New five-year cancer survival statistics show that U.S. patients with the disease "are living longer now than ever before," the head of the National Cancer Institute reported yesterday.

The latest numbers, considered the most complete data of this kind ever compiled by the federal government, indicate that at least 46 percent and possibly as many as 50 percent of white patients diagnosed with cancer from 1973 to 1979 are "curable," said NCI director Dr. Vincent T. DeVita Jr. The figures were reported in a preliminary form to the National Cancer Advisory Board.

The survival rate is up from the 41 percent, five-year figure earlier reported for white patients diagnosed with cancer in the period 1967 to 1973. DeVita said the increase indicates that the medical profession is having more success curing cancer than anyone realized.

Curable is defined as those patients who do not die of cancer within five years after their disease is diagnosed. NCI officials said that except for cancers of the breast, prostate and kidney, five-year survival is a reliable, though not absolute, indicator of the probability of long-term cure.

DeVita stressed that the 46 percent figure was "conservative" and that further analysis of the survival data may well show that 50 percent of white patients survive for five years or more.

For all races and sexes combined, the five-year survival rates are slightly lower. DeVita said that the new figures show a 45 percent survival for all groups, compared to 40 percent in the earlier studies.

John Young, head of NCI's demographic analysis section, later noted that the numbers were not strictly comparable but that they were the only national figures available.

The new survival data is the first to come from the so-called SEER program, short for Surveillance, Epidemiology and End Results, which collects information about the occurrence and outcome of cancer patients in a 10 percent sample of the population.

The early survival statistics were taken from a smaller federal follow-up conducted largely, said Young, at university teaching hospitals, and may not be as representative. He said he didn't know if the survival rate of this group was higher or lower because it might include more desperately ill patients but they might have received better treatment.

DeVita said that the new SEER survival data independently "confirm and exceed" recent estimates he has made for congressional testimony.

Based upon a review of published reports, he had said earlier this year that about 45 percent of the 785,000 patients diagnosed with serious cancer in 1980 may be curable using combined treatments of surgery, radiation and drugs. These figures did not include early skin or cervical cancers, which have a higher cure rate.

DeVita said yesterday that he was "surprised" to learn from the new SEER statistics that the improved survival rates date back at least to 1973 cases.

"For the past decade, we have been more successful in treating cancer patients than we thought, because of the lag time between diagnosis and therapy and five-year follow-up," he said in a short report to the board.

A 5 percent increase in the survival rate would mean that an additional 320,000 patients diagnosed between 1973 and 1981 would survive cancer for five years. This would be in addition to the 2.6 million patients expected to survive at the 41 percent level, said an NCI spokesman.

DeVita contended that the gains are "real and that they are due to the progress we have made in the treatment of cancer since 1970." He cited improvements in the treatment of cancers such as testicular, Hodgkin's disease, rectal, and breast cancer for women under 50.

He noted that the 46 percent figure is considered "preliminary" because a complete five-year follow-up is currently available only for patients diagnosed in 1973 and 1974 and only for all cancer types combined.

Young said yesterday it will be six to eight months before the complete analysis by type of cancer and geographic location is finished. Detailed information on black patients, who have previously suffered from poorer cancer survival than whites, will also be included.

While others praised the latest figures, they were criticized by one advisory board member, Dr. William Powers of Detroit's Harper Grace Hospital, who said they were based on "less than adequate data."