Pleural effusion

"Pleural disease" is a Level 1 topic from Section 2.1.5 in the first edition of the CICM Syllabus for the Second Part Examination, where the college leaves the question "which disease" sufficiently open for the same syllabus item to embrace empyema, pleural effusion, and theoretically pneumothorax or even mesothelioma. Question 13.3 from the first paper of 2008  presented the candidates with a characteristic film, and asked them to "list 4 clinical signs typically found on chest examination". Up until Question 16 from the first paper of 2020,  this had been the only engagement with pleural effusions the college has had, which is surprising given how much one can ask about. Is it transudative or exudative? What caused it? What are the radiological features? What tests would you order? And so forth. Acknowledging this, they ultimately asked a detailed question about pleural fluid analysis in Question 16 from the first paper of 2020.

The best resources for this:

• The atelectasis page from LITFL
• Light's own 1972 paper about his exudate vs transudate criteria (1972)
• Light's "Pleural effusion" article for NEJM (2002)

Of the formal published literature, nothing beats this BTS guidelines statement from Thorax (2003). Additionally, Quadri & Thomson (2002) have an excellent paper with an unfortunate name ("The art of pleural fluid analysis") which wrongly implies that some kind of creativity is required for this mundane biochemical pattern recognition exercise. Pleural fluid analysis is probably more scientific than that, and most normal people would agree that one would not benefit professionally from becoming renowned as the Picasso or Hieronymus Bosch of pleural fluid analysis.

Causes of pleural effusion:

Clinical features of pleural effusion

• Tachypnoea
• Unilaterally decreased chest expansion
• Stony dull percussion note
• Absent breath sounds
• Whispering pectoriloquy above level of effusion
• Bronchial breath sounds above level of effusion
• Apical impulse shift to left

Radiological features of a pleural effusion

What's that opacity on the Xray? It is denser than air, that's all you can say. The photons didn't penetrate it very well. Could be anything dense: could be tumour, could be pus, could be blood, could be semen. Anything.

The features listed in Radiopedia are:

• blunting of the costophrenic angle
• blunting of the cardiophrenic angle
• fluid within the horizontal or oblique fissures
  • 200ml of fluid is required to make a normal erect PA film look abnormal
  • About 50ml is enough to blunt the costophrenic angles 

Indications for thoracocentesis

• Do I even need to tap it, you might ask. Can't I just dry them out? You sure can. Another R.W. Light paper can be quoted: Approximately 75 percent of effusions due to congestive heart failure resolve within 48 hours with aggressive diuresis.
• Unilateral effusion
• Bilateral effusion which fails to respond to medical therapy
• Suspicion of empyema (or the expectation that one will form)
• Difficulty weaning ventilation (increased effort of breathing)
• Diagnosis (i.e. is it malignant?)

The appearance of the fluid gives clues, according to another paper by Light.

• Bloody fluid suggests malignancy (or haemothorax, duh)
• Straw-coloured fluid suggests transudate
• Milky-coloured fluids suggests chylothorax.

Biochemical features of pleural effusion

This is really the list of things you might want to order:

• Pleural fluid protein and LDH;  Serum protein and LDH. Light's own 1972 paper about his criteria is available online. Of course at that stage he never actually called them "Light's criteria". However, 30 years later in his article on pleural effusions for NEJM Richard W Light does refer to his own criteria as "Light's criteria".
  • In short, your effusion is exudative if:
    • The fluid to serum protein ratio is greater than 0.5
    • The fluid LDH is over 200 IU/L
    • the fluid LDH to serum LDH ratio os greater than 0.6
• Glucose:  an extremely low pleural fluid glucose suggests that something is consuming it. Low pleural fluid glucose suggests TB, pneumonia or malignancy.
• pH: this is a weird one. Everybody orders pleural fluid pH, and few understand what significance it has. According to the 2000 guidelines from CHEST, pH can determine the need for therapeutic drainage. Anormal pleural pH is about 7.60; a pH of <7.20 is equivalent to a positive gram stain in terms of identifying an effusion which requires drainage. Oesophageal rupture can also cause a low pleural pH.
• Amylase:  this is elevated in pancreatitis-related effusion and in oesophageal rupture
• Cholesterol: this reveals the effusion as a chylothorax
• Cell count

Cellular features of a pleural effusion

• Lymphocytosis = malignancy or tuberculosis. This cell count clue comes from another paper by Light. In his case series, of 31 exudative effusions with a lymphocytic predominance, 30 were due either to tuberculosis or neoplasm.
• Neutrophilia = parapneumonic effusion or PE.
• Eosinophilia (more than 10% eosinophils)  usually means there has recently been blood or air in the pleural space; however weird causes include drugs and environmental toxins (dantrolene, bromocriptine, nitrofurantoin, exposure to asbestos) or autoimmune causes eg.  Churg–Strauss syndrome.

Appearance of pleural fluid

In much the same way as the analysis of ascitic fluid can include a careful examination of its appearance, looking at your pleural fluid sample using your normal human eyes is possible. The question is, does it achieve anything? The answer may not surprise you. CICM examiners certainly think it does, as they included an appearance section in their answer to Question 16 from the first paper of 2020. Villena et al (2004) asked this question in a titanic 7-year-long prospective study from Madrid, during which 715 10-ml samples (that's 7 litres of pleural fluid) were scrutinised by numerous clinicians. The samples were described as "watery", "serous", "blood tinged", "bloody", "purulent", "milky" "turbid" and "brown".  Their appearance, it turned out, had virtually no relationship with their aetiology. "The appearance of the fluid should not be overemphasized as a diagnostic test", the investigators grated sourly. However, in order to please the cour of examiners, the following list (from Question 16) should be regurgitated into future exam answers:

• Clear, straw-coloured – more likely transudate (although still may be exudate)
• Blood-stained – malignancy, pulmonary infarction
• Yellow/green – rheumatoid Pus – empyema
• Turbid – inflammatory exudate

Culture of pleural fluid

It is inevitable that this potentially extremely turbid frank-pus-looking specimen will be decanted triumphantly into some kind of jar and sent to the laboratory with the expectation that the result will surely be a positive one, which will quickly identify the pathogenic organism and set the course for antibiotic success. This is not the inevitable outcome of a pleural aspirate culture, because:

• The fluid is often sterile following the use of antibiotics
• Aerobic organisms will have died in the middle of a large anoxic pool of fluid
• Anaerobic organisms are likely to die in the oxygen-rich culture bottle. If this thing was happy in the middle of an effusion,  it will probably be unhappy in the presence of fresh air and sunlight.
• Those organisms that survive the collection process will surely be killed by the millions of angry neutrophils who are still alive and well in the culture
• Even when something happens to survive long enough to grow in the sample, the result is some kind of resilient hero who outcompeted both the neutrophils and the other bugs on this specific agar, and is therefore potentially not a representative of what is happening in the pleural space.

The yield of such cultures can be increased by inoculating the culture into blood culture bottles (an improvement of about 20% in the case of Menzies et al, 2011) and Porcel (2011) reports some data where the yield from patients with "true" empyema was something like 70%. On the other hand, the same author (Porcel, 2010) found that unselected cultures only ever grow something in 3.2% of cases. When  Jiménez et al (2006) selected their patients more carefully and tested only patients with parapneumonic effusion, the yield was 19.3%, but changed management in only a minority of patients without affecting their outcomes in any way.