Immune Abnormalities in Fontan Protein-Losing Enteropathy: A Case-Control Study H. Sonali Magdo, DO1, Terri L. Stillwell, MD2, Matthew J. Greenhawt, MD3, Kathleen A. Stringer, PharmD4, Sunkyung Yu, MS5, Carlen G. Fifer, MD1, Mark W. Russell, MD1, and Kurt R. Schumacher, MD1 Objective To comprehensively characterize the immunologic characteristics of patients with protein-losing enteropathy (PLE) post-Fontan and compare them with patients without PLE post-Fontan.
Study design Patients with PLE post-Fontan and age-matched controls post-Fontan were prospectively studied with laboratory markers of immune function. Infectious history was obtained by interview and chart review. The groups’ demographics, cardiac history, immune characteristics, and infection history were compared using appropriate 2-group statistics. Results A total of 16 patients enrolled (8 patients with PLE and 8 controls). All patients with PLE had lymphopenia compared with 25% of controls (P = .01). All patients with PLE had markedly depressed CD4 T cell counts (median 58 cells/mL) compared with controls (median 450 cells/mL, P = .0002); CD4% was also low in the PLE group (12.3%) and normal in control (36.9%, P = .004). Both groups had mildly depressed CD8 T cells and normal to slightly elevated natural killer and B-cell subsets. A majority of patients with PLE (62.5%) had negative titers to measles, mumps, and rubella vaccination, compared with no control Fontan with a negative titer (P = .03). Despite profoundly low CD4 counts, the frequency of infection was not different between groups with no reported opportunistic infections. Conclusions Patients with Fontan-associated PLE have extensive quantitative immune abnormalities, particularly CD4 deficiency. These immune abnormalities are similar to those found in non-Fontan patients with PLE caused by intestinal lymphangiectasia. (J Pediatr 2015;167:331-7).
P
rotein-losing enteropathy (PLE) is a complication of the Fontan palliation for single-ventricle congenital heart disease. Although PLE occurs in only about 5% of patients post-Fontan, it confers marked morbidity and high mortality within 5 years of diagnosis.1-3 Fontan-related PLE is episodic with “flares” in symptom severity characterized by severe enteral protein loss leading to peripheral edema, effusions, and ascites.4,5 The pathogenesis of Fontan-related PLE is not completely understood, and it is unclear why it develops in some patients post-Fontan and not others.1,4,5 PLE has been described in many different diseases. PLE that results from some diseases, such as primary intestinal lymphangiectasia (IL) and constrictive pericarditis, has been associated with significant immune abnormalities, such as lymphopenia, with a selective loss of CD4 cells and severe hypogammaglobulinemia.6-9 However, patients with PLE associated with inflammatory diseases such as ulcerative colitis, celiac disease, and allergic PLE do not have these immune derangements.10-12 Immune abnormalities have been reported in Fontan-related PLE. In small case series or case reports, patients with PLE postFontan have reported reduced immunoglobulin levels,13,14 lymphopenia,11 and selective CD4 lymphocyte deficiency.14-16 However, these studies were limited in number of patients, the extent of immune characterization, and/or lack of a control group. As some lymphocyte abnormalities have been found in patients who have undergone neonatal thymectomy, and to control for the possibility that the Fontan circulation itself may influence the immune system, we sought to comprehensively characterize and compare the immune system in individuals receiving Fontan with and without PLE. We hypothesized that patients with PLE post-Fontan have distinct quantitative immune abnormalities compared with those without PLE and may share similarities with PLE caused by other diseases.
Methods This is a prospective, case-control study of patients with and without PLE post-Fontan. The study was approved by the University of Michigan Institutional
GI IL IVIG MMR NK PLE
Gastrointestinal Intestinal lymphangiectasia Intravenous immunoglobulin Measles, mumps, and rubella Natural killer Protein-losing enteropathy
From the 1University of Michigan Congenital Heart Center, C.S. Mott Children’s Hospital; 2University of Michigan Pediatric Infectious Disease; 3University of Michigan Allergy and Immunology; 4University of Michigan College of Pharmacy; and 5University of Michigan, Michigan Congenital Heart Outcomes Research and Discovery Ann Arbor, MI H.M. received a grant from University of Michigan Congenital Heart Center (G012107). The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2015.04.061
331
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Review Board. Informed consent was obtained from each patient if 18 years old or older, or from the subjects’ parent/legal guardian if less than 18 years old. Eligible subjects were identified from institutional records and approached for participation in the hospital or clinic setting. All patients enrolled were required to have single ventricle congenital heart disease and have undergone a total cavopulmonary anastomosis. A patient with Fontanassociated PLE was defined as having a clinical diagnosis of PLE made by a treating cardiologist, a history of hypoalbuminemia, and resultant clinical symptoms requiring previous hospitalization for PLE-specific treatment. An uncomplicated control was defined as any patient status post-Fontan palliation without significant comorbid disease including Fontan hepatopathy, plastic bronchitis, known hematologic or immunologic disorders, any other non-cardiac major organ dysfunction, or a previous diagnosis of PLE. Controls were matched to patients with PLE post-Fontan based on age (12 months). Exclusion criteria included concurrent febrile illness, antibiotic use, or systemic steroid use within 3 weeks of blood collection. Patients post-Fontan taking oral or inhaled budesonide therapy were not excluded. From a combination of chart review and patient interview, the following clinical information was gathered for each subject: underlying congenital heart disease, age at stage II palliation, date of Fontan, type of Fontan, length of hospital stay after Fontan, readmission within the first month after Fontan, history of neonatal thymectomy or thymectomy after 1 month of age, total or subtotal thymectomy, infection history, and immunization history. For subjects with PLE, information was obtained regarding the clinical status of the disease (whether symptoms were currently active) and whether albumin or intravenous immunoglobulin (IVIG) was administered within the last 4 weeks. For characterization of the immune system, the following serum tests were obtained: comprehensive metabolic panel, total protein electrophoresis, quantitative immunoglobulins, immunoglobulin G subclasses, serum C3, serum C4, serum CH50, antibody titers for tetanus, diphtheria, hepatitis B, and measles, mumps, and rubella (MMR). Complete blood counts with differential and B/T/natural killer (NK) cell counts were obtained from anticoagulated whole blood. Complete blood count was determined using Sysmex XE-5000 (Sysmex America, Inc, Mundelein, Illinois). Comprehensive metabolic panel was performed using standard laboratory techniques. Rubella and hepatitis B antibody levels were measured using chemiluminescent immunoassay. Measles and mumps antibody levels were measured using multiplex flow immunoassay. Tetanus and diphtheria antibody titers were measured with enzyme immunoassay. C3 and C4 levels were measured using immunoturbidimetry. CH50 was collected, placed on ice, and measured using enzyme immunoassay. B-cell, T-cell, and NK-cell subset counts were measured by direct immunofluorescence using flow cytometric techniques. Immunoglobulin A, G, M levels, and IgG subclasses were measured by nephelometry. Total protein electrophoresis was performed using capillary electrophoresis. All results were assessed based on the University of 332
Vol. 167, No. 2 Michigan’s standard reference laboratory values, with the exception of NK cell counts and percentages which were derived from the study “Lymphocyte subsets in healthy children from birth through 18 years of age.”17 Statistical Analyses Immune characteristics in each group were presented using frequency with percentage for categorical variables and median with range for continuous variables. Group comparisons between the PLE and non-PLE groups were made using Fisher exact test for categorical variables and Wilcoxon rank sum test for continuous variables. All analyses were performed using SAS v 9.3 (SAS Institute Inc, Cary, North Carolina), with statistical significance set at a P value of 500 cells/mL) (Table II). In contrast, 88% of the control group patients had a low CD4 cell count (median: 450 cells/mL), with a Magdo et al
ORIGINAL ARTICLES
August 2015
Table I. Comparison of patient and clinical characteristics in patients post-Fontan with and without PLE Characteristics
PLE (N = 8)
Controls (N = 8)
P value*
Male sex Age at stage II palliation, y Age at Fontan, y Type of Fontan
4 (50.0) 0.6 (0.4-1.7) 2.1 (1.5-4.5)
2 (25.0) 0.5 (0.4-1.1) 2.1 (1.6-11.1)
.61 .29 .80
Lateral tunnel Extracardiac
6 (75.0) 2 (25.0)
6 (75.0) 2 (25.0)
1.00
6 (75.0) 0 (0.0) 1 (12.5) 1 (12.5) 0 (0.0)
3 (37.5) 1 (12.5) 0 (0.0) 0 (0.0) 4 (50.0)
7 (87.5) 1 (12.5) 0 (0.0) 1 (12.5) 16 (9-145) 1 (12.5)
5 (62.5) 1 (12.5) 2 (25.0) 1 (12.5) 15 (6-32) 2 (25.0)
1.00 .62 1.00
7 (87.5)
6 (75.0)
1.00
5 (62.5) 2 (25.0) 1 (12.5)
6 (75.0) 2 (25.0) 0 (0.0)
1.00
7 (87.5) 8 (100.0) 8 (100.0)
8 (100.0) 7 (87.5) 8 (100.0)
1.00 N/A N/A
.31†
Type of ventricle
Right dominant systemic ventricle Left dominant systemic ventricle Indeterminant ventricle Heterotaxy Post-Fontan hospital length of stay, d Readmitted within 30 d of Fontan discharge Sternotomy as a neonate Thymectomy
As neonate Older than neonate Unknown
PLE (N = 8)
Controls (N = 8)
P value*
16.8 (10.7-19.1) 14.5 (13-17.3) .47 Hemoglobin, g/dl 2 (25.0) 0 (0.0) .47 Low hemoglobin 47.3 (32-55.6) 43.6 (38.3-49.9) .36 Hematocrit, % 2 (25.0) 0 (0.0) .47 Low hematocrit 7.2 (4.2-13) 5.9 (3-9.8) .27 WBC count, K/mL 0 (0.0) 1 (12.5) 1.00 Low WBC count 82.4 (69.6-89.4) 71 (51.8-78) .02 % Neutrophils 0 (0.0) 0 (0.0) N/A Low % neutrophils 5.6 (4.6-16.6) 20.4 (12.2-32.2) .01 % Lymphocytes 0.6 (0.3-1.0) 1.2 (0.6-1.8) .01 ALC, K/mL 8 (100.0) 2 (25.0) .01 Low % lymphocytes or ALC Lymphocyte subsets
.57
z
Vaccine history
Diphtheria/tetanus Hepatitis B MMR
Characteristics CBC with WBC differential
Cardiac diagnosis
HLHS Unbalanced AVSD DORV with HLV Tricuspid atresia Other
Table II. Comparison of CBC, WBC differential, lymphocyte subsets in patients post-Fontan with and without PLE
AVSD, atrioventricular septal defect; DORV, double outlet right ventricle; HLHS, hypoplastic left heart syndrome; HLV, hypoplastic left ventricle; N/A, not applicable. Data are presented as N (%) for categorical variables and median (range) for continuous variables. *P value from Fisher exact test for categorical variables and Wilcoxon rank sum test for continuous variables on comparison of each characteristic between the patients with PLE and healthy controls. †Comparison was made as HLHS vs all others and P value was from Fisher exact test. zComparison was made as right dominant systemic ventricle vs left dominant systemic ventricle or indeterminant ventricle and P value was from Fisher exact test.
normal CD4 percentage at 36.9% of lymphocytes. The patients with PLE tended to have a lower CD8 (cytotoxic T cell) count than the control group patients (median: 90.5 cells/mL vs 207 cells/mL, P = .07; normal range: 255-649), however, the CD8 percentage was similarly low in both groups (17% in the PLE group vs 16.2% in controls, P = .76; normal range 19%-29%). The majority of patients with PLE (75%) had reversed CD4:CD8 (
Study design Patients with PLE post-Fontan and age-matched controls post-Fontan were prospectively studied with laboratory markers of immune function. Infectious history was obtained by interview and chart review. The groups’ demographics, cardiac history, immune characteristics, and infection history were compared using appropriate 2-group statistics. Results A total of 16 patients enrolled (8 patients with PLE and 8 controls). All patients with PLE had lymphopenia compared with 25% of controls (P = .01). All patients with PLE had markedly depressed CD4 T cell counts (median 58 cells/mL) compared with controls (median 450 cells/mL, P = .0002); CD4% was also low in the PLE group (12.3%) and normal in control (36.9%, P = .004). Both groups had mildly depressed CD8 T cells and normal to slightly elevated natural killer and B-cell subsets. A majority of patients with PLE (62.5%) had negative titers to measles, mumps, and rubella vaccination, compared with no control Fontan with a negative titer (P = .03). Despite profoundly low CD4 counts, the frequency of infection was not different between groups with no reported opportunistic infections. Conclusions Patients with Fontan-associated PLE have extensive quantitative immune abnormalities, particularly CD4 deficiency. These immune abnormalities are similar to those found in non-Fontan patients with PLE caused by intestinal lymphangiectasia. (J Pediatr 2015;167:331-7).
P
rotein-losing enteropathy (PLE) is a complication of the Fontan palliation for single-ventricle congenital heart disease. Although PLE occurs in only about 5% of patients post-Fontan, it confers marked morbidity and high mortality within 5 years of diagnosis.1-3 Fontan-related PLE is episodic with “flares” in symptom severity characterized by severe enteral protein loss leading to peripheral edema, effusions, and ascites.4,5 The pathogenesis of Fontan-related PLE is not completely understood, and it is unclear why it develops in some patients post-Fontan and not others.1,4,5 PLE has been described in many different diseases. PLE that results from some diseases, such as primary intestinal lymphangiectasia (IL) and constrictive pericarditis, has been associated with significant immune abnormalities, such as lymphopenia, with a selective loss of CD4 cells and severe hypogammaglobulinemia.6-9 However, patients with PLE associated with inflammatory diseases such as ulcerative colitis, celiac disease, and allergic PLE do not have these immune derangements.10-12 Immune abnormalities have been reported in Fontan-related PLE. In small case series or case reports, patients with PLE postFontan have reported reduced immunoglobulin levels,13,14 lymphopenia,11 and selective CD4 lymphocyte deficiency.14-16 However, these studies were limited in number of patients, the extent of immune characterization, and/or lack of a control group. As some lymphocyte abnormalities have been found in patients who have undergone neonatal thymectomy, and to control for the possibility that the Fontan circulation itself may influence the immune system, we sought to comprehensively characterize and compare the immune system in individuals receiving Fontan with and without PLE. We hypothesized that patients with PLE post-Fontan have distinct quantitative immune abnormalities compared with those without PLE and may share similarities with PLE caused by other diseases.
Methods This is a prospective, case-control study of patients with and without PLE post-Fontan. The study was approved by the University of Michigan Institutional
GI IL IVIG MMR NK PLE
Gastrointestinal Intestinal lymphangiectasia Intravenous immunoglobulin Measles, mumps, and rubella Natural killer Protein-losing enteropathy
From the 1University of Michigan Congenital Heart Center, C.S. Mott Children’s Hospital; 2University of Michigan Pediatric Infectious Disease; 3University of Michigan Allergy and Immunology; 4University of Michigan College of Pharmacy; and 5University of Michigan, Michigan Congenital Heart Outcomes Research and Discovery Ann Arbor, MI H.M. received a grant from University of Michigan Congenital Heart Center (G012107). The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2015.04.061
331
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Review Board. Informed consent was obtained from each patient if 18 years old or older, or from the subjects’ parent/legal guardian if less than 18 years old. Eligible subjects were identified from institutional records and approached for participation in the hospital or clinic setting. All patients enrolled were required to have single ventricle congenital heart disease and have undergone a total cavopulmonary anastomosis. A patient with Fontanassociated PLE was defined as having a clinical diagnosis of PLE made by a treating cardiologist, a history of hypoalbuminemia, and resultant clinical symptoms requiring previous hospitalization for PLE-specific treatment. An uncomplicated control was defined as any patient status post-Fontan palliation without significant comorbid disease including Fontan hepatopathy, plastic bronchitis, known hematologic or immunologic disorders, any other non-cardiac major organ dysfunction, or a previous diagnosis of PLE. Controls were matched to patients with PLE post-Fontan based on age (12 months). Exclusion criteria included concurrent febrile illness, antibiotic use, or systemic steroid use within 3 weeks of blood collection. Patients post-Fontan taking oral or inhaled budesonide therapy were not excluded. From a combination of chart review and patient interview, the following clinical information was gathered for each subject: underlying congenital heart disease, age at stage II palliation, date of Fontan, type of Fontan, length of hospital stay after Fontan, readmission within the first month after Fontan, history of neonatal thymectomy or thymectomy after 1 month of age, total or subtotal thymectomy, infection history, and immunization history. For subjects with PLE, information was obtained regarding the clinical status of the disease (whether symptoms were currently active) and whether albumin or intravenous immunoglobulin (IVIG) was administered within the last 4 weeks. For characterization of the immune system, the following serum tests were obtained: comprehensive metabolic panel, total protein electrophoresis, quantitative immunoglobulins, immunoglobulin G subclasses, serum C3, serum C4, serum CH50, antibody titers for tetanus, diphtheria, hepatitis B, and measles, mumps, and rubella (MMR). Complete blood counts with differential and B/T/natural killer (NK) cell counts were obtained from anticoagulated whole blood. Complete blood count was determined using Sysmex XE-5000 (Sysmex America, Inc, Mundelein, Illinois). Comprehensive metabolic panel was performed using standard laboratory techniques. Rubella and hepatitis B antibody levels were measured using chemiluminescent immunoassay. Measles and mumps antibody levels were measured using multiplex flow immunoassay. Tetanus and diphtheria antibody titers were measured with enzyme immunoassay. C3 and C4 levels were measured using immunoturbidimetry. CH50 was collected, placed on ice, and measured using enzyme immunoassay. B-cell, T-cell, and NK-cell subset counts were measured by direct immunofluorescence using flow cytometric techniques. Immunoglobulin A, G, M levels, and IgG subclasses were measured by nephelometry. Total protein electrophoresis was performed using capillary electrophoresis. All results were assessed based on the University of 332
Vol. 167, No. 2 Michigan’s standard reference laboratory values, with the exception of NK cell counts and percentages which were derived from the study “Lymphocyte subsets in healthy children from birth through 18 years of age.”17 Statistical Analyses Immune characteristics in each group were presented using frequency with percentage for categorical variables and median with range for continuous variables. Group comparisons between the PLE and non-PLE groups were made using Fisher exact test for categorical variables and Wilcoxon rank sum test for continuous variables. All analyses were performed using SAS v 9.3 (SAS Institute Inc, Cary, North Carolina), with statistical significance set at a P value of 500 cells/mL) (Table II). In contrast, 88% of the control group patients had a low CD4 cell count (median: 450 cells/mL), with a Magdo et al
ORIGINAL ARTICLES
August 2015
Table I. Comparison of patient and clinical characteristics in patients post-Fontan with and without PLE Characteristics
PLE (N = 8)
Controls (N = 8)
P value*
Male sex Age at stage II palliation, y Age at Fontan, y Type of Fontan
4 (50.0) 0.6 (0.4-1.7) 2.1 (1.5-4.5)
2 (25.0) 0.5 (0.4-1.1) 2.1 (1.6-11.1)
.61 .29 .80
Lateral tunnel Extracardiac
6 (75.0) 2 (25.0)
6 (75.0) 2 (25.0)
1.00
6 (75.0) 0 (0.0) 1 (12.5) 1 (12.5) 0 (0.0)
3 (37.5) 1 (12.5) 0 (0.0) 0 (0.0) 4 (50.0)
7 (87.5) 1 (12.5) 0 (0.0) 1 (12.5) 16 (9-145) 1 (12.5)
5 (62.5) 1 (12.5) 2 (25.0) 1 (12.5) 15 (6-32) 2 (25.0)
1.00 .62 1.00
7 (87.5)
6 (75.0)
1.00
5 (62.5) 2 (25.0) 1 (12.5)
6 (75.0) 2 (25.0) 0 (0.0)
1.00
7 (87.5) 8 (100.0) 8 (100.0)
8 (100.0) 7 (87.5) 8 (100.0)
1.00 N/A N/A
.31†
Type of ventricle
Right dominant systemic ventricle Left dominant systemic ventricle Indeterminant ventricle Heterotaxy Post-Fontan hospital length of stay, d Readmitted within 30 d of Fontan discharge Sternotomy as a neonate Thymectomy
As neonate Older than neonate Unknown
PLE (N = 8)
Controls (N = 8)
P value*
16.8 (10.7-19.1) 14.5 (13-17.3) .47 Hemoglobin, g/dl 2 (25.0) 0 (0.0) .47 Low hemoglobin 47.3 (32-55.6) 43.6 (38.3-49.9) .36 Hematocrit, % 2 (25.0) 0 (0.0) .47 Low hematocrit 7.2 (4.2-13) 5.9 (3-9.8) .27 WBC count, K/mL 0 (0.0) 1 (12.5) 1.00 Low WBC count 82.4 (69.6-89.4) 71 (51.8-78) .02 % Neutrophils 0 (0.0) 0 (0.0) N/A Low % neutrophils 5.6 (4.6-16.6) 20.4 (12.2-32.2) .01 % Lymphocytes 0.6 (0.3-1.0) 1.2 (0.6-1.8) .01 ALC, K/mL 8 (100.0) 2 (25.0) .01 Low % lymphocytes or ALC Lymphocyte subsets
.57
z
Vaccine history
Diphtheria/tetanus Hepatitis B MMR
Characteristics CBC with WBC differential
Cardiac diagnosis
HLHS Unbalanced AVSD DORV with HLV Tricuspid atresia Other
Table II. Comparison of CBC, WBC differential, lymphocyte subsets in patients post-Fontan with and without PLE
AVSD, atrioventricular septal defect; DORV, double outlet right ventricle; HLHS, hypoplastic left heart syndrome; HLV, hypoplastic left ventricle; N/A, not applicable. Data are presented as N (%) for categorical variables and median (range) for continuous variables. *P value from Fisher exact test for categorical variables and Wilcoxon rank sum test for continuous variables on comparison of each characteristic between the patients with PLE and healthy controls. †Comparison was made as HLHS vs all others and P value was from Fisher exact test. zComparison was made as right dominant systemic ventricle vs left dominant systemic ventricle or indeterminant ventricle and P value was from Fisher exact test.
normal CD4 percentage at 36.9% of lymphocytes. The patients with PLE tended to have a lower CD8 (cytotoxic T cell) count than the control group patients (median: 90.5 cells/mL vs 207 cells/mL, P = .07; normal range: 255-649), however, the CD8 percentage was similarly low in both groups (17% in the PLE group vs 16.2% in controls, P = .76; normal range 19%-29%). The majority of patients with PLE (75%) had reversed CD4:CD8 (
