This is a picture of Josh Hardy. You might recall seeing his face on TV or on the #savejosh hashtag on Twitter a few years back. Josh, who was 7 at the time, had undergone a bone-marrow transplant for kidney cancer that left him hospitalized with a severe infection. He was dying.

There was an experimental drug called brincidofovir made by a small company called Chimerix that his doctors thought might work. But the company declined their repeated requests to provide it. His parents, being smart and well educated, rallied friends, who rallied their friends and their friends and so forth, until it seemed as though the entire Internet were behind them.

The ending was a good one: The company gave Josh the drug, and it worked and he got to go home.

But not all cases like this go so well. A powerful report from STAT this week provides a heartbreaking reminder that the reason experimental drugs are not available for anyone to use is because they are just that — experimental. And the chances that things will go wrong are as strong as that they will go right.

The report focuses on a highly controversial treatment for rare cancers being promoted by Texas doctor Stanislaw Burzynski and how, from 2011 to 2016, 37 members of Congress wrote to the Food and Drug Administration about his work.

Many of those letters asked the agency to give their constituents “compassionate use exemptions” to try his unapproved drugs. But what about the other letters?

They were reports about patients having bad reactions — or even dying — from his treatments. Burzynski has been cited by the FDA in “dozens of cases” of this sort, the science and medicine news site reported.

For example, Sandy Smith said that in 2007, her son Andrew was being seen at Burzynski’s clinic until she began looking into cases of other children who were treated for the same brain condition, DIPG, or diffuse intrinsic pontine glioma. She discovered that they all died. She pulled her son out and took him to the National Institutes of Health.

“Andrew lived another two years, an unusually long amount of time for a child with a DIPG diagnosis, and longer than any child who stayed with Burzynski survived,” Stat wrote.

Another boy, 5-year-old Nolan Hancock, died three months after he started Burzynski’s treatment.

These cases should serve as cautionary tales to parents to have realistic expectations about compassionate use and to beware of supposed miracle cures.

Read more:

Like our Health & Wellness page on Facebook for more news about the ins and outs of the human body and mind, essays and advice. You can sign up here for our newsletter.