Primary immunodeficiency diseases treated with immunoglobulin and associated comorbidities

Abstract

Background:

Primary immunodeficiency diseases (PIDD) consist of a heterogeneous group of disorders characterized by various aspects of immune dysregulation. Although the most universally recognized manifestation of PIDD is an increased susceptibility to infections, there is a growing body of evidence that patients with PIDD often have a higher incidence of lung disease, autoimmunity, autoinflammatory disorders, and malignancy.

Objective:

The purpose of this study was to better understand the noninfectious complications of PIDD by determining the comorbid disease prevalence across various age groups, genders, and immunoglobulin replacement types compared with the general population.

Methods:

A large U.S. insurance claims database was retrospectively analyzed for patients who had a diagnosis of PIDD and who had received intravenous immunoglobulin (IVIG) or subcutaneous immunoglobulin (SCIG). The prevalences of 31 different comorbid conditions in the Elixhauser comorbidity index were compared among the 3125 patients in the PIDD population to > 37 million controls separated by gender and by 10-year age cohorts.

Results:

In the PIDD population, statistically significantly higher comorbid diagnoses included chronic obstructive pulmonary disease–asthma in 51.5%, rheumatoid disease in 14%, deficiency anemia in 11.8%, hypothyroidism in 21.2%, lymphoma in 16.7%, neurologic disorders in 9.7%, arrhythmias in 19.9%, electrolyte disorders in 23.6%, coagulopathies in 16.9%, and weight loss in 8.4%.

Conclusion:

PIDD that require immunoglobulin replacement are associated with an increased risk of numerous comorbid conditions that affect morbidity and mortality. Recognition and increased awareness of these noninfectious complications can allow for better monitoring, care coordination, targeted treatments, and improved prognosis.

Predisposition to infections is a hallmark aspect of primary immunodeficiency diseases (PIDD), although autoimmunity, malignancy, and other features of immune dysregulation are being increasingly recognized as associated conditions. The presentation of PIDD is highly variable because different components of the immune system and nearly every organ system can be affected.^1–7 Common treatment for PIDD that involves humoral immunodeficiency consists of immunoglobulin replacement. This study included patients with PIDD who were on intravenous immunoglobulin (IVIG) or subcutaneous immunoglobulin (SCIG), which primarily captured individuals with PIDD who had humoral defects. Immunoglobulin replacement and improved antimicrobials have reduced infectious complications of PIDD; however, recognition and treatment of noninfectious complications is lacking.

Noninfectious complications likely vary, depending on the specific immunodeficiency, and have been found in 68–74% of patients with common variable immunodeficiency (CVID).^8,9 These noninfectious complications may be evident at presentation or may appear after the diagnosis.^8,10,11 Most previous studies compare the prevalence of comorbid conditions in PIDD versus controls but do not compare comorbid disease prevalence within age groups. This study more broadly assessed noninfectious complications of PIDD by comparing the prevalence of diseases in a predefined comorbidity index with the general population.

Unlike previous studies, we further stratified the prevalence of comorbid conditions in those who received IVIG versus SCIG because these patients may have had inherent differences in the underlying disease process. Categorizing the full spectrum of noninfectious complications of PIDD is still lacking in the field. This study aimed to explore comorbidities associated with PIDD within a large U.S. insurance claims database. We intended to better inform providers on the full spectrum of risk associated with PIDD to better coordinate care, provide targeted treatment, and improve overall prognosis.

METHODS

This study retrospectively analyzed insurance claims data from the PharMetrics Plus insurance claims database (IQVIA Real-World Data Adjudicated Claims, US) for PIDD diagnoses based on the International Classification of Diseases, Ninth Revision (ICD-9 codes: 279.00, 279.01, 279.02, 279.03, 279.04, 279.05, 279.06, 279.09, 279.12, 279.2) and the International Classification of Diseases, Tenth Revision (ICD-10 codes: D80.0, D80.1, D80.2, D80.3, D80.4, D80.5, D80.7, D81.0, D81.1, D81.2, D81.6, D81.7, D81.89, D81.9, D82.0, D83.0, D83.2, D83.8, D83.9) codes from 2012–2017. The database includes administrative health insurance claims of ∼200 insurance providers across the United States. The population included patients with two or more claims for PIDD ≥ 90 days apart and who received IVIG or SCIG. The IVIG cohort included patients with a claim for any of the top three prescribed IVIG therapies (Gamunex (Grifols, US), Gammagard (Baxalta, US), and Privigen (CSL Behring, US)). The SCIG cohort included patients with a claim for the main SCIG therapy (Hizentra (CSL Behring, US)). Patients were not included if they did not fulfill both criteria or if they switched therapy within 1 year after identification. The control population consisted of a 1-year (2016) sample of all PharMetrics patients with no claims of PIDD or use of IVIG/SCIG therapies. Additional exclusions included a lack of information with regard to age and/or gender.

Demographic and clinical characteristics, including age, gender, and the presence of Elixhauser comorbidity conditions, were collected. Elixhauser index categorizes comorbidities of patients based on International Classification of Diseases diagnosis codes and consists of 31 disease processes (congestive heart failure, arrhythmia, valvular disease, pulmonary circulation disorder, peripheral vascular disease, hypertension uncomplicated, hypertension complicated, paralysis, other neurologic disorders, COPD, diabetes uncomplicated, diabetes complicated, hypothyroid, renal failure, liver disease, peptic ulcer disease excluding bleeding, AIDS/HIV, lymphoma, metastatic cancer, solid tumor without metastasis, rheumatoid arthritis/collagen vascular disease, coagulopathy, obesity, weight loss, fluid and electrolyte disorders, blood loss anemia, deficiency anemia, alcohol abuse, drug abuse, psychosis, depression). The patients were subdivided into 10-year age cohorts to account for variability in disease prevalence across lifespans. The patients with PIDD were also subdivided into IVIG and SCIG groups and were compared with controls of similar age and/or gender. Analysis used χ² tests to highlight the differences in populations. Statistical software included SAS 9.3 (SAS Institute Inc., Cary, NC) and Prism (Proteogenomics Research Institute for Systems Medicine, San Diego, CA).

RESULTS

A total of 3125 patients (1297 male and 1828 female patients) met inclusion criteria with claims of PIDD and the use of IVIG (76%) or SCIG (24%). Similar to other large claims databases, female patients had a higher prevalence of PIDD overall.¹² The patients were subdivided into age groups (Table 1). The control population with no claims of PIDD or use of IVIG or SCIG consisted of >37 million patients, divided into gender and age cohorts, as seen in Table 2.

Table 1.

Primary immunodeficiency diseases population characteristics

IVIG = Intravenous immunoglobulin; SCIG = subcutaneous immunoglobulin.

Table 2.

Control population

Chronic obstructive pulmonary disease (COPD)–asthma was the most commonly observed comorbid condition in all age groups in both genders with PIDD on immunoglobulin therapy whether on IVIG or SCIG. Nearly every age and gender cohort had a statistically significant (p < 0.001) increased prevalence of COPD-asthma compared with controls (Fig. 1A; Supplemental Table S1). Of the patients with PIDD, 51.5% had COPD-asthma compared with 8.1% of COPD-asthma in the general population (Fig. 2A). This is further broken down into COPD-asthma in male patients on IVIG, at 40.9%; female patients on IVIG, at 56.2%; male patients on SCIG, at 51.7%; and female patients on SCIG, at 60.9%, compared with 7.7% of male controls and 8.5% of female controls (Fig. 2B).

Figure 1.

Primary immunodeficiency diseases (PIDD) comorbidities across various age groups in male versus female patients on intravenous immunoglobulin (IVIG) versus subcutaneous immunoglobulin (SCIG). (A) Chronic obstructive pulmonary disease (COPD)–asthma; (B) rheumatoid disease; (C) anemia (deficiency); (D) hypothyroidism; (E) lymphoma.

Figure 3.

Other disease associations with primary immunodeficiency diseases (PIDD) across various age groups in male versus female patients on intravenous immunoglobulin (IVIG) versus subcutaneous immunoglobulin (SCIG). (A) Neurologic disorders; (B) arrhythmias; (C) electrolyte disorders; (D) coagulopathies; (E) weight loss.

Rheumatoid arthritis and/or collagen vascular disease was found to be 7.8 times higher in the PIDD population (Fig. 2A). Also, rheumatoid disease was more common in female patients, with 18.7% of female patients on IVIG and 17.9% of female patients on SCIG compared with 7.9% of male patients on IVIG and 7.2% of male patients on SCIG with coexisting rheumatoid disease (Fig. 2B). Nearly every age and gender strata in the larger IVIG group had a statistically significant increased prevalence of rheumatoid disease compared with the controls; however, in the SCIG population, only the female patients between ages 31 and 70 years had statistically significant differences (Supplemental Table S1). Similar to the general population, many rheumatologic processes did not present until later years and predominately in female patients.

Other potential autoimmune categories within the Elixhauser index included anemia and hypothyroidism. Of the patients with PIDD, 11.8% had anemia compared with only 1.7% of the general population (Fig. 2A). Hypothyroidism was present in 21.2% of the patients with PIDD and in 5.7% of the patients in the control population (Fig. 2A). Almost all male and female patients in every age group who received IVIG had a statistically significant increased prevalence of anemia and hypothyroidism, but these trends were only statistically significant in a few strata in the smaller SCIG population (Supplemental Table S1).

Lymphoma appeared to be present in 16.7% of the PIDD population compared with 0.2% of the control population (Fig. 2A). The rates of lymphoma were higher in those on IVIG, with 28.2% of the male and 15.3% of the female patients, compared with those on SCIG with 4.9% of the male and 2.3% of the female patients (Fig. 2B). Almost all gender and age strata in the IVIG group had a statistically significant increased prevalence of lymphoma compared with the controls (Supplemental Table S1). The prevalence of lymphoma especially increased in the older age groups for both genders and both types of immunoglobulin treatment (Fig. 1E). Solid tumors and metastatic cancer did not show statistically significant differences until patients were > 40 years old, at which point solid tumors were two to three times higher and metastatic cancers were four to seven times higher in the IVIG group when compared with the general population (Supplemental Table S1). When analyzing the differences of lymphoma, solid tumors, and metastatic cancer between the SCIG and control populations, few statistically significant differences were seen (Supplemental Table S1).

Nonspecific neurologic disorders were found in 9.7% of the patients with PIDD compared with 1.6% of the general population (Fig. 2A). The prevalence was similar across gender and type of immunoglobulin therapy but was significantly higher in younger patients, with 15.5-25% in 0–10 year olds. Statistically significant changes started to disappear after the age of 30 years (Fig. 3A; Supplemental Table S2). Within the PIDD population, 19.9% of the patients had arrhythmia compared with 3.5% of the general population (Fig. 2A). The prevalence of arrhythmias was somewhat higher in those who received IVIG, with 24.4% of the male and 19.9% of the female patients, compared with those who received SCIG, with 13.7% of the male and 13.9% of the female patients (Fig. 2B). The prevalence overall increased with increasing age for all groups (Fig. 3B; Supplemental Table S2).

Figure 2.

(A) The prevalence of comorbidities in patients with primary immunodeficiency diseases (PIDD) on immunoglobulin. (B) The prevalence of comorbidities in patients with PIDD on immunoglobulin stratified into male versus female patients and intravenous immunoglobulin (IVIG) versus subcutaneous immunoglobulin (SCIG).

Electrolyte and/or fluid disorders were found in 23.6% of the patients with PIDD compared with 2.3% of the general population (Fig. 2A). The prevalence was approximately double for those who received IVIG versus SCIG. Nearly every gender and age strata in the IVIG population had a statistically significant change compared with the controls. In the SCIG population, the most consistent statistically significant changes were in the younger age groups (Fig. 3C; Supplemental Table S2). In the general population, coagulopathies were found in 0.8% versus 16.9% of the patients with PIDD (Fig. 2A). Higher and statistically significant changes were found in all gender and age strata in the IVIG group (Figs. 2B and 3D; Supplemental Table S2). Weight loss was more prevalent in PIDD, which consisted of 8.4% versus 0.9% of the general population (Fig. 2B). In the IVIG group, all gender and age strata had a statistically significant increased prevalence of weight loss compared with the controls (Fig. 3E; Supplemental Table S2).

DISCUSSION

This large epidemiologic study corroborated with the literature with regard to the presence and degree of lung disease associated with PIDD. Similar to other studies, respiratory disease is the most common organ-specific complication.⁸ Other large data registries found asthma to be the most common respiratory complication in patients with CVID.^9,13 The high prevalence of lung disease likely, in part, is related to the recurrent sinopulmonary infections that occur in many patients with humoral immunodeficiency, which resulted in impaired pulmonary function and structural damage.⁶ Early in the infectious-inflammatory process, reversible changes may be recognized clinically as asthma.

Persistent airway inflammation can lead to airway remodeling and irreversible obstructive lung diseases, such as COPD and bronchiectasis.^2,13 It can be difficult to determine the degree of impairment due to infection with colonization of the airways that leads to bronchiectasis versus inflammation from granulomatous and interstitial lung diseases.¹⁴ Most previous studies focus on adult patients but this study also highlighted the high prevalence of associated lung disease in children. Providers should closely monitor and manage these respiratory complications, expected to occur in the majority of patients with PIDD.

Autoimmune diseases are linked to certain subtypes of PIDD given the underlying basis of immune dysregulation and generally are known to be more frequent in female patients of increasing age.^10,15 This pattern is clearly seen in the PIDD population, with higher and more statistically significant rates of rheumatoid processes in female patients of middle-to-older age. This study showed similar results as the European Society for Immunodeficiency registry's cohort of 2700 patients with CVID and with the prevalence of autoimmunity, which exceeded that in the general population by a factor of 7.6.¹⁶ The literature showed autoimmune disease, including cytopenias, inflammatory arthritis, systemic lupus erythematous, and other autoimmune disorders, in ∼20–30% of patients with PIDD.^8,10,15

A diagnosis of autoimmune cytopenia precedes a diagnosis of PIDD in 60% of the patients.¹⁰ Therefore, the sevenfold increase in anemia observed in our PIDD population may suggest the potential presence of autoimmune anemia in the patients with PIDD. In addition, hypothyroidism was found to affect just over 20% of patients in our PIDD population, with some of these cases likely representing autoimmune processes. Previous studies show hypothyroidism as the most prevalent organ specific autoimmunity in immunodeficiency.^10,16

PIDDs are associated with a higher prevalence of malignancy, especially lymphoid tumors. Previous literature found lymphoma to occur in 3.4–3.8% of patients with PIDD,^6,16 which is consistent with prevalence of lymphoma seen in our SCIG population. The SCIG population is likely a better representation of patients with true primary immunodeficiency because SCIG is only approved on-label for PIDD (with only one exception: chronic inflammatory demyelinating polyneuropathy). IVIG is more broadly used by various subspecialists and is indicated for several other diseases, and may also be used for secondary immunodeficiencies due to malignancy or medication-induced immunosuppression.^17,18 The central nervous system manifestations of PIDD are not well known. The study suggests that infection and inflammatory neurologic disorders are the cause of most neurologic presentations in these patients.¹⁹ Some types of PIDD have core neurologic features, e.g., ataxia-telangiectasia. Neurologic symptoms, depending on the specific immunodeficiency, may vary from mild cognitive defects to severe disabilities.^20–22

Arrhythmias observed in the PIDD population may be related to infectious, autoimmune, and/or genetic mechanisms that disrupt cardiac function.²³ However, humoral immunodeficiencies are not classically associated with cardiovascular disease. There are syndromes such as 22q11.2 deletion and Jacobsen syndrome in which immunodeficiency and congenital heart disease coexist.^24,25 Otherwise, there is a lack of research that explores associations among cardiovascular diseases, e.g., arrhythmias with PIDD. This study found nearly 20% of the PIDD population to have arrhythmia, an association not previously described in the literature aside from a recent small study of 30 patients with CVID found to have a higher risk of developing arrhythmias based on electrocardiogram changes.²⁶

The Elixhauser index does not specifically include gastrointestinal inflammatory disease, a comorbidity well supported in the literature.^5,8,27 The Elixhauser index, however, does include electrolyte and/or fluid disorders and an increased prevalence in the PIDD population may be related to sequela of gastrointestinal involvement, such as diarrhea and/or malabsorption.¹⁴ Electrolyte and/or fluid disorders may also be related to kidney involvement, which is described only in isolated and/or sporadic case reports.²⁸ Infectious and inflammatory processes can result in fluid and/or electrolyte shifts, which possibly explained the increased prevalence in our cohort. Inflammation can also lead to altered coagulation with imbalance between pro- and anticoagulant states.^29,30

A higher prevalence of weight loss in PIDD is not unexpected because failure to thrive and weight loss are presenting features and warning signs of PIDD. The United States Immunodeficiency Network data demonstrated significantly higher prevalence of both pediatric and adult patients with PIDD and with underweight status.³¹ Weight loss in PIDD is often multifactorial and can be related to inflammatory or infectious gastrointestinal disorders.³¹ Similar to other chronic diseases, depression rates were found to be higher in PIDD. Results of studies have shown higher rates of mental health disease in PIDD.^32–35 PIDD can have a strong negative effect on quality of life (QOL).^32–36 Some major risk factors that lead to poor QOL include being female, older, and affected by chronic lung disease and chronic diarrhea.³⁴ The number and burden of comorbidities strongly associates with QOL.³⁶ Although immunoglobulin helps prolong survival and reduce morbidity, various other factors contribute to depression and lower QOL in PIDD.³⁴

Limitations

This claims database provided a cohort larger than previous comorbidity studies; however, insurance claims based on coding may not accurately reflect the clinical picture. For example, the Elixhauser index used for claims data does not separate COPD from asthma. Also some patients may not fulfill a clinical diagnosis of PIDD but may have been inaccurately coded to obtain immunoglobulin coverage for other secondary causes. To address some of these limitations, the population was further separated into IVIG versus SCIG groups because SCIG is more exclusively used by immunologists and likely a better representation of PIDD.^17,18 It is possible that immunoglobulin itself may contribute to an increased prevalence of some of the aforementioned disease processes, such as neurologic symptoms with headaches, electrolyte disorders, renal impairment, hemolytic anemia, and thrombotic events.³⁷

CONCLUSION

Future Directions

Health care providers need to recognize not only the associated infections but also the noninfectious characteristics of PIDD. Sometimes, other conditions can appear before infections and, in certain patients with PIDD, recurring infections may never occur. Therefore, patients may first come to the attention of other providers. A better understanding of noninfectious complications of PIDD can help providers more easily recognize potential comorbidities, refer to an immunology specialist, and decrease the delay in diagnosis and treatment. Immunologists can monitor for these comorbidities more closely and appropriately refer to other specialists. A multidisciplinary team approach is necessary for managing these patients with such heterogeneous immune defects and array of comorbidities.

Survival has improved in PIDD due to immunoglobulin and improved antimicrobials. There is minimal evidence with regard to treatments for the noninfectious comorbidities in PIDD.³⁸ Patients with PIDD continue to have reduced survival compared with age-matched controls.¹⁰ Those with at least one noninfectious complication have significantly higher mortality, estimated 11-fold risk, compared with patients with infections only.^8,10,11 Standard screening guidelines to best capture complications of this high-risk population do not currently exist in the United States. The British Society for Immunology developed 22 recommendations based on 57 consensus statements on managing noninfectious complications of CVID that recommended regular weight and lymph node assessments, monitoring of blood tests, and pulmonary function tests.³⁸ Future research is needed to develop and implement guidelines on screening of these noninfectious complications of PIDD. By anticipating and identifying patients expected to have a worse prognosis, providers can improve outcomes for patients with PIDD.