麻豆传媒

ATLANTA -- Thyroid nodules with atypia on repeat fine-needle aspiration (FNA) have different risks of malignancy depending on the type of genetic mutation, a retrospective cohort study found.

In over 200 patients who had atypia or follicular lesions of undetermined significance on repeat FNA, mutations in BRAF and PAX were associated with the highest risk of malignancy while mutations in KRAS represented lower-risk lesions, David S. Cohen, MD, from Southern California Permanente Medical Group in Harbor City, reported here.

As such, molecular testing of indeterminate nodules should be considered to help guide surgical decision in the future, said Cohen.

"I think that molecular profiling is a useful adjunct in terms of determining malignancy risk in certain patients," Cohen told 麻豆传媒. "That's important in terms of preoperative counseling. Whether the patient has an 80% or 90% chance of cancer versus something that may be lower may go into consideration for potentially what treatment is recommended."

As presented at the American Academy of Otolaryngology-Head and Neck Surgery Foundation (AAO-HNSF) meeting, his group studied 230 patients with thyroid nodules from Southern California Kaiser Permanente hospitals. All patients had atypia or follicular lesions of undetermined significance (Bethesda III) on two separate FNAs, had positive ThyGenX testing, and underwent thyroid surgery. Patients with microcarcinoma were included if it was on the side of the nodule that prompted FNA.

The prevalence of malignancy overall was 73.9% (170 patients). All 69 patients with a BRAF mutation, 85% of those with a PAX mutation (11 of 13 patients), 71% with an HRAS mutation (29 of 41 patients), and 63% with an NRAS mutation (51 of 81 patients) had a malignancy. Conversely, the rate of malignancy was only 33% (8 of 24) in those with a KRAS mutation. The difference in the percentage with malignancy between mutations was significant (P<0.01).

All 69 patients with a BRAF mutation had papillary cancer. Of the 11 malignancies in patients with a PAX mutation, 4 had follicular and 7 had papillary morphology. Malignancies in patients with HRAS and KRAS mutations were predominantly papillary (86% and 88%, respectively).

When assessed by mutation subtype, the malignancy rate was higher among patients with an HRAS Q61K versus a Q61R mutation, although the difference was not significant (91.7% versus 64.3%, P=0.12). The rate of malignancy for an NRAS Q61K versus a Q61R mutation was also not significant (69.2% versus 61.8%, P=0.78).

"An important point is that the generalization of the Bethesda FNA III classification, usually thought to be somewhere between a 5% and 15% rate of malignancy, in the age of molecular testing, is probably obsolete and not accurate," said Cohen. "Also importantly, not all positive molecular mutations carry the same risk. Just because you see that it says 'positive,' you shouldn't necessarily think 'cancer.'"

With a very high-risk mutation, such as BRAF or PAX, surgery may be recommended, he said, whereas patients with a KRAS mutation could potentially be observed or undergo lobectomy only.

"There are multiple things we look at in addition to molecular mutation positivity, even so much as the subtype of molecular mutation. On top of that, what is the overall clinical picture? Part of what our future studies will look at is synthesizing the ultrasound findings as well," he said.

"Another facet that's important in figuring out whether this can be implemented to personalize surgery for different patients is the negative predictive value," said Cohen. "It's important to make sure that those mutation-negative patients don't have a very high risk of cancer."