Late Effects of Childhood Cancer and Treatment

Since Sidney Farber proposed the first treatment for childhood cancer in 1948, the field of pediatric oncology has been constantly evolving. ^[1] In the decades following Farber’s landmark work, there has been an extraordinary evolution in the care of children with cancer, as reflected in current survival rates, which approach 90%. ^[2] As a consequence of these incredible survival rates, the number of childhood cancer survivors continues to grow, with current estimates surpassing 300,000.

Long-term studies of this population have brought to light specific adverse effects of treatment, which are often present years after treatment and thus are termed late effects. ^[3] These long-term follow-up studies are augmented well by the findings of effect-specific investigations. This article addresses particular late effects seen in specific body systems and discusses the various forms of care provided to cancer survivors.

It is now generally recognized that surviving childhood cancer requires follow-up care by an integrated team that includes qualified and invested specialists as well as primary caregivers. These teams deliver care with a risk-based approach, following a systemic plan for lifelong screening, surveillance, and prevention that incorporates risks based on the previous cancer, cancer therapy, genetic predispositions, lifestyle behaviors, and comorbid health conditions. ^[4]

Late effects following bone marrow transplantation are discussed in Bone Marrow Transplantation, Long-Term Effects.

Heart damage can occur secondary to radiation therapy (RT) that includes all or part of the heart within the radiation field. This includes radiation to the chest or thorax (including mantle, mediastinal and/or axillary treatments fileds), abdomen, spine or total body irradiation (TBI). It can also be caused by chemotherapeutic agents, especially the anthracycline drugs. Commonly used antracycline drugs are doxorubicin, daunorubicin, idrarubicin, mitoxantrone, and epirubicin. The pathophysiology of anthracycline cardiotoxicity includes the following:

• Free radical damage

• Focal fibrosis

• Dropout of muscle fibers

• Increased wall stress and afterload

• Dilated cardiomyopathy

Risk factors for cardiac toxicity

Risk factors for cardiac toxicity include the following:

• Sex

• Age at diagnosis (at the time of cancer therapy)

• Length of time since diagnosis

• Anthracycline regimen (including total dose of chemotherapy)

• Mediastinal or spinal irradiation (including total dose of radiation and amount of heart tissue irradiated)

• Treatment with other medications affecting heart function

• Presence of co-morbid conditions affecting heart function (eg, hypertension, obseity, dyslipidemia, congenital heart disease, diabetes mellitus)

Girls have about twice the risk of cardiac toxicity that boys do. Infants and toddlers with ALL or neuroblastoma who receive anthracyclines have more frequent echocardiographic abnormalities than older children who receive the same treatment. This suggests that the heart cannot increase its workload sufficiently to keep up with a growing child. ^[5] Studies in both adults and children have shown abnormal echocardiographic findings to be more frequent in patients monitored for more than 10 years than in those monitored for less than 10 years.

The cumulative dose of anthracyclines is known to correlate with presence of toxicity. Low cumulative doses (60-150 mg/m²) are used frequently in the treatment of ALL in children, with a low likelihood of subsequent left ventricular dysfunction. Higher-risk groups, including those with T-cell leukemia, are generally treated with moderate cumulative doses (300 mg/m²).

In adults, the incidence of CHF was 3% at a cumulative doxorubicin dose of 400 mg/m², 7% at a cumulative dose of 550 mg/m², and 18% at cumulative doses higher than 700 mg/m². In a review of 6493 children on Pediatric Oncology Group studies, the risk of developing CHF was 5 times higher with cumulative doses higher than 550 mg/m² than with lower doses. ^[6] This phenomenon has been best studied with doxorubicin; other anthracyclines have different threshold doses for cardiac toxicity.

Mild clinically insignificant wall motion abnormalities related to low or moderate doses of an anthracycline may become clinically relevant as ALL survivors age and experience other comorbid conditions that affect the heart muscle, such as diabetes, hypertension, and ischemic coronary heart disease. ^[7]

In addition to cumulative dose, delivery method and schedule play a role in toxicity. Fewer toxic effects were observed with intravenous (IV) infusion over 24 hours than with IV bolus or infusion over 15-30 minutes. Fewer toxic effects were observed with weekly infusions than with infusions every 3 weeks.

Cardiac toxicity from radiation therapy

Pericarditis can be acute, occurring during RT or years later, or chronic, with pericardial effusion or constrictive pericarditis. Children who receive mantle irradiation for Hodgkin disease have a 0-2.5% long-term incidence of pericarditis. In 1 study, as many as 43% of patients were found to have pericardial thickening on echocardiography; the rate was even higher in patients monitored for at least 6 years.

Late coronary artery disease with development of myocardial infarction is observed in patients who receive mantle irradiation for Hodgkin disease. Studies in children undergoing mantle irradiation for Hodgkin disease show a greater susceptibility to premature coronary artery disease in adolescents exposed to mediastinal radiation doses higher than 40 Gy.

Additionally, RT may cause stiffening and scarring of heart tissues increasing the risk for the development of arrhythmias. RT has also been shown to increase the risk of cardiomyopathy, valvular stenosis or insufficiency and pericardial fibrosis. Fibrosis of the myocardium or the coronary vessels increases the risk of myocardial infarction and death. In a study of 16 children treated with spinal irradiation for malignancies, 75% of the patients had a maximal cardiac index below the fifth percentile, and the group as a whole had significantly higher estimated posterior wall stress; 31% had pathologic Q waves in the inferior leads.

Cardiac toxicity from chemotherapy

Patients with anthracycline-induced cardiomyopathy usually present with symptoms of congestive heart failure (CHF), which may develop spontaneously or may be initiated by stressors such as extreme exertion, as in weight lifting or difficult labor. Pericarditis may also be present, further compromising cardiac function. Additionally, ventricular arrhythmias may occur.

Subclinical or mild toxic effects can be found in a significant number of treated children, depending on the methods used to assess damage. One study of children who received anthracyclines for acute lymphoblastic leukemia (ALL) showed that 57% had abnormalities of afterload or contractility on echocardiography.

Abnormalities are more frequent and more severe in patients who receive both RT and chemotherapy. However, the incidence of early and late CHF from anthracycline chemotherapy is still only 1-2%. Problems in predicting which patients are going to develop clinically significant heart disease are still recognized.

Cardiac monitoring tests

Cardiac monitoring tests include the following:

• Serial electrocardiography (ECG)

• Serial echocardiography (2-dimensional and M-mode)

• Multiple gated acquisition (MUGA) scan

• Endomyocardial biopsy

• Fasting glucose and lipid profile

Low QRS voltage and ST-T wave abnormalities occur late and are not useful for detecting early cardiac damage. Prolongation of the QTc interval may predict late cardiac decompensation. Because significant dysrhythmias can be asymptomatic and may be missed by routine ECG, 24-hour Holter monitor has been recommended as part of routine long-term follow-up. An ECG is done at baseline with entry into long-term follow up and repeated as clinically indicated.

Echocardiography, which is noninvasive and easy to perform in children, provides measurements of left ventricular shortening fraction, diastolic filling times, and end-systolic wall stress (a measure of afterload). Additionally, RNA or MUGA, which have been well studied in adults, measure left ventricular ejection fraction and are used to evaluate regional wall motion. All studies conducted with stress or exercise tend to show more abnormalities than those conducted at rest. An echocardiogram is done at baseline with entry into long-term follow up and then periodically done based on age at treatment, radiation dose and cumulative anthracycline dose. (See table)

Endomyocardial biopsy is invasive, requiring cardiac catheterization; however, it allows quantitation of cardiac toxicity that is predictive of later decompensation. It can also be used to differentiate between malignant infiltration, infection, and anthracycline toxicity.

Lastly, fasting glucose and lipid profile should be checked every 2 years and abnormal results managed appropriately.

Recommended Frequency of Echocardiogram or MUGA scan

Table. (Open Table in a new window)

 

*Age at time of first cardiotoxic therapy (Anthracycline or radiation, whichever was given first)

Pulmonary toxicity can be acute and lethal or, more commonly, insidious in onset over a period of months to years, resulting in pulmonary fibrosis, interstitial pneumonitis, restrictive lung disease and/or obstructive lung disease. Symptoms include a dry hacking cough, dyspnea on exertion, shortness of breath, wheezing and/or exercise intolerance.

Physical examination may reveal crackles in the lung bases and, rarely, a pleural friction rub. The chest radiograph may show infiltrates, though the findings are more often normal. Pulmonary function tests (PFTs) usually reveal evidence of restrictive lung disease with a decreased forced vital capacity or total lung capacity as well as decreased carbon monoxide diffusing capacity (DLCO).

Corticosteroids have been used with some success in radiation-induced pneumonitis. However, the results of corticosteroid therapy for pneumonitis secondary to bleomycin have varied widely. Radiation pneumonitis is associated with significant morbidity and mortality.

The pulmonary pathophysiologic processes for most chemotherapeutic agents and radiation therapy (RT) are thought to be similar. Because most alveolus formation and enlargement occurs in infancy and childhood, the effects of chemotherapy and RT may be more severe in children than in adults. These pathophysiologic processes include the following:

• The initial response is oxidative injury to the pulmonary capillary endothelium and pneumocytes

• An influx of granulocytes releases chemotactic actors, elastase, collagenase, and myeloperoxidase

• Lymphocytes and plasma cells then infiltrate, secreting growth factors that stimulate fibroblasts to deposit collagen

• Pulmonary fibrosis ensues

Pulmonary toxicity from radiation therapy

The most common causes of pulmonary toxic effects are as follows ^[8] :

• Mantle or mediastinal irradiation for Hodgkin disease

• Lung irradiation in children with lung metastases from sarcomas or Wilms tumor ^[9]

• Spinal irradiation in children with brain tumors

In mantle irradiation for Hodgkin disease, toxicity is dose related; 40-55% of children studied had abnormal findings on PFTs or abnormal DLCO, though most received chemotherapy as well as RT. Few were symptomatic.

A study from St. Jude Children’s Hospital prospectively evaluated 37 children with Hodgkin disease who received chemotherapy that included bleomycin and low-dose (ie, 18-20 Gy) involved-field RT. Decreases in vital capacity and DLCO were noted over the first 6 months, but these were followed by improvement. Only 1 patient was symptomatic, but DLCO per unit of alveolar volume was still decreased significantly in most patients at the 2-year follow-up visit.

Studies in children who received lung irradiation of 12-20 Gy for metastases from Wilms tumor also showed significant drops in total lung capacity and vital capacity 18-48 months after therapy, with worsening of function. On the other hand, in a group of children (mostly adolescents) with osteosarcoma, 16 Gy of whole-lung irradiation did not produce any long-term abnormalities in PFT findings.

In patients who were treated as young children, the appearance of restrictive lung disease may relate to inadequate growth of the chest wall and lung cavity after radiation therapy, a problem not observed in older pediatric patients.

Pulmonary toxicity from chemotherapy

Drugs such as bleomycin, busulfan, the nitrosoureas (carmustine and lomustine), and methotrexate can cause long-term toxic effects on the lungs. Effects are additive to or synergistic with those of RT.

Bleomycin is used most commonly in children with germ cell tumors and lymphomas. Pulmonary toxicity is related mainly to dose and increases exponentially with cumulative doses higher than 200 units (12-17% incidence in adults). Although DLCO is a rather insensitive predictor, most oncologists advocate monitoring it and stopping bleomycin if DLCO drops below 50% of predicted.

Nitrosoureas, especially carmustine (BCNU), have been shown to cause pneumonitis in children with brain tumors. One study reported a 35% mortality in children treated with BCNU and RT to the spine. One group found a significantly increased risk of BCNU pneumonitis when high-dose BCNU was administered within 120 days of mantle irradiation.

Methotrexate administered weekly by mouth for acute lymphoblastic leukemia (ALL) in adults and in patients with rheumatoid arthritis has been shown to cause restrictive lung disease. The incidence of this effect is probably lower than 1%.

As many as 40% of all long-term survivors of childhood cancer show evidence of endocrine toxicity. Irradiation of the hypothalamic-pituitary axis, thyroid gland, or gonads can affect growth and reproductive capabilities. Chemotherapy agents, specifically alkylating agents can affect ovarian and testicular function as well. ^[10]

The hypothalamus tends to be more sensitive to effects of radiation than the pituitary gland. Growth hormone (GH) is the first hormone to be affected, followed by gonadotropins and then adrenocorticotropic hormone (ACTH) secretion. This effect is related to the total dose and fraction size of radiation received. Age at the time of treatment is also a factor; younger patients are more sensitive to the GH-lowering effects of radiation than older children are.

Growth hormone deficiency

Short stature can result from growth hormone deficiency (GHD), hypothyroidism, and poor skeletal growth after radiation therapy (RT).

GHD is the most common toxic endocrine effect of RT caused by cranial radiation. ^[11] Other factors that increase the likelihood of developing GHD include higher radiation doses, surgery in the suprasellar region, pretransplant radiation (especially pretransplant cranial radiation) and TBI. Specifically, TBI ≥ 10 Gy in single fraction, ≥ 12 Gy fractionated, or total radiation dose of ≥ 18 Gy are associated with increased risk of GHD.

The incidence of GHD is 100% in children who receive more than 45 Gy for optic chiasm gliomas and as high as 75% in children who receive 29-45 Gy for medulloblastoma. As many as 50% of children who receive 24 Gy of prophylactic cranial RT (CRT) for acute lymphoblastic leukemia (ALL) develop GHD during the 2 years following treatment, whereas those who receive 18 Gy are less prone to GHD (0-14% incidence).

Although most children recover adequate hormone levels, they do not experience catch-up growth. Treatment with GH also does not result in catch-up growth, especially in children who also received spinal irradiation because of a direct growth-inhibiting effect on bone and soft tissue.

Spinal irradiation (and thoracic or abdominal irradiation that includes the spine) impairs growth by limiting the growth of vertebral bodies. It is estimated that a 10-year-old child who undergoes spinal irradiation loses 5.5 cm of final adult height, with proportionately more growth retardation expected in younger children.

Precocious puberty

Precocious puberty has been reported in some children after CRT. Female sex, younger age at treatment and radiation doses ≥ 18 Gy are associated with increased risk of precocious puberty. Younger age at the time of radiation increases the risk, and both sexes may be affected. Studies of girls with ALL report that the onset of puberty occurs about 1 year earlier in this population than in the general population.

Thyroid dysfunction

Damage to the thyroid gland is common after neck or mantle irradiation, as used in children with Hodgkin disease, or spinal irradiation, as used in children with brain tumors. ^{[12, 13]}

Hypothyroidism is the most common thyroid abnormality after cancer therapy. ^[14] Specifically, it is associated with focal radiation (cranial, orbital/eye, ear/infratemporal, nasopharyngeal) at ≥ 40 Gy or focal radiation plus TBI, the sum of which is ≥ 40 Gy. Compensated hypothyroidism (ie, elevated thyroid-stimulating hormone [TSH] and normal thyroxine levels) occurs in 14-75% of children irradiated for Hodgkin disease with doses of 40 Gy or more and about 9% of children who receive prophylactic CRT for ALL with doses of 24 Gy or more. Overt hypothyroidism occurs in 16-21% of Hodgkin disease patients and 2% of ALL patients after radiation doses of 24 Gy.

In a large study from Stanford involving children irradiated for Hodgkin disease, the incidence of compensated or overt hypothyroidism after 26 years of follow-up was 47%. It showed that children irradiated for Hodgkin disease and monitored for 26 years, the incidence of benign thyroid nodules was 3.3%, that of Graves disease was 3.1%, that of thyroid cancer was 1.7%, and that of Hashimoto thyroiditis was 0.7%. Whether chemotherapy contributes to hypothyroidism has been controversial. Although most studies did not show an increased risk, others question whether hypothyroidism occurs earlier in patients treated with both radiation and chemotherapy than in those treated with radiation alone. ^[15]

Despite prophylaxis with oral iodide, children who receive therapeutic doses of iodine-131 metaiodobenzylguanidine for relapsed neuroblastoma have developed primary hypothyroidism. Additionally, compensated or overt hypothyroidism is observed in 47-68% of children who undergo spinal irradiation for medulloblastoma.

Thyroid replacement is recommended even in those with compensated hypothyroidism because chronic stimulation of the thyroid gland by elevated TSH has been suggested, but not proved, to increase the risk of secondary thyroid cancer in humans.

Gonadal dysfunction

The degree of gonadal damage depends on the type and total dose of chemotherapy employed, as well as on the site and dose of RT received.

Of the chemotherapeutic agents, alkylating agents such as nitrogen mustard, procarbazine, and cyclophosphamide are the most damaging to the gonads. Males are at increased risk for delayed or arrested puberty, hypogonadism, oligospermia, azoospermia, and infertility. Spermatogenesis is impaired at lower doses compared to testosterone production. Prepubertal status does not protect against gonadal injury in males. Females are at increased risk for infertility, premature menopause and delayed or arrested puberty. Higher cumulative doses of chemotherapy and/or combination with radiation therapy specifically targeted to the abdomen/pelvis, testes, lumbosacral spine (ovaries) or brain increases risk for gondal dysfunction.

In 1 study, almost 30% of prepubertal boys had evidence of gonadal dysfunction with total doses higher than 400 mg/kg (12 g/m²), but these doses had no effect on prepubertal girls. Midpubertal and sexually mature boys frequently had gonadal dysfunction even with total doses as low as 100 mg/kg (3 g/m²). When girls receive chemotherapy during or after puberty, they are affected more severely but nevertheless are less sensitive than boys. ^[16]

Girls who undergo abdominal irradiation for Hodgkin disease or Wilms tumor (ie, ovaries in the radiation field) have a 50% incidence of ovarian failure if both ovaries are in the field and the dose is greater than 15 Gy; the rate is higher if alkylating agents are also used.

Early menopause is a major concern. In a large study (the Five Center Study), the average age at menopause was 31 years in women treated with abdominal irradiation and alkylating agents combined. Early menopause increases the risk of osteoporosis and heart disease at a younger age and also has implications for patient education and family planning. ^[17]

Fertility

Irradiation of the gonads can also affect fertility. Delivery of as little as 2-3 Gy to the testes causes 100% aspermia with no recovery after as many as 40 months of follow-up. This is an important consideration for boys undergoing testicular irradiation for testicular germ cell tumors or testicular disease from ALL, abdominal irradiation for advanced Hodgkin disease, or total-body irradiation with bone marrow transplant.

A 2009 report that surveyed more than 5000 patients in the Childhood Cancer Survivor Study (CCSS) found that fertility was decreased among female CCSS participants. More specifically, females aged 15-44 years who received a hypothalamic/pituitary radiation dose greater than 30 Gy, an ovarian/uterine radiation dose greater than 5 Gy, or CCNU or cyclophosphamide were less likely ever to become pregnant. ^[18]

Data on the offspring of childhood cancer survivors were collected in 12 series. The overall rate of congenital malformations was 3-4% (no different from rates in the general population and sibling controls). No increased rate of cancer was found in offspring of cancer survivors. These data regarding survivors’ offspring must be interpreted cautiously, however, because mean follow-up was only 11 years, and the survivors received less intensive therapy by today’s standards.

Females receiving radiation therapy to the spine, abdomen/pelvis, vaginal, bladder or TBI are at increased risk for uterine vascular insufficiency. This increases the risk for adverse pregnancy outcomes, including fetal malposition, premature labor, spontaneous abortion, low birth weight infants and infant death. Higher radiation dose to the pelvis, associated müllerian anomalies and prepubertal status at start of treatment increases the risk for vascular insufficiency. The incidence of stillbirth and premature delivery is higher than normal in women who underwent abdominal irradiation that included the uterus. ^[19]

Obesity

Several studies of children with ALL who received prophylactic CRT (especially doses higher than 20 Gy) noted increased body mass index (BMI), particularly in girls. ^[20] Depending on the study endpoint (4 years after therapy or the time when adult height was reached), 12-57% of the girls and 21-45% of the boys were obese. Some small studies suggested a link between glucocorticoids, elevated leptin levels, leptin insensitivity, and obesity even in children with ALL who did not receive CRT.

Current protocols use CRT only in very high-risk children (ie, those with central nervous system [CNS] disease or high white blood cell [WBC] counts at diagnosis). Most of these children have not reached final adult height yet; thus, their risk of obesity remains to be determined. Interventions for treating obesity in children have been well reviewed elsewhere. ^{[21, 22, 23]}

In addition to obesity studies, extensive research has studied the factors predisposing to increased BMI and associated activity levels. Childhood cancer survivors were less active than a sibling comparison group or an age- and sex-matched population sample. Survivors who are at risk for an inactive lifestyle should be considered a high priority for developing and testing intervention approaches. ^{[24, 25, 26]}

Most of the data regarding neurocognitive effects of therapy derive from 2 groups of patients: children with acute lymphoblastic leukemia (ALL) and children with brain tumors. Interpretation of early data in children was hindered by methodologic problems, small sample sizes, and relatively short follow-up intervals. However, subsequent research into late effects has brought advances in both neuropsychological measurement and neuroimaging techniques that now allow us to better understand how chemotherapy alone affects the developing brain.

In isolated use, chemotherapy is typically much less neurotoxic than cranial irradiation, though chemotherapy alone does appear to have subtle effects on specific neurocognitive functions, most commonly including attention and executive functioning, visual processing, and visual-motor functioning. These effects are more prominent in children who are younger at the time of treatment, and they occur in more boys than girls. ^[27]

A study by Phillips et al showed that new-onset memory impairment was associated with cranial radiation among CNS tumor survivors (relative risk [RR], 1.97) and alkylator chemotherapy at or above 8000 mg/m² among survivors of ALL who were treated without cranial radiation (RR, 2.80). The investigators also found that smoking, low educational attainment, and low physical activity were associated with an increased risk of new-onset memory impairment. ^{[28, 29]}

Studies using more comprehensive tests of neurocognitive function and academic achievement confirm the deleterious effects of radiation therapy (RT) on the young brain. Effects are much more severe in survivors of medulloblastoma than in survivors of ALL, principally because of the higher radiation doses used in medulloblastoma.

Factors that affect the severity of brain injury include the following:

• Age at diagnosis

• Radiation dose

• Type of cancer (eg, brain tumor)

• Type of surgery and concurrent treatment (eg, chemotherapy)

Frequent and extended school absences may also contribute to neurocognitive deficits. ^[30]

Neuropsychological effects in children treated for acute lymphoblastic leukemia

Abnormalities found on computed tomography (CT) and magnetic resonance imaging (MRI) of children with ALL who were treated with cranial RT (CRT) include leukoencephalopathy and mineralizing microangiopathy. ^[31]

Leukoencephalopathy, an abnormality of white matter, usually develops 4-5 months after combined CRT and intravenous (IV) methotrexate therapy; presenting symptoms and signs include seizures, lethargy, ataxia, slurred speech, and memory loss. Similar events can occur in children who receive high-dose IV and intrathecal methotrexate without CRT.

Generally, the severity of leukoencephalopathy is not correlated with the results of neuropsychological testing or academic performance, though deficits are frequently found. Pathologic mechanisms proposed include demyelination and vascular insufficiency. The former may explain the increased incidence of toxic effects in younger people.

Neuropsychological deficits are common in children who receive 18-24 Gy of CRT. Results of verbal intelligence quotient (IQ) and full-scale intelligence quotient (FSIQ) testing are lower in these children than in age-matched controls and in patients with solid tumors who did not receive CRT. Children irradiated with 24 Gy of CRT when younger than 7 years lose an average of 13-14 points on FSIQ.

Specific deficits have been shown on tests of fine motor functioning, visual-spatial functioning, nonverbal memory, and attention and concentration. Expressive language skills and verbal learning are largely unaffected. Academically, children most often have difficulties with mathematics; difficulties with reading and spelling are less frequent.

Results of a study from St. Jude Children’s Hospital revealed similar deficits in children with ALL who received intrathecal and IV intermediate-dose methotrexate with and without CRT. Significant decreases in FSIQ (about 20% of patients), verbal IQ, and mathematics scores (about 24-27% of patients) were observed in both groups of patients. Children younger than 4 years had greater decreases in FSIQ than older children did. Behavioral problems were no more common than in the general population.

A report from the St Jude lifetime cohort study determined rates, patterns, and predictors of neurocognitive impairment in adults decades after treatment for childhood acute lymphoblastic leukemia (ALL). Impairment rates across neurocognitive domains ranged from 28.6% to 58.9%, and those treated with chemotherapy only demonstrated increased impairment in all domains. In survivors who received no CRT, dexamethasone was associated with impaired attention and executive function. The impact of CRT was dependent on young age at diagnosis for intelligence, academic, and memory functions. Risk for executive function problems increased with survival time in a CRT dose-dependent fashion. In all survivors, self-reported behavior problems increased by 5% with each year from diagnosis. Impairment was associated with reduced educational attainment and unemployment. The authors concluded that this study demonstrates persistent and significant neurocognitive impairment in adult survivors of childhood ALL andwarrantsongoing monitoring of brain health to facilitate successful adult development and to detect early onset of decline as survivors mature. ^[32]

Neurocognitive deficits in children with brain tumors

Few large studies have been conducted on survivors of childhood brain tumors, though such patients show patterns of neurocognitive defects similar to those of children with ALL. ^[33] Age at the time of CRT and total radiation dose are the major factors determining the extent of deficits, and these children generally receive much higher doses of CRT than patients with ALL do. Deficits in short-term (or working) memory and focused attention reduce the ability to learn new things. With longer follow-up, survivors fall further behind their peers.

Irradiated survivors of medulloblastoma have significant deficits. In a study of 32 children who survived more than 5 years from diagnosis, all had an FSIQ lower than 90, and 4 had an FSIQ lower than 70 (ie, in the intellectually disabled range). Children treated when younger than 3 years had a lower mean FSIQ (65) than children treated when older than 3 years (80). Similar patterns were observed in verbal and performance IQ and achievement tests in reading, spelling, and mathematics: 38% were classified as learning disabled.

In another study, 60% of children treated when younger than 3 years had an FSIQ of less than 70, whereas only 10% of those treated when older than 3 years had an FSIQ of less than 70.

Several small studies compared neurocognitive functioning in children with medulloblastoma who were irradiated and children with cerebellar astrocytoma who were never irradiated. Deficits in verbal IQ, tests of attention, spatial memory, perceptual-motor coordination, and motor domains were noted in the former group. The latter group generally scored in the low-normal range for most tests. Both groups showed continued worsening over time, but the rate of progression was steeper in the medulloblastoma group.

Tumor location and surgery may also contribute to later neurocognitive deficits, though the significance of these factors has been more difficult to elucidate because of the small numbers of survivors in studies to date. For example, in 1 study, more than 50% of children who underwent surgery for craniopharyngioma had some form of memory impairment. Children with temporal lobe tumors had both verbal and nonverbal memory deficits.

Effect of total-body irradiation

Total-body irradiation is commonly used as part of the conditioning regimen for bone marrow transplantation. Typical doses range from 12 to 14 Gy. Significant concern exists about the use of total body irradiation in younger patients, especially those younger than 2-3 years.

One study showed minimal late neurocognitive defects in patients who were older than 6 years at the time of their bone marrow transplant and total-body radiation treatment. The patients aged 3-6 years had more pronounced abnormalities on neuropsychological testing, and the who were younger than 3 years at the time of transplant were the most heavily affected.

Survivors of childhood cancer are at risk for psychological problems stemming from the diagnosis of cancer, its treatment, and the multitude of physical late effects that may accompany survival. In addition, educational, occupational, and insurance issues complicate overall quality of life. Despite numerous methodologic problems, most studies have concluded that cancer survivors are generally well adjusted and that, as 1 report states, “specific difficulties exist within an overall context of normal emotional adjustment.”

Because of physical and neurocognitive deficits, many survivors of brain tumors have more difficulties than usual functioning in society. They are less likely to go to college (by about 10%), get married (only one third have married, whereas one quarter have been divorced), and be employed (only one half are employed and their salaries are significantly lower than those of other survivors or controls).

Psychosocial adaptation

Most studies have not found an increased incidence of psychiatric disorders in cancer survivors. A study by Lubas et al showed that survivors of childhood cancer have a similar risk of having suicidal thoughts compared with the general population, but appear to be at lower risk for suicidal behavior and suicidal death. ^{[34, 35]}

With the exception of children who survive brain tumors, childhood cancer survivors are no different from the general population in terms of educational attainment, marital status, and whether they live with their parents. Yet in a long-term surveillance study, survivors reported more adverse general and mental health, greater functional impairment, and more significant activity limitations as compared with siblings. Rates of marriage, college graduation, employment, and health insurance were all lower than for sibling controls. ^[2]

The Candlelighters Childhood Cancer Foundation surveyed 300 survivors and compared them with a control group of young adults. Most survivors reported feeling different from peers, although two thirds felt that the differences were more positive than negative. Self-reported health status was more often negative in survivors than in peers, and survivors had more worries about late effects such as second cancers and fertility issues. On the other hand, survivors had fewer general health worries than their peers.

Adjustment difficulties have been observed in some areas, most notably initiation and maintenance of interpersonal relationships. In 1 study, symptoms of posttraumatic stress disorder were noted in 12% of survivors surveyed. ^[36] Adolescent survivors have been found to be less anxious than their peers, with a tendency to employ avoidance strategies more often as a means of dealing with problems of adolescence.

Several studies have found that one of the strongest predictors of survivors’ adjustment was maternal coping.

Educational attainment and school problems

Children who undergo radiation therapy (RT) for acute lymphoblastic leukemia (ALL) or brain tumors can have neurocognitive deficits that affect ability to learn new material and performance in mathematics, reading, and spelling. In addition, school absence continues to be a problem even after children finish therapy. ^[37] With the exception of children with brain tumors, childhood cancer survivors’ rates of high school graduation and college attendance are similar to those of controls.

Occupational attainment and workplace problems

Rates of employment and salaries are similar between survivors and controls, with the exception that cancer survivors have a higher likelihood of being denied entry into the military services. In the past, cancer survivors often experienced problems in the workplace with respect to job discrimination, though more recent studies suggest that such problems are occurring less often.

Insurance issues

Barriers to obtaining health insurance include refusal of new applications, policy cancellations or reductions, higher premiums, waived or excluded preexisting conditions, and extended waiting periods. A study from North Carolina found that childhood cancer survivors were more likely to be denied health insurance than their siblings were. These problems seem to lessen as the survivors become older.

Childhood cancer survivors are more likely to be rejected for life insurance or required to pay higher premiums.

Genitourinary tract

Tubular damage and hypertension from renal artery stenosis can be observed with radiation doses higher than 20 Gy. Cisplatin and high-dose carboplatin frequently cause transient glomerular and tubular damage. Tubular damage may persist for months to years. Ifosfamide administered in high doses (> 90 g/m²) can cause a renal Fanconi syndrome, with electrolyte and protein wasting that can be severe, and eventual renal failure.

Bladder damage from acrolein, a metabolite of cyclophosphamide and ifosfamide, is usually observed during, not after, therapy. Regular use of mesna during the administration of high doses of cyclophosphamide and ifosfamide usually prevents long-term bladder damage. The risk of secondary bladder cancer in children treated with these agents may be increased if frequent damage occurs, as demonstrated by hematuria. Irradiation of the bladder can result in fibrosis and diminished capacity.

Gastrointestinal tract and liver

Transfusions increase the risk of viral hepatitis. Hepatitis C has been found in 4-8% of survivors tested, and more than half of those show evidence of chronic active hepatitis. Results of treatment with interferon have been generally disappointing so far.

Radiation to the abdomen can cause bowel adhesions and fibrosis, malabsorption, and lactose intolerance. The bowel is also in the field of spinal irradiation.

Ophthalmic structures

Radiation therapy (RT) is used in some patients with retinoblastoma and almost all of those with orbital rhabdomyosarcoma. Radiation retinopathy is a frequent complication of doses higher than 40 Gy administered for retinoblastoma. Cranial RT (CRT) for some brain tumors and total-body irradiation for bone marrow transplantation can also affect the eye and orbit.

Cataracts can develop even with fairly low doses of radiation. A single large dose is more cataractogenic than a similar but fractionated total dose. Younger children may be more susceptible to developing cataracts than older children. In a study of 102 survivors of orbital rhabdomyosarcoma, 82% of those irradiated developed cataracts, and 70% had decreased visual acuity.

Orbital hypoplasia, decreased tear production, keratoconjunctivitis, and ptosis or enophthalmos have also been observed after RT.

Aural structures

Hearing loss can result in difficulties with communication, speech and language acquisition, and development of learning skills. Platinum-based chemotherapy (cisplatin and high-dose carboplatin), aminoglycoside antibiotics, loop diuretics, and CRT all can damage the cochlea, resulting in significant sensorineural hearing loss.

Hearing loss due to platinum-based chemotherapy is caused by cumulative dose-related damage to the inner hair cells of the cochlea. Loss of high-frequency hearing (6-8 kHz) occurs most commonly, but at higher cumulative doses, the loss can extend into the speech range (ie, 1-2 kHz). Children are more sensitive to this damage than adults are. Receiving cisplatin after CRT causes synergistic hearing loss.

In a study from the Dana Farber Cancer Institute, 39 children with brain tumors received cisplatin 100 mg/m² every 3 weeks for 3 doses, followed by CRT and weekly vincristine. After the 3 courses of cisplatin, 20% had hearing loss in the 6- to 8-kHz range, 16% in the 3- to 4-kHz range, and 11% in the speech range.

Musculoskeletal structures

Hypoplasia or impaired growth of bone and muscle can occur after RT. Short stature, scoliosis, and asymmetric bone and soft tissue growth are potential problems after spinal or abdominal irradiation.

Osteonecrosis or avascular necrosis (AVN) has been reported after RT and after high-dose steroid therapy in children with acute lymphoblastic lymphoma (ALL). ^{[38, 39]} A study from the Children’s Cancer Group reported AVN in 111 of 1409 children (9.3%) treated for high-risk ALL. The incidence of AVN in girls aged 10-15 years was 19.2%, and that in boys aged 15-20 years was 20.7%, whereas the incidence in children younger than 10 years was only 0.9%.

Almost all patients were affected in weight-bearing joints, and 74% had multifocal disease. ^{[38, 39]} More than 99% presented within 3 years of their cancer diagnosis. Risk factors included age (ie, before or during puberty) and the number of courses of dexamethasone in the reinduction phase (ie, 1 or 2).

A study by van Atteveld et al found that adults aged 18 to 45 years who had survived childhood cancer had a three- to fivefold higher risk of a first fracture than adults of the same age and sex in the general population. Among younger adult childhood cancer survivors who had bone mineral density (BMD) results, 36% had low BMD (Z-score ≤ –1) at any site and 9.6% had very low BMD (Z-score ≤ –2) at any site. ^{[40, 41]}

Immune system

The long-term decrease in immunity after splenectomy or post-RT functional asplenia has been extensively documented, not only in the loss of opsonization but also in alterations of the levels of serum immunoglobulins.

Second malignant neoplasms (SMNs) in childhood cancer survivors (see Table 1 below) are caused by the carcinogenic effects of radiation, chemotherapy, or both. With either type of therapy, the risk of SMNs is related to the cumulative dose received. Genetic predisposition, such as mutation of the retinoblastoma (Rb) or p53 (ie, Li-Fraumeni syndrome) gene, also plays an important role.

Table. Second Malignancies in Survivors of Childhood Cancer (Open Table in a new window)

In several large studies of patients with SMNs, retinoblastoma was the most common primary cancer, followed by Hodgkin disease and soft tissue sarcomas. The most common sites or types of SMN include the brain, the breast, the thyroid, bone, and second leukemias. An estimated 50% of carriers of the Rb gene develop an SMN within 50 years from radiation; the risk to nonirradiated carriers is approximately 25% or less. ^{[45, 46, 42]}

For secondary leukemia due to alkylating agents (eg, nitrogen mustard, cyclophosphamide, ifosfamide, melphalan, and procarbazine), the mechanism is direct DNA damage. Risk is related to the cumulative dose. Acute myelogenous leukemia (AML) after therapy with these agents usually occurs 4-8 years after diagnosis of the primary cancer and is often preceded by myelodysplasia. Abnormality or loss of chromosome 5, 7, or both is common.

For secondary leukemia due to epipodophyllotoxins (eg, etoposide and teniposide), the mechanism is inhibition of topoisomerase II, which prevents repair of damaged DNA. Risk may be related to the cumulative dose, dose scheduling (ie, once weekly or twice weekly), or both. The data are not clear, but secondary AML following administration of these agents has been reported in as many as 12% of patients with acute lymphoblastic leukemia (ALL).

AML usually develops 2-4 years after diagnosis of the primary cancer. No myelodysplastic phase occurs. As an SMN, AML is usually French-American-British classification (FAB) subtype M4 or M5 (ie, myelomonocytic or monocytic), with abnormalities of chromosome arm 11q23 (MLL gene). Almost no patients survive.

For secondary solid tumors due to radiation therapy, risk is related to the cumulative radiation dose and the age at time of treatment. Unlike the risk of secondary leukemia, the risk of secondary solid tumor after radiation continues to increase over time, with a latency of 10-20 years or longer.

The most common SMNs are the following:

• Breast cancer (after mantle radiation for Hodgkin disease)

• Brain tumors (after cranial radiation therapy [CRT] for ALL)

• Soft tissue or bone sarcomas

• Thyroid cancer

• Bladder cancer

The risk to a specific site may be related to the proliferative capacity of that tissue (eg, bone and breast during puberty). ^{[43, 47, 48]}

Prevention of toxicity

Radiation therapy (RT) may be omitted (as in most children with acute lymphoblastic leukemia [ALL] on current protocols) or delayed (as in infants with brain tumors). In addition, techniques for limiting fields, including radiosurgery and conformal radiation, may lessen toxic effects on normal structures. For example, conformal radiation to the posterior fossa is being examined in children with medulloblastoma with the hope that the cochlea can be spared and ototoxicity lessened.

Cumulative doses of alkylating agents (eg, nitrogen mustard, cyclophosphamide, ifosfamide, and procarbazine) may be limited. Toxic effects of chemotherapy may be prevented or lessened with the use of chemoprotectants; for example, dexrazoxane may lessen both the cardiac toxicity of anthracyclines and the pulmonary toxicity of bleomycin. ^[49]

In a Dana Farber study, dexrazoxane was shown to decrease early cardiac toxicity in children with ALL who received doxorubicin. That study and several completed Pediatric Oncology Group studies showed no short-term adverse effects on survival. However, long-term results regarding the effectiveness of cardiac protection and disease-free survival await further follow-up. Current recommendations are to use dexrazoxane only in the setting of a controlled clinical trial.

Amifostine is also being studied as a potential protectant against cisplatin-induced acute bone marrow toxicity, nephrotoxicity, neurotoxicity (including ototoxicity), and radiation-induced damage to normal tissues. However, a small study in children with medulloblastoma and another study in children with germ cell tumors did not document any protective effect against cisplatin-induced ototoxicity. ^[50]

Healthy dietary and lifestyle habits, such as exercise and avoidance of alcohol, sun, and tobacco, should be promoted. Counseling regarding avoidance of risky behaviors, such as smoking and recreational use of drugs (especially use of cocaine by persons with cardiac risk factors) is recommended during follow-up visits with survivors.

Models of long-term follow-up

Current care

Because the likelihood and severity of adverse outcomes can be decreased by prevention or early detection, the international pediatric community recommends periodic lifetime survivor-focused health care. ^[51] Currently, most long-term follow-up (LTFU) programs are found at pediatric hematology/oncology care centers and include both pediatric oncologists and survivor-focused specialists such as psychologists, social workers, and highly trained nurses.

The LTFU clinics monitor general medical issues and refer to a network of specialists for specific needs, such as cardiac care or endocrine abnormalities. Often, patients are referred to a LTFU specialist within their center of care or simply stay with their primary oncologist if he or she does LTFU care. The transition from acute therapy to LTFU often occurs 1-2 years after treatment.

While the patient is being seen at the LTFU clinic, the team assesses for recurrence of malignancy, monitors and treats late effects, cares for any psychosocial concerns the patient or family may have, and provides targeted counseling and education on methods to optimize health and quality of life. ^[52]

Quality care has been defined as a risk-based approach to health care consisting of a systemic plan for lifelong screening surveillance and prevention that incorporates risks on the basis of the previous cancer, cancer therapy, genetic predisposition, lifestyle behaviors, and comorbid health conditions. ^[4]

Rationale for further development of care models

As the number of survivors increases, pediatric oncology providers will not be able to manage both patients currently in therapy and long-term survivors, simply because capacity constraints will not allow them to do so. Accordingly, there is a need for further development of models of care.

The main barrier to the implementation of longitudinal risk-based care through an LTFU program is that most late effects occur in adulthood, after most survivors have been lost to follow-up or are unable to return to the pediatric-based institution. ^[53] Barriers can be divided into 3 general types according to whether they originate from the survivor, the physician, or the health care system. (Note that these barrier types are not mutually exclusive and may coexist.)

Survivor-based barriers include the following:

• Lack of knowledge about late effects and future risk (this lack of knowledge is certainly not the patient’s fault; when most of them received care, late effects were as yet unknown)

• Lack of knowledge of details and therapy of cancer

Physician-based barriers include the following:

• Lack of capacity for survivor care within the treating institution

• Primary care physicians who are unfamiliar with the survivor population and their needs

• Poor communication between cancer centers and primary care physicians

Health care system–based barriers include the following:

• Survivors not always covered by public assistance after cancer treatment

• Screening, labs, tests not always covered by insurance ^[4]

The best transition to LTFU occurs when young adults, along with their families and trusted health care providers, participate in creating an individualized plan that is organized and comprehensive. Communication and collaboration among providers, families, and patients are essential. ^[54]

Characteristics of care models

Models of care for long-term survivors must be flexible enough to meet the needs of young children, adolescents, and young and older adults and must be sensitive to change throughout the life cycle. ^{[55, 56]} At present, no uniform, comprehensive roadmap for LTFU exists.

Models of care should cover the following:

• Risk for development of late effects

• Psychosocial functioning

• Need for health education

• Financial challenges associated with obtaining health and life insurance and employment ^[57]

Care models are of 3 basic types, as follows:

• Care delivered at a cancer center – This has historically been the primary model for delivery of care but is now being strained by the growing number of survivors

• Community-based care – Survivors have become increasingly more reliant on primary care providers for follow-up care; this leaves primary care physicians responsible for specialized continuing education to provide accurate and beneficial care

• Combined care, the responsibility for which is shared by the cancer center and the community – This appears to be the model that currently is most popular with caregivers; the general consensus is that without collaborative communication, optimal care cannot be attained

Three emerging models of transitional care have been defined:

• Cancer center-based model – This model is located at the pediatric treatment center or within its larger governing institution (university or healthy system). Transitional care is delivered within the same system as treatment was given and involves direct on-site collaboration of the pediatric and adult teams.

• Community-based model – This model is located in the office or clinic of the care provider, typically the primary care physician. It requires a coordinated transition of care that includes transfer of all pertinent clinical information and follow-up responsibility from the treatment center to the primary care provider.

• Hybrid model – With this model, care is transferred to the office or clinic of the primary care provider but relies on an ongoing interaction with the cancer treatment center.

Regardless of the model of care used, partnership with health care providers assessing a wide range of specialties is required to deliver optimal care to childhood cancer survivors. Pediatricians, pediatric oncologists, adolescent medicine physicians, internists, family medicine physicians, physician assistants, and nurse practitioners require ongoing education regarding the potential long-term effects for which survivors of childhood cancer are at risk.

In view of the expected increase in the number of survivors of childhood cancer, there is some urgency to the task of determining where LTFU should take place and who should provide the care. ^[57] The coming years should see a great deal of active research in this area.

Monitoring for late effects

Musculoskeletal system

Children receiving RT should have yearly physical examinations, including a scoliosis examination (especially if in pubertal growth spurt). Radiographs should be obtained as needed.

Breast

Girls receiving mediastinal RT should be taught self-examination of the breast. Beginning baseline mammograms at age 25-30 years or 10 years after RT in these patients has been recommended. Perform annual clinical examinations and repeat mammograms every 2-3 years, depending on breast tissue.

A study by Zaletel et al found a 15.2% cumulative incidence of secondary breast cancer (SBC) 40 years after chest radiotherapy in long-term survivors of childhood Hodgkin lymphoma. The investigators emphasized the importance of regular, long-term follow-up with breast cancer screening and self-examination, as all SBCs were diagnosed at an early stage and no SBC-related deaths were reported in the cohort. ^[58]

Brain

Children receiving cranial RT (CRT) should undergo neurocognitive testing at baseline and subsequently whenever the clinical need arises.

Neuroendocrine system

Monitor yearly growth curves in all children who received radiation. For children who receive RT to the hypothalamic-pituitary axis in doses higher than 29 Gy, assess bone age around the age of 9 years. Perform growth hormone (GH) stimulation testing about 2 years after therapy is completed, sooner if growth starts to decelerate.

Check luteinizing hormone (LH), follicle-stimulating hormone (FSH), testosterone or estradiol, prolactin, and morning cortisol levels at baseline and then as needed. Perform thyroid function studies (ie, thyroid-stimulating hormone [TSH], free thyroxine, and triiodothyronine) at baseline and then every 3-5 years as needed.

Ovary

In girls who receive alkylating agents or abdominal or pelvic RT, monitor FSH, LH, and estradiol levels as a baseline at age 13 years and as indicated in cases of delayed puberty, irregular menses, primary or secondary amenorrhea, or clinical signs and symptoms of estrogen deficiency. Continue to monitor menstrual histories yearly after therapy; elevated LH and FSH levels and low estradiol levels may indicate ovarian failure if menses do not occur and if signs of premature ovarian failure are present.

Hormone replacement therapy is necessary for girls who do not go through puberty or who have evidence of premature ovarian failure. ^[59]

Testes

In boys who receive alkylating agents or testicular or pelvic RT, check baseline LH, FSH and testosterone levels at age 14 years and then as clinically indicated. ^[59] Passage through puberty is rarely affected; only large doses of alkylating agents and high radiation doses (> 35 Gy) are likely to affect Leydig cells. Sperm analysis is the standard for evaluation of fertility, though elevated gonadotropin levels and small testes are excellent indicators of potential infertility.

Heart

In children receiving anthracyclines, high-dose cyclophosphamide, or mediastinal or spinal RT, perform echocardiography and electrocardiography (ECG) at baseline, then every 3-5 years after treatment; perform these tests more often if abnormalities are present. Perform Holter ECG every 5 years and then as needed in children who receive high cumulative doses of anthracyclines. Perform stress echocardiography, radionuclide angiocardiography (RNA), or multiple gated acquisition (MUGA) as needed or if screening echocardiography findings are abnormal.

Lungs

In children who receive chest or mediastinal RT, bleomycin, or carmustine (BCNU) or lomustine (CCNU), perform pulmonary function tests (PFTs) at baseline, then every 3-5 years as needed. Strongly counsel these children regarding smoking and exposure to pulmonary toxins.

Kidneys

In children who receive cisplatin, carboplatin, cyclophosphamide, ifosfamide, or abdominal RT, monitor creatinine and magnesium levels every 1-2 years. Measure creatinine clearance at baseline and then every 3-5 years as needed in patients who receive platinum-containing chemotherapy. Perform urinalysis annually in all children who receive ifosfamide or abdominal RT. In children who receive ifosfamide, also monitor serum phosphate and urine glucose and protein levels annually for evidence of Fanconi syndrome.

Bladder

In patients who receive cyclophosphamide, ifosfamide, or bladder RT, perform urinalysis annually to check for hematuria.

Thyroid

In children who receive RT to the neck, mediastinum, or spine, check TSH, free thyroxine, and triiodothyronine levels annually for 10 years. If a nodule or goiter is palpated, perform ultrasonography.

Liver

In children who have been treated with 6-mercaptopurine, methotrexate, actinomycin-D, or abdominal RT, perform liver function tests (LFTs) every 1-3 years. All survivors who received blood products, especially those who were treated before 1992, are at risk for hepatitis C and should be checked for all hepatitis viruses.

Gastrointestinal tract

In patients who have received abdominal RT, check stool annually for occult blood because of the risk of secondary cancer. Malabsorption may be reflected in a history of diarrhea, lactose intolerance, or failure to thrive and should be investigated.

Second malignant neoplasms

The Childhood Cancer Survivor Study, examining adherence to cancer screening guidelines by survivors at average risk for a second malignant neoplasm (SMN) and survivors at high risk, found that whereas cervical and breast cancer screening rates in average-risk survivors were reasonable, those in survivors at highest risk for colon, breast, or skin cancer were very low. ^[60]

Monitoring for SMNs varies according to the patient’s genetic predisposition and the treatment provided. For an early diagnosis, the survivor and the family must be informed regarding the risks, and they must be educated to understand that they can detect problems in the earliest stages. Annual visits to a health care provider serve the purpose of ongoing counseling and provide a baseline of health that will facilitate rational assessment of future signs and symptoms.