Abstract
Objective:
Our knowledge of symptom burden and functioning among adolescent and young adult (AYA; diagnosed ages 15-39) cancer survivors has been hindered by variability in health-related quality of life (HRQOL) measurement associated with developmental and disease heterogeneity among AYAs. We aimed to examine the variability in domain-specific aspects of HRQOL as a function of cancer type and developmental stage to clarify commonalities and differences using the NIH Patient-Reported Outcome Measurement Information System® (PROMIS®).
Methods:
572 AYAs were recruited by an online research panel using stratified sampling (treatment status: on vs. off; developmental stage: adolescents, emerging adults, young adults). Participants completed questionnaires that included socio-demographic characteristics, clinical history, and the adult version of the PROMIS-29. Generalized linear models were run for each HRQOL domain and included treatment status, developmental stage, and cancer type (hematologic vs. solid tumor) and their interactions as independent variables.
Results:
There were no significant differences in any HRQOL domain by cancer type, and few significant differences were observed in PROMIS domains between developmental groups among on-treatment AYA survivors. In contrast, off-treatment emerging adults and young adults reported significantly higher symptoms and worse functioning compared to adolescents (all ps ≤ .003).
Conclusions:
AYAs diagnosed in different developmental stages, particularly among off-treatment survivors, experienced diverse constellations of symptoms and functioning, and developmental stage was a more critical predictor of HRQOL than cancer type. These results suggest that supportive care interventions developed for AYA cancer survivors must be tailored and flexible by developmental stage and treatment status.
Keywords: Adolescent and young adult, cancer, functioning, health-related quality of life, survivorship, symptoms, oncology, psycho-oncology
Background
Six times as many individuals are diagnosed with cancer in adolescence and young adulthood (diagnosed ages 15-39) compared to childhood (diagnosed < 15 years).1 Adolescent and young adult (AYA) cancer survivors, inclusive of all individuals diagnosed with cancer regardless of treatment status, continue to experience significant impairment in health-related quality of life (HRQOL).2,3 AYA cancer survivors report significantly lower HRQOL compared to their healthy peers in some studies,4-6 but not all.7 Discussions of HRQOL across the AYA survivor population have primarily focused on cancer type,8,9 but there may be meaningful differences in HRQOL among AYA survivors diagnosed at different life stages.4,10,11
Adolescence and young adulthood are characterized by the development of autonomy, emotion regulation, and executive functioning.12-15 For example, risk-taking increases in early adolescence and then decreases in emerging adulthood (18-25 years) while emotion regulation improves across late adolescence and young adulthood.15,16 These changes occur in parallel to shifting parent-child relationships, beginning career paths, and childrearing.13,17 Developmental changes may differentially impact AYA survivors’ HRQOL, highlighting the need to examine HRQOL within the AYA developmental period.
The cancer types most commonly diagnosed among AYA patients also vary. Hematologic cancers are most common in adolescents (15-17 years) and emerging adults, and solid tumors are most common in young adults (26-39 years).18 Disease heterogeneity leads to variation in treatment exposure, which may be associated with differences in HRQOL outcomes.19,20 Within AYA cancer survivors, it is likely that both treatment exposure and developmental changes are influencing their HRQOL but potentially influencing different domains.
The overall objective of this study is to improve our understanding of HRQOL within the diverse (by both age and cancer type) AYA cancer survivor population. Guided by lifespan developmental frameworks, we categorized AYA cancer survivors into three developmental stages: adolescence (15-17 years); emerging adulthood (18-25 years); and young adulthood (26-39 years).17,21,22 First, we describe the symptom burden and functional impact of cancer in AYA survivors across seven domains of HRQOL overall and by treatment status. We hypothesized that the psychosocial HRQOL domains scores will reach the threshold for clinical relevance. Second, we examine the variability in domain-specific aspects of HRQOL as a function of cancer type, developmental stage, and treatment status. We hypothesize that the physical aspects of HRQOL will be more strongly associated with cancer type, and the psychosocial aspects of HRQOL will be more strongly associated with developmental stage.
Methods
Data Collection and Study Participants
We utilized a 2x3 stratified sampling design (on- vs. off-treatment; adolescents, emerging adults, young adults) to conduct an online, cross-sectional survey. AYAs were eligible to be included in the study if they were: (1) diagnosed with stage 0-IV cancer between the ages of 15-39, (2) were either currently receiving treatment or less than five years post-treatment, (3) were between the ages of 15-39 at time of survey, (4) could speak and read English, (5) had internet access, and (6) could provide electronic consent to the study prior to participation. Individuals were not eligible if they (1) were diagnosed with basal cell carcinoma.
We partnered with the online research panel Op4G to recruit and consent AYA cancer survivors. Op4G partners with national health-related non-profit organizations recruit their donors, volunteers, and the communities they support. This approach aids in recruiting panel members from high-, middle-, and low-socioeconomic strata. AYA survivors selected a non-profit organization for an Op4G donation as incentive for completing the survey. Participants were provided the opportunity to opt out of the survey at any time. To ensure data quality, we eliminated surveys indicative of invalid responding (e.g. survey completion time < one-third median completion time, etc.), and excluded surveys missing >10% of the items. The study procedures were approved by the Northwestern University Institutional Review Board (Protocol #: IRB00035377). The data that support the findings of this study are available from the corresponding author upon reasonable request.
Measures
Health-related quality of life.
HRQOL was assessed using the English-language, adult version of the Patient-Reported Outcomes Measurement Information System®-29 (PROMIS®-29) health status profile v2.0.23 The PROMIS-29 was completed unassisted by self-report, and assessed seven HRQOL domains (physical function, ability to participate in social roles and activities [social participation], fatigue, pain interference, sleep disturbance, anxiety, and depression) using four items for each domain. Internal consistency in the current sample (Cronbach’s alpha (α)) was high (Physical function α = 0.93; Fatigue α = 0.94; Pain interference α = 0.94; Sleep disturbance α = 0.85; Anxiety α = 0.93; Depression α = 0.93; Social participation α = 0.94).
The raw sum score of each PROMIS measure was converted to a T-score with a mean of 50 and SD of 10 using the PROMIS-29 profile v2.0 scoring table.24 A score of 50 is considered the general U.S. population mean. For all symptom and functioning domains, summary scores were not calculated when ≥25% of items were missing. Higher scores represent greater symptom burden for pain interference, fatigue, sleep disturbance, anxiety, and depression; and for physical functioning and social participation, lower scores represent worse functioning.
Research in cancer survivors has identified ranges of PROMIS T-scores to indicate minimally important differences (MIDs) between groups for most of the HRQOL domains: fatigue (2.5-5.0 points); pain interference (4.0-6.0 points); physical function (4.0-6.0 points); anxiety (3.0-4.5 points); and depression (3.0-4.5 points).25 MID cut-offs for social participation and sleep disturbance have not been established in cancer samples. Instead, we considered a one half standard deviation (5-points) difference in scores to be clinically relevant for these two outcomes. This approach has been used in previous studies with PROMIS measures.26
Sociodemographic and clinical characteristics.
Sociodemographic and clinical characteristics were captured through self-report. The sociodemographic characteristics included: AYA developmental stage (adolescence; emerging adulthood; young adulthood), race/ethnicity (Non-Hispanic White; Other race/ethnicity), sex (male; female), health insurance coverage (yes; no), and history of other comorbid conditions (e.g., diabetes, migraines, etc.), which were grouped into four categories (no conditions; 1 condition; 2-3 conditions; 4+ conditions).
Clinical characteristics included: cancer type (hematologic cancer; solid tumor), treatment status (on-treatment; off-treatment), history of chemotherapy (yes; no), history of surgery (yes; no), and history of radiation (yes; no). In order to examine differences in HRQOL between cancer types, we created a dichotomous cancer type variable grouping hematologic cancers (Hodgkin lymphoma, leukemia, myeloma, Non-Hodgkin lymphoma) and solid tumors (bladder cancer, bone tumors + sarcomas, brain cancer, breast cancer, central nervous system tumor, cervical cancer, colorectal cancer, esophageal cancer, head and neck cancer, hepatobiliary cancer, kidney cancer, lung cancer, melanoma, ovarian cancer, stomach cancer, testicular cancer, thyroid cancer) into separate categories. This was done in order to maintain sufficient power for subgroup analysis.
Statistical Analysis
We tested our first hypothesis by calculating the means and SDs for each HRQOL domain among the full sample and by treatment status. If the pain, fatigue, anxiety, sleep disturbance and/or depression scores were above 50 and exceeded the lower bound of the MID range (or half a standard deviation), we considered this HRQOL domain to be clinically impacted by cancer. Similarly, if the physical function and/or the social participation scores were below 50 and exceeded the lower bound for a MID (or half a standard deviation), this domain was considered clinically impacted by cancer.
To test our second hypothesis, we ran a generalized linear model (GLM) for each HRQOL domain. Each GLM included developmental stage, cancer type, and treatment status as the primary independent variables. We also tested the interaction terms between all three variables (developmental stage*cancer type, developmental stage*treatment status, cancer type*treatment status). If any of the interaction terms were found to be significant at the p ≤ .003 (Bonferroni-correction; .05/16), it was retained in the final model. Interaction terms that did not meet this threshold were removed. Each GLM also controlled for sociodemographic and clinical variables. All tests were two-sided, and analyses were conducted using SAS 9.4 (Cary, NC).
Results
Sample Characteristics
A total of 572 AYAs completed the survey with balanced representation by on-treatment (N=294, 51.4%) and off-treatment (N=278, 48.6%) survivors, and across the three developmental groups (adolescents: 189, 33.0%; emerging adults: 193, 33.7%; young adults: 190, 33.2%). Descriptive statistics stratified by cancer type and developmental stage are summarized in Table 1.
Table 1.
Sociodemographic and clinical characteristics among AYA cancer survivors, stratified by age at diagnosis
| Adolescents N = 189 |
Emerging Adults N = 193 |
Young Adults N = 190 |
Hematologic Cancers N=135 |
Solid Tumors N=437 |
Overall Sample N = 572 |
|
|---|---|---|---|---|---|---|
| Developmental Stage (N, %) | ||||||
| Adolescents | 73 (54.1) | 116 (26.5) | 189 (33.0) | |||
| Emerging Adults | 42 (31.1) | 151 (34.6) | 193 (33.7) | |||
| Young Adults | 20 (14.8) | 170 (38.9) | 190 (33.2) | |||
| Cancer Diagnosis (N, %) | ||||||
| Hematologic cancers | 73 (38.6) | 42 (21.8) | 20 (10.5) | 135 (23.6) | ||
| Solid tumors | 116 (61.4) | 151 (78.2) | 170 (89.5) | 437 (76.4) | ||
| Sex (N, %) | ||||||
| Male | 108 (57.1) | 105 (54.4) | 110 (57.9) | 69 (51.1) | 254 (58.1) | 323 (56.5) |
| Female | 81 (42.9) | 88 (45.6) | 80 (42.1) | 66 (48.9) | 183 (41.9) | 249 (43.5) |
| Current Living Situation (N, %) | ||||||
| Live alone | 4 (2.1) | 45 (23.3) | 74 (39.0) | 13 (9.6) | 110 (25.2) | 123 (21.5) |
| Live with others | 185 (97.9) | 148 (76.7) | 116 (61.1) | 122 (90.4) | 327 (74.8) | 449 (78.5) |
| Health Insurance Coverage (N, %) | ||||||
| No | 22 (11.6) | 42 (21.8) | 40 (21.1) | 18 (13.3) | 86 (19.7) | 104 (18.2) |
| Yes | 167 (88.4) | 151 (78.2) | 150 (79.0) | 117 (86.7) | 351 (80.3) | 468 (81.8) |
| Race/Ethnicity (N, %) | ||||||
| Non-Hispanic White | 118 (63.4) | 130 (68.4) | 142 (75.1) | 304 (70.5) | 86 (64.2) | 390 (69.0) |
| Other race/ethnicity | 68 (36.6) | 60 (31.6) | 47 (24.9) | 127 (29.5) | 48 (35.8) | 175 (31.0) |
| Treatment Location (N, %) | ||||||
| Academic medical center | 101 (53.4) | 82 (42.5) | 72 (37.9) | 66 (48.9) | 189 (43.3) | 255 (44.6) |
| Community- or office-based practice | 88 (46.6) | 111 (57.5) | 118 (62.1) | 69 (51.1) | 248 (56.8) | 317 (55.4) |
| Comorbidity (N, %) | ||||||
| 0 comorbid conditions | 6 (3.2) | 22 (11.4) | 28 (14.7) | 8 (5.9) | 48 (11.0) | 56 (9.8) |
| 1 comorbid condition | 66 (34.9) | 100 (51.8) | 66 (34.7) | 43 (31.9) | 189 (43.3) | 232 (40.6) |
| 2-3 comorbid conditions | 63 (33.3) | 48 (24.9) | 62 (32.6) | 51 (37.8) | 122 (27.9) | 173 (30.2) |
| 4+ comorbid conditions | 54 (28.6) | 23 (11.9) | 34 (17.9) | 33 (24.4) | 78 (17.9 | 111 (19.4) |
| Chemotherapy (N, %) | ||||||
| No | 116 (61.4) | 67 (34.7) | 41 (21.6) | 70 (51.9) | 154 (35.2) | 224 (39.2) |
| Yes | 73 (38.6) | 126 (65.3) | 149 (78.4) | 65 (48.2) | 283 (64.8) | 348 (60.8) |
| Radiation (N, %) | ||||||
| No | 49 (25.9) | 84 (43.5) | 83 (43.7) | 38 (28.2) | 178 (40.7) | 216 (37.8) |
| Yes | 140 (74.1) | 109 (56.5) | 107 (56.3) | 97 (71.9) | 259 (59.3) | 356 (62.2) |
| Surgery (N, %) | ||||||
| No | 99 (52.4) | 85 (44.0) | 57 (30.0) | 110 (81.5) | 131 (30.0) | 241 (42.1) |
| Yes | 90 (47.6) | 108 (56.0) | 133 (70.0) | 25 (18.5) | 306 (70.0) | 331 (57.9) |
Overall HRQOL in AYA Cancer Survivors
The overall means and standard deviations for each HRQOL domain in the full sample and stratified by treatment status can be found in Table 2. Five of the seven domains (anxiety, depression, fatigue, pain interference, and physical function) reached the clinically-relevant threshold in the overall sample. Additionally, this finding held true and was extended in the on-treatment survivors with all seven domains reaching the minimum threshold for clinical relevance. Among off-treatment survivors, only the psychosocial outcomes (anxiety, depression) met the clinically-relevant threshold.
Table 2.
Overall means and standard deviations of PROMIS®-29 T-scores, stratified by treatment status
| Overall Sample N = 572 |
On-Treatment N = 294 |
Off-Treatment N = 278 |
|
|---|---|---|---|
| Symptoms | |||
| Anxiety | 61.1 (11.3) | 66.5 (8.0) | 55.3 (11.5) |
| Depression | 58.6 (10.7) | 63.2 (8.0) | 53.7 (10.9) |
| Fatigue | 55.8 (10.2) | 60.2 (7.3) | 51.2 (10.8) |
| Pain Interference | 57.8 (9.1) | 62.2 (6.5) | 53.2 (9.2) |
| Sleep Disturbance | 54.0 (8.1) | 57.0 (6.5) | 50.8 (8.5) |
| Function | |||
| Physical Function | 41.8 (9.2) | 37.0 (6.8) | 47.0 (8.6) |
| Social Participation | 46.0 (9.6) | 41.2 (6.5) | 51.0 (9.8) |
Higher scores= greater symptom burden for pain interference, fatigue, sleep disturbance, anxiety, depression Lower scores= worse functioning for social participation, physical function.
Differences in HRQOL by Cancer Type and Developmental Stage
For all HRQOL domains, we only saw a significant interaction effect between treatment status and developmental stage. There was no difference between AYAs with hematologic cancers compared to AYAs with solid tumors across any physical domains (all p’s> .003; see Table 3). However, we saw statistically significant differences in some physical outcomes between treatment exposures (see Table 3). Although these differences were statistically significant, none of them exceeded the MID threshold.
Table 3.
Generalized linear model examining differences in physical quality of life outcomes by developmental stage, treatment status, and cancer type
| Physical Outcomes | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pain | Fatigue | Sleep Disturbance | Physical Function | |||||||||
| β | S.E.b |
p- value |
β | S.E. |
p- value |
β | S.E. |
p- value |
β | S.E. |
p- value |
|
| Developmental Stage*Treatment Status | ||||||||||||
| On-Treatment | ||||||||||||
| Emerging adults vs. young adults | −0.45 | 1.03 | 0.659 | −0.74 | 1.19 | 0.530 | −0.004 | 0.98 | 1.00 | 0.44 | 0.95 | 0.642 |
| Emerging adults vs. adolescents | −2.93 | 1.15 | 0.011 | −1.06 | 1.32 | 0.420 | −0.51 | 1.09 | 0.640 | 5.80 | 1.06 | < .001 |
| Young adults vs. adolescents | −2.47 | 1.13 | 0.029 | −0.32 | 1.30 | 0.804 | −0.51 | 1.07 | 0.637 | 5.36 | 1.04 | < .001 |
| Off-Treatment | ||||||||||||
| Emerging adults vs. young adults | −1.99 | 1.05 | 0.058 | −1.70 | 1.22 | 0.164 | −1.74 | 1.00 | 0.082 | 2.43 | 0.97 | 0.013 |
| Emerging adults vs. adolescents | 6.25 | 1.05 | < .001 | 9.11 | 1.22 | < .001 | 6.13 | 1.00 | < .001 | −5.97 | 0.98 | < .001 |
| Young adults vs. adolescents | 8.24 | 1.10 | < .001 | 10.81 | 1.26 | < .001 | 7.87 | 1.04 | < .001 | −8.40 | 1.02 | < .001 |
| Cancer Type (Ref= Solid Tumors)a | ||||||||||||
| Hematologic cancers | −1.02 | 0.81 | 0.208 | 0.81 | 0.94 | 0.388 | −0.02 | 0.77 | 0.982 | 0.23 | 0.75 | 0.767 |
| Chemotherapy (Ref = No chemotherapy) | ||||||||||||
| Received chemotherapy | 2.23 | 0.68 | 0.001 | 1.34 | 0.78 | 0.088 | 0.89 | 0.65 | 0.171 | −1.88 | 0.63 | 0.003 |
| Radiation (Ref= No radiation) | ||||||||||||
| Received radiation | 2.44 | 0.64 | < .001 | 0.36 | 0.74 | 0.632 | −0.15 | 0.61 | 0.810 | −1.49 | 0.59 | 0.012 |
| Surgery (Ref= No surgery) | ||||||||||||
| Had surgery | −0.29 | 0.70 | 0.683 | −1.18 | 0.81 | 0.150 | −0.85 | 0.67 | 0.204 | 0.54 | 0.65 | 0.408 |
| Comorbidity (Ref= No comorbid conditions) | ||||||||||||
| 1 comorbid condition | 0.91 | 1.08 | 0.399 | 0.05 | 1.25 | 0.971 | −0.87 | 1.03 | 0.400 | −1.07 | 1.00 | 0.287 |
| 2-3 comorbid conditions | 3.92 | 1.11 | 0.001 | 5.65 | 1.29 | < .001 | 3.80 | 1.06 | < .001 | −4.26 | 1.04 | < .001 |
| 4+ comorbid conditions | 6.60 | 1.22 | < .001 | 7.13 | 1.40 | < .001 | 6.97 | 1.16 | < .001 | −6.93 | 1.13 | < .001 |
All models were also adjusted for: race/ethnicity, sex, insurance status
Higher scores= greater symptom burden for pain, fatigue, sleep disturbance; lower scores= worse physical function
Bolded values= statistically significant result of p ≤ .003
Ref= Reference group
S.E.= Standard Error
Developmental stage emerged as a significant predictor of HRQOL across all HRQOL domains, particularly among off-treatment survivors (see Tables 3, 4 and Figures S1 through S7). Off-treatment emerging adults reported significantly higher and clinically-relevant anxiety (β=5.92, p< .001), depression (β=6.06, p< .001), pain (β=6.25, p< .001), fatigue (β=9.11, p< .001), and sleep disturbance (β=6.13, p< .001) compared with adolescents. Emerging adults also reported significantly worse and clinically-relevant social participation (β=−7.04, p< .001) and physical function (β=−5.97, p< .001) compared with adolescents. Young adults reported significantly higher and clinically-relevant anxiety (β=10.48, p< .001), depression (β=9.62, p< .001), pain (β=8.24, p< .001), fatigue (β=10.81, p< .001), and sleep disturbance (β=7.87, p< .001) compared with adolescents. Young adults also reported significantly worse and clinically-relevant social participation (β=−8.93, p< .001) and physical function (β=−8.40, p< .001) compared with adolescents. Among on-treatment survivors, there were few differences in HRQOL across the three developmental stages.
Table 4.
Generalized linear model examining differences in psychosocial quality of life outcomes by developmental stage, treatment status, and cancer type
| Psychosocial Outcomes | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Anxiety | Depression | Social Participation | |||||||
| β | S.E. | p-value | β | S.E. | p-value | β | S.E. | p-value | |
| Developmental Stage*Treatment Status | |||||||||
| On-Treatment | |||||||||
| Emerging adults vs. young adults | −1.75 | 1.28 | 0.172 | −0.31 | 1.26 | 0.800 | 1.16 | 1.06 | 0.273 |
| Emerging adults vs. adolescents | −4.66 | 1.43 | 0.001 | −2.90 | 1.41 | 0.040 | 3.98 | 1.18 | 0.001 |
| Young adults vs. adolescents | −2.91 | 1.40 | 0.038 | −2.58 | 1.38 | 0.062 | 2.82 | 1.16 | 0.015 |
| Off-Treatment | |||||||||
| Emerging adults vs. young adults | −4.56 | 1.31 | 0.001 | −3.56 | 1.30 | 0.006 | 1.89 | 1.08 | 0.079 |
| Emerging adults vs. adolescents | 5.92 | 1.32 | < .001 | 6.06 | 1.30 | < .001 | −7.04 | 1.08 | < .001 |
| Young adults vs. adolescents | 10.48 | 1.38 | < .001 | 9.62 | 1.34 | < .001 | −8.93 | 1.13 | < .001 |
| Cancer Type (Ref= Solid Tumors) | |||||||||
| Hematologic cancers | −1.01 | 1.02 | 0.322 | −1.86 | 1.01 | 0.066 | 0.61 | 0.84 | 0.468 |
| Chemotherapy (Ref = No chemotherapy) | |||||||||
| Received chemotherapy | 0.85 | 0.85 | 0.317 | 2.19 | 0.83 | 0.009 | −2.57 | 0.7 | 0.007 |
| Radiation (Ref= No radiation) | |||||||||
| Received radiation | 0.88 | 0.80 | 0.273 | 1.02 | 0.79 | 0.196 | −1.77 | 0.66 | 0.007 |
| Surgery (Ref= No surgery) | |||||||||
| Had surgery | −1.03 | 0.87 | 0.240 | −1.80 | 0.87 | 0.039 | 0.60 | 0.72 | 0.402 |
| Comorbidity (Ref= No comorbid conditions) | |||||||||
| 1 comorbid condition | 0.06 | 1.35 | 0.965 | 0.003 | 1.34 | 0.998 | −0.27 | 1.11 | 0.806 |
| 2-3 comorbid conditions | 5.28 | 1.39 | < .001 | 4.34 | 1.38 | 0.002 | −4.53 | 1.15 | < .001 |
| 4+ comorbid conditions | 7.44 | 1.51 | < .001 | 6.68 | 1.49 | < .001 | −6.67 | 1.25 | < .001 |
All models were also adjusted for: race/ethnicity, sex, insurance status
Higher scores= greater symptom burden for anxiety, depression; lower scores= worse social participation
Bolded values= statistically significant result of p ≤ .003
Ref= Reference group
S.E.= Standard Error
Discussion
AYA cancer survivors experience a clinically-relevant level of symptom burden. The adjusted T-scores for all AYA cancer survivors reached the threshold for a clinically-relevant difference compared to the general U.S. population for anxiety, depression, fatigue, pain, and physical function, and all seven HRQOL domains reached the threshold for a clinically-relevant difference in on-treatment survivors. Among the off-treatment AYA survivors, only the psychosocial outcomes (anxiety, depression) met the threshold for a clinically-relevant difference. Contrary to our hypothesis, we found no significant differences between hematologic cancers and solid tumors in any HRQOL domain assessed, but developmental stage was a consistent predictor of HRQOL across all domains in off-treatment AYA survivors.
Our results suggest that direct assessment of treatment exposure, rather than cancer type, may be more important in AYA survivors. We found that certain treatment exposures were associated with higher physical symptom burden and worse physical function, but were not associated with differences in psychosocial outcomes. This finding is distinct from previous research in non-AYA survivor populations that have found differences in psychosocial outcomes and overall HRQOL by treatment exposure,20 but is supported by previous research in samples of primarily AYA survivors that found no differences in distress by cancer type or treatment.27,28 Adolescence and young adulthood are stages of development characterized by numerous physiological and psychosocial changes, and these factors may be even more critical in survivors’ overall HRQOL than among older cancer patients.
Both off-treatment emerging adult and young adult survivors reported significantly greater symptoms and worse functioning compared to off-treatment adolescent survivors. For both developmental groups, the two domains with the highest adjusted mean T-scores were anxiety and depression. These two domains may be particularly impaired in older AYA survivors because they are likely to be balancing some of this life stage’s most stressful responsibilities including full-time work, parenting, and financial stressors (e.g. lack of health insurance, mortgage).29,30 Future interventions need to focus on the HRQOL domains most salient to AYAs and be flexible enough to target different domains across the AYA developmental period.
Study Limitations
The results of this study must be interpreted considering its limitations. First, this study was an anonymous, cross-sectional survey. We cannot directly test causality or confirm the accuracy of the clinical/sociodemographic details. However, the current study was the largest observational study to utilize a stratified sampling approach to ensure adequate representation and statistical power across the AYA developmental period. Second, AYA survivors were recruited through an online research panel which may limit generalizability. However, data collected through online research panels is comparable to data obtained from population-based estimates.31 Third, the PROMIS-29 was developed and validated for use with adult (≥ 18 years) cancer survivors and this could drive some of the variability we saw in the current analysis. However, previous work has demonstrated understanding of the adult PROMIS items among adolescent survivors.32,33 Fourth, the MID thresholds were developed in advanced-stage, adult cancer survivors and may not describe clinically meaningful differences among AYA populations. However, many of the observed differences in symptoms and functioning were higher or lower by more than half a standard deviation, a well-established benchmark for meaningful differences on PRO measures.26 Fifth, there may be unmeasured confounding variables not assessed in our current analyses (e.g. rural/urban status).
Clinical Implications
In planning for the transition from active treatment to survivorship,34,35 we need to utilize tailored approaches that consider developmental stage in addition to treatment exposure. Our results suggest that emerging adults and young adults are most at-risk for poor HRQOL outcomes after treatment, regardless of cancer type or treatment exposure. Developmental stage needs to be a piece of information used in the creation of risk-stratified survivorship care pathways in AYA survivors.36 Currently, few symptom monitoring tools, like smartphone apps, are tailored to the needs of older AYA survivors,37 and this must change.
Conclusions
Developmental stage has often been an ignored variable in AYA survivorship research, and needs to be better integrated into cancer care delivery. Additionally, existing HRQOL measures are limited in AYA survivors because they do not measure all important domains (e.g. body image) and are often validated in pediatric (< 18 years) or adult (≥ 18 years) populations, but not both. Linking efforts suggest that pediatric and adult PROMIS measures are comparable,32,33 but efforts are needed to improve these gaps.38 The current study represents a starting point for recognizing the importance of HRQOL and developmental stage in AYA survivorship care and research. Future efforts to improve HRQOL assessment, integrate flexible survivorship care approaches, and consider developmental stage in research continue to be needed.
Supplementary Material
Figure S1. Graph depicting the interaction between treatment status and developmental stage for anxiety
Figure S2. Graph depicting the interaction between treatment status and developmental stage for depression
Figure S3. Graph depicting the interaction between treatment status and developmental stage for social participation
Figure S4. Graph depicting the interaction between treatment status and developmental stage for pain
Figure S5. Graph depicting the interaction between treatment status and developmental stage for fatigue
Figure S6. Graph depicting the interaction between treatment status and developmental stage for sleep disturbance
Figure S7. Graph depicting the interaction between treatment status and developmental stage for physical function
Acknowledgments:
The authors would like to thank Eddie Ip for his help and guidance in conducting the statistical analyses for this study. Research reported in this publication was supported by pilot funding from the Department of Medical Social Sciences at the Northwestern University Feinberg School of Medicine and by the National Cancer Institute of the NIH under award number R01CA218398. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Footnotes
Conflict of Interest: The authors have no conflicts of interest to declare.
Data Availability Statement:
Data available on request from the authors.
References
- 1.Smith AW, Seibel NL, Lewis DR, et al. Next steps for adolescent and young adult oncology workshop: an update on progress and recommendations for the future. Cancer 2016; 122:988–999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Cimprich B, Ronis DL, Martinez-Ramos G. Age at diagnosis and quality of life in breast cancer survivors. Cancer Pract 2002; 10:85–93. [DOI] [PubMed] [Google Scholar]
- 3.Dunn J, Ng SK, Breitbart W, et al. Health-related quality of life and life satisfaction in colorectal cancer survivors: trajectories of adjustment. Health Qual Life Outcomes 2013; 11:46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Smith AW, Bellizzi KM, Keegan THM, et al. Health-related quality of life of adolescent and young adult patients with cancer in the United States: the adolescent and young adult health outcomes and patient experience study. J Clin Oncol 2013; 31(17):2136–2145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Husson O, Prins JB, Kaal SEJ, et al. Adolescent and young adult (AYA) lymphoma survivors report lower health-related quality of life compared to a normative population: results from the PROFILES registry. Acta Oncologica 2017; 56(2):288–294 [DOI] [PubMed] [Google Scholar]
- 6.Salsman JM, Garcia SF, Yanez B, Sanford SD, Snyder MA, Victorson D. Physical, emotional, and social health differences between posttreatment young adults with cancer and matched healthy controls. Cancer 2014; 120:2247–2254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Henderson JR, Kiernan E, McNeer JL, et al. Patient-reported health-related quality-of-life assessment at the point-of-care with adolescents and young adults with cancer. J Adolesc Young Adult Oncol 2018; 7(1):97–102 [DOI] [PubMed] [Google Scholar]
- 8.Erhardt MJ, Hochberg J, Bjornard KL, Brinkman TM. Long-term survivors of childhood, adolescent and young adult non-Hodgkin lymphoma. Br J Haematol 2019; 185:1099–1110. [DOI] [PubMed] [Google Scholar]
- 9.Nicklin E, Velikova G, Hulme C, et al. Long-term issues and supportive care needs of adolescent and young adult childhood brain tumour survivors and their caregivers: a systematic review. Psychooncology 2019; 28(3):477–487. [DOI] [PubMed] [Google Scholar]
- 10.Keegan THM, Lichtensztajn DY, Kato I, et al. Unmet adolescent and young adult cancer survivors information and service needs: a population-based cancer registry study. J Cancer Surviv 2012; 6:239–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wettergren L, Kent EE, Mitchell SA, et al. Cancer negatively impacts on sexual function in adolescents and young adults: the AYA HOPE study. Psychooncology 2017; 26:1632–1639. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Collins WA, Steinberg L. Adolescent development in interpersonal context. In: Eisenberg N, Damon W, Lerner RM, eds. Handbook of Child Psychology: Vol 3, Social, Emotional, and Personality Development. Hoboken, NJ: John Wiley & Sons, Inc. 2006:1003–1067. [Google Scholar]
- 13.McNeely C, Blanchard J. The teen years explained: A guide to healthy adolescent development. Baltimore, MD: Center for Adolescent Health at Johns Hopkins Bloomberg School of Public Health; 2009. [Google Scholar]
- 14.Piaget J Intellectual evolution from adolescence to adulthood. Hum Dev 2008; 51:40–47 (Reprint from Human Development 1972;15:1-12). [Google Scholar]
- 15.Eluvanthigal TJ, Hasan KM, Kramer L, et al. Quantitative diffusion tensor tractography of association and projection fibers in normally developing children and adolescents. Cereb Cortex 2007; 17:2760–2768. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chambers RA, Taylor JR, Potenza MN. Developmental neurocircuity of motivation in adolescence: a critical period of addiction vulnerability. Am J Psychiatry 2003; 160:1041–1052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Arnett JJ. Emerging adulthood: a theory of development from the late teens through the twenties. Am Psychol 2000; 55(5):469–480. [PubMed] [Google Scholar]
- 18.Close AG, Dreyzin A, Miller KD, et al. Adolescent and young adult oncology – past, present, and future. CA Cancer J Clin 2019; 69:485–496. [DOI] [PubMed] [Google Scholar]
- 19.Moradian S, Voelker N, Brown C, Liu G, Howell D. Effectiveness of internet-based interventions in managing chemotherapy-related symptoms in patients with cancer: a systematic literature review. Support Care Cancer 2018; 26:361–374. [DOI] [PubMed] [Google Scholar]
- 20.Lee ES, Lee MK, Kim SH, et al. Health-related quality of life in survivors with breast cancer 1 year after diagnosis compared with the general population: a prospective cohort study. Ann Surg 2011; 253:101–108. [DOI] [PubMed] [Google Scholar]
- 21.Halfon N, Hochstein M: Life course health development: an integrated framework for developing health, policy, and research. Milbank Q 2000; 80(3): 433–479. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Baltes PB, Staurdinger UM, Lindenberger U: Lifespan psychology: theory and application to intellectual functioning. Annu Rev Psychol 1999; 50: 471–507. [DOI] [PubMed] [Google Scholar]
- 23.Hays RD, Spritzer KL, Schalet BD, Cella D. PROMIS®-29 v2.0 profile physical and mental health summary scores. Qual Life Res 2018; 27(7):1885–1891 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Healthmeasures.net [online resource]. PROMIS adult profile instruments: A brief guide to the PROMIS® profile instruments for adult respondents. Scoring tables pp.16–17. [Cited on 2020 July 9]. Available from: http://www.healthmeasures.net/images/PROMIS/manuals/PROMIS_Adult_Profile_Scoring_Manual.pdf [Google Scholar]
- 25.Yost KJ, Eton DT, Garcia SF, Cella D. Minimally important differences were estimated for six Patient-Reported Outcomes Measurement Information System-Cancer scales in advanced-stage cancer patients. J Clin Epidemiol 2011; 64:507–516. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Norman GR, Sloan JA, Wyrwich KW. Interpretation of changes in health-related quality of life. the remarkable universality of half a standard deviation. Medical Care 2003; 41(5):582–592. [DOI] [PubMed] [Google Scholar]
- 27.Kwak M, Zebrack BJ, Meeske KA, et al. Trajectories of psychological distress in adolescent and young adults patients with cancer: 1-year longitudinal study. J Clin Oncol 2013; 31(17):2160–2166. [DOI] [PubMed] [Google Scholar]
- 28.Robert RS, Ottaviani G, Huh WW, Palla S, Jaffe N. Psychosocial and functional outcomes in long-term survivors of osteosarcoma: a comparison of limb-salvage surgery and amputation. Pediatr Blood Cancer 2010; 54:990–999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kaddas HK, Pannier ST, Mann K, et al. Age-related differences in financial toxicity and unmet resource needs among adolescent and young adult cancer patients. J Adolesc Young Adult Oncol 2019; Epub ahead of print. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Landwehr MS, Watson SE, Macpherson CF, Novak KA, Johnson RH. The cost of cancer: a retrospective analysis of the financial impact of cancer on young adults. Cancer Med 2016; 5(5):863–870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Liu H, Cella D, Gershon R, et al. Representativeness of the PROMIS internet panel. J Clin Epidemiol 2010; 63:1169–1178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Reeve BB, Thissen D, DeWalt DA, et al. Linkage between the PROMIS® pediatric and adult emotional distress measures. Qual Life Res 2016; 25:823–833. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Tulsky DS, Kisala PA, Boulton AJ, et al. Determining a transitional scoring link between PROMIS® pediatric and adult physical health measures. Qual Life Res 2018; 28:1217–1229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Yabroff KR, Gansler T, Wender RC, et al. Minimizing the burden of cancer in the United States: goals for a high-performing health care system. CA-Cancer J Clin 2019; 69:166–183. [DOI] [PubMed] [Google Scholar]
- 35.Naylor M, Berlinger N. Transitional care: a priority for health care organizational ethics. Hastings Cent Rep 2016; 46:S39–S42. [DOI] [PubMed] [Google Scholar]
- 36.Mayer DK, Alfano CM. Personalized risk-stratified cancer follow-up care: its potential for healthier survivors, happier clinicians and lower costs. J Natl Cancer I 2019; 11:442–448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Wesley KM, Fizur PJ. A review of mobile applications to help adolescent and young adult cancer patients. Adolesc Health Med Ther 2015; 6:141–148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Salsman JM, Danhauer SC, Moore JB, et al. Optimizing the measurement of health-related quality of life in adolescents and young adults with cancer. Cancer In press. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1. Graph depicting the interaction between treatment status and developmental stage for anxiety
Figure S2. Graph depicting the interaction between treatment status and developmental stage for depression
Figure S3. Graph depicting the interaction between treatment status and developmental stage for social participation
Figure S4. Graph depicting the interaction between treatment status and developmental stage for pain
Figure S5. Graph depicting the interaction between treatment status and developmental stage for fatigue
Figure S6. Graph depicting the interaction between treatment status and developmental stage for sleep disturbance
Figure S7. Graph depicting the interaction between treatment status and developmental stage for physical function
Data Availability Statement
Data available on request from the authors.