Drug trials and long-term side effects like PFS

A long-term medicine can be approved after only about 100 people have taken it for a year, and trials stop watching at the last tablet. So conditions like post-finasteride syndrome (PFS) often come to light only once millions use it.

That gap is filled by patients. Their reports to the medicines regulator are how lasting harms get noticed, and they fed the reviews behind today’s warnings on finasteride.

These conditions are more than sexual problems. Proposed criteria from 2022 say that with PFS, problems with thinking, depression and suicidality “can occur independently of any sexual difficulties”.

How much testing does a new medicine get?

Less than most people think. The international guideline for long-term medicines, ICH E1, expects about 1,500 people in total, and only 100 of them for at least a year.

The guideline is open about the limits. It says the safety review “is not expected to characterise rare adverse events”, meaning the rarer ones. Some serious side effects, it notes, “may occur only after drug treatment for more than 6 months.”

A study of medicines approved in the EU found that about 4 in 5 medicines for chronic use met even these minimum numbers. The rest were approved with fewer.

Why do lasting side effects show up so late?

Because trials rarely follow people after the last tablet. A trial asks whether a medicine works and what happens while people take it. When the trial ends, so does the watching.

The 2022 criteria put it plainly: “Clinical trials do not routinely include follow-up to assess the resolution of adverse sexual effects.” Nobody checked whether those effects went away.

A condition that starts or stays after stopping can’t show up in a study that has already stopped looking. It also takes time to link a symptom to a medicine a person took months or years before. Many people, and many doctors, never make that link.

Once a medicine is on sale, regulators rely on doctors and patients to report suspected side effects. Patterns in those reports lead to reviews, warnings and new leaflet lines.

Thalidomide, sold for morning sickness in the late 1950s, caused severe birth defects in thousands of babies. Diethylstilbestrol (DES), given to pregnant women for decades, was later linked to rare cancers in their daughters.

Vioxx, a painkiller, was withdrawn worldwide in 2004 after a trial showed more heart attacks and strokes. An FDA scientist, writing in the Lancet, estimated that 88,000 to 140,000 extra cases of serious heart disease had probably occurred in the US.

What happened with finasteride and antidepressants?

Both now carry warnings that came from reports, years after approval. Finasteride is used for hair loss and an enlarged prostate.

In 2024 the UK regulator, the MHRA, said sexual dysfunction with finasteride “has persisted even after treatment was stopped”. Nearly half of 426 UK Yellow Card reports of sexual problems recorded the person as not recovered. It also linked finasteride with depression and suicidal thoughts, and introduced a patient alert card.

For SSRI and SNRI antidepressants, the EU medicines regulator asked for a warning on every leaflet in 2019. It reads: “In some cases, these symptoms have continued after stopping treatment.” Patients call this PSSD. With PSSD, the 2022 criteria describe a lasting change in genital sensation, often numbness, and other symptoms such as emotional numbing and problems with thinking.

Nobody has measured how often either condition happens. That’s why every report counts.

What can I do?

These medicines help many people. If you have side effects, that’s a decision for you and your prescriber. For antidepressants, don’t stop suddenly. The finasteride leaflet says to stop and see a doctor if your mood changes.

If you’re about to start a medicine you’ll take for years, you can ask your GP a few plain questions.

If symptoms began on or after a medicine, write down the name, dose and dates, for example “Finasteride 1 mg from March 2021, last tablet June 2023”, and what changed.

Report it: it helps everyone

A doctor’s note helps you. A report to the medicines regulator helps everyone else: it is how patterns get noticed.

• No proof needed. A suspicion is enough, even years after stopping.
• Where: the Yellow Card scheme in the UK, MedWatch in the US.
• Use the right words. If symptoms continued after stopping finasteride or dutasteride, write Post 5-alpha-reductase inhibitor syndrome (MedDRA 10082430). After an antidepressant, write Post-SSRI sexual dysfunction (MedDRA 10086208). Post-accutane syndrome has no term yet, so say the symptoms continued after stopping.

Our reporting guide walks you through it

Our article on why reporting side effects matters explains what happens to a report.

What else do people ask?

Are medicines unsafe because of this? No. Approval means the benefits were judged to outweigh the known risks for most people. It doesn’t mean every long-term effect is known. That’s why monitoring continues after a medicine goes on sale, and why your report matters.

Can side effects really last after stopping? Official warnings now say so for finasteride, SSRIs and SNRIs. Read more in can side effects last after stopping? For some people it has lasted years. Nobody knows yet who will be affected, or whether it will ease.

Can I see the trial data for my medicine? Partly. The UK leaflet and the full prescribing information on the electronic Medicines Compendium list side effects seen in trials. Detailed trial data is often harder to get.

If something changed after a medicine and stayed, your report is how the next person gets a warning. Start with our PFS page or report it today.

Sources

1. ICH. E1: The extent of population exposure to assess clinical safety for drugs intended for long-term treatment of non-life-threatening conditions, EMA, step 5. Total “will be about 1500”; “Usually 300-600 patients should be adequate” for six months; “100 patients exposed for a minimum of one-year is considered to be acceptable”. The evaluation “is not expected to characterise rare adverse events” (under 1 in 1,000). Some serious effects “may occur only after drug treatment for more than 6 months.” International guidance applied in the EU, US and Japan.
2. Duijnhoven RG, Straus SMJM, Raine JM and colleagues. Number of patients studied prior to approval of new medicines: a database analysis. PLoS Med 2013. Medicines approved in the EU: “the minimal requirements are met by approximately 80% of new medicines approved for chronic use”.
3. Healy D, Bahrick A, Bak M and colleagues. Diagnostic criteria for enduring sexual dysfunction after treatment with antidepressants, finasteride and isotretinoin. Int J Risk Saf Med 2022;33(1):65-76. Proposed criteria from expert agreement. “Clinical trials do not routinely include follow-up to assess the resolution of adverse sexual effects.” Section 3 (PSSD): an enduring change in genital sensation; ancillary symptoms may include “emotional numbing” and “cognitive impairment”. Section 5 (PFS): cognitive impairment, depression and suicidality “can occur independently of any sexual difficulties”. Full text not re-opened on 22 October 2026 (captcha); wording as verified for other SIDEfxHUB articles.
4. MHRA. Finasteride: reminder of the risk of psychiatric side effects and of sexual side effects (which may persist after discontinuation of treatment). Drug Safety Update, 29 April 2024. UK.
5. EMA PRAC. New product information wording, PRAC recommendations on signals, 13 to 16 May 2019. EU leaflet wording for SSRIs and SNRIs.
6. Sibbald B. Rofecoxib (Vioxx) voluntarily withdrawn from market. CMAJ 2004;171(9):1027-1028. Withdrawal announced 30 September 2004.
7. CBC News. Study linking Vioxx to heart problems finally published, January 2005, reporting Graham DJ and colleagues in the Lancet: “An estimated 88,000-140,000 excess cases of serious coronary heart disease probably occurred in the U.S.A.” A modelled estimate.
8. Thalidomide scandal, Wikipedia, and National Academies, DES chapter (not re-opened on 22 October 2026, captcha).

Sources checked on 22 October 2026. AI-assisted writing and review, not clinical review.