Predictors of fluid overload in allogeneic hematopoietic cell transplant patients receiving post-transplant cyclophosphamide

Abstract

Background

Fluid overload (FO) commonly occurs during hospitalization for allogeneic hematopoietic cell transplantation (HCT). Grade 2–4 FO is associated with day +100 non-relapse mortality.1 Post-transplant cyclophosphamide (PTCY) for graft-versus-host disease prevention requires aggressive IV hydration to prevent hemorrhagic cystitis.

Materials and Methods

This is a single-center, retrospective, observational study conducted at an academic medical center via electronic chart review. Included patients received allogeneic HCT followed by PTCY on days +3 and +4. Patients were excluded for age < 18 years or incarceration. Primary endpoints are incidence of Grade 2–4 FO and associated risk factors. Descriptive and inferential statistics (i.e., Fisher's exact test, multivariable regression analysis) were used.

Results

Of 97 patients screened, 95 were included and 2 were excluded due to absence of weight measurements needed to grade FO. Median age was 60 years, 66.3% were male, 91.6% received reduced-intensity conditioning, 72.6% received haploidentical HCT, 44.2% were ECOG 0, and 11.6% had diastolic dysfunction. Incidence of grade 2–4 FO was 33.7% (n = 32). Univariate analyses found age (continuous; p = 0.04) and BSA < 1.7 m² (p = 0.006) as independent factors associated with grade 2–4 FO. Multivariable regression analysis found 3.3% higher risk with every 1-year increase in age ranging from f 20 to 78 years (OR 1.033, 95% CI 1.001, 1.006; p = 0.0453) and 82.8% lower risk with BSA ≥ 1.7 m² (OR 0.172, 95% CI 0.051, 0.588; p = 0.005) after adjusting for co-variates.

Conclusion(s)

Increasing age and BSA < 1.7 m² are risk factors associated with grade 2–4 FO during hospitalization for allogeneic HCT with PTCY.

Introduction

Allogeneic hematopoietic cell transplantation (HCT) is a curative treatment for both malignant and nonmalignant hematologic disorders. However, treatment-related organ toxicities, infections, graft-versus-host disease, relapse, and other complications contribute to high morbidity and mortality rates. Fluid overload (FO) is one such early complication after allogeneic HCT. Rondon et al. found that an association exists between grade 2–4 FO and Day +100 non-relapse mortality (NRM) in allogeneic HCT recipients. This study also found a higher incidence of grade 2–4 FO after a HLA-haploidentical transplant with tacrolimus, mycophenolate mofetil (MMF), and post-transplant cyclophosphamide (PTCY), compared to those who received a HLA-matched related or unrelated allogeneic HCT with tacrolimus and methotrexate. Thus, this study established FO as an important early prognostic factor after allogeneic HCT, as well as an FO grading criteria outside of the Common Terminology Criteria for Adverse Events (CTCAE). ¹

Use of PTCY is increasing as a GVHD prevention strategy. Luznik et al demonstrated low incidences of acute and chronic GVHD with tacrolimus, MMF, and PTCY after HLA-haploidentical allogeneic HCT. ² A phase 3 trial demonstrated GVHD-free, relapse-free survival benefit with PTCY, tacrolimus and MMF compared to tacrolimus and methotrexate in reduced intensity HLA-matched or mismatched (7/8) peripheral blood allogeneic HCT. ³ Administration of PTCY requires aggressive hydration to prevent hemorrhagic cystitis (HC). This additional hydration during allogeneic HCT hospitalization presumably increases the risk of FO.

A common hyperhydration strategy is normal saline given over 48 h at a rate greater than 200 mL/hour. Due to the risk of FO, some institutions have adopted different strategies, such as body surface area (BSA) based hydration (i.e., normal saline 2.5–3 L/m²) with or without mesna or forced diuresis. ⁴ However, there is currently no standard approach to fluid management with PTCY and risk factors for grade 2–4 FO. The purpose of this study is to describe the incidence and identify factors associated with grade 2–4 FO during hospitalization for allogeneic HCT with PTCY.

Materials and methods

Study design

This was a single-center, retrospective, observational study conducted at an academic medical center. Data was collected via electronic chart review. This study was given exempt status by the University of California, Davis Institutional Review Board (IRB).

Patient population

All patients received an allogeneic HCT followed by PTCY 50 mg/kg IV on Day +3 and +4 for GVHD prevention. Haploidentical, matched, and mismatched transplants were included. For supportive care, patients received sodium chloride 0.9% IV continuous infusion at 200 mL/hour on Days +3 and +4 for 48 h. Mesna 10 mg/kg IV bolus was administered prior to each PTCY dose, followed by mesna 60 mg/kg continuous infusion for 3 days from Day +3 through Day +5. Patients received diuretics as needed based upon provider discretion. Patients were excluded if they were less than 18 years old or incarcerated.

Study variables

The primary endpoint was incidence of grade 2–4 FO after receiving PTCY for GVHD prevention. The secondary endpoints were incidences of HC and Day +100 NRM.

Clinical assessment

FO was measured from pre-admission (within 7 days prior to admission) through Day +10 and evaluated based on previously published grading criteria (Table 1). ¹ HC was diagnosed based on clinical findings including but not limited to presence of hematuria, urinalysis, and imaging. Diastolic and systolic functions were assessed based on cardiologist-reported echocardiography results.

Table 1.

Fluid overload grading criteria. ¹

Grade	Criteria
1	Weight gain less than 10% from baseline, asymptomatic or mild edema, possibly requiring diuretic therapy or a decrease in intravenous fluid replacement
2	Symptomatic fluid retention, with or without weight gain at least 10% and less than 20% from baseline, requiring ongoing diuretic therapy
3	Weight gain at least 20% from baseline, not responding to diuretic therapy, with possible renal, pulmonary, or cardiac dysfunction requiring further treatment
4	Progressive dysfunction of more than 1 organ system or requiring intensive care

Statistical analysis

The Fisher's Exact test was used for univariate analyses. Multivariable regression analysis was used to identify risk factors associated with grade 2–4 FO while adjusting for covariates. Risk factors assessed included age, sex, donor status, BSA, ECOG performance status, baseline albumin level, baseline diastolic function, baseline creatinine clearance, and baseline Left ventricular ejection fraction (LVEF).

Results

After screening 97 HCTs, 95 HCTs were included and two were excluded due to insufficient weight measurements needed to accurately grade FO. Four patients received two allogeneic HCT followed by PTCY after each HCT. Notably, the median age was 60 years old, 66.3% were male, 63.1% received an allogeneic HCT for acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS), 91.6% received a reduced intensity conditioning regimen, and 72.6% received an HLA-haploidentical allogeneic HCT. In addition, 11.6% had diastolic dysfunction prior to allogeneic HCT (Table 2).

Table 2.

Baseline characteristics.

Characteristics		n (%)
Age, year, median (range)		60 (20, 78)
Sex	Male	63 (66.3)
	Female	32 (33.7)
Diagnosis	AML/MDS	60 (63.1)
	ALL	7 (7.4)
	CML/MPN	13 (13.7)
	HL/NHL/CLL/PTCL	8 (8.4)
	Other	7 (7.4)
Conditioning regimen	Reduced intensity	87 (91.6)
	Myeloablative	8 (8.4)
Disease status	CR1	40 (42.1)
	CR2+	15 (15.8)
	Other	40 (42.1)
Donor Status	Haploidentical	69 (72.6)
	MRD	2 (2.1)
	MUD	19 (20)
	MMUD	5 (5.3)
ECOG	0	42 (44.2)
	1	42 (44.2)
	2	10 (10.5)
	3	1 (1.1)
Left ventricular ejection fraction (LVEF), %, median (range)		60 (30, 77)
Creatinine clearance (CrCL), mL/min, median (range)		96 (51, 227)
Body surface area (BSA), m2, median (range)		1.92 (1.37, 2.52)
Albumin, mg/dL, median (range)		3.8 (2, 4.8)

Of 95 HCTs, 33.7% (n = 32) resulted in grade 2–4 FO. Of these 32 HCTs, 4.2% (n = 4) and 3.2% (n = 3) resulted in grade 3 and 4 FO, respectively (Figure 1).

Figure 1.

Incidence of fluid overload. Distribution of allogeneic HCTs in different grade of FO.

After univariate analysis, age as a continuous variable (p = 0.04) and BSA less than 1.7 m² (p = 0.006) were independently associated with grade 2–4 FO (Table 3). Other risk factors assessed did not correlate with risk for grade 2–4 FO.

Table 3.

Univariate analysis of factors associated with grade 2–4 fluid overload.

Predictor		FO Grade 0–1 n (%)	FO Grade 2–4 n (%)	p-value*
Mean age ± SD, years		52.5 ± 17.0	59.3 ± 13.3	p = 0.04
Sex	Female	18 (19.0)	14 (14.7)	p = 0.17
	Male	45 (47.3)	18 (19.0)
Donor Status	Haploidentical	45 (47.3)	24 (25.3)	p = 0.81
	Non-haploidentical	18 (19.0)	8 (8.4)
BSA, m2	<1.7	5 (5.3)	10 (10.5)	p = 0.006
	≥1.7	58 (61.0)	22 (23.2)
ECOG	0	29 (30.5)	13 (13.7)	p = 0.67
	≥1	34 (35.8)	19 (20.0)
Albumin, mg/dL	<4	42 (44.2)	25 (26.3)	p = 0.34
	≥4	21 (22.1)	7 (7.3)
Diastolic Function	No dysfunction or indeterminate	57 (60.0)	27 (28.4)	p = 0.50
	Any dysfunction	6 (6.3)	5 (5.3)
CrCL, mL/min	<60	3 (3.2)	2 (2.1)	p = 1.00
	≥60	60 (63.2)	30 (32.5)
Mean LVEF ± SD, %		59.5 ± 6.1	60.9 ± 7.0	p = 0.34

*p < 0.05 is statistically significant.

Multivariable regression analysis found that every 1-year increase in age was associated with a 3.3% increased risk of developing grade 2–4 FO (OR 1.033, 95% CI 1.001, 1.006, p = 0.0453). Additionally, BSA of at least 1.7 m² was associated with an 82.8% decreased risk of developing grade 2–4 FO (OR 0.172, 95% CI 0.051, 0.588, p = 0.005).

Considering that the purpose of aggressive hydration is to prevent HC, the incidence of HC was 2.1% (n = 2/95) occurring in one patient with Grade 1 FO and another with Grade 3 FO. Notably, both patients who developed HC also had concomitant BK virus infection. Of 91 patients who received HCT, 13.2% (n = 12) experienced Day +100 NRM, 9 of which had grade 2–4 FO.

Discussion

This study provides insight into real-world outcomes regarding incidence of grade 2–4 FO in patients who receive PTCY after allogeneic HCT. Contrary to the classical use of a calcineurin inhibitor with methotrexate or MMF as GVHD prevention, PTCY necessitates hyperhydration and mesna to prevent HC. ⁵ The outcomes of our study highlight a need to find proactive and patient-tailored preventative strategies to mitigate risk of grade 2–4 FO in patients who received PTCY.

The incidence of grade 2–4 FO in this study (33.7%) was slightly higher than what was reported by Rondon et al. (21%), possibly due to the differences in hydration strategies between the institutions. While this study was not powered to show statistically significant differences in Day +100 NRM, incidence of Day +100 NRM was numerically higher in those who experienced grade 2–4 FO compared to no FO, which is consistent with existing published data. ¹ Of note, Rondon et al. found that a baseline creatinine of 1 mg/dL was associated with significantly higher Day +100 NRM. ¹ Our study did not further evaluate the effect of renal dysfunction on NRM as only 5 patients had CrCL less than 60 mL/min, likely due to institution-specific eligibility criteria for baseline renal function when determining transplant candidacy.

Our study further identifies increasing age and lower BSA as factors associated with grade 2–4 FO. These findings are similar to those of a previous study conducted in patients who received ex-vivo CD34-selected allogeneic HCT without pharmacotherapy for GVHD prevention that identified age at least 55 years old and body mass index (BMI) less than 25 kg/m² as factors associated with grade 2–4 FO. ⁶ Although age was evaluated as a binomial variable and weight as BMI rather than BSA, findings in this study are comparable with our study, such that older patients or those with lower weight are at higher risk of developing grade 2–4 FO.

While reducing risk of grade 2–4 FO is important to prevent post-transplant complications, the incidence of HC still needs to be considered when strategizing how to optimize fluid management in patients who receive PTCY. A retrospective study of patients who received PTCY after allogeneic HCT reported a 31.4% incidence of HC despite hyperhydration with sodium chloride 0.9% at 250 mL/hour, forced diuresis, and a daily mesna dose of 320% of the daily PTCY dose. ⁷ In our study, the HC rate was only 2.1% despite less aggressive hyperhydration, as needed diuresis, and a lower daily mesna dose of 170% of the daily PTCY dose. This highlights the question of whether weight-based hydration (i.e., 150 mL/hour) or proactive (i.e., forced) diuresis can be considered to decrease the risk of grade 2–4 FO, especially in patients with FO risk factors, while also maintaining low rates of HC.

To prevent HC, most centers use some variation of hyperhydration, diuresis, and mesna. For example, Bolaños-Meade et al. utilized a weight-based hyper-hydration protocol in the phase 3 trial evaluating PTCY, tacrolimus, and MMF as GVHD prevention. ³ However, the optimal combination of hydration, diuresis, and mesna to minimize FO without permitting HC has not been studied. Further studies are needed to establish evidence-based fluid management strategies for PTCY.

Our study identified patient-specific factors associated with grade 2–4 FO to better inform decision-making regarding optimal fluid management strategies in PTCY recipients. However, this study had a few limitations with a small sample size and single center experience. Also, it did not account for variation in fluids or diuretics administered.

Conclusion

Increasing age and lower body surface area are associated with Grade 2–4 FO in patients receiving PTCY. Proactive and patient-tailored fluid management strategies should be conducted in at-risk patients to decrease the risk of grade 2–4 FO which has been associated with higher day +100 NRM.