C 1,000,000) genetically transmitted disease in which a. Surgical Management of Pulmonary Infections in Chronic Granulomatous Disease of Childhood

Transcription

1 Surgical Management of Pulmonary Infections in Chronic Granulomatous Disease of Childhood Helen W. Pogrebniak, MD, John I. Gallin, MD, Harry L. Malech, MD, Alan R. Baker, MD, Christopher A. Moskaluk, MD, William D. Travis, MD, and Harvey I. Pass, MD Thoracic Oncology Section, Surgery Branch, National Cancer Institute, Laboratory of Clinical Investigation, National Institute of Allergy and Infectious Disease, and Surgical Pathology Laboratory, Clinical Center, National Institutes of Health, Bethesda, Maryland Chronic granulomatous disease of childhood is an inheritable disorder of phagocytic cell respiratory burst resulting in recurrent, life-threatening, catalase-positive infections. The lung is the most common site of infection, and pulmonary disease is the primary cause of death in greater than 50% of children with chronic granulomatous disease. Still, the role of surgery in management of this disease remains undefined. Between 974 and 990, 9 patients with chronic granulomatous disease required 3 thoracic interventions at our institution. Patients ranged in age from 2.5 to 27 years (mean age, 5 years). Seventeen of 9 patients (89%) had had previous pulmonary infections. Patients presented as toxic (temperature > 38.5OC, chest pain, and cough) in 22 instances before the 3 procedures. Aggressive surgical intervention for diagnosis and extirpation of localized infections was undertaken with lobectomylpneumonectomy with or without other procedures (5), bisegmentectomy (2),segmentectomy with or without other procedures (5), or wedge with or without other procedures (3). In five instances, an empyema was drained; a chest tube for a sterile collection was placed in one instance. There was one intraoperative death, and 3 patients died 22 to 600 days postoperatively with overwhelming sepsis. The mean hospitalization was 0 days (range, 24 to 600 days). Wound complications occumd in 5 patients, requiring 7 separate anesthetic debridements. A change in therapy was dictated by the results of the procedure in 23 of 3 instances (74%). Thoracic surgeons must be aware of this rare cause of immunosuppression in these children and, due to the unusual nature of the pulmonary infections, should follow an aggressive approach in their diagnosis and management. (Ann Thorac Surg 993;55:844-9) hronic granulomatous disease (CGD) is a rare (: C,000,000) genetically transmitted disease in which a compromised inflammatory system must confront recurrent life-threatening infections with common, as well as unusual, catalase-positive organisms [ -3. The basic defect is the inability of the patient's phagocytes to respond to stimuli with a normal respiratory burst. This respiratory burst, through membrane and cytosolic NADPH (nicotinamide adenine dinucleotide phosphate [reduced form]) oxidase systems, produces superoxide and hydrogen peroxide, which will lead to killing of ingested microorganisms. The inheritance patterns, 60% X-linked, 35% autosomal recessive, and approximately % autosomal dominant, are characterized by specific defects in the genes coding for the four proteins that make up the NADPH oxidase systems [4-7. Diagnosis is most commonly made by the nitro blue tetrazolium test or direct measurement of superoxide production or NADPH oxidase activity. Most recently antibodies to the defective Presented at the Thirty-ninth Annual Meeting of the Southern Thoraac Surgical Association, Wesley Chapel, FL, Nov 57, 992. Address reprint requests to Dr Pass, Thoracic Oncology Section, Surgery Branch, National Cancer Institute/NIH, Building 0, Rm 2807, Bethesda, MD proteins, as well as complementary DNA probes, have become available for Western and Northern analyses. The patient with CGD is unable to combat infection by bacteria and fungi that contain catalase, an enzyme that degrades hydrogen peroxide. Granuloma formation, possibly due to inadequate degradation of microbial antigens or other inflammatory mediators by the defective phagocytes, will result from the ongoing inflammation. Before antibiotic prophylaxis was introduced [8, 9, the clinical course was characterized by repeated and protracted infections, with death usually in the first or second decade. The importance of the thoracic surgeon in the management of children with CGD is underscored when one considers that more than 80% of these children will present with life-threatening lung infections. These infections are particularly difficult to treat and may be the result of unusual organisms. The thoracic surgeon not only must assist in the diagnosis of the radiographic abnormalities, but also must decide whether therapeutic ablation of the involved segment should be performed in these frequently septic individuals. The National Institutes of Health (NIH) has had a long-standing interest in the characterization of the genetic and molecular biologic abnormalities as well as in the treatment of CGD in more than 60 patients [4-6, 9. This report describes the ap-

2 Ann Thorac Surg 993;55:8449 POGREBNIAK ET AL 845 Table I. Symptom Complex Parameter Not Toxic Toxic (0). T, ("C) t 0.2" ESR (mm/h) 53 C 9 86 t 8" Hgb (mg/dl).6 t WBC (03/d~) 0.6 C Antibiotic prophylaxis (No.) 9/9 6/22 Radiography Focal unilateral 7 (78%) 2 (54%) Focal bilateral (%) 2 (9%) Diffuse bilateral (%) 8 (36%)" pz < 0.05 versus not toxic. ESR = erythrocyte sedimentation rate; Hgb = hemoglobin level; T, = maximum temperature; WBC = leukocyte count. proach used at our institution in 9 patients with CGD requiring thoracic surgical intervention since 974. Material and Methods Patient Population Of 60 patients treated at the NIH for CGD, 9 have required thoracic surgical consultation and are the subject of this report. All 9 were referred to the NIH with a presumed diagnosis of chronic granulomatous disease. All had confirmation of the disease at the NIH by nitro blue tetrazolium test, WesterdNorthern blot analyses, or both. The median age was 5 f.7 years (range, 2.5 to 29 years) at the time of admission for pulmonary evaluation. The original diagnosis of CGD, however, was made at the age of 5.3 f.7 years in these children, and there was a delay in this diagnosis when one considers that the age of theirfirst infection was 2. * 0.8 years. The majority of the children were white (n = 4) and male (n = 6), and the inheritance paralleled well-established percentages in the literature (3 X-linked [68%], 6 autosomal recessive [32%]). Seventeen of the 9 patients (89%) had had a previous pulmonary infection, and all had at least two previous infection sites documented including skin (ll), mucosa (8), sinus (3), brain (2), bone (3), liver (3), or nodes SymptomslSigns at the Time of Thoracic Evaluation These 9 patients were evaluated for thoracic intervention in 3 instances. In case (3%) the patient was asymptomatic and presented with an incidental radiographic abnormality. The most common symptoms before the 3 explorations were pleuritic chest pain (58%), cough (52%), chills (45%), and dyspnea (35%). Analysis of the symptom complexes revealed two distinct types of presentation: asymptomatic or cough only (9 presentations) or toxic, consisting of 22 presentations with a triad of spiking temperatures greater than 38.5"C, chest pain, and cough (Table ). These two groups differed significantly with regard to maximum temperature recorded before the operation (37." * 0.2" versus 39." f 0.2"C; pz < 0.05) and sedimentation rate (53 * 9 versus 86 * 8 mm/min; p2 < 0.05). There were no differences between the two groups in the preoperative hemoglobin level, white blood cell count, or the percentage of patients on antibiotic prophylaxis. Roentgenographic Evaluation Three patterns of roentgenographic presentation were noted: focal unilateral disease, focal bilateral disease, and diffuse bilateral infiltrates (Figs -3). The focal infiltrates or abscesses were usually confined to one segment or a lobe, whereas diffuse disease presented as either infiltrative or reticular-nodular disease. Nineteen instances (6%) of focal, unilateral disease were recorded, 3 (9%) were focal and bilateral, and 9 cases (29%) were bilateral and diffuse. Eight of the 22 toxic patients had bilateral diffuse disease as opposed to one of the 9 nontoxic patients (pz < 0.05). Preoperative Fine Needle Aspirate or Bronchoscopic Evaluation Computed tomographic or fluoroscopic biopsy and bronchoscopy were used sparingly in this population of patients, primarily due to their age. Most recently, however, with the advent of smaller fiberoptic bronchoscopes, bronchoscopy was attempted more frequently for analysis of bronchoalveolar lavage. Fine needle aspirate was diag- Fig. Patient with chronic granulomatous disease and bilateral reticulonodular pattern, which proved to be Aspergillus. - A Fig 2. Bilateral focal disease in a patient with chronic granulomatous disease. Lingular resection revealed Pseudomonas cepacia.

3 846 POGREBNIAK ET AL Ann Thorac Surg 993;5584&9 Involved lung was cultured for bacteria, fungi, mycobacteria, LegioneZZa, and viruses. Histologic examination included hematoxylin and eosin, periodic acid-schiff, Ziehl-Neelsen, Giemsa, and methenamine silver stains. Fig 3. Unilateral focal disease. At operation chest wall involvement was managed with en bloc segmentectomy and chest wall resection. Microscopic examination revealed Wangiella dermatitidis. nostic in 2 of the 6 patients, whereas bronchoscopy jielded a microbiologic diagnosis in 5 of 5 attempts. Surgical Approach The three roentgenographic patterns guided intraoperative procedural decisions. An aggressive surgical approach was performed, if possible, in these patients both for diagnostic purposes, and, if possible, anatomical resection of the involved segment. In the cases where a formal resection was planned, single-lung ventilation was used, and bronchial blockade accomplished with Fogarty balloon catheterization (if standard double-lumen tubes were too large) preoperatively under bronchoscopic guidance. Children less than 5 years old who required formal resection (n = 6) had bronchial closure with interrupted nonabsorbable sutures. Other bronchial closures, as well as all wedge biopsies, were performed with the stapling device. Patients with diffuse bilateral disease, however, whether they were toxic or not, had lung biopsy only for diagnostic purposes. Of the patients with unilateral focal disease, all 6 with fluid collections had drainage only. In 9 patients a formal resection, either segmentectomy or lobectomy was performed, whereas in 4 of the patients with focal unilateral diseases a biopsy was performed for diffuse disease. All 3 patients with bilateral focal disease had formal resection. In these cases we thought that one infectious process was responsible for the entire clinical complex and if the most extensive site of disease could be eradicated, hopefully there would be added efficacy to the postoperative antibiotic therapy. Evaluation of Lung Samples Microbiologic examination was performed intraoperatively in selected cases as touch preparations and frozen sections. All specimens were evaluated by Gram stain, wet mount, fluorescent stain for acid-fast bacilli, toluidine blue stain, and Legionella direct fluorescent antibody stain. Results The median hospital time was days (range, 24 to 600 days). In 26 of 30 instances (87%) the patients were extubated within 24 hours of the operation. There was no mortality in the asymptomatic group of patients. In the patients with overt signs of septicemia (n = 22 instances) 6 formal resections, 5 drainages, and biopsies were performed. Four of these septic patients died in the hospital. One epinephrine-dependent patient, who had preoperative anuria from rhabdomyolysis and acidosis, died intraoperatively of septic shock due to Pseudomonas cepacia during attempted pneumonectomy. Another patient died 22 days after anterior segmental/ lingular resection for Pseudomonas cepacia. She had had a right lower lobectomy 3 months earlier for the same problem. Her course at the NIH was complicated by progressive adult respiratory distress syndrome and bacteremia with Pseudomonas cepacia. She died with a panlobar pneumonia. Two of the other patients died due to disseminated aspergillosis, one 46 days after open lung biopsy and the other 5 months after lower lobectomy. The latter patient had to be continually hospitalized for 20 months, requiring numerous operations for Aspergillus empyema and Aspergillus osteomyelitis. There were no postoperative bronchopleural fistulas, and 4 of the 3 cases were essentially complication free. Most patients were hospitalized for periods greater than month for antibiotic therapy and suffered medically related complications including adult respiratory distress syndrome (4), central line sepsis/thrombosis (4), Clostridia dificile infection (3),. reversible renal insufficiency (3), related hematologic problems (3), deep venous thrombosis and pulmonary embolus (3), and herpes zoster (). One patient required drainage of postoperative fungal empyema. He had Aspergillus invading the chest wall at the time of the original apical posterior segmentectomy, and reexploration, decortication, and antibiotics controlled this empyema. The patient presently remains without infection. Another patient suffered a phrenic nerve palsy, and 6 patients required multiple debridements of the superficial chest wound due to granuloma formation. These wound complications, characterized by heaped up, abundant hypercellular granulation tissue, usually occurred within 0 days of the operation and required 7 debridements under general anesthesia, with secondary closure in 4 of the 6 patients treated. All patients with wound problems had X-linked inheritance patterns. Histologic and Pathologic Results Eighteen of the 30 procedures with histologic specimens included an intraoperative frozen section. In all cases except, the frozen section diagnosis accurately repre-

4 Ann Thorac Surg 993;55:84&9 POGREBNIAK ET AL 847 Table 2. Microbioloxic Results Organisms Cultured Single species Aspergillus Pseudomonas cepacia Penicillium Wangiella Sarcinosporin Mycobacterium Serratia marcescens Nocardia Double species Pseudomonas aeruginosa and Aspergillus Mycobacterium and Aspergillus Triple species Klebsiella, Mycobacterium, and Aspergillus Paecilomyces, Penicillium, and Aspergill us No growth Number sented the type and extent of the inflammatory process present in the subsequent permanent sections. In a single specimen, the presence of a focal granulomatous process was missed by frozen section analysis due to sampling error. No microorganisms were identified in any of the frozen sections, although in ten of these specimens fungal elements were found on special stains of permanent sections. The pattern of inflammation present in the resected material varied to some extent, but a granulomatous component was present in each specimen. The patterns of granulomatous inflammation observed included a diffuse infiltration of epithelioid histiocytes as well as discrete well-formed granulomas with or without central caseation. Giant cell formation was present in the majority of specimens. A component of acute inflammation was present in the majority of specimens, usually intimately admixed with the granulomatous inflammation. In a third of the specimens pure abscess formation was observed. Although acute inflammation was present in every case in which bacteria were cultured from the tissue, a similar histology was present in many specimens that proved to be culture negative for bacteria. Abscess formation was observed in both bacterial and fungal infections. In 9 of the 30 specimens, microabscesses surrounded by palisading histiocytes were present. Microorganisms were found on special stains of permanent sections in 7 of 30 specimens. In each instance, the microorganism identified was a fungus. In general, fungal forms were rare, widely scattered, and often fragmented. Definitive speciation was not attempted, but if tubular septate fungal hyphae with acute-angled branching were present, a tentative diagnosis of Aspergillus infection was rendered. The microbiologic results are depicted in Table 2. The most common organisms cultured were Aspergillus spe- 6 cies and Pseudornonas cepacia. Four procedures revealed more than one organism. Of note were the unusual organisms encountered in some patients including Sarcinosporin and Wangiella. In only of the 30 specimens was a fungal organism grown in culture from tissue that did not reveal fungal elements by histopathology. In the nine specimens in which routine bacterial cultures were positive, and in the three specimens in which mycobacterial cultures were positive, these organisms were not identified on special stains of histologic sections. Postoperative Therapy A change in therapy was dictated by the results of the procedure in 23 of the 3 instances (74%) and chiefly included the institution of new antibiotics for unsuspected organisms. In the 6 cases where no growth was determined on the microbiologic cultures of the operative specimens, 4 patients had no change in therapy, had steroids added for granulomatous pneumonitis, and had cessation of antifungal coverage. All 6 of these patients recovered uneventfully. With the institution of proper antibiotics and other postoperative measures, the surviving patients' temperatures returned to normal (37.5"C) consistently, a mean of 8? days after the operative procedure. Comment Chronic granulomatous disease is an uncommon disorder. The frequency of pulmonary disease in this group of patients and its presentation with fungal pneumonia and unusual organisms commits the thoracic surgeon to be aware of this disease in the pediatric population. Unfortunately, however, as seen by the delay in diagnosis when compared with documentation of first infection, many children will have suffered potentially lethal septic complications necessitating long hospital admissions. As documented in this series, as well as others [0-6], fungal infections are the most frequent pathogen in the lung. In our series fungal organisms were responsible for the clinical course in 6 of 3 (52%), specifically Aspergillus species. Although not limited to the lung, CGDassociated Aspergillus infection most frequently presents as a pneumonia or an osteomyelitis, and is characterized by its indolent onset and slow progression in contrast to invasive aspergillosis in other immunocompromised patients. The radiologic pattern seen in our patients, either miliary-bilateral disease or localized pneumonitis, is common, as well as localized spread to the vertebral body or ribs. Pathologically, one sees granulomas with rare or abundant hyphae, and although the patient may be septic, vascular invasion is rare. Relapses are common after treatment with amphotericin B and in many cases 5-fluorocytocine or intraconazole must be added. A 26% mortality from Aspergillus has been reported in patients with CGD. Roback and associates [7] were first to propose an aggressive surgical policy in the management of these patients. In a report of 4 patients with CGD, 2 of the 4 patients who had unresolving pneumonic processes were

5 848 POGREBNIAK ET AL Ann Thorac Surg 993;55:844-9 salvaged with en bloc (ie, lobectomy) removal of the localized area and institution of the appropriate antibiotics. We have used an aggressive policy in the management of unresolving pulmonary disease since 983 for both diagnostic and therapeutic purposes. A report of the first 4 patients at the NIH with CGD in 983 revealed that no organisms were recovered in more than 55% of febrile episodes. Since that time, patients have been treated with prophylactic antibiotics [9]. In the present series 74% of the patients were being treated with trimethoprim and sulfamethoxazole before thoracic intervention. Overall, in the CGD population seen at NIH, prophylactic antibiotic administration has decreased the number of nonfungal infections and there has been no difference in the incidence of fungal infections when patients were receiving trimethoprim and sulfamethoxazole. The recent predominance of fungal infections has been due to the relative percent decrease in the number of nonfungal infections. In many instances, therefore, the study of thoracic disease and CGD becomes the management of the immunosuppressed child with undiagnosed pneumonitis, which could be fungal in origin. As detailed by the results of the biopsies, some of the infections found are extremely rarely described in humans. The institution of other treatments in an acutely ill patient chronically on prophylaxis then becomes problematic. To compound the issue, the pattern of presentation differs from patient to patient in that there appear to be two distinct groups of patients, one clearly toxic and the other with minimal symptoms. In both cases, however, rapid diagnosis of the underlying pulmonary problem is crucial to determine the most appropriate antimicrobial therapy. Semiinvasive techniques such as bronchoscopic biopsy, fine needle aspiration and biopsy, and bronchoalveolar lavage were attempted in selected cases, usually in older children, but had low yields. Hence, open techniques were required for more precise management. The radiographic pattern in CGD infiltrates partially dictated preoperatively the most efficient way to obtain the diagnosis. In patients with bilateral diffuse disease where anatomic resection could not remove the bulk of the disease, simple wedge excision was used and a diagnosis obtained in 0 of 3 instances (77%). Indeed, one could argue that in a select group of patients with diffuse disease the use of thoracoscopy may be justified to obtain a peripheral piece of tissue for diagnostic purposes. The most difficult intraoperative planning occurred in the situation of focal disease, where it was determined that the majority of the infiltrate could be anatomically excised. Some of these lobar or segmental infiltrates were very consolidated, which prohibited wedge biopsy due to prohibitive risk of bleeding, bronchopleural fistula, and possibly pleural contamination with fungal empyema. The aggressive resection approach, however, had to be tempered with the realization that these children would probably have numerous episodes of recurring pulmonary infiltrates emphasizing the need for lung preservation. This was reinforced in this series in that two-thirds of the patients with life-threatening infections required repeat thoracotomies. Moreover, general anesthesia in these children requires adequate pulmonary reserve in the future, for it may be necessary to surgically manage other manifestations of the disease including liver abscesses [8]. Fortunately, the majority of these patients had a dominant lesion that could be excised by segmentectomy. The lack of postoperative bronchopleural fistulas and the low rate of postoperative empyema validate the safety of this approach. Whether the resection of these parenchymally involved areas leads to more rapid recovery or allows the antibiotics to be more efficacious in these immune-compromised children is subject to speculation. Nevertheless, in our experience, it is extremely rare that the antibiotics alone completely eradicate the infection if it has caused considerable lung consolidation. There is no doubt, however, that formal resection in the overtly septic patient has a greater mortality. The contribution of operation to the mortality in these patients compared with the contribution of the overwhelming infection in these immunocompromised individuals, however, is difficult to sort out because of the 4 patients was in extremis at the time of his attempted resection and the others all had evidence of extrathoracic septicemia. Moreover, the only other in-hospital death occurred in a toxic patient who died 42 days after a wedge resection. The inability to "get ahead" of the sepsis is attested to by the inability to render these patients normothermic despite resection and antibiotics, whereas all other patients responded with normalization of their temperature after appropriate management. In addition, the extended hospitalizations for these patients are the consequence of postoperative antibiotic management and not a result of postresection complications. The only predictor of mortality in this series of patients appears to be toxic presentation as opposed to those who are asymptomatic or present only with a cough. Although it is stated in the literature that X-linked CGD is associated with more severe infection, we saw an equal distribution of deaths in our 4 patients, with 2 being X-linked and 2 being autosomal recessive. The postoperative management of these patients cannot be underestimated. Bronchoscopies are frequently required to ensure lobar expansion in these children and young adults, and the prolonged use of antibiotics and steroids may be necessary to diminish inflammatory responses associated with granuloma formation. In the setting of severe life-threatening illness, white cell transfusions may be used in some patients [4]. Recently interferon-gamma has been shown to decrease the frequency of infection of patients with chronic granulomatous disease either by restoring the ability of the leukocytes to produce hydrogen peroxide or through other systemic effects, which may contribute to host defense [6, 9, 20. Nevertheless, there is a cohort of patients who will escape the benefit of interferon therapy and will present with life-threatening infections. Therefore, an aggressive diagnostic and therapeutic approach to the management of nonresponding, undiagnosed, pulmonary infections should be taken in patients with CGD. This may require segmental resection of the localized infiltrate or wedge biopsy of the lung if the patients are not responding to

6 Ann Thorac Surg 993: POGREBNIAK ET AL 849 prophylactic antibiotics. Mortality will increase with overtly toxic patients; however, in the majority of cases the temperature will return to normal and the septic picture will resolve after limited resection and institution of the appropriate antibiotics. References. Berendes H, Bridges RA, Good RA. A fatal granulomatosis of childhood; the clinical study of a new syndrome. Minn Med 957;40: Johnston RB Jr, Baehner RL. Chronic granulomatous disease: correlation between pathogenesis and clinical findings. Pediatrics 97;4&73&5. 3. Johnston RB Jr, McMurray JS. Chronic familial granulomatosis. Report of five cases and review of the literature. Am J Dis Child 967; Gallin JI, Buescher ES, Seligmann BE, Nath J, Faither T, Katz P. Recent advances in chronic granulomatous disease. Ann Intern Med 983;99: Gallin JI. Phagocytic cells: disorders of function. In: Gallin JI, Goldstein IM, Snyderman R, eds. Inflammation: Basic Principles and Clinical Correlates. New York Raven, 988: Gallin JI, Malech HL. Update on chronic granulomatous diseases of childhood. Imrnunotherapy and potential for gene therapy. JAMA 990;263: Segal AW. The electron transport chain of the microbicidal oxidase of phagocytic cells and its involvement in the molecular pathology of chronic granulomatous disease. J Clin Invest 989;83: Weening RS, Kabel P, Pijman P. Continuous therapy with sulfamethoxazole-trirnethoprim in patients with chronic granulomatous disease. J Pediatr 983;03: Margolis DM, Melnick DA, Alling DW, Gallin JI. Trimethoprirn-sulfamethoxazole prophylaxis in the manage- ment of chronic granulomatous disease. J Infect Dis 990;62: 72M. 0. Mouy R, Fischer A, V ier E, Seger R, Griscelli C. Incidence, severity, and prevention of infections in chronic granulomatous disease. J Pediatr 989; Cohen MS, Isturiz RE, Malech HL, et al. Fungal infection in chronic granulomatous disease. Am J Med 98;7: Redmond A, Capre JJ, Biggart JD, Mackenzie WR. Aspergillosis involving bone. J Path Bact 965;89: Fems B, Jones C. Paraplegia due to aspergillosis. Successful conservative treatment of two cases. J Bone Joint Surg 985; Altman AR. Thoracic wall invasion secondary to pulmonary aspergillosis: a complication of chronic granulomatous disease of childhood. Am J Roentgen0 977; Kelly JK, Pinto AR, Whitelaw WA, Rorstad OP. Fatal Aspergillus pneumonia in chronic granulomatous disease. Am J Clin Pathol 986;86:23W. 6. Neijens HJ, Frenkel J, Keizer-Schrama S, Dzoljic-Dzoljic G, Jeradji M, van Dongen JIM. lnvasive Aspergillus infection in chronic granulomatous disease: treatment with itraconazole. J Pediatr 98;5: Roback SA, Weintraub WH, Good RA, et al. Chronic granulomatous disease of childhood: surgical considerations. J Pediatr Surg 97;6: Wall RT, Buzzanell CA, Epstein TA, et al. Anesthetic considerations in patients with chronic granulomatous disease. J Clin Anesth 990;230&. 9. Abramson SL, Lomax KJ, Malech HL, Gallin JI. Recombinant human interferon-gamma (rifn-gamma) and interleukin-4 (ril-4) regulate gene expression of several phagocyte oxidase components [Abstract]. Clin Res 990;38:236A. 20. The International Chronic Granulomatous Disease Cooperative Study Group. A controlled trial of interferon gamma to prevent infection in chronic granulomatous disease. N Engl J Med 99;324:509-6.