 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
SPECIAL ARTICLE |
Department of Pediatrics, University of Vermont College of Medicine, Burlington, Vermont
 |
ABSTRACT |
|---|
 TOP ABSTRACT THE PROBLEMNONMALIGNANT LATE EFFECTS OF...SECOND MALIGNANT NEOPLASMS...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
Key Words: late effects • childhood cancer • cancer therapy • cancer survivors
Abbreviations: CCSS—Childhood Cancer Survivor Study • RR—relative risk • CNS—central nervous system • BC—breast cancer • COG—Children's Oncology Group • H&P—history and physical examination • SMN—second malignant neoplasm • SIR—standardized incidence ratio • TC—thyroid cancer • AER—absolute excess risk • AML—acute myelogenous leukemia • MDS—myelodysplastic syndrome • t-AML/MDS, treatment-related acute myelogenous leukemia/myelodysplastic syndrome
Primary care physicians, a category that includes pediatricians, internists, family practitioners, and obstetricians/gynecologists, are and will be the health care providers for childhood cancer survivors. Therefore, it is imperative that they familiarize themselves with the late effects of cancer therapy to prevent or diminish them. Chemotherapy used to treat adult cancers but not those that occur in childhood will not be addressed, and the late effects of surgical therapy will not be discussed. In the last 5 years there has been only 1 comprehensive review in the primary care literature that dealt with the late effects of childhood cancer therapy.1 Since the publication of that article in 2002, a prodigious amount of information on this topic has become available.
|
THE PROBLEM |
|---|
|
TOPABSTRACTTHE PROBLEMNONMALIGNANT LATE EFFECTS OF...SECOND MALIGNANT NEOPLASMS...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
The long-term morbidity of childhood cancer survivors, now numbering 
270000 in the United States, was determined in a landmark Childhood Cancer Survivor Study (CCSS).6 Patients (10397) treated from 1970 to 1986 with a mean age of 26.6 years (range: 18–48 years) were interviewed and compared with their siblings. At least 1 chronic condition was present in 62.3%, and 27.5% had a severe life-threatening condition. The relative risk (RR) of a chronic condition in a survivor compared with a sibling was 3.3, and for a severe or life-threatening condition the RR was 8.2 compared with a sibling. The cumulative incidence of a chronic health condition 30 years after diagnosis was 73.4%, with a cumulative incidence of 42.4% for a severe, disabling, or life-threatening condition or death resulting from a chronic condition. Two or more chronic health conditions were seen in 37.6% of patients compared with 13.1% in siblings and 3 or more chronic health conditions were seen in 23.8% of patients compared with 5.4% in siblings. The RR of selected severe or life-threatening or disabling health conditions were as follows: major joint replacement if not part of therapy (54), congestive heart failure (15.1), second cancers excluding basal cell and squamous cell carcinomas (14.8), severe cognitive dysfunction (10.5), coronary artery disease (10.4), cerebrovascular accident (9.3), renal failure or dialysis (8.9), hearing loss uncorrected by hearing aid (6.3), legally blind or loss of an eye (5.8), and ovarian failure (3.5). Survivors of bone or central nervous system (CNS) tumors and patients with Hodgkin's lymphoma were at highest risk for these severe or life-threatening conditions and were also more likely to have multiple conditions.
Bone cancer survivors more frequently had severe musculoskeletal problems, congestive heart failure, and loss of hearing, whereas CNS cancer survivors were more likely to have cognitive dysfunction, seizures, and endocrinopathies. Hodgkin's lymphoma survivors were more likely to have coronary artery disease, cerebrovascular accidents, valvular heart disease, cardiomyopathy, second cancers (breast cancer [BC] in women), and lung and thyroid disease. Exposure to 1 of 5 treatment combinations was associated with a risk of having a severe or life-threatening or disabling condition that was 10 times the expected risk: (1) chest radiation and bleomycin, (2) chest radiation and anthracycline, (3) chest radiation and abdominal or pelvic irradiation, (4) anthracycline and an alkylating agent, and (5) abdominal or pelvic irradiation and an alkylating agent. Adverse psychosocial outcomes such as depression were not included in this analysis.
Approximately 95% of children aged 0 to 14 years are treated in Children's Oncology Group (COG) centers, and 65% are entered onto clinical trials.7 In another CCSS report to determine the type of medical care received over a 2-year period, 9434 patients were surveyed at a mean age of 26.8 years (range: 18–48 years). Eighty-seven percent reported general medical contact, 71.4% a general physical examination, 41.9% a cancer-related visit, and only 19.2% a visit to a cancer center.8
The advantages and problems of long-term follow-up clinics have been enumerated recently. In a survey sent to directors of 24 comprehensive long-term follow-up programs for pediatric cancer survivors in the United States and Canada, the following were listed as primary benefits: health care provided by clinicians familiar with long-term risks, provision of risk-based screening and surveillance, and targeted education for risk reduction and healthy lifestyles. Barriers to clinic functioning were inadequate resources and finances, low institutional commitment, lack of capacity to care for a growing population, difficulty with ongoing communication with community physicians, and lack of interest and awareness among survivors.9
In a study from Sweden, 335 childhood survivors of acute leukemia, lymphoma, or Wilms' tumor who were over 18 years of age were sent questionnaires 5 years or more after completion of therapy (response rate: 73%).10 Sixty percent had no regular follow-up visits, and 42% of these patients reported that they missed not having one. One third of the respondents were dissatisfied with the follow-up program, but only 3% with regular follow-up visits found them unnecessary. Complaints subjectively related to their diseases or treatment were reported by 47%. Thirty-four percent of all respondents did not miss having a regular follow-up visit, and neither perceived disease-related complaints nor radiation therapy was a predictor for having a scheduled follow-up visit.
In a report on 650 survivors of childhood cancer,11 it was determined that 40% had endocrine problems followed, in order, by hearing or vision difficulties, neurocognitive impairment, cardiopulmonary dysfunction, gastrointestinal disorders, second malignancies, and miscellaneous complications, whereas 30% had no problems. In another study of 290 survivors of childhood cancer,12 41% had endocrine problems, 26% developed organ toxicity, 17% had mobility problems, 15% demonstrated neuropsychological difficulties, 14% were infertile, 13% had vision or auditory difficulties, and 10% had cosmetic problems. Furthermore, pediatric cancer survivors face insurance-coverage difficulties13 and are at risk for unemployment, especially if they had CNS or brain tumors.14
All monitoring recommendations that follow are my own except as noted. The COG recommendations may be found on their Web site.15
|
NONMALIGNANT LATE EFFECTS OF CANCER THERAPY |
|---|
|
TOPABSTRACTTHE PROBLEMNONMALIGNANT LATE EFFECTS OF...SECOND MALIGNANT NEOPLASMS...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
Heart
Cardiomyopathy, subclinical left ventricular dysfunction, valvular disease, pericardial disease, and arrhythmias have been reported. Cardiac damage may be caused by anthracyclines (doxorubicin, daunorubicin, mitoxantrone, epirubicin, and idarubicin), and 60% of patients with childhood cancer are currently being treated with anthracyclines.16 There may be an increased risk of cardiac damage after anthracycline therapy for pregnant women (the 40% increase in blood volume that occurs during pregnancy may adversely affect the subclinical cardiomyopathy attributable anthracyclines, if present, as well as the cardiomyopathy of pregnancy, if that occurs), females, patients treated at <4 years of age, patients treated with other drugs that affect the heart, or patients for whom thoracic radiation is used.17
After anthracycline therapy, the risk of congestive heart failure is 0% to 16%, and for subclinical cardiomyopathy it is 0% to 57%.18 The risk of cardiac damage, most importantly, depends on the cumulative dose of anthracycline. After more than 6 years of follow-up, 57% of children with acute lymphocytic leukemia treated with doxorubicin 45 to 500 mg/m2 (mean: 360 mg/m2) had abnormalities of left ventricular afterload or contractility. However, 17% of patients who received only 1 dose of doxorubicin (45 mg/m2) developed elevated afterload, and patients who received a cumulative doxorubicin dose of only 228 mg/m2 had either increased afterload (59%), decreased contractility (23%), or both.19 Protection against anthracycline-induced cardiac toxicity resulting from free-radical damage may be afforded by dexrazoxane, a free-oxygen-radical scavenger.20 Other antineoplastic drugs may also cause cardiac damage, but to a lesser degree.21 Thoracic radiation may also be responsible for cardiotoxicity.22 The pathophysiology of both anthracycline- and radiation-associated cardiomyopathy has been described.22,23
Monitor the patient with a history and physical examination (H&P) and echocardiogram every year for 2 years and then every 2 years if normal and associated with a normal H&P. If the patient is pregnant, obtain an echocardiogram every trimester with at least 1 evaluation by a cardiologist if she has an abnormal echocardiogram.
Vasculature
Arterial damage resulting in stroke or myocardial infarction has been reported. Patients who receive cranial or thoracic radiation are at risk for these complications. In 1 study of childhood CNS malignancies, stroke occurred in 1.6% of the patients.24 A CCSS report25 identified 37 of 4828 leukemia survivors (mean age at diagnosis: 5.9 years) and 63 of 1871 brain tumor survivors (mean age at diagnosis: 7.7 years) diagnosed between 1970 and 1986 who developed late-occurring stroke, defined as presenting 5 or more years after the primary diagnosis. The mean interval from the first cancer diagnosis to late-occurring stroke was 9.8 years for the former and 13.9 years for the latter. When compared with siblings, the RR for late-occurring stroke was 6.4 for leukemia survivors and 29 for brain tumor survivors. A mean dose of cranial radiation therapy of 
30 Gy was associated, in a dose-dependent fashion, with an increased risk for both groups of survivors. In a pilot study,26 anthracyclines were shown to cause impaired endothelial function, which suggests that they may play a role in the progression of coronary disease. Venous damage is rare.
Monitor the patient with a yearly H&P.
Lung
Pulmonary fibrosis, restrictive-obstructive lung disease, and delayed interstitial pneumonia have been reported. The risk of these complications increases the younger the age at diagnosis. Pulmonary damage may be caused by thoracic radiation, which is dose and fractionation dependent, and significant abnormalities in pulmonary function have been observed after lung irradiation.27 Thoracic radiation when combined with bleomycin, actinomycin D, cyclophosphamide, vincristine, or adriamycin can produce radiation pneumonitis at much lower radiation doses.28 This synergistic effect is observed in abdominal organs when these drugs are given with abdominal radiation therapy.
A CCSS report29 on 12390 children 5 years postdiagnosis demonstrated a significant association between lung radiation and lung fibrosis (RR: 4.3), supplemental oxygen use (RR: 1.8), emphysema (RR: 2), recurrent pneumonia (RR: 2.2), chronic cough and shortness of breath for >1 month (RR: 2), exercise-induced shortness of breath (RR: 1.8), and abnormal chest wall development (RR: 5), as well as a significant association of lung fibrosis with the carmustine (RR: 1.4), bleomycin (RR: 1.7), busulfan (RR: 3.2), lomustine (RR: 2.1), and cyclophosphamide (RR: 1.6). Chest radiation was associated with a 3.5% cumulative incidence of lung fibrosis 20 years after diagnosis.
If the dose of bleomycin is <450 mg, 3% to 5% of patients are affected, and with >500 mg, 20% of patients affected28; with carmustine, up to 30% of patients develop pulmonary fibrosis with doses between 80 and 240 mg/m2 given every 6 to 8 weeks for >2 years (cumulative dose: 700–1800 mg/m2), and there is a marked increase in incidence at >1500 mg/m2.30 Methotrexate may cause pulmonary damage,31 whereas other chemotherapeutic agents not previously discussed are rarely responsible for pulmonary toxicity.28
Monitor the patient with an H&P and a pulmonary-function study every other year, unless the results are abnormal, and then every year. Obtain a chest computed tomography scan if there is a significantly abnormal pulmonary-function study result and/or clinical symptoms are present.
Gastrointestinal Tract
Fibrosis leading to partial or complete obstruction may occur as a result of radiation.32
Monitor the patient with an H&P yearly.
Spleen
Functional (radiation-induced) or anatomic (staging splenectomy, which is rarely, if ever, done today) asplenia may predispose the patient to sepsis, a lifelong problem.33
Prevent asplenia with immunizations against Haemophilus influenzae, Streptococcus pneumoniae, and Neisseria meningitidis if not previously immunized and counsel the patient to immediately report a fever or feeling ill to his or her physician. There is an ongoing debate regarding the use of prophylactic antibiotics.34 The Working Party of the British Committee for Standards in Clinic Hematology Task Force recommends, as do I, lifelong antibiotic prophylaxis,35 whereas the American Academy of Pediatrics states that such prophylaxis may be discontinued after 5 years of age.36
Liver
Hepatitis and cirrhosis can occur. Liver damage may be caused by methotrexate and 6-mercaptopurine37,38 as well as many other agents39 and contaminated blood products.40–42 Hepatitis B and C infections secondary to transfusion therapy are rarely seen today; however, for patients treated in the past this may still be a significant problem, especially the increased risk of hepatic cancer secondary to hepatitis C infection.
Monitor the patient with an annual H&P, and check aspartate aminotransferase and alanine aminotransferase levels yearly.
Kidney
Nephropathy, both glomerular and/or tubular damage, can occur.43 Kidney damage may be caused by radiation44 and ifosfamide45 as well as other antineoplastic drugs such as cisplatin.46
Monitor the patient with H&P, urinalysis, microscopic examination of the urine, complete metabolic panel (including magnesium), and blood pressure assessment yearly.
Bladder
Hematuria, cystitis, fibrosis, and dysfunctional voiding can occur. Bladder damage may be caused by radiation,47 cyclophosphamide, and ifosfamide. The metabolic byproduct of these 2 drugs is acrolein, which irritates the bladder, and the incidence of the adverse effects caused by cyclophosphamide and ifosfamide can be decreased with hydration and mesna, which binds acrolein. The incidence of bladder damage is 5% to 10% for cyclophosphamide48 and 20% to 40% for ifosfamide.49
Monitor the patient with H&P, urinalysis, and microscopic examination of the urine yearly.
Skeletal
Osteopenia, osteoporosis, avascular necrosis, spinal deformities, and other skeletal changes can occur. Skeletal damage may be caused by steroids,50 methotrexate, cranial radiation (decreased growth hormone with resultant abnormal bone metabolism), direct radiation to the bone, and cyclophosphamide/ifosfamide (gonadal damage leading to ovarian and/or Leydig cell dysfunction with resultant loss of bone mass).51,52
In patients with childhood acute lymphocytic leukemia, there may be a decrease in bone mineral density, the severity of which decreases with time after treatment.53 The significance of this finding in these patients is presently unclear.54 Bone mineral density has also been reported to be reduced in up to one third of survivors of childhood brain tumors, and the reasons are multifactorial, with craniospinal irradiation probably being the most important factor.55 The spinal deformities and other skeletal changes that may result after radiation therapy56 are seen less frequently now because of lower doses and newer radiation techniques.
Monitor the patient with a yearly H&P including scoliosis screening, and perform a bone density study once 2 or 3 years after therapy; if results are normal, it does not need to be repeated unless there are clinical symptoms and/or signs that suggest a problem.
Muscle
Atrophy may occur after direct radiation to the muscle.56
Monitor the patient with an H&P yearly.
Thyroid
Hypothyroidism or hyperthyroidism may occur after thoracic, cranial, or neck radiation. A CCSS study of 13674 patients with Hodgkin's lymphoma57 found an increased risk of hypothyroidism with an increased dose of radiation (>4.5 Gy), with older age (>15 years), with female gender, and <5 years after diagnosis. The RR for hypothyroidism in this study was 17.1, and it occurred in 25% of patients (50% at 20 years if >4.5 Gy was given). It developed at a mean of 7 years after diagnosis, and the median age at treatment was 14 years. The RR for hyperthyroidism was 8, and it occurred in 5% of the patients, with a mean age to development of 8 years.
In a study of 461 children treated for Hodgkin's lymphoma,58 43% developed hypothyroidism (47% of white patients and 21% of black patients) at a median of 2.9 years after therapy, and the risk was greater for female patients. In an older study of 1787 patients (35% younger than 22 years) with Hodgkin's lymphoma,59 overt and subclinical hypothyroidism was seen in 44% of the patients 20 years after therapy with >30 Gy and in 27% of the patients 20 years after therapy with 7.5 to 30 Gy. Most cases were identified during the second or third year after therapy. The majority of cases occurred in patients treated at 15 to 25 years of age, and the risk was increased in female patients. Hyperthyroidism was seen in 1.7% of the patients, and the risk was 7.2 to 20.4 times that seen in subjects without Hodgkin's lymphoma.
Monitor the patient with an annual H&P, and check thyrotropin and free-thyroxine levels yearly.
Growth and Development
Obesity
The prevalence of obesity after therapy for childhood acute lymphocytic leukemia is 16% to 56% and is caused by cranial radiation (growth hormone deficiency), steroid therapy, physical inactivity, and increased dietary intake.60 In adult survivors of childhood acute lymphocytic leukemia, cranial radiation at >20 Gy has been associated with obesity, particularly in girls who were treated at 0 to 4 years of age.61 Chemotherapy without cranial radiation may also lead to obesity in survivors of childhood acute lymphocytic leukemia.62
Monitor the patient with an H&P (weight determination) and dietary counseling yearly.
Short Stature
In patients with childhood acute lymphocytic leukemia, short stature may be caused by cranial radiation because of growth hormone deficiency,63 and there is an increased risk if the patient is <4 years of age. Early growth deceleration with bone age retardation is seen, and at the end of therapy 70% of patients will show a variable degree of catch-up growth; this catch-up growth can be complete 2 to 3 years after treatment in patients who did not receive cranial radiation, but it is usually incomplete in patients who did receive cranial radiation. Because growth is impaired in patients with acute lymphocytic leukemia who did not receive cranial radiation, chemotherapy and/or other factors may also be responsible for this problem.64 Maximizing final height with growth hormone treatment may be achieved if therapy is initiated at the earliest bone age that is clinically feasible.65
Monitor the patient with an H&P (height measurement) yearly and bone age films when clinically indicated.
Gonads
Gonadal failure can occur. Testicular or ovarian damage may be caused by radiation therapy (directly to the gonads or to the brain [hypothalamic-pituitary axis damage]) or alkylating agents (cyclophosphamide, nitrosourea, chlorambucil, ifosfamide, dacarbazine, thiotepa, melphalan, busulphan, carmustine, lomustine, cytarabine, or procarbazine). Sertoli cell function (sperm production) is impaired at lower drug doses when compared with impairment of Leydig cell function (testosterone production). The degree of gonadal impairment is related to the age at, dose of, and fractionation schedule for radiation therapy and the age at and dose of chemotherapy at the time of treatment.30,66,67
Prepubertal and adolescent girls are more resistant to alkylator-induced and radiation-induced failure because of increased numbers of follicles. Most young female patients treated with standard chemotherapy will retain ovarian function. The ovaries of younger female patients are more resistant to radiation injury than older ones, but >20 Gy produces failure in most female children. In prepubertal girls, >20 to 30 Gy may lead to failure or incomplete pubertal development. In boys, alkylator-induced Leydig cell failure requiring testosterone replacement is uncommon; however, germ cell damage and infertility are common. Gonadal failure is dose and fractionation dependent with radiation therapy; >3 Gy usually produces irreversible azoospermia. With <12 Gy, Leydig function is usually spared in prepubertal boys.30,66,67 Fertility may be preserved with sperm banking,68 ovarian transposition, or egg or ovarian tissue banking.69,70
Monitor the patient with an H&P (height measurement, weight determination, and Tanner assessment) and check follicle-stimulating hormone, luteinizing hormone, testosterone, and estradiol levels when clinically indicated. Follicle-stimulating hormone and luteinizing hormone levels are high when direct gonadal radiation and/or alkylator therapy is responsible for the gonadal failure and low when cranial radiation is responsible.
Central Nervous System
Intellectual or cognitive impairment involving mental processing, attentional or memory deficits, visual-spatial abilities, attention concentration, nonverbal memory, and somatosensory function may be caused by cranial radiation. There is an increase in impairment associated with female gender, decreasing age at therapy (<4 years), and increase in radiation dose.30 A recent study demonstrated that intrathecal methotrexate combined with either high-dose or very high-dose intravenous methotrexate did not result in poorer neurocognitive, cognitive, or academic outcomes in children treated for acute lymphocytic leukemia when compared with population norms.71
Monitor the patient with an H&P yearly and neuropsychological testing when clinically indicated.
Psychological maladjustment, mood disturbances, behavioral problems, somatic distress, academic underachievement, unemployment, and/or posttraumatic stress disorder may occur in 10% to 20% of long-term survivors of childhood cancer.72 In 1 study of 9535 patients from the CCSS group,72 44% reported at least 1 adverse health status domain, which included general health, mental health, functional status, cancer-related pain, and cancer-related anxiety/fears. An increased risk was seen in female patients and those patients with a lower income or educational attainment. In another study of 226 adult survivors of childhood cancer,73 29 patients (12.83%) reported suicidal ideation, although only 11 of them were significantly depressed according to the Beck Depression Inventory self-reporting questionnaire. Suicidal symptoms were related to cancer treatment and posttreatment mental and physical health. Elevated levels of stress may lead to multiple health-compromising behaviors such as smoking, lack of exercise, and/or failure to adhere to sun-protection recommendations.74
Monitor the patient with an H&P and a psychological screening test75 yearly.
The hypothalamic-pituitary axis may be affected by cranial radiation, which can lead to growth hormone, thyrotropin, adrenocorticotropic hormone, and/or gonadotropin deficiency. The risk is dose dependent, and growth hormone is the most sensitive to radiation.76 Except for growth hormone deficiency and premature sexual development, which may develop with doses as low as 18 Gy, this axis is rarely affected unless >40 Gy is used.77 Growth hormone deficiency has been reported to occur in 24% of childhood acute lymphocytic leukemia patients treated with cranial radiation and in 8% of such patients who did not receive cranial radiation.78 Hypothalamic dysfunction may also occur after chemotherapy alone, and the mechanism is not well understood. In a study of 31 childhood cancer survivors who received chemotherapy but no radiotherapy,78 central hypothyroidism was identified in 16 (52%), growth hormone deficiency in 15 (48%), and pubertal abnormalities in 10 (32%). In the latter group, 5 (19%) had gonadal failure, 3 (11%) had gonadotropin deficiency, and 2 (6%) had precocious puberty.
Monitor the patient with an H&P (height measurement and Tanner assessment) yearly and check growth hormone, thyrotropin, adrenocorticotropic hormone, and/or gonadotropin levels when clinically indicated.
Neuropathy may be caused by vincristine and can lead to impaired deep-tendon reflexes (most common), paresthesias, sensory symptoms, motor weakness, paralytic ileus, and cranial neuropathies.79 This neuropathy may be enhanced by the use of concurrent itraconazole.80 Thalidomide may cause a sensory neuropathy, and up to 50% of patients do not recover, whereas the sensory neuropathy caused by cisplatin may persist and can affect up to 20% to 60% of patients.79
Monitor the patient with an H&P (with a neurologic examination) yearly.
Leukoencephalopathy, defined as demyelination, white matter necrosis, calcification, and glial damage, may develop after CNS radiation and/or intrathecal methotrexate. It presents with seizures, focal motor signs, dementia, ataxia, and cognitive abnormalities. The risk correlates with methotrexate dose and route, age at treatment, and amount of radiation.79 A computed tomography scan of the brain may be used to diagnose this condition.
Eye
Cataracts may be observed and can be caused by radiation, steroids, and busulphan. Cranial radiation may result in keratoconjunctivitis.81
Monitor the patient with an H&P (with lens examination) yearly.
Ear
Sensorineural, high-frequency hearing loss may develop after cisplatin therapy, most often observed at a cumulative dose approaching 400 mg/m2.82 Ifosfamide and cranial radiation may exacerbate cisplatin-related hearing loss.83 New therapies for preventing cisplatin-induced ototoxicity are in the early stages of development. It is thought that cisplatin suppresses the formation of endogenous antioxidants that protect the inner ear against reactive oxygen species, and these new therapies are free-radical scavengers.84
Monitor the patient with a yearly H&P and hearing testing. Patients should be carefully evaluated for speech delay or abnormality as well as school performance, because either may be the first manifestation of hearing loss.
Teeth and Gums
Defective dentition, increased caries, root abnormalities, and periodontal disease can occur. These oral abnormalities may be caused by radiation to the area, and there is an increased risk if >2.5 Gy is used or the patient is <5 years of age. Even 0.4 Gy may cause some problems. Chemotherapy may also cause dental defects.85,86
Monitor the patient with a yearly H&P and dental examination by a dentist.
|
SECOND MALIGNANT NEOPLASMS RESULTING FROM CANCER THERAPY |
|---|
|
TOPABSTRACTTHE PROBLEMNONMALIGNANT LATE EFFECTS OF...SECOND MALIGNANT NEOPLASMS...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
Skin Cancer
Skin cancer is probably the result of decreased immunosurveillance secondary to radiation and chemotherapy coupled with sun exposure. Malignant melanoma and nonmelanotic skin cancers represent 10% to 20% of second malignancies after cancer therapy, and there is an increased risk of developing skin cancer over time. There is also an increased incidence of melanocytic nevi, a risk factor for malignant melanoma, in uncommon sites such as the palms and soles of children receiving chemotherapy.89
A CCSS study of 13132 patients reported 213 patients (1.6%) with nonmelanotic skin cancers, which accounted for the most commonly observed (41%) subsequent cancer.90 Forty-six percent of the patients had multiple occurrences, 90% had received radiation, and 90% of the cancers were in the radiation field. The median age of occurrence was 31 years (range: 7–46 years). Radiation therapy was associated with a 6.3-fold increased risk of skin cancer.
Monitor the patient with an H&P and prevent skin cancer by advising the patient to avoid the sun, wear a hat, cover the skin, and have a yearly skin examination by a dermatologist.
Breast Cancer
BC may develop after thoracic radiation, usually after treatment for Hodgkin's lymphoma. There are 120000 survivors of Hodgkin's lymphoma in the United States, and SMNs are the leading cause of death for long-term survivors of Hodgkin's lymphoma, with BC the most frequent solid tumor in women who were treated for Hodgkin's lymphoma.91 The annual excess incidence of BC increases as the age of patients who are post–Hodgkin's lymphoma therapy increases. The risk of BC after thoracic radiation for Hodgkin's lymphoma in women <30 years of age is elevated 4- to 56-fold92,93 depending on the dose of radiation and the age at treatment, with the highest risk in women who were treated at 10 to 20 years of age.92–98 The risk of BC increases with increasing dose of radiation, with each gray unit received by any breast increasing the excess RR of BC by 0.13.98
In a study of adult survivors of pediatric Hodgkin's lymphoma from the Late Effects Study Group, no increased risk of BC was found if <26 Gy was given,93 and current protocols do not usually use doses greater than this. The risk of BC in a patient >30 years of age at diagnosis of Hodgkin's lymphoma declined significantly, and there was almost no increased risk of BC in women who were treated after they were 30 to 40 years of age.98,99 The median time to development of BC is 15 years after therapy,94 and hormonal stimulation may be necessary for the development of radiation-induced BC, because 1 study showed a decreased risk of BC associated with ovarian damage that resulted from alkylating agents or radiation.91
Most secondary BCs occur in the field of radiation and are invasive ductal carcinomas, although 13% are ductal carcinoma in situ. There is an increased incidence of BC in the contralateral breast.100 Whether there is an increased incidence of BC in males who receive thoracic radiation has not been determined yet.
Monitor the patient by having her perform monthly breast self-examination, have an H&P yearly, and, in addition, obtain a yearly mammogram beginning 5 years after diagnosis; this recommendation is endorsed by others,100 because a number of BCs after therapy for Hodgkin's lymphoma have occurred as early as 5 years after therapy. The COG recommends a mammogram 8 years after therapy or at age 25, whichever is later.15 Breast MRI is more sensitive for detection of BC than ultrasound or mammography in women who are at increased risks of BC secondary to BRCA gene mutations.101 Although it is not clear whether this particular surveillance method will result in reduced mortality, I believe that breast MRI rather than mammography should be used in patients who received thoracic radiation, especially those who were treated when radiation doses were larger and delivery was not as refined.
Thyroid Cancer
Thyroid cancer (TC) may develop after head, neck, or thoracic radiation. In a CCSS study of 14054 patients, of whom 69 developed TC,102 the incidence of TC in patients with any first malignancy who had radiation therapy to the head, neck, or thorax increased with an increasing dose of radiation. There was an increased incidence at 20 to 29 Gy but a decrease in incidence if >30 Gy (ie, killing effect). The estimated RR of TC was 1.32/Gy for a dose <15 Gy. The increased or decreased risk of TC was more pronounced in patients who were diagnosed with a first primary malignant disease before the age of 10 years as opposed to those patients who were older than 10 years. Chemotherapy had no effect on the development of secondary TC. In this study, patients younger than 10 years of age had a substantially higher TC risk over the entire radiation-dose range then did those patients who were older than 10 years, which demonstrated the increased susceptibility of the thyroid gland in young patients. Of the 69 patients with TC, 42% had a first diagnosis of Hodgkin's lymphoma and 20% had a first diagnosis of leukemia. Other studies have demonstrated a similar increased risk for TC after radiotherapy (RR: 18.357; RR: 15.659; absolute excess risk [AER]: 1.492; SIR: 36.493; SIR: 35 at 0.5 Gy and 73 at 3.6 Gy103).
A thyroid nodule can be palpated in 4% to 7% of adults,104 and the incidence of nodules is increased if ultrasound is used for detection or the presence of nodules is determined at autopsy.105 At a mean of 11 years after therapy for Hodgkin's lymphoma, up to 44% of childhood cancer survivors who received head or neck radiation had detection of nodules when screened by ultrasound.106 In a report of 647 pediatric patients treated for Hodgkin's lymphoma,107 it was determined that 67 (10.4%) of these patients developed 1 or more nodules. Four patients had thyroid disease before diagnosis, and 19 (28%) had received thyroid hormone replacement therapy before the diagnosis of a nodule. Seven patients (10%) had TC diagnosed at a median of 16.2 years after therapy for Hodgkin's lymphoma (range: 8.4–23.7 years). Only 1 TC was found by ultrasound, and the rest were found by palpation or clinical symptoms (53 of 67 had ultrasound). Forty-one of the 67 patients had asymptomatic nodules detected only by ultrasound. Thirty-four of the patients had a clinical course and imaging characteristics that were consistent with benign nodules and had no biopsy. The median size of the nodules was 0.7 cm (range: 0.2–2.0 cm), and 30% resolved. The majority of TCs resulting from radiotherapy are papillary,108 and they have a low mortality rate even after metastases.109
Monitor the patient with an H&P and testing of thyrotropin and free thyroxine levels yearly, and obtain a thyroid ultrasound every 3 to 4 years if there was no change from the previous ultrasound; otherwise, perform it every year. The role of thyroid ultrasound screening is controversial.108 Ultrasound all palpable nodules. Obtain a fine-needle aspiration if the nodule is enlarging or is >10 mm or if hypoechoic areas are seen on ultrasound.
Leukemia
Treatment-related acute myelogenous leukemia (AML)/myelodysplastic syndrome (MDS), known as t-AML/MDS,110 may be caused by alkylating agents (which are also linked to bone and bladder cancers) and topoisomerase 2 inhibitors (etoposide [VP-16] and teniposide [VM-26]).111 The long-term survival rate of pediatric patients with t-AML/MDS treated with allogeneic stem cell transplantation is 15% to 24%.112 t-AML/MDS is one of the few late effects for which the incidence plateaus or reverts to normal after a period of time.
With alkylating agents, the risk of t-AML/MDS increases with increasing dose of and age at treatment. The incidence of t-AML/MDS is 0.8% to 2.8%, and the median latency period is 4 to 6 years (range: 1–20 years).111 There is usually a loss or deletion of chromosome 5 or 7.113 With topoisomerase 2 inhibitors, there is an increased risk of t-AML (M4 or M5 morphology)/MDS with increasing dose, and it usually occurs in younger patients. The cumulative risk is 0.5% to 18.4%, depending on the dose intensity, and the median latency period is 1 to 3 years (range: 0.5–4.5 years).111 There is usually rearrangement involving the MLL gene on chromosome band 11q23.114
The short-term use of granulocyte colony-stimulating factor after etoposide therapy may increase the risk of developing t-AML/MDS.115 Although the risk of t-AML/MDS after the use of anthracyclines is thought to be relatively low, 1 study suggests otherwise.116 Survivors of pediatric Hodgkin's lymphoma have a 4 to 175 times increased risk for developing t-AML/MDS.97 The AER of t-AML/MDS was 6.3 in a study of 32591 adult patients with Hodgkin's lymphoma,92 whereas the SIR was 174.8 in a Late Effects Study Group report of 1380 pediatric patients with Hodgkin's lymphoma.93
Monitor the patient with an H&P and complete blood count yearly.
Mildly abnormal complete blood count values (low white blood count, platelets, and hemoglobin and an elevated mean corpuscular volume) are common in survivors of childhood cancer; they tend to persist, are of questionable significance, and do not seem to reflect preleukemic changes.117 Telomere shortening resulting from chemotherapy, which leads to repeated cycles of hematopoietic regeneration, is thought to be responsible for the hematopoietic stem cell injury.118
Sarcoma of Bone and Connective Tissue
Bone and soft tissue sarcomas may occur after radiation therapy, and the risk is proportional to the dose and the concurrent use of alkylating agents. In a report from the British Childhood Survivors of Cancer Study of 13175 patients diagnosed between 1940 and 1983,119 the cumulative probability of developing bone cancer after radiation therapy for the entire cohort was 0.9% within 20 years after treatment. The risk was increased to 7.2% after hereditary retinoblastoma (a malignancy that predisposes a patient to the development of osteosarcoma), 5.4% after Ewing sarcoma, and 2.4% after other malignant bone cancers. The RR of sarcomas, with bone and connective tissue being the most common sarcomas, after pediatric Hodgkin's lymphoma therapy was reported to be 1.3 to 37.1.97 In another study, the AER for bone/connective tissue cancers after therapy for Hodgkin's lymphoma was 2.3 in patients <21 years of age, and the AER declined in patients who were older when treated.92 In a report of 11183 patients <21 years of age,120 Ewing sarcoma–family tumors rarely occurred after treatment of primary childhood cancer (1.3% of 479 second cancers). Most of these tumors did not seem to be related to radiation therapy, and long-term survival was possible.
Monitor the patient with an H&P yearly.
Second Carcinomas Other Than Breast, Thyroid, and Skin
The CCSS group, in the largest study to date of secondary carcinomas, identified 71 carcinomas (excluding breast, thyroid, and skin) in 13136 patients diagnosed from 1970 to 1986 at <21 years of age.121 The overall SIR for a subsequent adult-type carcinoma was 4 (a fourfold increased risk) and was significantly elevated for all primary cancer diagnoses except CNS neoplasms. Patients with SMNs were more likely to be older and to have a primary diagnosis of Hodgkin's lymphoma, soft tissue sarcoma, or neuroblastoma, a history of a first-degree relative with cancer, and a history of alcohol use. The overall cumulative incidence of developing a second carcinoma was 0.45% at 20 years of follow-up. Table 1 lists the type of second carcinomas with the unadjusted SIR for both the risk of the carcinoma and the male and female distribution.
|
Survivors of Wilms' tumor had an increased risk of developing colorectal (SIR: 25.4) and other gastrointestinal carcinomas (SIR: 18), as did the survivors of Hodgkin's lymphoma (SIR: 2.5 and 7.4, respectively). For most sites, patients <10 years of age at diagnosis had a greater risk of developing secondary carcinomas. Patients who were treated with topoisomerase 2 inhibitors (etoposide, teniposide) had an increased risk of developing lung cancer (SIR: 73.4 vs 1.8 in nonexposed patients), as did patients treated with alkylating agents (SIR: 7 vs 0 in nonexposed patients). Platinum therapy was associated with an elevated SIR for colorectal (SIR: 14.7) and kidney (SIR: 48.7) carcinomas. Radiation therapy was associated with an increased risk of all secondary carcinomas except for the reproductive organs, and the SIR was most marked for head and neck carcinomas (SIR: 18.5 vs 2.3 in nonradiated patients). The site of the second carcinoma occurred in the previously irradiated field for all (4 of 4) lung carcinomas, in 85% (17 of 20) for head and neck carcinomas, and in 71% (10 of 14) of gastrointestinal carcinomas.
Of 71 patients, 22 (33%) had second carcinomas in an area that was not exposed to radiotherapy, and 11 (50%) of the 22 had no previous radiation. Sixteen of these patients (73%) received alkylating agents. Four (5.6%) of the 71 patients had received neither chemotherapy nor radiation therapy. The authors of this study cautioned that the sample size was too small to perform adjusted analysis to determine the independent contribution of treatment and patient factors.
In addition to a yearly H&P, monitor for colon cancer in those patients who received abdominal radiation with colonoscopy every 10 years beginning 15 years after therapy or at age 35, whichever is later; this is a COG recommendation.15
Brain Tumors Resulting From Treatment of All Childhood Malignancies and All SMNs Resulting From Treatment of Brain Tumors
Brain tumors resulting from cranial radiation have a latency period of 9 to 10 years, and the younger the age at treatment, the greater the risk.122 The CCSS reported 116 subsequent CNS neoplasms in 14361 5-year survivors of childhood cancers.123 Gliomas (n = 30) occurred at a median of 9 years from the original diagnosis, and the SIR was 8.7. Meningiomas (n = 66) occurred at a median of 17 years from the original diagnosis. Other subsequent CNS tumors were primitive neuroectodermal tumor (n = 6) and CNS lymphoma (n = 1). After adjustment for radiation dose, neither original cancer diagnosis nor chemotherapy was associated with an increased risk.
A National Cancer Institutes Surveillance and End Results (SEER) analysis of 2056 survivors of childhood brain tumors revealed that there were 39 SMNs.124 There was a 4.7-fold increase in risk in patients treated from 1979 to 1984 and a 6.7-fold increase in risk in those treated after 1985; the difference was possibly related to treatment intensity with the increasing use of chemotherapy in addition to the standard radiation therapy. For astrocytomas, the 3 most common SMNs were "other tumors" (5), followed by fibrosarcomas (3) and melanomas (3). For primitive neuroectodermal tumors, the 2 most common SMNs were "other tumors" (3) and acute lymphocytic leukemia (2). For other gliomas, astrocytomas (3) were the most common SMN. The mean age at primary diagnosis was 9.5 years, and there was a median time to SMN diagnosis of 14.1 years.
Monitor the patient with an H&P (with neurologic examination) yearly.
Bone Marrow and Stem Cell Transplantation
In addition to the usual late effects of chemotherapy and radiation therapy, patients posttransplant are also at risk of developing chronic graft-versus-host disease and all its attendant complications.125–127 They are also at risk for developing secondary malignancies.128 In 1 study, the SIR for second malignancies was 5.4 (excluding posttransplant lymphoproliferative disorder, which usually develops within the first year after transplant), and the SIR for t-MDS/AML was 300.129
To facilitate the use of the information provided above, I have formulated a table (Table 2) that lists radiation-therapy site and chemotherapeutic agents and selected late effects resulting from their use. The reader may refer back to the section on each individual organ, both in the nonmalignant and malignant late-effects section, for a more detailed discussion of the late effects of cancer therapy on each organ system as well as recommendations for patient monitoring.
|
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTTHE PROBLEMNONMALIGNANT LATE EFFECTS OF...SECOND MALIGNANT NEOPLASMS...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
I believe that in addition to being followed by their primary care physician, all long-term survivors of childhood cancer therapy should attend a specialized late-effects clinic yearly and be evaluated by a member of the oncology team, either the physician or the pediatric oncology nurse practitioner, who initially treated them. A psychologist and social worker should always be present, and subspecialists should be available on or near the site as needed. Ideally, such clinics should be located in the same center in which the patient was initially treated but scheduled at a different location or time from the regular hematology/oncology clinic. Four possible models of late-effects clinics with their advantages and disadvantages have been described,130 and a clinic-based comprehensive care model for survivors of childhood cancer has been proposed.131 Recently, 11 review articles that dealt with a number of issues concerning cancer survivorship (including follow-up, late effects, models for delivering care, patient education, patient advocacy, employment and psychosocial concerns, health promotion, and research directions in survivors of pediatric and adult cancers) have been published.132
With no or less radiotherapy being delivered with newer equipment in better fractionation schedules, and the replacement of or the use of reduced doses of second-cancer–inducing chemotherapy, the late effects resulting from such current treatment will decrease when compared with those reported here. In fact, for the most recent protocols for some patients with Hodgkin's lymphoma, radiation therapy has been omitted or delivered at lower doses, as it has in the treatment of some abdominal neoplasms, whereas in childhood acute lymphocytic leukemia it is used only in high-risk patients. However, there are still many long-term survivors of childhood cancer therapy who were treated 10 to 50 years ago, and these patients need particularly close monitoring.
The risk estimates reported here, which were determined from patients treated a number of years ago, are probably too high for current use, because cancer therapy has changed dramatically in recent years; however, new cancer therapies used now or in the future will, in all likelihood, be associated with their own late effects. Therefore, patients who are treated with these new therapies must be monitored closely to assess the magnitude of any late effects. On the COG Web site15 there are patient-education handouts that described the late effects of cancer therapy, which should be given to patients during their first visit to the late-effects clinic. For readers interested in the embryonic field dealing with the late effects of cancer therapy in adult survivors of adult cancer, the Institute of Medicine report on this subject can also be found on the Web.133
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Address correspondence to Joseph D. Dickerman, MD, University of Vermont College of Medicine, Department of Pediatrics, 89 Beaumont Ave, Given D201, Burlington, VT 05405-0068. E-mail: joseph.dickerman{at}uvm.edu
The author has indicated he has no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTTHE PROBLEMNONMALIGNANT LATE EFFECTS OF...SECOND MALIGNANT NEOPLASMS...SUMMARY AND CONCLUSIONSREFERENCES |
|---|
