Feb 042014

On-site, intra-procedure cytopathologic examination of aspirated tissue during transbronchial needle aspiration (either by EBUS or "blind" approach) is probably helpful during bronchoscopy. Why wouldn't it be? You've got a trained professional there to tell you when you've made the diagnosis and can stop taking biopsies. Diagnostic yield should go up, complications down. Randomized trials have confirmed the latter (lower complications, fewer passes), but not the former (increased yields).

The problem with on-site pathology is, like any consultation, it's dependent on the availability and physical presence of another professional. But why can't pulmonologists just do their own cytopathologic analysis of tissue samples from bronchoscopic biopsies?

Dr. Michele Sediari (Italy) et al gave us an interesting answer to that question. Dr. Sediari got 54 hours of cytopathology training (18 x 3 hr sessions) and read a few pathology textbooks. Then she went head-to-head against a professional cytopathologist on 362 TBNA samples on 84 patients, with subsequent definitive analysis by another pathologist as the gold standard.

She did pretty darned well:

  • Sensitivity: Cytopathologist 95%, Pulmonologist 91%
  • Specificity: Cytopathologist 92%, Pulmonologist 72%
  • In cases of definite malignancy as the final diagnosis, the pulmonologist agreed with the cytopathologist 92% of the time.
  • Overall agreement was 81%, with the pulmonologist failing mostly on borderline cases ("unsure, probably [benign/malignant]")

In other words, the pulmonologist had a high sensitivity (low false negative), but only okay specificity (18% false positive), and a high certainty in malignant cases. All of which is just fine for a screening test that will be followed by a definitive evaluation. Sensitivity is what's most important in this situation.

The problem here is one of "scalability," as the I.T. people say.

I received zero formal pathology training in fellowship, although we walked down there a few times and it was usually edifying. While I don't think it should crowd out airway management, 54 hours of cytopathology would be a challenging but manageable carve-out of a three-year U.S. curriculum. In 18 clinical months, it's 3 hours a month. (The hardest part, at most places, might be getting the pathologists to work it in so they can still leave at 5 PM.) The most effective visual learning would probably happen during the tense moments around diagnosis on biopsy samples being collected and examined by the training physician.

However, in the U.S., the need and the will for this kind of change are probably lacking. For physicians moving on to high-resource settings, their cytopathology training would not go to good use (most trained physicians would probably still engage cytopath resources rather than do it all themselves -- it's more efficient). A better solution for pulmonologists practicing in underserved areas would be CME training subsidized by their centers. But I suspect most physicians would rather forgo that significant investment and keep doing it the old way: take a bunch of biopsy passes and send it off to pathology, and accept the slightly higher risk of complications suggested in randomized trials.

My problem (ever since 8th grade biology lab) is, I have never been able to work those microscope knobs to get the slide in focus. Also, those slide prep solvent fumes make me woozy. All in all, I'd say we have a fine division of labor as it stands, thank you.

Martina Bonifazi, Michele Sediari, Stefano Gasparini et al. The Role of the Pulmonologist in Rapid On-site Cytologic Evaluation of Transbronchial Needle Aspiration: A Prospective Study. Chest. 2014;145(1):60-65. doi:10.1378/chest.13-0756.

Get our weekly email update, and explore our library of practice updates and review articles.

PulmCCM is an independent publication not affiliated with or endorsed by any organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)


Can pulmonologists do their own on-site cytopathology during bronchoscopy?