Hypothalamic–Pituitary and Growth Disorders in Survivors of Childhood Cancer

An Endocrine Society Clinical Practice Guideline

Charles A. Sklar; Zoltan Antal; Wassim Chemaitilly; Laurie E. Cohen; Cecilia Follin; Lillian R. Meacham; M. Hassan Murad


J Clin Endocrinol Metab. 2018;103(8):2761-2784. 

In This Article

Short Stature/Impaired Linear Growth in Childhood Cancer Survivors

Epidemiology, Morbidity, and Mortality

Impaired growth is defined by a loss in height SD over time and may be transient or progressive. Short stature is characterized by a standing AH of >2 SD below the mean for age and sex. Growth impairment and short stature in childhood cancer survivors may result from: alterations in HP hormone secretion due to tumors in the suprasellar/optic pathway region, surgery or CRT involving the HP axis, primary hypothyroidism (resulting from thyroidal radiation or high-dose alkylating agent chemotherapy),[6,7] and radiation-induced impairment of spinal growth. The effects of both cranial and spinal radiation are dose- and time-dependent. Additional causes of growth impairment may include a malnourished state, growth suppressive effects of medications (e.g., glucocorticoids, tyrosine kinase inhibitors [TKIs]), and medications associated with accelerated epiphyseal/physeal closure, such as retinoids.

The prevalence of adult short stature ranges from ~9% in studies of childhood acute leukemia survivors[8–10] to as high as 40% among survivors of childhood brain tumors.[11]

Etiology and Clinical Manifestations

The major risk factors for impaired growth and short stature in cancer survivors are CRT, CSI, TBI, and younger age at the time of treatment (Table 1). Exposure to 18 to 30 Gy CRT may result in GHD and precocious puberty, whereas doses >30 Gy may result in multiple pituitary hormone deficiencies.[12,13] Exposure to CRT can also result in an earlier onset or altered tempo of puberty, including onset of breast development between ages 8 and 9 years, peak height velocity at age ≤10 years, and early menarche.[14–18] Importantly, children who have both GHD and concomitant early or precocious puberty may not demonstrate a significant growth deceleration due to the stimulatory effects of sex hormones on linear growth, and the treating endocrinologist might miss a diagnosis of GHD unless he/she is knowledgeable in this regard.

Boney structures previously exposed to radiation may be at risk for poor growth; this effect is potentially greater with higher radiation doses and younger age at exposure. Exposure to spinal radiation can result in disproportionate short stature due to impaired spinal growth, which helps differentiate spinal radiation–related growth impairment from the symmetrical impairment caused by other etiologies, such as GHD.

Systemic therapy with retinoic acid and its derivatives is associated with premature epiphyseal closure in both animal models and human studies of noncancer populations.[19] Studies of survivors of high-risk neuroblastoma reveal a significantly greater incidence of advanced bone ages in those treated with cis-retinoic acid.[20,21] This premature advancement and earlier closure of growth plates may explain, at least in part, the short AH seen in survivors treated with multimodality therapy that includes systemic cis-retinoic acid.

TKIs are targeted cancer therapies designed to disrupt specific signaling pathways involved in cellular growth and proliferation. Despite their intended specificity, nonselective, off-target effects on various protein kinases involved in chondrocyte accrual, as well as the GH/IGF-I signaling pathway, may result in growth deceleration and the potential for subsequent short AH.[22]

Diagnosis and Monitoring of Short Stature/Impaired Linear Growth in Childhood Cancer Survivors

   1.1 We recommend prospective follow-up of linear growth for childhood cancer survivors at high risk for short AH, namely those exposed to CRT, CSI, or TBI at a young age and those with a history of inadequate weight gain or prolonged steroid requirement. (1∣⊕⊕⊕O)

   1.2 We recommend measuring standing height and sitting height in childhood cancer survivors treated with radiation that included the spine (i.e., TBI, CSI, as well as radiation to the chest, abdomen, or pelvis). (1∣⊕⊕OO)

    Technical remark: Sitting height is measured directly using a sitting height stadiometer, and the lower segment can be determined by subtracting sitting height from standing height. Alternatively, the lower segment can be determined by measuring from the pubic symphysis to the floor, and the upper segment can be determined by subtracting leg length from height. The upper to lower segment ratio can then be calculated but differs depending on the method used and ethnicity. In situations where clinicians are unable to measure sitting height, measuring arm span and comparing it to standing height will provide an estimate of spinal foreshortening due to prior spinal radiation.

Evidence. The risk of growth impairment and adult short stature (height SD < −2 SD) in survivors of childhood leukemia is significantly higher among survivors treated before puberty, at younger ages, and at CRT doses >20 Gy.[10,23,24] Among studies of survivors of leukemia, lymphoma, and a broad group of pediatric cancers (e.g., osteosarcoma, Wilms' tumor, neuroblastoma, and soft tissue tumors of the head and neck), younger age at diagnosis and higher doses of CRT remained significant risk factors for adult short stature.[9,18] In a large study of 921 brain tumor survivors exposed to high-dose CRT, Gurney et al.[11] found that a significant number of adults diagnosed at younger ages had an AH <10th percentile, including 53% of those diagnosed before age 5 years, 46% of those diagnosed between 5 and 9 years, and 26% of those diagnosed between 10 and 20 years. Independent of age, those exposed to higher doses of CRT were more likely to have adult short stature than those not treated with CRT, with a threefold increased risk among those treated with >20 Gy and a fivefold increased risk among those treated with >60 Gy. These findings may be due to the development of multiple hormone aberrations, as detailed in subsequent sections of these guidelines.

Spinal radiation is an independent risk factor for short AH[10] and is associated with progressive growth impairment.[25,26] Survivors treated with higher doses of spinal radiation (>20 Gy) at younger ages, and to a larger volume of the spine, are at increased risk of short AH.[11,27] Compared with the proportionate short stature seen in GHD children resulting from CRT only, short AH associated with spinal irradiation results in disproportionate short stature, which is evident in the greater loss of spinal height SD relative to lower leg length SD.[28–30] This disproportionate growth may be evident as early as 1 year following spinal radiation and becomes progressively more evident during puberty.[26] Survivors treated with high-dose CSI (e.g., >30 Gy for medulloblastoma) demonstrate the most significant losses in seated and standing AH.[17,25,31]

Treatment of Short Stature/Impaired Linear Growth in Childhood Cancer Survivors

   1.3 We suggest against using GH in cancer survivors who do not have GHD to treat for short stature and/or poor linear growth following spinal irradiation. (2∣⊕OOO)

   1.4 We suggest against treatment with GH in children with short stature and/or impaired linear growth who are being treated with TKIs. (2∣⊕OOO)

Evidence. Studies on using GH to treat cancer survivors who do not have GHD are limited to a few small case series. In a study of 13 survivors of acute leukemia treated with cyclophosphamide and TBI, three of whom were not GH deficient, there was a progressive decline in height SD and impaired spine and leg growth despite GHT during a 3-year period.[32] In a report of 51 high-risk neuroblastoma survivors treated with multimodal therapy (including TBI), Cohen et al.[33] described GHT in seven of these survivors. One had GHD, and six were initially thought to have GH neurosecretory dysfunction. Although short-term response to GH was good, the long-term response was not; of the two that achieved AH, even the patient with GHD remained >2 SD below midparental height.

Studies of GHD childhood cancer survivors treated with GH who were exposed to spinal irradiation also suggest a reduced benefit of GHT after spinal irradiation. Ciaccio et al.[34] found that among GH-deficient medulloblastoma survivors treated with 26 to 38 Gy CSI, the mean adult standing height decreased from −1.38 SD to −1.9 SD at AH in those treated with GH, whereas the standing height of those not treated with GH decreased from −1.55 SD to −3.4 SD. However, spinal heights in both groups were similar at AH, that is, −4.56 SD and −4.85 SD, respectively. In another study of 100 survivors with GHD treated with GH, those exposed to CSI had significantly lower growth responses to GHT (4.2 cm/y vs 6.7 cm/y) and significantly greater height loss from time of radiation to AH (−3.6 SD vs −1.6 SD) than those not exposed to spinal radiation.[35] Any benefit to AH of GHT may be at the expense of further disproportionate growth.[17,31] Thus, full disclosure of this risk should be made and individual preferences considered when counseling survivors and their families about GHT in patients who have received spinal radiation treatment (RT).

Childhood cancer survivors exposed to TBI are also at risk for reduced spinal height, with the greatest risk among those treated at a younger age and with unfractionated TBI.[29,36,37] Members of this patient subgroup who receive GHT for GHD may improve AH by preventing further height loss;[38,39] however, they may experience a worsening disproportion due to lack of spinal height gain.[37,40]

TKIs (such as imatinib and dasatinib) are mainstay treatments for chronic myelogenous leukemia and other malignancies that possess the BCR-ABL1 fusion protein; patients generally are treated with them for the long term to maintain molecular remission.[41] Most studies report decreased growth in children who are using TKIs, with greater effects observed in prepubertal children and conflicting evidence of catch-up growth in pubertal children.[42–44] Although the precise mechanisms underlying the growth deceleration associated with TKI therapy are unknown, reports of low serum IGF-I levels in children on TKI therapy suggest a possible disruption of the tyrosine kinases involved in the GH signaling cascade.[44] Additional proposed mechanisms for growth failure include disrupted platelet-derived growth factor receptor-β, leading to altered recruitment and activity of chondrocytes.

We consider patients on continuous TKI therapy as having an active malignancy, as many will develop molecular evidence of persistent disease when TKI therapy is discontinued. Data on the safety and efficacy of GH use in these patients are quite limited;[41] therefore, we cannot generally recommend GHT in this setting.