Thalidomide and the cost of weak drug safety

International sales and the U.S. safety review

By the time the United States considered thalidomide, it was already available in dozens of countries. It had been sold as a sleeping medicine, and clinicians also used it for nausea during pregnancy.^[2] That second use exposed a different question: evidence about an adult patient's immediate response could not settle whether a developing fetus would be harmed. The later disaster was international, but the decisions about selling, studying and withdrawing the medicine occurred through different national systems.

On 12 September 1960, the William S. Merrell Company submitted a U.S. application for Kevadon, its proposed thalidomide brand.^[1] Frances Oldham Kelsey, a new FDA medical officer, was assigned the safety review. The existing 1938 law already required evidence of safety before a new drug could be marketed. Under that system an application could become effective if the agency did not act within the permitted period, making an active review consequential.

FDA's account of Kelsey's work describes repeated company pressure and repeated demands for scientifically reliable evidence.^[1] She did not accept international availability as sufficient proof. The National Library of Medicine's biography also records concern about adverse effects in patients who took the drug repeatedly.^[3] Her decision was to keep an inadequately supported application from becoming effective; it was not a claim that she had already reconstructed every fetal effect before clinicians reported the birth-defect pattern.

Kelsey’s review illustrates a decision rule rather than a retrospective prediction: the company had to supply adequate evidence, and unresolved safety concerns were grounds to keep asking for it. The later identification of fetal harm demonstrated the value of that scrutiny. It does not mean the agency needed to know every possible injury before requiring a stronger application.

The emerging clinical warning

During 1961, doctors in Germany and Australia linked thalidomide exposure with severe congenital abnormalities. A primary document lets readers see a warning as it was being communicated: William McBride's short letter in The Lancet reported unusual abnormalities among babies whose mothers had received thalidomide during pregnancy, as a sedative or for nausea.^[4] He asked whether other readers had observed similar cases. The letter made an emerging clinical signal available for comparison beyond his own practice.

McBride's reported percentage belongs to the patients he described. It is not a measured risk for every pregnancy exposed to thalidomide, and the letter alone is not a global case count. Its accompanying editorial note recorded that Distillers had withdrawn its thalidomide preparations pending investigation.^[4] FDA's retrospective history places the German withdrawal in November 1961.^[2] Together, these documents show warnings and withdrawals developing before the later historical accounts could describe the full scale of the disaster.

Investigational exposure despite withheld approval

Kelsey's review prevented the original Kevadon application from authorizing a general U.S. market launch. It did not mean that no American patient received the drug. FDA's history of its drug centre records that Merrell distributed more than two million tablets for investigational use while that route was largely unchecked by the law and regulations.^[5] Once the harm became known, FDA sought to recover supplies from physicians, pharmacists and patients. A count of tablets is a measure of distribution, not a count of patients or injured children.

The distinction matters when assessing what was prevented. The unsupported application remained outside the general market, while investigational exposure revealed another weakness in protection. An exact number of births prevented would require knowing how the drug would otherwise have been sold and used. The sources here do not establish that counterfactual. They do establish both the blocked market application and exposure despite it. FDA describes thousands of affected babies internationally; that wording should not be turned into a complete count of every affected pregnancy, death or miscarriage.

Changing the evidence and approval process

On 10 October 1962, President John F. Kennedy signed the Kefauver-Harris Drug Amendments.^[6] Manufacturers now had to demonstrate effectiveness as well as safety before marketing. FDA's centennial account identifies another important change: the agency had to give specific approval, rather than an application becoming effective through the passage of time.^[6] The amendments also strengthened control of clinical investigations, including informed consent, and required adverse-event reporting. The response addressed the quality of evidence and the process through which patients received experimental drugs.

The reform effort preceded the public crisis: Kefauver's hearings on pharmaceutical practices began in 1959, with a bill following in 1961.^[6] FDA's history describes thalidomide as helping revive that effort. The law grew from several years of investigation and political work. Attributing every change to a single reviewer or one revelation would erase that sequence.

A later approval under different conditions

Nor was thalidomide permanently excluded from U.S. medicine. Its 1998 Thalomid label documented a narrowly specified use for skin manifestations of erythema nodosum leprosum, a complication of leprosy, under a special distribution programme.^[7] The historical label combined that use with prominent warnings about fetal harm and registration requirements. It records a later decision under different conditions; it does not undo the earlier injuries or validate the original sedative application. Reading it alongside the 1961 warning and the 1962 reforms shows why a drug's indication, evidence and controls belong in the same history.