Abstract
The prognosis of steroid-refractory acute graft-versus-host disease (aGVHD) is poor and predictors of response and survival are unclear. In an exploratory analysis of 203 steroid-refractory aGVHD patients with prospectively collected GVHD data who received antithymocyte globulin, etanercept, or mycophenolate mofetil (MMF) as second-line treatment, we determined the predictors of day 28 response, 2-year overall survival (OS), and 2-year non-relapse mortality (NRM). To minimize the risk of finding false positive results, we used lasso regression, aggressively eliminating variables that are unlikely to be associated with outcome. Day 28 response to second-line therapy was 38% (complete response [CR] 23%), with a 2-year OS of 25% and a 2-year NRM of 62%. Factors associated with response were GVHD prophylaxis, organ involvement, and initial aGVHD to steroid-refractory aGVHD interval. Specifically, compared with cyclosporine (CsA)/MMF as GVHD prophylaxis, the odds ratio (OR) for calcineurin inhibitor/methotrexate was 0.8 and for CsA/prednisone was 0.6. The OR for aGVHD to steroid-refractory aGVHD interval ≥14 vs. <14 days was 1.3. The ORs for skin only involvement and gut or liver only involvement when compared with multi-organ involvement were 1.4 and 1.2, respectively. The only variable associated with worse survival was age, with a hazard ratio (HR) per decade of 1.04 for overall mortality. Similarly, age was the only variable associated with NRM (HR 1.02 per decade). When compared with CR, no response at day 28 increased the risk of death (HR: 2.4, 95% confidence interval: 1.5–3.7). In conclusion, using an underutilized statistical technique in the field of transplantation, we identified predictors of response and survival in steroid-refractory aGVHD. Our results highlight the importance of developing novel treatment strategies as current treatments yield poor outcomes.
Keywords: Steroid-refractory, Graft-versus-host disease, ATG, Etanercept
Introduction
Acute graft-versus-host disease (aGVHD) remains a major cause of morbidity and mortality after allogeneic hematopoietic cell transplantation (HCT)1. High-dose steroids are the mainstay of frontline therapy for aGVHD and the addition of other agents in this setting have not improved outcomes2,3. Steroid-refractory aGVHD (raGVHD) has a dismal prognosis and most patients succumb to organ failure or infection after a few months4–6. Using our prospectively collected, regularly updated, and strictly curated BMT database at the University of Minnesota, we asked the following questions in patients with steroid-refractory aGVHD: (i) What are the outcomes of steroid-refractory aGVHD? (ii) What baseline patient-, disease-, transplant-, and aGVHD-related characteristics are associated with steroid-refractory aGVHD outcomes? and (iii) Is steroid-refractory aGVHD response to second-line therapy associated with better overall survival (OS)?
Materials and Methods
Data from all first allo-HCT recipients at the University of Minnesota (1990–2016) who developed aGVHD were reviewed. aGVHD was graded by the Minnesota grading system, which uses standard clinical criteria derived from organ staging7 modified to include upper gastrointestinal aGVHD per the GVHD consensus conference8–10. All patients were to receive prednisone 60 mg/m2/day or methylprednisolone equivalent (divided in three doses) for 7 consecutive days, followed by daily prednisone for 7 days as initial therapy for aGVHD. Patients were maintained on therapeutic levels of cyclosporine (CSA), tacrolimus or sirolimus. Additionally, patients with skin acute GVHD were treated with topical 0.1% triamcinolone cream or 1% hydrocortisone cream (for facial rash) thrice daily. If a response to prednisone was observed, patients continued therapy with oral prednisone 60 mg/m2/day through day 14 and then commenced a taper of steroids over 8 weeks3,11. Response to therapy was recorded weekly in the University of Minnesota BMT Database. All GVHD data were retrospectively reviewed and adjudicated by four of the authors (MLM, DJW, SGH and AR).
Progression was defined as worsening GVHD in at least one organ within 28 days. Steroid-refractory aGVHD was defined as progression of aGVHD after 3 days of the initial treatment, no improvement after 7 days, or requirement for second-line treatment at any point during or after completion of steroid taper. Continuation and dosage of steroids after the diagnosis of steroid-refractory aGVHD was at the discretion of the treating physicians. We considered steroid-refractory aGVHD responsive to therapy at day 28 (+/−7 days) if the patient was alive, had improvement in GVHD stage in at least organ, and no worsening in other organs. Ten patients started steroid-refractory aGVHD therapy before day 3 of initial GVHD therapy; another 96 patients received steroid boosts alone, sirolimus alone, tacrolimus alone, or experimental agents. These 106 patients were not included in the final dataset. The remaining patients were selected for analysis. These patients received one of the 3 most commonly used treatments during the period of our study: antithymocyte globulin (equine ATG: 30 mg/kg/day given either in divided BID doses or once daily for 5 days)6, etanercept (25 mg twice weekly for 4 weeks), or mycophenolate mofetil (MMF).
The primary objective of this study was to study the clinical predictors of day 28 overall response, 2-year OS, and 2-year non-relapse mortality (NRM) after second-line therapy without emphasis on any formal comparisons between second-line therapies. The data were prospectively collected but retrospectively reviewed. Response was estimated with simple proportions, survival by Kaplan-Meier curves, and NRM by cumulative incidence treating relapse as a competing risk. Following the recommendations of the American Statistical Association, we reported confidence intervals around our point estimates to show precision without focusing on whether or not the intervals included a specific value such as 1.012. Since this analysis was exploratory in nature and we had many potential predictors for outcomes, we performed lasso (least absolute shrinkage and selection operator)13 regression (logistic for day 28 response and Cox for OS and NRM) rather than standard regression. Standard regression in such cases often results in model overfitting and/or spurious findings that may not be replicable in independent cohorts. The goal of lasso regression is to select only a subset (typically small) of the covariates in the final model by aggressively eliminating the other variables that are unlikely to be associated with the dependent variable. Lasso achieves the subset of the most influential predictors by forcing the sum of the absolute value of the regression coefficients to be less than a fixed value. This process forces the regression coefficient of certain variables to “shrink” to zero, thereby achieving a simple model with only a few likely correlates of the dependent variable among all covariates. Lasso regression is frequently used for the analysis of high throughput datasets. Lasso was implemented using the R package glmnet. 10-fold cross validation was used to estimate the optimal lasso constraint parameter that minimized prediction error (mean squared error for logistic regression and partial likelihood for cox regression).
We included the following variables in lasso regression: aGVHD grade and Minnesota risk (standard vs. high risk)14,15 at the initial diagnosis of aGVHD and steroid-refractory aGVHD, patient age (per increase by decade), patient gender (male vs. female), donor type (matched sibling vs. unrelated donor [URD] vs. umbilical cord blood [UCB]), underlying disease (acute leukemia vs. other malignancies vs. non-malignant disorders), GVHD prophylaxis (CsA plus MMF vs. calcineurin inhibitor [CNI] plus methotrexate [MTX] vs. ex-vivo T-cell depletion vs. others), HCT to initial aGVHD interval (continuous), initial aGVHD to steroid-refractory aGVHD interval (<14 vs. ≥14 days), steroid-refractory aGVHD organ involvement (multiple organs vs. skin only vs. gut or liver only), and steroid-refractory aGVHD treatment (ATG vs. etanercept vs. MMF), transplant year (5-year intervals), and conditioning intensity (myeloablative without total body irradiation [TBI] vs. TBI-based myeloablative vs. reduced-intensity). Graft source was not included due to its collinearity with donor type. After identifying the most important predictors of OS and NRM, a subsequent objective was to evaluate whether day 28 response is associated with OS and NRM. Cox regression used landmark analysis (excluding deaths prior to day 28 for OS and a few relapses prior to day 28 for NRM) with mortality measured from day 28 of steroid-refractory aGVHD therapy. We used SAS 9.4 (SAS Institute, Inc., Cary, NC) and R version 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria) for all analyses.
Results
A total of 203 patients met the eligibility criteria and were included in the analysis (Table 1). Baseline characteristics of this group were similar to the 106 excluded patients (59% male, median [range] age 40 [1–75] years, 43% before year 2000, 45% acute leukemia, 15% non-malignant). Organ stages at the time of initial diagnosis and steroid-refractoriness are shown in supplementary Table S1. As expected, stage 3–4 lower gut aGVHD was more common at the time of diagnosis of steroid-refractory aGVHD (38%) than initial aGVHD diagnosis (19%).
Table 1:
Day 28 response, 2-year OS, and 2-year NRM
| Variable | N (%) or value | Day 28 response % (95%CI) | 2-year OS % (95%CI) | 2-year NRM % (95%CI) |
|---|---|---|---|---|
| N | 203 | 38; 32–46) | 25 (25–32) | 65 (56–73) |
| Gender | ||||
| Male | 132 (65) | 40 (32–49) | 24 (17–32) | 65 (55–76) |
| Female | 71 (35) | 35 (24–48) | 28 (18–39) | 63 (50–77) |
| Age at transplant | ||||
| Median (range), years | 35 (<1–75) | |||
| <18 | 61 (30) | 34 (23–48) | 32 (21–44) | 61 (47–75) |
| 18–40 | 66 (33) | 36 (25–49) | 27 (17–38) | 61 (47–75) |
| >40 | 76 (37) | 43 (32–55) | 18 (11–28) | 71 (57–85) |
| Donor Type | ||||
| Matched sibling | 76 (37) | 39 (28–51) | 29 (19–39) | 59 (46–72) |
| URD WM | 26 (13) | 38 (20–59) | 14 (4–30) | 82 (60–100) |
| URD PM | 21 (10) | 33 (15–57) | 19 (6–38) | 76 (52–100) |
| URD MM | 34 (17) | 32 (17–51) | 32 (18–48) | 59 (40–78) |
| UCB | 46 (23) | 43 (29–59) | 24 (13–37) | 63 (46–80) |
| Graft Source | ||||
| BM | 109 (15) | 34 (25–44) | 27 (19–36) | 63 (51–74) |
| PB | 48 (65) | 44 (30–59) | 23 (12–35) | 69 (52–86) |
| UCB | 46 (23) | 43 (29–59) | 24 (13–37) | 63 (46–80) |
| Conditioning | ||||
| MAC, non-TBI | 30 (15) | 23 (10–42) | 33 (17–50) | 64 (44–84) |
| MAC, TBI | 132 (65) | 40 (32–49) | 26 (19–33) | 64 (54–75) |
| RIC | 41 (20) | 44 (29–60) | 20 (9–33) | 66 (48–84) |
| GVHD Prophylaxis | ||||
| CsA/MMF | 67 (33) | 52 (40–64) | 25 (16–36) | 63 (49–77) |
| CNI/MTX | 79 (39) | 34 (24–46) | 27 (17–37) | 65 (52–78) |
| T-cell depletion | 21 (10) | 33 (15–57) | 17 (4–36) | 73 (48–98) |
| CsA/prednisone ± ATG | 18 (9) | 28 (10–54) | 17 (4–37) | 78 (52–100) |
| Others | 18 (9) | 22 (6–48) | 39 (17–60) | 50 (26–74) |
| Sirolimus/MMF | 2 | |||
| MTX/ATG/prednisone | 12 | |||
| ATG | 4 | |||
| Underlying disease | ||||
| Non-malignant | 32 (16) | 34 (19–53) | 31 (16–47) | 69 (49–88) |
| ALL | 30 (15) | 47 (28–66) | 33 (18–50) | 57 (37–77) |
| AML | 47 (23) | 40 (26–56) | 19 (9–31) | 62 (45–78) |
| Lymphoma/CLL | 22 (11) | 50 (28–72) | 36 (17–56) | 50 (28–72) |
| MDS/MPN | 18 (9) | 28 (10–54) | 17 (4–37) | 67 (41–93) |
| Other Malignancy | 54 (27) | 33 (21–48) | 22 (12–34) | 75 (51–83) |
| Treatment of steroid-refractory aGVHD | ||||
| ATG | 166 (82) | 37 (29–45) | 24 (18–31) | 67 (58–76) |
| Etanercept | 10 (5) | 20 (3–56) | 30 (7–58) | 70 (38–100) |
| MMF | 27 (13) | 56 (35–75) | 33 (17–51) | 48 (28–68) |
| HCT to initial aGHVD interval | ||||
| Median (range), days | 30 (8–170) | |||
| <28 | 82 (40) | 33 (23–44) | 22 (14–31) | 68 (55–81) |
| ≥28 | 121 (60) | 42 (33–52) | 28 (20–36) | 62 (52–73) |
| Initial aGVHD to steroid-refractory aGVHD interval | ||||
| Median (range), days | 11 (3–98) | |||
| <14 | 113 (56) | 34 (25–43) | 28 (20–37) | 65 (54–76) |
| ≥14 | 90 (44) | 44 (34–55) | 22 (14–31) | 64 (52–77) |
| aGVHD grade at initial diagnosis | ||||
| I | 42 (21) | 36 (22–52) | 21 (11–35) | 67 (49–84) |
| II | 103 (51) | 41 (31–51) | 25 (17–34) | 63 (52–75) |
| III | 50 (25) | 38 (25–53) | 29 (17–42) | 65 (49–81) |
| IV | 8 (4) | 25 (3–65) | 25 (4–56) | 75 (40–100) |
| Steroid-refractory aGVHD grade | ||||
| I | 24 (12) | 42 (22–63) | 21 (8–39) | 58 (35–82) |
| II | 72 (35) | 42 (30–54) | 29 (19–40) | 57 (44–70) |
| III | 77 (38) | 35 (25–47) | 21 (13–31) | 72 (59–85) |
| IV | 30 (15) | 37 (20–56) | 30 (15–47) | 70 (50–90) |
| MN risk score at initial diagnosis | ||||
| Standard | 163 (80) | 37 (30–45) | 23 (17–30) | 67 (58–76) |
| High | 40 (20) | 43 (27–59) | 34 (20–49) | 56 (38–73) |
| MN risk score at the diagnosis of steroidrefractory aGVHD | ||||
| Standard | 118 (58) | 40 (31–49) | 26 (19–34) | 61 (50–72) |
| High | 85 (42) | 36 (26–48) | 24 (16–34) | 70 (57–82) |
| Steroid-refractory aGVHD organ involvement | ||||
| Multiple organs | 80 (39) | 28 (18–37) | 23 (15–32) | 66 (53–79) |
| Skin only | 70 (34) | 46 (34–58) | 29 (19–39) | 59 (45–72) |
| Liver only | 4 (2) | 25 (1–81) | 0 | 75 (37–100) |
| Gut only | 49 (24) | 47 (33–62) | 26 (15–39) | 70 (53–87) |
| Year of transplant | ||||
| 1990–1994 | 59 (29%) | 29 (17–41) | 31 (19–42) | 59 (44–74) |
| 1995–1999 | 41 (20%) | 32 (18–46) | 17 (8–30) | 73 (55–91) |
| 2000–2004 | 41 (20%) | 54 (39–69) | 32 (18–46) | 56 (39–74) |
| 2005–2009 | 20 (10%) | 55 (33–77) | 40 (18–62) | 40 (19–60) |
| 2010–2016 | 42 (21%) | 36 (21–51) | 13 (5–25) | 85 (66–100) |
aGVHD: Acute graft-versus-host disease; ALL: Acute lymphoblastic leukemia; AML: Acute myeloid leukemia; ATG: Antithymocyte globulin; BM: Bone marrow; CI: Confidence interval; CLL: Chronic lymphocytic leukemia; CNI: Calcineurin inhibitor; CsA: Cyclosporine; HCT: Hematopoietic cell transplantation; MAC: Myeloablative conditioning; MDS: Myelodysplastic syndrome; MM: Mismatched; MMF: Mycophenolate mofetil; MN: Minnesota; MPN: Myeloproliferative neoplasm; MTX: Methotrexate; NRM: Non-relapse mortality; PM: Partially matched; OS: Overall survival; PB: Peripheral blood; RIC: Reduced-intensity conditioning; TBI: Total body irradiation; URD: Unrelated donor; WM: Well-matched. HLA match definitions were according to published guidelines24.
Outcomes of steroid-refractory aGVHD
Day 28 response after the initiation of therapy for steroid-refractory aGVHD was complete response (CR) in 47 (23%; 95%CI, 18–30%), partial response (PR) in 31 (15%; 95%CI, 11–21%), no response (NR) in 88 (43%; 95%CI; 36–51%), and death by day 28 in 37 (18%) patients who were treated as non-responders. At 2 years after the diagnosis of steroid-refractory aGVHD, OS was 25% (95%CI; 20–32%), and NRM was 65% (95%CI; 56–73%). The most common cause of death was steroid-refractory aGVHD (69% of deaths), followed by relapse of the underlying malignancy (18%). 70% of patients dying of aGVHD had infection identified as a secondary cause of death.
Factors associated with steroid-refractory aGVHD outcomes
Variables associated with day 28 response were GVHD prophylaxis (odds ratio [OR] = 0.8 for CNI/MTX, 0.6 for CsA/prednisone±ATG/other, and 1.0 for T-cell depletion compared to the reference of CsA/MMF), initial aGVHD to steroid-refractory aGVHD interval (OR = 1.3 for ≥14 vs. <14 days), and organ involvement (OR = 1.4 for skin only and 1.2 for gut/liver only compared to multi-organ involvement). The coefficients for all other factors shrank to 0 in lasso regression. The only variable associated with OS after steroid-refractory aGVHD was age, with a hazard ratio (HR) per decade of 1.04 for overall mortality. Similarly, age was the only variable associated with NRM (HR = 1.02 per decade).
The results of conventional regression analyses are shown in supplementary Table S2. GVHD prophylaxis regimen (with CsA/MMF being the best) and steroid-refractory aGVHD organ involvement (with multiple organ involvement being the worst), but not initial aGVHD to steroid-refractory aGVHD interval were associated with day 28 response in conventional regression. In contrast to lasso regression, UCB was associated with a worse day 28 response than matched sibling donor in conventional regression (OR = 0.2, 95%CI = 0.4–0.7). Similar to lasso regression, age was associated with overall mortality and NRM in conventional regression. In contrast to lasso regression, transplant year between 2010–2016 was associated with higher overall mortality (HR = 2.5, 95%CI = 1.2–5.4) and NRM (HR = 3.3, 95%CI = 1.5–7.4) in conventional regression.
Relationship between day 28 response and mortality
2-year OS after steroid-refractory aGVHD was highest among patients with day 28 CR (45%, 95%CI: 30–58%), followed by those with a PR (35 [19–52]%), and NR (21 [13–30]%) (Figure 1A). Day 28 response and age (per decade; as the only variable associated with OS in lasso regression) were included in a regression model for OS (Table 2). In this analysis, every decade increase in age was associated with 10% increase in the risk of death. Overall, NR at day 28 more than doubled the risk of death (HR: 2.4, 95%CI: 1.5–3.7) compared with CR. This association was independent of age. Overall, the association between day 28 PR and mortality did not appear clinically important (compared with CR, HR: 1.1, 95%CI: 0.6–1.9). However, examination of survival curves in Figure 1A suggested a more apparent difference between CR and PR patients beyond 1 year, although the group sizes in the 1-year responders were modest (24 with CR and 14 with PR). Among the 3 late deaths in the CR group, 2 were after relapse, and 1 after a GVHD flare. Among the 4 late deaths in the PR group, 3 were after a GVHD flare and 1 was infection-related.
Figure 1: 2-year overall survival and non-relapse mortality by day 28 response.
Patients with a complete remission (CR) at day 28 of steroid-refractory acute graft-versus-host disease had similar 1-year, but better 2-year overall survival (A) and non-relapse mortality (B) compared to those with a partial response (PR) or no response (NR). Patients with NR did the worst.
Table 2:
Landmark Cox regression analysis of survival
| Factor | N | HR for overall mortality (95%CI) | N | HR for non-relapse mortality (95%CI) |
|---|---|---|---|---|
| Day 28 response | ||||
| CR | 47 | 1.0 (reference) | 46 | 1.0 (reference) |
| PR | 30 | 1.1 (0.6–1.9) | 30 | 1.3 (0.7–2.3) |
| NR | 89 | 2.4 (1.5–3.7) | 87 | 2.5 (1.5–4.1) |
| Age (per decade) | 1.1 (1.0–1.2) | 1.07 (0.96–1.20) | ||
CI: Confidence interval; HR: Hazard ratio
2-year NRM after steroid-refractory aGVHD was lowest among patients with day 28 CR (43%, 95%CI: 28–59%), followed by those with a PR (58 [38–79]%), and NR (70 [5883]%) (Figure 1B). Day 28 response and age (per decade; as the only variable associated with NRM in lasso regression) were included in a regression model for NRM (Table 2). Overall, NR at day 28 more than doubled the risk of non-relapse death (HR: 2.5, 95%CI: 1.5–4.1) compared with CR. This association was independent of age. Overall, the association between day 28 PR and mortality did not appear clinically important (compared with CR, HR: 1.3, 95%CI: 0.7–2.3). However, examination of the cumulative curves in Figure 1B suggested a more apparent difference between CR and PR patients beyond 1 year.
Discussion
In a large cohort of steroid-refractory aGVHD patients with prospective and consistent GVHD scoring, we observed a 38% day 28 overall response (23% CR). This disappointing response rate, while consistent with prior reports5,6,16, highlights the shortcomings of current therapies and the importance of novel therapeutics. In the period of this study, we used ATG as the drug of choice for steroid-refractory aGVHD when a clinical trial was not available. The poor response rates in steroid-refractory aGVHD are partly due to the high frequency of lower gut involvement in steroid-refractory aGVHD, 38% in this analysis. As expected5,17,18, steroid-refractory aGVHD was highly fatal, with only a quarter of patients surviving 2 years post-diagnosis.
Our study was novel in the approach we adopted from bioinformatics when the presence of many potential predictors increases the likelihood of finding false positive results. Lasso regression is a more conservative approach than conventional regression in that it more aggressively eliminates variables that are unlikely to be associated with outcome. This approach is particularly valuable in exploratory analyses like this study with no a priori hypotheses. Lasso regression identifies variables that are most likely to be associated with outcome in order to suggest future hypothesis-driven analyses using those variables. Lasso does not produce P values or confidence intervals, making it suitable for exploratory analysis without over-emphasizing false leads. In our analyses, conventional regression overall resulted in more complex models including more variables associated with outcome.
Although day 28 response in our series seemed dependent on the choice of treatment in univariate analysis, with MMF yielding the highest response (56%) and etanercept the lowest response (20%), confidence intervals were large and the associations did not hold in lasso regression. The largest previous study (n = 58) of etanercept in steroid-refractory aGVHD reported a response rate of 38% at day 2816. However, the extent of organ involvement was, as expected5,16,19, associated with day 28 response. Patients with only skin involvement had the best response rate, followed by those with only gut or liver involvement, followed by those with multi-organ disease. Our analysis suggests that response to steroid-refractory aGVHD may also depend on the GVHD prophylactic regimen used after transplant. This observation is worthy of further investigation in other cohorts as the GVHD prophylaxis regimens used at our institution are associated with graft source and conditioning intensity. We also found that a shorter interval (<14 days) between the initial diagnosis of aGVHD and steroid-refractory aGVHD predicts poor response. Patients with a combination of all high-risk features (CsA/prednisone ± ATG as GVHD prophylaxis, multi-organ involvement, and initial aGVHD to steroid-refractory aGVHD interval <14 days) had a day 28 response rate of 29%, whereas those with a combination of all low-risk features (CsA/MMF as GVHD prophylaxis, skin only steroid-refractory aGVHD, and initial aGVHD to steroid-refractory aGVHD interval ≥14 days) had a day 28 response rate of 71%.
We found that age and day 28 response were the only variables associated with OS. Each additional decade of age increased the risk of death by about 5–10%. Age was associated with OS in some previous reports20,21. The relationship with age and poor OS may be related to cumulative comorbidities or worsening physiological reserves due to the aging process. NR at day 28 more than doubled the risk of death compared to CR, consistent with previous reports19.
In conclusion, we used an underutilized statistical tool to identify the most likely correlates of steroid-refractory aGVHD treatment outcomes among many potential variables. We find that multi-organ involvement, short interval between initial and steroid-refractory aGVHD, and GVHD prophylactic regimens other than CsA/MMF are associated with poor response, while older age and poor response predict higher mortality. The group at highest risk for poor response and/or mortality includes elderly patients developing multi-organ steroid-refractory aGVHD within 2 weeks of the initial diagnosis of GVHD. We considered the 3 most commonly used agents for the treatment of steroid-refractory aGVHD in the period of our study. Future studies with other therapeutic agents would be valuable to assess the generalizability of our results. Most importantly, this analysis further emphasizes the disappointing results from current steroid-refractory aGVHD treatments and the need for novel therapeutics. Several such therapies are being tested in clinical trials22,23.
Supplementary Material
Highlights.
In 203 steroid-refractory acute GVHD patients, day 28 response was 38%
Age was the only predictor of survival
No response at day 28 increased the risk of overall mortality 2.4-fold
Acknowledgements
Research reported in this publication was supported by NIH grant P30CA077598 utilizing the Biostatistics Core shared resource of the Masonic Cancer Center, University of Minnesota and by the National Center for Advancing Translational Sciences of the National Institutes of Health Award Number UL1-TR002494. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Conflicts of interest: None
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Zeiser R, Blazar BR. Acute Graft-versus-Host Disease — Biologic Process, Prevention, and Therapy. N. Engl. J. Med 2017;377(22):2167–2179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Rashidi A, DiPersio JF, Sandmaier BM, Colditz GA, Weisdorf DJ. Steroids Versus Steroids Plus Additional Agent in Frontline Treatment of Acute Graft-versus-Host Disease: A Systematic Review and Meta-Analysis of Randomized Trials. Biol. Blood Marrow Transplant 2016;22(6):1133–1137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.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 2002;8(7):387–94. [DOI] [PubMed] [Google Scholar]
- 4.Westin JR, Saliba RM, De Lima M, et al. Steroid-Refractory Acute GVHD: Predictors and Outcomes. Adv. Hematol 2011;2011:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Xhaard A, Rocha V, Bueno B, et al. Steroid-refractory acute GVHD: lack of long-term improved survival using new generation anticytokine treatment. Biol. Blood Marrow Transplant 2012;18(3):406–13. [DOI] [PubMed] [Google Scholar]
- 6.MacMillan ML, Weisdorf DJ, Davies SM, et al. Early antithymocyte globulin therapy improves survival in patients with steroid-resistant acute graft-versus-host disease. Biol. Blood Marrow Transplant 2002;8(1):40–6. [DOI] [PubMed] [Google Scholar]
- 7.Glucksberg H, Storb R, Fefer A, et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HL-A-matched sibling donors. Transplantation. 1974;18(4):295–304. [DOI] [PubMed] [Google Scholar]
- 8.Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant. 1995;15(6):825–8. [PubMed] [Google Scholar]
- 9.Weisdorf DJ, Snover DC, Haake R, et al. Acute upper gastrointestinal graft-versus-host disease: clinical significance and response to immunosuppressive therapy. Blood. 1990;76(3):624–9. [PubMed] [Google Scholar]
- 10.Mehta RS, Cao Q, Holtan S, MacMillan ML, Weisdorf DJ. Upper GI GVHD: similar outcomes to other grade II graft-versus-host disease. Bone Marrow Transplant. 2017;52(8):1180–1186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hings IM, Filipovich AH, Miller WJ, et al. Prednisone therapy for acute graft-versus-host disease: short- versus long-term treatment. A prospective randomized trial. Transplantation. 1993;56(3):577–80. [DOI] [PubMed] [Google Scholar]
- 12.Wasserstein RL, Schirm AL, Lazar NA. Moving to a World Beyond “ p < 0.05.” Am. Stat 2019;73(sup1):1–19. [Google Scholar]
- 13.Tibshirani R Regression Shrinkage and Selection via the Lasso. J. R. Stat. Soc. Ser. B 1996;58:267–288. [Google Scholar]
- 14.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 2015;21(4):761–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.MacMillan ML, DeFor TE, Weisdorf DJ. What predicts high risk acute graft-versus-host disease (GVHD) at onset?: identification of those at highest risk by a novel acute GVHD risk score. Br. J. Haematol 2012;157(6):732–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ma CKK, García-Cadenas I, Fox ML, et al. Poor prognosis in patients with steroid refractory acute graft versus host disease treated with etanercept: a multi-centre analysis. Bone Marrow Transplant. 2018;53(11):1478–1482. [DOI] [PubMed] [Google Scholar]
- 17.Martin PJ, Rizzo JD, Wingard JR, et al. First- and second-line systemic treatment of acute graft-versus-host disease: recommendations of the American Society of Blood and Marrow Transplantation. Biol. Blood Marrow Transplant 2012;18(8):1150–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Deeg HJ. How I treat refractory acute GVHD. Blood. 2007;109(10):4119–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jagasia M, Greinix H, Robin M, et al. Extracorporeal Photopheresis versus Anticytokine Therapy as a Second-Line Treatment for Steroid-Refractory Acute GVHD: A Multicenter Comparative Analysis. Biol. Blood Marrow Transplant 2013;19(7):1129–1133. [DOI] [PubMed] [Google Scholar]
- 20.Ragon BK, Mehta RS, Gulbis AM, et al. Pentostatin therapy for steroid-refractory acute graft versus host disease: identifying those who may benefit. Bone Marrow Transplant. 2018;53(3):315–325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Socié G, Vigouroux S, Yakoub-Agha I, et al. A phase 3 randomized trial comparing inolimomab vs usual care in steroid-resistant acute GVHD. Blood. 2017;129(5):643–649. [DOI] [PubMed] [Google Scholar]
- 22.Groth C, van Groningen LFJ, Matos TR, et al. Phase I/II Trial of a Combination of Anti-CD3/CD7 Immunotoxins for Steroid-Refractory Acute Graft-versus-Host Disease. Biol. Blood Marrow Transplant 2019;25(4):712–719. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Fløisand Y, Lazarevic VL, Maertens J, et al. Safety and Effectiveness of Vedolizumab in Patients with Steroid-Refractory Gastrointestinal Acute Graft-versus-Host Disease: A Retrospective Record Review. Biol. Blood Marrow Transplant 2019;25(4):720–727. [DOI] [PubMed] [Google Scholar]
- 24.Weisdorf D, Spellman S, Haagenson M, et al. Classification of HLA-matching for retrospective analysis of unrelated donor transplantation: revised definitions to predict survival. Biol. Blood Marrow Transplant. 2008;14(7):748–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.