FNA Dropped From RDOG Study Due To High Rate of Insufficient Samples

August 1, 1996

DALLAS--Fine-needle aspiration (FNA) breast biopsies in women with mammographically apparent, nonpal-pable breast lesions were deemed impractical due to the high rate of insufficient samples in a Radiologic Diagnostic Oncology Group (RDOG) study, and were stopped 19 months into the trial, Etta D. Pisano, MD, said at the American College of Radiology (ACR) National Conference on Breast Cancer.

DALLAS--Fine-needle aspiration (FNA) breast biopsies in womenwith mammographically apparent, nonpal-pable breast lesions weredeemed impractical due to the high rate of insufficient samplesin a Radiologic Diagnostic Oncology Group (RDOG) study, and werestopped 19 months into the trial, Etta D. Pisano, MD, said atthe American College of Radiology (ACR) National Conference onBreast Cancer.

The multicenter study has to date enrolled 5,000 women with specificmam-mographic abnormalities who would ordinarily have had an openbiopsy procedure. The researchers expect to continue to enrollpatients in the study for at least one more year.

The women were initially randomized into four biopsy groups, toundergo either stereotactically or sonographically guided coreneedle biopsy with or without FNA. Some institutions were notequipped to provide ultrasound-guided biopsies, and women at thesecenters were randomized to stereotactic FNA plus core or stereotacticcore biopsy only.

Crossover was allowed if the lesion was inaccessible by the assignedsystem. "An example might be a cluster of calcificationsthat could not be seen by ultrasound or a lesion that was inaccessibleby stereotactic biopsy by virtue of its location far back in thechest wall or close to the nipple," Dr. Pisano said.

Not a Report of Accuracy

Accuracy was determined by either open surgical biopsy or follow-upmammography at 6, 12, and 24 months. "This is not a reportof the accuracy of either FNA or core needle biopsy," Dr.Pisano said, "only the insufficient sample rate seen withFNA."

A preliminary analysis of 398 women undergoing FNA across allinstitutions showed that 138 samples (35%) were insufficient,said Dr. Pisano, of the University of North Carolina School ofMedicine. Sample insufficiency was determined by two expert cytopathologistswho reviewed all FNA specimens: Dr. William J. Frable of VirginiaCommonwealth University and Dr. Nour Sneige of the Universityof Texas M.D. Anderson Cancer Center.

In July, 1995, after reviewing the data, an Independent Data Safetyand Monitoring Board recommended withdrawing FNA from the trial,and the procedure was dropped from the study at all institutionsin December, 1995. [See the box below for a dissenting opinionfrom a breast cancer surgeon.]

"It was felt that across all centers, FNA was impracticalfor the workup of mammographically apparent, nonpal-pable lesions,"Dr. Pisano said. The study continues regarding core needle biopsy,she added.

Dr. Pisano noted that there was some interesting variability inthe insufficient sample rate. The rate was significantly higherfor clustered calcifications than for masses (47% vs 28%, respectively).

When the cytopathologist was present for the procedure, the insufficientsample rate was significantly lower (19%) than when the cytopathologistwas not in the room (39%). However, Dr. Pisano pointed out that"the 19% rate is still too high."

The degree of suspicion had some value in predicting insufficientsample rate, with the more suspicious lesions tending to be lesslikely to be insufficient. "This is not too surprising giventhe fact that cancers tend to shed cells more readily than benignlesions," she said.

Dr. Pisano noted that before the centers were allowed to participatein the study, they must have performed at least 50 breast FNAs."There were a few institutions that seemed to perform well,"she said, and these centers will continue to do FNA biopsies,outside of the clinical trial.

FNA Insufficient Sample Rates: A Surgeon Disagrees

By Monica Morrow, MD

Fine-needle aspiration (FNA) is a well-established technique forthe diagnosis of palpable breast masses. Operator experience andhistology of the lesion being aspirated have both been shown toinfluence the insufficient sample rate.

A malignant FNA allows definitive treatment planning for palpablemasses, since 98% of palpable carcinomas are invasive tumors.

Its utility for mammographic lesions, where a high proportionof carcinomas will be intraductal tumors, is less clear, sinceFNA does not reliably distinguish between invasive and in situcarcinoma. Thus, an FNA diagnosis of malignancy does not allowa determination of the need for axillary node dissection.

The high rate of insufficient cytology samples in the RDOG studyprobably reflects lack of experience with the technique and agreater interest in core biopsy, since large FNA series in Europehave shown very low insufficient sample rates.

For example, Azavedo et al (Lancet 1:1033-1036, 1989) reportedonly 49 insufficient specimens in 2,594 stereotactically guidedfine- needle aspirates (1.9%), indicating that the technique canbe successfully applied to mammographic abnormalities.

[Dr. Morrow is associate professor of surgery and directorof the Comprehensive Breast Program, Northwestern University MedicalSchool.]