Pericardial effusion post-SCT in pediatric recipients with signs ... - Nature

Bone Marrow Transplantation (2011) 46, 529–538 & 2011 Macmillan Publishers Limited All rights reserved 0268-3369/11

www.nature.com/bmt

ORIGINAL ARTICLE

Pericardial effusion post-SCT in pediatric recipients with signs and/or symptoms of cardiac disease M Neier1, Z Jin2, C Kleinman1, L Baldinger1, M Bhatia1, E Silver1, C van de Ven1, E Morris1, P Satwani1, D George1, J Garvin1, MB Bradley1, J Schwartz3 and MS Cairo1,3,4 1 Department of Pediatrics, Medicine and Pathology, Morgan Stanley Children’s Hospital of New York-Presbyterian, Columbia University, New York, NY, USA; 2Department of Biostatistics, Columbia University, New York, NY, USA; 3Department of Pathology, Columbia University, New York, NY, USA and 4Department of Medicine, Columbia University, New York, NY, USA

The objective of this study was to assess the incidence, risk factors, outcome and impact on OS of pericardial effusion (PEF) in a cohort of 156 pediatric SCT recipients. The mean age was 8.15±6.25 years. In all, 74% of the patients had malignant disease and 35% of the patients received autologous stem cell grafts. Twenty-three subjects developed effusion at 2.75±3.54 months after SCT. The overall probability of developing a PEF after SCT was 16.9%. In the multivariate analysis of risk factors associated with time to PEF, increased age, allogeneic risk status and conditioning type, were all significant factors. In a multivariate analysis of time to death, PEF, CMV status and risk status were all independent risk factors. PEF, however, had the highest HR of 3.334. Of the 23 patients with PEF, 19 died (82.6%); however, none died as a direct result of pericardial tamponade. In summary, our results suggest that PEF is a significant risk factor for post transplant mortality. These results suggest a need for more frequent evaluation and monitoring for development of PEF. Studies are needed to determine the etiology of, and new therapeutic strategies for, PEF in the post-SCT population. Bone Marrow Transplantation (2011) 46, 529–538; doi:10.1038/bmt.2010.149; published online 28 June 2010 Keywords: pediatrics; SCT; pericardial effusion; survival; cardiac

Correspondence: Dr MS Cairo, Department of Pediatrics, Medicine and Pathology, Division of Pediatric Blood and Marrow Transplantation, Morgan Stanley Children’s Hospital of New York-Presbyterian, Columbia University, 3959 Broadway, CHN 10-03, New York, NY 10032, USA. E-mail: [email protected] This work was presented in part at the American Society of Pediatric Hematology/Oncology, Toronto, Canada, May 2007 and American Society of Blood and Marrow Transplantation, San Diego, CA, February 2008. Received 7 December 2009; revised 10 May 2010; accepted 13 May 2010; published online 28 June 2010

Introduction SCT is the treatment of choice for a variety of pediatric malignant and non-malignant conditions.1 Pediatric SCT recipients, however, may develop a variety of side effects and morbidities post-SCT. After SCT, recipients are at risk of non-engraftment, infection, GVHD and other specific organ complications, including cardiac toxicity. One study reported an incidence of life-threatening cardiac toxicity of less than 2% after SCT.2 Pericardial effusion (PEF), a potentially morbid condition, is becoming a more recognized and frequently described complication that may not be associated with cardiac toxicity per se.3–10 PEF after SCT was first reported in 1987 by Veys et al.10 They reported the case of a young woman with ALL who developed a PEF after conditioning for allogeneic BMT. They theorized that the effusion was secondary to cumulative cardiotoxicity from chemotherapy and radiation therapy.10 Einsele et al.3 also reported a patient with acute leukemia who developed a PEF after SCT. This patient, however, developed a PEF later in his post transplant course. At that time, the patient had evidence of both CMV infection and leukemia relapse, which were identified as the etiology of his PEF.3 There have been numerous other case reports6,8,9 but only one single retrospective review.5 In this review, Rhodes et al.5 reported that the incidence of PEF in pediatric recipients after SCT was approximately 4.4%. This study indicated an association between PEF, GVHD and allogeneic SCT. Other suggested etiologies of PEF have included infection, transplant conditioning, hemolytic uremic syndrome and relapsed disease.4,7,11–13 We therefore conducted a review of 200 consecutive SCTs in 156 pediatric recipients. We determined the time and incidence of PEF, factors associated with PEF and the impact of PEF on OS.

Materials and methods Patients We performed a retrospective Institutional Review Boardapproved review of 158 patients who received 200 SCTs at Morgan Stanley Children’s Hospital of New York-Presbyterian

Pericardial effusion in pediatric SCT recipients M Neier et al

530

between 1 January 2000 and 22 September 2005. Two of the 158 patients had a PEF before the transplant and were ineligible for analysis. Allogeneic SCT recipients were defined as either standard or high-risk. High-risk disease was defined as one of the following pre-transplant characteristics: (a) active leukemia in relapse, (b) active leukemia after induction failure, (c) leukemia in third CR or beyond, (d) progressive solid tumor and/or (e) second allogeneic SCT. All patients had to meet the minimum standards for adequate organ function, including appropriate glomerular filtration rate for age and adequate cardiac function measured by echocardiogram as described below.

HLA typing of patient and allogeneic donors HLA typing for patients receiving stem cells from matched family donors or umbilical cord blood donors was performed by serology for HLA-A and -B and by highresolution DNA typing for DRB1 as we previously described.14 Patients receiving unrelated adult donor PBSCs or BM also had high-resolution DNA typing of HLA-C and HLA-DQB1 as we have previously described.14 Matched family donors were required to be a 6/6 or 5/6 HLA match, and unrelated umbilical cord blood donor units had to be a 6/6, 5/6 or 4/6 HLA match. Umbilical cord blood donor units were also required to have a minimum of 1.5 107 nucleated cells of cryopreserved umbilical cord blood units per kg body weight of the recipient. Unrelated adult donors were required to have a minimum of 5 106 CD34 cells/kg and to be an 8/10, 9/10 or 10/10 HLA match (A, B, C, DRB1 and DQB1). For autologous hematopoietic SCT, PBSCs were collected using apheresis. The end point for successful PBSC harvesting was a collection of 5 106 CD34 cells/kg. For neuroblastoma patients, in case of detectable tumor cells in the product by immunocytology, CD34 cells were selected using the Isolex 300i system (Baxter Immunotherapy, Round Lake, IL, USA) or the CliniMacs system (Miltenyi Biotec, Auburn, CA, USA).15 Conditioning regimens Conditioning regimens were myeloablative in 71.6% of patients, including all autologous HSCT recipients, and non-myeloablative (reduced intensity allogeneic SCT) in 28.4% (Table 1). A variety of conditioning regimens were used. Myeloablative regimens consisted primarily of thiotepa/carboplatin, carmustine/etoposide/CY or were based in TBI or BU in combination with CY or melphalan. All RI regimens were fludarabine-based (Table 1). Engraftment Neutrophil engraftment was defined as the first of 3 days after the neutrophil nadir with an ANC 4500/mm3. Plt engraftment was defined as the first of three consecutive days showing a plt count X20 000/mm3, after a 7-day period of plts being X20 000/mm3 without transfusions. GVHD prophylaxis Tacrolimus and mycophenolate mofetil were administered as GVHD prophylaxis in the majority of allogeneic SCT Bone Marrow Transplantation

Table 1

Conditioning regimens Myeloablative

Total Thiotepa/Carboplatin Bu/Cy Bu/Mel TBI/Mel TBI/Cy Cy/VP-16/Carmustine Cy/Flu Cy/Other Other

Non-myeloablative N ¼ 111 39 17 3 18 11 10 6 4 3

Total Bu/Flu Cy/Flu Mel/Flu TBI/Flu

N ¼ 44 26 15 2 1

Abbreviations: Flu ¼ fludarabine; Mel ¼ melphalan; VP-16 ¼ etoposide. Flu/CY-based regimens were classified as myeloablative or non-myeloablative based on the dose of CY and/or the use of other agents. The base regimens listed above were variably combined with other agents. One patient (SCID) received no pre-transplant conditioning.

recipients, as we have previously described.14 Tacrolimus was administered i.v. at 0.03 mg/kg/day as a continuous infusion or orally at 0.12 mg/kg/day in two divided doses starting on the first day of conditioning or day 1 (protocoldependent). As we previously described, tacrolimus dosing was adjusted to maintain the tacrolimus steady state or trough concentrations of 5–20 ng/mL (whole-blood ELISA). Mycophenolate mofetil was administered daily starting on day 1 at a dose of 15 mg/kg/dose i.v. or orally twice daily, as we previously described. Beginning in January 2002, mycophenolate mofetil doses were adjusted to maintain mycophenolic acid trough concentrations within a reference range of 1–3.5 g/mL. Mycophenolate mofetil dosage was tapered or discontinued as per individual protocol, but not before day 28. Other GVHD prophylaxis regimens included CYA/methylprednisolone, CYA/methotrexate and CYA/ methylprednisolone/methotrexate. GVHD was graded according to the Glucksberg and Seattle consensus criteria.16 The ‘rule of 9’ or burn chart was used to estimate the extent of skin rash. If a clinical diagnosis of acute GVHD (aGVHD) or chronic GVHD (cGVHD) was made, histological confirmation was obtained whenever possible.

Supportive care All patients were hospitalized in protective isolation, defined as single hospital rooms with high-efficiency particulate air filtration system and reverse isolation requiring strict hand degerming and use of masks by staff for unrelated HSCT patients. Beginning in January 2001, liposomal amphotericin B (3 mg/kg/day i.v. over 2 h) was administered to all patients who received allogeneic SCT starting on day 0 until day 100, as we previously described.17 Fluconazole was used as anti-fungal prophylaxis for the remaining patients. Additional anti-infective prophylaxis against Pneumocystis carinii pneumonia, HSV and CMV, if indicated, was administered, as we have previously described.18 All patients received hematopoietic growth factors starting with sargramostim (GM-CSF 250 mg/m2/day i.v. daily) on day 0 within 3 h of stem cell infusion and continued until WBC count was 0.3 109 cells/L for two consecutive days. Patients were then


531

switched to filgrastim (G-CSF 10 mcg/kg/day i.v. daily) and tapered by 50% when the ANC reached 42.5 109 cells/L for two consecutive days, as we previously described.19

Cardiovascular surveillance Two-dimensional echocardiograms (Hewlett Packard Philips Sonos 5500 or 7500, Philips Medical Systems, Andover, MA, USA) were performed uniformly at baseline before each SCT and when patients had symptoms and/or signs of cardiac or pericardial disease post-SCT. Signs and symptoms prompting examination with echocardiogram included, but were not limited to, changes in basal heart rate, pulse pressure, blood pressure, shortness of breath, oxygen desaturation and/or chest pain. A single member (CK) of the Pediatric Cardiology Department reviewed all echocardiograms. The cardiologist reviewing the echocardiograms, although aware of the clinical status of the patients, was blinded to all other associated complications. PEF was defined as a recognizable accumulation of fluid within the pericardial space, resulting in a clear space between the visceral and parietal pericardium. The assessment of PEF was at best semi-quantitative, but was most often qualitative. Patients were enrolled in the study on meeting the eligibility criteria of shortening fraction of X25–28% by echocardiogram, or ejection fraction of X45–48% by radionuclide angiogram or echocardiogram and no evidence of PEF. Statistical analysis Continuous variables were presented as mean±s.d., and categorical variables were presented as percentages. Kaplan–Meier curves were constructed for the time to appearance of PEF and time to death. The univariate and multivariate Cox proportional hazards models were used to examine the effect of each covariate on time to effusion and time to death. aGVHD and cGVHD were treated as timedependent covariates. Only subjects who had an allogeneic SCT were considered whenever aGVHD or cGVHD was involved (n ¼ 102). In time to death analysis, effusion status was also treated as a time-dependent covariate. The covariates analyzed for risk to develop PEF and time to death included older age (continuous variable), gender, ethnicity, CMV status, ablative conditioning, allogeneic transplant, aGVHD or cGVHD, number of transplants, malignant disease, high-risk, unrelated cord blood donor source and degree of matching (6/6 or 10/10 vs o6/6 or o10/10). In addition, presence of PEF was included in the analysis for time to death. Those covariates with a P-value of o0.2 were included in the final multivariate Cox proportional hazards model. All statistical analyses were performed using SAS Version 9.1 Software (SAS Institute, Cary, NC, USA). The significance level was set to be 0.05.

Results Patient demographics Between 1 January 2000 and 22 September 2005, we identified 23 of the 156 patients who developed PEF after SCT. Patient demographic characteristics are summarized

Table 2

Patient demographic characteristics

Mean age in years (range)

8.15±6.25

Sex Male Female

88 (56.4%) 68 (43.6%)

Ethnicity White Black Hispanic Other

55 26 44 31

(35.3%) (16.7%) (28.2%) (19.8%)

Diagnosis (N ¼ 156) Malignant Non-malignant

116 (74.4%) 40 (25.6%)

Prior transplant history (N ¼ 156) 1 41

122 (78.2%) 34 (21.8%)

Source of stem cells Allo-SCT Auto-SCT

102 (65.4%) 54 (34.6%)

Match 6/6 or 10/10 p6/6 or 10/10 Autologous

46 (29.4%) 56 (35.9%) 54 (34.7%)

Donor source BM/PBSC Cord blood

100 (64.1%) 56 (35.9%)

Conditioning type RI MA None

44 (28.4%) 111 (71.6%) 1 (0.6%)

CMV status Autologous Negative Positive Allogeneic Recipient/donor Recipient/donor Recipient/donor Recipient/donor Recipient/donor Recipient/donor Recipient/donor Unknown

29 (18.6%) 22 (14.1%)

(negative/negative) (negative/positive) (positive/negative) (positive/positive) (negative/unknown) (positive/unknown) (positive/NA)

23 4 31 37 1 1 3 6

(14.7%) (2.6%) (19.9%) (23.7%) (0.6%) (0.6%) (1.9%) (3.8%)

Abbreviations: MA ¼ myeloablative; RI ¼ reduced intensity.

in Table 2. The mean follow-up time was 978.6±838.3 days. The mean age was 8.15±6.25 years, male to female ratio was 56 vs 44%. In all, 74% had malignant disease; 35% received autologous stem cell grafts and among the other 65% who received allogeneic grafts, 36% were X1 Ag mismatched.

Engraftment Among 156 subjects, 138 subjects (88.46%) achieved neutrophil engraftment at a mean±s.d. of 17.33±11.42 days after transplant, and 105 subjects (67.31%) achieved plt engraftment at a mean±s.d. of 32.25±38.56 days after transplant. Bone Marrow Transplantation


532

Acute and chronic GVHD The probability of developing grade II–IV aGVHD was 29.6% (95% CI 21.6–36.7%). In assessing cGVHD, there were 31 patients (19.9%) who were not evaluable because they died before day 100. The probability of developing cGVHD was 13.4% (95% CI 6.8–19.6%). In a univariate analysis of time to development of PEF, aGVHD and cGVHD did not have an increased risk, with a HR of 1.213 (P ¼ 0.7159) and HR of 0 (P ¼ 0.9933), respectively.

Probability of developing effusion

Probability of developing PEF and risk factors associated with it PEF was found in 18/102 (17.6%) patients receiving allogenic SCT and in 5/54 (9.3%) patients who received autologous SCT (P ¼ NS). The overall probability of developing a symptomatic PEF after SCT in pediatric recipients was 16.9% with 95% CI 10.3–23.0% (Figure 1). The mean time after SCT to detection of effusion for the 23 PEF subjects was 2.75±3.54 months. The mean follow-up time after detection of PEF was 11.55±18.04 months. In a univariate analysis, the risk to developing a PEF was significant (Po0.05) according to the following risk factors: increased age (HR 1.129, 95% CI 1.06–1.21), allogeneic high risk (HR 3.43, 95% CI 1.51–7.80) and donor source (cord blood vs peripheral blood or BM) (HR 3.84, 95% CI 1.66–8.90) (Table 3a). In a multivariate analysis of the risk of developing PEF (including factors with a Po0.2 in the univariate analysis), increased age, high-risk status and ablative conditioning type were all significant risk factors contributing to development of PEF, with a HR of 1.133, 95% CI 1.05–1.22 (Po0.0007); HR of 3.155, 95% CI 1.32–7.55 (Po0.01); and HR of 3.949, 95% CI 1.07–14.5 (Po0.01), respectively (Table 3b). TBI as part of the conditioning regimen was an independent risk factor for development of PEF. In the patients who received TBI as part of their conditioning regimen, the rate of PEF was 39.38%. In the group of patients who did not receive TBI, the rate of developing PEF was 15.59% (P ¼ 0.027). The HR of TBI in the univariate Cox proportional hazards model was significant,

1.0 0.9 0.8 0.7

with a value of 3.43, 95% CI 1.38–8.52 (P ¼ 0.008), but the HR was not significant in multivariate analysis after adjusting for age, race/ethnicity, risk, donor source and match, HR ¼ 1.47, 95% CI 0.53–4.06 (P ¼ 0.46). Of the 23 patients with PEF, 19 died (82.6%); however, none died as a direct result of pericardial tamponade and/ or surgical therapeutic intervention. Six patients who developed PEF after SCT required mechanical intervention for decompression of the effusion. Three of those patients underwent surgical intervention with a pericardial window, while three patients underwent pericardiocentesis. Five of the patients who required pericardiocentesis post-SCT had microbiological studies sent and all were negative for bacterial, viral and fungal cultures. Given the very small number of patients who had PEF sent for analysis, we are unable to draw any conclusions or detect any trends in the WBC count or protein of the fluid. All patients who required surgical intervention for the PEF eventually died; Table 3a

Univariate analysis of risk to develop pericardial effusion Hazard ratio

95% confidence interval

P-value

1.129

1.057–1.205

0.0003

Gender Female Male

1.0 (reference) 1.752

0.721–4.260

0.21


1.0 (reference) 1.739 2.391 1.924

0.467–6.477 0.801–7.137 0.557–6.650

0.41 0.12 0.30

CMV status Negative/negative Negative/positive


0.392–2.149

0.84

Transplant type Autologous Allogeneic


0.878–6.398

0.09

GVHD Acute Chronic

1.213 0.000

0.429–3.430 0.000–infinity

0.72 0.99

No. of transplants

1.098

0.433–2.786

0.84

Malignancy

1.048

0.413–2.658

0.92

Risk Standard High


1.511–7.802

0.003

Conditioning type RI MA


0.817–9.247

0.10

Donor source BM/PBSC UCB


1.655–8.903

0.0017

0.369–4.410 1.267–10.060

0.70 0.016

Age

0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

5

10

15

20

25

30

Months after SCT Figure 1

Probability of developing a PEF. Probability of developing symptomatic PEF following SCT in pediatric recipients as determined by the Kaplan–Meier method.

Bone Marrow Transplantation

Match Autologous Allogeneic 6/6 or 10/10 o6/6 or o10/10

1.0 (reference) 1.276 3.570



533 Table 3b Multivariate analysis of development of pericardial effusion (with covariates Po0.2 in univariate analysis) Hazard ratio

95% hazard ratio confidence limits

P-value

1.133

1.054–1.217

0.0007

significant risk factors for time to death, with Po0.03 and Po0.02, respectively.

Discussion Age Ethnicity White Black Hispanic Other

1.0 (reference) 1.039 2.253 1.269

0.250–4.318 0.705–7.196 0.314–5.127

0.96 0.17 0.74

Risk Standard High


1.318–7.553

0.01



1.073–14.533

0.04



0.556–16.097

0.20

0.446–7.463 0.242–14.726

0.40 0.54


1.0 (reference) 1.824 1.888


however, none of them died as a direct result of tamponade. In the entire group of patients with PEF, deaths were attributed to infection/sepsis (n ¼ 5), GVHD (n ¼ 3), persistent disease (n ¼ 8) and multi-organ failure (n ¼ 3). The characteristics of the patients with PEF are summarized in Table 4. Six of the patients with PEF who died had autopsies performed. The results are summarized in Table 5.

Survival Risk factors associated with survival. The mean follow-up of all subjects after SCT was 978.60±838.29 days. The probability of 3-year OS was 52.44% (95% CI 45.1–61.0%) (Figure 2). There was a significant difference in OS between patients with and without development of PEF as shown in Figure 2. PEF, as a time-dependent variable, had the highest HR of 3.345, 95% CI 1.76–6.35 (P ¼ 0.0002). In the univariate analysis of risk of death (Table 6a), there were many factors that were significant (Po0.2). Presence of PEF was the most significant risk factor for time to death with a P-value of o0.0001. Consistent with known risk factors for transplant, increased age, CMV status, transplant type (allogeneic), presence of cGVHD, high-risk status, donor source (umbilical cord blood) and match (o6/6 or o10/10) were all significant risk factors for time to death. In a multivariate analysis of time to death, PEF, CMV status and high-risk status all remained as independent risk factors. In a multivariate analysis of risk of death (Table 6b), PEF was the most significant risk factor (HR: 3.345, 95% CI 1.76–6.35) with a P-value of 0.0002. Positive CMV status and poor risk were also

As previously mentioned above, PEF has become a more recognized complication after SCT. In this study, we evaluated multiple factors that may contribute to the development of and risk of mortality from PEF post-SCT. We have conducted the largest study to date evaluating the role of PEF in pediatric SCT recipients. The probability of developing PEF was 16.9% overall, 17.6% in our patients receiving allogeneic SCT and 9.3% in those receiving autologous SCT. This was much higher than the 4.4% of patients previously described by Rhodes et al.5 Rhodes et al.,5 however, included only those patients with clinically significant PEF post-SCT.5 This previous study differs from ours in that we included all patients with PEF detected on echocardiography irrespective of whether or not they were clinically relevant. As so many SCT patients with PEF may be asymptomatic and additionally may not be routinely screened with echocardiograms, we theorize that this may be an under-diagnosed or under-recognized complication of SCT. If detected, however, PEF may be an important indicator of increased risk of morbidity and mortality post-SCT. Given the myriad of complications after SCT, it is difficult to tease out the risk factors associated with the development of PEF. As mentioned earlier, other studies have implicated allogeneic donor source, GVHD, infection and/or relapsed malignant disease as risk factors for developing PEF post-SCT.4,5,7,11–13 We evaluated these and other risk factors using univariate and multivariate analysis. Risk factors typically associated with a worse prognosis were also associated with a risk for developing PEF. We found that increased age, high-risk status and ablative conditioning type were all significant risk factors contributing to the development of effusion. In addition, TBI-based conditioning was an independent risk factor for development of PEF. Increased age has previously been shown to be a risk factor for SCT recipients,20–22 and it was a statistically significant risk factor for development of PEF in this study. We evaluated age as a continuous variable with increased risk as age increased. There are no large studies, however, evaluating the development of PEF in adult recipients. Our data suggest that there is a need to monitor and report PEF in adult and pediatric SCT recipients. We defined high-risk patients to be those having (a) active leukemia in relapse, (b) active leukemia following induction failure, (c) leukemia in third CR or beyond, (d) progressive solid tumor and/or (e) second allogeneic SCT. In our study, poor risk status was a significant risk factor for development of and risk of death from PEF after SCT. This was similar to the data taken from other studies that showed high-risk factors to increase a patient’s risk of relapse and death.23 Ablative conditioning was noted in our data to be an important risk factor for the development of PEF in SCT recipients. Our group recently summarized the experience Bone Marrow Transplantation


10.52

8.7

20.27

18.66

a

5a

6

7

20.38

7.4

16.41

5.56

12.21

7.29

17.9

15.02

13.03

14.59

21.67

17.61

10.93

9.59

10

11

12

13

14

15

16

17

18a

19

20a

21

22

23

AML

SAA

AML

ALL

SAA

ALL

NHL

NHL

NBL

SAA

NBL

SAA

SAA

HD

HLH

ALL

HD

HD

ALL

NHL

ALL

AML

ALL

Diagnosis

High

Average

Average

High

Average

High

High

High

High

Average

Average

Average

Average

High

Average

High

High

High

Average

Average

Average

High

Average

Risk status

5/6 UCB

6/6 sib

6/6 maternal BM

4/6 UCB

4/6 UCB

4/6 UCB

Auto

Auto

4/6 UCB

6/6 BM

Auto

4/6 UCB

6/6 sib

Auto

6/6 UCB

5/6 UCB

Auto

4/6 UCB

4/6 UCB

4/6 UCB

4/6 UCB

4/6 UCB

5/6 UCB

Type of transplant and HLA match

39

47

153

28

38

88

98

49

96

219

518

63

140

67

24

39

14

29

16

26

21

5

97

Time to PEF (days)

36

40

39

38

39

34

36

29

38

37

36

44

39

36

42

38

29

35

41

33

38

37

40

Average SF (%)

92

117

110

87

—

106

67

76

—

60

—

87

94

94

143

77

94

122

—

79

80

86

104

Average HR beat/min

Pericardiocentesis

Pericardial window

None

None

Pericardiocentesis

Pericardiocentesis pre-transplant None

None

None

None

None

Pericardial window

Pericardial window

None

None

None

None

None

None

None

None

None

Pericardiocentesis

Surgical intervention

CMV D-day+336 Sepsis D-day+18 Adenovirus D-day+28 Alive, secondary MDS day+2339 GVHD D-day+41 Persistent disease D-day+79 Aspergillus D-day+737 Persistent disease D-day+212 GVHD D-day+227 Persistent disease D-day+351 Multi-organ failure D-day+229 Multi-organ failure D-day+84 Alive day+1401 Alive day+1378 Persistent disease D-day+172 Alive day+1107 Persistent disease D-day+167 Persistent disease D-day+453 Fungal infection D-day+40 Liver GVHD D-day+94 Persistent disease D-day+305 Multiorgan failure D-day+159 Persistent disease D-day+74

Outcome/cause of death

Abbreviations: auto ¼ autologous; D ¼ death; HD ¼ Hodgkin’s disease; HLH ¼ hemophagocytic lymphohistiocytosis; HR ¼ heart rate; NBL ¼ neuroblastoma; PEF ¼ pericardial effusion; SAA ¼ severe aplastic anemia; SF ¼ shortening fraction; sib ¼ sibling. a TBI included in conditioning regimen.

0.74

9

8

4

11.17

18.33

3a

a

14.97

2

2.32

a

Age at transplant

Characteristics of patients with pericardial effusion

1a

Patient

Table 4


534


535 Autopsy findings of patients with pericardial effusion

Table 5 Patienta

Gross cardiovascular findings

Microscopic cardiovascular findings

Cultures

Other

1

Cardiomegaly, interstitial fibrosis, edema, mild hemosiderosis, pericardial effusion Serous effusion, septal myocyte hyperplasia Pericardium with multiple focal fibrinous exudates. Moderate ventricular dilatation

Mild diffuse interstitial fibrosis and edema

Negative

Evidence of prolonged pulmonary disease. Cardiac disease possibly consistent with toxicity or volume overload. No definitive cause of death identified

Septal myocyte hyperplasia Focal myocytolysis, interstitial fibrosis, focal myocyte hypertrophy

Negative. Weak positive Evidence of ARDS as cause of death stain for adenovirus Right lung positive for Death attributed to cardiac arrest secondary to cytomegalovirus pulmonary hemorrhage due to CMV pneumonitis. Evidence of hemophagocytic syndrome

3 5

CMV and adenovirus immunostains positive

11 (limited to lung only) 12

Pericardial effusion. Mild chronic pericarditis and fibrosis

19

Thickened pericardium, thrombus adherent to the aortic wall causing 490% occlusion

Marked fibrosis of the pericardium with a moderate lymphoplasmocytic infiltrate Fibrinous exudates and fungal hyphae on epicardial and pericardial surfaces. Aortic thrombus with fungal hyphae

22 (lung only)

Positive adenovirus culture from bladder tissue

Acute alveolar injury with acute and chronic hemorrhage and focal interstitial fibrosis likely secondary to adenoviral and cytomegalovirus infections Massive hematoperitoneum without noted site of bleeding, possibly due to severe ischemic damage of liver. Hemorrhagic cystitis

Morphological evidence of aspergillus although cultures negative

Pleural and pericardial effusions determined to be result of invasive fungal infection including involvement of aorta

Cultures negative

Diffuse alveolar damage, diffuse alveolar hemorrhage, small-vessel thrombi consistent with intravascular coagulopathy, fibrinous pleuritis

Probability of overall survival

Abbreviation: ARDS ¼ acute respiratory distress syndrome. a Patient number correlates with Table 4.

1.0 0.8 0.6 0.4 0.2 0.0 0

10

20 Months after SCT

30

40

Figure 2 Probability of OS. Probability of OS as determined by the Kaplan–Meier method in all patients (n ¼ 154).

of reduced intensity allogeneic SCT in pediatric recipients.24 We hypothesized that the risk of both aGVHD and cGVHD may be lower in recipients of reduced-intensity conditioning transplants and that the risk of bacterial infections may also be lower in the peri-transplant period after reduced-intensity conditioning. There may be an increased risk, however, of viral and fungal infections because of the increased need for immunosuppression in these patients. Patients in our study who received ablative conditioning were at higher risk for development of PEF; this may be related to the increased risk of GVHD in these

patients. Given that patients who receive ablative conditioning regimens are more likely to receive TBI, and TBI was an independent risk factor for development of PEF, TBI may be the inciting factor predisposing these patients to PEF. In our study, PEF had the most significant impact on risk of death post-SCT. CMV status and poor risk status, however, were also significant risk factors for mortality. Although CMV status increased the risk of death, there was no increased risk of development of PEF. As testing for CMV changed over the course of our study, we chose to evaluate patients based on their CMV status measured with titers rather than levels of infection/disease with PCR methods. One patient with a PEF, however, had clear evidence of active CMV disease. CMV reactivation has previously been shown to be a known risk factor for mortality in allogeneic SCT recipients.25 As previously described, our group used a prophylactic regimen of alternative day ganciclovir and foscarnet to prevent CMV activation in allogeneic SCT recipients.18 It may be that there is a common interplay of host factors and donor factors contributing to mortality in patients with PEF in SCT recipients. Given the relatively early incidence of PEF, it may be a manifestation of aGVHD in patients receiving allogeneic transplantation. GVHD is often associated with a cytokine surge that could be the inciting factor for the development of PEF.26 For example, Carlson et al.27 described the association between differentiated Th17 cells and the Bone Marrow Transplantation


536 Table 6a

Univariate analysis of time to death Hazard ratio 95% confidence interval P-value

Age Gender Female Male

1.039

1.002–1.077

Table 6b Multivariate analysis of time to death (with covariates Po0.2 in univariate analysis with all subjects, n ¼ 148) 95% hazard ratio confidence limits

1.008

0.966–1.051

0.72

Pericardial effusion No PEF Pericardial effusion


1.762–6.350

0.0002



1.066–3.221

0.03

Risk Standard High


1.127–3.420

0.02



0.473–2.219

0.95

Number of transplants

0.820

0.417–1.611

0.56

Chronic GVHD

1.240

0.515–2.990

0.63

0.686–3.198 0.901–6.535

0.32 0.08

Age 1.0 (reference) 0.904


1.0 (reference) 0.895 1.264 1.413

Pericardial effusion No PEF PEF




0.572–1.428

0.428–1.873 0.713–2.239 0.759–2.630

2.814–8.322

1.009–2.800

0.66

0.77 0.42 0.28

o0.0001

0.046



Transplant type Autologous Allogeneic


1.454–4.314

0.0009

GVHD Acute Chronic


1.272 2.139

0.729–2.222 0.880–5.197

0.40 0.09

Abbreviation: PEF ¼ pericardial effusion.

No. of transplants

0.664

0.371–1.188

0.17

Malignancy

1.320

0.759–2.297

0.33

Risk Standard High


1.287–3.327

0.003



1.556–3.890

0.0001

Match Autologous 1.0 (reference) Allogeneic 6/6 or 10/10 1.764 o6/6 or o10/10 3.335

0.930–3.347 1.868–5.956

0.08 o0.0001

0.663–1.883

0.68

Abbreviations: MA ¼ myeloablative; PEF ¼ pericardial effusion; RI ¼ reduced intensity.

development of a lethal form of GVHD If development of PEF is indeed related to aGVHD, however, it is more difficult to understand why patients receiving autologous SCT experience PEF. It is impressive to note that in Rhodes’ review, all patients with PEF had aGVHD or cGVHD.5 The patients with clinically insignificant PEF may also be experiencing a subacute or milder form of GVHD. Alternatively, PEF may be a harbinger of development of GVHD. We hypothesize that PEF postSCT may be related to a cytokine surge, regardless of the presence of aGVHD. This needs to be evaluated prospectively. In future prospective studies, it would be helpful to measure either the amount of serum cytokines or, in patients requiring intervention, the amount of cytokines in the pericardial fluid. Bone Marrow Transplantation

P-value

Hazard ratio

0.04

1.0 (reference) 1.481 2.426

Another possible explanation for the development of PEF post-SCT may be the large volume of i.v. fluids many of these patients receive. It is not feasible, however, to delineate the exact amount of fluids any patient received in the peri-transplant period. In addition, we did not note an increased amount of peripheral edema that would be related to fluid overload syndromes. Given the limitations of a retrospective study, it is not possible to quantify the exact amount of fluid intake each patient had 48–72 h before developing a PEF. Other authors have postulated infectious etiologies for development of PEF.28,29 However, we did not note an increased risk of CMV in patients developing PEF. Not all patients were routinely screened for other viral infections such as adenovirus, so it was not possible to include this as a risk factor. However, in evaluating the patients with clinically significant PEF requiring intervention, we did not note a trend towards increased serious bacterial or viral infections. There are some notable limitations with this study. Given the nature of pediatric SCT programs, this is a very heterogeneous population; for example, there are patients with both malignant and non-malignant diseases. In order to evaluate a more uniform population with significant numbers, one would need to perform a multi-institutional study. In addition, as this was a retrospective study, there were no standardized time points for performing routine echocardiograms. Given this limitation, we may have underestimated the number of patients with asymptomatic PEF as well as the timing of the onset of the PEF. As this was a retrospective study, there were some factors that we were unable to evaluate, such as total anthracyline dose;


537

however, cumulative exposure to anthracyclines was limited to o350 mg/m2. Another limitation of this study is that there was no standard timing of post-HSCT echocardiograms. In future studies, we would strongly consider performing echocardiograms at day þ 30, þ 60 and þ 100 after HSCT. While the etiology of PEF and its impact on mortality post-SCT is unclear, we are convinced that it is imperative to initiate closer screening for the development of PEF. We propose that patients with PEF may need closer monitoring for the development of GVHD or other complications. As most of these patients have clinically insignificant PEF, it is typically not necessary to surgically intervene. It could be important, though, to begin prospective studies to evaluate more closely the etiology of PEF and possible interventions to decrease mortality, such as more frequent screening with echocardiogram.

9

10

11

12

13

Conflict of interest The authors declare no conflict of interest.

14

Acknowledgements This work was supported in part by grants from the Pediatric Cancer Research Foundation, Brittany Baron Fund, Sonia Scaramella Fund, Bevanmar Foundation, Paul Luisi Foundation, Dreaming for Discovery and Cure Fund, and the Marisa Fund.

15

16

References 1 Cairo M, Heslop H. Pediatric blood and marrow transplantation: state of the science. Bone Marrow Transplant 2008; 41: 97. 2 Murdych T, Weisdorf DJ. Serious cardiac complications during bone marrow transplantation at the University of Minnesota, 1977–1997. Bone Marrow Transplant 2001; 28: 283–287. 3 Einsele H, Ehninger G, Vallbracht A, Kompf J, Schmidt H, Kandolf R et al. Isolated pericardial relapse following allogeneic bone marrow transplantation for acute myelogenous leukemia. Bone Marrow Transplant 1989; 4: 323–325. 4 Krishnan GS, Chaudhary V, Al-Janadi A, Ramanarayanan J, D’Silva KJ. BCNU toxicity presenting with a large pericardial and pleural effusion. Ann Transplant 2008; 13: 44–47. 5 Rhodes M, Lautz T, Kavanaugh-Mchugh A, Manes B, Calder C, Koyama T et al. Pericardial effusion and cardiac tamponade in pediatric stem cell transplant recipients. Bone Marrow Transplant 2005; 36: 139–144. 6 Saito Y, Matsushima T, Doki N, Tsumita Y, Takizawa M, Yokohama A et al. Pericardial graft vs. host disease in a patient with myelodysplastic syndrome following peripheral blood stem cell transplantation. Eur J Haematol 2005; 75: 65–67. 7 Seber A, Khan SP, Kersey JH. Unexplained effusions: association with allogeneic bone marrow transplantation and acute or chronic graft-versus-host disease. Bone Marrow Transplant 1996; 17: 207–211. 8 Silberstein L, Davies A, Kelsey S, Foran J, Murrell C, D’Cruz D et al. Myositis, polyserositis with a large pericardial effusion and constrictive pericarditis as manifestations of chronic

17

18

19

20

21

22

graft-versus-host disease after non-myeloablative peripheral stem cell transplantation and subsequent donor lymphocyte infusion. Bone Marrow Transplant 2001; 27: 231–233. Toren A, Nagler A. Massive pericardial effusion complicating the course of chronic graft-versus-host disease (cGVHD) in a child with acute lymphoblastic leukemia following allogeneic bone marrow transplantation. Bone Marrow Transplant 1997; 20: 805–807. Veys PA, McAvinchey R, Rothman MT, Mair GH, Newland AC. Pericardial effusion following conditioning for bone marrow transplantation in acute leukaemia. Bone Marrow Transplant 1987; 2: 213–216. Angelucci E, Mariotti E, Lucarelli G, Baronciani D, Cesaroni P, Durazzi SM et al. Sudden cardiac tamponade after chemotherapy for marrow transplantation in thalassaemia. Lancet 1992; 339: 287–289. Cereda M, Trocino G, Pogliani EM, Schiavina R. A case of cardiac localization of graft-versus-host disease after allogenic bone marrow transplantation. Ital Heart J 2003; 4: 60–63. Kondo M, Kojima S, Horibe K, Kato K, Matsuyama T. Hemolytic uremic syndrome after allogeneic or autologous hematopoietic stem cell transplantation for childhood malignancies. Bone Marrow Transplant 1998; 21: 281–286. Osunkwo I, Bessmertny O, Harrison L, Cheung YK, Van de Ven C, del Toro G et al. A pilot study of tacrolimus and mycophenolate mofetil graft-versus-host disease prophylaxis in childhood and adolescent allogeneic stem cell transplant recipients. Biol Blood Marrow Transplant 2004; 10: 246–258. Yamashiro DJ, Lee A, Bhatia M, Glade-Bender J, Qualter E, Militano O et al. Feasibility of autologous stem cell transplant followed by reduced intensity allogeneic stem cell transplantation for high risk neuroblastoma: a single institution pilot study. Biol Blood Marrow Transplant 2007; 13: 68–69 (abstract). Glucksberg H, Storb R, Fefer A, Buckner CD, Neiman PE, Clift RA et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HL-A-matched sibling donors. Transplantation 1974; 18: 295–304. Roman E, Osunkwo I, Militano O, Cooney E, van de Ven C, Cairo MS. Liposomal amphotericin B prophylaxis of invasive mold infections in children post allogeneic stem cell transplantation. Pediatr Blood Cancer 2008; 50: 325–330. Shereck EB, Cooney E, van de Ven C, Della-Lotta P, Cairo MS. A pilot phase II study of alternate day ganciclovir and foscarnet in preventing cytomegalovirus (CMV) infections in at-risk pediatric and adolescent allogeneic stem cell transplant recipients. Pediatr Blood Cancer 2007; 49: 306–312. Waxman IM, Militano O, Baldinger L, Roman E, Qualter E, Morris E et al. Sequential administration of sargramostim and filgrastim in pediatric allogeneic stem cell transplantation recipients undergoing myeloablative conditioning. Pediatr Transplant 2009; 13: 464–474. Anasetti C. What are the most important donor and recipient factors affecting the outcome of related and unrelated allogeneic transplantation? Best Pract Res Clin Haematol 2008; 21: 691–697. Cornelissen JJ, van Putten WL, Verdonck LF, Theobald M, Jacky E, Daenen SM et al. Results of a HOVON/SAKK donor versus no-donor analysis of myeloablative HLA-identical sibling stem cell transplantation in first remission acute myeloid leukemia in young and middle-aged adults: benefits for whom? Blood 2007; 109: 3658–3666. Goldstone AH, Richards SM, Lazarus HM, Tallman MS, Buck G, Fielding AK et al. In adults with standard-risk acute lymphoblastic leukemia, the greatest benefit is achieved from a matched sibling allogeneic transplantation in first complete remission, and an autologous transplantation is less effective Bone Marrow Transplantation


538 than conventional consolidation/maintenance chemotherapy in all patients: final results of the International ALL Trial (MRC UKALL XII/ECOG E2993). Blood 2008; 111: 1827–1833. 23 Doney K, Hagglund H, Leisenring W, Chauncey T, Appelbaum FR, Storb R. Predictive factors for outcome of allogeneic hematopoietic cell transplantation for adult acute lymphoblastic leukemia. Biol Blood Marrow Transplant 2003; 9: 472–481. 24 Satwani P, Morris E, Bradley MB, Bhatia M, van de Ven C, Cairo MS. Reduced intensity and non-myeloablative allogeneic stem cell transplantation in children and adolescents with malignant and non-malignant diseases. Pediatr Blood Cancer 2008; 50: 1–8. 25 Boeckh M, Nichols WG, Papanicolaou G, Rubin R, Wingard JR, Zaia J. Cytomegalovirus in hematopoietic stem cell transplant recipients: current status, known challenges, and future strategies. Biol Blood Marrow Transplant 2003; 9: 543–558.


26 Dean RM, Fry T, Mackall C, Steinberg SM, Hakim F, Fowler D et al. Association of serum interleukin-7 levels with the development of acute graft-versus-host disease. J Clin Oncol 2008; 26: 5735–5741. 27 Carlson MJ, West ML, Coghill JM, Panoskaltsis-Mortari A, Blazar BR, Serody JS. In vitro-differentiated TH17 cells mediate lethal acute graft-versus-host disease with severe cutaneous and pulmonary pathologic manifestations. Blood 2009; 113: 1365–1374. 28 Chung HT, Hsieh TC, Yu MC, Chang YS, Lo WC, Jaing TH. Staphylococcus aureus pericardial abscess in a child with betathalassemia major following double-unit unrelated cord blood transplantation. Pediatr Hematol Oncol 2007; 24: 275–279. 29 Schaumann R, Ponisch W, Helbig JH, Hegenbart U, Ackermann G, Hofmann J et al. Pericarditis after allogeneic peripheral blood stem cell transplantation caused by Legionella pneumophila (non-serogroup 1). Infection 2001; 29: 51–53.