Functional Neutrophil Disorders: Chronic Granulomatous Disease and Beyond

Summary

Since their description by Metchnikoff in 1905, phagocytes have been increasingly recognized to be the entities that traffic to sites of infection and inflammation, engulf and kill infecting organisms, and clear out apoptotic debris all while making antigens available and accessible to the lymphoid organs for future use. Therefore, phagocytes provide the gateway and the first check on host protection and immune response. Disorders in killing and chemotaxis lead not only to infection susceptibility but also to autoimmunity. We aim to describe chronic granulomatous disease and the leukocyte adhesion deficiencies as well as myeloperoxidase deficiency and G6PD deficiency as paradigms of critical pathways.

Chronic Granulomatous Disease

Since it was first reported in 1954 in children with hypergammaglobulinemia and infections, Chronic Granulomatous Disease (CGD), has been characterized by both infection susceptibility and dysregulation of inflammation^1–3 Genetically linked to 6 genotypes with similar a phenotype, the defect is in phagocyte production of reactive oxygen species (ROS) or superoxide, which is associated with ineffective microbial killing. This defect in microbial killing leads to specific bacterial and fungal infection susceptibilities. In the US, CGD occurs approximately in 1 in 200,000 live births⁴, but the numbers are likely higher. The inheritance can be either X-linked or autosomal recessive (AR) with increased rates of AR inheritance in areas of the world in which there are higher rates of consanguinity. In the US, the X-linked form accounts for approximately 65% of all cases.

ROS and CGD

CGD is due to mutations in any of the proteins that make up the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase complex itself or lead to its assembly. Impairment of the NADPH complex leads to an inability to transfer an electron from NADPH to molecular oxygen to form superoxide. This results in impaired killing in neutrophils, macrophages and monocytes. The NADPH complex per se is made up of six proteins. gp91^phox and p22^phox are lodged in the membrane of the secondary granules and together form the heterodimer cytochrome b₅₅₈; the cytoplasmic tail of gp91^phox (encoded by CYBB) binds flavin adenine dinucleotide (FAD), NADPH (on the C-terminus end of gp91^phox) and heme (on the N-terminus) permitting the transfer of an electron to molecular oxygen; p22^phox (encoded by CYBA) is bound to gp91^phox and is required for the expression of the heterodimer. Upon phagocyte activation, three cytosolic proteins p47^phox (encoded by NCF1), p67^phox (encoded by NCF2), and p40^phox (encoded by NCF4) are phosphorylated, aggregate and translocate to cytochrome b₅₅₈ along with RAC1/RAC2 at the membrane surface. The critical interaction is the binding of the p47^phox/p67^phox complex to p22^phox5. RAC1/RAC2 are GTPases which are bound to the complex, with RAC2 being the predominant phagocyte form. Electron transfer from NADPH to molecular oxygen results in a superoxide anion (O₂⁻), which is converted to hydrogen peroxide by superoxide dismutase (SOD). Hydrogen peroxide is then converted to HOCl (hypohalous acid; bleach in neutrophils because the halide is chlorine) by myeloperoxidase (MPO). The intravacuolar generation of superoxide changes the intravacuolar electrical charge, eliciting a K⁺ flux that enhances intracellular killing through proteases which had been delivered by the primary granule⁶ (Figure 1.). Another genetic cause of CGD is due to a deficiency of the chaperone protein required for the transport of the gp91^phox/p22^phox complex from the endoplasmic reticulum into a plasma membrane, named essential for reactive oxidant signaling (EROS, encoded by CYBC1)⁷. Neutrophil extracellular traps (NETs) appear to be dependent on ROS, therefore are also defective in CGD, which may also affect microbial killing^8,9. NETs are released decondensed nuclear chromatin coated with histones and proteases that require MPO for protease activation after nuclear elastases promote decondensation of chromatin. Defects of microbial killing via NETosis in CGD cells have been reversed with the addition of tamoxifen, which induced NET formation in a ROS independent manner and neutrophils treated with tamoxifen showed significant improvement in microbial killing⁹. NET formation was restored in a patient who underwent gene therapy for CGD¹⁰.

Figure 1.

Assembly of the NADPH complex leads to both the production of HOCl and the influx of potassium as a result of the net negative intraphagosomic charge resulting in activation of proteases. This leads to killing in intracellular organisms to which the patients with CGD are susceptible.

Genetics of CGD

The most common affected gene causing CGD in the United States is CYBB, located at Xp21.1, encoding gp91^phox. There are 681 mutations identified in CYBB so far¹¹. Overall, mortality is inversely linked to ROS production¹², with survival increasing in patients with mutations that allow more residual ROS production. Early missense mutations before amino acid 310 (except for His 222) tend to result in an intact protein with some residual ROS production and higher rates of overall survival, while those mutations after amino acid 310, which are in the NADPH, FAD, and heme binding domain, tend to be unable to support ROS production. In contrast, nonsense mutations, deletions and some splice mutations result in no protein production and have the lowest residual ROS production and lower overall survival¹². Additionally, liver abscess and liver disease as evidenced by platelet decline of >9000/mm3/y and elevation in alkaline phosphatase by >0.25U/L/y have been independently linked to mortality¹³. Given its location on Xp21, large deletions involving CYBB may also encompass other genes: telomeric deletions involving KX lead to McLeod syndrome (Kell antigen deletion), as CYBB and KX are < 50 kbp apart. In this deletion, initial red blood cell transfusion can result in formation of anti-Kx and anti-Km antibodies making future transfusions extremely dangerous¹⁴. Duchenne muscular dystrophy (DMD) and X-linked retinitis pigmentosa (RPGR) can also be impacted by large deletions in Xp21.1 as part of a contiguous deletion. Centromeric deletions can cause ornithine transcarbamylase deficiency, which has also been described in X-linked carriers^15–17. This is why chromosomal microarrays to detect the boundaries of large deletions are important in X-linked CGD.

The second most common CGD mutation in the US is in NCF1, the recessive gene encoding p47^phox and accounting for ~25% of cases. The vast majority of mutations are homozygous for a GT deletion at the start of exon 2¹⁸. Of ~900 mutations in NCF1 worldwide, about 80% were homozygous for the GT deletion. This common deletion occurs because most individuals carry two pseudogenes, each of which carries the deletion and flanks the functional gene. Recombination events in which a GT-deleted pseudogene replaces the functional NCF1 result in early truncation and no protein production^18,19. Despite this lack of protein, p47^phox deficient CGD typically has residual ROS, albeit not normal amounts, and is a milder infection susceptibility phenotype.

The second subunit of the cytochrome b₅₅₈ is autosomal recessive p22^phox (encoded by CYBA at 16q24) accounts for 7% of the mutations leading to US CGD¹⁸. Because it is required for gp91phox expression and stability, mutations in CYBA are clinically almost indistinguishable from mutations in CYBB, except that CYBA is autosomal recessive, leading to females with the X-linked phenotype. p67^phox is another autosomal recessive cytosolic component (encoded by NCF2, at 1q25.3), and the critical cytosolic factor for NADPH activation. P47^phox is the partner protein crucial for getting p67^phox into the NADPH complex, partly explaining why p67^phox deficiency is usually more severe than p47phox deficiency. p40^phox (encoded by NCF4 at 22q12.3) is another autosomal recessive member of the NADPH oxidase that serves a more regulatory role. The standard functional test, Dihydrorhodamine- 1,2,3 (DHR) oxidation, may be normal to PMA and ionomycin but abnormal to testing with fMLF or LPS. This form of CGD should be considered in patients with significant inflammatory bowel disease, even without infections^20,21. This diagnosis should be considered in patients severe inflammatory bowel disease, and in particular with very early onset inflammatory bowel disease (VEO-IBD). Essential for reactive oxidant signaling (EROS) is encoded by CYBC1 at 17q25.3. EROS is the critical chaperone for the gp91^phox/p22^phox complex to get out of the ER and into the secondary granule membrane. It is autosomal recessive and tends to have high levels of inflammatory bowel disease^7,22,23.

Clinical Manifestations

Infection

CGD infection susceptibility has been well described in numerous reviews^24–27. Infections typically occur in infancy or early childhood. However, later diagnoses occur in patients with significant residual superoxide production, which can be missed when the phenotype is less severe. Characteristic sites of infection include lungs, liver, skin and bone^24,25. The most common pathogens in the US are Staphylococcus aureus, Burkholderia cepacia complex, Serratia marcescens, Nocardia spp., and Aspergillus spp. In other parts of the world Salmonella spp, Mycobacterium tuberculosis and Bacille Calmette-Guerin (BCG) are also important. There are a few bacteria that are almost pathognomonic for CGD and should prompt diagnostic testing: Chromobacterium violaceum, Francisella philomiragia and Granulibacter bethesdensis^28–32.Both Chromobacterium and Francisella are gram negative organisms that are found in warm brackish water and can cause sepsis in CGD. Chromobacterium and Francisella are both well treated with carbapenems. Granulibacter, a fastidious gram-negative rod, can present as chronic or recurrent necrotizing lymphadenitis and can be particularly difficult to eradicate, and often responds to ceftriaxone^30,31. In the US there has been one case of fatal sepsis secondary to G. bethesdensis³³. European strains of Granulibacter have been more severe in presentation³⁴. Granulibacter requires special culture media, such as Charcoal yeast extract (CYE) and can be identified by 16S sequencing^30,35. In some instances, after parenteral therapy with ceftriaxone, longer term cefdinir has been helpful.

Staphylococcus aureus is a gram-positive bacterium that tends to cause deep seated infections in CGD including lymph nodes, skin, liver and bone. Liver abscess was previously treated with surgery and antibiotics. This approach was complicated by high rates of relapse and recurrence. The liver abscesses of CGD are not very amenable to percutaneous drainage as the abscess is often fibrocaseous and septated^36,37. The seemingly counterintuitive approach of simultaneous therapy with steroids and antibiotics for staphylococcal liver abscess has resulted in decreased need for surgery lower morbidity and mortality³⁶. However, diagnosis of the microbiological cause of liver abscess still requires definitive determination, such as percutaneous aspiration³⁶.

Infections caused by Burkholderia spp often cause pneumonia, but bacteremia can result in both sepsis and hemophagocytic lymphohistiocytosis (HLH)³⁸. B. gladioli and B. pseudomallei have also been reported in CGD^39,40

Aspergillus and other mold infections can be fatal in CGD. With the advent of highly effective antifungal prophylaxis and newer azoles, A. fumigatus is largely curable. However, the non-fumigatus Aspergillus species can be intrinsically resistant to amphotericin and many azole antifungal therapies and can be invasive, eroding through tissue planes⁴¹. In a review of non-Aspergillus infections in CGD the most frequent offenders were Rhizopus spp. and Trichosporon spp. and overall mortality was 26%⁴². Other molds that have been described to cause invasive and often fatal disease in CGD include A. viridinutans, A. nidulans, A. tanneri, Paecilomyces variotii, P. lilacinus, Neosartorya udagawae, and Phellinus spp.^41,43–46. In these cases, if there is isolated disease then surgical resection may be advisable. Because of the stark therapeutic divide, it is critical to definitively diagnose invasive fungal disease to the species level as medical therapies are not always sufficient and adjunctive treatments take time to arrange (hematopoietic stem cell transplant, allogeneic granulocyte infusions).

A variation on fungal pneumonia is “mulch pneumonitis” described in CGD presenting acutely, with fever and dyspnea, typically within 24 hours of exposure and with potentially minimal changes on conventional X-ray^47–51. The history often includes exposure to decaying matter such as mulching, landscaping, demolition, lawn debris or animal feed. Patients may present for dyspnea with what is called community acquired pneumonia. Left untreated, patients invariably return in respiratory distress with abnormal X-ray and widespread diffuse nodules on CT. This condition is life threatening. Treatment typically requires both antifungal (mold-active azole) and antibacterial (carbapenem) and corticosteroids⁵⁰.

Inflammatory complications

Gastrointestinal complications occur in about 50% of patients with X-linked CGD, appearing usually by 7 years, whereas in p47^phox CGD the rate is about 40% but the age of onset is 22 years⁵². Signs and symptoms can range from failure to thrive, irritability and diarrhea in the very young to urgency, frequency, tenesmus, iron deficiency anemia, diarrhea, and hematochezia in adults. Functional gastric, esophageal, and duodenal obstruction have also been described^53,54. Common radiographic findings can include peri-colonic fat stranding, mesenteric adenopathy and bowel wall thickening⁵⁵. CGD-related IBD does not affect overall mortality, but it does cause significant morbidity^12,56. IBD may be the initial presentation of any of the genotypes of CGD. In particular NCF4 deficiency often presents as IBD only, without infections^18,20,21. Typically, the colon is most affected, with the perirectal region most involved. IBD can progress from distal to proximal affecting the entirety of the colon^53,57. Disease is morphologically similar to Crohn disease (CD) with ulcerations, friability and erythema of the mucosa with acute colitis, and crypt abscesses^58,59. There is a high rate of perianal disease, with fistula formation and fissures⁵⁹. In addition to the colonic disease, video capsule endoscopy (VCE) in CGD found that 85% had some sort of small bowel involvement, not predicted by radiographic appearance⁶⁰. Therefore, comprehensive staging of CGD patients with IBD is indicated.

CGD IBD does not correlate with residual superoxide production^12,53. In addition, within families who all share the same mutation rates of IBD are discordant. A comparison of patients with CGD IBD against those with IBD without CGD found that patients with CGD IBD had many of the same genetic risks as in the general IBD population but their burden was lower overall, consistent with the idea that ROS deficiency is per se an important contributor to the development of CGD IBD⁶¹. Inflammasome activation and defective autophagy in the setting of elevated IL-1beta has also been implicated in a mouse model of CGD as a cause of CGD IBD⁶². Treatment of CGD IBD with IL-1 blockade has shown some small improvements in disease, but has not been sustained in small reports^62,63. The intestinal microbiome regulates immunity and inflammasome activation and may be important in the development of CGD IBD⁶⁴. The microbiota of CGD patients with and without IBD show different levels of alpha diversity in CGD patients compared with healthy controls^64,65. Moreover, the beta diversity of the fecal microbiome can distinguish CGD patients from healthy controls⁶⁵. The microbiome in symptomatic CGD IBD is distinct from that in CGD without IBD and can shift from one to the other upon institution of an elemental diet^66,67. There are ongoing trials targeting the microbiome in CGD, both through the use of fecal transplant and also using elemental diets.

Medical therapy for CGD IBD is directed at the dysregulated inflammation, ideally without increasing the risk of infection. Traditionally, the mainstay of therapy has been corticosteroids and lumenal non-steroidal gastrointestinal medications such as mesalamine, or sulfasalazine; for more refractory cases azathioprine is usually added^53,68. Unfortunately, long term steroids can lead to aseptic necrosis and osteopenia⁵³. Therapies aimed at Crohn IBD have been utilized with varying degrees of success. Despite improvement in fistulizing disease with TNF-α blockade in CGD IBD, fatal infections and increased infection susceptibility have been associated with their use^69,70. The humanized monoclonal antibody that targets the integrin α4β7, vedolizumab, has shown modest improvement in CGD IBD. Although it did not show significant sustained improvement it may have a role as a bridge to more definitive therapy^71,72. Ustekinumab, a monoclonal antibody targeting IL-12/IL-23 has been reported to improve CGD IBD but has been associated with some increase in infection susceptibility. However, it can take several months to see results and disease may recur if therapy is stopped and^73,74. We use a long steroid taper until the effects of ustekinumab can be established. There is still not an immunomodulatory therapy that is both safe and effective for severe CGD IBD. The only definitive therapy for CGD IBD is hematopoietic stem cell transplantation (HSCT)^75,76. Encouragingly, CGD IBD largely resolves with transplant and its presence does not impact transplant outcomes. Therefore, HSCT is an attractive option even for patients with inflammatory bowel disease as their sole CGD morbidity^75–78.

The liver is another part of the GI tract that can be affected by dysregulated inflammation, with complications including non-cirrhotic portal hypertension, nodular regenerative hyperplasia, and hepato-splenomegaly resulting in sequestration and thrombocytopenia^13,79.

Approximately one quarter of patients with CGD exhibit signs of pulmonary inflammatory complications^80,81. These can be severe enough to result in hypoxemia and pulmonary hypertension. In one European cohort, upwards of 77% of their patients had pulmonary abnormalities on computed tomography scanning, including 43% with bronchiectasis or fibrosis⁸². Unpublished data suggest that some of this can resolve with transplantation.

Urological manifestations of CGD can include eosinophilic cystitis as well as obstructive uropathy. Both are effectively treated with corticosteroids; however, the episodes can recur and may require low dose steroids over a prolonged period^54,83.

Chorioretinitis, uveitis, and ocular granulomata have also been described in CGD, most of which do not progress. However, there are reports of loss of vision sometimes resulting in enucleation. The etiology of this complication is poorly understood and no clear effective treatments have been described^80,84,85. Patients with CGD should be screened and followed routinely for this inflammatory complication.

Curative therapy

Much of the therapy for CGD is aimed at prevention and treatment of both infections and inflammatory complications. Currently the only available cure for CGD is HSCT, with overall survival >90% for children transplanted < 14 years of age^{76,77,86–88}. Resolution of inflammatory bowel disease and full engraftment with both matched related and unrelated donors have been reported, making transplant a viable option even in the setting of ongoing infection or inflammatory complications. Transplant has also been reported in highly Lyonized carriers, either in the setting of incurable infection or for the same inflammatory complications of the GI tract as in X-linked boys. Gene therapy for X-linked CGD has been performed with success, but has been complicated in some cases by loss of gene expression and in rare cases by development of myelodysplasia or leukemia, as has been the case with some other early efforts at gene therapy⁸⁹. However, some of the newer vectors and approaches are promising. Whether gene therapy can fully control the inflammatory complications of CGD remains to be demonstrated. Newer ex-vivo using technologies like CRISPR/Cas9 have shown promise and are on the cusp of trials in humans^90,91.

Myeloperoxidase deficiency

After the superoxide free radical is converted by superoxide dismutase to hydrogen peroxide (H₂O₂), MPO is the heme-containing enzyme necessary for its conversion to hypochlorous acid (HOCl). MPO is expressed early in myeloid differentiation and resides in the azurophilic (primary) granules of neutrophils and the lysosomes of monocytes. Located at 17q23, this is the most common inherited neutrophil disorder, occurring in approximately 1/2000–4000. It is rarely associated with infection despite its demonstrated role in oxidation of phagocytosed organisms⁹². There are case reports of invasive Candida infections and paracoccidioidomycosis infection, but these infections are typically associated with other co-morbidities such as diabetes mellitus^93–95. It is important to keep in mind that MPO deficiency is another of the causes of an abnormal DHR assay, which can on occasion be quite misleading. More commonly, MPO has been associated with inflammatory disorders given its role as a mediator of inflammation⁹⁶. Recent work in mouse models suggests that MPO is involved in mediating apoptosis and that in its absence inflammation is enhanced, offering a possible overlap between MPO deficiency, in which HOCl is not generated because of an enzyme defect and CGD, in which HOCl is not generated because of a substrate defect⁹⁷

Glucose-6-Phosphate Dehydrogenase Deficiency (G6PD)

Glucose-5-phosphate dehydrogenase deficiency (G6PD) belongs in the category of neutrophil disorders, as it is one of the first two reactions of the hexose monophosphate shunt pathway that generates NADPH. G6PD deficiency affects about 400 million persons worldwide with an increased prevalence in Africa, Asia and the Middle East, making it the most common X-linked disorder⁹⁸. The amount of NADPH available for use in a number of reactions, including for the NADPH oxidase, is dependent on G6PD, and therefore less or less effective enzyme results in insufficient NADPH and reviewed recently^11,99. G6PD deficiency is most commonly associated with hemolysis of red blood cells. Located at Xq23, G6PD deficiency can range from mild to severe, with 5 different classes of variants. Infections occur in the most severe form of G6PD deficiency and generally phenocopy CGD susceptibility¹⁰⁰. G6PD deficiency should be considered in a patient with severe hemolysis and CGD-associated infections.

Leukocyte Adhesion Deficiency

Defects in migration resulting in infection susceptibility and inflammatory complications can be due to one of the three types of Leukocyte Adhesion Deficiency (LAD). These defects can result in variable phenotypes as a result of impaired adhesion to the endothelium or transendothelial migration out of the bloodstream¹⁰¹. Neutrophil and monocyte adhesion to the endothelium and to other cells and to some pathogens is mediated in part by molecules collectively known as integrins [CD11a/CD18 (LFA-1), CD18/CD11b (CR3 or Mac-1), CD18/CD11c (CR4), CD18/CD11d] which are stored in intracellular vesicles and translocate to the leukocyte cell surface where they can interact with intercellular adhesion molecules (ICAMs) on the endothelial surface. Lymphocyte function-associated antigen (LFA-1) is a heterodimeric integrin present on the surface of all leukocytes that is essential for neutrophil recruitment to sites of tissue inflammation, but also for processes such as homotypic adhesion leading to T cell activation¹⁰². Mac-1 or Complement receptor (CR) 3 is essential in responding to chemotactic factors, such as the complement fragment C5a, IL-8, leukotriene B4 (LTB4), or the bacterial product formyl-methionyl-leucyl-phenylalanine (fMLF) and leukocyte adhesion to the endothelium^101,102. CR4 (also known as p150,95) mediates binding to lipopolysaccharide (LPS) directly as well as fibrinogen. It is also highly expressed on monocyte-derived dendritic cells¹⁰³. CD18/CD11d is an integrin whose expression is more complexly regulated and whose role in human health and disease is less well understood¹⁰³. There is a bidirectional interplay between the neutrophil (integrin expression) and the endothelium (ICAM expression) required for appropriate adhesion and extravasation. Neutrophil integrin expression is entirely dependent on the common subunit of all 4 integrins, CD18 encoded by ITGB2 at 21q22¹⁰⁴. Therefore, defects in CD18 expression or function lead to defects in any or all of the CD11 family molecules and therefore defects in adhesion, trafficking and killing. Disease severity in LAD-1 is determined by the amount of CD18 expression and categorized as severe (<2%), moderate (2%−30%), or mild (>30%). Inherited in a autosomal recessive manner, patients with LFA-1 severe mutations frequently present with delayed umbilical cord separation, poor wound healing, neutrophilia, oral ulcers, and aggressive gingivitis with accelerated apical bone loss^105,106. The severe periodontitis typically results in loss of all adult teeth by early adulthood. The mechanism of the inflammation associated with LAD1 is somewhat counterintuitive, as the inability of neutrophils to exit the vascular space leads to tissue neutropenia, which is associated with hypoinflammatory lesions, such as in neutropenia. However, tissue resident macrophages are constantly monitoring for adequate neutrophil levels in the tissue, and when they decline macrophages produce IL-23 which eventually leads to IL-17 and G-CSF production. Therefore, it is the local production of high levels of IL-17 as a result of macrophage IL-23 that leads to the severe inflammatory complications of LAD1, including severe gingivitis and inflammatory skin lesions¹⁰⁷. This latter observation led to the use of ustekinumab in the treatment of gingivitis and inflammation in LAD1¹⁰⁸. HSCT is curative for LAD1 but results are complicated by high rates of graft failure and graft versus host disease¹⁰⁹. In view of the persisting difficulties of HSCT there has been significant effort for gene therapy for LAD1, which has so far been successful with survival of all recipients past one year¹¹⁰.

LAD-2 is due to defect in the fucosylation of the protein CD15s (sialyl Lewis^X, SLe^X) on neutrophils, thus impairing the early low-avidity rolling step of neutrophil adhesion which is mediated by selectins on the endothelium. The molecular defect is in the guanine diphosphate (GDP)–fucose transporter-1 (FUCT1 or SLC35C1). This defect in glycosylation places LAD2 within the congenital disorders of glycosylation (CDG), making this CDGIIc. Inherited in an autosomal recessive manner, patients with LAD2 have the Bombay blood group phenotype, which is due to the absence of the H antigen, which is also dependent on fucosylation¹⁰⁵. There are many fucoslylated proteins, all of which are affected, so patients often present with short stature and cognitive deficits. The infection phenotype in LAD2 appears to improve with age and is usually less prominent in adulthood, but the number of cases reports is small. There may be a subset of cases who respond to fucose supplementation, as well¹¹¹. Because the integrin, phagocyte killing and T-cell receptor are still functional in this LAD2, the infectious complications tend to be less severe¹⁰⁶.

Initially thought to be a variant of LAD-1, LAD-III was first described in a Turkish patient with mutations in FERMT3, which encodes KINDLIN3, an adaptor protein expressed in hematopoietic cells that regulates integrin activation through binding to the short intracellular tails of β2 integrins¹¹². The process of KINDLIN3 binding to the intracellular portion of the integrins is called inside-out signaling, since the KINDLIN3 interaction strengthens integrin binding. Described in only a handful of patients and families, LAD III is autosomal recessive and presents with infections and a Glanzmann thrombasthenia phenotype bleeding tendency^113–115. Patients typically have a purpuric (“mulberry”) rash at birth and may have early fatal infections. LAD-III platelets have decreased binding to soluble fibrinogen, and do not respond properly to thrombin via thrombin receptors, resulting in improper platelet granule secretion via integrin activation and hence the purpuric rash. HSCT is curative for LAD3 and should be considered early in the course^{109,116–118}.

Table 1.

Genetic testing for clinicians who suspect a neutrophil disorder and the genes affected.

Disease and Protein	Gene	Diagnostic Test	Chromosome	Inheritance	Sequencing
CGD
- gp91	CYBB	DHR	Xp21.1	X-linked	Yes
- p22phox	CYBA	DHR	16q24.2	AR	Yes
- p47phox	NCF1	DHR	7q11.23	AR	Possible1
- p67phox	NCF2	DHR	1q25	AR	Yes
- p40phox	NCF4	May look normal2	22q12.3	AR	Yes
- EROS	CYBC1	DHR may be low	17q25.3	AR	Yes
- RAC2	Rac2		22q13	AD	Yes
G6PD	G6PD	Enzymatic activity	Xq28	X-linked	Yes
Myeloperoxidase deficiency		3DHR may appear abnormal	17q23	AR	Yes
LAD
- LAD1	ITGB2	Flow Cytometry4	21q22.3	AR	Yes
- LAD2	SLC35C1	Flow Cytometry4	11p11.2	AR	Yes
- LAD3	FERMT3	Flow Cytometry4	11q13.1	AR	Yes

¹May be difficult given that there are two pseudo genes which may miss the typical GT deletion

²If clinical suspicion is high – request DHR with serum-opsonized zymosan or E.coli rather than PMA. However direct sequencing is indicated for high clinical suspicion.

³Patient may erroneously be diagnosed with CGD, however direct visualization of neutrophils when stained for peroxidase is more specific.

⁴Abnormal flow cytometry for CD11a, CD11b, CD11c and CD18 in combination with clinical suspicion fitting the phenotype