Almost all the women who have had breasts removed as the result of a federal cancer-screening program really had cancer, a federal advisory committee has decided.

Though it said two months ago that 53 women apparently had breasts removed needlessly or hastily because of misdiagnosis, yesterday the committee told the National Cancer Institute that the number of clearly needless operations was only seven.

There were 30 more cases in which the committee's expert pathologists and local pathologists disagreed, but Dr. Oliver Beahrs of the Mayo Clinic said yesterday these "reflect the fact that diagnosing cell abnormalities can be very complex," and there is often no unanimity.

Beahrs headed the working group named to look at the results to date of the five-year-old program to screen women at 27 medical centers.

He said the group's pathologists found that for their original review they had often been sent the "wrong slides" - that is, breast tissue samples that failed to show the actual cancers - or other faulty information. As a result, he said, they made their second study.

But one critic of the way the screening program has been conducted - Dr. Sidney Wolfe, head of Ralph Nader's Public Citizen Health Research Group - said the cancer institute and the screening projects it finances have still not done enough to review all the breast surgery done as a result of supposed cancers first found by the 27 centers. He said more scrutiny would turn up more errors and more need for "quality control" in every hospital.

De. Diane Fink of the cancer institute called Wolfe "right in saying there are still cases that should be reviewed," and "we intend to do so."

Beahrs said his group's first review - 462 cases of supposedly minimal or very small breast cancers - determined that 66 were actually "benign" or non-cancerous.

But now, he said, it has been found that:

Two case reports were distorted by computer error.

Eleven women had only breast biopsies and no further surgery.

Sixteen women had "frank" or unmistakable cancers.

Another 30 were operated on in only a cautious two-stage procedure, with a biopsy one day and a breast operation later, often after review by a second pathologist. In these there was "almost a 90 per cent agreement" between local pathologists and detection center reviewers that the lesions were cancerous, though the Beahrs pathologists still disagree.

In only seven cases is there "real question as to whether the treatment was appropriate," and "we still need more information" on some of these.

Beahrs, Wolfe and Fink agreed that much of the problem is that mammography - breast X-raying - is turning up many very small or borderline lesions with which many pathologists are unfamiliar. They called for caution and no rush to surgery in such cases.

Another advisory panel has recommended that lesions of this kind "be reviewed by at least two pathologists" before any surgery.