I feel that having a life-threatening illness, being treated for it and surviving, can trigger strong emotions—anxiety, fear, anger, and sadness. If these feelings aren’t acknowledged, addressed, and treated, they can over time evolve into more chronic problems like panic disorders and depression.—Childhood cancer survivor1 (p34)
Cure-directed therapy given during the critical developmental phases of childhood and adolescence has resulted in well-documented acute and late effects that can affect healthy survivorship. For this reason, state-of-the-art pediatric oncology clinical care and research emphasize improving both the duration and quality of survival, in part by better understanding the lived experience—the health-related quality of life (HRQOL)—of young people treated for cancer. The recent shift toward including patient-reported outcomes (PROs) as endpoints in clinical trials dovetails with increasing recognition of HRQOL as a critical dimension of cancer survivorship.2–7
Quality of life (QOL) is a broad multidimensional concept that includes subjective evaluations of both positive and negative aspects of life. Health-related QOL encompasses components of overall QOL that can clearly be shown to affect health, including physical, psychological, and social domains. The construct also may include the impact of disease-specific and treatment-related symptoms on self-perceived functioning, as well as spiritual and existential aspects of the experience.8 Hinds et al9 have advanced a pediatric oncology-oriented definition of HRQOL, based on interviews with children and adolescents receiving treatment for cancer, as “an overall sense of well-being based on being able to participate in usual activities, to interact with others and feel cared about; to cope with uncomfortable physical, emotional and cognitive reactions; and to find meaning in the illness experience.”
Outcome studies of psychosocial HRQOL in pediatric cancer survivors have had mixed results that range from reports of psychological dysfunction and poor psychosocial HRQOL to findings of normal/near-normal scores on standardized measures and recent documentation of perceived higher HRQOL.2 These discordant study findings may be the result of using different instruments and different reporting sources (ie, self-report vs proxy report), variability in whether study subjects were treated in earlier eras before cancer treatment regimens were risk modified to minimize toxicity that might affect HRQOL, variability in the availability of supportive services and interventions, temporal variations in the timing of assessment, as well as variable population or individual/family characteristics that could directly and/or indirectly influence pathways to HRQOL. In addition, whether and which theoretical frameworks have been used to guide examination of psychosocial HRQOL have varied across studies. Thus, we do not yet have a solid evidence base to guide clinical interventions for children and adolescents who are exposed to the extensive stressors and adjustments related to a cancer experience.
Patients in the adolescent age group are at potentially increased risk for adverse impact on their HRQOL because they are confronted simultaneously with the dual challenges of having had cancer and accomplishing their developmental transition to emerging adulthood.10 Typical adolescence, spanning the time from puberty to maturity, is a period of accelerated physical, cognitive, and emotional growth and development that carries special vulnerabilities and risks. Particular challenges for adolescents during and after cancer treatment include issues of body image and sense of self, peer relationships, family relationships, sexual health, and health risk/promoting behaviors including adherence to treatment regimens and posttreatment heath surveillance recommendations. Although these issues may conceivably affect HRQOL, most research examining HRQOL in pediatric oncology has focused on young children undergoing treatment or on adult long-term survivors, rather than on adolescents.11
Young people may experience varying threats to HRQOL during the different phases of the cancer treatment-survivorship trajectory. The period of active cancer treatment, which can range from months to years, is often arduous, requiring numerous outpatient visits and periodic hospitalizations and leading to poor HRQOL compared with healthy peers.12,13 In the long-term posttreatment survivorship phase, as summarized in a recent systematic review of the literature, both negative and positive effects have been reported.14 Some of the reported determinants of HRQOL in this phase include physical health, perceived level of self-esteem, coping abilities, personality characteristics, hopefulness, social support, and overall experiences during treatment.2 In clinical practice, the immediate aftermath of cancer treatment is recognized as a time of increased stress, often marked by relief and celebration but also by feelings of anxiety, uncertainty, fear, and vulnerability. However, few studies have explored HRQOL in the transitional time between treatment completion and long-term survivorship.14
The theoretical framework for this study is drawn from the Pediatric Medical Traumatic Stress Model and from the Adolescent Resilience Model (ARM).15,16 The merged model incorporates ARM illness-related, family, and social protective/risk factors among the antecedent factors that combine with characteristics of the traumatic medical stress (ie, diagnosis and treatment of cancer) and the patient’s subjective interpretation of the Pediatric Medical Traumatic Stress Model to affect their experience during and after the peritrauma period. Coping was conceptualized as a mediator for the outcome variable of resilience/HRQOL. Covariates such as race/ethnicity, gender, socioeconomic status, and disease/treatment variables were considered as potential moderators in a moderated mediation path model (Figure). The aims of the present study were to assess adolescent cancer survivors’ psychosocial HRQOL within 6 months after treatment completion and to evaluate the relationships among demographic and clinical factors (eg, gender, age, race/ethnicity, treatment intensity, symptoms), coping skills, and psychosocial HRQOL.
This report represents a subanalysis from a prospective longitudinal study examining patient-reported psychosocial outcomes among childhood cancer survivors during the first 5 years of survivorship. Our focus is on the early transition to posttreatment survivorship, that is, within 6 months of completing treatment.
Patients 11 to 21 years of age were recruited between May 2005 and February 2007 from 3 pediatric cancer treatment centers. Patients became eligible for the study as soon as their cancer treatment was completed. Oncology advanced practice nurses identified potential participants according to eligibility criteria requiring that participants be disease-free, English speaking, and cognitively able to complete study questionnaires. After receiving the attending physician’s approval to approach the patient, the advanced practice nurses or research assistants introduced the study during the patient’s scheduled clinic visit. For patients younger than 18 years, signed assent was obtained from the patient and permission to participate was received from the parent. Patients older than 18 years provided consent on their own behalf.
Of 119 identified eligible patients, 102 (85%) were interviewed. Among those who declined to participate, the primary reason given (by 12/17 nonparticipants) was lack of interest. Comparison analyses showed that there were no statistically significant differences between nonparticipants and participants with respect to age, gender, race/ethnicity, and type or duration of treatment. Eight participants with incomplete information were excluded from analyses. The final analytic sample included 94 participants, 56 from Children’s Hospital Los Angeles (CHLA; Los Angeles, California), 27 from C.S. Mott Children’s Hospital (Ann Arbor, Michigan), and 11 from Miller Children’s Hospital (Long Beach, California).
Subjects participated in a 35-minute structured interview conducted by trained research assistants in-person or by telephone. In the interview, study team members collected information on demographics, patient-reported symptoms, physical and psychosocial QOL, and coping strategies. Disease and treatment information was abstracted from participants’ medical records. Before enrolling participants, study procedures were approved by each institution’s human subjects’ protection committee, in accordance with assurances filed with and approved by the US Department of Health and Human Services.
The study used PedsQL 4.0 Generic Core Scales–Adolescent and Young Adult Self-Report versions (23 items), which are developmentally appropriate questionnaires that measure dimensions of self-reported HRQOL.17 For each item, respondents indicate the extent to which it has been a problem (0 = never to 4 = almost always) during the preceding month. Items are reverse scored and linearly transformed to a 0-to-100 scale, with higher scores indicating better QOL. We calculated 2 summary scores: Physical Functioning (8 items) and Psychosocial Functioning (15 items, including emotional functioning, social functioning, and school functioning). The PedsQL 4.0 Generic Core Scales have been tested with pediatric oncology patients and demonstrate strong internal consistency and test-retest reliability; they also accurately differentiate between children with cancer and healthy controls.17 In the study sample, Cronbach’s α coefficients for the Physical and Psychosocial Functioning subscales were .89 and .87, respectively.
Using the Center for Epidemiological Studies Depression Scale (20 items),18 participants rated, along a 4-point ordinal scale ranging from “rarely or none of the time” (<1 day) to “most or all of the time” (5–7 days), how often they had symptoms during the previous week. A total score was calculated, with higher scores representing more elevated levels of depressive symptoms. The scale yields high internal consistency, adequate test-retest repeatability, and good construct validity in both clinical and community samples.19 The Cronbach’s α coefficient for the Center for Epidemiological Studies Depression Scale in the study sample was .91.
We used the PedsQL Multidimensional Fatigue Scale, General Fatigue subscale (6 items), which has the same format, instructions, Likert response scale, and scoring method as the PedsQL 4.0 Generic Core Scales.17 Items are reverse scored and linearly transformed to a 0-to-100 scale, with higher scores indicating fewer symptoms of fatigue. The scale has been tested with pediatric oncology patients and demonstrated strong internal consistency and test-retest reliability; it accurately differentiates between children with cancer on treatment and those off treatment and between children with cancer on treatment and healthy controls.17 The Cronbach’s α coefficient for the study sample was .90.
A single item rating of pain intensity was used.20 Using a visual analog scale, with 0 representing no pain and 10 indicating worst pain, subjects rated the average amount of pain experienced in the past 7 days. Visual analog scales have demonstrated good acceptability, responsivity, and validity among children older than 8 years; they met criteria for “well-established” measures in an evidence-based review of pediatric pain measures by a task force of the American Psychological Association in 2008.21
POSTTRAUMATIC STRESS SYMPTOMS
We used the University of California at Los Angeles Post-traumatic Stress Disorder Reaction Index for DSM-IV, Adolescent Version, a 22-item questionnaire that evaluates adolescent subjective reexperiencing, avoidance, and arousal reactions to traumatic experience. With this measure, which has been used previously in pediatric oncology research,22 subjects rate how much of the time during the past month they have experienced the symptom listed in relation to their cancer diagnosis (eg, “When something reminds me of cancer, I get very upset, afraid, or sad”). Using a 5-point Likert scale (0 = none and 4 = most of the time), a total posttraumatic stress severity score is calculated by summing 20 of the items. The University of California at Los Angeles Post-traumatic Stress Disorder Reaction Index has demonstrated internal consistency (α = .90) and good to excellent test-retest reliability.22 The Cronbach’s α coefficient in the study sample was .99.
The study used the Adolescent Coping Orientation for Problem Strategies Questionnaire (A-COPE), a 54-item generic questionnaire that evaluates strategies typically used by adolescents to cope with stressful situations.23 Respondents read statements that described behaviors for coping with problems and indicated, according to a 5 point Likert scale (1 = never to 5 = most of the time), how often they used each of the behaviors when facing difficulties or feeling tense. The A-COPE assesses 12 factors: ventilating feelings, seeking diversions, developing self-reliance, developing social support, solving family problems, avoiding problems, seeking spiritual support, investing in close friends, seeking professional support, engaging in demanding activities, being humorous, and relaxing. Internal consistency estimates vary across the 12 subscales and range from α = .50 to .75, with strong test-retest reliability (α = .75).24 In the study sample, the Cronbach’s α coefficients for the 12 subscales ranged from .51 to .78.
INTENSITY OF TREATMENT
Treatment intensity was graded using the Intensity of Treatment Rating Scale 2.0 (ITR-2), a valid and reliable scale that can be applied retrospectively across cancer diagnoses and treatments.25 Ratings were based on information retrieved from medical chart review, including disease, stage at diagnosis, treatment modalities, and relapse history by 2 study investigators (K.R., K.M.), with high κ for interrater reliability (κ = 1.0). Ratings were categorized by 4 levels of intensity: 1 = least intensive (eg, surgery only), 2 = moderately intensive (eg, pretransplant, chemotherapy, or radiation), 3 = very intensive (eg, ≥2 treatment modalities or metastatic disease), and 4 = most intensive (eg, relapse protocols, stem cell transplant). An updated version of the ITR was published after the completion of treatment intensity ratings in this study.26 Review of the ITR-3 revisions by the 2 study investigators who performed the treatment intensity ratings, however, determined that there were no changes in the ITR-3 that would alter the treatment intensity ratings for patients in this study.
Parental educational attainment was used as the indicator of socioeconomic status in this study. Although parental education level is only 1 aspect of overall socioeconomic status, it is commonly used in epidemiological studies and it has been shown to be a better predictor of children’s health than family income, number of parents, or family size.27
Data Analysis Plan
The study’s primary outcome variable was the PedsQL 4.0 Psychosocial Functioning summary score. Survivors’ race/ethnicity was classified into 3 categories: (1) white non-Hispanic, (2) Hispanic, and (3) other. Parental educational level was used as an indicator of socioeconomic status; when father’s education level was missing, mother’s educational level was substituted (correlation coefficient for mother’s and father’s education: r = 0.64, P < .0001). All symptoms (depression, fatigue, pain, and posttraumatic stress) were treated as continuous variables in univariate and multivariable analyses. For coping behaviors, we calculated mean subscale scores, and analyses were performed on these scores. Participants whose treatment intensity was graded as level 1 (least intensive) and 2 (moderately intensive) were grouped together in analyses because of the small number of participants (n = 2) in level 1.
Univariate and multivariable linear regression procedures were used to identify statistically significant clinical and demographic variables associated with psychosocial HRQOL. We included variables in the multivariable analyses that were found to be associated (P < .10) with psychosocial HRQOL in univariate analyses (ie, diagnosis of central nervous system [CNS] tumor, parent’s education, treatment with radiation therapy, duration of treatment, pain, fatigue, posttraumatic stress, avoiding problems, being humorous, seeking professional support, ventilating feelings), as well as variables with an a priori reason for inclusion (treatment center, gender, age at interview). Stepwise procedures were used to identify predictor variables that were statistically significant (P < .05 cutoff) for psychosocial HRQOL. The final (reduced) regression model controlled for treatment center and included only significant variables, yielding a model to explain the maximum amount of variation in the outcome. Data analyses were conducted using SAS statistical software (version 9.2; SAS Institute, Cary, North Carolina).
Characteristics of the Treatment Centers
The distribution of participants from CHLA (59%), C.S. Mott Children’s Hospital (29%), and Miller Children’s Hospital (12%) (Table 1) roughly approximates the proportions of newly diagnosed oncology patients at these large, medium, and small pediatric cancer treatment centers. Statistically significant differences were noted among the 3 centers for race/ethnicity (P < .0001), parental education (P = .04), type of diagnosis (P = .04), and treatment intensity (P = .03) (data not shown). Race/ethnicity and parental education were correlated, with a highly significant association between Hispanic ethnicity and lower parental education (P < .0001). Nearly all Hispanic patients (43/44) were recruited from the 2 California treatment centers. The percentage of patients with a diagnosis of leukemia from CHLA, Miller Children’s Hospital, and C.S. Mott Children’s Hospital was 37.5%, 63.7%, and 29.7% respectively. These differences reflect a larger proportion of patients with solid tumors at CHLA compared with the other 2 centers; CHLA has well-known programs for several rare malignancies included in the soft tissue tumor category (eg, neuroblastoma), and patients are referred from the Southern California region and beyond for these programs. Treatment intensity is reflected in this referral pattern as well, with CHLA having a higher proportion of patients receiving the most intensive treatment. There were no significant differences by institution in the outcome variable, mean PedsQL Psychosocial Functioning summary score. Therefore, all 3 treatment centers were retained in the dataset, and multivariable models were adjusted for treatment center.
Characteristics of the Study Participants
Treatment center and clinical and demographic characteristics of the sample are shown in Table 1. All patients had completed therapy within the past 6 months. Participants’ mean (SD) age at interview was 14.8 (2.74) years, with 54% of the participants 15 years or older. Fifty-two percent of participants were male adolescents and nearly half were Hispanic. Participants were diagnosed with a variety of cancers, most of which were treated with chemotherapy. Treatment was scored as “very” or “most” intensive in 70% of our sample. Length of treatment ranged from 1 to 119 months (mean [SD], 19.6 [18.43] months), with most patients receiving 1 to 12 months of treatment. There were 4 outliers with very long treatment duration; all 4 had relapsed disease and therefore received additional treatment.
The mean (SD) PedsQL Psychosocial Functioning Summary score for the study sample (Table 1) was 76.9 (14.53) (range, 43.3–98.3). Varni and colleagues28 have previously explored 1 standard deviation (SD) below the population mean as a meaningful cut point for an at-risk status for impaired HRQOL relative to the population sample. In this study, there were 17 subjects (18%) whose PedsQL Psychosocial Functioning summary scores were equal to or less than 1 SD below the mean (≤62.39).
As seen in Table 1, we found no significant differences in psychosocial functioning summary scores by treatment center, gender, race/ethnicity, age at diagnosis, or age at interview. Scores were significantly lower for patients diagnosed with CNS tumors (P = .01) and for patients treated with radiation therapy (P = .01). There was a positive association between parental education and psychosocial functioning score (P for trend = .01). There was a statistically significant negative association between treatment duration of 13 to 24 months and psychosocial functioning score (P < .009). As shown in Table 2, there were several significant differences in psychosocial functioning summary scores by all 4 patient-reported symptoms, physical functioning summary score, and 4 coping behaviors.
Pain, fatigue, depression, and posttraumatic stress were associated with lower psychosocial functioning scores (P < .0001); these symptoms also were significantly correlated with each other (r = 0.36–0.75) (Table 3). Higher self-rated physical functioning scores (Table 2) were associated with higher psychosocial functioning scores (P < .0001). Three coping behaviors had statistically significant negative associations with psychosocial functioning scores; these included avoiding problems (P < .003), seeking professional support (P = .01), and ventilating feelings (P < .0001). One coping behavior, being humorous, was positively associated with psychosocial functioning scores (P = .04).
In the final reduced multivariable model (Table 4), there were statistically significant positive associations with age at interview (P for trend <.002) and the coping strategy of being humorous (P = .03). There was a negative association between psychosocial functioning scores and the symptoms of pain (P = .03), fatigue (P < .0001), and posttraumatic stress (P < .0001). The final model explained 67% of the variance in psychosocial functioning scores among our sample of 94 adolescent cancer survivors within 6 months after treatment completion.
Although previous studies have examined psychosocial outcomes for young people during active cancer treatment or during long-term survivorship, the patient-reported experience of adolescents during the transition between treatment completion and survivorship (ie, within 6 months after treatment completion) has been relatively neglected, yet this phase might offer a window of opportunity when interventions may prove useful in minimizing the adverse psychosocial effects of the cancer experience and improving long-range HRQOL. Wakefield and colleagues14 systematically reviewed 19 studies (4 qualitative, 13 quantitative, and 2 using mixed methodology) of the psychosocial impact of completing childhood cancer treatment; of these, 11 included young people in the adolescent age range, but none had an assessment point earlier than 1 year after treatment completion. To our knowledge, only 2 studies, both conducted in the Netherlands, have addressed the HRQOL of adolescents shortly after treatment completion. Stam and colleagues29 found that adolescent patients scored worse than the general population did on motor functioning (walking, running, and endurance). Engelen and colleagues30 found that when adolescent cancer patients were compared with general and chronically ill groups, they self-reported lower physical well-being but better HRQOL on social scales. To these we add our study, the first to focus on adolescents’ psychosocial HRQOL, in association with symptom experience and coping strategies, in the stressful transitional phase soon after treatment completion.
On average, adolescent survivors in this study rated their overall psychosocial HRQOL as good. Subjects’ mean self-reported PedsQL Psychosocial Functioning scores soon after cancer treatment completion (76.9) fell between the mean scores reported for healthy children and children with cancer (Table 5). However, the range of scores in the study sample was wide (43–98), and 18% of psychosocial functioning summary scores in our sample were at least 1 SD below the mean, a cut point previously explored as a clinically meaningful indicator of risk for impaired HRQOL.28 This is a risk group that might benefit from early identification of, and attention to, modifiable factors that impact psychosocial HRQOL.
Findings from univariate analyses showed that survivors who had a diagnosis of CNS tumor or treatment with radiation rated their psychosocial HRQOL as significantly lower than did patients with other cancer diagnoses or types of treatment. In our study, this association lost significance in the multivariable model, but it may merit further exploration in longitudinal studies because it has been reported that psychosocial HRQOL may decline over time among childhood cancer survivors treated for CNS tumors.31 Lower HRQOL among individuals with CNS tumors or radiation treatment is consistent with a number of previous reports, based on assessments conducted several years to several decades after treatment.32–34 For example, Speechley and colleagues35 studied HRQOL in 800 survivors younger than 16 years compared with 923 age- and gender-matched controls and reported that psychosocial HRQOL was most compromised among survivors of CNS tumors and for patients treated with cranial radiation either alone or in combination with chemotherapy or chemotherapy and surgery. Decreased psychosocial HRQOL among survivors whose cancer or treatment involved the CNS has been attributed to their greater burden of neurocognitive deficits and social difficulties, including the impact on how they process thoughts and feelings, and their behavior.36
Whereas there was a trend in univariate analyses toward poorer HRQOL for those who received more intensive therapy, in the multivariable models, we found no significant association between investigator-rated treatment intensity and patient-reported psychosocial HRQOL. Although this finding is in contrast to a recent report of Hispanic adolescent/young adult cancer survivors in which those who had more intensive treatments had poorer psychosocial outcomes,37 it is not surprising given that the treatment intensity ratings were objective measures, whereas the HRQOL ratings were subjective in nature and thus likely to be influenced by multiple factors including emotions (such as anxiety), beliefs and perceptions about cancer, and treatment, as well as the broader ecological context of an individual’s life (such as family, social networks, and acculturation). An unexpected finding was a negative association between treatment duration of 13 to 24 months and psychosocial functioning summary scores. A possible explanation is that this finding was confounded by diagnosis. There were 16 subjects in this group, 11 of whom had CNS tumors or soft tissue tumors. Treatment regimens for these tumors tend to be intensive and have considerable disease- and treatment-related morbidity that could diminish HRQOL. In contrast, 14 of the 19 subjects with treatment durations of 25 to 36 months had leukemia, for which treatment regimens have specific phases that vary in intensity over time and may have had fewer complications that negatively affected HRQOL. This duration of treatment-psychosocial functioning association became nonsignificant in multivariable analyses.
In univariate analyses, strong negative correlations were found between psychosocial functioning and symptoms of pain, fatigue, posttraumatic stress, and depression, with all but depression remaining statistically significant in multivariable analyses. The inverse relationships between these symptoms and psychosocial functioning summary scores support clinical observations that individuals treated for cancer experience multiple concurrent symptoms and that these symptoms can have a negative effect on HRQOL.38 Symptom research in pediatric oncology patients is an area of emerging exploration, which has been focused primarily on the period of active treatment and on individual symptoms rather than on symptom clusters, which are characterized as being related, predictable, and synergistic. Docherty39 and Baggott et al38 reviewed the published symptom literature in pediatric oncology and found very few reports of symptom clusters, while identifying numerous methodological issues that impede cross-study comparisons of what symptoms constitute clusters, how symptoms may change over time, or what patient or treatment factors affect symptom occurrence or severity. To our knowledge, there have been no reports specifically addressing multiple symptoms experienced by adolescents in the early posttreatment phase of survivorship. Further research is needed to improve our understanding of the incidence and prevalence of early posttreatment symptom clusters in childhood cancer survivors, to elucidate the relationship between symptom clusters and biobehavioral factors, and to develop theoretically grounded, evidence-based pragmatic symptom interventions that may influence outcomes such as HRQOL.
Coping strategies used by adolescents to manage or modify stressful life events are often dichotomized as cognitive or problem-focused efforts versus behavioral or emotion-focused efforts, with problem-focused coping more likely to be associated with positive outcomes (eg, resiliency) and emotion-focused coping more likely to be associated with negative outcomes (eg, risky health behaviors).40 In this study’s univariate analyses, we found that the coping behaviors classified in the A-COPE subscales as avoiding problems and ventilating feelings were negatively associated with psychosocial functioning. These coping strategies would be classified as emotion focused and are congruent with what is termed defensive coping in the ARM, where they are conceptualized as risk factors negatively influencing resilience and QOL.41 Another coping behavior, seeking professional support, also was negatively associated with psychosocial functioning summary scores. Although causal direction cannot be determined from the association, it is plausible that those individuals with impaired psychosocial HRQOL would be more likely to seek professional support.
One coping behavior, being humorous, was positively associated with psychosocial functioning summary scores in univariate analyses, and it was the only coping variable that remained significant in the multivariable analyses. There is theoretical and anecdotal support, but little empirical research evidence, for adolescents’ use of humor as a coping strategy.42 The leaders of an adolescent cancer support group, Teen Impact, at CHLA, have often observed participants effectively using humor as a coping strategy. As one example, during a group outing to Disneyland, a teenage boy whose right arm was amputated at the shoulder got off the Space Mountain ride in front of a crowd, pointed at his amputation, and yelled, “Man, this ride is dangerous!” Using humor aligns with the ARM conceptualization of developmentally appropriate positive or “courageous coping” as a protective factor. In a systematic review of the literature on coping during adolescence, Garcia40 has advocated for further research to advance the conceptualization and measurement of adolescent coping, noting that coping is a complex construct worthy of examination because it can be a critical point of intervention in adolescents’ healthy development.
This study found a positive association between older age at interview and psychosocial functioning. Among healthy adolescent samples, age has been shown to mediate the relationship between psychosocial development and coping, but findings regarding age in adolescents with cancer are inconsistent.2 Evan and Zeltzer36 have noted that within the pediatric-adolescent age spectrum, older age can be either a risk or a protective factor. Previous studies have reported that older age is associated with greater perceived control in coping with cancer and a larger repertoire of coping strategies; in contrast, older adolescent age has been associated with poorer HRQOL and higher intensity of symptoms during treatment, possibly because of greater cognitive ability to comprehend life threat and other health status information, resulting in a net increase in external stressors.36 Perhaps, older age and better psychosocial functioning in our study are partly explained by recent neuroimaging research, which has shown that brain development continues after puberty and that among the last regions of the brain to achieve full functional maturation is the prefrontal cortex, an area crucial to self-control and affect regulation.43 According to Thompson,44 affect regulation refers to the extrinsic and intrinsic processes responsible for monitoring, evaluating, and modifying emotional reactions, especially their intensive and temporal features, to accomplish one’s goals. Further study is needed to determine whether older adolescents’ more advanced neurobiological development affects their understanding and perception of their cancer experience in ways that increase the likelihood of higher self-ratings of psychosocial HRQOL soon after treatment completion.
Limitations of this study include its cross-sectional design, which constrains any inferences about causal relationships. A volunteer bias is recognized as a threat to the generalizability of our findings because participants were not randomly selected and there may have been undetected differences between participants and nonparticipants. In addition, cognitively impaired survivors unable to complete study questionnaires were excluded. Such differences could have introduced bias if subjects who were either more likely or less likely to be doing well were more apt to participate. Overall, study subjects self-reported good psychosocial HRQOL. Several alternative possible explanations for this positive PRO have been suggested, including benefit finding, adaptive style, response shift, resilience, and social desirability bias.29,30
The strengths of this study include the use of a well-validated PRO measure of psychosocial HRQOL, the PedsQL 4.0 Generic Core Scales–Adolescent and Young Adult Self-report, which is composed of a series of age-specific questionnaires that now have been used with a variety of populations, including healthy children/adolescents. Patient self-report was used because discrepancies between patient and parent-proxy HRQOL ratings are known to occur. Although these differences may be minor, self-reports and parent-proxy reports are not interchangeable and investigators are encouraged to focus on the adolescent’s own perceptions of HRQOL.45,46 The study addresses several knowledge gaps relevant to childhood cancer survivorship. It is the first to focus on adolescents’ psychosocial HRQOL soon after treatment completion. The finding that a subset of adolescents may have increased risk for impaired HRQOL suggests that there is a need for assessment of psychosocial HRQOL during the early posttreatment phase. This phase might offer a window of opportunity for targeted interventions to mitigate the adverse psychosocial effects of the cancer experience and improve long-range HRQOL. In addition, there may be adolescents within the upper end of the range who report doing well but whose well-being should be validated to be sure they are not overestimating their functioning and underestimating their need for support and/or information,30 as well as to better understand the factors that promote positive adaptation. This is the first study to describe multiple symptoms in adolescent cancer patients in the early posttreatment phase of survivorship and their association with self-reported psychosocial HRQOL. Further study is needed to examine the incidence and persistence of posttreatment symptom clusters in this population, including unique patterns of symptoms that cluster, as well as risk factors for symptom clustering. As the initial assessment in a longitudinal study, the study provides a baseline for future comparisons that may help us better understand the dynamic nature of the psychosocial impact of cancer during the posttreatment survivorship trajectory. Ultimately, what is learned from research focused on psychosocial outcomes for adolescent survivors of childhood cancer will help drive evidence-based interventions that support successful transitions into adulthood and long-term survivorship for the large and growing population of adolescent cancer patients.
Although most patients rated their psychosocial HRQOL as good, a subset (18%) may have increased risk for impaired HRQOL. Protective factors identified in multivariable analyses included older age and the use of humor as a coping strategy. Risk factors included symptoms of pain, fatigue, and posttraumatic stress. Clinical assessment of psychosocial HRQOL using validated PRO measures during the early posttreatment phase is recommended for early identification of young people at risk for diminished psychosocial HRQOL. Longitudinal studies are needed to further explore risk and protective factors, as well as to identify and evaluate targeted interventions aimed at minimizing adverse effects of cancer treatment and maximizing healthy survivorship.
The authors gratefully acknowledge Susan Shannon, MSN, CPNP, and Marcia Leonard, RN, PNP, for coordinating the implementation of this study at their treatment centers and for their longstanding dedication to survivorship care and thank Tavo Zavala, BS, for sharing the anecdote from Teen Impact.
1. Keene N, Hobbie W, Ruccione K. Childhood Cancer Survivors: A Practical Guide to Your Future. 3rd ed. Bellingham, WA: Childhood Cancer Guides; 2012.
2. Cantrell MA. Health-related quality of life in childhood cancer: state of the science. Oncol Nurs Forum. 2007; 34 (1): 103–111.
3. Zebrack BJ, Zeltzer LK. Quality of life issues and cancer survivorship
. Curr Probl Cancer. 2003; 27 (4): 198–211.
4. Varni JW, Limbers C, Burwinkle TM. Impaired health-related quality of life in children and adolescents with chronic conditions: a comparative analysis of 10 disease clusters and 33 disease categories/severities utilizing the PedsQL 4.0 Generic Core Scales. Health Qual Life Outcomes. 2007; 5: 43.
5. McDougall J, Tsonis M. Quality of life in survivors of childhood cancer: a systematic review of the literature (2001–2008). Support Care Cancer. 2009; 17 (10): 1231–1246.
6. Klassen AF, Anthony SJ, Khan A, Sung L, Klaassen R. Identifying determinants of quality of life of children with cancer and childhood cancer survivors: a systematic review. Support Care Cancer. 2011; 19 (9): 1275–1287.
7. Lipscomb J, Gotay CC, Snyder CF. Patient-reported outcomes in cancer: a review of recent research and policy initiatives. CA Cancer J Clin. 2007; 57 (5): 278–300.
8. Katz ER, Burwinkle T, Varni JW, Barr RD. Health-related quality of life. In: Bleyer WA, Barr RD, ed. Cancer in Adolescents and Young Adults. Heidelberg, Berlin: Springer-Verlag; 2007: 387–400.
9. Hinds PS, Gattuso JS, Fletcher A, et al. Quality of life as conveyed by pediatric patients with cancer. Qual Life Res. 2004; 13 (4): 761–772.
10. Clinton-McHarg T, Carey M, Sanson-Fisher R, Shakeshaft A, Rainbird K. Measuring the psychosocial health of adolescent and young adult (AYA) cancer survivors: a critical review. Health Qual Life Outcomes. 2010; 8: 25.
11. Barakat LP, Marmer PL, Schwartz LA. Quality of life of adolescents with cancer: family risks and resources. Health Qual Life Outcomes. 2010; 8: 63.
12. Landolt MA, Vollrath M, Niggli FK, Gnehm HE, Sennhauser FH. Health-related quality of life in children with newly diagnosed cancer: a one year follow-up study. Health Qual Life Outcomes. 2006; 4: 63.
13. Sung L, Klaassen RJ, Dix D, et al. Identification of paediatric cancer patients with poor quality of life. Br J Cancer. 2009; 100 (1): 82–88.
14. Wakefield CE, McLoone J, Goodenough B, Lenthen K, Cairns DR, Cohn RJ. The psychosocial impact of completing childhood cancer treatment: a systematic review of the literature. J Pediatr Psychol. 2010; 35 (3): 262–274.
15. Haase JE. The adolescent resilience model as a guide to interventions. J Pediatr Oncol Nurs. 2004; 21 (5): 289–299; discussion 300–284.
16. Kazak AE, Kassam-Adams N, Schneider S, Zelikovsky N, Alderfer MA, Rourke M. An integrative model of pediatric medical traumatic stress. J Pediatr Psychol. 2006; 31 (4): 343–355.
17. Varni JW, Burwinkle TM, Katz ER, Meeske K, Dickinson P. The PedsQL in pediatric cancer: reliability and validity of the Pediatric Quality of Life Inventory Generic Core Scales, Multidimensional Fatigue Scale, and Cancer Module. Cancer. 2002; 94 (7): 2090–2106.
18. Radloff LS. The CES-D Scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977; 1: 385–401.
19. Weissman MM, Sholomskas D, Pottenger M, Prusoff BA, Locke BZ. Assessing depressive symptoms in five psychiatric populations: a validation study. Am J Epidemiol. 1977; 106 (3): 203–214.
20. Stinson JN, Kavanagh T, Yamada J, Gill N, Stevens B. Systematic review of the psychometric properties, interpretability and feasibility of self-report pain intensity measures for use in clinical trials in children and adolescents. Pain. 2006; 125 (1–2): 143–157.
21. Cohen LL, Lemanek K, Blount RL, et al. Evidence-based assessment of pediatric pain. J Pediatr Psychol. 2008; 33 (9): 939–955; discussion 956–937.
22. Steinberg AM, Brymer MJ, Decker KB, Pynoos RS. The University of California at Los Angeles Post-traumatic Stress Disorder Reaction Index. Curr Psychiatry Rep. 2004; 6 (2): 96–100.
23. Patterson JM, McCubbin HI. Adolescent coping style and behaviors: conceptualization and measurement. J Adolesc. 1987; 10 (2): 163–186.
24. Blount RL, Simons LE, Devine KA, et al. Evidence-based assessment of coping and stress in pediatric psychology. J Pediatr Psychol. 2008; 33 (9): 1021–1045.
25. Werba BE, Hobbie W, Kazak AE, Ittenbach RF, Reilly AF, Meadows AT. Classifying the intensity of pediatric cancer treatment protocols: the Intensity of Treatment Rating Scale 2.0 (ITR-2). Pediatr Blood Cancer. 2007; 48 (7): 673–677.
26. Kazak AE, Hocking MC, Ittenbach RF, et al. A revision of the Intensity of Treatment Rating Scale: classifying the intensity of pediatric cancer treatment. Pediatr Blood Cancer. 2012; 59 (1): 96–99.
27. Carozza SE, Puumala SE, Chow EJ, et al. Parental educational attainment as an indicator of socioeconomic status and risk of childhood cancers. Br J Cancer. 2010; 103 (1): 136–142.
28. Varni JW, Burwinkle TM, Seid M, Skarr D. The PedsQL 4.0 as a pediatric population health measure: feasibility, reliability, and validity. Ambul Pediatr. 2003; 3 (6): 329–341.
29. Stam H, Grootenhuis MA, Brons PP, Caron HN, Last BF. Health-related quality of life in children and emotional reactions of parents following completion of cancer treatment. Pediatr Blood Cancer. 2006; 47 (3): 312–319.
30. Engelen V, Koopman HM, Detmar SB, et al. Health-related quality of life after completion of successful treatment for childhood cancer. Pediatr Blood Cancer. 2011; 56 (4): 646–653.
31. Bhat SR, Goodwin TL, Burwinkle TM, et al. Profile of daily life in children with brain tumors: an assessment of health-related quality of life. J Clin Oncol. 2005; 23 (24): 5493–5500.
32. Zeltzer LK, Lu Q, Leisenring W, et al. Psychosocial outcomes and health-related quality of life in adult childhood cancer survivors: a report from the childhood cancer survivor study. Cancer Epidemiol Biomarkers Prev. 2008; 17 (2): 435–446.
33. Zebrack BJ, Gurney JG, Oeffinger K, et al. Psychological outcomes in long-term survivors of childhood brain cancer: a report from the childhood cancer survivor study. J Clin Oncol. 2004; 22 (6): 999–1006.
34. Schultz KA, Ness KK, Whitton J, et al. Behavioral and social outcomes in adolescent survivors of childhood cancer: a report from the childhood cancer survivor study. J Clin Oncol. 2007; 25 (24): 3649–3656.
35. Speechley KN, Barrera M, Shaw AK, Morrison HI, Maunsell E. Health-related quality of life among child and adolescent survivors of childhood cancer. J Clin Oncol. 2006; 24 (16): 2536–2543.
36. Evan EE, Zeltzer LK. Psychosocial dimensions of cancer in adolescents and young adults. Cancer. 2006; 107 (suppl 7): 1663–1671.
37. Kazak AE, Derosa BW, Schwartz LA, et al. Psychological outcomes and health beliefs in adolescent and young adult survivors of childhood cancer and controls. J Clin Oncol. 2010; 28 (12): 2002–2007.
38. Baggott C, Dodd M, Kennedy C, Marina N, Miaskowski C. Multiple symptoms in pediatric oncology
patients: a systematic review. J Pediatr Oncol Nurs. 2009; 26 (6): 325–339.
39. Docherty SL. Symptom experiences of children and adolescents with cancer. Annu Rev Nurs Res. 2003; 21: 123–149.
40. Garcia C. Conceptualization and measurement of coping during adolescence
: a review of the literature. J Nurs Scholarsh. 2010; 42 (2): 166–185.
41. Haase JE. Resilience. In: Peterson SJ, Bredow TS, ed. Middle Range Theories: Application to Nursing Research. Philadelphia, PA: Lippincott Williams & Wilkins; 2009: 326–362.
42. Dowling JS. Humor: a coping strategy for pediatric patients. Pediatr Nurs. 2002; 28 (2): 123–131.
43. Giedd JN, Blumenthal J, Jeffries NO, et al. Brain development during childhood and adolescence
: a longitudinal MRI study. Nat Neurosci. 1999; 2 (10): 861–863.
44. Thompson RA. Emotion regulation: a theme in search of definition. Monogr Soc Res Child Dev. 1994; 59 (2–3): 25–52.
45. Eiser C, Morse R. Can parents rate their child’s health-related quality of life? Results of a systematic review. Qual Life Res. 2001; 10 (4): 347–357.
46. Sattoe JN, van Staa A, Moll HA. The proxy problem anatomized: child-parent disagreement in health related quality of life reports of chronically ill adolescents. Health Qual Life Outcomes. 2012; 10: 10.