Abstract
Despite a well-established risk of chronic kidney disease (CKD) after allogeneic hematopoietic cell transplant (allo-HCT), the benefits of using nephrotoxic anti-infective agents to treat serious peritransplant infections often outweigh this risk. While there is no consensus on the optimal management of post-allo-HCT human herpes virus 6 (HHV6) reactivation, the nephrotoxic drug foscarnet is often used, although its long-term impact on renal function has not been established. We retrospectively reviewed 987 adult patients who underwent transplantation between 2002 and 2016, of whom 45.3% (n = 447) were exposed to foscarnet. The most frequent indications for foscarnet treatment were cytomegalovirus (n = 257, 57.5%) and HHV6 (n = 139, 31.1%). In the first 3 months post-transplant, patients exposed versus unexposed had similar rates of acute kidney injury and acute kidney failure (defined as 3 times baseline creatinine or <75% baseline estimated glomerular filtration rate [eGFR], 61.6% versus 58.7%, P =.42 and 28.1% versus 26.6%, P =.64, respectively). There was no difference in the eGFR at 3 months (P =.36), but patients treated with foscarnet had significantly lower median eGFRs (mL/min/1.73 m2) at 6 months (69.3, interquartile range [IQR] 51.4 to 92.8 versus 77.4, IQR 57.3 to 99.3; P =.009) and 12 months (67.8, IQR 52.7 to 85.0 versus 80.7, IQR 63.1 to 102.0; P < .001), respectively. There was also a significant difference in the decline in eGFR from baseline to12 months (median 32.8, IQR 14.6 to 53.2 versus 21.9, IQR6.4 to 37.4; P < .001), irrespective of the duration of foscarnet treatment. Multivariate analysis revealed that patients treated with foscarnet were more likely to experience a >30% decrease in eGFR from baseline to 12 months compared to those who were not (odds ratio, 2.30; 95% CI, 1.40 to 3.78; P =.001). We conclude that foscarnet use following allo-HCT had a profound impact on long-term renal function independent of other transplant-related factors.
Keywords: Hematopoietic stem cell, transplantation, Adult, Herpesvirus 6, Foscarnet, Renal insufficiency, Chronic
With improvement in supportive care for patients undergoing allogeneic hematopoietic cell transplantation (allo-HCT), the proportion of long-term survivors is rising. In addition, patients presently undergoing allo-HCT are older and more medically complex than in the past. It is becoming more important to address comorbid conditions and prevent complications in the peritransplant period that may have implications for long-term health and survival. For instance, more data are emerging about the effects of acute and chronic kidney injury in this population and their associations with morbidity and mortality [1].
A significant clinical dilemma in this population is the monitoring and treatment of human herpes virus 6 (HHV6) reactivation. HHV6 reactivation occurs frequently after allo-HCT and can be associated with myelosuppression, pneumonitis, hepatitis, and, most important, a life-threatening limbic encephalitis [2]. It is also associated with increased all-cause mortality [3]. The antivirals ganciclovir and foscarnet are both active against HHV6. However, ganciclovir has not been shown to improve survival or prevent encephalitis [4]. In contrast, retrospective data suggested that treatment with foscarnet is associated with a lower incidence of early death from HHV6 encephalitis [4].
Given the toxicity associated with foscarnet, guideline statements have not recommended for or against routine use of antivirals for HHV6 viremia without central nervous system symptoms [5]. Foscarnet was initially approved by the US Food and Drug Administration in 1991 to treat resistant cytomegalovirus (CMV) retinitis in terminally ill patients with AIDS, carrying a black box warning of renal failure. Despite this well-known risk of acute nephrotoxicity, foscarnet is often used in the allogeneic HCT setting since it is not associated with myelosuppression, as is the case with ganciclovir.
With a large population of patients at risk for complications from HHV6 reactivation following hematopoietic stem cell transplantation, our center has a longstanding practice of preemptive therapy using foscarnet. Because of the absence of published data describing the long-term impact of foscarnet on renal function after HCT, we performed a retrospective study to address this question.
PATIENTS AND METHODS
Data Collection
All patients receiving allo-HCT at Duke were identified using the Duke Enterprise Data Unified Content Explorer Portal [6], a web-based data query tool in the Duke Maestro Care (EPIC) Electronic Health Record for clinical research; Data Back to Centers tool from the Center for International Blood and Marrow Transplant Research Database; and an in-house Microsoft Access database. Institutional data were also abstracted by manual chart review. Collected data included serum creatinine (Cr), age, sex, race, diagnosis at transplant, stage and disease status at transplant, American Society for Blood and Marrow Transplantation (ASBMT) disease risk classification [7], graft source, donor type, degree of HLA match, transplant year, conditioning regimen, recipient and donor CMV status, engraftment date, exposure to calcineurin inhibitors for graft-versus-host disease (GVHD) prophylaxis, hematopoietic cell transplantation-specific comorbidity index [8], exposure to alemtuzumab and post-transplant cyclophosphamide, presence and grade of GVHD, date of death, and date of last clinical contact.
Foscarnet exposure, duration of exposure, and indication for treatment were ascertained using both manual chart review and medication use evaluation universe reports generated from the Duke Health Maestro Care Business Objects database. A medication use evaluation is a performance improvement tool that can analyze the process of medication prescribing, preparation, dispensing, administration, and monitoring [9]. HHV6 viral loads were obtained directly from the Viracor Clinical Diagnostic Laboratory.
Criteria for Eligibility
Inclusion criteria were age ≥18 years and receipt of an allogeneic transplant from June 2002 to February 2016 at Duke University Medical Center. Patients were excluded if they had undergone syngeneic transplant or if they did not have available Cr values or data on foscarnet exposure. In patients who underwent multiple transplants, only the first transplant was included in the analysis. The protocol was approved by the Duke University Medical Center Institutional Review Board.
Definitions
Estimated glomerular filtration rate (eGFR) was calculated based on the CKD-EPI formula [10]. Baseline eGFR was calculated using serum Cr measured prior to transplant. The Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease [11] criteria were used to define acute kidney injury (AKI) and acute kidney failure as a Cr increase of at least 2 and 3 times the baseline Cr, or >50% and >75% decrease in eGFR, respectively. eGFR at 3 months was calculated using the median of all creatinine measures taken within 15 days of day 90, eGFR at 6 months was calculated using the median of all creatinine measures taken within 45 days of day 180, and eGFR at 12 months was defined as the median of all measures taken within 45 days of day 365. Acute and chronic GVHD were defined as previously described in the National Institutes of Health consensus criteria and as reported to the Center for International Blood and Marrow Transplant Research [12].
Viral Surveillance and Foscarnet Treatment
HHV6 viremia was assessed by quantitative HHV6 DNA PCR weekly for the first 3 months in patients receiving umbilical cord blood transplantation or those who received alemtuzumab as part of the conditioning regimen. Viral loads were checked as clinically indicated (eg, cytopenias, encephalopathy) in all others. CMV, herpes simplex virus (HSV), and varicella zoster virus viral loads were not collected.
Foscarnet treatment for HHV6 reactivation was at the discretion of the managing physician. Factors considered in the decision to initiate treatment and treatment duration included donor source (matched versus mismatched), presence of cytopenias, presence of fever, and days after allo-HCT. Some patients were only treated with induction without maintenance (Figure 3). Rapid clearance of HHV6 viremia was often factored into this decision.
Figure 3.
Change in eGFR(mL/min/1.73 m2) at 12 months by duration of treatment. Dots indicate the median change in eGFR from baseline to 12 months. Bars indicate first and third quartiles of change in eGFR from baseline to 12 months.
In patients with Cockcroft-Gault creatinine clearance >1.4 mL/min/kg, foscarnet induction dosing was 90 mg/kg i.v. every 12 hours for 14 days and maintenance dosing was 90 mg/kg i.v. every 24 hours for 14 days or longer, based on the clinical scenario. Dosing was adjusted for renal function according to the foscarnet package insert [13].
Group Assignment
Inclusion in the “foscarnet-exposed” group was defined by the latest treatment dose within 45 days of the specified follow-up time point. Only patients treated before each follow-up time point were included in the associated analyses as foscarnet exposed. As such, for the 6-month analysis, the foscarnet-exposed group received foscarnet between 0 and 6 months following transplantation. For the 12-month analysis, the foscarnet-exposed group received foscarnet between 0 and 12 months following transplantation. In an attempt to screen out patients who were experiencing acute kidney injury as a consequence of foscarnet exposure, patients were excluded from the 6-month and 12-month analyses if foscarnet had been administered within the preceding 45 days for the 6-month analysis and within the preceding 90 days for the 12-month analysis. Descriptive characteristics and overall survival were based on foscarnet exposure any time after transplant.
Statistical Analysis
Patient characteristics were summarized with number (%) for categorical variables and median (interquartile range [IQR]) or mean (SD) for continuous variables. Chi-square or Fisher exact tests and Student t tests were used to test for differences between patients treated with and without foscarnet in categorical and continuous variables, respectively. Within-group difference of Cr and eGFR was tested using the Wilcoxon signed rank test, and between-group differences were tested with the Wilcoxon rank-sum test.
All patients included in the final analysis had baseline creatinine values, 833 had 3-month values, 680 had 6-month values, and 480 had 12-month values. Only patients with available data at each follow-up time point were included in the associated analyses.
Overall survival was defined as the time from transplant to death due to any cause. Patients who did not die were censored at date of last follow-up. Kaplan-Meier curves were used to visualize differences in survival, and the log-rank test was used to test for difference for treated versus untreated with foscarnet. A Cox proportional hazards model was used to estimate the association of treatment with foscarnet with overall survival after adjustment for age at transplant (years), acute GVHD grades II to IV, conditioning regimen, year of transplant, graft source, HHV6 and CMV status, and ASBMT disease risk. Hazard ratios (HRs) and 95% confidence intervals (CIs) are reported for this model.
Multivariate logistic regression models were used to estimate the association of treatment with foscarnet and the probability of a patient experiencing a decrease in eGFR from baseline greater than 30% at 6 months and 12 months. These models were adjusted for age at transplant, acute GVHD grades II to IV, chronic GVHD, race, conditioning regimen, donor type, treatment with calcineurin inhibitors, and HHV6 status. Odds ratios (ORs) and 95% CIs are reported for these models. A multivariate linear regression model was used to estimate the association of treatment with foscarnet on the percent change in eGFR from baseline to 12 months, after adjustment for baseline eGFR, age at transplant, acute GVHD grades II to IV, conditioning regimen, donor type, treatment with calcineurin inhibitors, and HHV6 status. Estimated effect size and 95% CIs are reported for this model.
All statistical analyses were conducted with SAS, version 9.4 (SAS Institute, Cary, NC). No adjustments were made for multiple comparisons.
RESULTS
Patients
There were 989 patients who were eligible for the study. Two patients were excluded for lack of patient-level data. Of the 987 included patients, 45.4% (n = 447) were treated with foscarnet (Figure 1). Patients treated with foscarnet were older with a median age of 52 versus 49 years, more likely to be CMV positive (72.5%, n = 324 versus 49.6%, n = 268), and more likely to receive an alternative donor graft (umbilical cord blood 25.5%, n = 114 versus 10.2%, n = 55; haploidentical 17.4%, n = 78 versus 6.7%, n = 36). Patients treated with foscarnet were less likely to receive grafts from matched related (21.9%, n = 98 versus 42.4%, n = 229) and matched unrelated donors (30.4%, n = 136 versus 36.9%, n = 199), and foscarnet-exposed patients were less likely to receive myeloablative conditioning (42.3%, n = 189 versus 59.4%, n = 321). Foscarnet-exposed patients were more likely to be African American (16.8%, n = 75 versus 12.0%, n = 65). ASBMT disease risk categories differed between foscarnet-exposed versus unexposed patients as well (Table 1).
Figure 1.
Flowchart of the study and analysis. Only patients with available data at each follow-up time point were included in the associated analyses. Foscarnet exposure was defined as exposure prior to the specified follow-up time point.
Table 1.
Patient Characteristics
| Characteristic | All Patients (N = 987) | Treated with Foscarnet | P Value | |
|---|---|---|---|---|
| No (n = 540) | Yes (n = 447) | |||
| Age, median (IQR), yr | 50(39–58) | 49(38–57) | 52(39–60) | .01 |
| Weight, median (IQR), kg | 82.5(70–94.8) | 83.1 (70.2–96.4) | 81.8(70–93.8) | .45 |
| Year of transplant, median (IQR) | 2010(2006–2013) | 2009(2006–2013) | 2010(2007–2013) | .005 |
| Sex, No. (%) | .09 | |||
| Female | 419(42.5) | 216(40) | 203 (45.4) | |
| Male | 568 (57.5) | 324(60) | 244 (54.6) | |
| Race, No. (%) | .03 | |||
| White | 831 (84.2) | 469 (86.9) | 362(81) | |
| Black | 140(14.2) | 65(12) | 75(16.8) | |
| Other | 16(1.6) | 6(1.1) | 10(2.2) | |
| Patient CMV, No. (%) | <.001 | |||
| Negative | 337(34.1) | 236(43.7) | 101 (22.6) | |
| Positive | 592(60) | 268 (49.6) | 324 (72.5) | |
| Unknown/indeterminate | 51 (5.2) | 32(5.9) | 19(4.3) | |
| Graft source, No. (%) | ||||
| BM | 59(6) | 40(7.4) | 19(4.3) | .04 |
| Umbilical cord blood | 169(17.1) | 55(10.2) | 114(25.5) | <.001 |
| PBPC | 759(76.9) | 445 (82.4) | 314(70.2) | <.001 |
| Donor type, No. (%) | ||||
| Haploidentical | 114(11.6) | 36(6.7) | 78(17.4) | <.001 |
| Matched related | 327(33.1) | 229 (42.4) | 98 (21.9) | <.001 |
| Matched unrelated | 335(33.9) | 199(36.9) | 136 (30.4) | .03 |
| Mismatched related | 25(2.5) | 12(2.2) | 13 (2.9) | .49 |
| Mismatched unrelated | 17(1.7) | 9(1.7) | 8(1.8) | .88 |
| Unrelated umbilical cord blood | 169(17.1) | 55(10.2) | 114(25.5) | <.001 |
| Conditioning regimen, No. (%) | <.001 | |||
| Myeloablative | 510(51.7) | 321 (59.4) | 189 (42.3) | |
| Nonmyeloablative/reduced intensity | 477 (48.3) | 219(40.6) | 258 (57.7) | |
| ASBMT disease risk, No. (%) | .05 | |||
| Low | 335(33.9) | 192(35.6) | 143 (32) | |
| Intermediate | 303 (30.7) | 146(27) | 157 (35.1) | |
| High | 270 (27.4) | 154(28.5) | 116(26) | |
| Unknown | 79(8) | 48 (8.9) | 31 (6.9) | |
| GVHD, No. (%) | ||||
| Acute grades II-IV | 476 (48.2) | 218(40.4) | 258 (57.7) | <.001 |
| Acute grade III | 199(20.2) | 86(15.9) | 113(25.3) | <.001 |
| Acute grade IV | 39(4) | 15(2.8) | 24 (5.4) | .04 |
| Chronic | 297(30.1) | 168(31.1) | 129 (28.9) | .44 |
| GVHD prophylaxis, No. (%) | ||||
| Alemtuzumab | 375(38) | 150(27.8) | 225 (50.3) | <.001 |
| Calcineurin inhibitors | 684(69.3) | 431 (79.8) | 253 (56.6) | <.001 |
| With post-transplant cyclophosphamide | 12(6.2) | 4(3.5) | 8 (9.8) | - |
| HCT-CI, median (IQR)* | 3(2–4) | 3(2–4) | 3(3–4) | .005 |
| HHV6, No. (%) | <.001 | |||
| HHV6+ | 277(28.1) | 62(11.5) | 215(48.1) | |
| HHV6− | 710(71.9) | 478(88.5) | 232 (51.9) | |
| Indication for foscarnet, No. (%) | - | |||
| CMV | 257(26) | - | 257 (57.5) | |
| HHV6 | 139(14.1) | - | 139 (31.1) | |
| HSV | 9(0.9) | - | 9(2) | |
| Varicella zoster virus | 2(0.2) | - | 2(0.4) | |
| Empiric | 12(1.2) | - | 12(2.7) | |
| CMV and HSV | 2(0.2) | - | 2(0.4) | |
| CMV and HHV6 | 23 (2.3) | - | 23 (5.1) | |
| HHV6 and HSV | 2(0.2) | - | 2(0.4) | |
| CMV, HHV6, and HSV | 1(0.1) | - | 1 (0.2) | |
BM indicates bone marrow; PBPC, XXX; HCT-CI, hematopoietic cell transplantation-specific comorbidity index; —, XXX.
HCT-CI was available for 595 patients.
While more patients treated with foscarnet experienced acute GVHD (grades II to IV: 57.7%, n = 258 versus 40.4%, n = 218), rates of chronic GVHD were similar to foscarnet-unexposed patients (P =.44). Patients treated with foscarnet were also more likely to have had GVHD prophylaxis in the form of alemtuzumab (50.3%, n = 225 versus 27.8%, n = 150) and less likely to have been treated with calcineurin inhibitors for GVHD prophylaxis (56.6%, n = 253 versus 79.8%, n = 431).
Acute and Long-Term Renal Injury
Patients in the group treated with foscarnet had a slightly lower median IQR baseline eGFR (mL/min/1.73 m2) at the time of transplant (99.9, IQR 79.3 to 117.5 versus 104.6, IQR 88.9 to 119.3; P =.01) (Table 2). Most patients, irrespective of foscarnet exposure, experienced a decline in renal function during allo-HCT (Figure 2). In patients treated with foscarnet (versus not) within 3 months post-transplant, the incidence of AKI (61.6%, n = 149 versus 58.7%, n = 437) and acute kidney failure (28.1%, n = 68 versus 26.6%, n = 198) were not statistically different (P = .42 and 0.64, respectively) (Supplementary Table S1). Patients treated with foscarnet versus not had a similar decrement in eGFR (mL/min/1.73 m2) from baseline to 3 months (−21.3, IQR −50.6 to 0.0 versus −24.3, IQR −44.7 to 5.4; P = .70) and baseline to 6 months (−23.5, IQR −48.3 to 6.7 versus −21.1, IQR −42.7 to 3.2; P = .13). However, between the 6- and 12-month time points, foscarnet-exposed patients’ median eGFR continued to decline, while unexposed patients’ eGFR improved (median at 12 months 67.8 [IQR 52.7 to 85.0] versus 80.7 [IQR 63.1 to 102.0]; P < .001) (Table 2, Figure 2). This resulted in a significant difference in the decline in eGFR from baseline to 12 months with a median change of −32.8 (IQR −53.2 to −14.6) in the foscarnet-exposed group compared to a median change of −21.9 (IQR −37.4 to −6.4; P < . 001) for the unexposed group (Table 2).
Table 2.
eGFR (mL/min/1.73 m2) in Foscarnet-Exposed versus Unexposed Patients
| Characteristic | No Foscarnet (3 Months, n = 631; 6 Months, n = 493; 12 Months, n = 313) | Foscarnet (3 Months, n = 202; 6 Months, n = 187; 12 Months, n = 167) | P Value* |
|---|---|---|---|
| Baseline | 104.6 (88.9–119.3) | 99.9(79.3–117.5) | .01 |
| 3 months | 72.7 (55.8–96.4) | 71.6 (50.7–95.7) | .36 |
| Δ baseline to 3 months | −24.3 (−44.7 to 5.4) | −21.3 (−50.6 to 0) | .70 |
| 6 months | 77.4 (57.3 to 99.3) | 69.3 (51.4 to 92.8) | .009 |
| Δ baseline to 6 months | −21.1 (−42.7 to 3.2) | −23.5 (−48.3 to 6.7) | .13 |
| 12 months | 80.7 (63.1 to 102.0) | 67.8 (52.7 to 85.0) | <.001 |
| Δ baseline to 12 months | −21.9 (−37.4 to 6.4) | −32.8 (−53.2 to 14.6) | <.001 |
Values are presented as median (IQR).
Wilcoxon rank-sum test P values.
Figure 2.
Change in eGFR over time by exposure to foscarnet. Histograms depicting eGFR distribution of patients exposed versus unexposed to foscarnet. The dotted red line represents the baseline median eGFR in both groups.
Foscarnet Treatment
The median time to foscarnet exposure was 90 days (IQR 55 to 199). There was no statistical difference in eGFR decline 1 year following transplantation among patients treated with foscarnet for <2 weeks, 2 to 4 weeks, or >4 weeks (Figure 3).
The most frequent indications for foscarnet treatment were CMV (n = 257, 57.5%) and HHV6 viremia (n = 139, 31.1%) (Table 1). In patients with HHV6 reactivation, the median HHV6 viral load present at the time of foscarnet initiation was 500 copies/mL (IQR188 to 3300). Within the foscarnet-exposed group, at 12 months, HHV6-positive patients had a significantly greater decline in median eGFR (mL/min/1.73 m2) from baseline than HHV6-negative patients (−41.4, IQR −57.1 to 24.0 versus −23.4, IQR −44.8 to 10.2; P = .004) (Supplementary Table S2).
Predictors of Long-Term Decline in Renal Function
Multivariate analysis identified foscarnet exposure (OR, 2.30; 95% CI, 1.40 to 3.78; P = .001), myeloablative conditioning (OR, 2.29; 95% CI, 1.23 to 4.25; P =.009), and matched unrelated donor versus matched related graft source (OR, 1.86; 95% CI, 1.17–2.97; P = .009) as the only statistically significant predictors of >30% decline in eGFR from baseline to 12 months following allo-HCT (Table 3). Other factors previously described to contribute to renal decline, including age (P = .25), black race (P = .99), acute and chronic GVHD (P = .15 and .06, respectively), and treatment with calcineurin inhibitors (P = .87), were not statistically significant predictors of >30% decline in eGFR at 12 months [1,14–17]. Baseline eGFR (OR, 1.01; 95% CI, 1.00 to 1.03; P =.01), age at transplant (OR, 1.03; 95% CI, 1.01 to 1.05; P =.004), haploidentical (versus matched related) graft source (OR, 2.75; 95% CI, 1.26 to 5.98; P =.01), umbilical cord blood versus matched related graft source (OR, 2.11; 95% CI, 1.13 to 3.92; P = .02), and treatment with calcineurin inhibitors (OR, 2.10; 95% CI, 1.12 to 3.94; P =.02) were predictors of >30% decline in eGFR at 6 months but not at 12 months following hematopoietic stem cell transplantation (Table 3). HHV6-positive status was a statistically significant negative predictor of >30% eGFR decline at 6 months (OR, 0.62; 95% CI, 0.40 to 0.98; P = .04) but not at 12 months (1.24; 95%CI,0.76 to 2.03; P =.40).
Table 3.
Adjusted Logistic Model Predicting a Decrease in eGFR of >30%
| Predictor | 6Months (n = 680) | 12Months(n = 480) | ||
|---|---|---|---|---|
| Odds Ratio (95% CI) | P Value | Odds Ratio (95% CI) | P Value | |
| Baseline eGFR (per mL/min/1.73 m2) | 1.01 (1.00–1.03) | .01 | 1.00(0.99–1.02) | .54 |
| Foscarnet exposure | 1.88(1.19–2.96) | .007 | 2.30(1.40–3.78) | .001 |
| HHV6+ | 0.62 (0.40–0.98) | .04 | 1.24(0.76–2.03) | .40 |
| Age at transplant | 1.03(1.01–1.05) | .004 | 1.01 (0.99–1.04) | .25 |
| Race (versus white) | ||||
| Black | 0.88(0.51–1.53) | .65 | 1.00(0.54–1.83) | .99 |
| Other | 0.60(0.14–2.61) | .49 | 0.53(0.13–2.10) | .36 |
| Acute GVHD grades II-IV | 0.92(0.63–1.33) | .65 | 1.37(0.89–2.09) | .15 |
| Chronic GVHD | 1.39(0.96–2.01) | .08 | 1.51 (0.99–2.30) | .06 |
| Myeloablative conditioning (versus non-myeloablative) | 3.08(1.77–5.37) | <.001 | 2.29(1.23–4.25) | .009 |
| Donor type (versus matched related) | ||||
| Haploidentical | 2.75(1.26–5.98) | .01 | 0.87(0.33–2.31) | .78 |
| Matched unrelated | 2.90(1.90–4.46) | <.001 | 1.86(1.17–2.97) | .009 |
| Mismatched related | 1.59(0.56–4.53) | .38 | 0.38(0.08–1.83) | .23 |
| Mismatched unrelated | 0.81 (0.19–3.42) | .78 | 2.87(0.71–11.51) | .14 |
| Umbilical cord blood | 2.11 (1.13–3.92) | .02 | 1.58(0.78–3.18) | .20 |
| Treatment with calcineurin inhibitors | 2.10(1.12–3.94) | .02 | 1.06 (0.53–2.13) | .87 |
An adjusted linear model was used to estimate the magnitude of eGFR decline attributable to each of the relevant covariates at 12 months post-HCT. Treatment with foscarnet was associated with a statistically significant 10% (95% CI, −15% to −6%; P < .001) decrement in eGFR from baseline to 12 months, compared to no foscarnet treatment. In comparison, myeloablative conditioning contributed a 11% (95% CI, −17% to −5%; P < .001) decrement, while chronic GVHD and matched unrelated versus matched related donor source each contributed a 5% decline (P = .01 and P = .03, respectively) (Table 4).
Table 4.
Adjusted Linear Model Predicting Percent Change in eGFR at 12 Months (n = 480)
| Covariate | Estimate (95% CI) | P Value |
|---|---|---|
| Baseline eGFR (per mL/min/1.73 m2) | −0.002 (−0.003 to 0.001) | <.001 |
| Treatment with foscarnet | −0.10(−0.15to 0.06) | <.001 |
| HHV6+ | −0.03 (−0.08 to 0.02) | .28 |
| Age at transplant (yr) | −0.003 (−0.005 to 0.001) | .008 |
| Race (versus white) | ||
| Black | 0.04 (−0.02 to 0.10) | .17 |
| Other | 0.06 (−0.08 to 0.19) | .40 |
| Acute GVHD grades II-IV | −0.03 (−0.07 to 0.01) | .16 |
| Chronic GVHD | −0.05 (−0.09 to 0.01) | .01 |
| Conditioning regimen (versus nonmyeloablative) | ||
| Myeloablative | −0.11 (−0.17 to 0.05) | <.001 |
| Donor type (versus matched related) |
||
| Haploidentical | −0.01 (−0.10 to 0.07) | .78 |
| Matched unrelated | −0.05 (−0.10 to 0.005) | .03 |
| Mismatched related | 0.03 (−0.09 to 0.15) | .57 |
| Mismatched unrelated | −0.09 (−0.23 to 0.04) | .17 |
| Umbilical cord blood | −0.02 (−0.09 to 0.05) | .65 |
| Treatment with calcineurin inhibitors | 0.02 (−0.05 to 0.08) | .62 |
Overall Survival
With a median follow-up of 71.8 months (95% CI, 69.3 to 74.1), the unadjusted 1-year survival rates (95% CI) for the foscarnet-unexposed versus foscarnet-exposed patients were 59.3% (55.0% to 63.3%) versus 49.9% (45.2% to 54.4%), respectively (log-rank P =.003) (Figure 4). Improved survival for foscarnet-unexposed patients was maintained at the 5-year timepoint (37.0%, range 32.7% to 41.2% versus 28.0%, range 23.7% to 32.4%). However, in the multivariate analysis of risk factors for survival, only older age at transplant (as a continuous variable) (HR, 1.01; 95% CI, 1.01 to 1.02; P < .001), ASBMT high disease risk category (compared to low risk) (HR, 1.49; 95% CI, 1.23 to 1.81; P < .001), and acute GVHD (grade II to IV) (HR, 1.25; 95% CI, 1.06 to 1.47; P =.009) were associated with statistically significant inferior overall survival. Later year of transplant (as a continuous variable) and chronic GVHD (HR, 0.39; 95% CI, 0.32 to 0.47; P < .001) were significantly associated with a more favorable overall survival (Table 5). Foscarnet exposure was not significantly associated with survival.
Figure 4.
Overall survival (n = 987). Kaplan-Meier estimates of unadjusted overall survival in patients who were exposed versus unexposed to foscarnet at any time after transplant.
Table 5.
Adjusted Overall Survival (n = 980)*
| Covariate | Hazard Ratio (95% CI) | P Value |
|---|---|---|
| Any foscarnet exposure | 1.16(0.97–1.39) | .12 |
| HHV6+ | 1.15(0.92–1.44) | .23 |
| Age at transplant (yr) | 1.01 (1.01–1.02) | <.001 |
| Race (versus white) | ||
| Black | 1.01 (0.81–1.27)) | .91 |
| Other | 0.59(0.30–1.15) | .12 |
| Acute GVHD grades II-IV | 1.25(1.06–1.47) | .009 |
| Chronic GVHD | 0.39(0.32–0.47) | <.001 |
| Myeloablative conditioning (versus nonmyeloablative) | 1.01 (0.84–1.21) | .94 |
| ASBMT disease risk (versus low) | ||
| Intermediate | 1.09(0.89–1.33) | .42 |
| High | 1.49(1.23–1.81) | <.001 |
| Unknown | 0.79(0.55–1.12) | .18 |
| Patient CMV status (versus negative) | ||
| Positive | 0.96(0.81–1.14) | .66 |
| Unknown | 0.84(0.57–1.25) | .39 |
| Graft source (versus PBPC) | ||
| BM | 0.96(0.65–1.42) | .84 |
| Cord | 0.98(0.77–1.25) | .86 |
| Year of transplant | 0.95 (0.93–0.97) | <.001 |
Seven patients were excluded from this analysis due to missing values for 1 or more values.
DISCUSSION
As previously described, we found that most allo-HCT recipients experience a significant decline in renal function during the first 3 to 6 months. At 6 months, we observed a nearly 22% median decline in renal function from baseline. The novel finding in this study is that after the 6-month time point, we observed a small recovery of renal function in patients not treated with foscarnet. Conversely, patients who were treated with foscarnet during the peritransplant period, irrespective of the duration of therapy, continued to experience a decline in renal function between months 6 and 12. This sustained decline in renal function in the foscarnet group suggests that these patients experienced irreversible nephrotoxicity in line with the drug’s suspected mechanism of tubular injury [18]. In the unexposed group, the relative rebound may reflect other acute nephrotoxicities experienced by these patients early in the post-transplant course that are, to some extent, reversible.
Foscarnet is a known nephrotoxic medication that is used frequently for treatment of herpesvirus infections following allo-HCT. Efficacy for the treatment of ganciclovir-resistant CMV and acyclovir-resistant HSV are clear. Foscarnet does not cause bone marrow suppression and is thus well suited to treat patients in the periengraftment period or those with severe pancytopenia. However, uncertainty surrounding the clinical implications of post-transplant HHV6 reactivation has led to considerable variability in its use among stem cell transplant centers for this indication. Our institution has historically taken an aggressive approach to treating HHV6 viremia with foscarnet to prevent marrow suppression and encephalitis in patients known to be at high risk for these complications. This allowed for a large retrospective study of the impact of foscarnet therapy on long-term renal injury, something that has not previously been investigated in the allo-HCT population.
While others have reported an association between foscarnet use and renal dysfunction, this is the first large study to convincingly link foscarnet to long-term renal impairment following allo-HCT [15,16,18–22]. In an Asian population of 216 adults who had received HCT in Singapore, duration of foscarnet was shown to be significantly associated with chronic kidney disease (CKD) development (defined by an eGFR <60 mL/min/1.73 m2 for more than 3 months), but it did not emerge as a significant independent predictor of CKD in a multivariate analysis [19]. In another cohort of 39 patients treated with foscarnet (22 solid organ transplant and 17 HCT) for ganciclovir-resistant CMV, over 50% (29/39) of patients had renal dysfunction by the end of foscarnet treatment and 28% (7/25) of patients had persistent renal impairment after 6 months [20]. These studies were limited by sample size and duration of follow-up but certainly allude to a negative effect of foscarnet on long-term renal function in HCT patients.
In our study, the percentage of patients experiencing AKI was similar between the foscarnet-exposed and unexposed cohorts. In each group, greater than 50% of patients experienced an episode of AKI in the first 3 months post-transplantation; this is consistent with known rates of AKI in allo-HCT recipients [14]. Treatment with foscarnet is most commonly initiated within the first 3 months after HCT, as HHV6 and CMV reactivation are early complications [21,22]. Since the incidence of AKI in the first 3 months was similar between groups, this endpoint clearly does not capture the progressive decline in renal function related to foscarnet exposure. To assess long-term CKD risks, we chose to look at a >30% decline in eGFR as this outcome is strongly associated with a 10-year risk of end-stage renal disease (>60%) and mortality (50%) in the general population [23]. It is also a better way to measure long-term renal function decline relative to a patient’s individual baseline. We found that foscarnet exposure more than doubles the odds of >30% decline in GFR at 1 year. Moreover, foscarnet exposure is associated with a 10% incremental decline in eGFR at 12 months. In comparison, myeloablative conditioning was associated with an 11% decline in eGFR at 12 months while chronic GVHD and matched unrelated graft source were associated with a 5% decline each. In this population, declines in eGFR have been associated with increased all-cause mortality, with the hazard of death increasing significantly as the eGFR approached 60 mL/min/1.73 m2 [1]. We expect these variables to be additive in their effects on renal function. In our analysis, however, foscarnet exposure did not emerge as associated with mortality after adjustment for other covariates.
Our multivariate model identified foscarnet exposure as a strong predictor of >30% eGFR decline from baseline to 12 months. In addition, our data confirm previously described predictors in the HCT population, including chronic GVHD, myeloablative conditioning, and matched unrelated donor source [1,14,15,24,25]. Total body irradiation has been shown in several studies to be a risk factor for renal injury, but we identified myeloablative conditioning, in general, as a strong predictor [26,27]. Interestingly, we found that calcineurin inhibitor use for GVHD prophylaxis was associated with >30% decline in eGFR at 6 months but not at 12 months. This suggests more of a reversible injury related to early calcineurin inhibitor (CNI) exposure that does not persist at 12 months. CNIs are known to cause renal injury through a variety of acute and chronic mechanisms. Some acute effects are mediated by hemodynamic changes to renal blood flow and are potentially reversible. Other effects lead to chronic, irreversible damage through immune- and vascular-mediated glomerulosclerosis, tubular atrophy, and interstitial fibrosis [28]. Some studies have shown that the number of episodes of toxic CNI levels as well as CNI use >6 months in this population is linked to CKD [15,16]. However, other studies have not shown long-term decline in eGFR to be related to CNI use in general [1,14].
We were surprised to find that foscarnet-exposed patients who were HHV6 positive had nearly double the decline in eGFR from baseline to 12 months than those who were HHV6 negative (Supplementary Table S2). However, given the heterogeneity of factors driving individual clinicians’ decision to treat with foscarnet, it is difficult to interpret how much of this difference can be attributed to treating HHV6 reactivation per se. Importantly, HHV6 positivity was not a statistically significant predictor of >30% decrease in eGFR at 12 months in the adjusted logistic model (OR, 1.24; 95% CI, 0.76 to 2.03; P = .40) (Table 3).
Our study has several limitations, especially pertaining to its retrospective design. The foscarnet-exposed and unexposed groups differed in important ways, including graft source, donor type, conditioning regimen, type of GVHD prophylaxis, and incidence of acute and chronic GVHD, among others. Associations were adjusted for these differences with logistic and linear regression multivariate analyses. Our population was at high risk for HHV6 and CMV reactivation. Fifty-five percent of the patients had either alemtuzumab as part of their bone marrow conditioning regimen or received an umbilical cord blood graft. Both approaches are known to delay cellular immune recovery, increasing the risk for viral reactivation and therefore increasing the risk for foscarnet exposure compared to other transplant centers. Since most haploidentical transplant recipients received in vivo T cell depletion with alemtuzumab during the study period, there were too few patients treated with post-transplant cyclophosphamide to make any meaningful statements about the associated risk of renal dysfunction. Finally, we were unable to compare the incidence of end-stage renal disease in our cohorts due to the paucity of long-term follow-up data. However, given the strong and consistent association in the literature between a 30% reduction in GFR over 2 years and the risk of end-stage renal disease and mortality [20], increasing the risk of this outcome in only 12 months is a tougher standard and portends an even worse decline in renal function at 24 months.
These data suggest that treatment with foscarnet is a previously undescribed independent predictor of long-term decline in renal function after allogeneic stem cell transplant [29]. In contrast to other nephrotoxins used in the early post-transplant period, foscarnet’s impact on renal function is irreversible. This irreversible injury only becomes apparent many months following allo-HCT.
Overall, our data suggest that foscarnet treatment should be reconsidered for indications (such as HHV6-viral reactivation) that do not have compelling prospective data [5,30], demonstrating its benefit outweighing this important long-term risk.
Supplementary Material
ACKNOWLEDGMENTS
Gena G. Foster and Michael J. Grant contributed equally to this work.
Financial disclosure: This research was funded in part through the Duke Resident Research Grant (G.G.F. and M.J.G.) and the Duke Cancer Institute through National Institutes of Health grant P30CA014236 (SMT).
Authorship statement: G.G.F. and M.J.G. designed the study, acquired the data, interpreted the data, and wrote the manuscript. M.H. designed the study, interpreted the data, and wrote the manuscript. S.M.T. designed the study, performed the statistical analysis, interpreted the data, and wrote the manuscript. D.R., K.C., C.F., and G.L. obtained the data. B.C. interpreted the data and critically reviewed the manuscript.
Footnotes
All authors critically reviewed and approved the final version of the manuscript.
SUPPLEMENTARY MATERIALS
Supplementary material associated with this article can be found in the online version at doi:10.1016/j.bbmt.2020.05.007.
Conflict of interest statement: There are no conflicts of interest to report.
REFERENCES
- 1.Hingorani S, et al. Changes in glomerular filtration rate and impact on long-term survival among adults after hematopoietic cell transplantation: a prospective cohort study. Clin J Am Soc Nephrol. 2018;13(6):866–873. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Agut H. Deciphering the clinical impact of acute human herpesvirus 6 (HHV-6) infections. J Clin Virol. 2011;52(3):164–171. [DOI] [PubMed] [Google Scholar]
- 3.Zerr DM, et al. Clinical outcomes of human herpesvirus 6 reactivation after hematopoietic stem cell transplantation. Clin Infect Dis. 2005;40(7):932–940. [DOI] [PubMed] [Google Scholar]
- 4.Ogata M, et al. Plasma HHV-6 viral load-guided preemptive therapy against HHV-6 encephalopathy after allogeneic stem cell transplantation: a prospective evaluation. Bone Marrow Transplant. 2008;41(3):279–285. [DOI] [PubMed] [Google Scholar]
- 5.Tomblyn M, et al. Guidelines for preventing infectious complications among hematopoietic cell transplantation recipients: a global perspective. Biol Blood Marrow Transplant. 2009;15(10):1143–1238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Horvath MM, et al. The DEDUCE Guided Query tool: providing simplified access to clinical data for research and quality improvement. J Biomed Inform. 2011;44(2):266–276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Armand P, et al. Validation and refinement of the Disease Risk Index for allogeneic stem cell transplantation. Blood. 2014;123(23):3664–3671. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sorror ML, et al. Hematopoietic cell transplantation (HCT)-specific comorbidity index: a new tool for risk assessment before allogeneic HCT. Blood. 2005;106(8):2912–2919. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Fanikos JK, Piazza G, Connors J, Goldhaber SZ. Medication use evaluation: pharmacist rubric for performance improvement. Pharmacotherapy. 2014;34(12):5S–13S. suppl 1). [DOI] [PubMed] [Google Scholar]
- 10.Levey AS, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604–612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bellomo R, et al. Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care. 2004;8(4):R204–R212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Filipovic AH, et al. National Institutes of Health consensus development project on criteria for clinical trials in chronic graft-versus-host disease, I: diagnosis and staging working group report. Biol Blood Marrow Transplant. 2005;11(12):945–956. [DOI] [PubMed] [Google Scholar]
- 13.AstraZeneca. FOSCAVIR Product Information. 2006. [Google Scholar]
- 14.Hingorani S. Renal complications of hematopoietic-cell transplantation. N Engl J Med. 2016;374(23):2256–2267. [DOI] [PubMed] [Google Scholar]
- 15.Al-Hazzouri A, et al. Similar risks for chronic kidney disease in long-term survivors of myeloablative and reduced-intensity allogeneic hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2008;14(6):658–663. [DOI] [PubMed] [Google Scholar]
- 16.Glezerman IG, et al. Long term renal survival in patients undergoing T-cell depleted versus conventional hematopoietic stem cell transplants. Bone Marrow Transplant. 2017;52(5):733–738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.McClellan W, et al. Racial differences in the prevalence of chronic kidney disease among participants in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) cohort study. J Am Soc Nephrol. 2006;17 (6):1710–1715. [DOI] [PubMed] [Google Scholar]
- 18.Deray G, et al. Foscarnet nephrotoxicity: mechanism, incidence and prevention. Am J Nephrol. 1989;9(4):316–321. [DOI] [PubMed] [Google Scholar]
- 19.Zhou W, et al. Long-term renal outcome after allogeneic hemopoietic stem cell transplant: a comprehensive analysis of risk factors in an Asian patient population. Clin Transplant.2017;31(4). [DOI] [PubMed] [Google Scholar]
- 20.Avery RK, et al. Outcomes in transplant recipients treated with foscarnet for ganciclovir-resistant or refractory cytomegalovirus infection. Transplantation. 2016;100(10):e74–e80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Betts BC, et al. Human herpesvirus 6 infection after hematopoietic cell transplantation: is routine surveillance necessary? Biol Blood Marrow Transplant. 2011;17(10):1562–1568. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Styczynski J. Who is the patient at risk of CMV recurrence: a review of the current scientific evidence with a focus on hematopoietic cell transplantation. Infect Dis Ther. 2018;7(1):1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Coresh J, et al. Decline in estimated glomerular filtration rate and subsequent risk of end-stage renal disease and mortality. JAMA. 2014;311 (24):2518–2531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Abboud I, et al. Chronic kidney dysfunction in patients alive without relapse 2 years after allogeneic hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2009;15(10):1251–1257. [DOI] [PubMed] [Google Scholar]
- 25.Touzot M, et al. Long-term renal function after allogenic haematopoietic stem cell transplantation in adult patients: a single-centre study. Nephrol Dial Transplant. 2010;25(2):624–627. [DOI] [PubMed] [Google Scholar]
- 26.Hingorani S, et al. Chronic kidney disease in long-term survivors of hematopoietic cell transplant. Bone Marrow Transplant. 2007;39(4):223–229. [DOI] [PubMed] [Google Scholar]
- 27.Miralbell R, et al. Renal insufficiency in patients with hematologic malignancies undergoing total body irradiation and bone marrow transplantation: a prospective assessment. Int J Radiat Oncol Biol Phys. 2004;58 (3):809–816. [DOI] [PubMed] [Google Scholar]
- 28.Naesens M, Kuypers DR, Sarwal. Calcineurin inhibitor nephrotoxicity. Clin J Am Soc Nephrol. 2009;4(2):481–508. [DOI] [PubMed] [Google Scholar]
- 29.Ellis MJ, et al. Chronic kidney disease after hematopoietic cell transplantation: a systematic review. Am J Transplant. 2008;8(11):2378–2390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ogata M, et al. Effects of prophylactic foscarnet on human herpesvirus-6 reactivation and encephalitis in cord blood transplant recipients: a prospective multicenter trial with an historical control group. Biol Blood Marrow Transplant. 2018;24(6):1264–1273. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.