A pilot study proposing an algorithm for pubertal induction in cerebral palsy

Study design

A two phase study conducted in patients with CP who have delayed puberty:

1. Retrospective review of clinical records and interviews with patients (or their carers) who have been treated with sex-steroids.

2. Prospective interventional trial of pubertal induction with a proposed algorithm in a cohort of CP patients (separate from phase 1).

Outcomes

Phase 1 outcomes include incidence of hypogonadism and delayed puberty in children with cerebral palsy, types of sex-steroid use and experience of those who were treated with sex-steroids. Phase 2 outcomes included tolerability of pubertal induction (measured by adverse events and progression to different tanner stages) and bone mineral density gain.

Participants

Phase 1: Participants were included if identified from either of the two tertiary paediatric hospital databases (up to 2018) as having CP and pubertal delay. Those treated with sex-steroids were invited to participate in a telephone interview about their experience.

Phase 2: Participants were recruited from outpatient clinics from January 2017 to December 2021 if they were diagnosed with CP (4 or 5 on the Gross Motor Function Classification System (GMFCS) [16] or level 2–3 if they had other complex needs such as multiple anticonvulsant use and were either pre-pubertal or had pubertal arrest. The cohort included girls >13 years and boys >14.5 years, who had absent onset of breast development in girls, testicular enlargement >4 mL in boys or no progressive pubertal change over at least 6 months. Exclusion criteria included use of oral glucocorticoids in the preceding 12 months, current vertebral fractures >Genant 3 found on spinal X-ray at screening, use of bisphosphonates or other bone modifying drugs in the last two years, contraindication to androgen or oestrogen use due to medication allergy, previous deep vein thrombosis or stroke related to coagulopathy. Both Institutional Ethics Committees approved the study.

Data collection and clinical measures

Phase 1 GMFCS was obtained from medical records. Participants were divided into two groups: predominantly ambulatory (GMFCS I–III) or non-ambulatory (GMFCS IV–V). Minimal trauma fracture was defined either as a self-reported or radiologically proven fracture occurring after a fall from standing height or less or a minimal trauma incident other than a fall (e.g. turning over in bed). Data were collected concerning current or past use of anticonvulsants, use of percutaneous endoscopic gastrostomy (PEG) feeding, age of commencement of sex-steroid therapy, the medication used, dosage and side effects if treated, Tanner stage, age of menarche for females, bone age reported by a radiologist using the Greulich and Pyle (GP) method, FSH, LH, oestradiol/testosterone, dual-energy X-ray absorptiometry (DXA), using a GE Lunar Prodigy (Madison, Wisconsin, software version 17). Interviews with patients who had sex-steroid treatment focused on bone and sexual health, medication compliance, tolerability and QoL (see Supplemental Material, Appendix 1). This was performed by one interviewer (AT).

Phase 2 Pubertal induction was commenced at baseline in a separate cohort of previously untreated patients. Pubertal status and progress were assessed every three months as per algorithm (Supplemental Material, Appendix 2). The primary outcome measure was bone density gain as assessed by lumbar spine height adjusted BMD over 24 months using DXA. Secondary outcomes were a reduction in fracture incidence as detected on spinal X-rays and self-reported long bone fractures, altered QoL, incidence of adverse events, pain and mobility using questionnaires. DXA scans were performed on a Hologic QDR 4500 Horizon DXA scanner (Hologic Inc., Bedford, MA).

Phase 2 puberty induction intervention

Induction of puberty was performed according to a proposed algorithm (Supplemental Material, Appendix 2). Boys were induced with transdermal testosterone commencing as 25 mg (½ sachet) Testogel^® 50 mg sachets daily or 25 mg (2 actuations) of Testogel^® pump, increasing slowly to 50 mg/day at 6 months, to a maximum of 100 mg/day at 12 months, to mimic normal pubertal progress. Girls were induced with transdermal oestrogen commencing at 0.5 mg oestradiol, increasing to 1 mg at 6 months to a maximum of 2 mg at 18 months or half a Climara^® 25 patch weekly, to a maximum of Climara^® 50 (3.8 mg oestradiol) patch at 18 months, to mimic normal pubertal progress. All dosing changes were based on clinical progress of puberty at each visit.

Patient characteristics

For phase 1, 20 patients identified with delayed puberty were included in the clinical review, 14 female and six male. Mean BMI was 18.2 ± 6.4 g/cm². Sixteen of the twenty (80 %) were non-ambulatory, with 50 % having a GMFCS of V. For phase 2, 10 participants were recruited into the study. Participant characteristics are shown in Table 1. Nine out of 10 participants were PEG fed in phase 2; all participants were taking adequate calcium supplementation whilst on this study. Investigators would monitor calcium and vitamin D supplementation (using dietary questions and blood tests) during this study and would intervene if either was low, but no intervention was necessary.

Puberty outcomes

In phase 1, the mean age of menarche was 17.8 ± 2.3 years, with one patient premenarchal at age 18. Eleven patients were treated with sex-steroids. Three had died at time of the study. Seven of the remaining eight families consented to an interview (five female, two male). All interviewees were mothers of the patients. Median time between treatment and interviews was 4 years (range 1–27).

Phase 2 participants utilised a proposed algorithm (Supplemental Material, Appendix 2) for pubertal induction with Tanner staging every three months (Table 2). Most patients (7/10) reached Tanner stage 3 by nine months. Only one participant reached Tanner stage 5 at 24 months.

Effects of hormone therapy, including adverse events

Phase 1: Eleven patients (55 %) received treatment. Reasons for not receiving treatment included parental concern regarding psychosexual effects of sex-steroids, late referral and better nutrition permitting onset of spontaneous puberty. All patients in phase 2 were treated with sex-steroid. Side effects and tolerability for both phases are detailed in Table 3. Worsening acne and irritability were reported in one male when oral testosterone undecanoate was increased from 160 to 240 mg daily. Carers of the two males denied any issues with aggression with testosterone replacement. For females using HRT, mild worsening mood (n=2) and possible improvement in cognition (n=1) were described. No difficulty with menstrual management was reported.

Phase 2: Participants tolerated treatment very well, with no reports of breast tenderness, no aggressive behaviour or inappropriate masturbation in public for either sex (previously reported by families as an adverse outcome). Three participants had mild skin irritation related to patches. Three female participants experienced intermittent vaginal bleeding. One male participant disliked the gel and was switched to IM therapy. Treatment adherence was monitored by a treatment diary showing a compliance rate of above 80 %.

Bone outcomes

For phase 1, LS BMD was too variable to draw any major conclusions, percentage change in BMD per year ranging from −4.8 to +36 % over 1–4 years for males and 0.54–5.76 % for females. Eleven patients from the phase 1 cohort had bone age X-ray, with a mean delay in bone age of 4.1 ± 2.2 years. Growth hormone assessment was not available for most patients, so contribution of growth hormone deficiency to the delay could not be excluded. Vitamin D levels were not regularly recorded. Eight patients (40 %) had a prevalent fracture, and five experienced multiple fractures, all with minimal trauma, predominantly involving the lower limb (55 %), the remainder involving the upper arm (n=2), clavicle (n=2) and spine (n=1).

In phase 2, DXA scans were challenging to measure in this population as many patients were unable to stay still for scanning. Only six patients could be scanned. DXA reports of participants who had withdrawn and stopped treatment were also collected. LS BMD% changes were negligible in phase 2, ranging from 0.15 to 0.24 % per year, remaining stable over 1–2 years in all who could be scanned. One participant was withdrawn from the study after 6 months due to two minimal trauma femoral fractures requiring bisphosphonate. One male had a left femoral fracture documented straight after starting testosterone therapy. No other phase 2 participants had documented vertebral or long bone fractures during the intervention period.

Pain and mobility outcomes

Pain and mobility outcomes could only be measured in the phase 2 cohort. For families who completed the questionnaires, baseline QoL scores (out of 120) ranged from 62 to 92 (n=5), and 24 month QoL scores ranged from 80.5 to 92 (n=5). Mobility was maintained in all previously ambulant patients (n=5), with IPAQ results showing that over 24 months, all ambulant participants could maintain the same number of minutes conducting activities such as walking, except one participant who doubled their activity minutes from baseline (30–60 min per week).

Limitations

For phase 1, limitations included the retrospective nature of the data, limited bone density data to support a benefit of HRT and the inability to contact several families to confirm the documented history. Similarly, for phase 2, despite intensive attempts over 4 years to recruit families across several paediatric disciplines of developmental medicine, neurology and general pediatrics, only 10 participants agreed; of those, loss to follow up, withdrawal from treatment due to contending priorities and death further reduced the numbers.

For those who were eligible, no one refused treatment, suggesting the uptake of non-invasive pubertal induction is appealing to parents, and they see the benefit of optimised bone health. This study used lumbar spine as a marker for bone health, which is the most commonly used marker but limits our understanding regarding other clinically relevant sites. Future research will be aimed at following these patients prospectively, to assess longitudinal effects of the intervention.

In summary, delayed puberty and hypogonadism may be poorly recognised in CP leading to late diagnosis. Both assessment and interventions are hampered by multiple complex factors. Assessment of bone health in CP should include screening for sex-steroid deficiency, which is more common in those with a poor functional status. Optimisation of nutritional status and replacement of sex-steroids or use of pubertal induction where indicated is likely to improve bone health and to reduce adult osteoporosis.