|Home | About | Journals | Submit | Contact Us | Français|
Chronic Granulomatous Disease (CGD) still causes significant morbidity and mortality. The difficulty in considering high-risk yet curative treatments, such as allogeneic bone marrow transplantation, lies in the unpredictable courses of both CGD and bone marrow transplantation in different patients. Some CGD patients may have frequent infections and/or suffer from granulomatous or autoimmune disorders necessitating immunosuppressive therapy, but also experience long periods of relative good health. However, the risk of death is clearly higher in CGD of all types, and the complications of CGD short of death can still cause significant morbidity. Therefore, with recent developments and improvements, bone marrow transplantation, previously considered an experimental or high-risk procedure, has emerged as an important option for patients with CGD. We will discuss the complications of CGD that result in significant morbidity and mortality, in particular the most common infections and autoimmune/inflammatory complications as well as their typical management. We will then discuss the status of bone marrow transplantation.
Chronic Granulomatous Disease results from defects in the NADPH oxidase complex resulting in an inability to produce the superoxide anion necessary for normal killing of bacterial and fungal microorganisms. In addition, this defect predisposes to granulomatous complications and autoimmune diseases. Mutations in at least five different genes involved in the assembly and activation of the NADPH oxidase can lead to CGD. 1 The gene encoding the enzymatic center of the NADPH oxidase, gp91phox, is on the X-chromosome and accounts of about 2/3 of the cases. Autosomal forms occur from mutations in p47phox, p67phox p22phox or p40phox, with the latter being the most recently described.2 In general, gp91phox deficient patients (X-linked CGD) are the most severely affected whereas patients with mutations in p47phox seem to have the best outcomes overall. Deficiency in p40phox may predispose to more GI disease and fewer infections.2 Specific mutations affect the severity of disease through the amount of residual NADPH oxidase activity.3 However, even among patients with similar NADPH oxidase mutations there can be widely different clinical outcomes. Therefore, the genetic type of CGD, the specific mutation, the patient’s own infection history, the presence of inflammatory or autoimmune complications and access to appropriate medical care all factor into what to expect from CGD in a particular patient’s case.
Despite the significant progress made in antibiotic and antifungal therapy and prophylaxis, patients with CGD still develop serious infections. Most large studies have shown an infection rate of around 0.15 to 0.3/year.4, 5, 6 The US National Institutes of Health (NIH) has followed more than 250 patients with CGD over almost 40 years, the majority of whom were diagnosed following infections of skin, lymph node, lung or liver. A small group of patients (~5%) were identified because of inflammatory lesions as their primary clinical event. The diagnosis was usually established early in life (median age of diagnosis 5.4 years), although a small proportion were diagnosed as adults. Notably, the majority of these later diagnoses were due to autosomal recessive forms of CGD.
Isolation of the microorganism causing infection in CGD is essential to rational and appropriate treatment, but it is not always feasible. In the last ten years, 80% of CGD patients at NIH with a pulmonary infection underwent some type of diagnostic procedure, either needle biopsy or bronchial lavage. Of these, 52% were successful in identifying a pathogen. Co-infection, such as fungal with bacterial, was found in less than 10% of biopsies. Viral infections appeared at similar rates as in the general population (unpublished data).
The majority of infections in North American CGD are due to four bacterial organisms (Staphylococcus aureus, Serratia marcescens, Burkholderia cepacia complex, and Nocardia species) as well as species of the fungus Aspergillus. Invasive aspergillosis has been a major cause of morbidity and mortality in CGD, but the advent of the newer azole antifungals has dramatically changed the treatment and outcome of these infections and shifted the intractable fungal infections to non-fumigatus Aspergillus species, dermateacous molds and hyalohyphomycosis such as paecilomyces.1, 7, 8
CGD patients may present without symptoms or with low grade fevers and only mild constitutional symptoms inconsistent with the extent of disease seen by imaging studies. Consequently, frequent imaging studies (e.g., CT, MRI), are recommended for clinical monitoring. The paradoxically dampened inflammation in response to some serious infections and the exaggerated responses to some non-infectious stimuli (see below) remains perplexing.
The lung was the most common site of disease in the NIH cohort and Aspergillus was responsible for ~40% of the culture positive cases. Chest scans and markers of acute inflammation (e.g., C-reactive protein and erythrocyte sedimentation rate) have proven useful in diagnosis and monitoring of fungal disease (unpublished data). The role for serology such as the β-D glucan and galactomannan assays are undefined in CGD, but when positive may be helpful to follow in some cases. North American studies have identified a much higher incidence of Burkholderia and Nocardia infections than in European reports, which in part may reflect the differences in diagnostic approaches and may also reflect environmental differences.4, 9
Emerging pathogens in CGD include Gram-negatives (eg. Granulibacter bethesdensis10), Gram-positives (eg Actinomyces11), and fungi (eg. Neosastorya udagawae12). Occurrence of these uncommon pathogens in CGD may provide clues to the critical pathways and functions of the NADPH oxidase.13–17
Liver abscesses are common in CGD.18 Thirty percent of NIH patients have had liver abscesses with 25% of these occurring more than once. Staphylococcus aureus was the organism most frequently cultured and surgical resection was the usual treatment. Percutaneous drainage was usually not helpful, as liver abscesses associated with CGD tend to develop multiple loculations. When resected, the lesions are a collection of microabscesses.18 Corticosteroids have been reported to be helpful in 2 cases of liver abscess.19 Other staphylococcal infections are typically confined to the skin or lymph nodes.20
Patients compliant with prophylaxis still develop skin infections, but these infrequently spread. Skin and soft tissue infections are caused by S. aureus, Klebsiella species, Serratia marcescens, Burkholderia cepacia complex and some fungi. Lymph node and skin infections have decreased overall and constitute only about 20% of the infections seen in NIH patients.
Antibacterial (trimethoprim/sulfamethoxazole) and antifungal (itraconazole) prophylaxis have significantly reduced the rates and severity of infections in CGD, but breakthrough infections still occur.21–23 Prophylactic antibiotics were used in 93% of NIH CGD patients, with trimethoprim/sulfamethoxazole the most frequent. Intolerance to sulfamethoxazole or other adverse events typically led to use of trimethoprim alone, cephalosporins, or quinolones.
Fungal prophylaxis was used by only 68% of the patients, though recommended for all CGD patients. Of these, 55% were on itraconazole, 30% on posaconazole and 15% on voriconazole. Typically patients on the latter two were receiving them after having been treated for an invasive fungal infection. There are no data in patients with CGD comparing voriconazole, posaconazole or itraconazole. A single center transplant study did show better outcomes with posaconazole as compared to itraconazole; however direct extrapolation to CGD patients may not be appropriate.24
Mild toxicity related to drugs was recorded in 36% of the overall NIH cohort, 15% of whom had photosensitivity, most likely due to voriconazole or trimethoprim/sulfamethoxazole. Severe photosensitivity leading to squamous cell carcinoma and melanoma has been reported with long-term voriconazole.25, 26 Patients receiving voriconazole should use aggressive sun protection. For patients with severe voriconazole-induced photosensitivity, despite sun avoidance, posaconazole causes less photoreactivity.
Interferon gamma (IFN) was shown in 1991 to be effective prophylaxis for CGD.27 However, use in Europe has been less than in the United States as non-randomized European data suggested less benefit from IFN use.28 Even in our own cohort, with the advent of better antifungals and more active oral antibiotics, the percentage on INF is only 36% due to intolerance or lack of access. Fevers, myalgias and irritability were reported as reasons for stopping the IFN in 13% of patients in one study. 29
Renal failure or severe dysfunction occurred in 3.5% of our patients, probably due to long-term amphotericin exposure before the advent of newer agents.
Dysregulated inflammation in CGD typically occurs in response to a trigger and may be due to either elevated pro-inflammatory or decreased anti-inflammatory mediators. Patients with CGD frequently experience inflammatory complications and some may develop autoimmune problems.30
Other than infection, a characteristic feature of CGD is granulomatous inflammation. CGD granulomas are typically non-caseating, composed of multinucleated giant cells, and can be found in multiple tissues, including brain, lungs, liver, spleen and gastrointestinal tract. When present in hollow viscera they can lead obstruction, such as the gastric outlet or ureteral obstruction that are relatively common in X-linked CGD. For most of these granulomas no pathogen is identified and they respond rapidly to steroids, suggesting that the inciting event is not an invasive infectious one. Surgical intervention should be avoided and corticosteroids, when used, are usually started at doses of 1 mg/kg/d and then tapered after one week. In many patients, the symptoms recur when the steroid dose is reduced, thus our current practice is to taper the corticosteroid dose gradually to around 0.1mg/kg/day on alternate days. Patients with recurring problems may be kept on low dose prednisone for years, which does not appear to increase infection rates or impair growth.6
A unique presentation in CGD is an acute pneumonitis due to the inhalation of mulch or other decayed organic matter (e.g., potting soil, hay, leaves). Exposure to a large burden of fungal elements and spores triggers an acute inflammatory response leading to fever, hypoxia, and diffuse infiltrates, usually beginning within one week of the exposure.31 Similar responses are seen in CGD mice exposed to live or even dead fungi, 32 indicating that some of this pathology is due to dysregulated inflammation rather than infection per se. Bronchoscopies and lung biopsies may yield one or more fungal pathogens, especially Aspergillus. In addition to rapid institution of antifungals, moderately high doses of prednisone (1 mg/kg/day), help prevent respiratory failure and may facilitate more successful healing.31, 33
Inflammatory lesions without demonstrated pathogens have also been noted in the lungs of CGD patients, and are characterized by discrete infiltrates on chest CT that wax and wane without intervention. In some patients diffuse pulmonary inflammation can progress to hypoxia and functional limitation.34 It is difficult to exclude infection despite negative cultures, cytology, nucleic acid testing and the lack of improvement in response to antibacterial or antifungal agents. However, in some cases empiric treatment beyond corticosteroids has included methotrexate. Progressive lung inflammation with augmented NF-kB activation and elevated pro-inflammatory cytokines has been recently demonstrated in CGD mice (p47- and gp91phox-deficient) following intratracheal challenge with zymosan or LPS.35
Inflammatory bowel disease characterized by granulomatous involvement of the bowel especially in the perirectal area is hard to distinguish pathologically from Crohn’s disease. However, the inflammatory bowel disease of CGD is typically limited to the bowel and unassociated with any of the extra-intestinal manifestations often seen in Crohn’s. In the NIH series 43% of X-linked and 11% of p47phox deficient patients had biopsy-proven symptomatic bowel disease.6 How many had active subclinical disease remains unknown. Other autoimmune diseases in CGD patients and carriers have included IgA nephropathy, anti-phospholipid syndrome, systemic lupus erythematosus, idiopathic thrombocytopenic purpura and juvenile idiopathic arthritis.7 An estimated 10% of the CGD patients followed at NIH have some autoimmune manifestation other than inflammatory bowel disease. The underlying etiology for this predisposition to autoimmunity remains unknown. Polymorphisms in a few genes have been loosely associated with inflammatory complications in CGD (MPO, mannose binding lectin, Fc receptors IIa, IIIa and IIIb, TNFa and IL-1 receptor).36
Dysregulated inflammation may play a role in the development of autoimmune complications in CGD. For example, normal NADPH oxidase activity plays regulatory roles in apoptosis 37–39 and macrophage clearance of apoptotic cells.40–42 Altered NADPH oxidase function may therefore lead to aberrant macrophage programming, impaired clearance of antigen and intracellular elements with further recruitment of neutrophils and prolonged production of IL-8, IL-1b, caspases and other pro-inflammatory cytokines. 43, 44 Persistence of CGD phagocytes during induced inflammation was reported in human X-CGD45 and in CGD mouse peritonitis.46, 47 These diverse studies suggest that the role of the NADPH oxidase in CGD extends far beyond the simple predisposition to infection.
Treatment for inflammatory and autoimmune complications in CGD is problematic since most agents are immune suppressive and immunity is already impaired in CGD. Many patients respond well to corticosteroids, but they may require prolonged courses. Sulphasalazine and azathioprine are useful steroid sparing agents. TNFa inhibitors such as infliximab are effective anti-inflammatories but may significantly increase the risk of severe and even fatal infections. 48 The risk of infection needs to be weighed carefully against the risks of uncontrolled mucosal inflammation or surgery that may be further complicated by persistent inflammation, abscesses, and fistulae formation at surgical sites. If TNF inhibitors are used, augmented prophylaxis and enhanced vigilance regarding exposures are mandatory. Methotrexate and plaquenil can be effective in those with arthritides or lupus-like problems.
Currently the only known cure for CGD is allogeneic hematopoietic cell transplantation. Historically this has only been reluctantly offered due to risks of procedure related morbidity and mortality. Additionally, unrelated donor transplants were riskier than sibling transplants, and the pool of donors was limited. From 1973, when the first CGD bone marrow transplant was performed, until now, 99 transplants, not including cord blood recipients, have been published, with the majority being single case reports. 49–70 However, of the 99 patients transplanted, 50 have occurred in the last 10 years compared to 49 in the prior 27 years. With the advent of non-myeloablative regimens, the risks surrounding transplant have decreased and have permitted transplantation in patients with ongoing infections. Additionally, more transplants are being performed using unrelated donors. Notably, the first transplant ever performed for CGD used an unrelated donor and to date 22 patients have been transplanted using unrelated donors with the majority being performed within the last ten years.
Hematopoietic stem cell transplant has been more frequently offered to European patients with CGD than in North America. The first large report of bone marrow transplantation for CGD was from a group of European centres describing the results in 27 patients transplanted from 1985 to 2000 (7 of whom were described previously in single case reports).65 HLA matched sibling donors were used for 25 of these cases and the majority received a myeloablative, busulfan-based regimen. In nine patients transplanted during a refractory infection, there were 2 graft failures and severe Graft versus Host Disease (GvHD) in 3, with one patient dying as a result.
In the largest North American study published to date, Horwitz et al. reported the outcomes of 10 patients who received a fully matched sibling donor transplant with a non-myeloablative conditioning regimen of fludarabine, cyclophosphamide, and antithymocyte globulin (ATG).69 Stem cell products were T-cell depleted and donor lymphocyte infusions were given post transplant to augment engraftment. Eight patients engrafted, but one had significant GvHD resulting in death, with one additional patient dying 18 months post transplant with pneumococcal sepsis despite full myeloid engraftment. Of the non-engrafted patients, both survived and went on to re-transplant with one dying subsequently. Long term follow up in the engrafted patients show stable mixed chimerism in two, including donor lymphoid engraftment of less than 50% in one patient but continued myeloid engraftment, with more than 10 years follow up. All surviving patients with engraftment remain phenotypically well with no evidence of CGD related autoimmune complications or infections.
In 2009, a survey of North American centers treating patients with CGD performed in conjunction with the Center for International Blood and Marrow Transplant Research (CIBMTR) found 59 patients had undergone allogeneic transplant for CGD with 71% survival overall. Three of these patients had survived beyond 10 years, but outcome data are not published. As transplant methods have changed, efforts are underway to comprehensively compile the North American CGD transplant experience both retrospectively and prospectively.
Two other large single-centre studies were recently published, both from European centres. Soncini et al. described the results in 20 patients transplanted from 1998 to 2007, one of whom was previously reported as a single case study. Patients ranged in age from 15 months to 21 years. Ten of those were with matched sibling donors, nine using bone marrow and one receiving cord blood. The remainder received transplants from matched and single mismatched unrelated donors, including one cord blood transplant. The follow up ranged from 4 to 117 months; 18 patients survived with continued normal neutrophil function (90%) and two died from pretransplant fungal infections. The majority of the patients received a busulphan/cyclophosphamide conditioning regimen with alemtuzumab added for those receiving unrelated donor products.70 Schuetz et al. also reported 12 patients, 9 of whom received grafts from unrelated donors. The majority received busulfan/cyclophosphamide with or without either ATG or alemtuzumab. Two patients had graft failure and 5 patients had Grade 1 or 2 acute GvHD. At a mean follow up of 53 months, 9 of the 12 were alive including 7 of the 9 recipients of matched unrelated transplants, all with stable engraftment, including one patient with mixed chimerism. 62
Most recently a European consortium reported good engraftment rates and minimal GvHD using a non-myeloablative busulfan and fludarabine based regimen for both matched related and matched unrelated donors. The intravenous busulfan dose was targeted to achieve an AUC between 45–65mg/h and either ATG or alemtuzumab was added along with mycophenolate mofetil for graft versus host disease prophylaxis. Of their 24 patients, 9 had matched unrelated donors. Eight patients developed grade 1 acute GvHD, one patient had Grade 2 and one patient had chronic GvHD of the skin only, which responded to treatment. Their only death post transplant was due to pneumonia giving an overall survival to date of 96%. 71
Preliminary data from the NIH, also suggest that intravenous busulfan may be an integral part of transplant conditioning for patients with CGD. The doses we used were lower than used by Gungor et al., at least based upon the AUCs measured. Fludarabine was not a part of the regimen but alemtuzumab was. 300cGy of total body irradiation(TBI) was also administered to patients receiving unrelated donor grafts. Eleven patients were described by NIH, 9 received unrelated donor products and patients ranged in age from 3 to 32. There was one failure to engraft using an unrelated cord blood product, and late graft rejection occurred in one patient who received an unrelated donor product. The remainder had almost one hundred percent myeloid engraftment with excellent NADPH oxidase function. There were only two patients with GvHD, both in the skin (one grade 1 and one grade 2). One patient died from renal dysfunction unrelated to transplant, the rest are alive and well, including both patients with graft failure, giving an overall survival of 10/11 or 91%. Notably 9 of the 11 patients had ongoing infection at the time of transplant, and 4 received granulocyte transfusions during the peritransplant neutropenia period.72
The first cord blood transplant for CGD was an 8 year old male transplanted with an unrelated donor matching at 5 of 6 loci published in 1999 by Nakano et al.73 He was conditioned with 10 Gy TBI, ATG, and cyclophosphamide, but died at day 51 due to infection. Seven subsequent patients have been reported as having received cord blood products, either from related or unrelated donors.54, 64, 70, 74–77 Three of the patients have required second transplants. One patient received his initial cord blood product for his re-transplant. All appear to have done well, even when a cord product was used for both transplants. More recently with advanced genetic and fertility techniques, three cases of preimplantation selection have resulted in live births of siblings who have provided either cord blood and/or bone marrow. The patients who received these products appear to be doing well.78, 79
Allogeneic stem cell transplantation for CGD is becoming more common and reflects increased overall transplant success. Survival has increased from approximately 85% prior to 2000 to 90–95%, based on recently reported outcomes and our own results, even with the use of unrelated donors. In fact, outcomes with perfectly HLA matched unrelated donors appear to approach if not equal those using HLA sibling donors. This suggests that donor availability should not be limiting for transplant of CGD patients.
Even for those without a matched unrelated donor, cord blood products are proving to be a reasonable alternative, and are being used more frequently. Even in adults, double cord products have had good engraftment rates, at least in the setting of leukemia. 80 In one study, the non-relapse mortality was slightly higher for recipients of double cord products compared to those receiving matched unrelated or matched related donor products. Studies will be needed in CGD to determine if a double cord transplant is preferable to an unrelated donor transplant.81 Although one published case used a haploidentical donor, the patient rejected requiring a second transplant.54.
Both peripheral blood stem cells and marrow have been used successfully, and the choice for CGD patients currently depends on donor and center preferences. Data from transplants in aplastic anemia suggest that bone marrow products result in less GvHD; however cell dose can be a limiting factor.82 Older patients with CGD often have splenomegaly and/or hepatomegaly, thereby requiring a larger cell dose. Although T cell depletion of products has been used in transplants for patients with CGD, the incidence of GvHD with donor lymphocyte infusion is significant, as seen in the first NIH series.69 In vivo or in vitro T-cell depletion with alemtuzumab appears to result in less GvHD without significantly affecting engraftment, although the need for viral monitoring is prolonged.
Some transplant centers prefer myeloablative transplant regimens.65 Although graft rejection is more likely with a reduced intensity conditioning regimen, the risk of GvHD, particularly acute GvHD, and regimen related toxicity, appears to be reduced with the non-myeloablative regimens. 83–85 This type of conditioning also allows transplantation during ongoing infection, with fewer infection related deaths. Further, those who have rejected after receiving reduced intensity conditioning have for the most part gone on to successful second transplants. On the other hand, patients with McLeod syndrome, (Kell antigen deficiency due to contiguous gene deletion of XK which is found next to the CYBB gene), who have red cell antigen sensitization should be considered for a myeloablative regimen, or at least pretreatment with rituximab, to limit red cell incompatibility as the availability of McLeod matched blood is extremely limited. Elimination of B cells with anti-CD 20 therapy prior to transplant diminishes the risk of transfusion reactions and makes red cell management easier during the transplant period prior to conversion to donor blood type.51 Those without preexisting red cell antibodies however, have successfully undergone non-myeloablative transplant.64 Most successful regimens in CGD patients appear to include busulfan. Some consider fludarabine necessary as well; however, the experience at NIH does not support this.
The question remains of which CGD patient to transplant? Given the current success rates, some favour transplanting all CGD patients who have an appropriate donor at the earliest opportunity. The recent data from Kuhns et al, showed that patients with very low superoxide production had worse long term survival than those with higher levels of NADPH oxidase activity suggesting that these patients might be considered appropriate candidates for early transplantation, particularly if a sibling matched donor is available. 3 However, even within this subgroup, there are patients who do relatively well for prolonged periods. Elevated alkaline phosphatase, a history of liver abscesses, and decline in platelet count reflecting portal hypertension are adverse prognostic indicators.86 These patients may also be considered for early transplantation.
Even with improved survival and longevity due to better infection and inflammation management, complications and their consequences can accumulate over time. However, transplant outcomes are probably better before infectious and inflammatory damage accumulate. Transplant has reversed some of the inflammatory and autoimmune complications associated with CGD and may prevent their development.65 Therefore, patients with significant inflammatory or autoimmune disease should also be at least evaluated for transplant, preferably at a center with experience in CGD transplantation. Those who have an active infection should not be summarily excluded, as non-myeloablative regimens have been successful even in this setting. Additionally, granulocyte transfusions may be helpful during the transplant period for those with active infections and do not appear to impact on engraftment. 87 For those with a prior history of infections, including fungal infections but no active infection, the necessity to use granulocytes is not clear. Patients who are being considered for transplant should not receive granulocytes prior to transplant (as opposed to during) so as to avoid the development of HLA alloimmunization.
While overall CGD life expectancy is still less than the general population even with the best current care, the strides in infection and inflammation management over the last decades have been significant. Allogeneic hematopoietic transplantation may have unanticipated consequences, and even the reduced intensity regimens may pose unknown long-term risks. Although there has been strong interest and progress in gene based therapies, it has not been shown to be curative at this point and has been reviewed elsewhere.88 Further, even ex vivo gene therapy appears to require some form of conditioning, so cytoreductive agents may still be needed.89–91
However, allogeneic transplantation has also improved dramatically over the last decade, due to improved conditioning regimens and GvHD prophylaxis, high resolution sequence based matching, and improved pre-, peri- and post- transplant management. It has become a successful and sensible option for many patients with CGD, that will likely treat and prevent both infectious and inflammatory complications. While further studies will be required to determine optimal timing, donor selection, and long-term efficacy in these patients, hematopoietic stem cell transplant is finally coming of age as a curative treatment of CGD.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.