some scenarios. 4 Additionally, the stability of a 24-h tacrolimus infusion can pose a risk for drug expiring while infusing. Furthermore, extension tubing requirements for imaging, lab result delays, and timing of oral conversion can lead to suboptimal use of resources and overall complicate the patient's care. In response to these concerns, we implemented IIV tacrolimus at a starting dose of 0.015 mg/kg (IBW) twice daily over 4 h. Given the paucity of data, to our knowledge this is the first reported retrospective comparison of CIV to IIV tacrolimus.
Objectives
The primary objective was to evaluate the safety of IIV tacrolimus in comparison to CIV with respect to development of nephrotoxicity and neurotoxicity. The secondary objectives were to compare incidence of grade II-IV aGVHD and any-stage cGVHD, outcomes at 180 days post-transplant (day +180) including relapse and overall survival, tacrolimus level results, time to neutrophil and platelet engraftment, and reactivation of cytomegalovirus (CMV) and Epstein-Barr virus (EBV).
Methods
This retrospective, single-center review obtained data via the electronic medical record for all patients greater than or equal to 18 years of age who received an allogeneic stem cell transplant within our institution between January 1, 2020, and December 31, 2022. All patients were followed until at least day +180 post-transplant or death, whichever occurred first. Patients were evaluated for nephrotoxicity and neurotoxicity incidence while receiving CIV or IIV tacrolimus (until conversion to oral tacrolimus). Patients were excluded from evaluation if they were previously on oral tacrolimus and required conversion back to intravenous for any reason. Tacrolimus levels obtained within 24 h of tacrolimus initiation or samples drawn in error (i.e., during running infusion) were excluded.
Nephrotoxicity was defined as serum creatinine two times greater than pre-tacrolimus baseline or greater than 2 mg/dL.4,5 Neurotoxicity was identified via provider documentation or brain magnetic resonance imaging (MRI) of a mild (tremor, paresthesia, headache, visual disturbance) or severe (seizure, posterior reversible encephalopathy) side effect. Neutrophil engraftment was defined as an absolute neutrophil count (ANC) greater than or equal to 0.5 × 109/L for three consecutive days; platelet engraftment was a count greater than or equal to 20 × 109/L for 3 consecutive days without transfusion support. Outcomes assessments at day +180 analyzed the incidence of aGVHD and cGVHD, incidence of relapse detected, and survival. One-year overall survival probability was estimated. CMV and EBV reactivation were classified as reactivation levels that required treatment per institutional standard.
Statistical analyses included Pearson's chi-squared test or Fisher's exact test for categorical variables and a two-sample t-test or Wilcoxon Rank Sum test for continuous variables; a p-value < 0.05 was considered statistically significant. RStudio V.2022.07.0 (R Foundation for Statistical Computing, Vienna, Austria, URL https://www.R-project.org/) was used for the following analyses. Overall survival probability was calculated using Kaplan-Meier curves. Cumulative incidence of aGVHD, cGVHD, and relapse, accounting for a competing risk of death, were calculated using the Gray's test and reported as incidence with a 95% confidence interval (95% CI).
Results
Fifty-one unique patients were eligible for evaluation – 23 in the CIV group and 28 in the IIV cohort. Baseline demographic data is summarized in Table 2.
Population baseline data and characteristics.
Significant p-Values are in bold type.
Other races included Asian and American Indian.
Other diagnoses included: mixed acute leukemia, chronic myelomonocytic leukemia, chronic neutrophilic leukemia, chronic lymphocytic leukemia, and hairy cell leukemia.
Four patients received cells from haploidentical donors.
ALL: acute lymphoblastic leukemia; AML: acute myeloid leukemia; ATG: antithymocyte globulin; BM: bone marrow; CIV: continuous infusion; CMV: cytomegalovirus; D/R: donor/recipient; GVHD: graft-versus-host disease; IIV: intermittent infusion; MAC: myeloablative conditioning; MMF; mycophenolate mofetil; MMRD: mismatched related donor; MMUD: mismatched unrelated donor; MRD: matched related donor; MTX: methotrexate; MUD: matched unrelated donor; PBSC: peripheral blood stem cell; PTCy: post-transplant cyclophosphamide; RIC: reduced intensity conditioning; SD: standard deviation.
Overall, the average age was 56.7 (range, 21–75) years and a majority were male (61%). Most patients received myeloablative conditioning (57%) with common regimens being fludarabine/busulfan (cumulative AUC 20,000 µMol x min) or a radiation-based regimen plus fludarabine, etoposide, or cyclophosphamide. Acute myeloblastic leukemia (41%) or myelodysplastic syndrome (24%) compromised the majority of transplant indications. Donor type was predominantly matched unrelated (59%) followed by matched related (27%), mismatch related (7.8%), and mismatch unrelated (5.9%). Peripheral blood stem cell grafts were used in 98% of patients. GVHD prophylaxis was comprised of a tacrolimus backbone plus antithymocyte globulin/mini-dose methotrexate (59%), mini-dose methotrexate alone (27%), or post-transplant cyclophosphamide plus mycophenolate mofetil (14%). The only significant difference at baseline between groups was CMV serostatus; 88% of recipients were CMV seropositive.
Safety
Figure 1 illustrates primary safety outcomes.
Primary safety outcomes. Intermittent infusion (IIV) n = 28, continuous infusion (CIV) n = 23.
Day +180 overall survival probability and number at-risk. The figure illustrates overall survival probability at day +180 between continuous infusion (CIV) and intermittent infusion (IIV) groups - IIV 96% (95% CI 90-100) vs CIV 87% (95% CI 74-100), p = 0.22.
Nephrotoxicity occurred in 4 patients (7.8%), two in each group (p > 0.9). Of these, three met the predefined criteria of a serum creatinine > 2 mg/dL and one had creatinine doubled from baseline. All four patients were also receiving at least two of the following concomitant nephrotoxins: intravenous acyclovir, contrast, diuretics, or vancomycin. Both patients in the CIV group received cyclophosphamide-based GVHD prophylaxis and required hemodialysis secondary to suspected multidrug-induced AKI. Neither of the IIV patients required intermittent hemodialysis, and each received GVHD prophylaxis consisting of antithymocyte globulin and mini-dose methotrexate.
Mild neurotoxicity was observed in 11 IIV patients and 7 CIV patients (39% vs 30%, p = 0.5). Of those with neurotoxic adverse events, headache was the most reported symptom (66.7%) followed by tremor (27.8%) and paresthesia (5.6%). No severe neurotoxic events occurred.
Efficacy
Secondary outcomes are described in Table 3. Acute GVHD grades II-IV at day +180, when censored for death, was observed in 10 (40% [95% CI 17.4–56.4]) patients in the IIV group compared to 10 (45.5% [95% CI 20.1–62.8]) in the CIV group (p = 0.69). Three additional patients in the IIV group were found to have aGVHD of unknown stage. No grade IV aGVHD occurred in any patient. Systemic steroids were needed in 75% and 57% of IIV and CIV patients, respectively (p = 0.2).
Secondary outcomes.
Significant p-Values are in bold type.
Three additional patients were not evaluable due to unknown GVHD grade.
One patient did not meet criteria for ANC engraftment.
Three patients did not meet criteria for PLT engraftment.
Total number of tacrolimus doses in CIV and IIV were 424 and 996, respectively.
Total number of tacrolimus levels in CIV and IIV were 240 and 351, respectively.
aGVHD: acute graft-versus-host disease; ANC: absolute neutrophil count; cGVHD: chronic graft-versus-host disease; CI: confidence interval; CIV: continuous infusion; CMV: cytomegalovirus; EBV: Epstein-Barr virus; IIV: intermittent infusion; IQR: interquartile range; LOS: length of stay; PLT: platelets; OS: overall survival; SD: standard deviation.
The incidence of cGVHD at day +180 was 18.2% (95% CI 2.4–31.3) and 9.8% (95% CI 0–21.8) in IIV and CIV cohorts, respectively (p = 0.35). Five patients in the IIV group and 6 patients in the CIV group had detectable relapse by day +180–18.3% (95% CI 2.4–31.6) IIV vs 27.3% (95% CI 6.1–43.7) CIV, p = 0.34. Differences in overall survival at day +180 were not statistically significant: IIV 96% (95% CI 90–100) vs CIV 87% (95% CI 74–100) vs), p = 0.22, Figure 2). Estimated 1-year overall survival between IIV and CIV groups was 68.5% and 82.6%, respectively (p = 0.37).
The overall incidence of CMV and EBV reactivation was 33% and 17%; p-value comparison between groups was 0.2 and 0.3, respectively (Table 3).
Length of stay was a median of 27 days, ANC engraftment 15 days, and platelet engraftment 17 days with no statistical difference in any category. The mean tacrolimus level in the IIV group was 8.0 ng/mL compared to 8.5 ng/mL in the CIV group (p = 0.22). More tacrolimus doses were administered and more levels were obtained in the IIV group: median number of doses 17 versus 36 (p < 0.001) and median number of levels 9 vs 12 (p = 0.017).
Discussion
To our knowledge, this is the first report to compare IIV to CIV tacrolimus in allogeneic stem cell transplant recipients. In this study, there were no significant differences among our study population, aside from CMV serostatus. Nearly 1000 IIV doses were evaluated which is one of the largest assessments of IIV tacrolimus administration in adults. Nephrotoxicity and neurotoxicity safety events were comparable between groups, and no significantly discernible differences were identified in transplant outcomes which resulted in an acceptable use profile for IIV tacrolimus.
Prior publications comment on the safety risks involved with twice daily IV dosing but are single arm evaluations and compare only to historical data. In our study, we note no significant differences between IIV and CIV administration methods with respect to nephrotoxic and neurotoxic events. Using similar defining criteria, rates of nephrotoxicity in our IIV population were low (7.1%) compared to existing IIV tacrolimus reports (10–42%).3–6 Given the limited incidence and sample size we are unable to appropriately draw correlations between potential influencing factors such as GVHD prophylaxis strategies. However, it is important to note that two or more concomitant nephrotoxic agents were present in each nephrotoxicity case.
Regarding neurotoxicity, the incidence of mild self-limiting events such as headache or tremor (39%) was comparable to existing literature (24–36%)3,4; direct comparisons are challenging given the subjective nature of neurotoxicity reporting and documentation. Unlike a majority of the current IIV data, there were no severe neurotoxic events observed in this study. Ultimately, in our study the safety risks of IIV tacrolimus were not different than CIV.
Efficacy outcomes were not compromised with the use of IIV tacrolimus. The follow up duration of 180 days post-transplant matches the longest follow up time studied in the single arm IIV reports. The cumulative incidence of grade II-IV aGVHD was not different between groups (p = 0.69). Among the IIV cohort, data aligned with day +180 aGVHD incidence reported by Bacopoulos et al. (40% vs 35%). 3 The development of cGVHD was numerically lower in the CIV group but not statistically significant. This may be due, in part, to a higher percentage of patients in the CIV group experiencing relapse than the IIV group, but this was not statistically significant (p = 0.36 and 0.22, respectively). Compared to Hashmi et al., 4 grade II-IV aGVHD rates were similar but relapse was more common in our findings, perhaps due to the extended follow up period of day +180 compared to day +100. We cannot attribute any observed numerical difference or infer any correlation in relapse and survival to the tacrolimus administration strategies used. Length of stay and cell engraftment were not affected by IIV versus CIV.
In terms of tacrolimus data, we observed similar average tacrolimus results between groups. The target therapeutic goal range for all of the patients in our study was within the broad general goal of 5–15 ng/mL. 7 Only one patient who developed nephrotoxicity had a single level greater than 15 ng/mL (15.3 ng/mL). Among the entire population, no patients had levels greater than 20 ng/mL which is known to be a significant risk factor for development of nephrotoxicity, increasing risk 2.2-fold.8,9
Additionally, methodology of tacrolimus level draws and goal ranges while on CIV is inconsistent. There remains unknown evidence of equivalence between CIV and IIV levels. The former is obtained as a steady-state trough but results following adjustments may be difficult to interpret given the CIV half-life of nearly 24 h as reported in heart transplant data. 10 The latter introduces the presence of peaks and troughs which requires appropriate level timing and correlation. Nakamura et al. attempted to address this by proposing a correction factor and suggested CIV exposure targets may need to be higher than IIV or oral goal ranges, 11 but more investigation is needed to elucidate this theory. Given the implementation of a different administration approach, it is noteworthy that closer level monitoring was performed in the IIV group with a median of 3 more levels obtained per IIV patient compared to those who received CIV. As expected with a change from once daily continuous to twice daily intermittent administration, the number of doses was essentially doubled in the IIV group.
There is no current consensus on tacrolimus IV to oral conversion timing in stem cell transplant patients. By using an IIV approach, the oral conversion was able to be timed at the next due tacrolimus dose (every 12 h) which aligns with manufacturer recommendations for solid organ transplant indications. Additionally, per the package insert, IV tacrolimus once diluted is stable for 24 h. 10 The implementation of twice daily dosing eliminated concern for 24-h CIV expiration and allowed for simultaneous preparation of two 4-h bags with a reasonable in-date timeframe. Notably, there are data to support extended stability of IV tacrolimus up to 9 days. 12 Adoption of this dating would limit risk of drug waste and may allow for multiple days of dosing to be batched, especially for centers that adjust dose based on CIV rate instead of total drug in bag. For others, batching opportunities would be limited to periods between level monitoring after which doses could change.
Our IIV tacrolimus duration of 4 h was based on the existing guardrails within our center's infusion pump software which limited our ability to administer over a shorter period and aligns with the approach used by Hashmi et al. 4 They found a 42% incidence of acute kidney injury (defined as a serum creatinine ≥ 0.3 mg/dL within a 48 h period) and a 10% incidence of creatinine > 2.0 mg/dL. Neurotoxicity defined as headache and tremor were reported at lower rates, possibly given the subjective nature, but Hashmi et al. did note two severe neurotoxic events (seizure) which may be multifactorial. In contrast, Skeens et al. 5 and Hayek et al. 6 utilized 2 and 3-h infusion durations, respectively, each with varied results relating to adverse events and outcomes of interest. A study comparing 4-h IIV versus shorter time periods would best draw conclusions on if infusion durations affect outcomes.
Although not directly captured in our data, the adoption of IIV tacrolimus has led to minimization, if not elimination, of many logistical challenges with CIV administration. The decrease in tacrolimus line time from 24 h to 8 h a day allows opportunity for more patient mobility and standardized timing for tacrolimus level draws and oral conversion, while providing less compatibility concerns with other intravenous medication and risk of pump programming errors. Our institution will continue to utilize IIV as our standard method of intravenous tacrolimus administration.
This study was limited in its single-center, retrospective nature as well as the smaller overall sample size despite similarities between group size and characteristics. While nephrotoxicity incidence was able to objectively be assessed, neurotoxicity may be underreported due to reliance on patient reporting and caregiver documentation within the medical record. Within our community hospital, tacrolimus monitoring is led by pharmacists in collaboration with the transplanting provider. Variation in clinical judgement could have affected dose adjustments and level frequency which, in part, may have limited events of nephrotoxicity secondary to elevated levels. Further comparisons, ideally in a prospective setting, between CIV and IIV tacrolimus methods are needed to elucidate any potential differences in safety and clinical outcomes.
Conclusion
IIV tacrolimus is comparable to CIV in terms of nephrotoxicity and neurotoxicity events, while also maintaining similar outcomes at day +180 regarding aGVHD, cGVHD, relapse, and overall survival. Neither engraftment nor length of stay were delayed or prolonged with the use of IIV tacrolimus. Administration burden and logistical considerations were relieved by adopting a less tedious management approach. Given the study results and benefits observed, our institution will continue to utilize twice daily infusions as our standard IV tacrolimus method.
Footnotes
Authorship Contributions
MW, ZG, and SM conceptualized the project. MW performed a literature evaluation, collected, and analyzed data. JG, ZG, and SS assisted in the interpretation of results. MW wrote the first draft which was then revised, edited, and approved by JG, BM, ZG, SS, and SM.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
