Abstract
The Endothelial Activation and Stress Index (EASIX) is a prognostic tool that uses common clinical laboratory values and has been shown to predict non-relapse mortality (NRM) and overall survival (OS) at the onset of acute graft-versus-host disease (GVHD) after allogeneic hematopoietic cell transplantation (HCT). We hypothesized that EASIX calculated at different time points pre- and post- HCT may predict NRM and OS, and that EASIX calculated at onset of GVHD may predict response to steroids. We evaluated the EASIX score pre- and post-HCT in 152 patients with lymphoid malignancies undergoing unmodified reduced intensity conditioning (RIC) alloHCT with uniform GVHD prophylaxis. In multivariate analysis, EASIX calculated pre-HCT was significantly associated with higher NRM (HR=1.64, p=0.009) and lower OS (HR=1.33, p=0.046). Furthermore, EASIX calculated at day 30 and at day 100 was associated with increased NRM (HR=1.65, p<0.001; and HR=1.65, p<0.001) and decreased OS (HR=1.27, p=0.018; and HR=1.49, p<0.001), independent of HCT-CI, disease and conditioning regimen. Our study shows that high EASIX scores at various time points pre- and post-HCT are significantly associated with poorer overall outcomes. EASIX provides an independent and easily accessible tool to predict outcomes that can be complementary to other measures of risk stratification for patients undergoing HCT.
Keywords: Endothelial damage, Allogeneic hematopoietic cell transplantation
Introduction
Endothelial damage has been linked to numerous complications early after allogeneic hematopoietic cell transplantation (HCT) that lead to higher morbidity and non-relapse mortality (NRM)1, such as sinusoidal obstruction syndrome (SOS)2, thrombotic microangiopathy (TMA)3, diffuse alveolar hemorrhage4 and graft-versus-host disease (GVHD)5–7. Acute GVHD remains the main cause of non-relapse mortality (NRM) after HCT, particularly in patients with steroid-refractory acute GVHD who have a dismal prognosis8. Steroid-refractory GVHD has been linked to TMA, a microvascular disorder characterized by renal injury in combination with hemolytic anemia and microthrombosis with platelet consumption that is associated with high mortality rates11. In recent years, both clinical tools and circulating biomarkers have helped identify patients with a higher risk of steroid refractory GVHD12–14, in order to more accurately predict or mitigate post- transplantation acute toxicities.
With the aim to develop an easy tool that could be used in clinical practice to predict endothelial dysfunction and TMA after HCT, Luft and colleagues developed the Endothelial Activation and Stress Index (EASIX) based on routine laboratory parameters (high creatinine, high lactate dehydrogenase [LDH] and low thrombocyte counts) that links endothelial damage, TMA, and steroid-refractory GVHD18–20. Endothelial dysfunction contributes to renal failure, glomerulonephritis, and transplantation-related glomerulonephropathy, providing a link to creatinine levels21. Elevated serum LDH in the setting of endothelial injury is related to direct release from endothelial cells and circulating blood cells22. Furthermore, endothelial activation is also associated with clotting and low platelet counts22–23. EASIX calculated pre-HCT (EASIX-pre) has been significantly associated with a diagnosis of TMA, increased NRM, and poorer overall survival (OS) after HCT16. In addition, EASIX calculated at onset of any grade of acute GVHD (EASIX-GVHD) in patients who underwent reduced intensity conditioning (RIC) HCT, has been significantly associated with increased NRM and inferior OS16. Recently, EASIX-pre has been established as a predictor of individual risk of mortality after HCT independently from established clinical criteria in a large cohort of patients18.
In the present study, we examined the EASIX score at different time points pre- and post-HCT in patients with lymphoid malignancies undergoing reduced intensity HCT with uniform GVHD prophylaxis in order to expand on prior studies by validating the ability of the score to predict HCT outcomes such as NRM and OS, as well as examining EASIX as a predictor of response in acute GVHD.
Material and Methods
This retrospective study includes consecutive patients with lymphoid malignancies aged over 18 years who underwent a first unmodified HCT at Memorial Sloan Kettering Cancer Center (MSKCC) between April 2008 and May 2017 using a uniform GVHD prophylaxis of sirolimus/tacrolimus and low/dose methotrexate25. HLA typing was done using high-resolution DNA sequence-specific oligonucleotide typing for the HLA-A, -B, -C, -DRB1, and -DQB1 loci. Written informed consent for treatment was obtained from all patients and donors. Approval for this retrospective review was obtained from the Institutional Review and Privacy Board.
All patients were treated with one of the following previously described RIC or non-myeloablative (NMA) conditioning regimens26, at the discretion of the treating physician: cyclophosphamide (50 mg/kg once) and fludarabine (25 mg/m2 daily for 5 days); cyclophosphamide (50 mg/kg once), fludarabine (25 mg/m2 daily for 5 days) and total body irradiation (TBI, 200 cGy); cyclophosphamide (50 mg/kg once), fludarabine (30 mg/m2 daily for 5 days), thiotepa (5 mg/kg daily for 2 days) and TBI (200 cGy daily for 2 days) TBI27; melphalan (70 mg/m2 daily for 2 days) and fludarabine (30 mg/m2 daily for 5 days); rituximab (375 mg/m2 pre-conditioning), cyclophosphamide (50 mg/kg), fludarabine (25 mg/m2 daily for 5 days) and TBI (200 cGy); fludarabine (30 mg/m2 daily for 4 days) and busulfan (0.8 mg/kg daily for 2 days); fludarabine (30 mg/m2 daily for 4 days), busulfan (0.8 mg/kg daily for 2 days), and rituximab (375 mg/m2). GVHD prophylaxis consisted of sirolimus and tacrolimus that were started on day −3, followed by MTX 5 mg/m2 on days +1, +3 and +625. Doses were adjusted to maintain target serum trough levels of 3–12 and 5–10 ng/mL for sirolimus and tacrolimus, respectively. Tapering of sirolimus and tacrolimus on an alternating schedule typically began at day 60 in the absence of GVHD or relapse, with the goal of complete discontinuation by 6 months, or earlier in patients not in complete remission (CR) at the time of transplantation.
All patients received standard supportive care for prevention of sinusoidal obstruction syndrome and opportunistic antimicrobial prophylaxis according to standard institutional guidelines. Granulocyte colony-stimulating factor was initiated on day +7 and continued until absolute neutrophil count recovery. Comorbidities pre-HCT were assessed using the HCT-specific comorbidity index (HCT-CI)28. Cause of death was determined using a standard algorithm29.
GVHD and TMA assessment
GVHD was diagnosed clinically, confirmed pathologically whenever possible, and classified according to standard criteria30, 31. Patients who engrafted were evaluable for acute GVHD. Patients who relapsed or received donor lymphocyte infusions were censored for GVHD at that time. We also assessed the refined Minnesota risk classification at onset of any grade of acute GVHD and responses to first line treatment of acute GVHD17. Responses to treatment were determined comparing the initial acute GVHD stage and grade in each organ to the best recorded stage and grade at day 28 (±7 days) after treatment was started. Complete response (CR) was defined as the complete resolution of acute GVHD manifestation in all organs, without need for secondary GVHD therapy. Partial response (PR) was defined as improvement in GVHD stage in any initially affected organ, without resolution in all organs, worsening in any other organ or need for secondary GVHD therapy. No response (NR) was defined as the same severity of GVHD in any organ or death, or the addition of secondary GVHD therapy before day 28. A flare of acute GVHD before day 28 and requirement of secondary therapy were also considered to have NR. Progression was defined as worsening of GVHD in at least 1 organ. Assessment of TMA was made retrospectively based on published diagnostic criteria20.
EASIX Formula and Time points
The EASIX formula (LDH [U/L] X CREAT [mg/dL]/ platelet [109 cells/L]) was calculated at different time points pre and post-HCT19. EASIX prior to HCT (EASIX-pre) was calculated from day −30 to day −10 before the start of the conditioning regimen. EASIX post-HCT was calculated on the closest day to day 30 and day 100 post-HCT (EASIX-d30 and EASIX-d100). We also calculated EASIX at onset of any grade of acute GVHD post-HCT (EASIX-GVHD).
Biostatistics
A log transformation using base 2 (log2) was applied to all EASIX variables to reduce skew and to ease interpretation so that a one-unit increase in log2 EASIX is associated with a doubling (one-fold increase) of EASIX on the original scale. Baseline clinical and demographic variables, as well as platelet transfusion values were compared across log2 EASIX quartiles using Kruskall Wallis rank sum tests and Fisher’s exact tests. Kaplan-Meier analysis was utilized to evaluate OS, and cumulative incidence (CI) analyses were used to estimate NRM, relapse and acute GVHD. For all EASIX assessments post-HCT, a landmark analysis was implemented at given time points: day 30 and day 100. Cox proportional hazards models were used for each endpoint to evaluate all covariates, using cause specific hazard ratios for NRM, relapse and acute GVHD. Relapse alone or relapse and death were considered competing risks for the endpoints of NRM and acute GVHD, respectively. Death in the absence of relapse was used as a competing event for the relapse endpoint. All analyses were conducted using R version 3.5.1
Results
Patients’ characteristics and evolution of EASIX during time
Patients’ characteristics for the whole population and based on EASIX-pre quartiles are shown in Table 1. The median age at transplantation was 55 years (range, 24–79 years). All patients were treated for lymphoid malignancies, with 67% non-Hodgkin lymphoma, 16% chronic lymphocytic leukemia, 15% Hodgkin lymphoma, and 3% acute lymphoblastic leukemia. At time of transplantation, most patients had chemosensitive disease: 78 patients (57%) were in CR and 46 patients (34%) were in PR; the other 12 patients (9%) had stable or progressive disease. All patients except two received peripheral blood stem cells as allograft source. Fifty-nine patients (43%) received a transplantation from an HLA-identical sibling donor, while the remaining 77 patients received an unrelated donor transplantation (HLA-matched in 66 [49%] patients, and HLA-mismatched in 11 [8%] patients). The HCT-CI was 0 in 44 patients (32%), 1–2 in 34 patients (25%) and ≥ 3 in 58 patients (43%). Regarding specific co-morbidities that may be associated with endothelial dysfunction, 10 patients had cardiac dysfunction (cardiomyopathy: N = 5; coronary artery disease: N = 5), and one patient had peripheral vascular disease. No patients were noted to have moderate-severe renal dysfunction. Forty-four patients had received a prior autologous-stem cell transplantation.
Table 1.
Patients’ characteristics for whole population and based on EASIX pre- transplantation.
| Patient characteristics | Overall cohort n (%) |
Q1 | Q2 | Q3 | Q4 | p-values |
|---|---|---|---|---|---|---|
| Sample size | 136 | 34 | 34 | 34 | 34 | |
| Median age at transplantation, years (range) | 55 (24–79) | 53 (24–75) | 54.1 (24–68) | 58.6 (27–58) | 55.6 (28–78) | 0.6 |
| DISEASE | 0.2 | |||||
| NHL | 91 (67%) | 19 (56%) | 24 (71%) | 23 (68%) | 25 (74%) | |
| CLL | 21 (15%) | 3 (8.8%) | 5 (15%) | 7 (21%) | 6 (18%) | |
| HD | 20 (15%) | 9 (26%) | 5 (15%) | 4 (12%) | 2 (5.9%) | |
| ALL | 4 (2.9%) | 3 (8.8%) | 0 (0%) | 0 (0%) | 1 (2.9%) | |
| STATUS AT TRANSPLANTATION | 0.09 | |||||
| CR | 78 (57%) | 24 (71%) | 20 (59%) | 17 (50%) | 17 (50%) | |
| PR | 46 (34%) | 9 (26%) | 11 (32%) | 16 (47%) | 10 (29%) | |
| SD+PD | 12 (8.8%) | 1 (2.9%) | 3 (8.8%) | 1 (2.9%) | 7 (21%) | |
| HCT-CI SCORE | 0.3 | |||||
| 0 | 44 (32%) | 9 (26%) | 17 (50%) | 10 (29%) | 8 (24%) | |
| 1–2 | 34 (25%) | 10 (29%) | 7 (21%) | 9 (26%) | 8 (24%) | |
| >=3 | 58 (43%) | 15 (44%) | 10 (29%) | 15 (44%) | 18 (53%) | |
| SEX | >0.9 | |||||
| Male | 94 (69%) | 24 (71%) | 22 (65%) | 23 (68%) | 25 (74%) | |
| Female | 42 (31%) | 10 (29%) | 12 (35%) | 11 (32%) | 9 (26%) | |
| HLA | 0.2 | |||||
| MRD | 59 (43%) | 15 (44%) | 18 (53%) | 15 (44%) | 11 (32%) | |
| MUD | 66 (49%) | 18 (53%) | 14 (41%) | 13 (38%) | 21 (62%) | |
| MM | 11 (8.1%) | 1 (2.9%) | 2 (5.9%) | 6 (18%) | 2 (5.9%) | |
| REGIMEN (%) | 0.6 | |||||
| Chemotherapy | 29 (21%) | 10 (29%) | 7 (21%) | 6 (18%) | 6 (18% | |
| TBI | 107 (79%) | 24 (71%) | 27 (79%) | 28 (82%) | 28 (82%) | |
| CONDITIONING REGIMEN (%) | 0.2 | |||||
| Cy/Flu/TBI | 34 (25%) | 8 (24%) | 10 (29%) | 8 (24%) | 8 (24%) | |
| Cy/Flu/TBI/Rituximab | 65 (48%) | 15 (44%) | 13 (38%) | 20 (59%) | 17 (50%) | |
| Mel/Flu | 23 (17%) | 10 (29%) | 6 (18%) | 4 (12%) | 3 (8.8%) | |
| Cy/Flu/Thio/TBI | 8 (5.9%) | 1 (2.9%) | 4 (12%) | 0 (0%) | 3 (8.8%) | |
| Other | 6 (4.4%) | 0 (0%) | 1 (2.9%) | 2 (5.9%) | 3 (8.8%) | |
| STEM CELL SOURCE | >0.9 | |||||
| PBSC unmodified | 135 (99%) | 34 (100%) | 34 (100%) | 34 (100%) | 33 (97%) | |
| BM unmodified | 1 (0.7%) | 0 (0%) | 0 (0%) | 0 (0%) | 1 (2.9%) | |
| Median CD34 cell dose (range) | 7.33 (0.78, 15.20) | 7.39 (0.78–12.02) | 6.30 (3.30–8.68) | 7.28 (1.40–15.20) | 7.68 (1.87–11.24) | |
| INDICATION | 0.08 | |||||
| Initial Allo HCT | 92 (68%) | 23 (68%) | 26 (76%) | 26 (76%) | 17 (50%) | |
| Allo HCT post-Auto | 44 (32%) | 11 (32%) | 8 (24%) | 8 (24%) | 17 (50%) | |
| EASIX MEDIAN VALUES | ||||||
| Median EASIX-pre (range) | 1.19 (0.39, 27.82) | 0.57 (0.39–0.78) | 1.02 (0.81–1.19) | 1.36 (1.19–1.72) | 3.00 (1.79–27.82) | |
| Median EASIX-day30 (range) | 2.1 (0.3, 317.6) | 1.4 (0.4–6.0) | 1.8 (0.5–11.7) | 2.7 (0.5–317.6) | 4.7 (0.3–21.9) | |
| Median EASIX-day100 (range) | 2.1 (0.3, 97.9) | 1.3 (0.3–18.5) | 2.0 (0.8–20.8) | 2.4 (0.5–66.6) | 4.0 (0.8–97.9) | |
| Median EASIX-GVHD (range) | 2.2 (0.4, 67.4) | 1.4 (0.5–21.6) | 1.3 (0.8–6.2) | 1.7 (0.4–67.4) | 4.0 (0.5–47.1) |
NHL, non-Hodgkin lymphoma; CLL, chronic lymphocytic leukemia; HD, Hodgkin disease; ALL, acute lymphoid leukemia, CR, complete remission; PR, partial remission; SD, stable disease; PD, progression of disease; NMA, non-myeloablative; RIC, reduced intensity; Cy, cyclophosphamide; Flu, fludarabine; TBI, total-body-irradiation; Mel, melphalan; Thio, thiotepa; PBSC, peripheral blood stem cell; BM, bone marrow; MRD, matched related donor; MUD, matched unrelated donor; MM: mismatched donor.
Other characteristics are shown in table 1.
The median (range) of EASIX-pre score for each quartile was 0.57 (0.39–0.78) for quartile 1 (Q1), 1.02 (0.81–1.19) for quartile 2 (Q2), 1.37 (1.21–1.72) for quartile 3 (Q3) and 3.00 (1.79–27.83) for quartile 4 (Q4). When comparing baseline and treatment characteristics, there were no significant differences noted between patients in the four quartiles. In particular, the graft dose ranged from 0.78 – 15.2 CD34/kg and there was no significant difference in CD34 dose between the 4 quartiles. Furthermore, we did not find a significant correlation between CD34 dose and EASIX at day 30 (p=0.09).
Since the platelet count is part of the EASIX formula, we next examined the number of platelet transfusions patients received. One hundred and ten patients had a transfusion before day 100, and 108 patients had a transfusion before day 30. We found that the number of transfusions received by day 30 was associated with the EASIX score at day 30. The median EASIX score (IQR) was 1.6 (1.1, 2.5) in patients who received no transfusion (n=44), 2.1 (1.5, 4.9) in those receiving one (n=24), 2.3 (1.4, 3.8) in those receiving 2–4 (n=50) and 4.4 (2.0, 10.1) in those receiving > 5 (n=34; p<0.001). Furthermore, when examining transfusions by EASIX quartiles, patients in the 4th quartile had significantly more transfusions (3.0; IQR 2.0, 4.0) than those in the other 3 quartiles (1.0 [0.0, 3.8], 1.0 [0.2, 4.0], and 1.0 [0.0, 3.8], for Q1, Q2 and Q3, respectively; p=0.007). These findings are consistent with a higher degree of thrombocytopenia in patients with higher EASIX scores despite receiving transfusions.
Acute graft-versus-host disease (GVHD)
Eighty-eight patients developed acute GVHD (Table 2). The 1-year cumulative incidences of grades 1–4, 2–4, and 3–4 acute GVHD were 56.6% (95% CI, 48.3–64.1), 44.1% (95% CI, 36.0–51.8) and 10.5% (95% CI, 6.3–16.0), respectively. When grouped according to the refined MN grading system, 65 patients (73.8%) were classified as standard-risk (SR-GVHD) and 23 (26.1%) as high-risk (HR-GVHD). The day 28 response rate (CR+PR) was 72.6% in SR-GVHD patients and 47.8% in HR-GVHD.
Table 2a.
GVHD Organ Stage in 88 Patients at Onset of Treatment.
| GVHD organ Stage at onset of diagnosis | 1 | 2 | 3 | 4 |
|---|---|---|---|---|
| Skin | 28 (18%) | 17 (11%) | 6 (4%) | 0 |
| Liver | 4 (3%) | 0 | 2 (1%) | 0 |
| Lower GI | 17 (11%) | 13 (85%) | 2 (1%) | 5 (3%) |
| Upper GI | 28 (18%) | - | - | - |
Some patients experienced more than 1 event.
Thrombotic microangiopathy (TMA)
Thirteen patients (2.5%) developed TMA. In the nine patients (69%) who also developed acute GVHD, none achieved a CR after first line GVHD treatment (PR=4, NR=2 and progression=3). Seven patients with TMA died, with GVHD the most common cause of death (n=3), followed by relapse (n=2), and organ failure/toxicity (n=2).
Overall Survival, Relapse and Non-Relapse Mortality
With a median follow-up in surviving patients of 5.4 years (range, 0.8–10), the 1 year and 3-year OS rates were 84.2% (95% CI, 77.3–89.1) and 67.9% (95% CI, 59.6–74.8), respectively. Cumulative incidence of relapse at 1 and 3 years was 19.1% (95% CI, 13.3–25.7) and 27.4% (95% CI, 20.5–34.8), respectively. At last follow up, 62 patients had died. The NRM rates at 1 and 3 years were 7.9% (95% CI, 4.3–12.9) and 16.4% (95% CI, 10.9–22.8), respectively. Three patients died before day 100 (treatment toxicity = 1, disease progression = 1; and acute GVHD = 1). The primary causes of death were relapse/progression of disease in 28 patients (45.2%), GVHD in 26 patients (41.9%), organ failure/toxicity from treatment in 5 patients (8.1%), and other in 2 (3.2%).
EASIX and outcomes
We looked at associations between HCT outcomes and EASIX determined at different time points. In univariate analysis, EASIX-pre was significantly associated with increased NRM (HR=1.60 [95% CI, 1.15–2.23], p=0.008) and grade 1–4 and 2–4 acute GVHD (HR=1.33 [95% CI, 1.08–1.64], p=0.006; and (HR=1.32 [95% CI, 1.05–1.67], p=0.019, respectively), but not with grade 3–4 acute GVHD (HR=1.14 [95% CI, 0.70–1.85], p=0.588) or OS (HR=1.28 [95% CI, 0.98–1.66], p=0.074) (Table 3). Furthermore, EASIX-pre was not associated with risk of TMA (HR=1.43 [95% CI, 0.93–2.21], p=0.107) or relapse (HR=0.78 [95% CI, 0.54, 1.14], p=0.192). EASIX-d30 and EASIX-d100, were also significantly associated in univariate analysis with higher NRM (HR=1.76 [95% CI, 1.38–2.25], p<0.001 and HR=1.61 [95% CI, 1.27–2.03], p<0.001; respectively) and lower OS (HR=1.39 [95% CI, 1.15–1.68], p=0.001 and HR=1.39 [95% CI, 1.16–1.67], p<0.001; respectively) (Figure 1). The other risk factor significantly associated with higher NRM only on univariate analysis was HCT-CI ≥3 (HR=2.81 [95% CI, 1.17–6.75], p=0.044). HCT-CI ≥3 was also associated with lower OS (HR=3.35 [95% CI, 1.71–6.58], p<0.001). The use of TBI in the conditioning regimen was significantly associated with higher OS (HR=0.52 [0.30–0.91], p=0.027).
Table 3.
Univariate and multivariate analysis.
| VARIABLE | UNIVARIATE ANALYSIS | MULTIVARIATE ANALYSIS | |||
|---|---|---|---|---|---|
|
OS HR (95% CI) p-value |
NRM HR (95% CI) p-value |
Acute GVHD Grade HR (95% CI) p-value |
OS HR (95% CI) p-value |
NRM HR (95% CI) p-value |
|
| Pre-EASIX | 1.28 (0.98, 1.66) 0.074 | 1.60 (1.15, 2.23) 0.008 | 1–4 1.33 (1.08,1.64) 0.006 2–4 1.32 (1.05–1.67) 0.019 3–4 1.14 (0.70–1.85) 0.588 |
1.33 (1.01,1.74) 0.046 | 1.64 (1.14,2.34) 0.009 |
| EASIX d30 | 1.39 (1.15, 1.68) 0.001 | 1.76 (1.38, 2.25) <0.001 | 1.27 (1.04,1.54) 0.018 | 1.65 (1.29,2.11) <0.001 | |
| EASIX d100 | 1.39 (1.16, 1.67) <0.001 | 1.61 (1.27, 2.03) <0.001 | 1.49 (1.22,1.81) <0.001 | 1.65 (1.282.12) <0.001 | |
| OTHER FACTORS | |||||
|
C. Regimen
TBI |
0.52 (0.30, 0.91) 0.027 |
0.60 (0.26, 1.35) 0.233 |
0.71 (0.34, 1.47) 0.363 |
||
|
HLA
Mistmatched |
3.75 (1.29, 9.87) 0.063 |
2.68 (0.91, 7.90) 0.175 |
|||
|
HCT-CI
1–2 ≥3 |
1.76 (0.80, 3.87) 3.35 (1.71, 6.58) <0.001 |
1.52 (0.55, 4.19) 2.81 (1.17, 6.75) 0.044 |
1.21 (0.49, 2.97) 3.36 (1.66, 6.79) <0.001 |
1.04 (0.34, 3.17) 2.32 (0.94, 5.69) 0.094 |
|
Figure 1.
EASIX-pre, at day 30 and at day 100 predict higher NRM (A) and lower OS (B).
In multivariate analysis, EASIX calculated pre-HCT was significantly associated with higher NRM (HR=1.64, p=0.009) and lower OS (HR=1.33, p=0.046). Furthermore, EASIX calculated at day 30 and at day 100 was associated with increased NRM (HR=1.65, p<0.001; and HR=1.65, p<0.001) and decreased OS (HR=1.27, p=0.018; and HR=1.49, p<0.001), independent of HCT-CI, disease and conditioning regimen.
We next analyzed the relationship between EASIX and day 28 GVHD responses to EASIX-GVHD quartiles (Figure 2). The day 28 CR rate was 42% among 19 patients in EASIX-GVHD quartile 1 (Q1) in contrast to 5% among 21 patients on EASIX-GVHD quartile 4 (Q4).
Figure 2. GVHD Steroid Responses by day 28 based EASIX-GVHD quartiles.
EASIX was calculated at onset of acute GVHD and patients were divided in 4 quartiles. GVHD responses to treatment were assessed at day 28.
We also assessed causes of death by EASIX-d30 quartiles (Figure 3). Most deaths occurred in the EASIX-d30 Q4 group (22 of 38 patients died) and were primarily due to GVHD (50%) or relapse (27%).
Figure 3.
Causes of death by EASIX-d30 quartiles.
Discussion
In a single center cohort of patients with lymphoid malignancies treated with reduced intensity conditioning and a uniform GVHD prophylaxis, we examined the EASIX score at various time points pre and post HCT by dividing patients into 4 quartiles based on their score and showed that most patients in EASIX-pre quartile 1 (Q1) or quartile 4 (Q4) remain in the same quartile over time (day 30 and day 100). In contrast, patients in EASIX-pre Q2 and Q3 tend to vary more between different quartiles during transplantation. Furthermore, we showed that higher values of EASIX-pre, EASIX-d30 and EASIX-d100 were all independently associated with lower OS and higher NRM. Finally, we showed a difference in responses to 1st line therapy in patients with acute GVHD based on EASIX-GVHD.
Our data concur with previous analyses showing that EASIX is an independent predictor for NRM and inferior survival. In the original study by Luft et al.18, EASIX evaluated at one timepoint, EASIX-GVHD, was a predictor of higher NRM and lower OS. Recently, EASIX-pre has also been validated as a predictor of lower OS after alloHCT18, and was associated both with high-grade fluid overload and OS32. This suggests that EASIX can be considered a dynamic variable that may have unique implications along the pre- and post-HCT time course. Furthermore, we and others have now expanded the application of the EASIX score to treatment with chimeric antigen receptor T cells and shown that it can predict severe cytokine release syndrome and neurotoxicity. 33, 34
Our study also showed that EASIX-pre was significantly associated with development of acute GVHD grade 1–4 and 2–4 in the univariate analysis. In the last few years, a number of potential biomarkers have been proposed for acute-GVHD and validated in a number of cohorts15 13–14. The biomarker that seems the most promising in the setting of acute GVHD is ST2 (suppression of tumorigenicity 2). ST2, calculated at day 7 and at day 14 post-AlloHCT, alone35–37 or in combination with TIM3 (T-cell immunoglobulin mucin-3) 36–38, is a prognostic marker for acute GVHD and NRM. Reg3α (regenerating islet-derived 3-α) and TIM3 have also been validated as prognostic biomarkers of acute GVHD, measured at day 7 and 14 post-AlloHCT39, 40. In the setting of TMA, ST2 was shown to be a reliable early biomarker of TMA diagnosis independent of acute-GVHD in several cohorts41,42. However, none of these biomarkers or panels of biomarkers have been validated for routine clinical practice. Furthermore, in the study by Luft et al, ST2 did not correlate with EASIX-pre.43 However, correlations at other time points, and at onset of GVHD in particular were not studied. The EASIX formula represents an easily accessible tool that can be used outside of the clinical trial setting and could be used in the future to develop new risk-adapted GVHD diagnosis and treatment strategies in AlloHCT. The use of EASIX in this setting will require further validation in well-designed prospective clinical trials.
Our study did have some limitations, the main one being the retrospective nature of the study. While this is a single center study, our results are consistent with and expand on the results from several other centers. We also chose to study a uniform population in terms of disease and GVHD prophylaxis to decrease potential variability in the study cohort. Finally, some outcomes in particular, such as TMA, had a low event rate in our series and the study was therefore underpowered to detect any association with EASIX. However, a recent study by Luft and colleagues reported an association between EASIX-pre and TMA.43
Overall, our study confirms the main utility of the EASIX formula approach based on the use of routinely available standard clinical laboratory parameters. While these parameters are non-specific and there may be many confounding factors, such as transfusions, infections or cytotoxic therapies that may affect each individual parameter, the score has now been validated across multiple centers and datasets. In the current study, we further expand on this work and illustrate the dynamic aspects of EASIX over time during the HCT journey as well as showing that the score retains its predictive value over time.
Table 2b.
GVHD Clinical Grade in 88 Patients at Onset of Treatment.
| A | B | C | D | |
|---|---|---|---|---|
| I | 20 | 12 | 0 | 0 |
| II | 0 | 28 | 6 | 0 |
| III | 0 | 12 | 5 | 0 |
| IV | 0 | 0 | 0 | 5 |
Acknowledgments
This research was supported in part by National Institutes of Health award numbers P01 CA23766 and NIH/NCI Cancer Center Support Grant P30 CA008748. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Dr. Sanchez-Escamilla was supported by Research Institute of Marques de Valdecilla (IDIVAL) Lopez-Albo-Wenceslao Grant (WLA17/03).
Footnotes
Financial Disclosure Statement (COI)
Dr. Perales reports honoraria from Abbvie, Allovir, Astellas, Bristol-Myers Squibb, Celgene, Equilium, Exevir, Incyte, Karyopharm, Kite/Gilead, Merck, Miltenyi Biotec, MorphoSys, Novartis, Nektar Therapeutics, Omeros, OrcaBio, Takeda, and VectivBio AG, Vor Biopharma. He serves on DSMBs for Cidara Therapeutics, Medigene, Sellas Life Sciences, and Servier, and the scientific advisory board of NexImmune. He has ownership interests in NexImmune and Omeros. He has received institutional research support for clinical trials from Incyte, Kite/Gilead, Miltenyi Biotec, Nektar Therapeutics, and Novartis. MS served as a paid consultant for McKinsey & Company, Angiocrine Bioscience, Inc., and Omeros Corporation; received research funding from Angiocrine Bioscience, Inc., and Omeros Corporation; served on ad hoc advisory boards for Kite – A Gilead Company; and received honoraria from i3Health and Medscape for CME-related activity.
COI
The other authors declare no conflict of interests.
References
- 1.Palomo M, Fernandez-Aviles F, Ghita G, et al. , “Endothelial Dysfunction after Hematopoietic Stem Cell Transplantation: Role of the Conditioning Regimen and the Type of Transplantation,” Biol. Blood Marrow Transplant, vol. 16, no. 7, pp. 985–993, 2010. [DOI] [PubMed] [Google Scholar]
- 2.McCuskey RS, Wang X, McCuskey MK, et al. , “Embolization by sinusoidal lining cells obstructs the microcirculation in rat sinusoidal obstruction syndrome,” Am. J. Physiol. Liver Physiol, vol. 284, no. 6, pp. G1045–G1052, 2015. [DOI] [PubMed] [Google Scholar]
- 3.Koreth J, Alyea E, Byun BC, et al. , “Impact of Thrombotic Microangiopathy on Renal Outcomes and Survival after Hematopoietic Stem Cell Transplantation,” Biol. Blood Marrow Transplant, vol. 24, no. 11, pp. 2344–2353, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Abdulai RM, Cartin-Ceba R, Afessa B, and Clinic M, “Diffuse Alveolar Hemorrhage In Hematopoietic Stem Cell Transplant Recipients,” Chest, pp. 4–5, 2007. [Google Scholar]
- 5.Biedermann BC, Sahner S, Gregor M, et al. , “Mechanisms of disease Endothelial injury mediated by cytotoxic T lymphocytes and loss of microvessels in chronic graft versus host disease,” Lancet, vol. 359, pp. 2078–2083, 2002. [DOI] [PubMed] [Google Scholar]
- 6.Biedermann BC, “Vascular endothelium and graft-versus-host disease,” Best Pract. Res. Clin. Haematol, vol. 21, no. 2, pp. 129–138, 2008. [DOI] [PubMed] [Google Scholar]
- 7.Mir E, Palomo M, Rovira M, et al. , “Endothelial damage is aggravated in acute GvHD and could predict its development,” Bone Marrow Transplant, no. May, pp. 1–9, 2017. [DOI] [PubMed] [Google Scholar]
- 8.Ferrara JLM, Levine JE, Reddy P, and Holler E, “Graft-versus-host disease,” Lancet, vol. 373, no. 9674, pp. 1550–1561, 2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Anderlini P, Champlin R, Khouri IF, et al. , “Steroid-Refractory Acute GVHD: Predictors and Outcomes,” Adv. Hematol, 2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Luft T, Dietrich S, Falk C, et al. , “Steroid-refractory GVHD : T-cell attack within a vulnerable endothelial system,” Blood, vol. 118, no. 6, pp. 1685–1693, 2011. [DOI] [PubMed] [Google Scholar]
- 11.George JN, Li X, Mcminn JR, et al. , “Thrombotic thrombocytopenic purpura-hemolytic uremic syndrome following allogeneic HPC transplantation: a diagnostic dilemma.,” Transfusion, vol. 44, no. 2, pp. 294–304, 2004. [DOI] [PubMed] [Google Scholar]
- 12.Zeisbrich M, Becker N, Benner A, et al. , “Transplant-associated thrombotic microangiopathy is an endothelial complication associated with refractoriness of acute GvHD,” Bone Marrow Transplant, vol. 52, no. 10, pp. 1399–1405, 2017. [DOI] [PubMed] [Google Scholar]
- 13.Ali AM, Dipersio JF, and Schroeder MA, “The Role of Biomarkers in the Diagnosis and Risk Stratification of Acute Graft-versus-Host Disease : A Systematic Review,” Biol. Blood Marrow Transplant, vol. 22, no. 9, pp. 1552–1564, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Levine JE, Braun TM, Harris AC, et al. , “A prognostic score for acute graft-versus-host disease based on biomarkers: A multicentre study,” Lancet Haematol, vol. 2, no. 1, pp. e21e–29, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Paczesny S, “Acute Graft-Versus-Host Disease Prognosis: Are Biomarkers Ready for Preemptive Clinical Trials?,” Clin. Chem, vol. 63, no. 10, pp. 1561–1563, 2017. [DOI] [PubMed] [Google Scholar]
- 16.Paczesny S, “Biomarkers for posttransplantation outcomes,” Blood, vol. 131, no. 20, pp. 2193–2205, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.MacMillan ML, Robin M, Harris AC, et al. , “A refined risk score for acute graft-versus-host disease that predicts response to initial therapy, survival, and transplant-related mortality,” Biol. Blood Marrow Transplant, vol. 21, no. 4, pp. 761–767, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Luft T, Benner A, Jodele S, et al. , “It Is Easix to Predict Non-Relapse Mortality (NRM) of Allogeneic Stem Cell Transplantation (alloSCT),” Blood, vol. 128, no. 22, p. 519, 2016.27226436 [Google Scholar]
- 19.Luft T, Benner A, Jodele S, et al. , “EASIX in patients with acute graft-versus-host disease: a retrospective cohort analysis,” Lancet Haematol, vol. 4, no. 9, pp. e414–e423, 2017. [DOI] [PubMed] [Google Scholar]
- 20.Ruutu T, Barosi G, Benjamin RJ, et al. , “Diagnostic criteria for hematopoietic stem cell transplant-associated microangiopathy: results of a consensus process by an International Working Group,” Haematologica, vol. 92, no. 01, pp. 95–100, 2007. [DOI] [PubMed] [Google Scholar]
- 21.Verma SK, and Molitoris BA, “Renal Endothelial Injury and Microvascular Dysfunction in Acute Kidney Injury,” Semin. Nephrol, vol. 35, no. 1, pp. 96–107, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Joist JH, and Baker RK, “Loss of 111Indium as indicator of platelet injury,” Blood, vol. 58, no. 2, pp. 350–353, 1981. [PubMed] [Google Scholar]
- 23.Coppo P, Schwarzinger M, Buffet M, et al. , “Predictive features of severe acquired ADAMTS13 deficiency in idiopathic thrombotic microangiopathies: The French TMA reference center experience,” PLoS One, vol. 5, no. 4, p. e10208, 2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Moreau D, Timsit J-F, Vesin A, et al. , “Platelet Count Decline: An Early Prognostic Marker in Critically Ill Patients With Prolonged ICU Stays,” Chest, vol. 131, no. 6, pp. 1735–1741, 2007. [DOI] [PubMed] [Google Scholar]
- 25.Ceberio I, Devlin SM, Sauter C, et al. , “Sirolimus, tacrolimus and low-dose methotrexate based graft-versus-host disease prophylaxis after non-ablative or reduced intensity conditioning in related and unrelated donor allogeneic hematopoietic cell transplant,” Leuk. Lymphoma, vol. 56, no. 3, pp. 663–670, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bacigalupo A, Ballen K, Rizzo D, et al. , “Defining the Intensity of Conditioning Regimens : Working Definitions,” Biol. Blood Marrow Transplant, vol. 15, no. 12, pp. 1628–1633, 2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hobbs G, Kaur N, Hilden P, et al. , “A novel reduced intensity conditioning regimen for patients with high-risk hematological malignancies undergoing allogeneic stem cell transplantation,” Bone Marrow Transplant, vol. 51, no. 7, pp. 1010–1012, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Sorror ML, Logan BR, Zhu X, et al. , “Prospective Validation of the Predictive Power of the Hematopoietic Cell Transplantation Comorbidity Index : A Center for International Blood and Marrow Transplant Research Study,” Biol. Blood Marrow Transplant, vol. 21, no. 8, pp. 1479–1487, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Copelan E, Casper JT, Carter SL, et al. , “A Scheme for Defining Cause of Death and Its Application in the T Cell Depletion Trial,” Biol. Blood Marrow Transplant, vol. 13, no. 12, pp. 1469–1476, 2007. [DOI] [PubMed] [Google Scholar]
- 30.Rowlings PA, Przepiorka D, Klein JP, et al. , “IBMTR Severity Index for grading acute graft-versus-host disease: Retrospective comparison with Glucksberg grade,” Br. J. Haematol, vol. 97, no. 4, pp. 855–864, 1997. [DOI] [PubMed] [Google Scholar]
- 31.MacMillan ML, Weisdorf DJ, Wagner JE, et al. , “Response of 443 patients to steroids as primary therapy for acute graft-versus-host disease: Comparison of grading systems,” Biol. Blood Marrow Transplant, vol. 8, no. 7, pp. 387–394, 2002. [DOI] [PubMed] [Google Scholar]
- 32.Saliba RM, Rondon G, Chen J, et al. , “Endothelial Activation and Stress Index (EASIX) Is Associated with Fluid Overload and Survival in Recipients of Allogeneic Stem Cell Transplantation,” Biol. Blood Marrow Transplant, vol. 25, no. 3, pp. 32–33, 2019. [DOI] [PubMed] [Google Scholar]
- 33.Pennisi M, Sanchez-Escamilla M, Flynn JR, et al. , “Modified EASIX predicts severe cytokine release syndrome and neurotoxicity after chimeric antigen receptor T cells,” Blood Adv, vol. 5, no. 17, pp. 3397–3406, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Greenbaum U, Strati P, Saliba RM, et al. , “CRP and ferritin in addition to the EASIX score predict CAR-T-related toxicity,” Blood Adv, vol. 5, no. 14, pp. 2799–2806, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Kanakry CG, Bakoyannis G, Perkins SM, et al. , “Plasma-derived proteomic biomarkers in human leukocyte antigen-haploidentical or human leukocyte antigen-matched bone marrow transplantation using post-transplantation cyclophosphamide,” Haematologica, vol. 102, no. 5, pp. 932–940, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Ponce DM, Hilden P, Mumaw C, et al. , “High day 28 ST2 levels predict for acute graft-versus-host disease and transplant-related mortality after cord blood transplantation,” Blood, vol. 125, no. 1, pp. 199–206, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Vander Lugt M, Braun T, Hanash S, et al. , “ST2 as a Marker for Risk of Therapy-Resistant Graft-versus-Host Disease and Death,” N. Engl. J. Med, vol. 369, no. 6, pp. 529–539, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Zaid MA, Wu J, Wu C, et al. , “Plasma biomarkers of risk for death in a multicenter phase 3 trial with uniform transplant characteristics post-allogeneic HCT,” Blood, vol. 129, no. 2, pp. 162–171, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Mcdonald GB, Tabellini L, Storer BE, et al. , “Plasma biomarkers of acute GVHD and nonrelapse mortality : predictive value of measurements before GVHD onset and treatment,” Blood, vol. 126, no. 1, pp. 113–121, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Hartwell MJ, Levine JE, Ferrara JLM, et al. , “An early-biomarker algorithm predicts lethal graft-versus-host disease and survival Find the latest version : An early-biomarker algorithm predicts lethal graft-versus-host disease and survival,” JCI Insight, vol. 2, no. 3, pp. 0–9, 2018. [Google Scholar]
- 41.Rotz SJ, Luebbering N, Dixon BP, et al. , “In vitro evidence of complement activation in transplantation-associated thrombotic microangiopathy,” Blood Adv, vol. 1, no. 20, pp. 1632–1634, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Rotz S Dandoy C Davies S, “ST2 and Endothelial Injury as a Link between GVHD and Microangiopathy,” N. Engl. J. Med, vol. 376, no. 12, pp. 1189–1190, 2017. [DOI] [PubMed] [Google Scholar]
- 43.Luft T, Benner A, Terzer T, et al. , “EASIX and mortality after allogeneic stem cell transplantation,” Bone Marrow Transplant, vol. 55, pp. 553–561, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]