PET scans are an important way to discriminate between metabolically active nodules, suggestive of cancer but sometimes representing inflammation or infection, and non-PET-avid lesions that are felt much likely to represent cancer. They are also a cornerstone of "clinical" staging by imaging and patient exam (vs. "pathologic" staging by surgery to clarify where cancer is or isn’t). PET scans, however, require rapid metabolic activity to highlight a cancer, so there’s good reason to question whether PET scans can be used effectively for the BAC subset of lung cancers, which, as I’ve described in prior posts (here, for instance), can be remarkably indolent.

No specific prospective (planned ahead of time) studies of PET scanning in BAC have ever been done, but there are now enough BAC patients in other trials for us to look retrospectively at meaningful numbers, so we can now begin to assess the feasibility of PET scans for diagnosis and staging of BAC.

By way of background, one key study demonstrated that PET scans for lung cancerhad a sensitivity of 97% (meaning 97% of lung cancers would be found to be abnormal on a PET scan), and a specificity of 78% (so 78% of the time, the abnormality would be from cancer and not another source, like infection or inflammation) (abstract here). Breaking out 7 BAC patients from a 29 patients (and 30 total lung adenocarcinomas), Higashi and colleagues from Japan (abstract here) found that only 3 of the 7 registered positive results on the PET scan. In contrast, only one of the other 23 non-BAC adenocarcinomas remained undetected on PET scans. This same study also reported that there is a stepwise increase in the mean (bascially the average) standard uptake value (SUV, the measure of metabolic intensity on a PET scan) from BAC tumors (mean SUV 1.36), to well-differentiated adenocarcinomas (mean SUV) to moderately differentiated adenocarcinomas (mean SUV 4.63). Heyneman and colleagues (abstract here) reviewed their results from a tumor registry over a 6-year period, during which time they identified 15 patients who had a proven BAC lesion and had also undergone a PET scan. Of these, nine were considered positive. In particular, focal BAC was more likely to lead to a false negative PET scan than the more widespread pneumonic form (in which most or all of an entire lung lobe is full of BAC).

Continuing along this theme, a Korean trial by Kim & colleagues (abstract here) reported on 9 BAC patients among 48 with lung cancer. In this population, the mean SUV was significantly lower among the BAC patients compared with those who had a non-BAC lung cancer.

(Click on image to enlarge)

Based on these results, the authors argued that BAC there is a problematic risk for false negative results (a study reported as negative but actually really having the disease) on PET scanning.

Overall, the results do suggest that while BAC can sometimes be identified on a PET scan, these cancers are disproportionately over-represented among the lung cancers that remain PET negative. Because of this, negative PET scan results should not be interpreted as ruling out cancer in people with CT findings suggestive of a BAC picture. Instead, this is a setting where we can't be reassured and need additional studies, potentially follow-up imaging or a biopsy, to clarify what's going on.