Late diagnosis and advances in genetics of chronic granulomatous disease

Chronic granulomatous disease (CGD) is a rare primary immunodeficiency of innate immunity, leading to increased susceptibility to bacterial and fungal infections and to dysregulated inflammatory responses. The disease is caused by defects in the genes encoding any of the components of the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase complex, which is essential for the intracellular killing of pathogens by phagocytes [1]. Two recent articles published in Clinical and Experimental Immunology offer important contributions regarding several aspects of the disease [2, 8].

Barkai et al. [2] have reported on the clinical, genetic and functional characterization of 16 Israeli patients diagnosed with CGD in adulthood [https://doi.org/10.1111/cei.13474]. In their analysis, X‐CGD accounted for a minority (31.2%) of the patients. The authors reported two categories of adult patients: those who had a milder clinical phenotype and those who had a classic phenotype with severe infectious and inflammatory complications reflecting a profoundly impaired neutrophil function. Although CGD is usually diagnosed in childhood, patients may also be diagnosed later in life as adults. This delay in diagnosis is attributed to the rarity of the disease, and hence its low awareness among physicians. The authors reported that the highest common factor leading to a correct CGD diagnosis was recent diagnosis of a family member, followed by suspicion on the part of an expert other than an immunologist, diagnosis by an immunologist and incidental findings through whole exome sequencing (WES) during a search for metabolic disease. The different misdiagnoses of CGD identified were, mainly tuberculosis, followed by sarcoidosis and inflammatory bowel disease.

A number of recent reports have provided further information on adult patients with CGD [3, 4]. Overall, onset in adulthood did not predict less severe disease. Indeed, despite appropriate prophylaxis, infection rates in adults were similar to those reported in children and they were the first cause of death. Inflammatory complications, especially pulmonary and digestive, seem to predominate during adulthood and have a profound impact on CGD‐associated morbidity and treatment. These conditions are especially difficult to manage because immunosuppressive treatments are challenging in patients with a severe immunodeficiency. Moreover, surviving patients often experience growth failure, severe organ dysfunction, chronic dyspnea, non‐infectious liver disease and autoimmune manifestations. Adults with CGD showed decreased wellbeing when compared with the general population, in both the physical and mental dimensions, possibly due to incoming co‐morbidities over time. Furthermore, despite modern management, there is an ongoing mortality associated with the disease. The study by Barkai et al. stresses the need to increase awareness of this congenital immunodeficiency among clinicians of different specialties who might be treating undiagnosed adult patients with CGD; it is likely that CGD in adults is more common than currently recognized.

Lastly, the authors discuss the role of HSCT (hematopoietic stem cell transplant) as a curative modality for adult patients with CGD. HSCT remains the only curative treatment for CGD. However, despite the fact HSCT has improved greatly in recent years, it is not performed for all patients. There is currently wide variability in published indications for transplant in patients with CGD.

Given apparently better outcomes with HSCT over those with conventional therapy in survival and quality‐of‐life measures, some authors have recommended a transplant for all patients with absent NADPH‐oxidase enzyme activity if a matched donor can be identified [5]. The current most problematic decision, especially in adults with CGD, relates to whether and when to treat with HSCT. In adults, criteria are more difficult to apply due to higher rates of organ dysfunction and reported increased mortality from HSCT. Promising results in recent years [5, 6, 7] suggest that HSCT is safe and effective when delivered in a specialist center and should be carefully considered as a potentially curative option for younger adult CGD patients with an appropriate donor.

In the second of these recent articles in the journal on this topic, Mollin et al. [8] described 16 X‐CGD patients from a clinical, functional and genetic view point [https://doi.org/10.1111/cei.13520]. Through a detailed characterization, the authors provide an overview of the main features related to X‐CGD and describe many aspects regarding the disease phenotype associated with variable NADPH oxidase 2 (NOX2) expression and NADPH‐oxidase activity due to different genetic variants.

In this cohort study, the clinical history and the inability of phagocytes to generate ROS (reactive oxygens species) were followed by the molecular analysis of the CYBB gene by cDNA or genomic DNA direct sequencing. In particular, the authors reported two interesting rare mutations, the c.‐67delT in the promoter region and the deep intronic c.253‐1879A>G, which activates a cryptic acceptor splice site, causing the insertion of 124 intronic nucleotides from intron 3 into the mRNA. This type of mutation in CGD patients had already been reported by other authors [9, 10, 11, 12]. Both patients were described as an X91⁻ CGD phenotype, which is typically variable depending on the localization of the mutations in CYBB. The authors suggested that the mutation in the promoter region, which leads to a residual NOX2 expression in eosinophils, protected the patient against infections. In the other case, the deep intronic mutation results in a faint but significant NADPH oxidase activity due to a residual normal mRNA expression that probably determined a mild phenotype in this late‐diagnosed patient. Moreover, in the rare case of a patient with two missense variants in the CYBB gene, the authors showed that each mutation is pathogenic and that these mutations differently impair NOX2 expression, although both compromise NADPH‐oxidase activity. Finally, a new three‐dimensional model of the dehydrogenase domain of NOX2 was applied in the study of missense mutation effects and their impact on NOX2 expression and NADPH‐oxidase enzyme activity.

This work shows how many aspects need to be investigated to establish the actual pathogenicity of a genetic variant. It is generally true that pathogenicity should be determined by the entire body of evidence as a whole. CGD constitutes a good model to study all possible consequences of a genetic variant from molecular to functional defects and phenotype and, even in the absence of an evident molecular defect, functional studies should be performed. Nevertheless, a clear genotype–phenotype correlation has not always been observed in CGD, and further studies will be required to account for this discrepancy.

Although many advances have been made in the field of molecular diagnostics, it is not surprising that 50–75% of patients do not receive a genetic diagnosis after WES. Indeed, there are many under‐estimated gene variants in known and unknown genes that complicate the craft of the molecular geneticist. High‐throughput sequencing, in particular tNGS (targeted NGS) or WES, fail to recognize the effect of mutations in deep intronic or elusive splicing defects, in enhancer and promoter regions or in regions with duplicated sequences or pseudogenes [13, 14, 15]. These types of variant have only been identified through more in‐depth or specific genetic tests, and are often weighed in the second step of data analysis. Only a careful analysis and a high causal relationship between the gene variant and the clinical phenotype can warn the clinical geneticist of the possible effect of these variants.

Functional validation should be used, when possible, to establish the pathogenicity of all potentially causal variants and clarify their relationships in other PID (primary immunodeficiency diseases) as well. Unfortunately, these essential studies generally remain unavailable or impractical for the evaluation of most variants of uncertain significance (VUSs) in genes that are less obviously related to the phenotype, despite the invaluable help of many bioinformatics tools.