Seven-year-old David Hemmis of Rockville was having more than 100 epileptic seizures a day and had to wear a helmet to protect himself before Johns Hopkins Hospital doctors began giving him a drug called sodium valproate. Now his seizures have disappeared.

Six-year-old Michael Raymond in La Plata, Md., started taking the drug at Johns Hopkins last December. Doctors say he is now improving and hope that by increasing his dosage they can control his seizures too.

Now the Hopkins doctors are having to tell the parents of 25 children on the drug that the drug supply in limited. Without broader Food and Drug Administration clearance, the doctors say they can't keep giving it indefinitely.

The reason is that the FDA, under federal laws and regulations intended to protect the public from ineffective and dangerous compounds, has approved the drug's use only in experimental situations. Before it will authorize more widespread use, the FDA has damanded more test results.

The FDA's caution has spawned a nationwide protest by doctors, parents, and epilepsy study groups who say that the drug, sodium valproate, has been widely used in Europe for 10 years on about 200,000 patients with marked success.

A panel of experts assembled by the Epilepsy Foundation of America told the FDA on Feb. 9 that "further delays . . . would constitute callous negligence." The group urged the FDA's "immediate approval" of the drug for general use.

Both a congressionally created National Commission for the Control of Epilepsy and the FDA's own Advisory Committee on Neurologic Drugs - in short, a wide array of the country's leading experts on epilepsy - have agreed with that recommendation. They recommended last year that FDA approve the drug's use by any qualified doctor.

The FDA's position - as explained by spokesman Wayne Pines - is that no matter how many patients have been treated in Europe or the United States, "we need to see the result of two well-controlled trials" that prove, not just assert, that the drug is safe and effective.

Thus far, safety is not the issue, Pines said, since reports of serious adverse reactions so far have been few.

As to effectiveness, FDA officials have accepted the results of a study done in Japan of the drug in handling one type of seizure: petit mal or absence seizure, which is characterized by brief loss of consciousness and is common in children. But the FDA refused in December to accept the results of studies of the drug conducted at the University of Virginia Medical Center in Charlottesville. Those studies were jointly conducted by Dr. Henry Dreifuss, of the University, and Dr. J. Kiffen Penry, chief of the epilepsy branch at the National Institute of Health in Bethesda.

But another unusual circumstance in this case is that both the national epilepsy commission last August and the FDA advisory committee last October unanimously recommended approval on the basis of the same test results that the FDA found inadequate.

"It's very strange for FDA not to accept the results of one of its own advisory committees," said Dr. Penry at NIH.

Dr. Eli Goldensohn of Columbia University, chairman of the panel that came to a similar conclusion this month said, "The trouble is that the FDA is charged with keeping bad drugs out, not (letting) good drugs in.

"No one considers this drug very dangerous, though it's not perfect," he said. "And there's no question in my mind but that it's been effective in Europe." EPILEPSY, From C1>

Another basic trouble, said sources inside the FDA and out, is that no American drug company even started testing the drug in order to give the FDA the information it says it needs until 1974.

That tardiness is partly the result of the fact that drug companies are profit-making institutions and their interest in introducing a new drug can be influenced by the potential for sales, according to several authorities, including the FDA's Dr. Marion Finkel, associate director of new drug evaluation.

One study says it takes a firm at least five to seven years and $30 million to bring any new drug to the market.

"It's true there are hundreds of thousands of people in the United States who could benefit from this drug," said Dr. Penry, "but compared with the market for drugs for hypertension or antibiotics, that's relatively few."

The pharmaceutical company that is financing the research into the drug in an effort to win FDA approval for its distribution is Abbott Laboratories.

Abbott was the only one of 10 companies to say yes when the drug's French patent holder tried to find an American manufacturer.

Of his work with Dreifuss at Charlottesville, Penry said, "We've tried to help FDA by doing the studies which drug companies haven't been doing. "In the past two year," Dreifuss said, "We've given the drug to 60 patients, and it has been very effective."

In one group of 23 individuals who were switched between valproate and another drug - without their or their doctors or nurses knowing which they were getting at any time - valproate proved clearly superior, Penry said.

On Jan. 30 Abbott Laboratories gave the FDA more information on the latest results from these and other tests, including some in Florida. The FDA told Abbott at the time that the new Florida data may permit it to approve the drug for at least petit mal seizures.

"We expect to finish our review of this data and should have a decision by the beginning of March," said Dr. Finkel.

Last August - when the national epilepsy commission made its report to President Carter - its medical members said they had been assured that approval would probably be forthcoming by Jan. 1. So now many doctors are anticipating another delay or an approval that is too limited to be practical.

Also, many doctors think the drug should be approved for a far broader range of seizures. The Epilepsy Foundation panel has urged that the drug be approved as either the sole or adjunctive - meaning additional - treatment in petit mal seizures and adjunctive treatment in other seizure types.

Many doctors also deplore a current FDA ruling by which about 5,000 doctors may give valproate to no more than two patients each. FDA says it made this interim rule to help patients pending further reviews.

Pamela McGarvey, director of research administration for the Epilepsy Foundation, said, however, that "there are 2 million people with epilepsy in the U.S., only half of whose seizures are controlled by present drugs. This means that, at the most, 10,000 patients can benefit from this ruling."

"Valproate isn't a panacea," said Dr. John Freeman of the Johns Hopkins Seizure Clinic. "It is not a cure. But it is effective in many persons."

In LaPlata, Michael Raymond's mother said, "When we were at Hopkins last month, we were told they had just enough of the drug to last to March 1.

"Michael's doctor told us that if the FDA didn't approve the drug, he might be allowed to keep only two patients on it. He asked us how he was to decide which two to keep."