CPA is more common in the developing countries due to the high prevalence of pulmonary tuberculosis [1
]. Alongside this there has been an increase in the number of invasive aspergillosis cases in developed countries related to immunosuppression in cancer patients having intense chemotherapy and for other autoimmune diseases [16
]. Like most published series [10
] tuberculosis was the commonest underlying disease in our series.
Here we report outcomes from a wide variety of surgical procedures for chronic pulmonary aspergillosis ranging from simple wedge resection to bilateral lung transplantation in patients with pulmonary aspergillosis. Most cases of CPA are managed medically in view of the high mortality and morbidity [17
], with the major exception of single aspergillomas. Simple aspergilloma cases are rare, most of our patients having CCPA. Surgery in CCPA is reserved for cases with complications or those who fail medical management. In our national centre, surgery has been reserved for patients with unilateral localised disease, failure of medical treatment and to deal with complications provided that their respiratory and performance status were adequate. Bilateral disease was not considered a contraindication per se; one patient with multiple aspergillomas underwent double lung transplantation for associated cystic fibrosis. Patients with good surgical indications, but poor nutritional status were considered for PEG feeding to improve their preoperative nutritional status. Two patients in our series had PEG feeding preoperatively. Preoperative optimisation of their respiratory status with intensive physiotherapy was routine.
Patients who presented with haemoptysis underwent bronchial artery embolization first. In patients with failed embolization, surgery was performed. Previous studies have found embolization to be ineffective or that it progressed to airway bleeding due to the existence of multiple feeder arteries from the chest wall [6
]. Some authorities even recommend prophylactic excision of pulmonary aspergillomas because of risk of massive haemoptysis [6
]. More recent angiographic series of bronchial artery embolization indicate that in ~90% of cases, control of bleeding is achieved, but that 30-50% patients rebleed by 3 years, especially if CCPA is not controlled with antifungal therapy [19
Most of our patients underwent pulmonary resection. In cases with localised unilateral disease, a wedge resection was carried out where possible to preserve lung tissue. Some authorities believe that because of the saprophytic nature of the organism parenchymal preservation is preferable provided that the rest of the underlying lung is healthy [11
]. These patients who underwent a localised resection had a better outcome and in most of them postoperative antifungal therapy could be stopped. The more complex group of patients in our series who underwent surgery to deal with complications like empyema and haemoptysis had a worse outcome (Figure ).
Survival curve of the patients who underwent surgery for simple and chronic pulmonary aspergillosis.
As others have found, patients with simple aspergillomas had better outcomes when compared to the CCPA group [9
]. Compared to other series [9
] we had better results in terms of intraoperative and 30 day mortality. We would attribute our preoperative optimisation regimes and postoperative care schedules as two leading reasons for this. Several of the studies were from Asia where patients came from rural areas with a high incidence of underlying lung disease. Results of different series for the surgical treatment of this condition have been summarized in Table .
Results of different studies concerning surgically treated cases of Aspergilloma
One patient described here underwent completion pneumonectomy for aspergillosis in the remaining lung following a previous lobectomy for lung cancer. This patient subsequently developed bronchopleural fistula which was managed conservatively. Completion pneumonectomy has previously been reported as having a high morbidity and mortality in the presence of infection [24
]. In these debilitated and often immunocompromised patients, a thoracoscopic approach has previously been described to be associated with better outcome and shorter hospital stay [25
]. In our series four patients (13%) underwent thoracoscopic surgery. All of them had simple aspergillosis but two of them needed conversion to a full posterolateral thoracotomy because of intrapleural adhesions. None of the patients with CCPA were suitable for a thoracoscopic approach.
Based on our experience we have concluded that surgery should be reserved for the following group of patients: Unilateral localised disease, failure of medical treatment or to deal with complications. In addition, we have summarised the risk factors for three groups of poor outcome, namely post-operative Aspergillus empyema, space infection and a general poor outcome leading to death (Table ). We cannot quantify these risks precisely and some are additive, others not.
The aims of antifungal therapy are to prevent Aspergillus
empyema and to prevent recurrence of CPA post-surgery, or at least progression, if residual disease remains. Discrete Aspergillus
nodules or simple aspergillomas that are resectable without any spillage of aspergilloma contents into the pleural space probably do not require any antifungal therapy [26
]. If given, it was discontinued after surgery if a complete resection had been done. Adjuvant antifungal pharmacotherapy does not improve the results of surgical treatment for isolated pulmonary aspergillosis where a full curative resection has been carried out [26
]. In the event that such spillage occurs unexpectedly, then washout of the pleural space with either amphotericin B deoxycholate or taurolidine is a reasonable, if unproven, measure to prevent Aspergillus
empyema is a difficult to treat entity, probably requiring long term antifungal therapy and may lead to pleural fibrosis and a significant restrictive pulmonary defect if only a lobectomy or wedge resection is done.
We also recommend peri-operative antifungal therapy for patients with multi-cavity disease, in whom surgery is done because of significant haemoptysis or in an attempt to improve quality of life, or in patients in whom resection of the primary disease is likely to lead to pleural spillage. Our standard practice is to start voriconazole 2 weeks before surgery, check plasma concentrations before surgery and adjust dose if necessary and continue IV therapy through the peri-operative period. Numerous drug interactions need to be considered, including marked prolongation of sedation post-operatively. We have seen patients unconscious for 24 hours after surgery because of the excessive effect of midazolam for example. If patients are on an azole prior to surgery and there is a risk of azole resistance, we use micafungin 150 mg daily pre- and peri-operatively. In either circumstance, if no spillage occurs, we will stop antifungal therapy shortly after surgery, but give at least 2 months therapy if here has been spillage to minimize the risk of pleural aspergillosis. If patients have residual cavitary disease, we will treat post-operatively long term to prevent recurrence, just as we do in patients who do not undergo surgery [27
In some patients lobectomy or pneumonectomy is possibly hazardous but a surgical necessity. Management of postoperative complications in this difficult patient group can be challenging. Creation of a modest sized fistula (the larger the better) is sometimes one approach. This is particularly so in patients with other medical problems and/or poor respiratory function, and amounts to a palliative procedure.
In patients unfit for a lobectomy drainage of the cavity (especially if there is a concurrent bacterial infection), pre-operative PEG feeding is a good first step. Following a stabilization, these patients could proceed to a definitive procedure if medically fit thereafter.
A difficult group of patients are those with CCPA who are left with a persistent space following resection. These space problems might be dealt with a pectoralis flap, modest thoracoplasty, or both. Use of a muscle flap reduces the extent of a thoracoplasty, which is helpful for later functioning of the chest. The advantage of the muscle flaps is that it often nearly fills the cavity initially, and then atrophies, leaving considerable space. Later post-op chest x rays show a small finger of muscle, but this is believed to be sufficient to keep Aspergillus from recolonising the cavity. Only one patient in our series underwent a thoracoplasty procedure. We planned thoracoplasty on another patient following resection; however as the patient was not fit for general anaesthesia and it couldn’t be carried out.
Postoperative follow-up continued for least 12 months, at 2 months initially and then 4 monthly, with Aspergillus IgG titres and a chest radiograph at a year. If there were no residual abnormalities, they were followed up for 3 years. Patients were asked to contact us on discharge if they had any new symptoms to allow us to identify expeditiously and treat patients with recurrence. Aspergillus IgG antibody titres usually fell very slowly following surgery, and would often level off remaining persistently elevated.