A Practical Approach to Pleural Infection

Abstract

Pleural infection encompasses a spectrum of disease that can present significant challenges in clinical practice. Despite better understanding of the underlying pathophysiology and microbiology, outcomes for patients remain poor. The use of antibiotics and chest tube drainage continue to be the mainstay of treatment, with surgery often reserved for those not responding to initial medical therapy. However, at present, the optimal management strategy for individual patients—including the role of early surgical and/or intrapleural therapy—is not clear. In this article, we provide an overview of the pathophysiology, diagnosis and management of pleural infection, highlighting current concepts and key practice points to aid the reader in caring for this important and often complex group of patients.

Key Summary Points

Pleural infection is a significant cause of morbidity and mortality worldwide.

Although typically seen in the context of bacterial pneumonia, the definition encompasses pleural space infection or empyema (frank pus) arising from any cause.

Maintaining a high index of suspicion is key, especially in individuals who fail to respond to treatment for pneumonia and/or have evidence of a new effusion on chest imaging.

Drainage of the infected pleural fluid along with antibiotics remains central to management, but often presents challenges in practice requiring individualised care.

Chest drains, regardless of size, must be flushed regularly to maintain patency and facilitate effective drainage.

Early discussion with multi-disciplinary team members is essential to providing optimal treatment, including early surgical intervention where appropriate.

Introduction

Pleural infection is a common condition encountered in clinical practice that dates to the time of the ancient Egyptians, around 3000 BC [1]. The early practice of open thoracic drainage, attributed to Hippocrates ~ 2500 years ago [2], was replaced by closed chest tube drainage in the nineteenth century, leading to substantial improvements in survival [3]. Nonetheless, pleural infection remains a significant cause of morbidity and mortality worldwide [4], which has claimed the lives of several notable physicians over the years, including Guillaume Dupytrens and Sir William Osler [5].

Typically defined as bacterial invasion and proliferation within the pleural space [6], pleural infection encompasses a spectrum of disease, ranging from the accumulation of free-flowing fluid to the presence of frank pus, termed empyema [4, 7]. The most common cause of pleural infection is bacterial pneumonia, where the descriptions ‘simple’ and ‘complicated’ parapneumonic effusion are often used to denote reactive or secondarily-infected fluid, respectively [8]. Up to ~ 50% of patients presenting with pneumonia may have an associated pleural effusion, with a smaller proportion of these (~ 5–10%) progressing to true pleural space infection [9–11]. However, it is recognised that around 30–40% of patients with pleural infection have no radiological evidence of pneumonia [12, 13], such that de novo infection within the pleural space is a possibility. This may be seen, for example, in cases of hepatic hydrothorax (effusions secondary to advanced liver disease) known as spontaneous bacterial empyema, which has a particularly poor prognosis [14]. Other potential causes of pleural infection are underlying lung cancer, oesophageal rupture, penetrating chest trauma, transdiaphragmatic spread from an abdominal source, or iatrogenic infections in the context of medical or surgical intervention [15].

The term ‘pleural infection’ is therefore broad, but in clinical practice most commonly refers to a complicated parapneumonic effusion, or empyema from any cause. Worryingly, for reasons that are not entirely clear, the incidence of pleural infection in the Western World appears to be increasing [16–18]. In one UK-based study, the incidence increased from 6.4 cases per 100,000 in 2008 to 8.4 per 100,000 in 2017, with the greatest proportional increase observed in patients over the age of 60 [19]. Mortality remains high, with rates of up to 10% in the first 30 days and up to 30% within 12 months, respectively [12, 17]. Notably, even in the absence of pleural infection, a ‘simple’ (i.e. reactive) parapneumonic effusion is associated with an increased length of hospital stay and higher mortality compared to pneumonia alone [20], suggesting that the mere presence of fluid within the pleural space is a significant factor.

Pathophysiology

The classic description for an evolving parapneumonic effusion is summarised in Fig. 1, based on a ‘tri-phasic’ progression from an exudative stage, to a fibrinopurulent stage, and finally to an organised collection with the pleural space [7, 21]. The process may be similar for non-parapneumonic causes of pleural infection, with bacteria entering the pleural space through either direct translocation (including from trauma or organ rupture) or haematogenous spread [22]. However, this model is likely an oversimplification, and the precise mechanisms underpinning the development and clinical course of pleural infection for individual patients may vary and are not fully established [21].

Fig. 1.

Stages of an evolving parapneumonic effusion [7, 21]. During the ‘exudative’ stage, inflammation of the lung parenchyma (i.e. pneumonia) extends to involve the visceral pleura, resulting in increased permeability of the cells lining the pleural membrane. Free-flowing fluid, rich in neutrophils, subsequently migrates into the pleural space. A ‘fibrinopurulent’ stage develops over hours to days, characterised by increasing fibrin deposition and membrane formation (septations) within the pleural space. The fluid becomes turbid and bacterial replication occurs. As fibroblasts proliferate, the pleural fluid becomes increasingly ‘organised’ (e.g. over the course of 2–3 weeks) and fibrin membranes become thickened and loculated. A pleural ‘rind’ may develop on the surface of the lung, resulting in lung entrapment and failure to expand. LDH lactate dehydrogenase

Microbiology

The causative bacteria associated with pleural infection differ slightly from uncomplicated pneumonia, despite the close association between these conditions (see Table 1). Notably, ‘atypical’ pneumonia species are rarely encountered within the pleural space [23], such that antibiotic cover for these organisms is not usually required. Historically, community-acquired sources (especially streptococcal species) have been implicated in the majority of cases (> 65%), with hospital-acquired sources (including Gram-negative organisms and staphylococcal species) contributing to the remainder [24, 25]. Staphylococcus aureus is now the most common pathogenic organism overall, with many cases (up to 60%) linked to methicillin-resistant species [25]. Nonetheless, it is increasingly recognised that pleural infection is complex and likely polymicrobial in nature (especially in the absence of underlying streptococcal pneumonia), with recent metagenomic analyses reporting the incidence of more than one organism in approximately 60–80% of cases [26, 27].

Table 1.

Common organisms associated with pleural infection [24, 25]

Community-acquired	Hospital-acquired
Streptococci	Staphylococcus
▪ S. viridans (including S. anginosus) ▪ S. pneumoniae Staphylococcus aureus ▪ Methicillin-sensitive Gram-negative species ▪ Enterobacteriaceae (including Klebsiella) ▪ Pseudomonas	▪ Methicillin-resistant Gram-negative species ▪ Enterobacteriaceae (including Klebsiella) ▪ Pseudomonas Streptococci ▪ S. viridans Enterococci

In a sub-analysis of the bacteriological profile of patients recruited to the MIST-1 trial (the First Multicenter Intrapleural Sepsis Trial [28]), non-Streptococcal species were associated with an increased length of hospital stay, while Staphylococcus aureus infection was linked to an increased mortality at 12-months [24]. This finding has recently been confirmed in the TORPIDS study [27] (based on the analysis of pleural fluid samples acquired during the PILOT study [29]), utilising next-generation sequencing (NGS) of recombinant 16S ribosomal RNA (a gene present in all bacteria) to determine the aetiology of pleural infection. Notably, while Staphylococcal and Enterobacteriaceae are associated with increased mortality, patients with Streptococcus anginosus infection or the presence of anaerobes appear to have improved survival outcomes compared to other species [27], consistent with the findings of a recently published 10-year retrospective observational study of positive pleural fluid microbiology [30]. The authors suggest this may reflect an increased susceptibility of Streptococcus anginosus to conventionally adopted antibiotic regimens, leading to earlier effective treatment in this group of patients [30]. As such, the identification of responsible organism(s) underpinning pleural infection may not only help guide appropriate antimicrobial therapy but also hold potential for future stratification of patients into high- or low-risk categories. While NGS is not currently used routinely, this represents a clear avenue for identifying the diversity of bacterial pathogens associated with pleural infection (particularly when conventional cultures are negative) that could help facilitate tailored management decisions [31, 32].

Other rarer, but important, causes of pleural infection include Mycobacterium tuberculosis (TB) and fungal organisms [33, 34]. Pleural effusions relating to TB are typically paucibacillary and lymphocytic, caused by a type IV hypersensitivity reaction to the tuberculous bacilli (‘TB pleuritis’). Frank empyema arising from TB is rare, but should be considered in at-risk individuals (e.g. those from areas of high TB prevalence, or the immunosuppressed) [4]. Although not tested routinely, in some centres pleural fluid adenosine deaminase (ADA) levels may be used to support the diagnosis, with elevated levels providing a sensitive marker for pleural TB infection [35]. Fortunately, most positive fungal cultures are likely to represent contaminants, and true fungal pleural infection rarely occurs but is associated with significant morbidity and mortality [36]. It should be considered particularly in the immunosuppressed (where testing for co-existent HIV is also advised [37]), or following oesophageal perforation or abdominal surgery [34].

Practical Box 1.

• The use of blood culture bottles, in addition to conventional universal containers, can substantially increase the yield of positive pleural fluid microbiology (up to 60%). Sending additional fluid samples of ~ 5–10 ml in blood culture bottles is therefore recommended.

• If fungal or TB infection is suspected, cultures need to be specifically requested. In some centres, ADA levels may also be used as a sensitive marker for pleural TB infection.

• Testing for co-existent HIV infection is advised, particularly in the immunocompromised or if fungal or TB infection is suspected.

Diagnosis and Investigation

The diagnosis of pleural infection requires a high degree of suspicion and is typically made based on a combination of clinical features, supportive radiology, and confirmatory pleural fluid analysis.

Clinical Features

The most characteristic presentation of pleural infection is a young patient with an acute respiratory illness (e.g. cough, sputum, breathlessness, pleuritic chest pain, fever), often with a history of non-resolving pneumonia, and the presence of a new pleural effusion on conventional chest radiography [8]. However, presentation can vary considerably, particularly in the elderly population who may present much more insidiously with fewer clinical manifestations. Non-specific symptoms, such as weight loss, anorexia, and fatigue, may be mistaken for other pathology, such as malignancy, leading to challenges in early diagnosis and initiation of appropriate treatment [37].

Several factors have been identified as independent predictors for the development of pleural infection [10, 38–40], which may support the clinical suspicion (see Table 2). Nonetheless, at present, there are no validated clinical prediction tools from which to reliably identify patients with pneumonia who are at risk of progressing to pleural space infection.

Table 2.

Risk factors and comorbidities associated with pleural infection [10, 38–40]

Independent risk factors	Associated comorbidities
Respiratory disease Current smokers Increasing age Alcohol excess Intravenous drug use Immunosuppression Gastro-oesophageal reflux disease Diabetes	Malnutrition Frailty Previous stroke Ischaemic heart disease Chronic renal failure Poor oral hygiene

Imaging

Chest radiography (CXR) is often the first-line investigation for patients with suspected pleural infection, and may demonstrate the presence of an effusion in addition to other pathology, such as consolidation. In the early stages, a typical crescenteric appearance with fluid meniscus may be seen (Fig. 2), which can become ‘lentiform’ (bi-convex) with increasing organisation and loculation of the pleural space [41]. This can create an obtuse angle with the chest wall, with loss of the normal gravitational fluid dependence. Importantly, CXR has poor sensitivity for detecting small-volume effusions which can often be ‘missed’, particularly in the context of basal consolidation [42]. The use of lateral CXR films may help to identify smaller effusions [43], but these are not performed routinely.

Fig. 2.

Diagrammatic representation of (A) the normal appearance of the lungs within the chest cavity, (B) a right-sided effusion (arrow) with typical fluid level, or meniscus, and (C) a ‘lentiform’ fluid collection tracking up the lateral chest wall (arrow), characteristic of an empyema

Computed tomography (CT) is widely used in clinical practice, yet its specific role in the investigation of pleural infection remains unclear and it is not an essential component of current guidelines [4]. There are no pathognomonic CT appearances that confirm the diagnosis. However, features such as increased pleural fluid volume and the ‘split-pleura sign’ (i.e. contrast-enhancement of thickened visceral and parietal pleura, separated by fluid) may help to differentiate pleural infection from ‘simple’ parapneumonic effusions [44]. Notably, the presence of pleural contrast enhancement along with fluid microbubbles, increased extra-pleural fat attenuation, and/or fluid volume ≥ 400 ml, have previously been identified as CT characteristics that predict the need for chest drainage [45], but these require further validation before they can be reliably used for clinical decision-making.

Bedside thoracic ultrasound (TUS) has become an essential imaging modality within pleural medicine and, specifically, for the investigation of pleural effusions. Its sensitivity and specificity to identify fluid are superior to CXR (94.5% vs. 67.7% and 97.9% vs. 85.3%, respectively) [46], while also enabling characterisation of the effusion and suitability for intervention [4]. Features such as echogenicity and the presence of septations or loculations have been postulated as highly suggestive of pleural infection [47]. However, none of these features are unique to infection and may be seen in other conditions, such as malignant effusions, mesothelioma, or effusions arising from rheumatoid disease [4]. It remains unclear whether these TUS appearances can accurately predict patient outcomes, such as the need for surgical intervention, or length of hospital stay [48]. As with other imaging modalities, it is not yet possible to diagnose or exclude pleural infection based on TUS appearances alone, and findings must therefore be interpreted in the broader clinical context.

Practical Box 2.

• Effusions that lose the typical fluid meniscus on conventional chest radiography should raise the suspicion of pleural space infection and warrant further evaluation.

• Although septations and loculations are highly suggestive of pleural space infection, these features may also be present in other conditions (e.g. malignancy). The clinical context is therefore key to interpreting these changes, alongside pleural fluid analysis.

Pleural Fluid Analysis

Pleural fluid analysis by thoracocentesis remains central to the diagnosis of pleural infection and can help guide appropriate management strategies (see Table 3). In the presence of frank pus, no further biochemical analysis is necessary (and indeed is considered a contraindication to gas machine testing [49]); in this circumstance, drainage of fluid should be attempted if safe to do so [4]. If the aspirated pleural fluid is non-purulent, this should be assessed initially for pH along with pleural lactate dehydrogenase (LDH), glucose and protein levels, to risk-stratify patients according to the likelihood of pleural infection.

Table 3.

Decision algorithm for suspected pleural infection based on pleural fluid analysis^a

Fluid appearance	Recommended approach
Purulent (i.e. frank pus)	Proceed to chest drain insertion (if safe to do so)
Non-purulent	Check pleural pH
	If ≤ 7.20	HIGH likelihood of infection → proceed to chest drain insertion
	If 7.21–7.39	INTERMEDIATE likelihood of infection → assess pleural LDH and glucose: 1. if LDH ≥ 900 IU/L and glucose ≤ 4 mmol/L, consider chest drain insertion (particularly if septations on TUS) 2. if LDH < 900 and glucose normal/high, no immediate indication for chest drain – continue to monitor clinically
	If ≥ 7.40	LOW likelihood of infection → no immediate indication for chest drain – continue to monitor clinically

LDH lactate dehydrogenase, TUS thoracic ultrasound

^aAdapted from British Thoracic Society guidelines, 2023 [4]

Importantly, in multi-loculated effusions, the pleural fluid pH can vary according to which locule has been sampled [50] and may therefore not be representative of the entire pleural space. In these circumstances, interpretation and clinical decisions (e.g. regarding insertion of chest drain) should not be based on the pH alone but guided by additional factors such as the clinical presentation and pleural fluid LDH and/or glucose levels, even if the pH value is ‘normal’. Conversely, a low pleural pH is not unique to pleural space infection and can be seen in other conditions, including malignancy and rheumatoid disease, as well as being influenced by the presence of local anaesthetic or residual heparin in blood gas syringes (both of which are acidic) [4, 49].

Microbiological analysis can confirm the presence of pathogenic organisms. However, standard cultures (i.e. those sent in universal containers alone) have a relatively poor diagnostic yield, in the region of 20–40% [51, 52]. Sending additional pleural fluid samples (e.g. 5–10 ml) in blood culture bottles can increase this yield to around 60% [52], and is therefore recommended within current guidelines [4]. Further improvements in microbiological yield (by up to 25%) may be achieved using ultrasound-guided pleural biopsies, with an initial feasibility study (the AUDIO trial) demonstrating positive cultures by this method even after administration of antibiotic therapy [51]. No adverse events were reported in this study, indicating the potential clinical utility of this technique as a supportive diagnostic tool, e.g. in instances where conventional culture methods are negative. Nonetheless, at present, pleural biopsy is not employed routinely for the investigation of pleural infection.

It is important to recognise that the diagnosis of pleural infection is not always clear-cut and, in many cases, uncertainty remains despite undertaking pleural fluid analysis. Utilising all available information, including the clinical context, available imaging, and results of biochemical tests, is therefore paramount: if the clinical suspicion persists, treatment should not be delayed. Table 4 provides a summary of key features which, cumulatively, may add weight to the diagnosis of pleural infection and prompt timely and appropriate therapeutic intervention.

Table 4.

Biochemical and radiological markers supporting the diagnosis of pleural infection [4, 10, 45]

Biochemical	Radiological
Serum	CT
▪ CRP > 100  ▪ Albumin < 30  ▪ Platelets > 400	▪ Pleural contrast enhancement  ▪ Pleural microbubbles  ▪ Increased attenuation of extra-pleural fat  ▪ Pleural fluid volume ≥ 400 ml
Pleural fluid	Ultrasound
▪ pH < 7.2  ▪ Glucose ≤ 4  ▪ LDH ≥ 900	▪ Presence of septations and/or loculations  ▪ Echogenicity

Two novel biomarkers have recently been identified, known as soluble urokinase plasminogen activator receptor (suPAR) and plasminogen activator inhibitor-1 (PAI-1). Both proteins are involved in fibrin formation, and elevated levels of suPAR and PAI-1 are found within infected pleural fluid and are associated with higher degrees of septation [48, 53, 54]. Increased concentrations of PAI-1, in particular, have been linked to patient outcomes such as length of hospital stay and 12-month mortality, offering potential utility as a pleural biomarker for infection [48]. However, their exact role in the diagnosis and management of pleural infection remains unclear and these are not currently used outside of research settings.

Practical Box 3.

• A low pleural fluid pH (< 7.20) is not always related to infection: the pH may be artificially lowered by the presence of local anaesthetic or heparin in blood gas syringes, as well as by malignant- or rheumatoid-related effusions.

• In multi-loculated effusions, the pH can vary according to which pocket of fluid has been sampled; a ‘normal’ pH does not necessarily exclude pleural infection in this circumstance.

• The typical ‘milky’ appearance of chylous or pseudochylous effusions can sometimes be mistaken for pus. Testing for pleural fluid triglyceride and cholesterol levels can help differentiate these conditions, as can centrifugation of the pleural aspirate if required (causing separation of cell debris and the clear supernatant in pleural infection) [55].

Management

The management of pleural infection is multifaceted, typically comprising medical treatment in the form of antibiotics, drainage of the infected pleural space where possible, select use of intrapleural therapies and, in suitable patients, surgical intervention. A simplified approach to the management of pleural infection, based on current guidance [4], is outlined in Fig. 3. Decisions are often challenging, especially for individuals who have multiple comorbidities or in the context of a heavily-organised pleural space, such that involvement of the wider multidisciplinary team (including respiratory physicians, microbiologists, cardiothoracic surgeons and, where relevant, palliative care colleagues) is key.

Fig.3.

Suggested algorithm for management of pleural infection, based on current British Thoracic Society guidance [4]. ICD intercostal chest drain, IPC indwelling pleural catheter, LAT local anaesthetic thoracoscopy, VATS video-assisted thoracoscopic surgery. *Consider early surgical referral in the presence of a complex pleural space (e.g. extensive thickening or solid component) or concerns regarding clinical stability. ^{^}Consider removing ICD when there is minimal residual fluid collection; consider switching to oral antibiotics once clinical improvement and a reduction in inflammatory markers has been achieved

Antibiotic Therapy

Antibiotics are often commenced empirically when pleural space infection is suspected. The choice of therapy should be guided by local policies and microbiology advice, but typically broad-spectrum agents targeting both Gram-positive and Gram-negative aerobes (particularly in hospital-acquired infections) as well as anaerobes is a sensible approach until culture results are available [25, 56]. Notably, recent work suggests that, aside from co-trimoxazole, the penetrance of commonly employed antibiotics within the pleural space is equivalent to that in the blood [57]. As indicated previously, ‘atypical’ organisms have little role in the aetiology of pleural infection, such that cover for these is unlikely to be necessary [23].

There is no firm position on when to switch from intravenous to oral antibiotic therapy in patients with pleural infection. However, this is usually considered when there has been objective evidence of clinical, radiological and biochemical improvement (e.g. cessation of fever, improving appearances on chest radiography, and a reduction in inflammatory markers). In practice, this should be assessed at 48 h following initiation of treatment [4], with the oral route usually feasible within 5–7 days provided there has been good clinical response. The overall duration of antibiotic therapy for pleural infection is typically in the region of 4–6 weeks [1], with recent European guidelines advocating at least 3 weeks [56]; however, this is largely based on expert consensus with no robust data to support such recommendations. Two recent trials have explored the feasibility of reducing antibiotic duration to 2–3 weeks, which may be appropriate in select patients with lower severity illness [58, 59]. However, both of these studies were underpowered, and there is currently insufficient evidence to recommend reduced durations routinely. Although previously studied [60, 61], there is no robust evidence to support the use of intrapleural antibiotic therapy in pleural infection [56].

Steroids

Since parapneumonic effusions are inflammatory in nature, it has been postulated that the use of corticosteroids may attenuate the immune response to prevent progression and/or improve clinical outcomes in such patients [62]. The recent STOPPE trial was the first randomised trial to assess this effect, comparing intravenous dexamethasone to placebo (normal saline) in patients hospitalised with pneumonia with evidence of pleural effusion. This study showed no benefit in clinical parameters, reduction in size of effusion or need for intervention in the steroid group [63]. There is therefore currently no evidence to support the use of steroid therapy in the management of pleural infection.

Practical Box 4.

• Response to initial treatment should be assessed within the first 48 h to help determine whether further intervention is necessary.

• A total duration of 4–6 weeks antibiotics is typically advised, but should be guided by the clinical, biochemical and radiological response, ability to drain the pleural space, and available microbiology.

• There is currently no evidence to support the routine use of steroids or intrapleural antibiotic therapy in the management of pleural infection.

Drainage of Pleural Fluid

Once the diagnosis of pleural infection is made, current recommendations are to insert an intercostal chest drain (ICD) as soon as feasible to allow effective drainage of the infected fluid [4, 56]. Usually, this is performed via the Seldinger technique under ultrasound guidance, though drain insertion by blunt dissection is also appropriate in certain circumstances (e.g. in a heavily organised collection). The size of chest drain employed remains an area of contention [64]. Many advocate the use of larger bore drains to help facilitate effective drainage of infected fluid. However, there are limited data comparing different drain sizes: the best available evidence suggests there is no difference in patient outcome when using a small-bore drain (defined as ≤ 14F) versus a larger bore drain (> 14F, and typically ≥ 18F), though larger drains may cause increased pain, particularly when these are inserted by blunt dissection [65]. Data from the MIST-1 trial showed that small-bore drains were as effective as large-bore drains but, crucially, all were flushed thrice daily with 30 ml saline, suggesting this is an important component to maintaining drain patency, particularly for smaller calibre drains [66]. It is therefore recommended that all drains are flushed regularly (e.g. at least 3–4 times/day) to avoid blockage [28, 66, 67].

In practice, the drain should remain in situ until drainage output reduces and there is evidence of clinical improvement; this is typically indicated by resolution of fever and improvement of inflammatory markers, though there are no absolute criteria [1, 56]. Radiological features, such as clearing of pleural opacification and reduced collection volume, are helpful markers of successful drainage and improving infection which have been utilised in clinical trials [66, 68]. If patients fail to respond clinically within 48 h, further imaging with CT should be considered to assess drain position, the extent of disease, and to facilitate discussion with cardiothoracic surgical colleagues regarding suitability for surgical intervention [4].

Other options for drainage include trials of therapeutic aspiration [69, 70] (e.g. in patients who are not suitable for ICD placement or surgical intervention), or the use of indwelling pleural catheters (IPCs) in some circumstances, for instance chronic infections not amenable to surgery [71]. The use of repeated therapeutic aspiration, in particular, offers a potential practical advantage by facilitating drainage of multiple fluid pockets (loculations) in an ambulatory setting without the requirement for chest tube placement, which can be distressing for some patients. The IMPACT trial [72] is currently in set-up and represents the first large-scale randomised trial to compare outcomes (such as length of hospital stay) in patients managed with therapeutic thoracentesis versus chest tube placement. However, at present, there is a lack of robust data concerning the efficacy of these approaches to inform guideline recommendations, and they may be considered pragmatically when other options are limited.

Practical Box 5.

• A small-bore chest tube (i.e. < 14F) is generally suitable for drainage of infected pleural fluid. However, regardless of drain size, a crucial element of management is to implement regular flushing (e.g. 30 ml saline at least 3–4 times/day) to maintain patency and prevent blockage of the drain.

• The use of therapeutic aspiration and indwelling pleural catheters may be considered in select cases, e.g. where insertion of an ICD is challenging, or in chronic pleural space infections not amenable to surgery.

Intrapleural Therapy

In some circumstances, insertion of an ICD alone is not sufficient to drain the infected pleural space, particularly when this has become multi-loculated through progressive fibrin deposition and membrane formation [73]. The MIST-2 trial determined that the combination of an intrapleural fibrinolytic agent (tissue plasminogen activator; tPA) with human DNase led to significant improvements in radiographic appearance, rates of surgical referral, and length of hospital stay compared to placebo or either agent alone [66]. On the strength of this study, the ‘MIST protocol’ treatment regimen has now become an established part of clinical practice for the management of poorly-draining infected pleural effusions [4, 56] (see Practical Box 6). This strategy should be considered early (e.g. within the first 24–48 h) in patients who have complex, multi-loculated effusions or who fail to respond to initial medical management in this time [4]. There is, however, no clear evidence that intrapleural therapy improves mortality compared to placebo [66, 74].

Of note, there has been a global shortage of tPA in recent years, limiting its availability for intrapleural administration. While there is no evidence to support the use of DNase monotherapy in pleural infection, several studies have reported on the efficacy of urokinase as an alternative to tPA in combination with DNase [75, 76]. As such, the British Thoracic Society currently recommend the use of urokinase 100,000 units in place of tPA where this is required [77]. Alternative strategies, employing reduced dose or once daily tPA in combination with DNase, have previously shown to be both safe and effective in the management of pleural infection [78, 79], and may therefore be considered.

Reduced-dose tPA has additional theoretical, as well as pragmatic, benefits for patients in whom the risk of bleeding—a recognised complication of intrapleural therapy [80]—is a concern. However, in a recent large multi-centre, retrospective observational study, there was no reported difference in bleeding outcomes with reduced (i.e. 5 mg) versus full (i.e. 10 mg) dose tPA [81]. The overall risk of bleeding in this study was approximately 4%; supporting data derived from the MIST-2 study [66]. Importantly, the risk of bleeding is significantly increased in patients receiving concomitant therapeutic anticoagulation (~ 10%); it is therefore recommended that anti-coagulation medication is reviewed and suspended if safe to do so for at least 2 days prior to administration of intrapleural fibrinolytic therapy [56, 81]. Where bleeding risk remains a concern, large-volume saline irrigation, i.e. 250 ml saline three times per day for 3 days, may be considered as an alternative therapeutic option [68]. However, larger randomised trials are necessary to fully confirm the efficacy of this approach.

Beyond bleeding, significant complications associated with intrapleural therapy are rare, although pain is a frequently reported symptom in up to 40% of patients [81].

Practical Box 6.

• Typical administration of intrapleural therapy, based on the MIST-2 study protocol [66], is as follows:

  1. 10 mg tPA (alteplase) in 30 ml 0.9% NaCl (saline), followed by a 10-ml saline flush, then
  2. 5 mg DNase (dornase alpha) in 30 ml water, followed by a 10-ml saline flush, then
  3. Clamp drain for 1 h, then re-open to allow free drainage

  This process should be repeated twice per day (i.e. approximately every 12 h) for 3 days (i.e. a maximum of 6 doses of intrapleural therapy should be administered).

• Urokinase 100,000 units may be used instead of tPA if required, e.g. as is currently recommended in view of the global shortage of tPA. Alternative strategies may include once daily or reduced dose (i.e. 5 mg) tPA, or large-volume saline irrigation (i.e. 250 ml saline thrice daily for 3 days).

• Concomitant anti-coagulant medication should be reviewed and suspended if safe to do so prior to administration of intrapleural fibrinolytic therapy to reduce the risk of bleeding complications.

Surgery

The role and timing of surgery for pleural space infection remains controversial. However, it is recognised that around 20% of patients will fail to respond to medical therapy alone [28, 66], typically those with residual sepsis and/or a complex pleural collection with pleural thickening and ‘trapped lung’ [29, 82]. Previous consensus statements from European and American surgical societies have advocated for the early consideration of surgery in such patients, where timely intervention may be associated with improved outcomes [41, 83]. Key to success is the selection of suitable patients and a technique that will facilitate optimal debridement of infected tissue and restoration of lung expansion (i.e. ‘decortication’ of the fibrinous pleural rind) [56]. In practice, video-assisted thoracoscopic surgery (VATS) is the preferred method, offering reduced pain, risk of post-operative air leak, length of hospital stay and overall morbidity compared to open thoracotomy [84–87]. Nonetheless, up to 14% of patients undergoing VATS may require conversion to an open procedure [86], with initial delays to surgery a recognised predictor [88, 89]. Determining which patients will benefit from early surgical intervention versus initial medical management therefore remains a key area of interest. The recent MIST-3 trial [90] confirmed the feasibility of randomising patients with pleural infection to early VATS or intrapleural enzyme therapy, demonstrating a potential reduction in median length of hospital stay with both interventions (7 days each) compared to standard care (10 days). This approach is supported by a recent US-based study [91], which found a trend (non-significant difference) towards reduced length of stay in patients receiving early surgical intervention versus intrapleural enzyme therapy (median 5 days vs. 11 days, respectively). A larger randomised trial is necessary to fully establish these findings: the forthcoming MIST-4 trial aims to directly compare the clinical efficacy of early VATS or intrapleural enzyme therapy [92], while the FIVERVATS trial will compare outcomes (including length of stay) in patients receiving VATS or intrapleural enzyme therapy as first-line treatment [93]. Results of these studies have potential to impact and reshape current management pathways.

Local Anaesthetic (‘Medical’) Thoracoscopy

The use of local anaesthetic thoracoscopy (LAT) offers an alternative, less-invasive approach to VATS in centres where there is both facility and expertise to deliver this procedure [56]. This may be particularly appealing for patients with complex pleural collections who are considered too high risk for surgery (e.g. frail, elderly patients with comorbid disease), but would nonetheless benefit from intrapleural intervention. Several observational studies have been performed, with a recent meta-analysis demonstrating a pooled success rate of 85% when LAT was used either first-line or following initial attempted drainage with ICD [94]. As such, LAT may offer a pragmatic option in certain patients. However, to date, there is no robust evidence or randomised trial data to support its routine use. Moreover, a recent UK-based feasibility study (the SPIRIT trial) has demonstrated failure to successfully randomise patients to LAT (vs. standard chest drain) for the management of pleural infection [95], bringing into question the wider applicability of this approach beyond specialist centres.

Risk Stratification

Current British Thoracic Society guidelines recommend calculating the RAPID score for all patients with pleural infection [4]. This score is based on five parameters: renal function (serum urea level); age; presence of purulent pleural fluid; source of infection (i.e. community or hospital acquired); and dietary status (serum albumin), and has been prospectively validated to risk stratify patients into low- (score 0–2), medium- (score 3–4), or high-risk (score 5–7) groups according to 3-month mortality [29, 96]. Notably, in low-risk patients, mortality is 2.3% at 3 months, compared to 9.2% and 29.3% in medium- and high-risk patients, respectively [96]. Such information could be beneficial in guiding discussions with patients and families regarding prognosis [4]. However, at present, it is unclear how the RAPID score can directly impact routine clinical care and decisions regarding optimal treatment of pleural infection. This represents an important area for future research.

Conclusion

Pleural infection is a common but often challenging condition encountered in clinical practice that continues to carry a high risk of mortality. Maintaining a high index of suspicion is key, particularly in patients presenting with new or recent pneumonia who fail to respond to initial treatment. Prompt recognition and investigation with pleural fluid analysis is fundamental to ensuring timely diagnosis and intervention. While medical treatment in the form of antibiotics and chest drain insertion remains central, in select patients early surgical discussion for debridement and decortication should be considered. At present, it is not possible to reliably predict who will benefit most from such early intervention, or indeed who will progress to pleural infection in the first place. These represent key areas of research that will better inform our understanding of the pathogenesis and management of pleural infection, enabling improved care that can be tailored to individual patient needs.

Author Contributions

Both authors contributed to the conception and design of this work. Initial preparation, collection of material, and first draft of the manuscript was performed by Steven J Smith. Subsequent editing and revision of the manuscript and preparation of Tables and Figures was performed by Benjamin J Pippard. Both authors commented on revised versions of the manuscript, and read and approved the final submitted manuscript.

Funding

No funding or sponsorship was received for this study or publication of this article.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.

Declarations

Conflict of Interest

Steven J Smith and Benjamin J Pippard declare that they have no competing interests.

Ethical Approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.