Post-Transplant Events

High-dose corticosteroids with or without etanercept for the treatment of idiopathic pneumonia syndrome after allo-SCT

Article metrics

Abstract

Idiopathic Pneumonia Syndrome (IPS) is a common complication after allo-SCT and results in high mortality rates. Conventional treatment for IPS typically includes supportive care and high-dose corticosteroids (CS). Data suggests that TNF-α is important in the pathogenesis of IPS and that the TNF-α inhibitor etanercept may be useful for IPS treatment. We performed a retrospective comparison of consecutive patients treated at our center for IPS with CS only from 1999 to 2003 (group 1, n=22) or CS plus etanercept from 2004 to 2007 (group 2, n=17). In all, 18% of patients in group 1 vs 53% in group 2 were successfully taken off respiratory support and discharged from the hospital (P=0.039). OS was significantly better for recipients of CS plus etanercept (P=0.003). The estimated survival at 28 days and 2 years after IPS was 36.4% (95% CI 17–56%) and 9.1% (95% CI 2–25%) for group 1 and 88.2% (95% CI 61–97%) and 18% (95% CI 4–38%) for group 2, respectively. Our retrospective comparison suggests that the addition of etanercept to CS for IPS improves response rates and OS. However, outcomes remain limited in both groups, highlighting the need for more effective interventions to treat early and late complications of IPS.

Introduction

Idiopathic Pneumonia Syndrome (IPS) describes a spectrum of acute non-infectious lung disorders that occurs in 5–25% of all hematopoietic SCT (HSCT) recipients.1, 2, 3, 4, 5, 6 The 1993 National Heart Lung and Blood Institute consensus workshop defined IPS as progressive respiratory failure with multi-lobar pulmonary infiltrates on chest radiography or computed tomography, which are determined to be non-infectious and non-cardiogenic in nature.1 IPS typically occurs within 100 days of HSCT and is associated with high mortality rates ranging between 50 and 80%.1, 3, 4, 5, 6, 7 Historically, standard treatment approaches to IPS included supportive care, broad-spectrum antibiotics and the administration of intermediate to high-dose corticosteroids (CS).1, 2, 3, 4, 6, 8

Risk factors for developing IPS include myeloablative conditioning, TBI, advanced age and prior GVHD.3, 6, 7, 9 Once IPS occurs however, all risk groups do poorly.6 IPS has been attributed to a myriad of pulmonary insults, including cytotoxic effects of the conditioning regimen, cell mediated immune injury, inflammatory cytokines and occult infections.5, 6, 10, 11 Mouse models have provided key insights into the pathogenesis of IPS and suggest a central role for donor T-cell-derived TNF-α as a mediator of lung injury.2, 5, 7, 12, 13, 14, 15 These observations have led to the use of etanercept (Enbrel; Amgen, Thousand Oaks, CA, USA), a soluble TNF-α-binding protein, in small numbers of patients with IPS.7, 9, 16, 17 The largest reported series to date includes 15 patients with IPS after HSCT treated with combination CS plus etanercept; response rates were high with a 28-day survival rate of 73% but poor long-term outcome (1-year survival 20%). Measurements of bronchoalveolar lavage (BAL) fluid and plasma reported elevations of TNF-α with IPS onset and corresponding reductions with response to therapy.5

Our own institutional standard in the treatment of IPS has evolved over several years. Prior to 2004, our standard approach was supportive care and high-dose CS. Given the poor outcome of these patients and the evolving literature on the role of TNF-α in this disease, we changed our practice in January 2004 to include a combination of etanercept with high-dose CS in the treatment of these patients. Here we report our institutional outcomes of consecutive patients with IPS treated with and without the use of etanercept.

Materials and methods

Patients and study design

Patients with treated IPS were identified through the retrospective review of medical records in a manner approved by the Institutional Review Board of the Hospital of the University of Pennsylvania. The goal of this retrospective cohort study was to compare outcomes of patients with IPS who were treated with CS or CS plus etanercept.

We reviewed the charts of 446 consecutive adult allogeneic HSCT recipients treated at our center from January 1999 to December 2007. In all, 39 patients (8.7%) were identified with confirmed or suspected IPS. From January 1999 to December 2003, 22 consecutive patients were treated with high-dose CS for IPS (group 1) according to our standard practice. From January 2004 to December 2007, 17 consecutive patients with IPS received CS in combination with etanercept (group 2).

Definitions

Confirmed or suspected cases of IPS were based on NHLBI working group criteria.1 Briefly, these criteria require evidence of widespread alveolar injury (including multi-lobar infiltrates on chest radiography, symptoms and signs of pneumonia and evidence of increased alveolar-to-arterial oxygen gradient) with no other suspected or identified cause for these findings, such as heart failure or bacterial, fungal or viral infection based on cultures or other assays. Attempts were always made to rule out other causes of pulmonary toxicity and infection and all patients were on empiric antibiotics for bacterial and fungal pathogens. BAL was performed in some but not all cases and was not part of the criteria for inclusion in this analysis.

GVHD was graded according to the modified Glucksberg criteria.18 Neutrophil engraftment was defined as the first of three consecutive days that an ANC was greater than 500/μL.

IPS treatment

Prior to January 2004, patients with a diagnosis of IPS were treated with 1 g of methylprednisolone for 3 days followed by a taper of 50% every 3 days until patients were on 1 mg/kg/day. Subsequent tapering of steroids was determined by the patient's physicians dependent on the clinical scenario. After January 2004, patients with a diagnosis of IPS received the same course of methylprednisolone with the addition of etanercept given at 25 mg s.c. twice weekly for 4 weeks for an intended course of eight doses.7

Statistical analysis

The primary objective of this retrospective review was to compare the two treatment groups with respect to survival at day +28 after initiation of treatment for IPS. Secondary objectives included identifying differences in hospital length of stay, days on mechanical ventilation, use of supplemental oxygen, 6-month and 2-year OS, treatment-related complications and causes of death.

We compared distributions of categorical baseline variables and outcomes between treatment groups using Fisher's exact test. We computed 95% exact Clopper–Pearson confidence intervals for binomial proportions. We computed medians of continuous uncensored distributions and compared distributions of uncensored continuous variables between treatment groups using the Mann–Whitney test. We tested for differences in censored continuous distributions using the log-rank test. All analyses were performed in SAS version 9.2 (SAS Institute, Cary, NC, USA).

Results

Patient characteristics

The characteristics of the 39 patients diagnosed with IPS during the study period are summarized in Table 1. Between January 1999 and December 2007, the incidence of IPS at our center was 8.7% with a median time of onset of 12 days (range 0–46 days) after HSCT. During the time period of this study, 161 patients received reduced intensity conditioning regimens prior to SCT with four cases of IPS (2.5%). In contrast, 288 patients received conventional myeloablative therapy accounting for 35 cases of IPS with an incidence more typical of the reported literature at 12%. Among patients who developed IPS, 90% received conventional myeloablative conditioning regimens and 10% received reduced intensity conditioning regimens.

Table 1 Patient characteristics

TBI (1200 cGy) was used in 67% (26/39) of all cases and in 74% (26/35) of all cases that received myeloablative conditioning. Prophylaxis for GVHD included a calcineurin inhibitor with MTX in 77% and a calcineurin inhibitor with methylprednisolone in 13%; all recipients of thiotepa/CY/TBI conditioning received grafts depleted of T cells and CYA with CS as further GVHD prophylaxis.19

Except for the year of transplant, patients with IPS treated with CS alone (group 1) vs those treated with CS plus etanercept (group 2) were largely comparable, except that the stem cell source was peripheral blood in 32% of patients in group 1 and 88% patients in group 2 (P<0.001) largely due to practice changes during this time period. Recipients of unrelated donor grafts before 2004 were conditioned with thiotepa, CY and TBI, and received CYA and CS as GVHD prophylaxis accounting for the differential use of this regimen and GVHD prophylaxis in group 1.19 The median engraftment time was 14 days for group 1 and 16.5 days in group 2 (P=0.63). There were no significant differences in recipient age, indication for HSCT, use of sibling vs unrelated donor grafts, HLA matching, conditioning regimen intensity, sex mismatching or CMV status.

Overall, 25 of 39 patients underwent BAL evaluations including 54% patients in group 1 and 76% of patients in group 2 (P=0.194).

IPS presentation

The presentation of IPS was similar between both groups of patients. Patients in group 1 experienced the onset of IPS at a median of 11.5 days after HSCT (range 0–46 days) whereas patients in group 2 experienced the onset of IPS at a median of 12 days after HSCT (range 3–26 days; P=1). Patients in group 2 were more likely to develop IPS within 7 days of engraftment; IPS occurred within 7 days of engraftment in 9/22 patients (40.9%) in group 1 compared with 13/17 (76%) in group 2 (P=0.05). Infection preceded the onset of IPS in 13/22 patients (68.4%) in group 1 and in 12/17 patients (70.6%) in group 2; P=1.00. The median time between infection and the onset of IPS was 4 days (range 1–12 days) and 2 days (range 1–17 days) for groups 1 and 2, respectively; P=0.30.

In all cases, aggressive attempts were made to rule out other causes of hypoxia and pulmonary infiltrates, and infection was not considered the cause of pulmonary infiltrates. Blood and sputum cultures were performed in all cases. No prior infection was found in 6/22 patients in group 1 and 5/17 patients in group 2. Bacterial infections occurred most often in both groups, with 7/22 patients in group 1 and 10/17 patients in group 2. Infection data was unavailable for 9 patients in group 1. There were no cases of mycobacterial infections identified. As noted, comparisons of infections in this study is difficult as the etanercept-treated patients survived longer often on immune suppression and therefore were at longer risk of infection.

All patients had imaging with chest X-ray and CT imaging was done on 2/22 patients in group 1 and 14/17 patients in group 2. Unfortunately, CT imaging data was unavailable for 10/22 patients in group 1, however 5/10 patients with unknown CT imaging had confirmed diagnosis of IPS by a negative BAL procedure. BAL was performed in 12/22 (54%) patients in group 1 and 13/17 (76%) patients in group 2 (P=0.495).

IPS treatment

All patients were initially treated with 1 g of i.v. methylprednisolone for 3 days followed by a 50% dose reduction every 3 days until a dose of 1 mg/kg/day was reached. Steroids were further tapered or discontinued at the discretion of the treating physician depending on response and additional factors such as the presence or absence of GVHD. Group 1 received a median duration of 12 days (range 3–44 days) of CS therapy compared with 43 days (range 11–168 days) in group 2 (P=0.003). This difference was largely due to the longer survival of patients in group 2.

In group 2, etanercept was started in all patients within 3 days of initiating CS treatment and continued for a median of six of eight planned doses of etanercept (range 2–8 doses).

Response to treatment

The duration of mechanical ventilation was similar in both groups; group 1 received a median of 10 days (range 0–43 days) and group 2 received a median of 10 days (range 0–32 days, P=0.456) of mechanical ventilation. In group 1, 17/22 (77%) patients were intubated (data unavailable for 4 patients and 1 patient did not require intubation), compared with 12/17 (71%) patients in group 2 (P=0.721). However, more patients were terminally extubated in group 1 (14/22; 64%) compared with group 2 (3/17; 18%; P=0.008). Only 4/22 (18%) patients in group 1 as opposed to 9/17 (53%) in group 2 (P=0.039) were successfully taken off respiratory support and discharged from the hospital.

Patients in both groups had similar durations of oxygen support (nasal cannulae or mechanical ventilation) requiring a median duration of 16 days (range, 1–65 days) and 18 days (range, 5–58 days) for groups 1 and 2, respectively (P=0.593). Group 1 required additional oxygen support after intubation for a median of 15 days compared with 11 days in group 2. The durations of oxygen support and mechanical ventilation were influenced by both improvement with successful extubation as well as terminal extubation and death.

GVHD and other transplant-related complications

Overall, 56% (95% CI 40–72%) of all patients developed GVHD. Grade III–IV acute GVHD developed in 41% (95% CI 26–58%) of all patients. The likelihood of developing any GVHD and grade III/IV GVHD was similar between groups (P=1.00) as noted in Table 2. In addition, the risk of GVHD developing before or after occurrence of IPS was similar in both groups (P=0.69).

Table 2 Transplant-related complications

Other transplant-related complications developing in these patients are outlined in Table 2. Patients in group 1 were more likely to develop veno-occlusive disease (group 1, 10/22 patients; group 2, 2/17 patients; P=0.012) and acute renal failure (group 1 13/22 patients; group 2 2/17 patients; P=0.007).

Length of stay

Patients in group 1 had a median length of stay in the intensive care unit of 10 days (range 1–43 days) compared with 9 days (range 0–34 days) in group 2 (P=0.667). Recipients of CS alone had a median total hospital stay of 25.5 days (range, 7–63 days), whereas recipients of CS plus etanercept had a median total hospital stay of 43 days (range, 21–99 days, P=0.012). The longer length of stay was likely due to the prolonged survival of etanercept recipients; as mentioned earlier, 4/22 patients (18.2%) in group 1 were discharged from the hospital compared with 9/17 patients (52.9%) in group 2 (P=0.039).

Survival

Kaplan–Meier curves of the estimated distribution of survival after onset of IPS for the two groups appear in Figure 1. A logrank test comparing the survival curves between groups gave a P-value of 0.003. The estimated survival probability 28 days after IPS for all patients was 59% and was significantly better (P<0.001) for recipients of CS plus etanercept (group 2; 88.2%, 95% CI 61–97%) compared with recipients of steroids alone (group 1; 36.4%, 95% CI 17–56%). Day 180 survival after IPS onset for all patients was 15% without a statistically significant difference between groups; estimates are 9.1% (95% CI 2–25%) for group 1 vs 23.5% (95% CI 7–45%) for group 2 (P=0.374). The estimated 2-year survival was 13% for all patients including 9.1% (95% CI 2–25%) for group 1 and 18% (95% CI 4–38%) for group 2 (P=0.636).

Figure 1
figure1

Kaplan–Meier estimates of OS following etanercept and CS. A logrank test comparing the survival curves between groups gave a P-value of 0.003. The estimated survival probability at day 28 for recipients of CS plus etanercent (group 2; 88.2%, 95% CI 61–97%) was significantly better compared with recipients of steroids alone (group 1; 36.4%, 95% CI 17–56%) (P<0.001).

Cause of death

In group 1, multi-system organ failure (n=10; 45%) was the most common cause of death, followed by respiratory failure (n=7; 32%). In group 2, multi-system organ failure accounted for death in 29% of patients and respiratory failure accounted for death in 35% of patients. GVHD was the cause of death in one patient in group 1. These differences in cause of death between these groups were not statistically significant (P=0.614). Of note, only 1 patient in group 1 had relapse of disease at the time of death and no patients in group 2 experienced relapse of disease. In addition, there was no significant difference in early cause of death (death prior to day +100) between group 1 and group 2 in terms of multi-system organ failure (P=0.22) and respiratory failure (P=0.768).

Discussion

The treatment of IPS has been generally limited to high-dose CS, aggressive supportive care and empiric antimicrobial therapy, but patient outcomes have been poor, with reported mortality rates of 50–94% and rapid progression to death often within 9 to 14 days of onset.4, 6, 20, 21

Although the pathogenesis of IPS is not well defined, TNF-α production from donor monocytes, macrophages and T cells appears to have an integral role in the development of IPS.2, 15, 22 To determine if anti-TNF therapy could improve outcomes in patients with IPS after allo-SCT, etanercept in combination with CS has been reported in several small case series.7, 9, 17 More recently, Yanik et al.5 reported the results of a prospective trial involving 15 IPS patients treated with CS plus etanercept. Complete response, defined as the elimination of oxygen support, was observed in 67% of patients with a median time to response of 7 days. Additionally, 11/15 patients (73%) were alive at day +28 and 60% survived to hospital discharge. Unfortunately, only 20% of patients survived beyond 8 months; most deaths within 100 days were related to IPS and infection and deaths beyond day 100 were related to GVHD, relapse and heart failure, findings remarkably similar to our findings. Initial improved survival rates without improvement in long-term survival have been observed in other smaller series as well.9, 23

Based on poor outcomes with high-dose CS and emerging data regarding the pathophysiology of IPS, in 2004 we began treating patients having a clinical diagnosis of IPS with a combination of high-dose CS and etanercept. In this study, we have compared outcomes in 22 consecutive patients treated at our center with high-dose CS to 17 recipients of similar doses of CS plus etanercept. The groups were treated during different time periods but were otherwise generally well matched except that patients in the earlier time periods were more likely to have received conditioning therapy with thiotepa, CY and TBI with T-cell depleted stem cell grafts, and to have received BM rather than a PBSC grafts. The incidence of IPS during the two time periods was similar suggesting that other practice patterns did not change the risk of IPS and were unlikely to influence the cause of pulmonary toxicity or response to therapy. Our patient population was also typical of patients in other reports who develop IPS in regards to overall incidence (8.7%), median time of onset (12 days), and incidence after the use of conventional myeloablative conditioning regimens (12%).6, 7, 9, 17, 21

To date, we are not aware of other reports directly comparing high-dose CS therapy to CS therapy plus anti-TNF-α treatment for IPS. Our analysis shows that combination therapy with CS and etanercept improves survival compared with CS alone (Figure 1, P=0.003). The duration of oxygen therapy, time spent in the intensive care unit and days on mechanical ventilation were similar in groups 1 and 2. The lack of observed differences in these outcomes may be explained by early deaths in recipients of CS only (group 1). This would translate into a shorter duration of mechanical ventilation and length of stay in the intensive care unit rather than a response to therapy. This hypothesis is supported by the fact that recipients of CS plus etanercept (group 2) had a longer stay in the hospital and a greater proportion of patients survived to hospital discharge. As some patients did not require mechanical ventilation, it was not possible to calculate ‘ventilator-free days’.24 Although the duration of total CS therapy was found to be significantly different between group 1 (12 days) and group 2 (43 days) (P=0.003); this difference is also likely explained by early mortality after IPS onset in group 1 rather than better responses.

The incidence of grade III/IV acute GVHD was similar in both groups. Furthermore, the occurrence of GVHD prior to IPS onset and distribution of organ involvement were also similar in both groups. Whereas etanercept has been hypothesized to minimize GVHD and is being used in phase-II clinical trials as prophylactic therapy, its effect on preventing severe GVHD remains unknown. In addition, the number of patients in our study was quite small to get an accurate estimate of the impact of etanercept. Finally, the early death in group 1 and longer survival in group 2 (and thus, higher risk of severe GVHD) would have been competing risks. Therefore we cannot make any conclusions about the role of etanercept in preventing or treating GVHD.

Despite a survival benefit with the addition of etanercept to CS, the overall prognosis for these patients remains poor with only 18% of patients in group 2 surviving at 2 years after IPS onset. This long-term survival data is similar to outcomes reported by others in both untreated patients with IPS and in recipients of anti-TNF therapy and CS.9, 16, 25

Etanercept has a half life of elimination of 72–132 h (3–5 days), however it is not known if ongoing exposure is necessary to prevent recurrence of IPS and if a longer duration of use would lower the risk of recurrent IPS. Tun et al.17 described two cases of late-onset IPS after non-myeloablative allo-SCT. Both patients responded to initial etanercept therapy; however each patient experienced a relapse of dyspnea and infiltrates 1 month after etanercept discontinuation. Re-initiation of etanercept therapy produced a rapid resolution of clinical and radiographic symptoms. In our present study, four patients who had previously demonstrated a response to etanercept treatment experienced a relapse of respiratory distress at a median of 41 days after the last dose of etanercept. None of the patients were restarted on etanercept.

Our study also found that patients treated with CS only (group 1) were more likely to develop veno-occlusive disease (P=0.04) and acute renal failure (P=0.003). This raises the concern that this patient population was sicker and at higher risk for organ toxicity. This seems unlikely as (1) patients characteristics were similar; (2) the incidence of IPS was similar among all transplant patients during the time periods studied; (3) the presentation and time of onset of IPS were similar; and (4) outcomes in our patients were similar to other reported case series. The reason for the higher incidence of veno-occlusive disease in group 1 is not known. It is possible that etanercept affected the development of these complications in patients treated in later years, a hypothesis that would best be tested in randomized trials and through comparisons of larger numbers of patients. However, it may be due to differences in conditioning therapies used in this group of patients, different patient selection, or other difficult-to-identify differences.

The major limitations to this analysis are the retrospective nature of the study, the comparison of patients treated during different time periods and the small numbers of patients studied. There were some differences in the conditioning regimen and GVHD prophylaxis given to these patient groups; these may have influenced the risk of IPS and other organ toxicity (and hence survival) but it is unclear if they would affect the response to IPS therapy. Due to practice changes during our study period, there was a difference in stem cell source between the two groups, with greater use of PBSC in group 2. According to the recently published statement on IPS by the American Thoracic Society, the impact of stem cell source on the development of IPS has not been fully evaluated.26 The small number of patients treated in each group precludes a detailed analysis of characteristics that might predict for, or be associated with, response and survival.

The diagnosis of IPS was made on clinical grounds after ruling out other possible causes and was confirmed by chart review for this study. Although examination of BAL fluid is not a requirement for a definition of IPS, and hence BAL was not required, it was performed in 59% of patients in group 1 and 76% of patients in group 2 (P=0.495). There were a number of reasons BAL may not have been performed. Several patients declined and often patients were clinically unstable and could not tolerate BAL without subsequent intubation. In addition, the sensitivity of BAL, particularly for fungal infections is often low, and bacterial and viral pneumonias were often ruled out on clinical grounds and through other less invasive testing (quantitative CMV testing, testing for common respiratory viruses, sputum samples, blood cultures, etc). All patients had chest X-rays and CT imaging was done to identify other sources of infection if clinically indicated. Heart failure was possible to exclude via clinical findings and echocardiograms. As there was no significant difference in survival in patients who did or did not undergo BAL, we believe that performance of BAL was not a measure of lower acuity and is unlikely to explain differences in outcomes between patient groups. It is also notable that the recently closed randomized trial studying the role of etanercept for IPS did not require BAL for a diagnosis of IPS.

Another issue was that we retrospectively identified only patients with a diagnosis of IPS; it was not possible to determine how many patients were evaluated for IPS and found to have other causes of lung injury or pulmonary infiltrates. Furthermore, we cannot definitively exclude the possibility that other supportive care practice changes in later years accounted for the improved responses and survival differences in recipients of etanercept (such as improved antibiotics, diagnostic methods and changes in intensive care support). Despite these limitations, the early outcomes for recipients of etanercept and CS were significant and better than our historical experience and from other published outcomes. We could not identify any other obvious change in practice to account for these differences.

The relationship of infection to development of IPS is complex. As highlighted by others, it is critical to rule out the possibility of infectious causes of pulmonary toxicity.1, 6, 16 Syndromes such as GVHD and infection may elicit an immune response and activate inflammatory cytokines that may contribute to lung injury and multi-organ failure syndromes.4 Therefore, it is logical that infections may be associated with subsequent IPS; infections were common in our patients. There were no differences in infections between the treatment groups however (P=1.0). The time from the first positive blood or urine cultures to the onset of IPS was also similar between both groups. In all cases, the identified infection was not thought to be the cause of pulmonary infiltrates. We believe that any identified or presumed infection was ‘under control’ at the time of etanercept. Of course, many patients were critically ill due to interstitial pneumonitis, but it was the clinical impression that known infections were being appropriately treated.

One noted difference in our patient groups was the timing of IPS in relationship to engraftment. In group 1, 9/22 patients (41%) had neutrophil engraftment within 7 days of onset of IPS compared with 13/17 patients (77%) in group 2 (P=0.05). The reason for this difference is unclear, as there was no difference in median time to engraftment between groups 1 and 2 (14 days vs 16.5 days; P=0.63). Peri-engraftment respiratory distress syndrome (PERDS) has been included as a clinical subset of IPS in various reports.8, 21, 26 PERDS is characterized by fever, dyspnea, bilateral interstitial infiltrates, and most notably, early onset within 5 to 7 days of engraftment after autologous transplant and there is likely significant overlap between these syndromes.21 IPS is also often described as occurring near the time of engraftment.5 PERDS may also be treated with high doses of CS though reported outcomes may be better than other manifestations of IPS with reported mortality rates of 26%.27 Although PERDS has been described to be more responsive to CS alone, TNF-α secreted by donor T cells is still a significant contributor to lung injury in the acute setting. In the murine model, acute pulmonary toxicity within the first 2 weeks of HSCT is caused by the influx of host monocytes and donor T cells into the lungs. The etanercept group may have an increased incidence of the PERDS subtype, which could have influenced the onset of IPS and response to treatment.

Though a randomized phase III trial comparing these approaches was initiated by the Blood and Marrow Transplant Clinical Trials Network (Clinicaltrials.gov identifier NCT00421174), the trial was closed early due to slow accrual. Definitive randomized results will not be available in the foreseeable future, and for now we must rely on phase II data and retrospective comparisons to guide treatment considerations.

Despite the retrospective nature of our analysis, our results suggest that treatment of IPS with combination high-dose CS and anti-TNF directed therapy with etanercept results in improved OS compared with treatment with CS alone, and leads to more frequent discharges from both the intensive care unit and the hospital. Unfortunately, long-term survival remains limited regardless of treatment. This clearly demonstrates that even more effective interventions are needed to treat IPS after allogeneic SCT.

References

  1. 1

    Clark JG, Hansen JA, Hertz MI, Parkman R, Jensen L, Peavy HH . NHLBI workshop summary. Idiopathic pneumonia syndrome after bone marrow transplantation. Am Rev Respir Dis 1993; 147 (6 Pt 1): 1601–1606.

  2. 2

    Cooke KR, Hill GR, Gerbitz A, Kobzik L, Martin TR, Crawford JM et al. Tumor necrosis factor-alpha neutralization reduces lung injury after experimental allogeneic bone marrow transplantation. Transplantation 2000; 70: 272–279.

  3. 3

    Crawford SW, Longton G, Storb R . Acute graft-versus-host disease and the risks for idiopathic pneumonia after marrow transplantation for severe aplastic anemia. Bone Marrow Transplant 1993; 12: 225–231.

  4. 4

    Kantrow SP, Hackman RC, Boeckh M, Myerson D, Crawford SW . Idiopathic pneumonia syndrome: changing spectrum of lung injury after marrow transplantation. Transplantation 1997; 63: 1079–1086.

  5. 5

    Yanik G, Ho V, Levine J, White E, Braun T, Antin J et al. The impact of soluble tumor necrosis factor receptor etanercept on the treatment of idiopathic pneumonia syndrome after allogeneic hematopoietic stem cell transplantation. Blood 2008; 112: 3073–3081.

  6. 6

    Fukuda T, Hackman R, Guthrie K, Sandmaier B, Boeckh M, Maris M et al. Risks and outcomes of idiopathic pneumonia syndrome after nonmyeloablative and conventional conditioning regimens for allogeneic hematopoietic stem cell transplantation. Blood 2003; 102: 2777–2785.

  7. 7

    Yanik G, Hellerstedt B, Custer J, Hutchinson R, Kwon D, Ferrara JL et al. Etanercept (Enbrel) administration for idiopathic pneumonia syndrome after allogeneic hematopoietic stem cell transplantation. Biol Blood Marrow Transplant 2002; 8: 395–400.

  8. 8

    Afessa B, Peters S . Noninfectious pneumonitis after blood and marrow transplant. Curr Opin Oncol 2008; 20: 227–233.

  9. 9

    Keates-Baleeiro J, Moore P, Koyama T, Manes B, Calder C, Frangoul H . Incidence and outcome of idiopathic pneumonia syndrome in pediatric stem cell transplant recipients. Bone Marrow Transplant 2006; 38: 285–289.

  10. 10

    Cooke KR, Yanik G . Acute lung injury after allogeneic stem cell transplantation: is the lung a target of acute graft-versus-host disease? Bone Marrow Transplant 2004; 34: 753–765.

  11. 11

    Clark JG, Madtes DK, Martin TR, Hackman RC, Farrand AL, Crawford SW . Idiopathic pneumonia after bone marrow transplantation: cytokine activation and lipopolysaccharide amplification in the bronchoalveolar compartment. Crit Care Med 1999; 27: 1800–1806.

  12. 12

    Cooke KR, Kobzik L, Martin TR, Brewer J, Delmonte J, Crawford JM et al. An experimental model of idiopathic pneumonia syndrome after bone marrow transplantation: I. The roles of minor H antigens and endotoxin. Blood 1996; 88: 3230–3239.

  13. 13

    Piguet PF, Grau GE, Collart MA, Vassalli P, Kapanci Y . Pneumopathies of the graft-versus-host reaction. Alveolitis associated with an increased level of tumor necrosis factor mRNA and chronic interstitial pneumonitis. Lab Invest 1989; 61: 37–45.

  14. 14

    Shankar G, Bryson JS, Jennings CD, Kaplan AM, Cohen DA . Idiopathic pneumonia syndrome after allogeneic bone marrow transplantation in mice. Role of pretransplant radiation conditioning. Am J Resp Cell Mol Biol 1999; 20: 1116–1124.

  15. 15

    Carlson M, West M, Coghill J, Panoskaltsis-Mortari A, Blazar B, Serody J . In vitro-differentiated TH17 cells mediate lethal acute graft-versus-host disease with severe cutaneous and pulmonary pathologic manifestations. Blood 2009; 113: 1365–1374.

  16. 16

    Yanik GA, Ho VT, Levine JE, White ES, Braun T, Antin JH et al. The impact of soluble tumor necrosis factor receptor etanercept on the treatment of idiopathic pneumonia syndrome after allogeneic hematopoietic stem cell transplantation. Blood 2008; 112: 3073–3081.

  17. 17

    Tun HW, Wallace KH, Grinton SF, Khoor A, Burger CD . Etanercept therapy for late-onset idiopathic pneumonia syndrome after nonmyeloablative allogeneic hematopoietic stem cell transplantation. Transplant Proc 2005; 37: 4492–4496.

  18. 18

    Przepiorka D, Weisdorf D, Martin P, Klingemann HG, Beatty P, Hows J et al. 1994 Consensus conference on acute GVHD grading. Bone Marrow Transplant 1995; 15: 825–828.

  19. 19

    Bunin N, Aplenc R, Leahey A, Magira E, Grupp S, Pierson G et al. Outcomes of transplantation with partial T-cell depletion of matched or mismatched unrelated or partially matched related donor bone marrow in children and adolescents with leukemias. Bone Marrow Transplant 2005; 35: 151–158.

  20. 20

    Zhu K-E, Hu J-Y, Zhang T, Chen J, Zhong J, Lu Y-H . Incidence, risks, and outcome of idiopathic pneumonia syndrome early after allogeneic hematopoietic stem cell transplantation. Eur J Haematol 2008; 81: 461–466.

  21. 21

    Yanik G, Cooke K . The lung as a target organ of graft-versus-host disease. Semin Hematol 2006; 43: 42–52.

  22. 22

    Hildebrandt G, Olkiewicz K, Corrion L, Chang Y, Clouthier S, Liu C et al. Donor-derived TNF-alpha regulates pulmonary chemokine expression and the development of idiopathic pneumonia syndrome after allogeneic bone marrow transplantation. Blood 2004; 104: 586–593.

  23. 23

    Frangoul H, Koyama T, Domm J . Etanercept for treatment of idiopathic pneumonia syndrome after allogeneic hematopoietic stem cell transplantation. Blood 2009; 113: 2868–2869; author reply 2869.

  24. 24

    Schoenfeld D, Bernard G . Statistical evaluation of ventilator-free days as an efficacy measure in clinical trials of treatments for acute respiratory distress syndrome. Crit Care Med 2002; 30: 1772–1777.

  25. 25

    Frangoul H, Koyama T, Domm J . Etanercept for treatment of idiopathic pneumonia syndrome after allogeneic hematopoietic stem cell transplantation. Blood 2009; 113: 2868–2869.

  26. 26

    Panoskaltsis-Mortari A, Griese M, Madtes DK, Belperio JA, Haddad IY, Folz RJ et al. An Official American Thoracic Society Research Statement: noninfectious lung injury after hematopoietic stem cell transplantation: idiopathic pneumonia syndrome. Am J Respir Crit Care Med 2011; 183: 1262–1279.

  27. 27

    Capizzi SA, Kumar S, Huneke NE, Gertz MA, Inwards DJ, Litzow MR et al. Peri-engraftment respiratory distress syndrome during autologous hematopoietic stem cell transplantation. Bone Marrow Transplant 2001; 27: 1299–1303.

Download references

Acknowledgements

This work was supported in part by the following grants: NIH K24 CA11787901 (DLP) and University of Pennsylvania Cancer Center Core Grant: P30-CA016520 (DFH, KST)

Author contributions: RT, NF, JD, MV and DLP designed research; RT, NF, JD, MV and DLP performed research; DFH and KST designed the statistical section and performed the statistical analyses; RT and DLP wrote the paper; all authors analyzed data and reviewed and edited the manuscript.

Author information

Correspondence to D L Porter.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • Idiopathic Pneumonia Syndrome
  • etanercept
  • TNF-α
  • allo-SCT

Further reading