Surgery for Lung Infection
When medical therapy has been optimised and an operation is the right next step

• Surgery is the last step, not the first. The decision to operate is multidisciplinary, made with respiratory medicine and microbiology/infectious diseases after medical therapy has been optimised; referrals come from those teams.
• Localised bronchiectasis is for symptom control. Disease confined to one lobe on CT, with recurrent exacerbations despite airway clearance and antibiotics for acute episodes, on a respiratory-medicine referral. Resection is usually a lobectomy.
• Aspergilloma is operated for symptoms or growth. A fungal ball in a cavity with an air-crescent sign; haemoptysis can be life-threatening. Bleeding is controlled first by bronchial artery embolisation, with antifungal therapy continued around the operation.
• Abscess and tuberculosis surgery follow medical failure or destruction. A prolonged antibiotic course for abscess; weeks of anti-tuberculous therapy under TB/ID direction for tuberculosis. Both are usually open operations.
• Spillage is the technical enemy. Across aspergilloma, abscess and hydatid, the operative priorities are to avoid spilling infected contents into the pleural space and to mobilise the lung so no fixed residual cavity remains.

Most lung infection never reaches a surgeon, and that is as it should be. Antibiotics, antifungals, anti-tuberculous therapy, physiotherapy and interventional radiology resolve the great majority of cases. What distinguishes the small number that come to surgery is not that they are more severe in isolation, but that they have declared themselves against optimised medical treatment — disease that remains localised but symptomatic, that is enlarging, that has destroyed a segment of lung, or that is bleeding despite every non-surgical measure.

This is why the operation is best understood as the earned step at the end of a pathway. The surgeon’s discipline is as much in knowing when not to operate, and in confirming that the medical groundwork has genuinely been laid, as in the operation itself. The judgement about whether antimicrobial therapy has been adequate, and whether the disease has truly failed it, belongs with the physicians managing the infection — respiratory medicine, microbiology and infectious diseases, and tuberculosis services. For this reason Mr Okiror takes these referrals from physician colleagues rather than as patient self-referrals, and works alongside the referring team rather than taking over the patient’s medical care.

A second principle runs through every one of these operations: do not flatten them into a single template. Each condition has its own medical preparation and its own threshold for surgery, set out below. Where the disease has a recognised national service — the National Aspergillosis Centre for complex fungal disease, the UCLH Echinococcosis multidisciplinary team for hydatid — those services are part of the decision, not an afterthought.

Most bronchiectasis is diffuse and managed medically for life, and surgery has no part in it. The exception is the patient whose disease is confined to a single lobe on CT and who remains symptomatic with recurrent chest infections, sometimes with haemoptysis, despite good medical care. The British Thoracic Society’s 2019 bronchiectasis guideline reserves surgical resection for exactly this group — localised disease, in a symptomatic patient, where medical management has not controlled the burden of infection.

A point worth making to referrers, because it differs from the abscess and tuberculosis pathways below: these patients are not usually on a continuous or prolonged antibiotic course before surgery. Bronchiectasis is managed with airway clearance and physiotherapy, with antibiotics given intermittently for acute exacerbations. Antibiotic therapy around the operation is perioperative cover, not a course that has to be “failed” first. The case for surgery rests on the pattern of recurrent localised infection and its effect on the patient, assessed on a respiratory-medicine referral.

The operation is usually a lobectomy to remove the affected lobe, and it is often begun robotically. Bronchiectatic lungs can, however, be densely adherent from years of inflammation, and the operation is converted to open surgery where adhesions make a keyhole approach unsafe. The aim is complete removal of the diseased lobe with the lung fully mobilised so the remaining lung expands to fill the space. The mechanics of lobectomy and the robotic approach are described on the robotic lobectomy and segmentectomy pages; where haemoptysis is a feature, bronchial artery embolisation is the first measure, as set out in central airway interventions.

An aspergilloma is a fungal ball within a lung cavity — a tangle of fungal hyphae, mucus and cellular debris, most often colonising a pre-existing cavity such as an old tuberculous space. On CT it characteristically shows an air-crescent sign: a rim of air separating the fungal ball from the cavity wall. Referrals come from respiratory medicine or from microbiology and infectious diseases. Surgery is indicated for symptomatic disease — and the dominant symptom is haemoptysis, which in this condition can be life-threatening — or, in an asymptomatic patient, for a lesion that is enlarging on serial imaging.

Two things must be in place before and around any operation. First, haemoptysis is controlled medically and radiologically: significant bleeding is treated initially by interventional radiology with bronchial artery embolisation, which stabilises the patient and converts an emergency into a planned procedure. Second, the patient is established on antifungal therapy — the triazole antifungals, such as voriconazole or itraconazole — which is continued around the operation and for a period afterwards. Dosing and duration are an antifungal-prescribing decision made with the microbiology/infectious-diseases team, not a website matter; the point for referrers is that surgery does not replace antifungal treatment, it is layered on top of it.

These are difficult operations, almost always performed open through a thoracotomy, because the cavity and surrounding lung are bound down by dense adhesions and the pleural space is often obliterated. The technical priorities are to remove the lesion — usually by lobectomy — without spilling fungal contents into the pleural space, and to mobilise the lung fully so that no fixed residual cavity is left, since a persistent space is itself a harbinger of further infection. A surgical review in the Journal of Thoracic Disease reports recurrence after resection of around 7%, and confirms that complex aspergillomas — those with an obliterated pleural space and a fixed hilum — are materially harder than simple ones. Complex and chronic pulmonary aspergillosis cases are discussed with the National Aspergillosis Centre at Wythenshawe Hospital in Manchester, the NHS England-commissioned national service, which holds a weekly remote multidisciplinary meeting for difficult cases. The patient-facing companion on coughing up blood (haemoptysis) covers the bleeding symptom in lay terms.

A lung abscess is a localised collection of pus within the lung. The great majority resolve with antibiotics alone, and the medical course is genuinely prolonged — several weeks of treatment. Surgery is considered only for the minority who fail adequate medical management: the patient who remains unwell with continuing fevers and with raised or rising inflammatory markers despite a proper antibiotic course. The reviews on lung abscess place the surgical group in this small refractory minority, and the threshold for operating is set by the failure of medical treatment, not by the presence of the abscess itself.

Where a patient is failing medical therapy, the next step is considered carefully. Image-guided drainage by interventional radiology can be appropriate, but it carries a real risk of spillage of pus into the pleural space, which can convert a contained problem into an empyema — the pleural-space infection covered on the empyema and pleural infection page. When surgery is required it is almost always a lobectomy and almost always open, through a thoracotomy, because the inflamed tissue planes are densely adherent and the priority is controlled removal of the abscess-bearing lobe without contaminating the chest. As with the other infective resections, the lung is mobilised so that the remaining lung fills the space and no residual cavity is left.

Tuberculosis is fundamentally a medical disease, and surgery has a narrow, well-defined supporting role. Any operation is undertaken only after the patient has been on anti-tuberculous medication for at least several weeks, and always under the direction of microbiology/infectious diseases or the respiratory TB specialists. The surgical literature is consistent that the role of surgery is selective and adjunctive to drug therapy, and that timing and drug cover are dictated by the TB physicians.

Within that frame, surgery addresses three situations. The commonest is the pleural manifestation: a tuberculous pleural collection that needs VATS debridement, or an organised tuberculous empyema requiring decortication — the operation to peel the fibrous rind off the lung so it can re-expand. The decortication operation, its staging and its evidence base are described in detail on the empyema and pleural infection page and are not repeated here. The second is lung resection for a localised, destroyed lung — a segment or lobe rendered non-functional and a continuing source of infection despite adequate treatment. The third is haemoptysis that persists despite embolisation, where resection of the bleeding focus becomes necessary after interventional radiology has been exhausted. These operations are usually performed open. Throughout, the surgeon works to the TB team’s treatment plan rather than independently of it.

Hydatid (echinococcal) disease of the lung is very rare in UK practice and is usually seen in patients with relevant geographic exposure. It presents as a cyst within the lung, and where that cyst needs removing, surgery is undertaken only once the patient is established on adequate antiparasitic medical treatment. Because this is a parasitic infection that almost no general thoracic service sees regularly, it is managed in liaison with microbiology and with the Echinococcosis multidisciplinary team at the Hospital for Tropical Diseases, UCLH — the national hub-and-spoke service for this condition, which has established links to cardiothoracic surgery for the thoracic cases.

Lung hydatid is also managed differently from the more familiar liver and spleen disease, and that distinction matters. The peri-operative antiparasitic strategy is designed to minimise the risk of cyst rupture while ensuring adequate antimicrobial cover through the operative and post-operative period, and it is planned case-by-case with the UCLH team — the specific drugs, doses and durations are a specialist prescribing decision, not a website matter.

The operation is almost always open. The overriding technical concern is to avoid spilling the cyst contents — the scolices — into the pleural space, because spillage can seed secondary infection and provoke an acute allergic reaction. The field is carefully isolated, a scolicidal agent such as hypertonic saline may be instilled into the lesion before it is dealt with, and the lung is mobilised so that no fixed residual space remains. As with every condition on this page, the surgery sits inside a medical pathway directed jointly with the infection specialists.

Mr Okiror’s wider thoracic practice is predominantly robotic and keyhole, and where the tissue allows — most often in localised bronchiectasis — an infective resection is begun that way. But it would be misleading to present infection surgery as routinely minimally invasive. Previous inflammation obliterates the natural planes between the lung, the chest wall and the mediastinum and binds them with dense adhesions, and in that setting a keyhole approach can be neither safe nor complete.

For that reason aspergilloma, lung abscess, a tuberculous destroyed lung and hydatid disease are usually open operations. The approach is chosen for safety and completeness rather than for the smallest scar. In infected tissue the priorities are the same across conditions: controlled dissection, avoiding spillage of infected material into the pleural space, and fully mobilising the lung so that no fixed residual cavity is left to harbour further infection. A planned open operation that achieves those goals is a better operation than a keyhole one that does not.

Disclosures

This page is referrer and clinician information, not medical advice for any individual case. Mr Lawrence Okiror is a Consultant Thoracic and Robotic Surgeon at Guy’s and St Thomas’ NHS Foundation Trust, with private practising privileges at London Bridge Hospital and The Lister Hospital Chelsea. Surgery for lung infection is undertaken in partnership with respiratory medicine and microbiology/infectious diseases, and, for the conditions noted, with the National Aspergillosis Centre (Manchester University NHS Foundation Trust) and the Echinococcosis multidisciplinary team at the Hospital for Tropical Diseases, UCLH. He has no commercial relationships relevant to this content. Antimicrobial, antifungal and antiparasitic prescribing decisions are made by the responsible specialist teams. Decisions about referral, surgery and treatment should be made on a case-by-case basis after appropriate clinical evaluation.