Abstract
Background
Previous longitudinal evidence has suggested that patients aged > 75 years may be associated with a poorer Health Utility Index Mark 3 (HUI-3) trajectory. We compared the longitudinal health utility scores (HUS) of older (>75 years) versus younger patients with head and neck cancers across various clinical subgroups and identified factors associated with decline in quality of life.
Methods
The study prospectively collected clinical information and follow-up data from head and neck cancer patients. The HUI-3 scale was used to assess changes in health-related quality of life at baseline, post-surgery, mid-radiation therapy, and at 3, 6, and 12 months. A linear mixed model was employed to evaluate the changes in HUS and age categories over time.
Results
We included 859 patients with head and neck cancer, of whom 647 were included in the longitudinal analytic cohort. In descriptive analyses, patients aged > 75 years showed less complete recovery in mean HUS after early treatment-related decline compared with younger patients. In the primary mixed-effects model, younger age groups had more positive average HUS slopes than patients aged > 75 years. However, these age-related differences were attenuated and were no longer statistically supported after adjustment for clinical stage, Eastern Cooperative Oncology Group (ECOG) performance status, and treatment modality. Subgroup findings were exploratory and were not consistently supported across all model-based comparisons.
Conclusion
Patients aged > 75 years may be at risk for slower or incomplete post-treatment recovery in HUI-3–derived health utility. These findings should be interpreted cautiously given potential selection bias, missing data, and attenuation after covariate adjustment. Larger studies with detailed geriatric and domain-specific functional measures are needed to confirm these patterns.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12955-026-02551-7.
Keywords: Head and neck cancer, Elderly patients, HUI-3 questionnaires, Time-dependent analysis
Background
In Canada, head and neck cancer (HNC) represents a clinically important cancer burden, with approximately 8,200 individuals expected to be diagnosed with HNC and 2,300 expected to die from the disease in 2026 [1]. HNC comprises a heterogeneous group of malignancies arising from the pharynx (nasopharynx, oropharynx, hypopharynx), larynx, nasal cavity, lips and oral cavity, paranasal sinuses, and salivary glands [2]. Primary treatments for HNC include surgery, radiotherapy, and chemotherapy, delivered alone or in combination. As overall survival from HNC is relatively favorable compared with other cancers, long-term treatment-related toxicities pose significant concerns [3]. Complications affecting speech, swallowing, and facial appearance can profoundly impact HNC patients’ mental health and physical functioning and reduce their health-related quality of life (HRQoL) both during and after their recovery, and may prevent some patients from returning to pre-treatment functioning [4–6].
Older adults may be particularly vulnerable to persistent or delayed functional decline after treatment due to a higher burden of comorbidity and frailty, lower physiologic reserve, poorer baseline swallowing function, reduced treatment tolerance, and slower recovery [7, 8]. With population aging, the number of elderly cancer patients is expected to increase [9, 10]. However, evidence regarding age-related differences in HRQoL outcomes after HNC treatment remains mixed: earlier studies reported similar HRQoL impacts across age groups [11–13], whereas more recent studies suggest greater functional decline and supportive care needs among older patients [14–16]. These discrepancies may reflect differences in follow-up duration, survivorship/attrition bias, heterogeneity in HRQoL instruments with variable sensitivity to key domains such as swallowing and speech, and residual confounding related to treatment intensity, comorbidity, and frailty.
As a preference-based summary measure, the Health Utilities Index Mark 3 (HUI-3) generates a health utility score (HUS) that captures multiple domains of HRQoL on a common utility scale, providing a clinically interpretable and complementary outcome for longitudinal comparison across age groups [17]. Previous longitudinal work has suggested that patients aged > 75 years may experience poorer HUI-3-derived HUS trajectories [18]. However, the clinical implications of this observation remain incompletely characterized. In particular, it remains important to clarify the timing and persistence of post-treatment HUS changes across age strata and to identify the disease- and treatment-related contexts in which older patients experience clinically meaningful decrements. Therefore, the present study was designed to extend prior work by focusing on age-stratified longitudinal HUS trajectories after HNC treatment and by examining clinical contexts associated with persistent post-treatment decrements. Specifically, we aimed to (i) characterize time-dependent changes in HUS following treatment and (ii) identify clinical contexts associated with persistent post-treatment decrements.
Methods
Patient eligibility
Patients were recruited to a prospective observational study from August 2014 to March 2019 at the Princess Margaret Cancer Centre in Toronto, Canada. Eligible participants were adults (≥18 years) with a newly diagnosed primary HNC. Patients were excluded if they had recurrent disease or a second primary tumor at enrollment, were treated with palliative intent, or were unable to complete patient-reported questionnaires independently. Detailed psychiatric diagnoses and literacy level were not systematically collected.
The recruited cohort comprised all eligible patients enrolled in the study. For the longitudinal analysis, we defined a longitudinal analytic cohort comprising patients with a baseline assessment and at least one follow-up patient-reported outcome assessment. Patients without baseline data or without any follow-up assessment were not included in the longitudinal analytic cohort. This study was approved by the University Health Network Institutional Research Ethics Board (REB #07–0521).
Data collection and measurement tools:
Clinico-demographic characteristics were collected by self-report at baseline (primary HNC diagnosis) and supplemented by medical chart review. Patient-reported quality-of-life data were assessed using the HUI-3 questionnaire. HUS was derived from the HUI-3, a validated instrument that measures eight attributes: vision, hearing, speech, ambulation, dexterity, emotion, cognition, and pain, each with 5 or 6 levels. Responses to the HUI-3 questionnaire were converted into attribute levels and then transformed using the standard HUI-3 multi-attribute utility algorithm to generate an overall HUS. The HUI-3-derived HUS has a theoretical range from −0.36 to 1.00, where 1.00 represents perfect health and 0 represents the death anchor; values below 0 represent health states considered worse than death. Thus, 0 was not interpreted as the lowest possible HUS value. Lower HUS values indicate poorer health-related quality of life.
HUI-3 assessments were collected at prespecified timepoints defined relative to diagnosis and treatment milestones (Table S1), including baseline before treatment, post-surgery when applicable, mid-radiotherapy, and post-treatment follow-up at 3, 6, and 12 months after completion of primary treatment. Because actual questionnaire completion dates did not always coincide exactly with the scheduled assessment time points, each HUI-3 assessment was assigned a time value based on the interval from the baseline HUI-3 assessment to the date of questionnaire completion. For model fitting and reporting, time was expressed in years, with baseline coded as 0. This approach allowed participants with irregular assessment timing to contribute all available longitudinal observations while preserving the temporal ordering of assessments. Only questionnaires with complete HUI-3 data were scored; questionnaires with incomplete HUI-3 responses were treated as missing for HUS estimation. Among patients included in the longitudinal analytic cohort, missing observations occurred across follow-up time points. Reasons for missingness were not systematically recorded for all missed assessments; when known, they included patient death, transfer of care, and missed questionnaire administration during clinic follow-up.
Baseline Eastern Cooperative Oncology Group (ECOG) performance status was recorded at diagnosis as part of routine prospective clinical assessment by the treating team. ECOG was graded on the standard 0–4 scale (0 = fully active; higher scores indicate increasing functional limitation) and was dichotomized a priori as 0 versus ≥1 for subgroup analyses.
The modified Charlson Comorbidity Index (mCCI) was derived at baseline using comorbidity information abstracted from medical records [19]. Comorbidities present prior to initiation of primary cancer treatment were coded, weighted according to the established Charlson scoring algorithm, and summed to generate the mCCI score. For subgroup analyses, comorbidity burden was categorized a priori as 0–1 versus ≥2 to contrast low versus higher comorbidity burden and to support stable estimates.
Statistical analysis
To facilitate clinically interpretable comparisons, we stratified age a priori into <65, 65–75, and >75 years. We used 65 years to reflect a commonly applied clinical threshold for “older adults” in oncology, and we separated > 75 years to represent an “older-old” subgroup in whom comorbidity burden and reduced functional reserve are more prevalent. The 65–75-year group was included as an intermediate stratum to evaluate potential graded age associations.
HUS over time was first summarized descriptively by age group. The primary analyses used linear mixed-effects models to evaluate associations between age group and longitudinal HUS. Time was modeled as a continuous variable, coded in years from the baseline HUI-3 assessment, with baseline assigned a value of 0. For follow-up assessments, time was calculated using the interval between the baseline HUI-3 assessment and the date of questionnaire completion, allowing participants with irregular assessment timing to contribute all available longitudinal observations. The model included fixed effects for age group, time, and the age group × time interaction, with patient-level random intercepts to account for within-patient correlation. The age group × time interaction estimated whether the average rate of change in HUS differed across age strata. The reference age category was set to >75 years, allowing direct comparison of the >75-year group with the <65-year and 65–75-year groups. A continuous linear time specification was selected to estimate average change over follow-up and to preserve model parsimony. HUS trajectories were also examined descriptively across scheduled assessment windows, and model residuals were evaluated. Because the linear specification may not capture short-term nonlinear changes around treatment milestones, model coefficients were interpreted as average longitudinal changes rather than strictly linear individual recovery patterns.
Exploratory subgroup analyses were conducted based on gender, clinical stage, treatment modality, disease site, ECOG performance status [0 vs ≥ 1] and mCCI [0–1 vs ≥ 2], and to compare the HUS trajectories among different age categories within each clinical subgroup.
We conducted a sensitivity analysis of our main model adjusting for stage, ECOG, and treatment modality to assess robustness. All variables were chosen for analysis a priori, based on clinical expertise. Model residuals were examined, and marginal and conditional R2 values were computed to assess goodness-of-fit.
Linear mixed-effects models used all available complete HUS observations and assume that missing outcome data were missing at random conditional on observed covariates, observed HUS history, assessment timing, and patient-level random effects. Because this assumption may not fully hold if missingness was related to death, transfer of care, inability to complete questionnaires, or unobserved clinical deterioration, results were interpreted cautiously.
Results
The recruited cohort comprised 859 HNC patients, including 495 patients aged < 65 years, 288 patients aged 65–75 years, and 76 patients aged > 75 years. Baseline clinical characteristics of the recruited cohort are shown in Table 1. Approximately 80% of patients were male across all age groups. Clinical stage distribution was broadly similar across age strata. Treatment modality differed by age group: among patients aged < 65 years, chemoradiotherapy (CRT) (41%) and surgery (39%) were most common primary treatments, whereas radiotherapy (RT) alone was the most frequent among those aged 65–75 years (48%) and >75 years (68%). Oropharyngeal cancer was the most common disease site across all age groups. Most younger patients (<65 years) had an ECOG performance status of 0, whereas the majority of older patients (aged 65–75 and >75) had an ECOG score of ≥1. Across all age strata, the most common mCCI category was 0–1.
Table 1.
Clinical characteristics of head and neck cancer patients across three age groups
| Covariate | Full Sample (n = 859) | <65 (n = 495) | 65–75 (n = 288) | >75 (n = 76) | p-value | |
|---|---|---|---|---|---|---|
| gender | Female | 182 (21) | 99 (20) | 66 (23) | 17 (22) | 0.61 |
| Male | 677 (79) | 396 (80) | 222 (77) | 59 (78) | ||
| stage 7th | I/II/III | 350 (41) | 188 (38) | 124 (43) | 38 (50) | 0.10 |
| IV | 506 (59) | 304 (62) | 164 (57) | 38 (50) | ||
| Missing | 3 | 3 | 0 | 0 | ||
| treatment | RT only | 286 (33) | 97 (20) | 137 (48) | 52 (68) | <0.001 |
| CRT | 270 (31) | 204 (41) | 65 (23) | 1 (1) | ||
| CT only | 2 (0) | 2 (0) | 0 (0) | 0 (0) | ||
| Sx ± Adj | 300 (35) | 192 (39) | 85 (30) | 23 (30) | ||
| Missing | 1 | 0 | 1 | 0 | ||
| disease site | Lip & Oral Cavity | 206 (24) | 125 (25) | 61 (21) | 20 (26) | 0.014 |
| Larynx | 144 (17) | 64 (13) | 60 (21) | 20 (26) | ||
| Oropharynx | 403 (47) | 239 (48) | 136 (47) | 28 (37) | ||
| others | 106(12) | 67 (14) | 31 (11) | 8 (11) | ||
| ECOG | 0 | 358 (50) | 259 (62) | 85 (36) | 14 (23) | <0.001 |
| 1+ | 352 (50) | 156 (38) | 150 (64) | 46 (77) | ||
| Missing | 149 | 80 | 53 | 16 | ||
| mCCI | 0–1 | 643 (89) | 383 (93) | 210 (84) | 50 (75) | <0.001 |
| 2+ | 87 (12) | 29 (7) | 41 (16) | 17 (25) | ||
| Missing | 129 | 83 | 37 | 9 | ||
| smoking | Current | 131 (18) | 87 (22) | 40 (16) | 4 (6) | 0.021 |
| Former | 359 (50) | 189 (47) | 130 (53) | 40 (65) | ||
| Never | 222 (31) | 127 (32) | 77 (31) | 18 (29) | ||
| Missing | 147 | 92 | 41 | 14 | ||
| drinking | Current | 315 (46) | 186 (48) | 105 (46) | 24 (40) | 0.23 |
| Former | 180 (26) | 92 (24) | 71 (31) | 17 (28) | ||
| Never | 185 (27) | 112 (29) | 54 (23) | 19 (32) | ||
| Missing | 179 | 105 | 58 | 16 | ||
| highest educ | Elementary/Middle School | 130 (18) | 51 (12) | 54 (22) | 25 (37) | <0.001 |
| Post-Secondary | 359 (49) | 214 (52) | 123 (50) | 22 (33) | ||
| Secondary School | 188 (26) | 119 (29) | 56 (23) | 13 (19) | ||
| Vocational/Technical/Other | 51 (7) | 29 (7) | 15 (6) | 7 (10) | ||
| Missing | 131 | 82 | 40 | 9 | ||
| race | Caucasian | 623 (84) | 341 (81) | 223 (89) | 59 (88) | 0.021 |
| Other | 115 (16) | 79 (19) | 28 (11) | 8 (12) | ||
| Missing | 121 | 75 | 37 | 9 | ||
Of the 859 recruited patients, 647 patients had both a baseline HUI-3 assessment and at least one follow-up HUI-3 assessment and were included in the longitudinal analytic cohort. This analytic cohort included 365 patients aged < 65 years, 226 patients aged 65–75 years, and 56 patients aged > 75 years. A total of 212 patients were not included in the longitudinal analytic cohort because they did not meet these longitudinal data requirements. The derivation of the longitudinal analytic cohort is shown in Figure S1.
HUS data availability by assessment time point and age group in the longitudinal analytic cohort is summarized in Table S2. Missing HUS observations were present at each follow-up assessment time point. Because post-surgery/pre-adjuvant and mid-radiotherapy assessments were tied to treatment-specific milestones, lower availability at these time points partly reflects differences in treatment pathways and assessment applicability, rather than follow-up attrition alone.
Baseline characteristics of patients included and not included in the longitudinal analytic cohort are compared in Table S3. Included and excluded patients were broadly similar with respect to sex, clinical stage, treatment modality, disease-site, ECOG performance status, and modified CCI. These comparisons did not show strong statistical evidence of baseline imbalance between included and excluded patients; however, selection bias due to non-random missingness cannot be excluded.
Figure 1 Ashows age-stratified mean HUS trajectories relative to baseline. Across all age groups, mean HUS decreased during the active treatment or early post-treatment period. By the 3-month follow-up, mean HUS appeared to recover toward baseline among patients aged < 65 years and 65–75 years, whereas mean HUS among patients aged > 75 years remained lower than baseline over subsequent follow-up. In the primary linear mixed-effect model, there was statistical evidence that the average annual change in HUS differed by age group. Compared with patients aged > 75 years, patients aged < 65 years had a more positive average annual HUS slope (difference in mean annual change: +0.10, p = 0.01), as did patients aged 65–75 years (difference in mean annual change: +0.12, p < 0.01; Fig. 1B).
Fig. 1.
Time-dependent analysis of HUS in HNC patients across different age categories
In sex-stratified analyses, there was statistical evidence that the HUS slope differed by age category among males (p for age × time < 0.01). Compared with men aged > 75 years, men aged < 65 years and 65–75 years had more positive average annual HUS slopes (difference in mean annual change: +0.12 for <65 vs > 75, p < 0.01; +0.14 for 65–75 vs > 75, p < 0.01; Figs. 2A and 2B). Among females, there was no statistical evidence that HUS slopes differed across age categories (p for age × time = 0.92; Fig. 2C). Therefore, female subgroup results were interpreted as showing no statistical evidence of age-related differences in longitudinal HUS change.
Fig. 2.
Time-dependent analysis of HUS among HNC patients across different age groups in gender subgroups
In analyses stratified by clinical stage, there was statistical evidence of age-category differences in HUS slopes among patients with stage IV disease. Compared with patients aged > 75 years, patients aged < 65 years and 65–75 years had more positive average annual HUS slopes (difference in mean annual change: +0.12 for <65 vs > 75, p = 0.04; +0.13 for 65–75 vs > 75, p = 0.04; Figs. 3A and 3B). Among patients with stage I–III disease, statistical evidence of an age-category difference was observed for the comparison between patients aged 65–75 years and those aged > 75 years (difference in mean annual change: +0.12, p = 0.04; Fig. 3C). Other age-group comparisons in the stage I–III subgroup did not provide statistical evidence of differences in HUS slopes and were not interpreted directionally.
Fig. 3.
Time-dependent analysis of HUS among HNC patients across different age groups in stage subgroups
In analyses stratified by treatment modality, among patients receiving RT alone, there was statistical evidence of a more positive average annual HUS slope in patients aged < 65 years compared with those aged > 75 years (difference in mean annual change: +0.13, p = 0.02; Fig. 4A–C). The comparison for patients receiving CRT was not performed because only one patient aged > 75 years received CRT, precluding a stable age-stratified estimate in this treatment subgroup. Other treatment-modality subgroup comparisons were interpreted cautiously because of limited subgroup sample sizes and lack of consistent statistical evidence across comparisons.
Fig. 4.
Time-dependent analysis of HUS among HNC patients across different age groups in treatment subgroups
Across disease subsites, descriptive plots suggested numerical heterogeneity in HUS trajectories by age group (Fig. 5A–C). However, in the mixed-effects models, there was no statistical evidence that HUS trajectories differed by age category within oral cavity, oropharyngeal, or laryngeal cancer subgroups (Table S4). Accordingly, these subsite-specific findings were interpreted as exploratory, and no definitive age-related differences in HUS trajectories were inferred from these subgroup analyses.
Fig. 5.
Time-dependent analysis of HUS among HNC patients across different age groups in disease sites, mCCI, ECOG subgroups
In the mCCI subgroup analysis, among patients with mCCI = 0–1, there was no statistical evidence of age-category differences in HUS slopes (difference in mean annual change: +0.08 for <65 vs > 75, p = 0.09; +0.09 for 65–75 vs > 75, p = 0.06; Fig. 5D; Table S5). These results were therefore interpreted as showing no statistical evidence of between-group differences in longitudinal HUS change in the lower comorbidity subgroup. Among patients with mCCI ≥ 2, there was statistical evidence that HUS slopes differed by age category, with more positive average annual HUS slopes in younger groups compared with patients aged > 75 years (+0.29 for <65 vs > 75, p < 0.01; +0.25 for 65–75 vs > 75, p = 0.02; Fig. 5E; Table S5).
In the ECOG performance status subgroup analysis, among patients with ECOG performance status 0, there was statistical evidence of a more positive average annual HUS slope in patients aged < 65 years compared with those aged > 75 years (difference in mean annual change: +0.16, p = 0.04; Fig. 5F; Table S6). Other age-group comparisons in this subgroup did not provide statistical evidence of differences in HUS slopes. Among patients with ECOG performance status ≥ 1, there was no statistical evidence of differences in HUS trajectories across the three age groups (Fig. 5G, Table S6).
In the sensitivity analysis additionally adjusting for clinical stage, ECOG performance status, and treatment modality, there was no statistical evidence that HUS change over time differed by age group. The adjusted age-by-time interaction estimate was +0.05 for patients aged < 65 years compared with those aged > 75 years (SE = 0.03, p = 0.07) and +0.04 for patients aged 65–75 years compared with those aged > 75 years (SE = 0.03, p = 0.16; Table S7). These findings indicate that the age-related differences observed in the primary model were attenuated after additional adjustment for key clinical covariates. Overall model fit indices are shown in Table S8.
Discussion
This longitudinal study evaluated HUS trajectories after curative-intent treatment for HNC across prespecified age strata. Descriptively, patients aged > 75 years showed less complete recovery in mean HUS after the early treatment-related decline compared with younger patients. However, model-based findings were not uniformly consistent across all subgroup analyses, and the sensitivity model adjusted for clinical stage, ECOG performance status, and treatment modality did not provide statistical evidence of differential HUS change over time by age group. Therefore, the observed age-related patterns should be interpreted cautiously and considered hypothesis-generating rather than confirmatory. Given the observational design, age-related differences in treatment selection, and potential non-random attrition, residual confounding and selection bias cannot be excluded.
These findings align with survivorship literature suggesting that HRQoL in older adults often worsens during treatment and improves afterward, although trajectories vary across cancers, instruments, and follow-up windows. A recent systematic review of longitudinal QoL in older adults reported heterogeneous survivorship trajectories—most commonly a decline during active treatment followed by partial recovery—and highlighted comorbidity burden and mobility limitations as recurring correlates of poorer or deteriorating QoL [20]. In HNC, prospective evidence suggests that overall HRQoL in older adults may be broadly comparable to younger patients in the early post-treatment period, while later survivorship can reveal domain-specific vulnerabilities in the oldest patients, particularly physical functioning, fatigue, role limitations and geriatric-specific concerns such as anxiety and challenges maintaining life goals [21, 22].
Potential mechanisms for less favorable post-treatment HUS in the oldest patients include reduced physiologic reserve, higher comorbidity or frailty burden, and lower capacity to compensate for treatment-related functional impairments [23–26]. However, the present subgroup findings should be interpreted cautiously. Exploratory analyses suggested possible variation in HUS trajectories by sex, ECOG performance status, and comorbidity burden, but statistical evidence was not consistent across all subgroup comparisons. For example, age-related differences were not statistically supported within the ECOG ≥ 1 or mCCI 0–1 subgroups. Floor effects may partly explain these findings if patients with worse baseline functional status had limited measurable room for further decline on utility-based outcomes. Selection processes may also attenuate observed age differences if older patients with poorer baseline function who proceeded with treatment and contributed longitudinal patient-reported outcomes represented a relatively fitter subset. These explanations should be considered hypotheses that warrant confirmation in larger cohorts with more granular baseline characterization.
Beyond the total comorbidity counts, the type and functional domain of comorbidity conditions may be relevant because different comorbidities map onto distinct attributes captured by the HUI-3, including ambulation, pain, emotion, cognition, and speech. Consistent with this, prior geriatric oncology studies have reported associations between frailty or geriatric assessment deficits and worse HRQoL courses after HNC treatment. These considerations support further evaluation of comorbidity profiles and geriatric domains, rather than reliance on aggregate comorbidity indices alone, when assessing post-treatment HRQoL risk in older patients with HNC [27, 28].
In exploratory subsite analyses, descriptive HUS patterns varied across disease sites. However, the mixed-effects models did not provide statistical evidence that HUS trajectories differed by age category within oral cavity, oropharyngeal, or laryngeal cancer subgroups. Therefore, no definitive conclusions can be drawn regarding subsite-specific age differences in longitudinal HUS trajectories. Although differences in functional burden [29–31], treatment exposure [32–34], and rehabilitation needs may plausibly contribute to heterogeneity across subsites, these explanations remain hypothesis-generating and need to be confirmed in larger cohorts with detailed domain-specific functional measures.
This study has several limitations. As an observational analysis, the reported associations may be influenced by residual confounding, including age-related differences in treatment selection and baseline health status. In addition, inclusion in the longitudinal analytic cohort required completion of the baseline assessment and at least one follow-up questionnaire. Older or more vulnerable patients, including those with mild cognitive impairment, severe functional decline, transfer of care, or early death, may therefore have been underrepresented, limiting the generalizability of the findings. The linear mixed-effects models specification assumed approximately linear change within the modeled intervals and may not have captured more complex non-linear recovery patterns, given the limited number of observed assessment time points and the absence of measurements between them. Finally, attrition, loss to follow-up, and death may have introduced selection bias through informative dropout, particularly at later follow-up time points.
Conclusions
In conclusion, older adults with HNC, particularly those aged > 75 years, may experience more persistent post-treatment decrements in HUI-3–derived health utility than younger patients. These findings highlight the importance of incorporating longitudinal patient-reported outcomes into survivorship care and suggest that baseline health complexity and functional status may help stratify risk for slower or incomplete recovery. Future studies should confirm these patterns in larger cohorts with detailed domain-specific functional measures and evaluate whether risk-adapted follow-up and multidisciplinary supportive care (e.g., rehabilitation, nutrition, symptom management, and geriatric assessment) can improve post-treatment recovery in the most vulnerable older patients.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
The authors would like to thank the Princess Margaret Cancer Centre Head & Neck Translational Program, with philanthropic funds from the Wharton Family, Joe’s Team, and Gordon Tozer, for their generous support.
Author contributions
Danni Cheng designed the study, performed experiments, analyzed data, and drafted the manuscript. Danni Cheng and Katrina Hueniken contributed to data analysis and interpretation. Danni Cheng, Katrina Hueniken, Andrew Hope, Shao Hui Huang, Anna Spreafico, David P Goldstein, Scott V Bratman, Yu Zhao, Jianjun Ren and John R de Almeida assisted with methodology and provided critical revisions. Wei Xu and Geoffrey Liu supervised the study, acquired funding, and revised the manuscript. All authors read and approved the final manuscript.
Funding
This study was supported by the Princess Margaret Cancer Centre Head & Neck Translational Program, with philanthropic funds from the Wharton Family, Joe’s Team, and Gordon Tozer; and National Natural Science Foundation of China (CDN) (82501390); the Science and Technology Department of Sichuan Province (CDN) (2024NSFSC1513); Sichuan University Interdisciplinary Innovation Fund (CDN); the Postdoctor Research Fund of West China Hospital, Sichuan University (CDN) (2025HXBH004).
Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval
This study was approved by the University Health Network Institutional Research Ethics Board (REB), Toronto, Canada (REB #07–0521).
Consent to participate
Written informed consent was obtained from all participants prior to enrollment and prior to completion of any study-related questionnaires and/or data collection.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Brenner DR, Gillis J, Demers AA, Ellison LF, Ng C, Zhang SX, et al. Projected estimates of cancer in Canada in 2026. CMAJ. 2026;198(14):E526–34. 10.1503/cmaj.252152. [DOI] [PMC free article] [PubMed]
- 2.Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clinicians. 2023;73(1):17–48. 10.3322/caac.21763. [DOI] [PubMed] [Google Scholar]
- 3.Chow LQM, Cancer N. Head and neck cancer. N Engl J Med. 2020;382(1):60–72. 10.1056/NEJMra1715715. [DOI] [PubMed] [Google Scholar]
- 4.van der Laan HP, van der Schaaf A, Van den Bosch L, Korevaar EW, Steenbakkers RJHM, Both S, et al. Quality of life and toxicity guided treatment plan optimisation for head and neck cancer. Radiother Oncol. 2021;162:85–90. 10.1016/j.radonc.2021.06.035. [DOI] [PubMed] [Google Scholar]
- 5.Araújo ALD, Moraes MC, Pérez-de-Oliveira ME, Silva VMD, Saldivia-Siracusa C, Pedroso CM, et al. Machine learning for the prediction of toxicities from head and neck cancer treatment: a systematic review with meta-analysis. Oral Oncol. 2023;140:106386. 10.1016/j.oraloncology.2023.106386. [DOI] [PubMed] [Google Scholar]
- 6.Sher DJ, Moon DH, Vo D, Wang J, Chen L, Dohopolski M, et al. Efficacy and quality-of-life following involved nodal radiotherapy for head and neck squamous cell carcinoma: the INRT-AIR phase II clinical trial. Clin Cancer Res. 2023;29(17):3284–91. 10.1158/1078-0432.CCR-23-0334. [DOI] [PubMed] [Google Scholar]
- 7.Dickstein DR, Powers AE, Vujovic D, Roof S, Bakst RL. Clinical and therapeutic considerations for older adults with head and neck cancer. CIA. 2023;18:409–22. 10.2147/CIA.S366155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bras L, de Vries J, Festen S, Steenbakkers RJHM, Langendijk JA, Witjes MJH, et al. Frailty and restrictions in geriatric domains are associated with surgical complications but not with radiation-induced acute toxicity in head and neck cancer patients: a prospective study. Oral Oncol. 2021;118:105329. 10.1016/j.oraloncology.2021.105329. [DOI] [PubMed] [Google Scholar]
- 9.Mensah ABB, Mikare M, Mensah KB, Okyere J, Amaniampong E-M, Poku AA, et al. Impact of cancer diagnosis and treatment: a qualitative analysis of strains, resources and coping strategies among elderly patients in a rural setting in Ghana. BMC Geriatr. 2023;23(1):540. 10.1186/s12877-023-04248-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zimmermann C, Swami N, Krzyzanowska M, Leighl N, Rydall A, Rodin G, et al. Perceptions of palliative care among patients with advanced cancer and their caregivers. CMAJ Can Med Assoc J J Assoc Medicale Can. 2016;188(10):E217–27. 10.1503/cmaj.151171. [DOI] [PMC free article] [PubMed]
- 11.Derks W, De Leeuw RJ, Hordijk GJ, Winnubst JA. Quality of life in elderly patients with head and neck cancer one year after diagnosis. Head Neck. 2004;26(12):1045–52. 10.1002/hed.20113. [DOI] [PubMed] [Google Scholar]
- 12.Laraway DC, Lakshmiah R, Lowe D, Roe B, Rogers SN. Quality of life in older people with oral cancer. Br J Oral Maxillofac Surg. 2012;50(8):715–20. 10.1016/j.bjoms.2012.01.010. [DOI] [PubMed] [Google Scholar]
- 13.Van Der Schroeff MP, Derks W, Hordijk GJ, De Leeuw RJ. The effect of age on survival and quality of life in elderly head and neck cancer patients: a long-term prospective study. Eur Arch Otorhinolaryngol. 2007;264(4):415–22. 10.1007/s00405-006-0203-y. [DOI] [PubMed] [Google Scholar]
- 14.Bruijnen CP, de Groot LGR, Vondeling AM, de Bree R, van den Bos F, Witteveen PO, et al. Functional decline after surgery in older patients with head and neck cancer. Oral Oncol. 2021;123:105584. 10.1016/j.oraloncology.2021.105584. [DOI] [PubMed] [Google Scholar]
- 15.VanderWalde NA, Fleming M, Weiss J, Chera BS. Treatment of older patients with head and neck cancer: a review. The Oncologist. 2013;18(5):568–78. 10.1634/theoncologist.2012-0427. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Thomas CM, Sklar MC, Su J, Xu W, de Almeida JR, Gullane P, et al. Evaluation of older age and Frailty as factors associated with depression and postoperative decision regret in patients undergoing major head and neck surgery. JAMA Otolaryngol Head Neck Surg. 2019;145(12):1170. 10.1001/jamaoto.2019.3020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Furlong WJ, Feeny DH, Torrance GW, Barr RD. The health Utilities Index (HUI®) system for assessing health-related quality of life in clinical studies. Ann Med. 2001;33(5):375–84. 10.3109/07853890109002092. [DOI] [PubMed] [Google Scholar]
- 18.Ren J, Pang W, Hueniken K, Haddad G, Hope A, Huang SH, et al. Longitudinal health utility and symptom-toxicity trajectories in patients with head and neck cancers. Cancer. 2022;128(3):497–508. 10.1002/cncr.33936. [DOI] [PubMed] [Google Scholar]
- 19.Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Educ Chang Clin Epidemiol. 1994;47(11):1245–51. 10.1016/0895-4356(94)90129-5. [DOI] [PubMed] [Google Scholar]
- 20.Vinckier C, de Nooijer K, Smets T, Du Cheyne H, Decoster L, Van den Block L. Changes in quality of life throughout the illness trajectory of older adults with cancer: a systematic review. The Oncologist. 2025;30(9):oyaf223. 10.1093/oncolo/oyaf223. [DOI] [PMC free article] [PubMed]
- 21.Berg M, Silander E, Bove M, Johansson L, Nyman J, Hammerlid E. The effect of age on health-related quality of life for head and neck cancer patients up to 1 year after curative treatment. J Geriatr Oncol. 2022;13(1):60–66. 10.1016/j.jgo.2021.06.009. [DOI] [PubMed] [Google Scholar]
- 22.Lafont C, Paillaud E, Bertolus C, Baron M, Caillet P, Bouvard E, et al. Geriatric assessment and quality of life for 2 years in older patients with head and neck cancer. JNCI: J The Natl Cancer Inst. 2024;116(5):758–63. 10.1093/jnci/djae027. [DOI] [PubMed] [Google Scholar]
- 23.Moon DH, Chera BS, Deal AM, Wang Y, Muss HB, VanderWalde NA. Clinician-observed and patient-reported toxicities and their association with poor tolerance to therapy in older patients with head and neck or lung cancer treated with curative radiotherapy. J Geriatr Oncol. 2019;10(1):42–47. 10.1016/j.jgo.2018.05.016. [DOI] [PubMed] [Google Scholar]
- 24.Dinapoli L, Colloca G, Di Capua B, Valentini V. Psychological aspects to consider in breast cancer diagnosis and treatment. Curr Oncol Rep. 2021;23(3). 10.1007/s11912-021-01049-3. [DOI] [PMC free article] [PubMed]
- 25.Ginzac A, Dubois S, Hager M-O, Kwiatkowski F, Passildas J, Biau J, et al. Quality of life for older patients with cancer: a review of the evidence supporting melatonin use. Aging Clin Exp Res. 2020;32(12):2459–68. 10.1007/s40520-020-01532-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hjelmeland IHH, Drageset J, Nordvik Ø, Beisland EG. Quality of life in home-dwelling cancer patients aged 80 years and older: a systematic review. Health Qual Life Outcomes. 2022;20(1):154. 10.1186/s12955-022-02070-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.De Vries J, Bras L, Sidorenkov G, Festen S, Steenbakkers RJHM, Langendijk JA, et al. Association of deficits identified by geriatric assessment with deterioration of health-related quality of life in patients treated for head and neck cancer. JAMA Otolaryngol Head Neck Surg. 2021;147(12):1089. 10.1001/jamaoto.2021.2837. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kirkhus L, Šaltytė Benth J, Grønberg BH, Hjermstad MJ, Rostoft S, Harneshaug M, et al. Frailty identified by geriatric assessment is associated with poor functioning, high symptom burden and increased risk of physical decline in older cancer patients: prospective observational study. Palliat Med. 2019;33(3):312–22. 10.1177/0269216319825972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ahn S-H, Hong HJ, Kwon SY, Kwon KH, Roh J-L, Ryu J, et al. Guidelines for the surgical management of laryngeal cancer: Korean Society of Thyroid-head and neck surgery. Clin Exp Otorhinolaryngol. 2017;10(1):1–43. 10.21053/ceo.2016.01389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Baird BJ, Sung CK, Beadle BM, Divi V. Treatment of early-stage laryngeal cancer: a comparison of treatment options. Oral Oncol. 2018;87:8–16. 10.1016/j.oraloncology.2018.09.012. [DOI] [PubMed] [Google Scholar]
- 31.Parmar A, Macluskey M, Mc Goldrick N, Conway DI, Glenny A-M, Clarkson JE, et al. Interventions for the treatment of oral cavity and oropharyngeal cancer: chemotherapy. Cochrane Database Systematic Rev. 2021;2021(12):CD006386. 10.1002/14651858.CD006386.pub4. [DOI] [PMC free article] [PubMed]
- 32.Karlsson T, Tuomi L, Finizia C. Effect of voice rehabilitation following radiotherapy for laryngeal cancer – a 3-year follow-up of a randomised controlled trial. Acta Oncologica. 2022;61(3):349–56. 10.1080/0284186X.2021.1995891. [DOI] [PubMed] [Google Scholar]
- 33.Karlsson T, Tuomi L, Finizia C. Do effects of voice rehabilitation in patients irradiated for laryngeal cancer remain 5 years postradiotherapy? J Voice. 2024;S0892-1997(24)00074-2. 10.1016/j.jvoice.2024.03.002. [DOI] [PubMed]
- 34.Wolf F, Spoerl S, Gottsauner M, Klingelhöffer C, Spanier G, Kolbeck C, et al. Significance of site-specific radiation dose and technique for success of implant-based prosthetic rehabilitation in irradiated head and neck cancer patients—A cohort study. Clin Implant Dent Rel Res. 2021;23(3):444–55. 10.1111/cid.13005. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.