Aspergillus is a ubiquitous soil-dwelling organism found in organic debris, water, dust, compost, foods, spices, unfiltered air, ventilation systems, horizontal surfaces, and ornamental or rotted plants (Walsh 1989, Anaissie, 2002). Despite the existence of more than 100 species of Aspergillus, only a few are known to be pathogenic for humans. Aspergillus fumigatus, Aspergillus flavus, and Aspergillus niger are the most common mould pathogens of humans, causing both life-threatening invasive diseases in immunocompromised patients and allergic diseases in patients with atopic immune systems. Other species, like Aspergillus terreus, Aspergillus clavatus, Aspergillus niveus, and Aspergillus nidulans, have rarely been reported to cause disease in humans (Stevens, 2000, Soubani, 2002). The frequency and relative importance of these infections are on the rise in all developed countries, which is most probably related to increased numbers of immunocompromised patients. These increased numbers are because of improved survival from AIDS, more intensive cytotoxic therapy for malignancies, more transplantation for organ dysfunctions, and better therapy and prophylaxis for candidal infections (Stevens, 2000, Marr, 2002). About 4% of patients dying in modern European teaching hospitals have invasive aspergillosis (Groll, 1996, Vogeser, 1997) and it is the leading infectious cause of death in leukaemia and bone marrow transplant patients (Pannuti, 1991).
Aspergillus, like other filamentous fungi, is primarily acquired from an inanimate reservoir, usually by the inhalation of airborne spores. The organism grows best at 37°C, and the small spores (2 to 3 micrometres) are easily inhaled and deposited deep in the lungs, leading to a variety of clinical syndromes. Nosocomial transmission of Aspergillus infection occurs mostly during or after hospital construction or renovation, and usually in severely immunocompromised patients (Lutz, 2003). Although the lung is the most frequent site of disease, on rare occasion post-operative invasive Aspergillus infections have been described; as shown below, most of these infections involved immunocompetent patients. The purpose of this article is to summarize important aspects related to the clinical presentation, diagnosis, treatment and prophylaxis of post-operative aspergillosis.
Primary cutaneous aspergillosis
Myers and Dunn reported in 1930 the case of a 37 year-old farmer who presented with an indolent ulcerative lesion on the back of the right hand of two years’ duration. Two years before, the patient had developed a ‘pimple’ on his right thumb, which was lanced and which discharged pus. He attributed his trouble to husking corn before the ‘pimple’ healed. The lesion presented itself as an ulcer with a crater about 15 mm in diameter with an elevated indurated purplish border from 7 to 9 mm wide. Minute pustules were visible along the edge of the ulcer from which thick creamy pus could be expressed. Only Staphylococcus was obtained on blood agar cultures made from this pus. When the pus from the lesion was planted on Sabouraud’s medium, an organism identified as Aspergillus terreus was observed. Similar lesions were produced by subcutaneous inoculation of the organisms in laboratory animals. Rapid healing of the hand lesion occurred after the application of 1 per cent copper sulphate solution alternating with 1 per cent copper subacetate ointment.
Other reports of primary cutaneous aspergillosis have been rarely published, mainly affecting immunocompromised patients. These infections are frequently characterized by necrotizing ulcerations, usually near an area of local skin trauma resulting from intravenous access, arm boards, burns or tape (Frank, 1933, Bruck, 1971, Nash, 1971, Caro, 1973, Salisbury, 1974, Prystowsky, 1976, Charlie, 1978, Stone, 1979, Estes, 1980, Langlois, 1980, Allo, 1987). Lai et al (Lai, 1993) reported the case of a cutaneous infection caused by A. flavus that primarily invaded the graft skin and its underlying free muscle flap in a 60 year-old insulin-dependent diabetic female. This patient was effectively treated with the combination of oral keroconazole, topical ketoconazole ointment and surgical debridement.
Post-operative/traumatic aspergillosis: the first reports
Frank and Alton (Frank, 1933) seems to have been the first to describe post-operative wound infection caused by AspergiIlus spp. They reported in 1933 the case of a 40 year-old woman who was operated for abdominal tumour and had a normal post-operative course. At the end of about sixteen days, the dressing was removed and revealed the appearance of being studded with black powder. A 3.7 cm ulcer was observed at the end of the wound, with sharply defined edges, as well as pustules varying in size in the skin under the gauze dressings covering the wound. The infection did not extend beyond the gauze pad used as dressing, and there were no systemic manifestations of the infection. Following exposure of the wound field to air and washing with iodine and alcohol, the apparent ulcers and pustules had entirely disappeared by the next morning. The surgical dressings covered with a dark powder were sent to the laboratory and revealed the growth of Aspergillus niger. The authors speculated that the skin could have been contaminated by this organism following the preparation of the patients’ skin in the operating room. They also commented about two previous similar cases in their medical centre, each of them exhibiting similar pustulation confined absolutely to the area of the gauze dressing and healing rapidly after exposure and painting with 3 per cent tincture of iodine followed by alcohol and dressing.
In 1988, Gordon et al reported a case of an otherwise healthy 11 year-old boy that developed a mixed fungal infection (Absidia, Rhizopus, and Aspergillus species) of his abdominal wall. The child was trapped in a silage machine for nearly 2h and suffered massive contamination and extremity lacerations with vegetable debris. The patient recovered after a long hospitalization with eventual debridement of almost his entire abdominal wall and multiple treatments with hyperbaric oxygen.
Sawyer et al reported in 1992 the case of a 79 year-old immunocompetent male who suffered blunt abdominal trauma and a ruptured duodenum while cutting down a tree. On the sixth post-operative day, an area of necrosis and slough was noted in the subcutaneous portion of the superior aspect of his abdominal wound; this was locally debrided and sent for culture. Culture results revealed Staphylococcus aureus and Aspergillus flavus. Despite treatment with systemic antibiotics and amphotericin B, the patient died from sepsis. A role of the A. flavuswound sepsis in his demise was not clear, though the infection was poorly controlled at the time of death. Given the environmental niche of the Aspergillus species, the authors supposed that direct inoculation from a piece of decaying log could have been the source of the patient’s contamination.
Carlson et al reported in 1996 the first case of infection of multiple abdominal viscera due toAspergillus fumigatus in a 37 year-old female patient with a laparostomy for Crohn’s disease. This patient was effectively treated with a 4-week course of amphotericin B. Sampling of air from the patient’s environment yielded one isolate of A. fumigatus that matched the patient’s isolates, suggesting that this patient acquired her infection from air in the intensive care unit. The authors concluded that it seems likely that a particularly high risk of acquisition ofAspergillus infection may occur in patients with laparostomies.
Lutz et al (Lutz, 2003) reported an outbreak of Aspergillus infection involving 6 patients at a tertiary care hospital during a 2-year period. All cases clustered in the operating theatre during a single 12-day period, and the investigation identified mold contamination in an operating theatre air-handling system. This outbreak was due to contamination of insulation in variable airflow volume (VAV) units that had deteriorated after becoming wet. These units were located downstream of final filters, and, therefore, conidia released from mold growing on insulation were not filtered out before entering the operating room. The source of this outbreak was identified through use of a closed-space video camera that allowed visualization of otherwise inaccessible ductwork.
Aspergillosis and cardiovascular surgery
1. Endocarditis
Aspergillus as the infectious agent in endocarditis or aortitis following open-heart surgery was first reported by Hadorn in 1960, who described the case of a 24 year-old male who developed aortitis with aortic rupture due to Aspergillus infection of the aorta following removal of a congenital subaortic valvular stenosis. Numerous blood cultures were negative, and the diagnosis was finally established by histopatological examinations of large emboli removed from the iliac arteries and by the growth of Aspergillus fumigatus in culture. The author concluded that the association of the mycotic vegetations with the aortotomy sutures, combined with the absence of bronchopulmonary lesions, indicated contamination of the aorta with A. fumigatus introduced at the time of operation.
In 1964, Newman and Cordell reported a case of Aspergillus endocarditis following mitral valvulectomy in a 24 year-old male with rheumatic mitral stenosis three years post-valvulotomy. Numerous blood cultures were negative, but one blood culture taken before the patients’ death finally revealed A. fumigatus after the patient died. Again, the authors believed that the organism was introduced either during operation or during the post-operative diagnostic or therapeutic procedures.
Amoury and associates (Amoury, 1966) in their report on bacterial endocarditis complicating intracardiac prostheses, described 2 patients with Staphylococcal endocarditis who had superimposed Aspergillus endocarditis. In 1967, Leffert and Hackett described the case of a 26 year-old male who had coarctation of the aorta surgically corrected 15 years before and was admitted due to severe aortic insufficiency. This patient was submitted to heart surgery for valve replacement, and he died of sepsis forty-four days after surgery. Necropsy findings revealed the presence of thrombotic material attached to the aorta composed of fungal elements typical of Aspergillus. Comparing to the previews reports, Leffert and Hackett described some common factors: (1) young patients without evidence of prior or concomitant bronchopulmonary aspergillosis; (2) antibiotic therapy just prior to or immediately after operation; (3) post-operative course consistent with bacterial endocarditis, but with negative blood cultures; (4) large vegetations of Aspergillus associated with sutures placed through the endocardium or aorta; and (5) extensive local necrosis and invasiveness of fungi with propensity to embolization late in the disease.
Gage et al. reported in 1970 three cases of A. fumigatus and one case of A. glaucusendocarditis following cardiac surgery. As the infections occurred in a three-year period, attention was focused on the operating room, its personnel and equipment, operative techniques, traffic, and ventilation. Aspergillus was found in an occasional swabbing from room surfaces in both areas and a culture plate exposed to the air in the operating room. Outside of the hospital, Aspergillus was identified in two areas in close relation to the operating room.Aspergillus fumigatus was found in the pigeon excreta on the ledges of the windows. In addition, the external roof of the operating room, two stories high and frequently covered by water, had a luxuriant growth of moss from which A. fumigatus was cultured. In the following years, other cases of endocarditis caused by Aspergillus species following open-heart surgery were reported (Kammer, 1974, Petheram, 1976, Drexler, 1980).
The prevalence of Aspergillus endocarditis seems to be increasing in the hospital population (Ellis, 2001). Several conditions predispose to Aspergillus endocarditis including underlying cardiac abnormalities, prosthetic heart valves, indwelling central venous catheters, prolonged use of broad-spectrum antibiotics, and intravenous drug use (Pierotti, 2002, El-Hamamsy, 2004). Identifying the source, establishing the diagnosis, and treating Aspergillus endocarditis remain highly challenging and is often met with little success. Three cases of Aspergillus aortic valve endocarditis in immunocompetent patients who had undergone previous aortic valve replacement were recently described by El-Hamamsy et al (El-Hamamsy, 2004). Despite attempts at combined medical and surgical therapy in this study, mortality was 100%. The origin of the infections was probably related to intraoperative seeding by airborne spores, emphasizing the importance of operating room sterility and the use of high-efficiency particulate air filters. Although no tangible sources of Aspergillus were found, the final report suggested reinforcing sterility rules and procedures by restricting personnel movements in and around the operating room, strict use of surgical scrubs in the vicinity of the operating room environment, and limiting door opening during the actual operations. No case ofAspergillus endocarditis was observed in their medical center since these rules have been imposed. The authors addressed the importance of considering the diagnosis of Aspergillusendocarditis in those patients presenting with culture-negative endocarditis after valve replacement.
2. Aortitis
The first report of aortitis caused by Aspergillus in the English literature seems to have been described by Malcolm et al. in 1971. They reported the case of a 34 year-old man who had undergone aortic valvotomy and subsequently developed leg pains, migratory arthralgias, periarticular swelling, and general malaise. Mild intermittent pyrexia, evanescent petechiae, splinter haemorrhages, and peripheral small artery occlusion characterized the early course in hospital. Dramatic popliteal artery occlusion led to surgical recovery of embolic material packed with mycelia of Aspergillus flavus, but the patient died despite amphotericin B therapy. Necropsy revealed aortitis and hyphae morphologically consistent with Aspergilli were demonstrated in the wall of a saccular dilatation of the ascending aorta close to non-absorbable sutures.
Sanchez-Recalde et al (2003) described eight cases of Aspergillus infection of the ascending aorta after cardiopulmonary bypass surgeries during a 25-year period. All patients were male, immunocompetent, and presented with prolonged fever. Initial symptoms appeared between the immediate post-operative period and up to two years after surgery. Ante-mortem diagnosis was established in only three patients for whom transthoracic echocardiography was suggestive of aortic pseudoaneurysm and was confirmed by thoracic computed tomography or aortography. All patients had negative peripheral blood cultures. The authors suggested that the surgical trauma such as aortotomy may have damaged the aortic wall, which could have been contaminated by airborne fungal spores. As a consequence of an inflammatory process, the aortic wall could have been disintegrated and weakened by the subsequent formation of an aneurysm. All patients except one died.
3. Infection of vascular prosthesis
Glotzbach (1982) seems to have been the first to describe Aspergillus infection of a vascular prosthesis, in 1982. He reported the case of a 71 year-old male patient who developed A. terreus infection of pseudoaneurysms and contiguous vertebral osteomyelitis. The patient died despite treatment with amphotericin B, and the exact origin of the A. terreus in this patient was unclear. In 1985, Brandt et al reported the case of a 73 year-old immunocompetent male with a mycotic pseudoaneurysm of an aortic bypass graft and contiguous vertebral osteomyelitis due to A. fumigatus. The patient was successfully treated with resection of the infected graft and administration of amphotericin B in a total dose of 2 g. Similar to the previous case described by Glotzbach (1982), this patient presented with back pain approximately two years after placement of the vascular graft. As this patient had no pulmonary disease, the authors speculated that the infection of the aortic graft or suture line could have occurred intra-operatively as a result of inoculation with airborne fungal spores.
Among fourteen cases of graft fungal infections reviewed by Doscher et al (1978) from 1966 to 1986, only four were aortic bypass infections due to Aspergillus. Aguado et al reported in 1992 the case of a 65 year-old male with aortobifemoral graft infection caused by A. fumigatus who died despite treatment with amphotericin B. In the same paper, the authors reviewed previous published cases of Aspergillus infections of thoracic or abdominal aortic grafts (Echene, 1980, Liesegang, 1980, Glotzbach, 1982, Anderson, 1984, Hargrove, 1984, Nussaume, 1990). Mean age of the eight patients included in the review was 54 years old, and there were no obvious differences in the clinical presentation of fungal and bacterial aortic graft infections. The site of the aortic graft was aortobifemoral in six cases and thoracic in the remaining two. Three of these eight patients had also pulmonary tuberculosis. Four patients developed peripheral embolic phenomena involving different organs and extremities. Five of six patients with abdominal aortic grafts infected by Aspergillus developed an aortic pseudoaneurysm. The diagnosis was made on the basis of results of a biopsy of the contiguously affected vertebral disk space (in three cases) and of findings of culture of the excised aortic graft or the peripheral arterial embolus (in five cases). In two cases the diagnosis of aortitis due to Aspergillus was made of the basis of histological data and was not confirmed by culture. The infecting organism was known in the remaining cases: A. fumigatusin five cases and A. terreus in one case. Aspergillus aortitis usually occurred on the suture line of a previous aortotomy. Amphotericin B was administered in seven cases for a total dose that ranged from 400 mg to 2 g in a period of 2-6 weeks; monotherapy with 5-fluocytosine was used in one patient. Infected aortic grafts were excised from all the patients who had intra-abdominal infections, while patients with thoracic infection had a thoracic aortic replacement with rerouting of the blood flow.
Collazos et al (2001) found 13 cases in 2001 published about prosthetic vascular graft infection due to Aspergillus species (Echene,1980, Lifschultz, 1982, Glotzbach, 1982, Anderson, 1984, Hargrove, 1984, Brandt, 1985, Nussaume, 1990, Motte, 1993, Sanchez Lastres, 1993, Wandschneider W, 1995, Garcia Fuster, 1999, Collazos, 2001). All patients were male and none was immunosuppressed. This finding indicated that the grafts were contaminated at the time of surgery, since invasive and generalized Aspergillus spp. infections are uncommon in immunocompetent patients, and previous foci of infection were not identified in the cases reported. The median age was 60.5 years, and the median time from the placement of the graft to diagnosis was 8 months, a period appreciably longer than that observed in Candida spp. graft infections, which is usually less than 6 weeks (Doscher, 1987). Fever was absent in about one-half of the cases, and blood cultures were positive in only two of the eight patients for whom this information was reported. Aspergillus was identified in the graft in all cases. As in Candida infections, (Doscher, 1987) the aortic location at either the thoracic or the abdominal level was the most common site of involvement. Optimal treatment for this condition consisted of a combination of surgery and antifungal agents. In situ replacements or no excision of the infected graft was associated with recurrence of the infection and death in all cases, whereas all patients who underwent extra-anatomic bypass through a clean field survived (Collazos, 2001).
Marroni et al (2001) reported the case of a 64 year-old male who died of septic shock 10 days after surgery for removal of a previous Dacron graft infected by A. flavus. The authors’ conclusions were that there were no obvious differences in the clinical presentation of fungal and bacterial infections, particularly if the clinical presentation of Staphylococcus epidermidis(long incubation time, mild systemic symptoms, and absence of fever) is considered.
4. Sternal wound infections
Although cases of Aspergillus sternal wound infections following intravenous drug abuse have been published, (Ahmad, 1981, Walker, 1991) most of the cases have been reported following sternotomy (Attah, 1979, Wellens, 1982, Richet, 1992, Vandecasteele, 2002). Barzaghi et al (1994) reported two cases of sternal osteomyelitis due to A. fumigatus in non-immunosuppressed patients after cardiac surgery. The authors speculated that the source of the infection could have been related to direct contamination during repeated surgical procedures, facilitated by transient decrease in immune defence during cardiopulmonary bypass. In both cases, itraconazole was administered without adverse effects. Although pharmacological treatment, together with appropriate surgery, was followed by wound healing and microbiologic eradication in the first case, the efficacy of the treatment could not be assessed in the second patient, who died of bacterial superinfection before itraconazole therapy discontinuation.
Richet et al documented in 1992 that six patients who had underwent open heart surgery during a 21-month period developed A. fumigatus sternal wound infection. To assess potential risk factors associated with these infections, they performed a case-control study; at multivariate analysis, chronic lung disease was the only independent risk factor and the best predictor of A. fumigatus sternal wound infection. The mean interval between the surgical procedure and the onset of A. fumigatus sternal wound infection was 56.3 days. All cases had clinical evidence of severe sternal osteomyelitis; all patients required sternectomy, reconstructive surgery, and long term amphotericin B therapy. No patient died. Despite an extensive investigation, no environmental source for A. fumigatus was identified. A. fumigatus, however, grew from the bronchial washing of one patient at the time the sternal wound infection was diagnosed, and a prospective study showed that the rate of A. fumigatuscolonization among open heart surgery patients was the same as the rate of sternal wound infections caused by A. fumigatus. These data suggested that patients with chronic lung diseases and respiratory colonization with A. fumigatus may be at increased risk for A. fumigatus sternal wound infections after open heart surgery.
Vandecasteele et al (2002) reported the clinical data for 9 patients affected during an outbreak of Aspergillus flavus sternal wound infections after cardiac surgery in a Belgium secondary care hospital. In contrast with bacterial infections, which usually occurs early, the median time for A. flavus sternal wound infections was 14 days after surgery, and the onset of infection was insidious in 8 of the 9 cases, with only moderate systemic involvement. However, local infection tended to be destructive and refractory to therapy, and relapse was quite common (3 of 9 patients had relapses). In 2 patients, the infection evolved to septic shock and death. The authors concluded that although rare, Aspergillus infection must be considered in the differential diagnosis of slowly progressive but destructive wound infections, culture-negative pleural effusion, and culture-negative mediastinitis that occurs after cardiac surgery. Treatment requires a combined surgical and antifungal approach.
Heinemann et al (2004) reported the results of an extensive environmental sampling performed after the outbreak of A. flavus sternal surgical site infection after cardiac surgery, described above. 118 Results showed massive contamination by A. flavus: more than 100 cfu per contact plate were frequently observed in some areas of the surgical ward. Strains from patients and from the hospital environment selected all over the surgical ward were typed by random amplification of polymorphic DNA (RAPD), using two different primers (ERIC-1, BG-2). All these strains showed the same genotype, proving the clonal single-source of the environmental contamination and the intra-operative acquisition of A. flavus in the sternal surgical site infection outbreak. These data highlighted the importance of molecular typing to prove nosocomial acquisition of A. flavus infections. This outbreak was controlled after extensive environmental sampling, the elimination of the source of the spores, the decontamination of the premises and the improvement of the hygiene precautions in the surgical ward.
5. Post-operative mediastinitis
Post-operative mediastinitis caused by Aspergillus seems to have been first published in a series of surgical sternal wound infections in 1990 (with no additional details) (Jeevanandam, 1990). The first well documented case of A. fumigatus mediastinitis was reported by Byl et al. in 1993, in a 51 year-old heart transplant recipient male who developed inflamed sternal wound in the third week after transplantation. Cultures obtained from the wound yielded A. fumigatus in heavy growth, and the patient was submitted to superficial debridement of the sternal wound. Because the patient remained well, no systemic antifungal therapy was added. However, he had fever and suddenly died 48 days after transplantation. Post-mortem examination revealed anterior mediastinitis with a ruptured pseudoaneurysm of the ascending aorta. No other site of Aspergillus infection was found. The authors concluded that the intraoperative seeding of the surgical field by Aspergillus seems the most probable route of infection.
Bronchial infections
In 1960, Kelemen and Novak reported the occurrence of bronchial stump aspergillosis after the resection of a tuberculous lung, finding the ascospore of Aspergillus nidulans around the silk thread. In 1961, Franke and associates reported that 6 cases of bronchial stump aspergillosis were discovered among 51 cases in which silk thread was used but no such cases were encountered among 104 in which stainless steel wire was used as the suture material. Sawasaki et al (1969) reported 10 cases of bronchial aspergillosis, 9 of which were a secondary infection to the bronchial stump after pulmonary resection in patients with pulmonary tuberculosis. The period from the operation to the onset of the disease ranged from 6 to 12 months, and most of cases were caused by A. oryzae, followed by A. tamarii. The treatment consisted of the following: the bronchoscopic removal of the silk thread and the putrid mass of hyphae, the bronchoscopes application of germicidal and antifungal drug (0.02 per cent solution of phenylacetic quicksilver) or the inhalation, the ingestion or the intravenous administration of an antifungal drug. The authors performed an experimental study to test their hypothesis, and concluded that silk thread as a suture material played an important role in causing bronchial stump aspergillosis. The local reaction, often developing in the ulceration around the silk thread and lasting several months, may have induced local infection, especially of the pathogenic fungi. The bronchial change to pulmonary tuberculosis, although slight, may have somewhat facilitated these infections and the gluten in the silk thread became an appropriate medium for the fungus.
Nunley et al (2002) described an association between saprophytic fungal infection of the bronchial anastomosis and the occurrence of serious airway complications such as fatal haemorrhage, symptomatic bronchial stenosis and/or bronchomalacia following lung transplantation. Saprophytic fungal organisms are easily airborne and thus have ready access to the airway lumen and the relatively ischemic bronchial anastomosis. As saprophytes are organisms that obtain their nourishment from nonliving organic matter, the ischemic and necrotic airway debris at the bronchial anastomosis provides a fertile environment for their proliferation. Other authors have also reported an increase risk of complications including haemorrhage, tracheobronchitis, and pneumonia as consequences of post-transplantation saprophytic fungal infections (Kramer, 1991, Bertocchi, 1995, Guillemain, 1995, Patterson, 1999).
Breast surgery
Williams et al (1983) reported the case of 39 year-old previously health female with A. nigerfungal colonization associated with bilateral inflatable silicone breast implants. The authors speculated that the presence of bilateral Aspergillus colonization in this patient may have represented a truly rare occurrence of massive operative contamination. Regardless of the source of contamination, the authors commented that the fungi were able to colonize within the implant and capsular space without significant clinical signs. In vitro studies have demonstrated that many species of Aspergillus are inhibited by physiologic concentrations of oestradiol. This may explain the predominance of Aspergillus in male patients with prosthetic heart valve implants (Kramer, 1974, Echene, 1980).
Orthopedic surgery
Tack et al (1982) reported in 1982 the occurrence of four patients with Aspergillusosteomyelitis over a one-year period. Two of their patients had been submitted to laminectomy, one to left hip replacement and one patient had idiopathic thrombocytopenic purpura and was in treatment with steroids. Two of these cases were related to A. fumigatus, one to A. flavus and other to A. niger. In the same paper, the authors described fifteen other cases of Aspergillus osteomyelitis that had been published in the medical literature (Cawley, 1947,Tobler, 1954, Shaw, 1963, Redmond, 1965, Wolfson, 1969, Seres, 1972, Fujak, 1974, Grossman, 1975, Prystowsky, 1976, Simpson, 1977, Seligsohn, 1977, Casscells, 1978, Ingwer, 1978, Attah, 1979, Roselle, 1979, Langlois, 1980). Most of these patients were immunocompromised patients. Among children, chronic granulomatous disease was most common. Among adults, immunosuppressive drug therapy was the most important predisposing condition. Although the presence of a prosthetic device or surgical wound occurred in four of these fifteen patients, most of the cases of Aspergillus osteomyelitis were not related to previous surgery, but resulted from contiguous spread from a pulmonary infection, contiguous skin infection and haematogenous spread. Most of the patients were treated with amphotericin B, some of them combined with 5-fluocytosine and surgical debridement. With the exception of one patient, who died from staphylococcal septicaemia, all deaths were directly related to Aspergillus infection.
Mawk et al (1983) described three cases of Aspergillus infections of the lumbar vertebral interspace. Two patients responded to vigorous surgical therapy in combination with prolonged administration of antifungal agents, but the third patient died of complications related to the antibiotic therapy. In 1991, Peters-Christodoulou et al (1991) reported the case of 46 year-old male presented with A. fumigatus spondylodiscitis 2.5 months after disc surgery at L4-L5 for a herniated nucleous pulposus. The patient was successfully treated with itraconazole in combination with surgical debridement of the disc space. Although the authors speculated that the infection could have been caused by direct inoculation of the disc space during surgery, surveillance cultures of the air-conditioning system of the operating room revealed no aspergilli, and no construction works were carried out in the vicinity of the operating theatre at the time the disc surgery was performed.
Neurosurgery
Involvement of the central nervous system in aspergillosis may occur with haematogenous dissemination from a remote extracranial site, usually the lungs, and direct extension from nasal or paranasal foci along the invaded bone structures of the skull base and the vascular anastomosis by which the paranasal sinuses communicate with the cavity (Morioka, 1990, Siddiqui, 2004).
Cerebral aspergillosis after a neurosurgical operation is a rare disease. Visudhiphan et al(1973) reported in 1973 two cases of cerebral aspergillosis following neurosurgery. The first patient was a 13 year-old girl submitted to craniotomy for the treatment of a brain abscess, from which no organisms were seen in the histological section. This patient received a longer course of steroids to treat uncontrolled increased intracranial pressure, and four months later she suddenly developed high fever, headache, vomiting, and lethargy, which rapidly lead to death. Culture of the cerebrospinal fluid taken before death revealed Aspergillus fumigatus, and post-mortem examination revealed a 5 cm brain abscess underneath the craniotomy site.Aspergillus fumigatus was identified in this abscess, and examination of the middle ears and sinuses disclosed no abnormalities. The second patient was a 13 year-old boy submitted to surgical treatment of a craniopharyngioma. The post-operative course was complicated with diabetes insipidus, meningitis and persistent fever, with several cultures negative for bacteria and fungi. In spite of empirical treatment with antibiotics and amphotericin B for 5 weeks, the patient died 6 weeks after surgery. Autopsy examination revealed a large aneurysm of the upper end of the basilar artery, which has ruptured and probably caused death. Microscopic examination revealed that the aneurysm was mycotic in nature; the numerous hyphae were identified as Aspergillus, and the fungi were not identified elsewhere in the body. The origin of these infections was not elucidated.
Shapiro and Tabaddor (1975) reported the case of a 36 year-old female who presented with a three-week history of bizarre behaviour, lethargy and a mild right hemiparesis. She became stuporous and required external ventricular drainage. Contrast ventriculography demonstrated hydrocephalus and a left thalamic tumour encroaching upon the lateral and third ventricles. Following ventriculo-peritoneal shunting, radiotherapy was started. Two months later the focal symptoms progressed and a craniotomy was performed. The thalamic mass proved to be a glioblastoma multiforme. She died from progressive growth of this tumour six weeks after craniotomy, and autopsy showed two granulomas containingAspergillus organisms in the frontal and parietal lobes. These lesions were entirely unexpected post-mortem findings; the authors speculated that their location in the tracts of shunt apparatus can be related to the introduction of the organism with the shunt tubing or colonisation of these foreign bodies by blood borne fungi occurring in a debilitated patient. Two years later, Feely et al (1977) reported two cases of Aspergillus infection complicating trans-sphenoidal yttrium-90 (Y90) implant for ablation of the pituitary gland. It is possible that these infections could have been originated from the sphenoid sinus, facilitated by the presence of foreign material. The patient who survived had the screw removed and received prolonged treatment with amphotericin B.
In 1978, Galassi et al reported the case of a 59 year-old woman who had diabetes and was submitted to intracranial surgery for the treatment of a meningioma. About 1 year after the surgery, the patient’s condition deteriorated rapidly; at admission, she was stuporous, with right hemiparesis and aphasia. A second operation was performed; when the very thick dura was opened, four nodular formations were found extending from the internal surface of the dura into the Sylvian fissure. Pathological findings of these abscesses revealed multiple granuloma with septate mycotic hyphae and some hyaline spores, and culture of the cerebrospinal fluid revealed A. fumigatus. The patient died of diffuse meninoencephalitis about 3 months after this second operation. Since their patient lacked radiological or post-mortem extracerebral involvement, the authors supposed that direct inoculation of Aspergillus through the ethmoid sinus during previous neurosurgical procedure may have occurred.
Mielke et al (1981) reported the case of a 58 year-old woman patient with acromegaly who underwent operation for resection of a tumour in the sella turcica. Ten months after neurosurgery, the patient insidiously developed a mild proptosis without chemosis, followed by the sudden onset of blindness and ophthalmoplegia. Left carotid angiography showed a small aneurysm pointing inferiorly in the cavernous sinus (this was in the region of the tumour spread through the dura mater). The patient suddenly died, and post-mortem studies revealed chronic meningitis with demonstrable fungal hyphae, shown by mycological cultures to be Candida albicans and Aspergillus fumigatus. Extensive inflammatory granulation tissue with microabscess formation containing mycotic hyphae was present throughout the cavernous and sphenoid sinuses, and there was no evidence of mycotic infection within the thoracic or abdominal viscera. The authors reviewed the literature on such aneurysms at that time.
Letscher et al (1997) first reported the case of post-operative intracranial epidural A. fumigatus abscess in a 20 year-old male victim of trauma, who suffered from an orbit roof fracture and a frontal sinus posterior wall fracture. This patient was aggressively treated with combined surgical and medical approach, including systemic itraconazole use and instillation of amphotericin B into the sinus cavity. The source of A. fumigatus in this case was unknown – possibilities included contamination of the frontal sinuses before the trauma, spread from the nasal cavity through the ethmoidal fracture, nosocomial contamination during previous surgical procedure, and contamination via the frontonasal duct which remained permeable after cranialization of the sinus.
Viriyavejakul et al (1999) presented the case of a 10 year-old girl who developed cerebral aspergillosis presenting with massive intracerebral haemorrhage three weeks after neurosurgery for craniopharyngioma resection. Craniotomy was again performed, and the patient died two days after surgery. At autopsy, the middle cerebral artery at the ruptured site showed fungal colonies invading the artery. The morphology of the organism was highly suggestive of Aspergillus spp.; culture was not performed. The histological findings of the heart and lungs were not remarkable.
Piotrowski et al (1990) reported the case of a 40 year-old woman who died of massive intraventricular haemorrhage three weeks after neurosurgery for the treatment of a subarachnoid haemorrhage. In the immediate neighbourhood of the clipped aneurysm,Aspergillus abscess developed and the wall of the left cerebral posterior artery was also infiltrated by fungi. The fungal infection then spread over the basal meninges and invaded the right vertebral artery, causing the fungal aneurysm that led to the final and fatal bleed. As the abscesses were found mainly around the operative site, and as there was no other source for a mycotic infection discovered in the patient, one must assume that the fungal infection occurred post-operatively. Most likely the prolonged preoperative use of antibiotics, in combination with steroid therapy, facilitated the propagation of the Aspergillus infection.
The first case of Aspergillus infection into the cavity of a chronic subdural haematoma after burr-hole surgery seems to have been published by Morioka et al (1990). They reported the case of an 83 year-old man with no history of diabetes mellitus or other immune disorder who underwent burr-hole evacuation of a bilateral chronic subdural haematoma. About three years after surgery, the patient presented with dysarthria and weakness of the left upper limb. Focal motor seizures involving the left side of the face occurred. A computed tomography scan showed a contrast-enhancing nodular lesion in the right frontal lobe extending over the subdural space. A new craniotomy was performed, and the intracerebral mass was completely excised. Histopathological studies showed a granuloma with abscess cavities, and culture of the surgical specimen revealed Aspergillus fumigatus. After removal of the mass, the cavity was repeatedly irrigated with saline containing amphotericin B, associated with oral flucytosine. There was no evidence of disease in any organ that might have been responsible for secondary infection. This favourable outcome in this patient was associated to the complete excision of the lesion and to the absence of immunosupressive conditions.
Other authors have also reported cases of cerebral aspergillosis following neurosurgery (Lihara, 1990, Takeshita 1992, Sharma 1992, Guiot, 1993) published an unusual case of aspergillotic aneurysm that resulted from post-operative meningitis without any other recognizable predisposing factors. Darras-Joy et al (1996) reported three cases of nosocomial cerebral aspergillosis that occurred after neurosurgery, two of which survived after exhaustive surgical treatment and medical treatment with high doses of amphotericin B (one liposomal), flucytosine and itraconazole. It has been suggested (Sanchez, 1995) that therapy with high-dose itraconazole may be an attractive strategy for the treatment of cerebral aspergillosis.
Ophthalmological surgery
Ocular aspergillosis usually presents as keratitis and rarely as endophthalmitis. Of allAspergillus species, isolated post-operative A. niger endophthalmitis is an unusual occurrence after cataract surgery (Pettit, 1980, Kermani, 2000, Brar, 2002). Aspergillus niger as a cause of scleritis associated with endophthalmitis has also been described (Jager, 1994). The authors suggested that any late post-operative inflammation should be taken seriously and that patients must be carefully monitored for an underlying infective aetiology.
Mendicute et al (2000) reported 8 patients who developed keratomycosis soon after cataract surgery. Mean patient age at diagnosis was 73.6 years. Culture analysis revealed 7 cases ofAspergillus fumigatus and 1 of A. flavus. After medical treatment with antifungal agents, 6 cases resolved and 2 required evisceration. All 8 patients developed keratomycosis in the immediate post-operative period and were receiving topical treatment with dexamethasone 0.1% 4 times a day as a routine post-operative treatment. In addition, one patient had diabetes mellitus type II and one patient was receiving immunosuppressive treatment after a kidney transplant. All patients had corneal ulceration with a grey-white stromal infiltration around the incisional area as a clinical sign of keratomycosis and early treatment. The diagnosis of keratomycosis is based on clinical suspicion and biomicroscopic findings. In most cases, biomicroscopy makes it possible to observe epithelial ulceration with elevated edges and grey or white stromal infiltration. Fibrinous material, which could develop into endophthalmitis, may also be observed in the anterior chamber. The diagnosis is confirmed by properly obtaining samples from the edges of the stromal corneal ulcerations (Mendicute, 2000). Performing corneal biopsies can be useful in these situations, especially when the scrapings from the edges of the ulcers have not allowed identification of the causal agent (Ishibashi,1986). Collagen shields soaked in antifungal agents and lamellar keratectomy can be useful in selected cases (Mendicute, 2000)
Corneal fungus infections have been described in relation to cataract surgery, (Mendicute, 1995) keratoplasty, (Kloess, 1993) glaucoma, and radial keratotomy (Heidermann, 1995). The incidence of corneal mycotic infections has been reported to be between 15% and 20% in cases of stromal keratitis (Liesegang, 1980). Factors associated with corneal infections are the fungus itself, previous trauma, the immunologic state of the host, and administration of steroidal agents (Mendicute, 2000).
Over the last decade, a number of advances have occurred in cataract surgery: one of the most important is the development of the self-sealing sutureless wound. However, this wound architecture may leave a potential space if both the roof and floor of the tunnel are not in tight apposition; such a space can lead to the formation of an abscess cavity in the event of infection. This may alter the clinical picture, course of the disease, and outcome compared with infection of the conventional cataract surgery wound. Garg et al (2003) reported seven consecutive patients who underwent cataract surgery in different locations in India and developed microbiologically proven fungal infection of the surgical wound. Organisms identified were A. flavus (n=2), A. terreus (n=2), Aspergillus spp. (n=2), and Candida albicans (n=1). Although no specific source of infection was found in this series, the possibility of contamination of surgical instruments, infected conjunctival sac, and airborne infection cannot be ruled out.
Renal surgery
One case of renal pelvis infection due to Aspergillus flavus following a pyelolithotomy has been described (Davies, 1987). Pathological examination showed a nidus of Aspergillus around suture material persisting from a pyelolithotomy operation. It proved intractable and a nephrectomy was eventually necessary.
Treatment
Because of the scarcity of available information, optimal treatment of surgical infections due toAspergillus is unknown. Survival seems to depend on the excision of the infected tissue; in cases involving prosthesis, the new prosthesis should be placed in a non-infected field. Concomitant use of systemic antifungals is also crucial. Although there is no data favouring one antifungal drug over the others. It can be presumed that the response rates of voriconazole will be at least as good as amphotericin, given the results of the randomised study (Herbrecht, 2002). Caspofungin and micafugnin may also be useful.
Prevention
Operating room air may contain micro-organisms, dust, aerosol, lint, skin squamous epithelial cells, and respiratory droplets (Hambraeus, 1988). The microbial level in operating room air is directly proportional to the number of people moving in the room (Ayliffe, 1991). Therefore, efforts should be made to minimize personnel traffic during operations. In addition, operating rooms should be maintained at positive pressure with respect to corridors and adjacent areas (Lidwell, 1986).
Most modern operating theatres have conventional ventilation with filtered air, using filters with an efficiency of 80-95% to remove airborne particles down to 5 micrometres (Dharan, 2002). However, these filters are not sufficient to contain Aspergillus spores (2.5-3.0 micrometres in diameter), which require better technology. Laminar airflow is designed to move particle-free air over the aseptic operating field at a uniform velocity (0.3-0.5 m/sec), sweeping away particles in its path. This air-flow can be directed vertically or horizontally, and recirculated air is passed through a HEPA filter. Laminar air-flow systems with HEPA filter which remove airborne particles of 0.3 micrometres and above with 99.97% efficiency are generally used for orthopaedic and other implant surgery (Babb, 1995). These systems have been used in protective environment to help reducing the risk for health-care–associated airborne infections (e.g., aspergillosis) in high-risk patients (Barnes, 1989, Rhame, 1991, Walmsley, 1993, Pittet, 1994, Sehulster, 2004). However, data that demonstrate a survival benefit for patients in protective environment with laminar airflow are lacking, and few data support the use of laminar airflow systems elsewhere in a hospital (Roy, 1997).
In surgical site infections, it has been suggested that laminar airflow and ultraviolet germicidal irradiation as adjunct measures may reduce surgical site infection risk for certain operations. Neither laminar airflow nor UV light, however, has been conclusively shown to decrease overall risk for these infections (Sehulster, 2004). Although theoretically attractive, the use of these systems cannot, at this moment, be recommended for the prophylaxis of post-operative aspergillosis. Since no study has been systematically designed to evaluate this subject, it is not known if the universal use of these technologies will decrease the incidence of these rare infections. Furthermore, these strategies are limited by the expenses of both constructing and maintaining these systems and the lack of consensus about the level at which the risk can be numerically defined for aspergillosis (Dharan, 2002).
Since outbreaks of post-operative aspergillosis have occurred in units already using an HEPA filter, (Lutz, 2003) periodic maintenance of ventilation systems by the engineering department seems to be warranted. Efforts to limit excess humidity and moisture in the infrastructure and on air-stream surfaces in the ventilation system can minimize the proliferation and dispersion of fungal spores and waterborne bacteria throughout indoor air (Sehulster, 2004). For optimal performance, filters require monitoring and replacement in accordance with the manufacturer’s recommendations and standard preventive maintenance practices (Sehulster, 2004). Finally, microbiological sampling has its place in the investigation of epidemics, validation of changes in products and procedures in the maintenance of operating theatres (cleaning, disinfection, and ventilation) and education (Humphreys, 2004).
Nosocomial aspergillosis continues to occur despite air filtration, thus suggesting that there may be other hospital sources of spores. Fungi can inhabit water distribution systems, including those of hospitals, and may cause nosocomial infection, Aspergillus species have been recovered from the same hospital and from 2 other hospital water systems in Little Rock, Ark. Additional evidence that the opportunistic moulds are waterborne comes from the serious infections caused by Aspergillus species and Pseudallescheria boydii after near-drowning incidents in other healthy individuals. Water precautions need to be introduced in hospitals caring for patients at risk for opportunistic mycosis. Larger controlled studies, however, are needed to determine the role of water in the transmission of aspergillosis (Soubani, 2002, Anaissie, 2002). The report by Richet (1992) also suggests that the patients may be their own source of infection in some circumstances, as has been well demonstrated for staphylococcal infections.
Conclusions
While invasive Apergillus spp. infections usually occur in immunocompromised patients, cases of troublesome surgical site infections have been reported in immunocompetent patients. As shown in this article, several organs may be involved, depending on the surgery performed. These infections are usually indolent, in some cases occurring several months after surgery. In most of these cases, the source of the spores seems to have originated from the air during the surgical procedure; contamination from bronchopulmonary lesions or through hematogenous dissemination is also possible. Although not systematically studied, it is recommended to treat these patients aggressively with combined medical and surgical therapy, and the prophylaxis of these rare infections should include special care with the ventilation system in the operating room.
Prepared by:
Alessandro C. Pasqualotto
Post-doctoral Research Associate
School of Medicine
University of Manchester
Manchester, UK
January 2005