Limitations on Physical Performance and Daily Activities among Long-Term Survivors of Childhood Cancer FREE

Long-term survivors of childhood cancer may be affected by treatment-associated impairments that limit physical performance and restrict participation in activities of daily living. Prevalence of these impairments and their association with particular types of cancer or treatment methods are not known.

Patients surviving childhood cancer for at least 5 years were compared with their siblings who did not have cancer. All survivors were at an increased relative risk for persistent impairment, but survivors of brain and bone cancer were particularly likely to report performance limitations and restrictions in routine activities.

Despite substantial improvements in treatment of childhood cancer, profound alterations in quality of life often persist into adulthood.

—The Editors

Profound advances in treatment efficacy and a concomitant increase in the number of long-term survivors have generated a need for investigations into the adverse effects of childhood cancer and its treatment (1–5). Surgery, chemotherapy, and radiotherapy are essential to providing a long-term cure in many children with malignant and nonmalignant neoplasms. Surgical procedures can result in structural and functional changes, which can have long-term implications in young children who are growing and developing. Chemotherapy and radiotherapy are cytotoxic and thus can have deleterious effects on developing organ systems (6–18). Long-term survivors of childhood cancer are at risk for many late effects, including early death, second cancer, and organ system dysfunction (19–22). Endocrine, musculoskeletal, neurologic, sensory, cardiac, and pulmonary impairments also have been documented (1, 19–20, 22–29).

A malignant disease or its treatment may cause long-term impairments in intellectual, emotional, and physical domains that diminish functioning (performance limitations). Performance limitations can restrict the survivor's ability to participate fully in daily activities necessary for self-care, home management, or work (participation restrictions). Delayed consequences of therapy are of special interest in children with cancer because of their young age and because of the potential remaining years of life that may be affected (30). Performance limitations and participation restrictions have been reported in as many as 69% of survivors of childhood cancer (31–33).

The aims of this analysis were to estimate the prevalence of performance limitations and participation restrictions among a cohort of long-term survivors of childhood cancer. We compared these results with similar findings in a comparative group of siblings of survivors. In addition, we evaluated whether performance limitations and participation restrictions were affected by cancer type; treatment methods; or delayed effects of the cancer, its treatment, or both.

The Childhood Cancer Survivor Study (34) is an ongoing epidemiologic follow-up study of survivors of childhood cancer who received their primary treatment at 1 of 26 collaborating institutions. Eligibility was restricted to patients who received their diagnosis between 1970 and 1986; whose age at diagnosis was 20 years or younger; and who survived at least 5 years after receiving the diagnosis of one of the following diseases: primary brain cancer, leukemia, Hodgkin disease, non-Hodgkin lymphoma, kidney tumor, neuroblastoma, soft-tissue sarcoma, or malignant bone tumor. The Human Subjects Research Review Committees at the University of Minnesota (the study coordinating center) and each collaborating institution approved the protocols and documents of the Childhood Cancer Survivor Study. Participants provided informed consent to participate in the study and separate consent to allow the release and abstraction of medical records, including treatment records.

Siblings of the survivors were recruited as a comparison group. Of the enrolled survivors who indicated that they had a living sibling, 5800 whose randomly generated 5-digit registration numbers ended with an odd digit were asked to seek permission for the investigators to contact the sibling who was closest in age to them. Four thousand six hundred ninety-one siblings (80.9%) agreed to be contacted.

Data collection was conducted by abstracting treatment-related information from medical records and by having participants, or a parent if the participant was younger than 18 years of age at the time of the survey, complete the Long Term Follow Up Study Questionnaire (35), a 24-page document that contains questions regarding the presence and timing of medical conditions, the presence of physical limitations, and the presence of current participation restrictions.

Performance limitations and participation restrictions were the primary outcome variables for this analysis. Physical performance was scored by adding the answers to a series of 6 questions about the participant's performance of particular physical activities during the past 2 years. We asked, “Over the past 2 years, how long (if at all) has your health limited you in each of the following activities: 1) vigorous activities like lifting heavy objects, running, or participating in strenuous sports; 2) moderate activities like moving a table, carrying groceries, or bowling; 3) walking uphill or climbing a few flights of stairs; 4) bending, lifting, or stooping; 5) walking 1 block; and 6) eating, dressing, bathing, or using the toilet?” Scores of 1 to 3 were assigned to each of the 6 questions. A score of 1 indicated that the participant was limited for more than 3 months, a score of 2 indicated that the participant was limited for 3 months or less, and a score of 3 indicated that the participant was not limited at all. A lower score indicated a greater degree of limitation. On the basis of graphical examination of the distribution of physical performance for siblings and survivors, this continuous variable was dichotomized by using a natural cutoff value at or below the 10th percentile of the distribution of the sibling group. Survivors and siblings at or below this cutoff point were classified as having a performance limitation.

Participation restrictions were evaluated in 3 separate categories: limited personal care skills; limited routine activities; and poor health preventing school or work attendance. A participation restriction was considered present if participants had a positive response to the following questions: 1) “Because of any impairment or health problem, do you need the help of other persons with personal care needs, such as eating, bathing, dressing, or getting around the home?” 2) “Because of any impairment or health problem, do you need the help of other persons in handling routine needs, such as everyday household chores, doing necessary business, shopping, or getting around for other purposes?” 3) “Does any impairment or health problem keep you from holding a job or attending school?”

Cancer type, treatment groups, and late effects were considered risk factors of interest for these analyses. Information about the original diagnosis was obtained from the treating institution, and information on primary therapy was abstracted from the medical record. Because most survivors had had surgery, therapy was grouped into broad categories that included surgery only; radiation with or without surgery (radiation); chemotherapy with or without surgery (chemotherapy); chemotherapy and radiation with or without surgery (chemotherapy with radiation); treatment other than surgery, radiation, or chemotherapy; or unknown. Late effects were defined as dichotomous indicators based on participants' self-reporting of answers to questions that asked if a doctor or other health care professional had ever told them that they had a particular condition. A late effect was considered prevalent if a participant answered “yes” to condition-specific questions. Late effects included impairment to any body system that occurred from the onset of the initial therapy to the time when the baseline questionnaire was completed. Participants who answered “don't know” or who left a question blank were included in the “no” or “not present” category for that variable. Late effects were grouped as follows according to major body system: endocrine impairments; bone and joint health; neurologic impairments; sensory impairments; cardiac impairments; and pulmonary impairments, as previously described (25, 27, 36).

The prevalence of performance limitations and participation restrictions among survivors was compared with that among the participating siblings, using generalized estimating equations with a Poisson distribution and a log link (37–39), and was reported as a standardized risk ratio (RR). These methods were used in all models to account for the possible correlation between survivors and siblings from the same family (40). The Poisson distribution with a log link was selected to allow model convergence with conservative CIs. Cancer type, treatment group, and presence of a late effect were also evaluated in relation to the outcome variables in generalized estimating equations. All models were adjusted for the participant's age at interview and sex. The model evaluating treatment was also adjusted for cancer type. Because the participant's age at diagnosis and race and ethnicity were not independent predictors of the outcomes, and because they did not appreciably alter the estimates, they were not included in the final models. To confirm the results we obtained using generalized estimating equation models, we also completed matched analyses for each outcome, limiting these models to only survivors with a sibling in the comparison group. To evaluate the sensitivity of the results to dichotomization of the physical performance variable, this outcome measure was also evaluated as a continuous variable comparing survivors with siblings overall and by cancer type in models again adjusted for age, sex, and intrafamily correlation. We used SAS, version 9.1 (SAS Institute, Inc., Cary, North Carolina), for all analyses.

The Childhood Cancer Survivor Study is funded by grant CA 55727 from the National Cancer Institute, Bethesda, Maryland, with additional support for data collection and personnel provided to the University of Minnesota from the Children's Cancer Research Fund. The funding source had no role in the design, conduct, or reporting of the study or in the decision to submit the manuscript for publication.

Among the 20 720 five-year survivors identified by the collaborating institutions, 14 372 (69.4%) were enrolled and completed an interview at the time of this analysis, 3189 (15.4%) declined to participate, and 3159 (15.2%) were lost to follow-up and were never offered enrollment. Among the 4691 eligible siblings, 3839 (83.9%) were enrolled and completed an interview, 551 (11.7%) declined to participate, and 301 (6.4%) were lost to follow-up and were never offered enrollment. Among the enrolled participants, 1880 (13.1%) did not consent to medical record abstraction and 1011 (7%) were not alive at the baseline interview. Participants who died before the baseline interview, although eligible to participate in the Childhood Cancer Survivor Study through proxy report, were not included in these analyses because the questions about performance and participation asked about activity levels during the past 2 years.

Participants did not differ from nonparticipants regarding sex, cancer diagnosis, age at diagnosis, age at survey, and type of treatment (34). This analysis included 11 481 survivors who consented to medical record abstraction and who were alive at the time of survey completion and 3839 siblings (Figure), including 3266 survivor–sibling pairs.

The characteristics of the study participants are shown in Table 1. The median age at survey completion was 23 years (range, 8 to 47 years) for the 11 481 survivors and 26 years (range, 5 to 56 years) for the siblings. Forty-eight percent of survivors and 52% of siblings were female. Most survivors and siblings were white (88.3% and 87.5%, respectively). The most common diagnosis was leukemia (33.9%), and the most common treatment method was chemotherapy with radiation (53.8%). Most survivors (63.2%) received their cancer diagnosis at 9 years of age or younger, and when interviewed, 88.8% had survived at least 10 years since the diagnosis.

The percentages of survivors and siblings with self-reported performance limitations or participation restrictions and the standardized RRs adjusted for age, sex, and intrafamily correlation with 95% CIs are shown in Table 2. After adjustment for age and sex, survivors (19.6%) were nearly twice as likely to report performance limitations when compared with siblings (11.8%) (RR, 1.8 [CI, 1.7 to 2.0]). When evaluated as a continuous variable, after adjustment for age and sex, mean physical performance score also differed between survivors (mean, 16.6 [SD, 2.8]) and siblings (mean, 17.2 [SD, 2.1]) (P < 0.001). Participation restrictions were less common than performance restrictions among survivors (range, 2.9% to 7.9%) and siblings (range, 0.5% to 1.5%); however, survivors were 4.7 times (CI, 3.0 to 7.2 times) more likely than siblings to report restricted abilities to perform personal care or routine activities like shopping or housework (RR, 4.7 [CI, 3.6 to 6.2]). Survivors were 5.9 times (CI, 4.5 to 7.6 times) more likely to report that poor health prevented them from attending school or work compared with siblings. Similar to findings in our overall analysis, a matched analysis that included only those survivors with complete data who had a sibling in the comparison group yielded RRs of 1.6 (CI, 1.4 to 1.9) for performance limitations, 5.2 (CI, 2.8 to 9.5) for restricted personal care skills, 4.0 (CI, 2.7 to 5.9) for restricted routine activities, and 5.2 (CI, 3.5 to 7.7) for poor health preventing school or work attendance.

The percentage of survivors reporting performance limitations was 19.6% among all cancer types, with the highest percentages reported among survivors of bone (36.9%) and brain cancer (26.6%) (Table 2). The survivors of bone and brain cancer also had the lowest mean physical performance scores (mean, 15.3 [SD, 3.3] and 15.8 [SD, 3.5], respectively). In an age- and sex-adjusted model, compared with siblings, survivors of bone cancer were 2.9 times (CI, 2.6 to 3.3 times) as likely, survivors of brain cancer were 2.5 times (CI, 2.2 to 2.8 times) as likely, and survivors of Hodgkin lymphoma were 1.8 times (CI, 1.6 to 2.0 times) as likely to report a performance limitation (Table 2).

The prevalence of participation restrictions also varied by cancer type. Survivors of brain cancer were most likely to report restrictions in personal care activities (10.5%) or routine activities (20.9%) and were most likely to report that poor health prevented them from attending school or work (20.0%). Among survivors of bone cancer, 11.2% reported that poor health prevented them from attending school or work (Table 2). The standardized RRs comparing the 3 dimensions of participation limitations among survivors and siblings clearly indicated that, although all survivors had an increased likelihood of adverse participation outcomes, survivors of brain cancer were at the greatest risk for a functional loss. When compared with siblings, survivors of brain cancer had 17 times (CI, 10.9 to 26.5 times) the risk for a restriction in the ability to perform personal care skills, 14.9 times (CI, 11.2 to 19.8 times) the risk for a restriction in the ability to participate in routine activities, and 15.6 times (CI, 11.8 to 20.5 times) the risk for not participating in school or work (Table 2).

After adjustment for age, sex, and cancer type, survivors who received radiation were more likely to report limitations in physical performance when compared with survivors who received only surgery (RR, 1.4 [CI, 1.1 to 1.7]). Those who received radiation were also more likely to report restrictions in their ability to participate in self-care activities (RR, 1.7 [CI, 1.1 to 2.6]), routine activities (RR, 1.9 [CI, 1.4 to 2.6]), or work or school (RR, 1.8 [CI, 1.3 to 2.3]) than were those who received surgery only. The chemotherapy and radiation group was more likely than the surgery-only group to report limitations in physical performance (RR, 1.4 [CI, 1.2 to 1.7]) and restrictions in their abilities to do activities required for self care (RR, 2.3 [CI, 1.5 to 3.4]) and to have difficulty participating in routine activities like shopping and housework (RR, 2.3 [CI, 1.7 to 3.1]). This group was also more likely than the surgery-only group to report that their poor health prevented them from attending school or work (RR, 2.3 [CI, 1.3 to 8.1]).

The percentage of survivors and siblings reporting specific late effects is shown in Table 3. Late effects of the neurologic system were highly prevalent, with 42.6% of survivors versus 18.8% of siblings reporting at least 1 problem in this category. Prolonged pain or an abnormal sensation in the trunk or extremities was the most commonly reported neurologic impairment among survivors (31.1%) and siblings (14.4%). Weakness in the arms or legs was reported by 9.2% of survivors. Survivors also frequently reported late effects of the endocrine, cardiac, and pulmonary systems; hypothyroidism (11.2%), chest pain with exertion (10.4%), and chronic cough (7.4%) were the most prevalent. The standardized RRs comparing the prevalence of late effects among survivors and siblings are also shown in Table 3. Survivors were more likely than siblings to report late effects in every category. Survivors were 6.8 times (CI, 5.7 to 8.2 times) more likely to report any endocrine problem and 8.5 times (CI, 6.5 to 11.2 times) more likely to report any musculoskeletal problem than were siblings. The risk for any neurologic impairment among survivors was 2.4 times (CI, 2.2 to 2.5 times) that of siblings. Sensory problems were 4.3 times (CI, 3.5 to 5.3 times) more common among survivors than in siblings. The risks for late effects of both the cardiac and pulmonary systems among survivors were at least twice as high as those among siblings (RRs, 2.0 [CI, 1.8 to 2.2] and 2.4 [CI, 2.1 to 2.8], respectively).

In an analysis limited to survivors only, late effects were associated with performance limitations and with participation restrictions in all categories. After adjustment for age, sex, and each medical category of late effects, survivors with musculoskeletal impairments versus those without musculoskeletal impairments (RR, 1.9 [CI, 1.7 to 2.0]) and survivors with neurologic impairments versus those without neurologic impairments (RR, 2.0 [CI, 1.9 to 2.2]) had the highest risk for reporting a performance limitation. Survivors with neurologic (RR, 7.6 [CI, 5.5 to 10.6]) or sensory impairments (RR, 2.7 [CI, 2.1 to 3.4]) versus those without such impairments had the highest risk for reporting a restriction in their ability to perform personal care skills. Neurologic and sensory impairments were also associated with restrictions in the survivor's ability to participate in routine activities (RRs, 5.6 [CI, 4.6 to 6.9] and 2.6 [CI, 2.3 to 3.0], respectively) and to attend work or school (RRs, 5.3 [CI, 4.5 to 6.4] and 2.0 [CI, 1.7 to 2.3], respectively).

In this study, a statistically significant proportion of long-term survivors of childhood cancer reported late effects that interfered with physical performance and restricted participation in daily activities. The greatest risk for performance limitations was observed among those who survived brain cancer, bone cancer, neuroblastoma, soft-tissue sarcoma, or Hodgkin lymphoma. Survivors of brain and bone cancer had the highest prevalence of participation restrictions.

Our results are supported by studies suggesting that individuals who survive specific types of childhood cancer are at risk for long-term deficits that limit physical performance and restrict participation in activities of daily living. Previous reports indicate that survivors of brain cancer are particularly vulnerable to neurologic sequelae and poor physical outcomes (27, 41–46). In our study, survivors of childhood brain cancer were 2.5 times more likely than siblings to report performance limitations and up to 17 times more likely to report a participation restriction. Balance deficits, ataxia, paralysis, and difficulty walking have been reported among survivors of brain cancer in many studies that used various study designs (27, 41–46).

Survivors of bone cancer and soft-tissue sarcoma are vulnerable to physical limitations and participation restrictions because surgical procedures can result in damage to, or loss of, structures required to execute movement. In our study, more than one third of the survivors of bone cancer reported physical limitations and 11% reported that poor health restricted their ability to attend work or school. Children who underwent amputations and those who had limb-sparing surgery were at risk for physical performance limitations (47). Limb-sparing procedures can be complicated by frequent infection, delayed healing, nonhealing fractures, delayed weight-bearing, and early immobility (48), all of which are associated with a loss of long-term function (49–58).

Performance limitations may also be problematic in survivors of neuroblastoma, primarily because such patients can have a multitude of late effects of the musculoskeletal and neurologic systems. In our study, approximately 6% of survivors in this category reported difficulty with routine activities like shopping or housework. In the Childhood Cancer Survivor Study, most long-term survivors of neuroblastoma are probably those who had low- or intermediate-risk disease and were treated in an era that included radical surgery, aggressive chemotherapy, and possibly orthovoltage (x-ray) radiation (59–64). Previously reported treatment complications include musculoskeletal impairments (scoliosis, kyphosis, hypoplasia and fibrosis of bone and soft tissue, and slipped capital femoral epiphysis) (59–61) and neurologic problems (paresthesias, paresis, paraplegia, and neurogenic bladder) (59–61, 65). Both musculoskeletal damage and neurologic dysfunction can be related to damage from the tumor itself or from the surgical procedure used to remove the tumor (62–64, 66–67).

In our study, the survivors of Hodgkin lymphoma were nearly twice as likely as siblings to report performance limitations but were not as likely as survivors of brain cancer, bone cancer, or neuroblastoma to report difficulty with activities of daily living, routine activities, or interference with work or school attendance. Late effects in survivors of Hodgkin lymphoma are more likely to be impairments of the cardiac or pulmonary systems (23, 68–69); initial damage may present with subtle symptoms that interfere only with activities requiring vigorous or sustained physical performance, which are easily avoided in today's environment.

Our study has several methodologic limitations. It should be noted that treatment methods have changed for many types of cancer since the participants in this study received their diagnoses. However, our results are applicable to the thousands of survivors of childhood cancer who were treated between 1970 and 1986. Participation bias must also be considered. If participants were healthier or more physically capable than eligible nonparticipants, our proportions of physical limitations and participation restrictions are probably underestimated. Conversely, if participants were more motivated to participate in this study because they had deficits, our proportions of physical limitations and participation restrictions are probably overestimated. The classification algorithm for performance limitations must also be considered when interpreting the RRs. A performance limitation was determined by dichotomizing a variable originally scored on an integer scale with a 12-point range. A natural cutoff point based on the distribution of the data was used for the dichotomy; however, as in any measure of this type, it is possible that participants nearest to the cutoff point on either side of the dichotomy may have been classified incorrectly. In addition, participants who answered “don't know” or who left a question blank for any of the outcome variables or the late effect variables were included in the “no” or “not present” category for that variable. Participants were more likely than siblings to answer “don't know” or to leave a blank response (total siblings vs. total survivors, 0.65% vs. 3.39%; P <0.001). If these participants had a late effect and chose not to report it, or inadvertently omitted it, our prevalence values may be underestimated.

Because of the nature of the questionnaire, we could not characterize the severity of performance limitations or activity restrictions among survivors and were unable to account for the possibility that participants may have occasionally (rather than regularly) missed work or that they may have needed help with shopping but not with household chores. Specific measures of impairments related to physical performance and activity tolerance (for example, muscle strength and aerobic capacity) would allow development of targeted intervention programs. It is also possible that factors other than cancer type, treatment, age, sex, or late effects explain performance limitations and participation restrictions among survivors.

Despite these limitations, our analysis adds important data to the literature on the prevalence and type of functional limitations experienced by long-term survivors of childhood cancer. Although not all late effects can be avoided or eliminated, their impact on physical performance and participation in routine activities can be influenced by rehabilitation measures designed to restore function or remediate loss in physical performance that may lead to participation restrictions. Adult survivors of childhood cancer are at risk for performance limitations and participation restrictions many years after treatment. They should be monitored for functional loss throughout their lives and referred for appropriate rehabilitation services when indicated.