To: Medicines and Healthcare products Regulatory Agency
6 April 2026
Please accept this letter from CAN-SG as a contribution to your review of the PATHWAYS trial, as outlined in your correspondence with King’s College London (KCL). In that letter you asked whether there were any further matters which impinge upon safety or efficacy that should also be considered alongside the issues set out in your letter.
We provide further information and observations regarding risks identified in your letter—particularly to fertility, bone, and brain development—while also highlighting additional risks not explicitly addressed. These include psychosocial development, sexual function, gynaecological problems, pregnancy risk, and the broader implications of cross-sex hormones on physical and mental health. We consider the impact of proposed protocol changes and outline significant concerns about the trial’s design brought up by points in your letter.
We also provide some references and two annexes, which we hope will provide useful information for your review of the trial and the risk-benefit assessment in view of the MHRA’s role regarding the approval of clinical trials of investigational medicinal products (CTIMPs) in the UK. Central to MHRA approval is a robust risk–benefit assessment, ensuring that participants are not exposed to harm without sufficient justification or likelihood of meaningful benefit.
Risks to Child Development
UK law places special emphasis on protecting children’s development. Developmental processes involve complex interactions between the growing child’s mind and body and the environment (experience expectant and experience dependant), are vulnerable in ways that stable systems may not be, and often time dependent. Adolescent sex hormones are integral to these developmental processes (Baker et al 2024)
Risks to development of fertility and sexual function, mental and physical health including bone health, cognitive and brain development are not short-term concerns but may have lifelong consequences. Such risks demand exceptionally strong justification, typically limited to serious or life-threatening conditions.
Gender incongruence does not meet this threshold. Therefore, exposing children to interventions that may impair their development requires clear and compelling evidence of benefit, which is currently lacking.
The proposed protocol changes—including raising the minimum age to 14—do not alter this fundamental concern.
Raising the age threshold will not significantly reduce risk for many children
While increasing the minimum age of participation may appear to reduce exposure and therefore risk, it is unlikely to do so significantly:
- New NHS proposals to delay cross-sex hormones until age 18 may extend exposure.
- Male adolescents will continue to take GnRHa if they start oestrogen, until or unless they have orchidectomy
- Individuals who identify as non-binary (a growing cohort according to the Cass Review) may remain on GnRHa indefinitely.
Thus, raising the age threshold does not substantially reduce duration of treatment or cumulative risk for many.
Risk–Benefit Balance
Fertility
The likely pathway from puberty blockers to cross-sex hormones carries a high risk of permanent infertility. Clinical trial regulations say trials should not proceed if there is a risk of disabling injury. Permanent infertility is a disabling injury.
Mitigation strategies such as fertility preservation are inadequate:
- Not feasible at early stages of puberty i.e. Tanner stage 2 and 3. (Laidlaw et al 2025) (MHRA letter to KCL said this just applied to T2 but it applies to T3 also.)
- Invasive, distressing, and rarely accepted. Worldwide studies show virtually zero uptake in girls
- Limited success rates
- Misleads girls that fertility preservation will enable them to “have children in future” (in the words of the Participant Information Sheet) when GnRHa followed by cross sex hormones likely to make that impossible due to effects on reproductive system.
- “Pausing” GnRHa to allow development of gametes (as suggested by lead investigator Emily Simonoff in BMJ) unlikely to be acceptable to trial participants as it will entail progression of puberty, the very thing they are taking GnRHa to avoid.
- Raising age to 14 will still include many in T2 and 3 due to variable onset of puberty
Human Rights
The MHRA letter raises concerns about human rights. Fertility loss constitutes a serious and irreversible harm, raising ethical and human rights concerns. Protection against harm to fertility is a fundamental human right as reflected in Article 8 “respect for family life” and Article 12 “right to found a family” of the European Convention on Human Rights.
The ethical precepts encapsulated by the concept of “a child’s right to an open future” explained in the article below make it unlikely a parent could consent to their child’s future infertility unless for exceptional reasons such as cancer treatment. (Jorgensen et al 2024)
Referring to Gillick Competence (see MHRA letter) is inappropriate. Gillick Competence does not apply to a CTIMP: in a CTIMP the child assents and the parents or guardians consent. Even using Gillick as a guide, the child would need to “fully understand the nature of the treatment, the options, the risks involved and the benefits.” And “…be capable of making a reasonable assessment of the advantages and disadvantages of the treatment proposed, so the consent, if given, can be properly and fairly described as true consent.” (Gillick v West Norfolk & Wisbech AHA, 1986)
Children and adolescents are unlikely to fully comprehend or weigh the long-term implications of infertility. Evidence shows that desire for children often increases with age, underscoring the difficulty of informed consent in this context. In the UK the average age of first birth is 29, which suggests that most adults are not thinking about having children until their late twenties.
Sexual and Reproductive Development
Puberty blockers disrupt normal sexual development (Laidlaw et al 2025):
- Underdevelopment of sexual organs.
- Potential impairment of sexual function and sensation.
- Unknown effects on psychosexual development.
These risks are not addressed in the protocol or participant information materials.
Gynaecological Risks
Evidence suggests that GnRHa use in later puberty may increase adverse effects, including emotional symptoms and vasomotor disturbances. (Jorgensen 2026)
Evidence-based strategies to reduce adverse effects of GnRHa in gynaecological practice are directly relevant, and their omission from the PATHWAYS protocol may expose participants to foreseeable and preventable harm.
Pregnancy Risk
The trial requires some participants to switch to less effective and less acceptable contraceptive methods, increasing the risk of unintended pregnancy.
This raises significant safeguarding concerns and conflicts with trial exclusion criteria: “Individuals of child-bearing potential who are at risk of pregnancy during the trial”
Barrier methods are less acceptable and less effective than hormonal contraceptives, even with optimal use, which is unlikely in pre-16 girls. One of the criteria for Gender Incongruence is “Strong dislike of sexual anatomy” making it even less likely that the cap or sponge will be acceptable. Girls’ ability to insist on condom use in partners is not reliable.
This is a safeguarding issue that the MHRA must take seriously. The “options” of termination or continuing with the pregnancy are both potentially traumatic.
Reviewing the contraceptives proposal is a responsibility of the MHRA as stated in 4 September REC minutes:
“The Committee was not assured that the current contraceptives proposal is suitable but recognised that this would form part of the MHRA’s review. Any request for action and/or information in this area which comes from the MHRA should be reflected in the protocol and all information sheets.”
Bone Health
The MHRA letter says that triptorelin may result in persistent and potentially permanent bone structural changes. There is no reason to think this is only likely beyond 12 months, and not before 12 months. Long term consequences of bone changes are under-studied.
Evidence shows that GnRHa treatment leads to reductions in bone density within the first 12 months.
Raising the entry age does not eliminate risk, as bone development continues into late adolescence.
- Evidence from GIDS EIS shows some participants to be osteopenic before 12 months.
- Withdrawal criteria based on harm detection do not prevent initial damage
- Withdrawal after possible irreversible harm detected is too late
Professor Michael Biggs published an analysis of the bone density data from the GIDS Early Intervention Study (EIS) at 24 months. (Biggs 2021).
He has provided us with unpublished analysis of bone density data from the EIS at 12 months which we provide in Annex A. This shows 12 month mean bone density Z score at osteopenic levels of -1. This means that for some children bone density was much lower than -1 by 12 months. MHRA could analyse the EIS data to understand how severely affected some children were by that point. It is not known whether bone density recovers with or without natural or exogenous hormones, or what the long-term osteoporosis and fracture risks are. These are the gaps in evidence that need to be filled, not more research telling us what we already know.
Brain and Cognitive Development
Adolescence is a critical period for brain maturation, influenced by sex hormones.
- GnRHa may disrupt neurocognitive development.
- Animal studies suggest possible permanent damage
- Proposed monitoring measures, such as MRI scans, are insufficient:
- Not all trial participants will have scans, just those in CONNECT.
- Lack defined outcomes.
- Detect harm only after it occurs.
- In gynaecological trials of short-term GnRHa therapy, impairments in memory and learning have been reported by approximately 30% to 45% of participants (Jorgensen 2026)
- As PATHWAYS participants will remain on GnRHa for two years or longer during critical educational periods, sustained difficulty with memory and learning during this time could have lasting academic and vocational consequences.
The primary risk is disruption of neurocognitive developmental processes, which may not be detectable through imaging.
Please see Annex B for a critique of use of BOLD fMRI in this context written by a cognitive neuroscientist.
“BOLD fMRI has been used in studies of adolescents to investigate how treatments affect brain function and development. In these settings it serves mainly as an exploratory or mechanistic research biomarker, rather than a primary safety tool. BOLD fMRI signals are indirect measures of neural activity, show substantial individual variability, and lack validated thresholds that define harmful change, particularly in developing brains where activation patterns normally evolve with age.”
Additional Risks Not Addressed in MHRA letter
“Locking-In” Effect
There is concern that puberty blockers may influence identity development, potentially “locking in” what would otherwise have been a transient cross-sex identity, increasing progression to medical transition.
In the 2022 letter to NHSE, Dr Cass warned that “we have no way of knowing whether, rather than buying time to make a decision, puberty blockers may disrupt that decision-making process”.
Dr Cass highlighted this issue as a “critically important unanswered question” for researchers. This was raised by the REC. The researchers said they were “acutely aware” of the risk and it was a question the trial was designed to answer. Yet it does not appear in the protocol as a primary, secondary or exploratory objective. This represents a significant omission in trial design.
That the treatment under study could exacerbate or prolong the condition for which the treatment is being given suggests a serious risk of iatrogenic harm that the MHRA must include in its risk/benefit calculation.
Raising the entry age to 14 will not mitigate this risk. As Cass noted, “children and young people are on a developmental trajectory that continues to their mid-20s” and “the Review heard accounts from young adults and parents about young people who felt certain about a binary gender identity in teenage years and then became more fluid in young adulthood or reverted to their birth-registered gender”.
Risks of the Full Treatment Pathway
Given the high likelihood of progression from puberty blockers to cross-sex hormones, the relevant risk profile should include the likely treatment pathway.
Cross-sex hormones carry potential long-term risks:
- Mental health: This recent study based on a comprehensive 25-year analysis of all Finnish adolescents who received “medical gender reassignment” suggests such interventions do not improve mental health and may make it worse (Ruuska et al, 2026)
- Cardiovascular and metabolic effects.
- Increased cancer risk.
- Increased mortality.
These must be considered in the overall risk–benefit assessment.
Lack of Evidence of Benefit
Trial inclusion criteria assume potential benefit.
- Clinical Eligibility Criterion 5: Leading clinician considers there is a reasonable prospect of benefit from GnRHa for puberty suppression.
- Research Eligibility Criterion 1: Clinician in CYPGS believes GnRHa treatment for 2 or more years may be helpful to the CYP.
No basis on which clinicians could make such judgments of potential benefit is provided.
Systematic reviews, including the Cass Review, find no convincing evidence of benefit from puberty blockers in this population. The trial’s design is unlikely to resolve this:
- No proper placebo control – all receive the intervention
- Reliance on subjective short term outcome measures.
- Short duration unable to address long-term evidence deficiency
- Built in placebo/nocebo effect biasing results
It is implausible that a randomized trial will reveal benefits not suggested by existing observational evidence.
Trial Design Concerns
Trial design is relevant to MHRA risk/benefit considerations because a poorly designed trial exposes participants to risk without knowledge gain. There have already been many serious criticisms of the trial design, including submitted by us, so it is not necessary to reiterate them all here. The MHRA letter to the KCL investigators has raised some specific design issues:
Cohort Heterogeneity
The trial population is highly heterogeneous in:
- Age and pubertal stage.
- Mental health status.
- Neurodiversity, sexual orientation and social context e.g. adverse childhood events.
This variability limits the ability to draw meaningful conclusions and further weakens the study design. Raising the age to 14 will not meaningfully reduce the heterogeneity due to variation in timing of onset of puberty.
Lack of Clear Rationale
Objections to the MHRA suggestion of raising the trial entry age to 14 include that it undermines the rationale for the trial. This exposes the lack of clear scientific rationale linking the intervention to the condition being treated. Without this, the research question is poorly defined, and the trial risks exposing participants to harm without meaningful scientific gain.
The protocol describes three possible rationales for use of GnRHa based on previous practice, but does not state which rationale the trial is using:
- Relief of distress from development of secondary sex characteristics
- Improved adult wellbeing from prevention of development of secondary sex characteristics allowing transition and better “passing” in adulthood
- Time to explore identity without development of secondary sex characteristics
We are informed that the trial is not studying effects of “transitioning” and for that reason the word “transition” was removed from the Participant Information Sheets (see REC minutes). The protocol includes outcome measures for “distress” and “dysphoria” but neither distress nor dysphoria is mentioned in the child or parent Participant Information Sheets. Given that the PIS is required to explain the purpose of the trial the failure to mention these calls into question whether relief of distress or dysphoria is the purpose.
Only the “time to explore” rationale appears in participant information materials, which suggests this is the purpose of the trial. Yet as a rationale it is very weak:
- “Exploration” suggests varied outcomes. But evidence from previous trials show this is highly unlikely as over 95% (98% in the GIDS EIS) continue in a transgender identification.
- The trial lacks measures to evaluate it
- Contrary to the MHRA letter, treatment choice post-trial is not a secondary endpoint and does not appear in the statistical analysis section.
- No rationale is provided for how adolescent identity exploration could be facilitated by an absence of adolescent sex hormones, which are integral to adolescent neurocognitive and psychosocial development. (Baker et al 2024)
Conclusion
Proceeding with the trial risks exposing children to harm without sufficient justification and fails to meet the standards required by clinical trial regulations.
- Significant and potentially irreversible risks.
- No convincing evidence of benefit.
- Absence of clinical equipoise.
- Weak and inadequate outcome measures.
- Fundamental flaws in design and rationale.
It is not ethical to subject adolescents to hormonal interventions in a research trial unless the state of the evidence suggests a favourable risk/benefit profile for the studied intervention, and the researchers have well- grounded confidence that the foreseeable risks and burdens have been adequately assessed and can be satisfactorily managed. The state of the science does not support a favourable risk/benefit profile, nor does it give researchers a basis for confidence that the risks of hormonal intervention can be satisfactorily managed. None of the suggested protocol changes addresses these core issues.
Dr Louise Irvine, Clinical Advisory Network on Sex and Gender April 2026
References
Baker, Amanda & Galvan, Adriana & Fuligni, Andrew. (2024). The connecting brain in context: How adolescent plasticity supports learning and development. Developmental Cognitive Neuroscience. 101486. 10.1016/j.dcn.2024.101486.
Biggs, Michael. “Revisiting the effect of GnRH analogue treatment on bone mineral density in young adolescents with gender dysphoria” Journal of Pediatric Endocrinology and Metabolism, vol. 34, no. 7, 2021, pp. 937-939. https://doi.org/10.1515/jpem-2021-0180
Jorgensen SCJ, Athéa N, Masson C. Puberty Suppression for Pediatric Gender Dysphoria and the Child’s Right to an Open Future. Arch Sex Behav. 2024 May;53(5):1941-1956. doi: 10.1007/s10508-024-02850-4. Epub 2024 Apr 2. PMID: 38565790; PMCID: PMC11106199.
Jorgensen SCJ. Safeguarding Gaps in the PATHWAYS Trial: Insights from GnRH Analogue Use in Gynecology. J Sex Marital Ther. 2026 Mar 11:1-8. doi: 10.1080/0092623X.2026.2641107. Epub ahead of print. PMID: 41811248.
Laidlaw MK, Lahl J and Thompson A (2025) Fertility preservation: is there a model for gender-dysphoric youth?. Front. Endocrinol. 16:1386716. doi: 10.3389/fendo.2025.1386716
S.-M.Ruuska, K.Tuisku, T.Holttinen, and R.Kaltiala, “Psychiatric Morbidity Among Adolescents and Young Adults Who Contacted Specialised Gender Identity Services in Finland in 1996–2019: A Register Study,” Acta Paediatrica (2026): 1–9, https://doi.org/10.1111/apa.70533.
Annex A: Mean bone density results at 12 months from the Tavistock GIDS Early Intervention Study
The bone density results from the GIDS EIS were analysed by Professor Michael Biggs in a letter published in the Journal of Pediatric Endocrinology and Metabolism analysing the results of the GIDS EIS on bone density at 2 years. In it he noted that “after 2 years of GnRHa, up to a third of patients had abnormally low bone density, in the lowest 2.3% of the distribution for their sex and age. A few patients recorded extremely low values, in the lowest 0.13% of the distribution.”
Unpublished analysis of the same data, which Professor Biggs has shared with us (see below), show serious reduction in bone density in some of the children at 12 months, with the mean bone density at or below -1, which is osteopenic. The mean baseline bone density was low (perhaps due to co-existing eating disorders or other health or lifestyle problems) but dropped at 12 months to the osteopenic range.
The graphs below are extremely worrying and don’t give grounds for confidence in the safety of GnRHa even by 12 months. They indicate that all children given GnRHa are at risk of significant reduction in bone density, into the osteopenic range, at 12 months. If the mean is -1 that means that several children would have bone densities much lower than that, in the osteopenic or osteoporotic range. Rather than put children at risk again the data from the EIS should be analysed to quantify the degree of reduction in bone density by 12 months for individual children.
Annex B Critique of BOLD fMRI as a safety measure
BOLD fMRI alone is generally not the best primary method to monitor cognitive safety. It is indirect, variable, and difficult to interpret at the individual level. It can be useful as a research biomarker but is weak as a standalone safety signal.
Key reasons:
- BOLD is an indirect physiological proxy, not cognition itself
BOLD measures blood-oxygenation changes linked to neural activity, not neuronal injury or functional impairment directly. Signal shifts can reflect: task differences or engagement, developmental change (especially in adolescents), scanner variability, motion, or preprocessing choices, vascular/metabolic effects unrelated to cognition. Therefore, detecting a “change from baseline” does not reliably indicate harmful cognitive impact.
- Poor specificity for individual clinical decisions
BOLD fMRI is for group-level inference in research, but: individual test–retest variability is substantial, there are no widely accepted clinical thresholds defining “adverse” BOLD change, normal developmental trajectories involve changing activation patterns. Using it as a withdrawal criterion risks false positives or false reassurance.
- Cognitive safety is better assessed functionally
From a safety perspective, regulators and clinical neuropsychology typically prioritize: Standardized neuropsychological testing (memory, executive function, processing speed, attention), Behavioral/educational outcomes, Patient-reported cognitive symptoms, Clinician assessment.
- Where BOLD can help:
BOLD fMRI may be reasonable as: an exploratory secondary endpoint, mechanistic research tool, hypothesis-generating biomarker, adjunct to cognitive testing. It is not usually utilised as a primary safety monitoring trigger unless validated for that specific intervention and population. For cognitive safety monitoring in a clinical trial, validated neuropsychological measures + clinical assessment are the standard.
Summary
BOLD fMRI, has been used in studies of adolescents taking medications such as stimulants, antidepressants, antipsychotics, and some hormonal treatments to investigate how treatments affect brain function and development. In these settings it serves mainly as an exploratory or mechanistic research biomarker, for example examining task activation or connectivity changes rather than a primary safety tool. BOLD signals are indirect measures of neural activity, show substantial individual variability, and lack validated thresholds that define harmful change, particularly in developing brains where activation patterns normally evolve with age. Because of this limited specificity and interpretability at the individual level, regulators and clinical trials typically rely on direct functional measures, standardized neuropsychological testing, clinical assessment, behavioral or educational outcomes, and symptom monitoring for cognitive safety, with imaging used only as a supplementary measure.