Chronic granulomatous disease with inflammatory bowel disease: Clinical presentation, treatment, and outcomes from the USIDNET Registry

Abstract

Background:

Chronic granulomatous disease (CGD) is an inborn error of immunity caused by defects in the phagocytic NADPH oxidase complex, leading to increased susceptibility to infection and inflammatory autoimmune diseases. Up to 50% of patients have gastrointestinal (GI) involvement and meet diagnostic criteria for inflammatory bowel disease (CGD-IBD).

Objective:

We analyzed CGD patients from the United States Immunodeficiency Network (USIDNET) registry to determine whether IBD changes the presentation, treatment, and outcomes of CGD patients.

Methods:

A retrospective evaluation of CGD cases from the USIDNET registry was completed. CGD-IBD was defined as the presence of any major physician-reported inflammatory, non-infectious GI disease manifestation. Demographic information, conditions, infections, antimicrobial therapies, immunomodulator use, and hematopoietic stem cell transplantation data were analyzed.

Results:

Of 194 patients with a diagnosis of CGD, 96 met criteria for IBD, and 98 were categorized in the non-IBD group. CGD-IBD patients had an increased rate of infection compared to the non-IBD group (0.66 versus 0.36 infections/patient/year). Enteric organism infections were more common in IBD patients. Immunomodulators were used at a significantly higher percentage in IBD patients compared to non-IBD patients (80% versus 56%, p < 0.001). Of the entire CGD cohort, 17 patients died (8.8%), with no significant difference between IBD and non-IBD patients (p = 1.00).

Conclusion:

Infectious events, enteric organism infections, and use of immunomodulatory drugs were higher in IBD than non-IBD patients; however, mortality was not increased. Patients with CGD and concurrent IBD are at increased risk for disease complications, supporting the importance of early recognition, diagnosis, and treatment.

INTRODUCTION

Chronic granulomatous disease (CGD) is a monogenic inborn error of immunity (IEI) in which a defect in the NADPH oxidase complex leads to inadequate killing by phagocytes (1, 2). The NADPH oxidase complex is composed of multiple subunits (gp91phox, p22phox, p47phox, p67phox, and p40phox) required for generation of reactive oxygen intermediates (ROIs). The majority of cases are X-linked recessive due to mutations in the CYBB gene (encoding gp91phox), with the remainder of cases due to autosomal recessive mutations in CYBA, NCF1, NCF2, NCF4, and the recently identified CYBC1, encoding for a protein required for gp91phox and p22phox expression. Loss of function of any of these genes results in defective phagocytic superoxide radical generation, leading to recurrent and life-threatening characteristic bacterial and fungal infections. In addition to an increased susceptibility to infection, patients with CGD can develop heightened inflammatory responses and generate granulomas around sites of infection or inflammation (2–5).

Up to 50% of patients with CGD have gastrointestinal (GI) involvement, with estimates ranging from 30%–60% (6–9). Many of these patients meet diagnostic criteria for inflammatory bowel disease (CGD-IBD), with diagnoses dependent upon endoscopy and biopsy (10–14). IBD manifestations can present prior to the onset of serious infections and may manifest as the initial disease presentation of CGD, however, the development of GI symptoms may not occur until adolescence (7, 8, 15, 16). The treatment of CGD-IBD patients is challenging, as immunosuppressant therapies for IBD lower the threshold for infection (8, 17–19). Multiple studies have demonstrated that hematopoietic stem cell transplantation (HSCT) leads to complete resolution of CGD, including IBD colitis (16, 20–24).

Patients with CGD and concurrent IBD are at increased risk for complications, supporting the importance of early recognition and diagnosis. The United States Immunodeficiency Network (USIDNET) is a research consortium and registry of primary immunodeficiency disease (PIDD) patients established specifically to support scientific research in the field of PIDDs (25). We elected to analyze CGD patients from the USIDNET cohort to determine whether a diagnosis of IBD may change the presentation, treatment, and outcomes of CGD patients, as compared to those without IBD.

METHODS

PATIENT POPULATION

A retrospective evaluation of data from the USIDNET registry was completed. The USIDNET includes 42 enrolling centers across North America, including sites in 23 states and large academic institutions. The primary data source for entry in the registry is the patient’s clinical record.

The registry database was queried for data entries through 2019 for a diagnosis of CGD, yielding 199 patients of which 96 had IBD, 98 did not have IBD and 5 were excluded due to indeterminate IBD status. Inclusion of subject data in the USIDNET registry requires informed consent as well as Institutional Review Board approval for each center. Queried data were deidentified by the USIDNET.

The initial query included data regarding basic demographic information, genotypes, symptoms and conditions, infections, antimicrobial therapies, use of immunomodulatory drugs, surgical interventions, and hematopoietic stem cell transplantation (HSCT). Additional queries were submitted to the primary submitting institutions to clarify genotypes on seven patients.

CLINICAL PHENOTYPE SCORING

All reported conditions and symptoms related to the GI tract were reviewed and scored. The primary outcome, CGD-associated inflammatory bowel disease (CGD-IBD), was a binary variable defined as the presence of any major physician-reported inflammatory non-infectious disease manifestation of the gastrointestinal tract, including Crohn disease, ulcerative colitis, and/or endoscopy findings consistent with IBD. Specific disease terms and diagnoses of Crohn disease or colitis were reported by the submitting provider to the USIDNET database. Additional GI disease findings that met inclusion criteria for CGD-IBD included the following conditions: GI fistulas, GI strictures, GI obstruction, and proctitis (Table 1).

Table 1.

Gastrointestinal symptoms and features specific to CGD patients with IBD.

IBD GI Disease Manifestations	Number of IBD Patients (n=96)
Anal fissure	1
Crohn disease/colitis*	79
Fistula	22
Granuloma	14
Obstruction	16
Proctitis	4
Stricture	12

^*Disease terms as submitted by providers to the USIDNET database.

The non-IBD CGD group was defined as patients without any criteria or conditions consistent with IBD and without other inflammatory manifestations of the gastrointestinal tract. Non-inflammatory (non-IBD) GI symptoms included abdominal pain, constipation, gastroesophageal reflux, and gastritis (Table 2). Perirectal abscesses were not categorized as an IBD condition, as we were unable to rule out infection as a cause of the abscess. Chronic diarrhea was not considered a condition which met criteria for IBD, as other infectious or noninflammatory causes could not be ruled out from the available data. The presence of increased eosinophils in the GI tract was not considered in the scoring of IBD versus non-IBD categories.

Table 2.

Additional gastrointestinal symptoms in IBD versus non-IBD groups, divided by type of manifestation.

GI Disease Manifestations	Number of Patients (n=194)	IBD Patients (n=96)	Non-IBD Patients (n=98)
Abdominal pain	52	42	10
Chronic diarrhea	24	24	0
Constipation	17	11	6
GERD	32	25	7
Gastritis	7	5	2
Perirectal abscess	21	14	7
Eosinophilic esophagitis	5	5	0

STATISTICAL ANALYSES

The data analysis for this paper was generated using SAS software, Version 9.4 of the SAS System for Windows. (Copyright © 2016 SAS Institute Inc. SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc., Cary, NC, USA). Basic demographic characteristics of all CGD patients are listed in Table 3. Current age and age at diagnosis are summarized as medians and interquartile ranges (IQR), as the distributions were not Gaussian, and were compared using a Wilcoxon rank-sum test. Sex and race/ethnicity distributions between the two groups were compared using the chi-square test. Patients with X-linked versus autosomal recessive genotypes were compared using the chi-square test. Mortality between groups was analyzed with an adjusted chi-square test (Yates correction).

Table 3.

Patient demographics.

	Overall (n=194)	IBD (n=96)	Non-IBD (n=98)
Median age (Yrs; range)	24 (1–52)	24 (5–51)	23 (1–52)
Sex
Male	169 (87.1%)	86 (90%)	83 (84.7%)
Female	25 (12.9%)	10 (10.4%)	15 (15.3%)
Inheritance
XL	123 (63.4%)	65 (67.7%)	58 (59.2%)
AR	47 (24.2%)	19 (19.8%)	28 (28.6%)
Unknown	24 (12.4%)	12 (12.5%)	12 (12.2%)
Median Age (Yrs) at CGD
Diagnosis	1.5 (0–24.3)	1.5 (0–24.3)	1.3 (0–22)
XL	1.0	1.5	0.9
AR	3.4	2.0	4.0
Unknown age	N=51	N=18	N=33
Median Age (Yrs) at Onset of Symptoms	0.5 (0–19.6)	0.5 (0–19.6)	0.7 (0–14.6)
XL	0.4	0.5	0.4
AR	1.7	0.5	3.1
Unknown age	N=52	N=22	N=30
Race/Ethnicity
White/Caucasian	121 (62.4%)	66 (68.8%)	55 (56.1%)
Hispanic or Latino	13 (6.7%)	5 (5.2%)	8 (8.2%)
Black/African American	28 (14.4%)	9 (9.4%)	19 (19.4%)
Asian or Pacific Islander	5 (2.6%)	5 (5.2%)	0 (0%)
Other/More than one race	5 (2.6%)	2 (2.1%)	3 (3.1%)
Unknown/Not reported	22 (11.3%)	9 (9.4%)	13 (13.3%)
HSCT (%)	61 (31.4%)	35 (36.5%)	26 (26.5%)
Immunomodulator therapy (%) *	132 (68.0%)	77 (80%)	55 (56%)
Mortality	N=17 (8.8%)	N=8 (8.3%)	N=9 (9.2%)
Median age of death (Yrs)	21.8 (2.5–48.1)	19.3 (2.5–48.1)	23.9 (6.2–37.4)
XL	10 (58.8%)	4 (50%)	6 (66.7%)
AR	4 (23.5%)	2 (25%)	2 (22.2%)
Unknown	3	2	1
HSCT	8 (47.1%)	4 (50%)	4 (44.4%)

^*Immunomodulator therapy p = 0.0003, indicating a statistically significant difference between IBD and non-IBD patients. P-values for Sex, Inheritance, HSCT, and Mortality were not statistically significant (p > 0.05). P-values for Age at Onset of Symptoms and Age at CGD Diagnosed were unable to be calculated given the amount of missing data. P-value for Race/Ethnicity was not calculated due to the predominance of white/Caucasian subjects.

Differences in the proportion of patients with X-linked CGD and autosomal recessive CGD, and in the outcomes between the IBD and non-IBD groups were compared using the chi-square test. Notably, genotype information was missing in 12.4% of the CGD cohort (n = 24). A simple logistic regression analysis was completed to determine the likelihood of use of immunomodulatory drugs in the IBD versus non-IBD categories. For our analysis, comparisons were performed assuming that measurements were missing at random. However, this is an important limitation of this analysis and is considered in the interpretation of the data.

RESULTS

BASIC DEMOGRAPHIC AND CLINICAL CHARACTERISTICS

The USIDNET query provided information for 199 total patients with a diagnosis of CGD. Information was provided by 27 providers from 26 enrolling centers across the United States (see Supplementary Table E1).

Data on sex and year of birth were available for all patients. There was a predominance of males (87.1%) in the total CGD cohort, with no significant difference in the distribution of males and females between the IBD and non-IBD groups (p=0.31; Table 3). The median age of all CGD patients was 24 years (range 1–52 years). Of the 142 patients with age of symptom onset or age of diagnosis data available, the first symptoms of CGD occurred at a median age of 0.5 years (range 0–19.6 years), while the median age of diagnosis was 1.5 years (range 0–24.3 years). Patients without IBD developed symptoms of CGD at a median age of 0.7 years (range 0–19.6 years) and were diagnosed at a median age of 1.3 years (range 0–24.3 years), whereas patients with IBD developed symptoms of CGD at a median age of 0.5 years (range 0–14.6 years) and were diagnosed at a median age of 1.5 years (range 0–22 years). The median interval between age of onset and age of diagnosis was 1 year for IBD patients and 0.6 years for non-IBD patients. The majority of subjects in the CGD cohort were reported as white/Caucasian (62.4%; Table 3); the second most common ethnicity or race was black/African American (14.4%), followed by Hispanic/Latino (6.7%), Asian/Pacific Islander (2.6%), and other/more than one race (2.6%). There were 22 patients with unknown ethnicity/race or ethnicity/race that was not reported (11.3%; Supplementary Table E2). Trends were similar between the IBD and non-IBD groups, but notably, all Asian patients in the cohort were found to have IBD (n=5).

Survival status was available in 189 patients (97.4%; Table 3). Five patients had unknown survival status. As of the year 2018, a total of 17 patients had died (8.8%), with a median age at death of 21.8 years (range 2.5–48.1 years). For one patient, age of death was unknown. Seven of these patients had CGD-IBD. The majority of patients who died had X-linked CGD (58.8%), with no significant difference between IBD and non-IBD patients (p=ns). Eight of the 17 patients who died had received HSCT, of which 4 had CGD-IBD. Causes of death among transplanted patients included acute graft-versus-host disease (aGvHD) complicated by infections or multiple organ failure (n=2), respiratory failure (n=2), infections and organ failure (n=2), uremia and electrolyte unbalances (n=1), and unspecified complications of HSCT (n=1). Causes of death among the remaining 9 patients who had died without receiving a prior HSCT, included infections (n=4), malignancies complicated by vascular events (n=2), fatal hemorrhage during pneumonectomy (n=1), and car accident (n=1). The cause of death was not known in one patient (Supplementary Table E3).

GENOTYPE

The majority of CGD patients were found to have X-linked disease with a mutation in the CYBB gene (63.4%), followed by autosomal recessive (AR) mutations in NCF1 (13.9%), CYBA (7.7%), and NCF2 (2.6%; Supplementary Table E4). Genotype frequency between the IBD and non-IBD groups was not significantly different (p=0.147). The genetic etiology of CGD was not known for 24 patients (12.3%). There were more males than females in every genotype group with the exception of NCF2 mutations (3 females and 2 males).

GASTROINTESTINAL-RELATED SYMPTOMS AND CONDITIONS

Crohn disease and inflammatory colitis are the most common inflammatory gastrointestinal conditions among CGD patients. Every CGD patient with chronic diarrhea was found to have IBD.

Crohn disease or colitis were the most common inflammatory gastrointestinal disease manifestations in the IBD-CGD group (n=79) followed by GI fistulas (n=22; Table 1). Additional non-inflammatory GI symptoms were shared by both the IBD and non-IBD groups, including abdominal pain, constipation, GERD, and gastritis (Table 2). Five IBD patients were noted to have eosinophilic esophagitis. The most common GI disease manifestation in the non-IBD-CGD group was abdominal pain (n=10). Every patient noted to have chronic diarrhea had other findings consistent with IBD, except for the five excluded patients, as described in the methods. In addition, all subjects with a diagnosis of eosinophilic esophagitis also met criteria for IBD. Although perirectal abscesses and GERD were more common in the IBD group, non-IBD patients were also noted to have perirectal abscesses and GERD.

INFECTIONS

CGD Patients with inflammatory bowel disease have an increased frequency of infections.

Infectious events were categorized by type or location of infection (Table 4). The IBD group had an average of 83% more infections per patient, per year, as compared to the non-IBD group: 0.66 versus 0.36, respectively. The most common type of infection in both groups was lower respiratory tract (26.8% IBD, 33.3% non-IBD), followed by skin infections (15.1% IBD, 17.4% non-IBD). Perirectal abscesses caused by infection were more frequent in the IBD group (8.2% IBD, 2% non-IBD). There were many unknown types of infections in each group (e.g., organism was known, but location of infection was not known), with 9.7% in the IBD group and 6% in the non-IBD group (data not shown).

Table 4.

Infectious events in IBD versus non-IBD groups, categorized by type/location of infection.

Type of Infection	IBD (% of total IBD infectious events)	Non-IBD (% of total Non-IBD infectious events)
Infections/Patient/Year *	0.66	0.36
Lower respiratory tract (pneumonia, lung abscess)	271 (26.8%)	166 (33.3%)
Skin (superficial, subcutaneous, abscess)	153 (15.1%)	87 (17.4%)
Upper respiratory tract (OM, URI, sinusitis, pharyngitis, tracheitis)	90 (8.9%)	58 (11.6%)
Perirectal abscess	83 (8.2%)	10 (2.0%)
Lymphadenitis	67 (6.6%)	35 (7.0%)
Abscess, unspecified	54 (5.3%)	25 (5.0%)
Liver abscess	39 (3.8%)	31 (6.2%)
Genitourinary (UTI, pyelonephritis, renal abscess, genital, prostatitis)	39 (3.8%)	9 (1.8%)
Bone/joint (osteomyelitis, septic arthritis)	32 (3.2%)	9 (1.8%)
Infectious diarrhea/colitis	28 (2.8%)	20 (4.0%)
Neurologic (meningitis, encephalitis, brain abscess)	24 (2.4%)	5 (1.0%)
Viremia	17 (1.7%)	2 (0.4%)
Other (pericarditis, peritonitis, conjunctivitis)	8 (0.8%)	8 (1.6%)
Bacteremia	8 (0.8%)	2 (0.4%)
Fungemia	2 (0.2%)	2 (0.4%)
Unknown	98 (9.7%)	30 (6.0%)
Total	1013	499

OM = otitis media; URI = upper respiratory infection; UTI = urinary tract infection

^*Number of years = Reported age minus age at diagnosis of CGD

Data regarding the timing of infections, including whether infections occurred before or after immunomodulator use, or before or after HSCT, were not available. Thus, the average number of infections per patient was calculated in the IBD and non-IBD groups, then further subcategorized into whether there was a history of immunomodulator use and whether the patient underwent HSCT (Figure 1). Non-IBD patients, on average, had fewer infections per patient as compared to IBD patients in every category. IBD patients who underwent HSCT had the highest average number of infections per patient, and non-IBD patients who had not received immunomodulator therapies had the fewest average number of infections per patient. Notably, given the limitations of the registry, the number of infections per patient per year was calculated by identifying the amount of time between age of diagnosis of CGD and the most recent age reported in the database for each patient, which would provide the duration of observation for each patient.

Figure 1.

Average number of infectious events per patient based on specific therapies. Patients with IBD had more infections than those without IBD in every category. IBD patients who underwent HSCT had the highest average number of infections per patient, and non-IBD patients who had not received immunomodulator therapies had the fewest average number of infections per patient.

Infections with enteric organisms are more common in IBD patients; Aspergillus is the most commonly identified source of infections in CGD patients with and without inflammatory bowel disease.

The specific causal organisms of infections were evaluated, with a total of 78 different pathogens identified by the submitting centers (Table 5 and Supplementary Table E5). Six of these microbes were representative of CGD-characteristic infections, including Aspergillus, Staphylococcus species, Serratia, Nocardia, Burkholderia, and Candida (Table 5). A causal micro-organism was not identified in 48.3% of all infectious events. Among the infections with an identified pathogen, the most common was Aspergillus in both the IBD and non-IBD groups, with about 10% of the total infections being attributed to the fungus. Regarding the frequency of infections caused by enteric organisms, there were 14 Enterobacter infections in the IBD group, compared to one in the non-IBD group. Similarly, there were 18 and 8 occurrences of Klebsiella infection in the IBD and non-IBD groups, respectively. Overall, infections with enteric organisms (Enterobacter, Enterococcus, Escherichia, and Klebsiella) were more common in patients with IBD than without. With regards to viral infections, herpes simplex virus (HSV) infection was reported 9 times in the IBD group while none were reported in the non-IBD group (Supplementary Table E5). Cytomegalovirus (CMV) infection was the most frequent viral infection in both groups, with 11 occurrences in the IBD group and 5 occurrences in the non-IBD group.

Table 5.

Characteristic CGD pathogens.

Organism	IBD (% of total IBD infectious events)	Non-IBD (% of total Non-IBD infectious events)
Aspergillus	105 (10.4%)	48 (9.6%)
Staphylococcus species	67 (6.6%)	36 (7.2%)
Serratia	38 (3.8%)	25 (5.0%)
Nocardia	37 (3.7%)	15 (3.0%)
Burkholderia	30 (3.0%)	20 (4.0%)
Candida	25 (2.5%)	16 (3.2%)

Prophylactic antibacterial and antifungal medications were used at similar frequencies in the IBD and non-IBD groups. Trimethoprim-sulfamethoxazole (TMP-SMX) was the most commonly prescribed antimicrobial, followed by itraconazole (Table 6). Additional medications used for prophylaxis are noted in Supplementary Table E6. Interferon-gamma (IFN-γ) prophylaxis and treatment are discussed below, as the use of this drug for prophylaxis versus treatment could not be distinguished.

Table 6.

Anti-Infective Prophylaxis

Drug	IBD (n=96)	Non-IBD (n=98)
Trimethoprim-Sulfamethoxazole	53 (55.2%)	43 (43.9%)
Itraconazole	34 (35.4%)	30 (31.3%)
Posaconazole	15 (15.6%)	11 (11.2%)
Voriconazole	5 (5.2%)	7 (7.3%)

IMMUNOMODULATOR THERAPY

Immunomodulators are frequently used in CGD patients with and without inflammatory bowel disease.

Use of immunomodulatory drugs was frequent in both groups (Table 3 and Supplementary Table E7), but significantly higher in IBD patients compared to non-IBD patients (80% versus 56%, p < .001). The most commonly prescribed immunomodulator overall was IFN-γ, with reported use in 41.2% of CGD patients, and with similar utilization in the non-IBD group (40.8%, versus 41.7% in CGD-IBD). The presence of IBD, as compared to no IBD, increased the odds of immunomodulator use overall (OR = 3.17; 95% Confidence Interval: 1.67–6.03, p < .001). Systemic corticosteroids were also frequently used (41.2% of all CGD patients) although there was no available data on treatment duration or dosing. Immunomodulators were utilized more frequently in the IBD compared to non-IBD group. In the non-IBD group, monoclonal antibodies and JAK inhibitors were only utilized four times, compared to 29 occurrences in the IBD group. Limited data was available regarding the indication or diagnosis for which the immunomodulators were given, or whether these were administered before or after HSCT. As shown in Figure 1, patients who had received immunomodulatory treatment were found to have a higher average number of infections as compared to those who did not receive these medications.

HEMATOPOIETIC STEM CELL TRANSPLANTATION

All patients were categorized as having undergone HSCT or not. Approximately one-third of all patients underwent HSCT (31.4%). A larger percentage of patients with IBD received transplant as compared to non-IBD patients, however this difference was not statistically significant (36.5% versus 26.5%, respectively; p = .137). The majority of patients who underwent HSCT had XL mutations (72.1%), with XL-CGD HSCT patients constituting 22.7% of the entire CGD cohort (Table 7). Transplanted patients were also found to have a greater average number of infections (Figure 1).

Table 7.

Hematopoietic stem cell transplantation (HSCT) data.

Mutation	IBD (% of total IBD patients)	Non-IBD (% of total non-IBD patients)	Total (% of HSCT patients)	% of Total CGD Cohort
X-linked	28 (29.2%)	16 (16.3%)	44 (72.1%)	22.7%
Autosomal recessive	3 (3.1%)	8 (8.2%)	11 (18.0%)	5.7%
Unknown	4 (4.2%)	2 (2.0%)	6 (9.8%)	3.1%
Total	35 (36.5%)	26 (26.5%)	61	31.4%

DISCUSSION

This study is a detailed analysis of the characteristics of inflammatory bowel disease among CGD patients using the largest registry of CGD patients in North America. Approximately half of the total 199 patients in this USIDNET registry were found to have IBD, a finding aligned with aforementioned estimates of 30%–60% from multiple studies (3, 7–9, 26–30). Published reports indicate that many CGD-IBD patients are diagnosed with inflammatory bowel disease years prior to receiving a CGD diagnosis (7, 13, 26, 28–33). This observation highlights the need to exclude the diagnosis of CGD in any patient who presents with early-onset inflammatory bowel disease. In this study, chronic diarrhea and eosinophilic esophagitis were features only present in patients with IBD. Perirectal abscesses and GERD were also seen more frequently in the IBD group.

Within the last 15 years, mortality rates in CGD cohorts have ranged from 13%–20%, with patients surviving into later decades of life as treatments improve (3, 7, 24, 26, 30, 31). In the USIDNET CGD cohort, a total of 17 patients (8.8%) died; a lower rate compared to other cohorts. Importantly, there were no differences in mortality between CGD patients with IBD and those without IBD. Larger studies are needed to further investigate the effect of IBD on mortality among CGD patients given the associated comorbidities.

The USIDNET cohort of CGD patients with IBD was found to have more infections compared to those without IBD. This finding remained consistent when comparing transplant status or immunomodulator use. With the data available, we are unable to determine whether increased infections may correlate to preceding use of immunomodulatory drugs, or other factors such as oxidative burst activity. However, it is certainly possible that the increased use of immunomodulatory drugs, other than IFN-γ, in the CGD-IBD group could have contributed to increased infections. In agreement with other CGD cohorts, lower respiratory infections were the most common type of infection, and Aspergillus species were the most commonly isolated organisms (24, 27–31, 34–36). Most notably, infections with enteric organisms (e.g., Enterobacter, Enterococcus, Escherichia, and Klebsiella) were more common in patients with IBD than without, a novel finding within a published CGD cohort. It is possible that this finding would not be present if more infectious data were available. However, considering that the percentages of missing organisms for infectious events in IBD versus non-IBD patients were comparable (47.6% versus 49.7%), we conclude that this is a key finding that deserves further exploration.

The majority of CGD patients are prescribed antimicrobial and antifungal prophylaxis to prevent invasive bacterial and fungal infections (2, 37–39). In our North American cohort, trimethoprim-sulfamethoxazole and itraconazole were most commonly used. The immunomodulatory drug IFN-γ was utilized in 41% of the USIDNET CGD cohort. There was no significant difference in the frequency of use between the IBD and non-IBD groups (42% versus 41%, respectively). With the available data in this study, we were not able to discern whether IFN-γ modulates IBD disease positively or negatively, thus highlighting the need for further studies.

The treatment of patients with CGD and IBD is complicated by increased susceptibility to infections and the concurrent need for immunosuppression. This has been evident in multiple treatment regimens, especially those using anti-TNF medications, such as infliximab. Infliximab is commonly used in patients with Crohn disease. Although this medication has been shown to improve symptoms and sequelae in patients with CGD-IBD, it has also been noted to come with a significant increased risk for infections (17, 44). Many immunologists believe that anti-TNF medications should be contraindicated in CGD-IBD patients (13, 17, 44, 45), although there were seven (7.1%) anti-TNF treated CGD-IBD patients (7.1%) in this cohort. Data regarding the timing of the use of immunomodulators was not available, and thus, we were unable to determine a relationship for the indication for use in CGD or, in some patients, HSCT. TNF-α inhibitors aside, other immunomodulators have been used to treat CGD colitis with varying levels of success, and often in combination with corticosteroids. Vedolizumab, a gut-selective anti-integrin antibody approved for the treatment of typical Crohn disease and ulcerative colitis (40, 41), was used in 11 CGD-IBD patients (11.5%). Anakinra, an IL-1 receptor antagonist, has also been used in CGD-IBD patients and shown to be effective, however one study noted that the early efficacy of the medication faded with time (42, 43). This may explain why only 2.1% of CGD patients in this cohort were treated with anakinra. Additional clinical trials are needed to demonstrate efficacy and safety of various immunomodulators in CGD.

Patients with severe forms of CGD due to absent ROI production have a significantly increased risk of life-threatening infections (3). In these patients, HSCT is a clear option for cure. In patients with hypomorphic mutations and residual NADPH oxidase activity, the benefit of HSCT may be less clear. Our study highlights that CGD patients with IBD appear to have an even larger burden of disease, whether due to infections or due to the significant morbidity associated with IBD and immunomodulatory treatments. Thus, the decision for moving forward with HSCT must involve balancing the potential risks associated with HSCT with the potential of cure of CGD and remission of IBD.

Since our study is a cross-sectional analysis of a patient registry, it has several limitations. As with many registry-based studies, the entries span decades of medical care and may not be updated as patient trajectories, standards of care for CGD, and treatment outcomes change. Many immunomodulators currently available were not available in prior decades, as is the knowledge of potential adverse effects unique to CGD with certain drugs such as TNF-inhibitors. As patients are mostly recruited at large academic centers in the U.S. and the National Institutes of Health, this group may represent CGD patients with more severe or complicated disease, and results may not be generalizable to all patients with CGD. The description of infectious pathogens is limited to patients with available culture results, which may bias estimates of the prevalence of GI or other pathogens in both groups. Finally, because of a cross-sectional design, we could not establish temporal relationships between important variables such as medications, infections, or transplant. Thus, the infection rate and number of infectious events reported may not truly span the presumed length of care between the age of diagnosis of CGD and the most recently reported age in the database. Nevertheless, this study leverages the resources of the largest registry of CGD patients in North America to increase attention to the importance of IBD diagnosis and treatment in CGD, relevant clinical aspects of this subset of patients, and important questions for future research.

FUTURE DIRECTIONS

This analysis of disease presentation, treatment, and disease outcomes in the CGD-IBD and CGD-non-IBD subjects enrolled in the USIDNET revealed that infectious events and immunomodulator use were higher in patients with IBD. However, mortality was not increased. Larger studies aimed to determine optimal IBD treatment modalities are greatly needed.

In the literature, symptoms of CGD-IBD are highly variable and range from mild abdominal discomfort to severe diarrhea or hematochezia leading to hospitalization. Symptoms at the time of presentation can be subtle, including isolated diarrhea. Collectively, this evidence suggests significant phenotypic heterogeneity of CGD-IBD presentation, and even mild symptoms could indicate underlying chronic GI inflammatory disease. Given the limitations of a registry-based study, more research is needed to develop specific predictive algorithms for IBD in CGD patients. However, this study demonstrated that all CGD patients with chronic diarrhea were found to have IBD. Elicitation of this portion of the clinical history would be important with each patient visit and could be one component in development of practice guidelines on IBD screening in CGD in order to enhance the clinical care of this patient group.

Our findings reinforce the notion that IBD and CGD frequently coexist, point out some of the major phenotypic features of CGD patients with GI inflammatory conditions, highlighting the need for further investigations in this topic.

Highlights Box:

1. What is already known about this topic:

A significant proportion of patients with chronic granulomatous disease (CGD) have inflammatory bowel disease (IBD).

2. What does this article add to our knowledge?

CGD patients with IBD may have more infections as compared to those without IBD, notably with enteric organisms. More CGD patients with IBD require immunomodulator treatment than CGD patients without IBD.

3. How does this study impact current management guidelines?

Early recognition and diagnosis of IBD in CGD patients is crucial in order to decrease infectious complications.

Abbreviations:

AR

Autosomal recessive

AV

Arteriovenous

CGD

Chronic granulomatous disease

CMV

Cytomegalovirus

DMARD

Disease modifying antirheumatic drug

IEI

Inborn error of immunity

GI

Gastrointestinal

GERD

Gastroesophageal reflux disease

GVHD

Graft-versus-host-disease

HHV

Human herpesvirus

HSCT

Hematopoietic stem cell transplant

HSV

Herpes simplex virus

IBD

Inflammatory bowel disease

IFN-γ

Interferon-gamma

IL-1

Interleukin-1

IQR

Interquartile range

PIDD

Primary immunodeficiency disease

ROI

Reactive oxygen intermediates

RSV

Respiratory syncytial virus

TMP-SMX

Trimethoprim-sulfamethoxazole

TNFα

Tumor necrosis factor-alpha

USIDNET

The United States Immunodeficiency Network

6-MP

6-mercaptopurine

VRE

Vancomycin-resistant enterococcus

VZV

Varicella zoster virus