Mar 042015

The American College of Radiology has started implementing its new system for risk-stratifying the findings on low dose chest CTs performed for lung cancer screening. The ACR's new system is called Lung-RADS, and it emulates the ACR's familiar Bi-RADS's 0-4 scoring system for mammography:

  • Lung-RADS 0: Incomplete, meaning previous chest CTs are still being located for comparison, or part of the lungs cannot be properly visualizes.
  • Lung-RADS 1: Negative. No nodules are seen, or definitely benign nodules are seen (with complete, central or popcorn calcifications or fat in a benign pattern). Risk of cancer <1%; continue annual low dose chest CT screening.
  • Lung-RADS 2: Benign appearance; nodules are present that are low risk (e.g., new solid nodules <4 mm or stable <6 mm). Risk of cancer <1%; continue annual chest CT screening.
  • Lung-RADS 3: Probably benign, repeat chest CT in 6 months (e.g., new solid nodules 4 mm to <6 mm). Risk of cancer 1-2%.
  • Lung-RADS 4A: Suspicious, 5-15% risk of cancer (e.g. new solid nodule 6 mm to <8 mm), repeat chest CT in 3 months or get PET/CT if  solid component is ≥ 8 mm
  • Lung-RADS 4B: More suspicious, >15% risk of cancer (e.g. new or growing solid nodule ≥ 8 mm). Obtain PET/CT and/or biopsy.

LungRADS's risk stratification is based on the Fleischner society guidelines. It has not yet been prospectively validated, but radiology leaders predict LungRADS will enter standard use among radiologists. (Anecdotally, this has already occurred.) LungRADS classification data could also be included in the data reporting registry required by CMS for Medicare payment eligibility.

The ACR hopes that LungRADS's variables and modifiers including nodule size, consistency, and other factors will help discriminate cancer from benign findings, reducing the high false positive rate of 40% seen in the National Lung Screening Trial (NLST).

For example, to be considered "suspicious," nodules must be at least 6 mm diameter for a solid nodule under Lung-RADS, compared to just 4 mm in the NLST.

A study in Annals of Internal Medicine applied the Lung-RADS criteria retrospectively to chest CTs in the NLST. The false positive rate was reduced by about half (27% to 13%); however, sensitivity was also reduced (79-85% using Lung-RADS, compared to 94% using the NLST criteria).

In plain English, all that means the ACR made the decision to let a few early lung cancers grow for a few extra months, to vastly reduce the total number of scans performed, including false positives. In theory, this shouldn't be dangerous, as the ~5 mm nodules that were demoted from "suspicious" (in NLST) to "probably benign" in Lung-RADS will be re-scanned again in 6 months. This scheme will still identify the vast majority of lung cancers at an early stage.

But not all, of course.

Changing the cutoffs for positivity from the NLST (where the survival benefit was actually shown) is a gutsy move by ACR. It's in the interest of public health and promotes the intelligent use of limited health care resources.

It's also quite timely. By implementing Lung-RADS before community-based lung cancer screening gets underway, there will be no real-world comparison cohort screened using the NLST criteria. By creating the new professional standard by fiat, ACR thus shields its members and other physicians (like pulmonologists) from liability.

Sometimes it's good being the cartel.

Lung CT Screening Reporting and Data System (Lung-RADS™), ACR website. 

Performance of Lung-RADS in the National Lung Screening Trial: A Retrospective Assessment, Annals of Internal Medicine 10 February 2015 online.

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)


Lung-RADS launches: New system for lung cancer screening chest CTs