The World Health Organization defines adolescents as "young people between the ages of 10 and 19 years."1 Many of those within this age group who are diagnosed with cancer will benefit from the development of aggressive treatment protocols and therefore experience significantly reduced risk of mortality. Based on data collected between 1975 and 2010 through the Surveillance, Epidemiology and End Results Program of the National Cancer Institute, the 5-year relative survival rate among those aged 10 to 14 years, regardless of cancer type, increased from 58.6% to 83.9%.2 Among those aged 15 to 19 years, the 5-year relative survival rate increased from 67.9% to 84.5%.2 In light of the improved likelihood of survival, attention has increasingly been paid to health-related quality of life, a multicomponent concept encompassing functioning in the physical, emotional, social, and cognitive domains.3 Given the changes that occur in all of these domains during the transition from childhood to young adulthood, the experience of cancer at this time of life is likely to differ with regard to symptom manifestation and relative distress.4
From a developmental perspective, adolescence is a time of establishing an identity outside the immediate family, increased identification with peers, and improved capacity to consider abstract concepts.5 A diagnosis of cancer reinforces an adolescent's dependence on caregivers, with the physical consequences of the disease and treatment alone serving as reminders of profound helplessness. Interrupted and unpredictable school attendance, coupled with outcomes such as hair loss and drastic weight change, may leave adolescents reluctant to engage in age-expected socialization. Adolescents may begin to understand the severity of their disease, as well as the implications of potential treatment-related side effects, but may be unable or unwilling to make necessary lifestyle changes or assume increased responsibility for their health.5 Given the unique challenges cancer presents at this particular stage of life, special care should be taken in assessing how symptoms impact adolescents' quality of life.
One of the most frequently reported and feared symptoms of cancer is pain.6 Pain may be related to the disease itself, treatments (surgery, chemotherapy, radiotherapy), procedures (placement of central line, lumbar punctures), or some combination thereof. Interest in children's experience of pain has seen remarkable growth in the past 30 years, following publications highlighting the discrepancies between analgesic and opioid administration among children and adults who had undergone identical surgeries or had the same diagnosis.7,8 Pain can be managed through a variety of interventions, including psychological (eg, cognitive-behavioral therapy, acceptance and commitment therapy, etc), physical (eg, massage, application of heat or cold, positioning, etc), and pharmacological (eg, opioids, acetaminophen, nonsteroidal anti-inflammatory drugs, etc). Mechanisms that contribute to pain (surgery, chemotherapy, radiation) may also provide relief, although it is presently difficult to determine beforehand whether such disease-directed treatment will increase or decrease pain. When used alone or, more likely, in combination, the aforementioned techniques may effectively manage pain; however, it is estimated that less than 50% of all cancer patients have good pain control.6 The most effective combination of interventions is currently determined on a case-by-case basis.
While it would seem obvious that the more pain is reduced, the better the quality of life, the exact relationship between pain and quality of life is not understood at present. Some adolescents may report the expected inverse relationships (eg, when pain is low, quality of life is highly affected). But assessing the presence or absence of pain or dichotomizing pain and quality of life into "low" or "high" categories may not be enough; considering aspects such as frequency, severity, intensity, and duration yields more precise information. For instance, quality of life may be significantly associated with 1 aspect (eg, severity) and not another (eg, duration).9 Others may report significant pain but also maintain what they perceive to be an acceptable or even good quality of life.10 And some may indicate that even a small amount of pain negatively impacts their quality of life to an unacceptable degree.10 The subjective nature of pain makes it difficult to develop standardized, generalizable descriptions, therefore reducing the opportunity to quantify and correlate reductions in pain with improvements in quality of life.
In order to better understand the relationship between pain-related symptoms and their impact on quality of life, further examination of the types and trajectories of cancer-related pain is warranted. The purpose of this systematic review is to examine self-reported pain among adolescents diagnosed with and treated for leukemia or a brain tumor to better understand the types of pain reported, as well as the methods used to assess pain.
Methods
Electronic searches of MEDLINE, PubMed, and the Cochrane Database of Systematic Reviews since inception to May 2014 were conducted to identify all relevant studies (Figure). Leukemia and brain tumors were chosen because they are the 2 most frequently occurring cancers in those younger than 20 years and have been the focus of multiple studies.2 However, studies that included only patients with a brain tumor, only patients with leukemia, or only patients with either a brain tumor or leukemia were rare, as researchers often included patients with other cancer diagnosis to improve the overall sample size and allow for statistical analyses. When studies included participants with diagnoses other than a brain tumor or leukemia, pain-related data were often presented in the context of the diagnosis, in which case only those data related to patients with diagnoses of interest were included in this review. Given that publications do not always provide information on specific types of leukemia or brain tumors, these were examined as 2 broad categories and not differentiated by type. Pain was assessed both in patients who were in active treatment and those who had transitioned to long-term survivorship.
In accordance with the World Health Organization definition, adolescents were defined as being 10 to 19 years of age at the time of self-report.1 The provided age range of studies was considered, and those with participants aged 10 to 19 years were selected. To account for the fact that there is currently no agreed-upon definition of exactly what time frame constitutes the period of adolescence, those studies with a reported mean/median patient age within the target age range were also selected. In addition, studies that reported an age range that included, but was not exclusive to, patients aged 10 to 19 years were selected. The benefit of including studies where patients aged 10 to 19 years comprised at least some part of the sample was determined to outweigh the inherent risk of reviewing data from patients younger or older than those in the target age range (see Limitations for further discussion of the methods).
Results
A total of 30 articles met inclusion criteria and were included in the review (Table). Of these, 11 of the 30 articles (37%) reported pain-related information in adolescents with leukemia, and 5 of the 30 (16%) reported pain-related information in adolescents with brain tumors. Fourteen studies (47%) included both patients with leukemia and those with brain tumor. The articles were published between 1987 and 2014, with the majority (n = 25) being published in or after 2000. Seven studies reported age ranges exclusively within 10 to 19 years, whereas 11 reported a mean/median patient age within the target age range. Twelve studies reported an age range that included, but was not exclusive to, patients aged 10 to 19 years. Reported pain was classified into 1 of 5 categories: pain reported before diagnosis or specifically related to disease, pain related to treatment, pain related to procedures, pain reported after treatment completion, and nonspecific pain/pain related to many, nondelineated variables.
Pain Reported Prior to Diagnosis or Related to Disease
Pain that was reported to have occurred before a diagnosis had been obtained or was described as specifically being a result of the cancer itself was reported in a total of 10 articles.11-20 The majority (n = 5) included samples with various diagnoses.11,12,14,15,17 Pain was assessed most frequently through interview (n = 5), followed by retrospective chart review (n = 3), and 11-point numerical rating scale (NRS). Of a sample consisting exclusively of patients with leukemia, nearly all reported pain at the time of diagnosis.13 Pain was reported to be present a median of 33 days prior to diagnosis by patients with leukemia.16 In 3 studies with both patients with brain tumor and patients with leukemia, 49% to 63% of patients reported pain at time of diagnosis.12,14,15 Similarly, authors of another study including patients with mixed diagnoses noted that it was the patients with newly diagnosed disease who most frequently reported pain.18 Headache was frequently reported as the primary or secondary symptom among patients with a brain tumor, and patients experienced median prediagnostic symptomatic intervals of 20.5 days (no range given), 60 days (range, 0-3010 days), and 75 days (range, 5-730 days).18-20 In 2 studies, brain tumor location was not significantly correlated with time from initial symptom presentation until diagnosis; however, 1 study reported a drastically reduced median prediagnostic symptomatic interval for infratentorial tumors (36 days) than for supratentorial tumors (113 days).18-20
Results were conflicting as to whether pain decreased upon initiation of treatment. Whereas reported pain in 1 sample of adolescents with leukemia did not significantly decrease upon beginning treatment, in another sample of patients with the same diagnosis, pain resolved a median of 3 days following initiation of treatment.13,16 In 2 samples of mixed brain tumor and leukemia patients, pain related to the disease was more frequently endorsed by those the farthest out from diagnosis, suggesting that it was not relieved by treatment.14,15 Pain caused by the disease itself, whether leukemia or a brain tumor, was reported as being less severe than pain caused by either treatments or procedures at a single assessment and at 3 of 4 time points.11,17 Pain intensity ratings provided by a mixed sample of patients with brain tumor and leukemia did not differ across 4 time points, including at the time of the interview, in the past 24 hours, in the last 7 days, and in the last 4 weeks.11 One study found sex differences between girls and boys in ratings of past, but not current, pain intensity, with girls reporting more intense past pain.11
Pain Related to Treatment
Pain related to treatment was reported in 9 articles.11,12,14,15,17,21-24 The majority of the studies (n = 8) included samples with various diagnoses.11,12,14,15,17,21,23 Pain was assessed most frequently through interview (n = 4), followed by 11-point NRS (n = 2). Treatment included chemotherapy, radiation, and bone marrow transplant.11,17,22-24 Treatment was reported to cause more pain than either the disease itself or procedures in samples with mixed diagnoses.14,15,17,21 In a study of patients exclusively with leukemia, 56% reported pain in the evening related to chemotherapy medication.22 In 1 study of patients with either leukemia or a brain tumor, pain was reported more frequently at the beginning and end of a treatment period, but less so in the middle.12 For other patients in a sample of mixed diagnoses, treatment-related pain actually increased in reported severity when assessed at a specific time point, in the last 24 hours, in the last 7 days, and in the last 4 weeks.17 Treatment-related pain was more frequently reported by those farther out from time of diagnosis, regardless of specific diagnosis.15,17 Treatment-related pain was frequently reported in the abdomen and legs and was often experienced in conjunction with mucositis.14,17,21,22
Pain Related to Procedures
Pain related to procedures was reported in 7 articles, with all articles including samples with various diagnoses.11,12,14,15,17,23,25 Pain was assessed most frequently through interview (n = 5), followed by 11-point NRS (n = 3). Procedures associated with pain included lumbar punctures, needle injection, central line or port catheter access, or insertion of nasogastric tube.11,14,15,17,23,25 Procedure-related pain was reported to be less severe than treatment-related pain when assessed at a specific point in time, within the past 24 hours, within the past 7 days, and within the past 4 weeks in a sample of mixed diagnoses.11 In a similar sample, patients reported that procedure-related pain was less severe than treatment-related pain but more severe than disease-related pain over the past month.17 Procedural pain decreased in a sample of patients with either a brain tumor or leukemia, and in another study of patients with similar diagnoses, it was reported to be worse among those who were less than 5 months from time of diagnosis.14,15,25
Pain Reported After Treatment Completion
Pain did not immediately cease upon treatment completion and was reported in 5 articles.26-30 The majority of studies (n = 3) included patients with various diagnoses.26-28 Pain was most frequently assessed through interview (n = 2) and the Health Utilities Index (n = 2). Participants were off-treatment anywhere from 1 month to more than 8 years. Pain was reported to more frequently cause mild disability in leukemia or brain tumor survivors than either moderate or severe disability.26 Chronic pain was cited as significantly limiting theactivity levels of those in a sample that included both brain tumor and leukemia survivors a median of 5 years after diagnosis.27 Among survivors within 6 months of treatment completion for leukemia or a brain tumor, pain was significantly correlated with fatigue (P < .0001), depression (P < .001), and posttraumatic stress (P < .001). In a sample of patients treated exclusively for a brain tumor a mean of almost 3 years prior, nearly one-third of participants reported ongoing pain.29 Other brain tumor survivors a similar length of time from treatment experienced pain in the form of new-onset, severe, unilateral headaches associated with nausea, hemiparesis, aphasia, or visual loss.30
Nonspecific Pain/Pain Related to Many, Nondelineated Variables
The majority of articles (n = 11) simply assessed the self-reported presence or absence of pain at 1 or more points in time, without further examining whether pain was specifically related to the cancer itself, treatment, and/or procedures.13,31-40 All 11 articles included patients with various diagnoses. Pain was assessed most frequently through the Memorial Symptom Assessment Scale (n = 5), followed by interview (n = 2). In patients with leukemia, pain did not increase or decrease over the course of chemotherapy.31 In a sample including brain tumor and leukemia diagnoses, pain was one of the most frequent, intense, and distressing symptoms reported by adolescents a week after chemotherapy administration.33 In a similar study with mixed diagnoses, pain was constant throughout the chemotherapy cycle, with the majority of adolescents reporting it in the moderate to severe/very severe range.34
In a sample of hospitalized patients with leukemia, pain was initially reported in 45.3% of study participants; however, over 5 days, the number reporting pain decreased significantly.32 The study authors noted that on days when patients reported pain, they also reported increased physical and psychological symptoms, as well as a higher global distress index.32 In patients with leukemia, reported pain was significantly associated with reduced hip flexion range of motion. In a qualitative study that included patients with leukemia, the authors noted how frequently cancer was equated with suffering and short-term pain for long-term gain.36 Pain was the second most frequently reported symptom after lack of energy by adolescents with leukemia or a brain tumor, with more than 80% of the sample reporting moderate to very severe intensity.37 When patients were separated into diagnostic categories, pain was endorsed by 50% of those with a brain tumor compared with 43.8% of those with leukemia.37 In a study that included adolescents with a diagnosis of leukemia, pain management was identified as an unmet service need.40 The specific need for pain-management support was strongly associated with lower overall health-related quality of life, as well as decrements in the domains of physical, emotional, social, and work/school functioning.40
Pain Management Strategies
While the purpose of this review was not to assess interventions designed to reduce pain in adolescent cancer patients, the paucity of research in this area could not be helped but be noted. There is much room for growth in the areas of applying known evidence-based treatments to decrease pain, evaluating the effectiveness of novel therapies in alleviating pain, or comparing the efficacy of pharmacological versus nonpharmacological treatment. In the studies included in this review, adolescents self-reported a handful of pain-management strategies. Beginning treatment was reported to resolve cancer-related pain in a median of 3 days in a study of patients with leukemia.16 Pharmacological pain-management strategies included the use of opioids and analgesics.11,13-17,21,25,36 Nonpharmacological pain-management strategies included rest/sleep, deep breathing, hot pack/massage, distraction, empathy, and the using humor, seeking professional support, and avoiding problems.13,21,28,36 Only pharmacological strategies were reported for reducing procedural pain, and these included EMLA (eutectic mixture of local anesthetics) for venipuncture/cannulation of a vein and before inserting needle into subcutaneous intravenous port, general anesthesia for lumbar puncture, general anesthesia for bone marrow aspirate, and sedation.14,15,17
Discussion
This systematic review sought to examine self-reported pain among adolescents with a diagnosis of a brain tumor or leukemia, although the lack of agreement on a specific age range included in the definition of "adolescent" impacted the researchers' ability to conduct such a review. In the majority of studies (n = 11), pain was nonspecific or related to many, nondelineated variables. The presence, and occasionally the severity or intensity, of pain was reported without further examination of contributing factors. Pain was next most frequently reported (n = 10) as occurring prior to diagnosis or specifically related to the disease itself. Studies also included self-reported treatment- and procedure-related pain (n = 9 and n = 7, respectively), as well as pain that occurred after treatment was completed (n = 5). With regard to patient care, the results of this review point to the need for regular pain assessment, beginning at the time of diagnosis and continuing throughout treatment and into survivorship. When working in a chronic and recurrent pain model, PedIMMPACT (Pediatric Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials) recommends assessing outcomes in the domains of pain intensity, physical functioning, emotional functioning, role functioning, symptoms and adverse events, global judgment of satisfaction with treatment, sleep, and economic factors.41 It is expected that the influence of these domains would differ over time, thus necessitating changes in clinical management.
Very few of the studies reviewed were limited to a single diagnosis, and, particularly in regard to the field of pain in adolescent oncology, providing the specific cancer diagnoses included in the study samples would allow for greater comparison of within and between diagnosis-group experiences of pain. Being able to parse out what is similar and what is different about pain across diagnoses could foster the development of general pain-management interventions that require only minor alterations to be most effective for patients with specific diagnoses. Some disease-specific aspects of pain were notable. For patients with leukemia, pain was present and had been present for an extended period prior to diagnosis. Two studies reported conflicting results on whether pain resolved with initiation of treatment. While treatment-related pain was specifically associated with chemotherapy, the trajectory of pain over a course of chemotherapy was unpredictable. Pain was noted to decrease over time in hospitalized patients. Overall, pain was frequently reported, and among survivors, pain management was identified as an unmet service need.
For adolescents with a brain tumor, pain was present 20 to 75 days prior to diagnosis. Tumor location was not always significantly correlated with time from initial symptom presentation to diagnosis, but 1 study found considerably reduced prediagnostic symptom interval for infratentorial tumors. Among survivors of a brain tumor, pain was reported years after treatment completion, sometimes in the form of new-onset headaches associated with nausea, hemiparesis, aphasia, or visual loss.
In the majority of studies, pain was assessed via interview (n = 9), followed by the Memorial Symptom Assessment Scale (n = 6), retrospective chart review (n = 4), 11-point NRS (n = 4), Adolescent Pediatric Pain Tool (n = 3), visual analog scale (n = 3), Health Utilities Index (n = 2), and other measures of single usage. More consistent inclusion of established measures across studies could increase the availability of data suitable for meta-analyses while also providing information on which aspects of the pain experience are understudied, therefore providing the rationale for developing new measures or altering existing ones. The results of this review were notable in that pain was more frequently assessed via open-ended questions and interviews than with a visual analog scale, which is designated as a well-established pain assessment measure for adolescents based on excellent evidence from psychometric testing.41 The Patient Reported Outcomes Measurement Information System (PROMIS) measures, for example, can be useful in obtaining self-reports of how pain interferes with functioning, as well as pain intensity and quality.45
Given the prevalence of pain in the population of interest, self-report measures appear to be an effective way of beginning a conversation about pain and normalizing ongoing assessment. As treatment progressed, adolescents reported a decrease in the belief that their pain could be managed more effectively.14,38 In 2 studies of samples with mixed cancer diagnoses, the majority of adolescents (and their families) expected that complete symptom relief simply was not achieveable.17,38 According to interviews with patients and family members, initially, any and all efforts were made to combat pain; as time went on, however, a sense of resignation came to pervade.38 Lack of communication with healthcare providers was typical, as were efforts to minimize the severity of symptoms.31,38 In a study that included patients with leukemia or a brain tumor, adolescents did not ask for help managing pain because they assumed hospital staff were too busy.15 Systematic assessment of pain using self-report measures might reduce the burden on adolescents to volunteer information on pain when they have little hope of symptom alleviation. Many evidence-based treatments for pain exist, and studies applying these treatments specifically to adolescents with cancer-related pain are sorely needed. While a review of such treatments is beyond the scope of this review, the interested reader is directed to Hickman et al,42, Thane,43 and Eccelston et al.44
Limitations
While this attempt to systematically review self-reported pain in adolescents with leukemia or a brain tumor appears to be the first of its kind, it is not without shortcomings, and limitations should be noted. While diagnoses of leukemia and brain tumor were selected because they are the most common cancers in those younger than 20 years, this selection may have limited the literature reviewed by excluding articles focused more generally on "cancer" in adolescents or that focused on a specific type of cancer and associated pain (ie, sarcomas). The variability of the definition of "adolescent" across the literature made it difficult to determine what range of ages should be addressed in this review. While the WHO definition was ultimately used, the authors recognize that this may be viewed alternatively as overinclusive or underinclusive.
When studies included participants with diagnoses other than a brain tumor or leukemia and presented pain-related data in the context of the diagnosis, great care was taken so that only those data related to patients with diagnoses of interest were reported. Data were not always presented by diagnosis, however, so it is possible that some of the reported types of pain, as well as pain-management strategies, may have arisen from adolescents with diagnoses other than leukemia or a brain tumor. While it is standard practice to review only published studies, this does have the consequence of perpetuating the publication bias. Excluded were the pain experiences of those adolescents with leukemia or a brain tumor who did not participate in research studies or who participated in studies that have not yet concluded, been written up, or published. Such experiences may have served to confirm or deny many of the findings presented here. While the authors sought to draw attention to the different types of self-report measures used in reviewed studies, an attempt to account for the variability of such measures was not made, thereby making it difficult to directly compare adolescents' pain in a standardized manner across measures.
Conclusion
Pain is a frequently reported symptom among adolescents with leukemia or a brain tumor. It may be reported prior to diagnosis and associated with the disease itself, occur in the context of treatment, or be related to 1 or more specific procedures. Pain trajectories may or may not vary over time, and given the effectiveness of treatments, distress surrounding a particular symptom may be overlooked in favor of achieving a cure or simply accepted as part of the disease and treatment course. Ultimately, it is hoped that the rigorous assessment of pain after diagnosis and before and after treatment and procedures will lead to improved pain-management plans incorporating behavioral and/or pharmacological strategies, thereby resulting in improved quality of life for adolescents with cancer.
References