Posted: June 6, 2014

In recent years much has been made of the need for a better screening test for lung cancer.  Several large trials have recently been concluded and show that the mortality from lung cancer can be reduced by 15 – 20 % with annual CT scanning and diagnosing very small tumors.  If you limit screening to those at highest risk – typically people with a long smoking history – you can increase the yield from the test; i.e., the chances of diagnosing the disease go up, when compared to screening the general population.  The biggest objections to CT for lung cancer screening have been the cost and the accuracy.   The cost is coming down, as hospitals create attractive pricing for the test to the point where it is roughly equivalent to mammography.   Accuracy has been a problem because a high percentage of people who have abnormal CT scans don’t have cancer, just scar tissue.  Those people have to undergo a lot of further testing, up to and including in some cases lung surgery to remove what turns out to be benign disease.

Recently in the Journal of Clinical Oncology (click here for a link to the abstract; Dr. Stark can provide the article by email if you wish) Dr. Sozzi and colleagues from Italy announced that they have devised a blood test that helps sort out this problem.  They identified small fragments of RNA, genetic material, specific for lung cancer, and have devised a way to detect tiny amounts of this material in the blood of people with early stage lung cancer.   They used this test in people enrolled in an Italian trial looking at CT as a screening tool.  What they found was that in people with abnormal CT scans, if the lung cancer RNA blood test was negative there was a 99% chance that they did not have lung cancer.  If the RNA test was positive, that was not definitive for cancer, and more testing – as if the blood test had not been done – was required.  Nonetheless this test should prevent most lung biopsies and dramatically reduce the cost of finding a lung cancer.  Furthermore, the complications from lung biopsy and surgery, which can be severe, will be eliminated in most screened people.


Dr. Stark weighs in, “We are close to a screening paradigm for lung cancer that will greatly enhance accuracy and reduce the cost and danger of unnecessary lung biopsies.  In theory the blood test could replace CT by flipping the paradigm: doing the blood test first and then the CT only if the blood test is abnormal.  Someone needs to test this approach.”

Update 2020: when Dr. Stark published this posting in 2014, he assumed that tumor-associated RNA in the blood would be the next great thing in the early diagnosis of lung cancer.  Sadly there is nothing new to report.  No one has embarked on a large prospective trial comparing this test to low-dose screening CT scan, or the value of adding the blood test to screening CT.  Dead end?  Maybe.