A Registry Study of 240 Patients with X-Linked Agammaglobulinemia Living in the USA

We conducted a detailed descriptive analysis of the USIDNET XLA cohort. We wanted to characterize the types of infections, co-morbid conditions, and sub-conditions within different organ/body systems. We reported means and standard deviations on continuous variables for highly skewed continuous variables, and we reported median and inter-quartile ranges. For categorical variables, we reported counts and proportions. We further examined the effects of IgG dose on survival, infection type, and complications using multivariable linear and generalized linear regression models. All analyses were conducted within R statistical software [13].

Data on 240 patients with XLA living in the USA were entered between the years 1981 and 2019 (Table 1). While the registry was created in 1999, data from prior visits based on medical records were included which dated back as early as 1981. Patients were born between 1945 and 2017 (median, 1991; range, 1945–2017). One hundred and forty-one patients (58.7%) were < 18 years of age.

Race was indicated in 204 patients as follows: White, 148 (72.5%); Black/African American, 23 (11.2%); Hispanic, 20 (9.8%); Asian or Pacific Islander, 6 (2.9%); and other or more than one race, 7 (3.4%).

Living status was reported for 178, of whom 158 (88.8%) were alive at the time of this analysis. One hundred and twelve (62%) patients reported a family history of immune deficiency among 180 with documented family history. Of those patients with a reported family history, a primary immune deficiency (PID) was reported by one other family member in 17.5%, 2 other family members in 10.4%, and more than 2 other family members in 7.9% of patients. BTK was the reported affected gene in 229 (95%) of the patients. However, the BTK pathogenic variant was only documented in 25 patients (Supplemental Table 12).

The median age at last entry was 15 years (range, 1–56). The median age of disease onset was 0.8 years (range, birth–22), with a median age at diagnosis of 2 years (range, birth–29), and the median length of time with diagnosis of XLA was 10 years (range, 1–56). The mean BMI for adult patients in this cohort was 25.6 (SD = 6.61) while in the USA, the average adult male has a BMI of 26.6 (https://​www.​cdc.​gov/​nchs/​data/​nhanes/​databriefs/​adultweight.​pdf).

The racial/ethnic breakdown reported by this USIDNET cohort (White, 72.5%) differs from what is currently being reported by the US Census Bureau (https://​www.​census.​gov/​quickfacts/​fact/​table/​US/​PST045221). Therefore, a subset analysis was performed to look at differences between White and non-White. For the limited data that was available, no difference was identified for age of death, age of disease onset, age at diagnosis, age at start of IgGR, family history, and survival (Supplemental Table 14).

Concerning treatment modalities, 221 (92%) patients were reported as receiving IgGR. For 138/221 patients, the mode of delivery was reported as intramuscular (IM) in 1 patient, intravenous (IV) in 92 patients, and subcutaneous (SC) in 45 patients. We were unable to determine the number of patients who may have been on more than one mode of IgG delivery (i.e., IV switched to SC or SC switched to IV and IM), due to the limitations of the dataset. Fifty-eight patients were reported as receiving antibiotic prophylaxis as shown in Supplemental Table 13. Two patients were reported as having received palivizumab. The IgG status was available for 49/58 patients as follows: 30 (61%) were on IVIG and 19 (39%) were on SCIG. The indication for prophylaxis and length of prophylaxis could not be extracted from the dataset. Nineteen patients (7.9%) were on treatment with immunomodulatory drugs (Supplemental Table 11). Eighty-six (35.9%) patients had undergone a total of 179 surgical procedures (data not shown), two had undergone hematopoietic cell transplantation, and one patient was reported as living with an unknown status for the other. Liver transplantation had been reported for 2 patients, both of whom were alive (Table 1).

In this cohort, 222 patients (92.5%) reported infections, with 1054 total infections reported in the Registry (Supplemental Table 2). Most patients (56.2%) reported pneumonia, while similar numbers reported sinusitis (54.6%) and otitis media (54.2%). Common infections included skin infections (29.2%), conjunctivitis (25%), and abscess (10%). Ten percent of patients reported bacteremia (7.5%) or sepsis (2.5%). Opportunistic infections were recorded in 8.3%, but details of these infections could not be clearly ascertained. Seven and a half percent of patients had meningitis, while bacterial arthritis and encephalitis were noted in 5.8% each. When comparing infections prior to diagnosis versus those after diagnosis, pneumonia was more commonly noted before diagnosis, followed by otitis media and sinusitis. However, after the diagnosis of XLA was established and despite IgGR, patients continue to experience episodes of pneumonia, but sinus and ear infections were more frequent (Fig. 1A). When evaluating post- vs pre-diagnosis differences in prevalence for each type of infection, only meningitis and bacteremia/sepsis had reduction after the diagnosis of XLA was made (Fig. 1B).

Medical conditions other than infections were reported in 193 patients (80.4%), with 1243 total conditions reported (supplemental Table 1). After diagnosis, 178 patients reported any medical condition: 89 patients (50%) reported four or more conditions, 40 (22%) reported 1 condition, 28 (15.7%) reported two conditions, and 21 (11.8%) reported 3 conditions (data not shown).

XLA patients were reported as having different organ systems involved (Table 2). The highest number of patients (51.3%) experienced respiratory illnesses, followed by gastrointestinal (40.0%), neurological (35.4%), and musculoskeletal (28.3%) problems (Table 2). Other conditions reported included hematologic (19.2%), dermatologic (18.3%), endocrine (10.8%), cardiovascular (7.9%), psychiatric (7.5%), and renal involvement (7.1%). (Supplemental Tables 3–10). When comparing organ system involvement before and after diagnosis of XLA, respiratory and gastrointestinal manifestations continued to be reported, but both neurological and musculoskeletal complications appeared to be more frequently noted after the diagnosis of XLA was made and despite therapies (Fig. 2A). When post- vs. pre-diagnosis differences were investigated, the three body systems with the largest increases after diagnosis were gastrointestinal, respiratory, and neurological (Fig. 2B). Among patients with respiratory disease, lower respiratory disease, asthma, bronchitis, and COPD were most frequently reported. In this cohort, twenty-two patients reported bronchiectasis (9%) (Supplemental Table 3). Among gastrointestinal manifestations, diarrhea, abdominal pain, and GERD were frequent. Inflammatory bowel disease, including Crohn’s disease, colitis, and ulcerative colitis, was seen in 14 patients (14/240; 5.8%; Supplemental Table 4). Among patients with neurologic complications, more than half had reported cognitive disabilities, speech delay, mental retardation, and difficulties with educational academic activities. Thirteen of these patients were affected by seizures/epilepsy. Among this cohort, 5 were reported as having had paralytic polio, all born prior to 1991 (Supplemental Table 5). Most patients with musculoskeletal symptoms described arthritis, arthralgias, or generalized aches and pains. Five patients are reported as having dermatomyositis. (Supplemental Table 6). Interestingly, of the five patients reported as having dermatomyositis, two had died and had experienced encephalitis and encephalopathy respectively; another patient had reported “enterovirus” among the infectious history. Sinusitis and no infections were documented for the other two patients. For those patients with hematologic organ involvement (46/240), all but one had some form of cytopenia; in 6 patients, more than one cytopenia was reported, and 1 patient had a B cell lymphoma (Supplemental Table 7). The most common endocrine abnormality was short stature with or without growth hormone disorder (Supplemental Table 8). Psychiatric manifestations included attention deficit disorder with hyperactivity followed by depression, behavioral problems, and anxiety (Supplemental Table 9). Almost 8% of patients reported a cardiac condition (Supplemental Table 10). Less commonly affected organs were the kidneys and liver (data not shown).

Symptoms of allergic disease such as chronic rhinitis, eczema/atopic dermatitis, and urticaria were reported in 28% of patients. Food allergies were reported in 7 patients and 44 patients reported drug allergies. Anaphylaxis was reported in 4 patients (2 to shellfish, 1 to bee venom, and 1 to sulfamethoxazole-trimethoprim) (data not shown).

Rare conditions included 2 patients with bronchiolitis obliterans with organizing pneumonia (BOOP), 4 with appendicitis, and 1 patient with neurofibromatosis. Among this cohort, 3 (1.25%) patients reported malignancy (B cell lymphoma, angiosarcoma, and craniopharyngioma).

In this cohort, 58 patients (24%) received antibiotic prophylaxis, and 44 of them were reported as being on continuous treatment. Regression analysis showed that patients on continuous prophylactic antibiotics were more likely to have an increased number of conditions (rate ratio (RR), 1.17; 95% CI, [1.10, 1.24]; p < 0.001). We did not detect a significant correlation between infection and continuous prophylactic antibiotic use (OR, 1.2; 95% CI, [005, 11.0], p = 0.9; data not shown). Most (92%) patients reported receiving gamma globulin treatment at some point.

Headaches were the most common IgG therapy reactions reported in 220 patients. Serious reactions included 2 patients with anaphylaxis, while thromboembolism, renal dysfunction, and aseptic meningitis were reported in 1 patient for each condition. In this cohort of patients, after controlling for age at the time of first IgG treatment, no significant association between IgG dose (mg/kg) and IgG level (mg/dl) and years of survival was observed since initial treatment (data not shown). IgG replacement therapy had a significant positive impact on survival (years lived since diagnosis). Patients in whom a response to “ever receiving IgGR” after diagnosis was either “no” or “not documented” (N = 11) had a median survival of 16 years, compared to 51 years for those who ever received IgG therapy (N = 180). Cox regression showed that IgG therapy was associated with 89% reduction in risk of death (hazard ratio = 0.11, p < 0.001; Fig. 3C).

The use of immunomodulators was reported in 8% of patients. Of note, three of the patients were on more than one immunomodulator (Supplemental Table 11).

Four patients had received an organ transplant: two underwent HCT and two had a liver transplant. One of the two patients treated with HCT reported it as treatment for the primary immunodeficiency and was alive. The living status of the other patient who had undergone HCT was not reported. One liver transplant patient reported indication for end-stage cirrhosis and the other for portal hypertension; both were reported as living.

Of 178 patients with reported living status, 158 (88.8%) were living and 20 (11.2%) were deceased. For 191 patients in whom sufficient data was available (age at diagnosis, living/death status, and most recent office visit date), Kaplan–Meier survival analysis showed that 75% of patients had survived to 25 years beyond diagnosis (Fig. 3A). There was no significant difference in survival between those who started IgG before or after 2 years of age, which corresponds to the median age at diagnosis among this cohort (Fig. 3B).

The reported median age of death was 21 years (range, 3–56.7 years). On average, these patients were diagnosed at a median age of 2.7 years of age (range, 0.8–8.7 years) and had lived to a median age of 17.1 years (range, 2–50.7) since diagnosis. In 14 of these 20 (70%) families, a history of XLA was known, whereas IgGR status was not known for 6/20 deceased patients. The most common underlying co-morbid conditions in these deceased patients were neurologic (75%), followed by respiratory (70%) (Table 3). Causes of death were primarily infections. Most of the deaths reported occurred before 1993, the year the BTK gene was discovered. Since the year 2000, only one death is reported among this USIDNET cohort.

The Lansky and Karnofsky scales of daily living were collected for 46 and 30 patients respectively. The mean Karnofsky score was 90.6 ± 10.99 and the mean Lansky score was 92.38 ± 21.3.