Why ignored comparisons, flawed expert evidence and restricted data continue to distort the HPT debate

The 1975 question that was never properly answered

In May 1975, Dr Annelise Pritsche, of the German health agency BGA, wrote to Dr Karl Knorr of the University of Ulm.  Dr Knorr was a recognised expert in women’s health.

There is no doubt that Dr Pritsche contacted Dr Knorr because of increasing publicity expressing concerns that HPT’s marketed as Duogynon in Germany and Primodos in the UK were causing birth defects.  That month, an article detailing the evidence had been published in the Sunday Times.  

Dr Pritsche asked Dr Knorr a very pertinent question: 

“Do you consider it possible that Duogynon injections can cause damage to the developing foetus which, when the injection does not lead to miscarriage, it can cause malformations?”

Hormone pregnancy tests and the known risk of miscarriage

This question was asked in the context of the by then widespread knowledge that HPT’s which had been introduced in the 1960s were associated with miscarriages or threatened miscarriage.  There was evidence that some women were deliberately taking overdoses with the intention of causing an early abortion.  Against this background, it could be thought a fairly obvious speculation that if the HPT did not cause a miscarriage, nonetheless there was damage which became evident when the surviving foetus was born.

Misoprostol and the emergence of a recognisable pattern of defects

We do not know what answer, if any, Dr Knorr gave to this enquiry.  However, we do know now, with confidence, that the answer should have been in the affirmative.  From the 1980’s onwards, Misoprostol was being used to induce early abortion, described as the morning-after pill.  However, the use of Misoprostol did not always have the intended result.  Children born after the use of Misoprostol in early pregnancy were noted to have birth defects.  There was a pattern to these defects which was described in a comprehensive review in 2002 by Philip and others.  This created an opportunity to assess the controversial question of whether HPT’s were teratogenic by reference to a comparison with the defects caused by Misoprostol.  

The Tummler Study (2014): Striking findings, swiftly discounted

The first occasion upon which this issue appears to have been considered was in 2014, in a paper published by Tummler and others.  This paper was published with financial support from the German Federal Institute for Drugs and Medical Devices.  The data analysed was of 296 case reports of defects where the mother had been known to take an HPT.  They found an increase in some birth defects.  Most significantly, there was a 37-fold increase in bladder exstrophy, a rare defect.  They described this as “a most remarkable result.”  However, they discounted it with the opaque comment that their findings “may be based on a selection bias”.

Interestingly, the authors were alive to the relevance of a comparison of the birth defects observed in the group studied with those caused by Misoprostol.  They explained that the possible abortifacient effect of HPT’s could be relevant.  However, they stated that there was no evidence of an increase in bladder exstrophy associated with Misoprostol.  This statement was inaccurate.  A study by Orioli and others in 2000 showed that there was an increase in bladder exstrophy associated with Misoprostol at a similar level to that found in the Tummler paper.  It appears that the authors of the Tummler paper did not consider the Orioli paper, nor the Philip paper, neither of which are listed as references.

The MHRA Expert Working Group: Data without comparative analysis

In 2017 an Expert Working Group reported to the MHRA on the association between  HPT’s and birth defects . As part of the research for this Report a dataset of 244 cases of birth defects where an HPT had been taken was assembled . Although limited analysis of this dataset was undertaken surprisingly no consideration was given to a comparison with the misoprostol data . This was all the more remarkable as there was clear evidence that certain defects including limb reduction and cardiac were clearly more prevalent in each group .

Scientific Advice to the CSM: Relevance not recognised

When the work by Professor Danielsson and others was referred to the Committee on Safety in Medicines in 2023, the scientists advising the committee did not undertake such exercise, nor (it would seem) appreciate the relevance of the same.

In 2019, the UK victim group began litigation, which was subsequently dismissed when they were abandoned by their lawyers, Pogust Goodhead.  However, in the course of litigation, expert evidence was served on behalf of the manufacturers, Bayer, and the Government.  Professor Anthony Scialli inaccurately stated in a written report that Misoprostol did not cause limb reduction defects.  

Expert evidence was also served from a Professor Friedman who commented upon the lack of evidence in the HPT group of a recurrent pattern of multiple congenital anomalies, as was seen with Misoprostol.  Whilst this is an accurate statement, it ignores the important point that the scientists advising the EWG did not analyse the pattern of multiple malformations.  Neither did the authors of the Tummler paper.

The pattern question which the MHRA will not answer

In 2025, a request was made on behalf of Professor Danielsson and other scientists who are continuing to research the presentation of defects in HPT cases, for sight of the raw data, which was considered by the EWG. This would reveal the pattern of defects to include multiple malformations.  This request has been refused by the MHRA.

Denial by design: Confirmation bias in regulation and industry

The response of the manufacturers and regulatory bodies to the assertion that HPT’s caused birth defects could be described as one of complex and sustained denial.  This process reveals clear evidence of confirmation bias.  Evidence is adapted to suit a favoured conclusion.  Lines of enquiry which might lead to a different conclusion are avoided or obstructed.

This issue will not go away.  An authoritative and unbiased review of the evidence is long overdue.