A fluid relationship: Calcineurin inhibitors and pericardial effusions

Hematopoietic stem cell transplantation (HSCT) is a common and effective treatment for multiple malignant and non‐malignant pediatric conditions. Graft‐versus‐host disease (GVHD) is a common complication of HSCT that can be prevented with prophylactic use of calcineurin inhibitor (CNI) immunosuppressants. A complication of HSCT and CNI use is pericardial effusion (PEF), which is frequently treated by CNI discontinuation with or without surgical intervention. No studies to date have evaluated the management of PEF without CNI discontinuation as a means of preventing GVHD flares.

cardiovascular toxicity that must be closely monitored. 4A complication associated with the use of CNI post-HSCT that can cause significant morbidity is pericardial effusion (PEF).Previous studies have reported the incidence of PEF post-HSCT from 4.4% to 37.8%. 5,6nservative management of PEF generally involves discontinuing the CNI, but more severe cases often additionally require surgical intervention. 2Although CNI discontinuation has proven effective in treating PEF, this intervention also runs the risk of inciting or exacerbating GVHD.Replacement of the CNI with other agents (e.g., mycophenolate mofetil or corticosteroids) could be advantageous. 7wever, this requires titration and may provide more or less immunosuppression than needed when the patients are already at high risk of acute or chronic GVHD flare or exacerbation.
Interventions for PEF include medical management or surgical intervention such as pericardiocentesis or pericardial window with placement of a surgical pericardial drain.Pericardiocentesis is often preferred because the equipment is more readily available, and the procedure is less invasive. 8Pericardiocentesis can also typically be performed with the patient under conscious sedation without general anesthetic and positive pressure ventilation, both of which can further worsen the physiologic effects tamponade physiology.A recent study in adults with PEF showed that pericardiocentesis and pericardial window have very similar outcomes with few complications. 8However, pericardiocentesis had a much higher rate of PEF re-accumulation than pericardial window.Re-accumulation could be reduced by the placement of a drain. 8Medical therapies for pericardial effusions are typically tapered to best modulate the suspected etiology.For patients with suspicion of an inflammatory process, such as related to pericarditis or post-pericardiotomy syndrome, most commonly non-steroidal anti-inflammatory drugs (NSAID) are used.Outside of inflammatory processes, medical treatments related to fluid imbalance or renal impairment present a particularly challenging situation.The patient may have renal impairment making diuresis challenging while, paradoxically, diuresis would likely benefit them from an effusion resorption standpoint.Both surgical approaches have been used previously to treat PEF associated with HSCT for pediatric patients. 1,9However, no studies to date have fully evaluated the efficacy of conservative or surgical intervention in treating PEF without CNI discontinuation.In this study, we performed a retrospective analysis to compare outcomes in pediatric HSCT recipients who received conservative or surgical intervention for PEF with or without CNI discontinuation.

| Study design
We performed a single-institution retrospective study of 194 pediatric patients undergoing allogenic HSCT at Riley Hospital for Children in Indianapolis, Indiana over a 10-year period from May 1, 2012, through June 1, 2022.Patients were excluded if they did not receive CNI prophylaxis for GVHD.We then specifically looked at patients who developed a PEF at any timepoint post-HSCT within the 10-year timeframe.
Descriptive statistics were used to analyze demographic data and were summarized by frequency and percentage for categorical variables.Quantitative statistics comparing age were performed using t-test.Statistics comparing mortality rates and PEF development with tacrolimus versus cyclosporine were performed using Fisher's exact test.Differences were considered statistically significant for p < .05.

| Data collection
Demographic data such as age, gender, ethnicity, and diagnosis were obtained from patient medical records.Additionally, information regarding stem cell transplant course and PEF diagnosis and management was reviewed and collected.All reviewed data were originally collected as part of each patient's routine clinical care.This study was approved by the institutional review board of Indiana University.

| Echocardiographic screening
All patients undergoing HSCT received echocardiograms prior to HSCT for baseline cardiac anatomy screening.Echocardiograms were performed at the time of concern for PEF.Additional echocardiograms were performed as needed to monitor for PEF resolution or re-accumulation after surgical draining, and at scheduled annual posttransplant follow-up appointments.

| Patient characteristics
Of the 194 pediatric HSCT recipients, 18 were excluded for not receiving CNI prophylaxis for GVHD.The male-to-female ratio for the remaining patients was 1.6:1.The average age of patients at HSCT was 9.8 years (range: 2 months-27 years).Approximately 70.5% underwent HSCT as part of treatment for malignant disorders, and 29.5% for nonmalignant disorders.For GVHD prophylaxis, 34.7% (61/176) of patients received tacrolimus and 65.3% (115/176) received cyclosporine.Patient characteristics for this cohort are summarized in Table 1.

| PEF incidence
Six patients were identified as having developed a PEF post-HSCT and were included in our analysis.See Table 2 for patient characteristics and demographics.The average age at HSCT for these patients was 8.1 years (range: 1-17 years).Although this average age is younger than that of the total cohort, the age difference was not statistically significant (p = .55).From the total number of patients undergoing allogenic HSCT with CNI prophylaxis for GVHD, the incidence of developing PEF post-HSCT at our institution was 3.4% (6/176).All of the patients who developed PEF received tacrolimus as GVHD prophylaxis.PEF incidence was 9.8% (6/61) for the patients receiving tacrolimus and 0% (0/115) for the patients receiving cyclosporine (p = .0015).The average time from day 0 of HSCT to development of PEF was 92 days (range: 58-128 days).

| Management of PEF
Echocardiographic measurements of PEF size ranged from mild to large; one (16.6%)patient with mild PEF, one patient with moderate PEF (16.6%) and four patients with a large PEF (66.6%).Two patients had associated pericardial tamponade with the PEF diagnosis.Most patients (n = 5, 83%) were treated with pericardiocentesis, and two of the six patients (33%) required a pericardial window, as shown in Table 3.Of the five patients that required pericardiocentesis, two (40%) required a second pericardiocentesis.For both patients, the time from first to second pericardiocentesis was 11 days.One of these two patients additionally went on to receive a pericardial window.For those who required pericardiocentesis, the average time for drainage tube requirement was 5.4 days (range: 2-15 days).
Two patients received NSAID therapy as a treatment for PEF.
One patient received indomethacin, with colchicine added later, and the other patient received ibuprofen.One patient was receiving dasatinib post-HSCT, but this was discontinued with the finding of PEF due to the known association between this tyrosine kinase inhibitor and PEF.
Most of our patients (n = 5, 83%) developed a PEF after their first and only HSCT, while one patient developed a PEF after a second HSCT due to relapsed ALL, as shown in Table 2.
Pericardial fluid samples collected by pericardiocentesis were sent for laboratory testing.All samples were described as clear serous fluid and were subsequently sent off for infectious disease testing.None were determined to contain infectious organisms as a potential cause of the PEF.

| Management of GVHD
All six patients received tacrolimus as the CNI for GVHD prophylaxis, as shown in Table 3.Five of the six patients (83%) continued tacrolimus as their GVHD prophylaxis after developing the PEF, TA B L E 1 Characteristics of pediatric HSCT patients who received CNI for GVHD prophylaxis.while one patient was switched from tacrolimus to sirolimus due to concurrent diagnosis of PEF and transplant-associated thrombotic microangiopathy (TA-TMA).Switching GVHD prophylaxis medications did not cause a flare in this patient's known GVHD.
Three patients had a history of acute GVHD prior to the development of PEF.Two had previously been diagnosed with grade 2 acute skin GVHD, and the other with grade 3 acute gut GVHD.At the time of PEF diagnosis, the skin and gut GVHD in these patients were grade 0 and quiescent.The patient with quiescent gut GVHD was on oral prednisone and oral budesonide at the time the PEF was noted.One patient with a history of acute skin GVHD later went on to develop chronic GVHD.However, this occurred approximately 3 months after weaning off tacrolimus and 4 months after PEF diagnosis.No patients developed acute or chronic GVHD while being treated for their PEF.

| Other complications
Patient 5 developed TA-TMA 29 days post-HSCT.This patient had already completed treatment with eculizumab, with resolution of the TA-TMA, before developing the PEF.The PEF in this patient was not thought to be associated with TA-TMA.Patient 6 was diagnosed with TA-TMA on the same day the PEF was noted and was treated with eculizumab.
Of note, Patient 1 developed PEF during their second HSCT for relapsed disease and was receiving tacrolimus as GVHD prophylaxis at that time.For their first HSCT, cyclosporine was utilized as GVHD prophylaxis.This patient was diagnosed with disseminated toxoplasmosis shortly after noting the PEF.Subsequent cardiovascular instability precluded performing surgical drainage of the PEF.
Echocardiography reported the PEF as mild, and repeat echocardiograms showed no change in size, supporting the decision to avoid surgical intervention.This patient ultimately passed away due to complications associated with the disseminated infection.
Among the cohort of 170 patients who did not develop a PEF, 46 ultimately passed away (27.1%).This is in comparison to one death out of six patients who developed a PEF (16.7%).However, this difference was not statistically significant (p = 1.00).

| DISCUSS ION
Our retrospective analysis aimed to review the association between CNI use and the development and management of PEF.Patients who receive allogenic HSCT will often require CNI use as part of their GVHD prophylaxis regimen.Tacrolimus is a commonly used CNI and is frequently associated with electrolyte derangements, nephrotoxicity, neurotoxicity, and cardiovascular toxicity, such as PEF. 4,10,11PEF has previously been reported to occur in about 4.4%-37.8% of HSCT patients. 5,6Our data show an incidence of There is a broad range of etiologies for pericardial effusions post-HSCT, including relapsed malignancy, medications, and infections.
In our retrospective review, none of the pericardial effusions were related to malignancy.Patient 1 was noted to have a rhinovirus-positive respiratory viral panel 2 days prior to PEF diagnosis.This patient was also diagnosed with disseminated toxoplasmosis shortly after PEF development.Patient 3 additionally had respiratory symptoms the day prior to PEF diagnosis with negative infectious laboratory workup when the PEF was noted.After pericardiocentesis, infectious disease testing of the pericardial fluid was negative.Although unlikely, infection could have been a confounding factor in both cases.
One of the six patients was noted to have renal impairment around the time of PEF diagnosis.This was noted by stable elevation of creatinine from baseline in patient 5.This is likely due to resolving renal impairment secondary to TA-TMA that required dialysis 2 months prior.However, it is important to note that uremia from renal impairment can lead to osmotic load that could result in a PEF.Versluys et al. 2 found that young age at the time of HSCT was a risk factor for the development of a PEF.Concurrently, a study by Jaing et al. 12 showed the opposite pattern, where patients with PEF post-HSCT were older than patients without PEF.Similar to Versluys et al., 2 on average our patients who developed PEF were younger than those who did not (8.1 vs. 9.8 years), as shown in Table 1.
However, this difference was not statistically significant (p = .55).
4][15] TA-TMA is known to be correlated with CNI therapy, and often changing GVHD prophylaxis treatment is required as an adjunct to treatment along with eculizumab. 2 From our cohort, one of the six patients (16.7%) developed TA-TMA at the time of PEF diagnosis.This patient was noted to have a large pericardial effusion, the longest number of chest tube days (15 days) in our cohort, and was one of two patients who required a pericardial window.Although this patient was switched from tacrolimus to sirolimus, it is possible the cardiac procedures and treatment of TA-TMA with eculizumab would have been sufficient to improve his PEF.This may suggest that TA-TMA-associated PEF involves multimodal treatment.Of note, another patient was previously diagnosed with TA-TMA on day 29 post-HSCT, completed treatment with eculizumab, and TA-TMA screening lab values (LDH, haptoglobin and sc5b9) normalized before being diagnosed with PEF on day 128 post-HSCT.Therefore, a correlation between the two conditions is unlikely for this patient.
Literature findings have supported the discontinuation of CNIs with the diagnosis of CNI-induced PEF. 2,16Versluys et al. 2 noted that they stopped CNI's in eight of their 12 patients who developed PEF.
Of those eight, five patients still required pericardiocentesis. 2 It is not our institutional standard to discontinue CNI if a PEF is diagnosed as there is often not another good non-CNI option for GVHD prophylaxis, and pericardiocentesis or pericardial window is frequently needed to resolve the PEF.The notion of not stopping CNIs after diagnosis of PEF post-HSCT is only hinted at in the literature and is an important option to consider in patients who are at risk for an acute GVHD flare or the development of GVHD.Given that there are other options that can aid in resolving the PEF (pericardiocentesis, pericardial window, NSAIDs, and colchicine) it is important to consider continuing the CNI if the patient can tolerate it.Interestingly, all the patients in our study who developed PEF received tacrolimus for GVHD prophylaxis.To our knowledge, there are no data published that suggest an increased risk of developing PEF post-HSCT with tacrolimus versus cyclosporine.The CNI chosen to prevent GVHD at our institution is standardized based on the donor, conditioning, and on provider preference.In more recent years there has been a growing body of literature to support tacrolimus as prophylaxis against GVHD over cyclosporine in certain types of transplants. 17Our data suggest that there may be an increased risk of PEF with the use of tacrolimus versus cyclosporine, but this will need to be confirmed with larger retrospective and/or prospective studies.
Pericardial window was placed in two of the six patients, but the majority were treated with pericardiocentesis alone.One patient did not require any intervention and two patients required a second pericardiocentesis.At our institution, most patients are managed initially with pericardiocentesis and pericardial drain placement.Surgical pericardial window creation and drain placement is pursued for patients with recurrence or prolonged drain placement without improvement in drain output.Cardiac follow-up for our patients who developed PEF showed normal anatomy and function on echocardiogram evaluation.
The mortality rate for our patients who developed a PEF (n = 1, 16.7%) was lower than the mortality rate for patients who did not develop a PEF (n = 46, 27.1%).This difference, however, was not statistically significant.The cause of death for the one patient with PEF was secondary to disseminated toxoplasmosis, not related to the PEF.Although some studies have correlated PEF with higher mortality post-HSCT, 1,6 a separate study determined that PEF had no effect on overall mortality. 12Our results support the conclusion that PEF does not correlate with higher post-HSCT mortality.
The major limitation of our retrospective study is the small number of patients diagnosed with PEF.It is possible that some of the noted differences would be statistically significant with a larger sample size.Another limitation of this study is that our institution does not have a protocol for consistent post-HSCT echocardiograms.
Because of this, it is possible that minor, asymptomatic cases of PEF were missed.
In conclusion, we found that continuation of CNI therapy for GVHD prophylaxis did not negatively impact the disease course of PEF in patients post-HSCT in the absence of TA-TMA.This was demonstrated by the resolution of the PEF for each of our patients without adverse outcomes related to the PEF diagnosis.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors have no financial disclosures relevant to this article.
There was no funding obtained for this study.The authors have no conflicts of interest relevant to this article to disclose.
Demographic characteristics and transplantation regimens for patients who developed PEF.