The National Cancer Institute has little control over millions of dollars in federal research money and the private contractors who spend it, a new government audit has found.

The auditors found, for example, that:

The institute frequently pays little attention to the cost of various research projects. In one case, the agency gave a contract to care for test animals to a Maryland laboratory whose bid was $400,000 higher than an equally qualified competitor.

Officials at the cancer institute often do not make their own estimates of how much these projects should cost. The officials summarized, in just a few lines, the costs for a 10-year, $40-million contract to build cancer therapy facilities.

Many contracts for cancer research are filled with "nebulous requirements that are confusing or vague." For example, the institute awarded a $2.9 million contract to develop various antitumor agents without saying which chemicals would be needed and in what amounts.

The report by the inspector general of the Department of Health and Human Services examined 57 cancer contracts worth more than $190 million from 1978 through 1981. The report has not yet been made public.

The study was requested by Sen. Orrin G. Hatch (R-Utah), chairman of the Labor and Human Resources Committee, to determine whether any progress has been made since the inspector general sharply criticized the agency's contracting procedures in a 1978 report.

The HHS report is the final version of a draft study used by Hatch to criticize the cancer institute at Senate hearings last spring. The final report says many of the same contracting problems still exist although the cancer institute has made some improvements and has adopted a new plan for corrective action.

Paul Van Nevel, a spokesman for the cancer institute, said the allegations were "old stuff."

"All the deficiencies," he said, "have been corrected and all the recommendations have been made, although some of this won't be apparent until a subsequent review is done."

But the audit said that officials at the NCI, the National Institutes of Health and the Public Health Service have failed to follow up on a similar pledge they made in 1978 to review many of the cancer agency's procedures. "I don't know why it wasn't done earlier," Van Nevel said, "but the institute is now determined that it's going to have the cleanest contract operation in the government."

The billion-dollar agency has come under public scrutiny recently for failing to tell doctors and patients about the dangers of some experimental cancer drugs. The NCI director, Vincent T. DeVita, defended the total chemotherapy program at congressional hearings, but said that in the future his agency would try to report such side effects more quickly.

The new audit found that in more than half the cases examined, cancer institute officials relied mainly on cost estimates provided by the contractors themselves.

The institute provided few details, for example, for a $40 million plan by the University of California, University of Washington and Fox Chase Cancer Center to build facilities for a new form of cancer treatment called neutron therapy.

"There were no specific estimates for labor hours, construction materials and patient treatment costs," the report said. When a special federal review board later concluded that the staff salaries and patient costs were too high, the report said, NCI officials simply ignored the findings.

While the auditors found many of the cancer contracts too vague, they said that other contracts are written so narrowly that they tend to favor the same researchers year after year. In one such case, only one firm decided to challenge Litton Bionetics of Kensington when its $912,582 contract to maintain animals for cancer research came up for renewal.

The NCI contracts office refused to give information about the project to other potential bidders, including a Texas firm that offered to do the work for $400,000 less. Both firms were rated competent to do the work. However, the NCI later awarded the contract to Litton after deciding that moving the animals from the firm's Maryland laboratory might be harmful, a fact not mentioned in the original proposal.

In one case after another, the auditors said the cancer institute has given out contracts despite a negative rating by a team of peer reviewers, the outside experts hired by NCI to evaluate the scientific merit of each project.

A $600,000 survey of cancer morbidity at a research firm in Colombia was approved even after the peer reviewers found that it "suffered from inadequate attention to protection of human subjects."

When the peer reviewers flatly recommended that three contracts for community-based cancer centers be terminated, high-level NCI officials called a closed-door meeting, decided that this would hurt the program and overruled the experts without any explanation.

And a breast cancer study at the New England Medical Center was renewed for $149,989 after the reviewers found that doctors were not cooperating and the hospital was not including the right patients in the study. Robina Folland, an official at the medical center, said this might have occurred at other hospitals involved in the study but that her hospital had included more than the required number of patients.

In other cases, the report said, contracts have been wrongly classified as nonresearch projects to avoid having to undergo peer review.

Far too frequently, the auditors said, the institute has rushed to make hasty or incomplete awards, especially toward the end of the fiscal year. In one example, NCI officials allowed six prospective contractors to bill the government for $257,390 in costs, even while they were still discussing what work would be included in their contracts.

Some of the agency's contracting officers do not have enough scientific training to supervise highly technical contracts, the auditors said. In many cases, they said, these officers do not know what their contractors are doing and fail to move quickly to resolve the problems they encounter.

They are also hampered by excessive workloads, with 11 project officers assigned to supervise 276 cancer contracts. For example, one project officer with only a bachelor's degree in biology is responsible for 54 biomedical research contracts and serves as secretary of two highly technical cancer committees. As a result, this official was able to make only two or three on-site inspections a year for all 54 contracts.

One of the sites the official failed to visit was Stanford University, which later suspended the principal researcher on a $288,734 cancer study for "significant professional misconduct."

The inspector general'sreport also raised questions about some private researchers. The auditors found four principal researchers on cancer contracts who supposedly had committed more than 100 percent of their time to various federal research projects. One of the researchers, a physician, devoted as little as 12 days a year to his cancer contract without the knowledge of the NCI.

In another case, the Centre National de la Research Scientifique in Paris won a noncompetitive contract by listing a prominent scientist as its principal researcher, even though the scientist was working for another firm and had not yet been offered the job. This kind of practice, the auditors said, could lead to "a scientific 'bait and switch' routine where a recognized scientist could obtain an award, but the research would be conducted by a lesser-known and -qualified associate."

The center, hired to study a form of nasal cancer, was supposed to submit financial reports and vouchers on a monthly basis. But the auditors found that the Centre National never handed in a single report or voucher and that NCI officials never requested the information. Instead, they paid the center the full $382,140 without asking any questions.

Officials at the French research center could not be reached for comment.

NCI spokesman Van Nevel said the institute now insists on more detailed and timely contract proposals and requires a special review of all contracts over $500,000. He said NCI officials now must make independent estimates of the cost of each project and will not allow expenditures without written approval.

Van Nevel also said the institute has issued new guidelines for peer review and has moved to reduce the workload of its project officers. All contractors will be asked for regular reports on their finances and how they are spending their time, he said.

Some of these reforms have fallen short in the past. In fiscal 1979, the cancer institute's financial services branch looked at contract proposals and identified $10 million in questionable or unsupported costs. But the branch had no time to review more than half the proposals submitted that year, amounting to $52 million.

The auditors also criticized one NCI official for improperly releasing technical and financial information about upcoming contracts to two trade publications, which they said "may have given some potential contractors an advantage and harmed the government's position in negotiations."