Pleural biopsy (multiple): a negative result did not rule out TB or malignancy.

Clinical bottom line (level 4)

  1. Tuberculosis and malignancy were very common in patients with pleural effusions.
  2. A positive or suspicious histology result following multiple pleural biopsies made tuberculosis or malignancy far more likely.
  3. A negative histology result could not safely exclude tuberculosis or malignancy.
Mungall et al: Thorax 1980; 35: 600-602
Expires April 2003

The study

Setting: two acute hospitals, UK

55 patients (aged range 26 to 90 years, 53% male) presenting with pleural effusions

?independent ?blinded reference standard, applied in all patients from a consecutive ?appropriate spectrum.
Reference standard:
  • final diagnosis
Diagnostic test: pleural biopsies (performed by four operators using an identical technique.) Five biopsies (2-10) were taken through one puncture site, and the Abrams needle was withdrawn each time. Pneumothorax was prevented by holding a swab to the chest wall between biopsies.

The evidence

pre-test probability of tuberculosis or malignancy: 75%, (95% CI: 63% to 86%)

diagnostic test TB or malignancy no TB or malignancy LR
(95% CI)		 post-test probability
positive pleural biopsy 19 0 inf
(1.8 to inf) 		100%
suspicious pleural biopsy 14 0 inf
(1.3 to inf) 		100%
negative pleural biopsy 11 11 0.25
(0.15 to 0.42) 		50%
total 44 11

  • Multiple pleural biopsies gave a diagnosis in 88% of patients with TB and 72% of patients with carcinoma.
  • Complications were four 'shallow' pneumothoraces and two episodes of surgical emphysema. Historically pneumothorax occurs in about 30% of FNA.
  • For the diagnosis of malignancy, the improvement with multiple biopsies was 29% (from 59 {single} to 88% multiple).

Comments

  1. There was referral bias probably present since patients from both district general and specialist chest hospitals were not separately identified.
  2. Can one extrapolate from four operators (skill level not defined, to multiple operators)? Probably not, and the number of biopsies performed in the specialist chest hospital is not separately given. I doubt that the complication rate in routine operators would be as low as reported. Paradoxically the pleural fluid cytology diagnoses were rather lower than one would expect even in routine laboratories. In addition the paper did not mention use of any ancillary cytologic or histologic techniques, i.e. immunostaining.

Citation

  1. Mungall IP, Cowen PN, Cooke NT, et al: Multiple pleural biopsies with the Abrams needle. Thorax 1980; 35: 600-602
Search Terms: pleural effusion in Cochrane
Contributor: Donald Stanley and Chris Ball, April 2000
Reviewer:

Clinical Question.
Patient pleural effusion
Intervention or Exposure pleural biopsy
Outcome diagnosis of TB and malignancy