The Older Adult with Locoregionally Advanced Head and Neck Squamous Cell Carcinoma: Knowledge Gaps and Future Direction in Assessment and Treatment

Abstract

Older adults with squamous cell carcinomas of the head and neck (HNSCC) pose unique treatment and supportive care challenges to oncologists and other cancer care providers. The majority of patients with HNSCC present with locoregionally advanced (LA) disease, for which combined modality treatment integrating chemotherapy and radiation therapy is often necessary to maximize tumor control. However, applying these approaches to an older population with concomitant comorbidities and higher risk for functional impairments remains challenging, and is exacerbated by the paucity of studies involving older adults. The purpose of this paper is to identify knowledge gaps in the evaluation and management of older adults with HNSCC and their caregivers, particularly those undergoing concurrent chemoradiation, through a review of the literature conducted by clinicians, researchers, and patient advocates. The findings highlight the importance of a geriatric assessment and the therapeutic paradigms and challenges relevant to this population. Furthermore, we identify the need for additional research and interventions related to key supportive care issues that arise during and after treatment in older adults with LA-HNSCC. Based on our findings, we prioritized these issues to guide future patient-oriented research endeavors to address these knowledge gaps and thus better serve this growing patient population.

INTRODUCTION

Management of locoregionally advanced head and neck squamous cell carcinoma (LA-HNSCC) in older adults is a common, but challenging, clinical scenario for cancer providers. Maximizing the likelihood of tumor eradication in LA-HNSCC generally requires multimodality therapy.^1,2 However, toxicity from both cancer-related symptoms and the short- and late-term toxicities from treatment can be substantial in older adults, who more commonly have significant comorbidity burden, less robust nutritional status, decreased performance status, and impaired kidney, hepatic, hematologic, pulmonary, and cardiovascular function than younger patients. Older patients have been observed to have higher rates of severe late toxicity following concurrent chemoradiation,³ and higher risks of acute toxicity from cisplatin-based chemotherapy.⁴ Moreover, subgroup analyses of clinical trials have shown no survival benefit for therapeutic intensification strategies that have improved outcomes in LA-HNSCC like concomitant chemotherapy, concomitant cetuximab, or altered fractionation in older adults undergoing radiotherapy,^1,5,6 likely due treatment related morbidity and difficulty delivering full therapeutic doses without treatment interruptions.

Although HNSCC predominantly occurs in older patients, clinicians have very little data upon which to base their treatment decisions for this patient population, given older adults have often been under-represented or excluded altogether from HNSCC prospective trials.⁷ Moreover, given that the epidemiology of HNSCC in older patients has dramatically changed over the past decade, with rapid increases in oropharyngeal cancers, likely due to HPV-related malignancies, and concomitant decreased incidence of tobacco- and alcohol-related cancers arising from the oral cavity, larynx, and hypopharynx,⁸ the limited data available may be less applicable to older patients seen in current practice. This evidence vacuum, combined with projected increases in absolute number of older patients with HNSCC,⁹ creates a situation in which it is imperative to develop high level evidence to drive treatment paradigms for older patients.

Because aging is an inherently heterogeneous process, it is likely that some older patients would tolerate and benefit from standard-of-care multimodality therapy to a similar degree as their younger counterparts, while others will be more vulnerable to cancer treatment-related complications. However, identification of the subset of patients most likely to benefit from more aggressive treatment remains challenging in clinical practice. A comprehensive, geriatrics-oriented evaluation and treatment paradigm needs to be further studied and implemented for older adults with HNSCC (Figure 1). Additionally, it is possible that advances in radiation, surgery, and supportive care have improved the tolerability of multimodality therapy in older adults in the modern era. Thus, contemporary studies that are more representative and inclusive of older adults are needed to determine if treatment paradigms currently used in younger healthier patients will be equally beneficial for those patients who are older and/or more vulnerable to treatment-related toxicities. Therefore, Geriatric Oncology research will need to address the following questions for older adults with HNSCC.^10,11 Which older patients would benefit from “standard” cancer treatment approaches? Are there interventions that could improve the therapeutic ratio for both survival and quality of life (QOL) relative to the risks?

Figure 1.

Visual Framework for Treatment Decision-Making in Older Adults with Cancer

METHODS

The purpose of this review is to summarize current knowledge gaps for future research in three areas: 1) Geriatric assessment (GA) in the older patient with HNSCC and subsequent management based on the assessment findings; 2) Therapeutics for older adults with LA-HNSCC with a focus on Chemoradiation (CRT); and 3) Supportive care issues related to physical, neurocognitive, and psychosocial factors affecting older adults with LA-HNSCC. This review was carried out by the Cancer and Aging Research Group Head and Neck Cancer (CARG-HNC) group. CARG-HNC is a group of like-minded clinicians and scientists dedicated to improving the evidence base for the prevention, evaluation and treatment of older patients with head and neck cancer. The group includes radiation oncologists, medical oncologists, surgeons, geriatric medicine specialists, dentists, nurses, social workers, psychologists, and patient advocates interested in caring for older adults with head and neck cancer. The group has monthly conference calls with the goals of identifying knowledge gaps that exist for older patients with HNC and promoting multidisciplinary research. For this review, PubMed literature searches were performed by the authors (volunteer group members) searching for articles related to geriatric assessments, treatment of older patients with HNSCC, and supportive care for older patients with HNSCC from 1990 to November 2016. Although there continues to be debate on what chronological age is considered “old” or “elderly”, and despite the authors’ belief that age 70 is a more clinically significant cutoff, for the sake of this review, the authors considered studies that defined older age as 65 years or older, in order to not exclude significant research performed on this topic.

WHAT IS THE ROLE OF A GERIATRIC ASSESSMENT (GA) IN OLDER ADULTS WITH HNSCC?

A GA is a multidisciplinary evaluation of an older adults that commonly involves assessing domains beyond physical health alone, including, but not limited to, mental health, functional status, cognition, and mobility. The use of a GA in an outpatient non-oncology setting has been linked to outcomes such as improved functional status, fewer falls, and even improved mortality.^12–15 Only recently has the GA been introduced into the evaluation of older adults with cancer in order to integrate functional age into the decision-making process for cancer treatment. For example, impairments in certain domains, such as Instrumental Activities of Daily Living (IADLs), which include the ability to provide or arrange for one’s own transportation; use of the telephone; managing finances; shopping for and preparing meals; doing housework; and managing medications have been linked to worse overall survival, higher risk for hospitalization/longer length of stay, and post-operative morbidity.¹⁶ Similarly, the augmentative effects of several comorbid conditions, often referred to as multimorbidity, can have a significant impact on survival of older adults with cancer as well.¹⁷ A GA often incorporates and synthesizes evaluation of these health domains such as IADLs to attain a better assessment of an individual patient’s functional age or “level of fitness”.¹⁷ Some of the earliest studies in evaluating the ability of a GA to risk-stratify for likelihood of chemotherapy toxicity among older adults were performed by the Cancer and Aging Research Group (CARG)^18–20 and by the Moffitt group.²¹ These studies highlight that the findings obtained through GA prior to initiation of chemotherapy were more helpful in identifying patients at higher risk for chemotherapy-related adverse events than other commonly used measures in oncology practice such as performance status (Karnofsky or ECOG). In both studies, a modified risk model derived from a GA outperformed the “traditional” measure of performance status oncologists typically use in determining “fitness” for cancer therapy. Both of these risk models are now available as online calculators for practicing oncologists to incorporate into their clinical practice (CARG: http://www.mycarg.org/Chemo_Toxicity_Calculator; Moffitt: https://www.moffitt.org/eforms/crashscoreform). However, it is important to note that older adults with HNSCC were not included in or explicitly excluded from [i.e., those receiving definitive radiotherapy (RT)] these studies. As a result of this exclusion, the applicability of these results to older adults with HNSCC remains unclear.

Many experts advocate that older cancer patients should undergo a GA prior to cancer treatment (i.e. baseline GA).²² Unfortunately, the increase in this aging patient population, the growing shortage of geriatricians, and the time required for a comprehensive, multidisciplinary evaluation have made a formal GA impractical for the practicing oncologist. In response, several screening tools that could determine who were at “high-risk” thus warranting a subsequent in-depth GA have been published.²³ These tools tend to be shorter and consist of mostly patient-reported items, which can be easily completed by patients and/or their caregivers before or during clinic. Of all the screening tools, the G8 at baseline has shown fairly reliable correlation with GA findings.^24,25 The G8 includes 8 items, most of which are patient-reported: age; decreased oral intake; weight loss in past 3 months; mobility; neuropsychological problems; body mass index; taking 3 or more medications daily; and self-reported overall health. It was eventually adopted in a study of older adults age ≥65 years with mostly LA-HNSCC.^26,27 This study evaluated the G8 along with a second instrument, the Vulnerable Elders Survey-13 (VES-13), compared to the “gold standard” Comprehensive GA at baseline prior to RT or CRT. Details of this study and a side-by-side comparison of the CARG GA, G8, and VES-13 are illustrated in Table 1 and 2 respectively.^18,24,28 The G8 appeared to correlate better than the VES-13 with GA findings at baseline, and thus was deemed to be a good way to distinguish fit from vulnerable patients as a “screening tool”. Not surprisingly, despite development of GA domain impairments from week 0 to 4 of RT/CRT across all types (i.e. repeated GA measurements at week 4), the impact of treatment on HR-QOL domains were more pronounced in those deemed vulnerable at the start of treatment compared to those deemed fit. This relationship was maintained in their study with longer follow-up.²⁷

Table 1.

Reported or Published GA-Based Studies in Older Adults with Head and Neck Cancer

Completed Study	Sample Size	Age	PS	Comorbidity, Other Baseline Non-Cancer Factors	GA tool + domains	Primary HNC Site + Stage	Cancer Treatment	Outcomes Evaluated	Salient Study Findings
Neve et al. 201631	N=35	≥65 (median =74; range 65–93)	N/A	N/A	Baseline G8 & CGA; CGA domains not well delineated	9% OP 34% OC 3% HP 6% LX OTHER 29% cutaneous 6% salivary 14% other 63% Stage III–IV 91% curative intent	Definitive RT 4 Definitive CRT 1 Neoadjuvant/Adjuvant RT: 15 Surgery Alone: 25	Potential association of G8 or CGA with treatment-related outcomes	49% vulnerable by baseline G8 Trend toward longer post-op LOS with vulnerable G8 scores in surgical patients and fewer RT completion rates in the vulnerable groups. Limited by small sample size.
Pottel et al. 2014,26 201527	N=100 (96 evaluable)	≥65 (median 70, range 65–86)	N/A	None	Baseline and Week 4: G8 Screening + CGA: ADL, IADL, MNA, MMSE, GDS, Tinetti Gait & Balance, CIRS-G; QOL measured by EuroQOL-5 dimensions (EQ-5D)	15% OC 22% OP 10% HP 2% NP 33% LX 6% UP 69% Stage III–IVB	Definitive RT alone: 40% Definitive CRT: 29% (21% cisplatin, 8% cetuximab) Adjuvant: RT: 21% CRT: 10% (8% cisplatin, 2% cetuximab)	Serial changes in G8, CGA and association with survival including quality-adjusted survival using the EQ-5D	Baseline G8 abnormal: 68% (score ≤14) Baseline CGA abnormal: 72% (≥2 domain impairments) Week 4 G8 abnormal: 92% Week 4 CGA abnormal: 84% G8 Fit vs Vulnerable predicted worse OS, quality-adjusted survival along with trends lasting up to 36 months post-treatment
VanderWalde ASCO 2015 29	N=46	≥65 mean=72.5 (65–92)	Karnofsky PS ?	N/A	Baseline CARG GA:	61% SCCHN 39% Lung	54% RT alone 46% CRT	Evaluate CGA association with IADL scores and Poor Tolerance to Treatment (PTT): Hospitalization during treatment or 4–8 weeks post-tx, >3 day tx delay, any change to RT or chemo, or death from any cause.	Findings: Frailty defined by CGA or IADL did not predict PTT. Those receiving CRT had more PTT (57% vs 24%, 0.03) vs. RT alone and trend toward more G tube use (53 vs 21%, 0.05). Trend seen IADL impairment in SCCHN was associated with higher incidence of gastrostomy tube use (86 vs 45%, P=0.9).
Kwon 2016 32	N=165	>=65 (median =71, range 65–84)	ECOG PS: 0: 75 (45.5) 1: 77 (46.7) 2: 12 (7.3) 3: 1 (0.6)	CCI=0: 101 (61.2) CCI>=1: 64 (38.8)	CGA vs. develop frailty index Not available except online access; Factors In final model frailty: Age >=75 years; Smoking >= 20 pack-years; ECOG >=2; CCI>=1; +dental problems; Voice handicap index >=8; MDADI <70 (impaired self-reported swallowing); Beck depression inventory II >=14	Primary sites: 38% LX 24% HP 20% OC 18% OP I/II: 41% III–IVB: 58%	26% Surgery alone 27% surgery + adjuvant RT or CRT 19% definitive RT 16% definitive CRT 6% IC + RT or CRT 6% Other	2-year morbidity and mortality; median follow-up duration = 30.3 months. Morbidity mainly defined as any Grade >=3 adverse event requiring hospitalization (mostly aspiration/pneumonia, dysphagia, respiratory difficulty)	Low-risk (0–1) + intermediate risk (2–4); vs high risk (5–9) index was better discriminatory power predicting 2-year morbidity and mortality than fit vs frail from baseline CGA, more so than mortality (0.716 vs 0.043; 0.024 vs 0.001). overall 2-year mortality among 3 novel groups: 7% vs. 17% vs. 59%.

Abbreviations: ADL, activities of daily living; CCI, Charlson Comorbidity Index; CGA, comprehensive geriatric assessment; CARG: Cancer and Aging Research Group;CIRS-G, Cumulative Illness Rating Scale for Geriatrics; CRT, chemoradiation therapy; ECOG PS, Eastern Cooperative Oncology Group Performance Status; GDS, Geriatric Depression Scale; HP, hypopharynx; IADL, instrumental activities of daily living; IC, induction chemotherapy; LX, larynx; MDADI, MD Anderson Dysphagia Index; MMSE, Mini-Mental Status Examination; MNA, Mini-Nutritional Assessment; NP, nasopharynx; OC, oral cavity; OP, oropharynx; QOL, quality of life; RT, radiation therapy; SCCHN, squamous cell carcinoma of the head and neck; UP, unknown primary.

Table 2.

Comparison of select geriatric assessment tools

	CARG GERIATRIC ASSESSMENT (Hurria et al. 2005)18	G8 SCREENING TOOL (Soubeyran et al. 2014)24	VES-13 SCREENING TOOL (Saliba et al 2001)28
DOMAIN	YES/NO (SPECIFIC MEASURE IF ONE USED)
AGE	Y	Y	Y
COGNITIVE STATUS	Y (BOMC)	Y	N
NUTRITIONAL STATUS	Y (unintentional weight loss; BMI)	Y (decreased food intake, weight loss, BMI)	N
FUNCTIONAL STATUS	Y (IADL OARS; ADL MOS)	N	Y (modified IADL/ADL)
FALLS/MOBILITY	Y (NUMBER OF FALLS/6 months; TIMED UP AND GO*)	Y (bed/chair-bound, able to get out of bed but not out of house, goes out of house)	N
SELF-REPORTED OVERALL HEALTH	Y	Y	Y
SOCIAL STATUS	Y (MOS SOCIAL SUPPORT & SOCIAL ACTIVITY)	N	N
PHYSICAL HEALTH/COMORBIDITY	Y (OARS PHYSICAL HEALTH)	N	N
MEDICATION REVIEW (POLYPHARMACY)	Y	Y (taking more than 3 meds/day—yes/no)	N
MOOD	Y (GDS)	N	N

Abbreviations: ADL, activities of daily living; BMI, body-mass index; BOMC, Blessed Orientation-Memory-Concentration; GDS, Geriatric Depression Scale; IADL, instrumental activities of daily living; MOS, Medical Outcomes Survey; OARS, Older Americans Resources and Services;

Other smaller studies incorporating a GA or specific GA domains in the evaluation of older adults with LA-HNSCC undergoing RT or CRT have also been conducted. Two of these to highlight are those by VanderWalde et al. and Neve et al.^29–31 The former study included both non-small-cell lung cancer and HNSCC patients undergoing RT or CRT. Among the older adults with LA-HNSCC, a trend toward higher gastrostomy tube use was observed among those with more IADL impairment.²⁹ Similar to the findings of study by Pottel et al.,²⁷ they found that although all patients had decreases among QOL measures during RT, those with intact IADLs at baseline generally recovered back to baseline post-RT, whereas those with baseline IADL impairments continued to decline.³⁰ Neve et al. showed that the G8 screening tool demonstrated that patients with abnormal baseline G8 scores (“vulnerable”) had worse post-operative outcomes for those undergoing HNSCC surgery and lower RT completion rates for those receiving RT or CRT.³¹ Both of these studies demonstrate that an abbreviated GA tool can have predictive capacity for specific cancer-related outcomes. The limitations of these studies, including small sample sizes and inherent heterogeneity of cancer type, cancer stage, and treatment paradigms, make clinical application difficult. On the other hand, it is possible that larger studies, perhaps one including a more HNSCC-specific assessment (e.g., dysphagia, aspiration, enteral feeding tube use, presence of tracheostomy), would allow the development of risk stratification that can be employed by clinicians for therapeutic decision-making. For example, Kwon et al. has recently developed a novel HNSCC frailty index that incorporated some disease-specific items such as the M.D. Anderson Dysphagia Index (MDADI) that were more predictive of 2-year morbidity and mortality than a “traditional” GA at baseline.³²

There are several ongoing studies, mainly in Europe, which are utilizing a GA to stratify risk and look at several clinical outcomes besides survival (Table 3). The EGESOR trial (NCT02025062) is obtaining baseline GA prior to any HNSCC treatment for adults age ≥65 years.³³ This is one of the largest studies to date evaluating baseline GA and how it may be associated with IADL impairments and other geriatric-oriented outcomes over time following HNSCC treatment. The same GA is serving as the baseline for the three ELAN trials: FIT (NCT01864772); UNFIT (NCT01884623); and RT (NCT01864850). In the first two trials, the GA will risk-stratify patients with recurrent/metastatic disease as “fit” or “unfit” and assign them multi-agent or single-agent palliative-intent chemotherapy, respectively. In the RT trial, the GA will be utilized to discern differences in how older adults age ≥70 years tolerate different definitive RT schedules with a primary outcome of locoregional control. However, QOL and ADL status up to 18 months post-RT will also be evaluated. Although there is an older-adult-specific HNSCC trial open in the U.S. (NCT00904345), it does not incorporate a GA or enumerate geriatric-oriented outcomes. Its primary outcome is evaluating serial molecular changes in adults age ≥70 years or otherwise “unfit” for platinum-based CRT and thus will receive concurrent cetuximab-based CRT in addition to evaluating for locoregional control, survival, and toxicities.

Table 3.

Ongoing GA-Based Studies in Older Adults with Head and Neck Cancer

Ongoing Study	Sample Size	Age	PS	Comorbidity, Other Baseline Non-Cancer Factors	GA tool + domains	Primary HNC Site + Stage	Cancer Treatment	Outcomes Evaluated	Other Outcomes
EGESOR Brugel 2014	N=640; planned)	≥65	N/A	N/A	ADL IADL Falls in past 6 months 1-leg stand TUG MNA Weight loss/low BMI MMSE GDS CIRS-G ≥5 meds/day	Any SCCHNabout to undergo staging	Any	6-Month composite: Death, 2-point decrease in IADL from baseline, ≥10% decrease in baseline body wt. Secondary outcomes: each primary outcome individually; PFS; in-hospital death; unplanned hospitalization; post-surgery length-of-stay; discharge to home vs. nursing home; final cancer treatment plan; QOL (EORTC QLQ-HN35; EORTC QLQ-C30); CTCAE for chemotherapy toxicity; Costs; Treatment feasibility; Measured at 6 mos., 12 mos., 24 mos.	Other: Evaluate CGA-based interventions including social work referral, neuropsychiatric evaluation/treatment, etc.
ELAN trials: ELAN UNFIT	N=164	≥70	ECOG PS 0–2	N/A	Baseline CGA ((see EGESOR))	Recurrent/Metastatic SCCHN: OC, OP, HP, LX	Cetuximab vs. Methotrexate	Primary outcome: Failure Free Survival: time to progression, death, treatment discontinuation, ≥2 point loss in IADL by 16 months	Secondary: OS, PFS
ELAN FIT	N=82	≥70	ECOG PS 0–1	N/A	Baseline CGA	Same as above	Carboplatin + 5-Fluorouracil+Cetuximab→ Cetuximab maintenance	Primary: Objective Tumor Response at 12 weeks and Lack of IADL (2 point loss) at 1 month after chemo completion	Secondary: Best tumor response at 6 weeks post-treatment OS PFS Duration of Response During Cetuximab Maintenance Toxicities ADL/IADL QOL (EORTC QLQ-C30, QLQ-HN35)
ELAN RT	N=202	≥70	ECOG PS 0–1	N/A	Baseline CGA	Stages II–IVB Unresectable/inoperable Any SCCHN type Cannot receive any chemotherapy/systemic therapy	70 Gy in 35 fx vs. 30 Gy in 10 fx, 2 week break, 25 Gy in 10 fx	Primary: Locoregional control 6 months post-RT	Secondary: Treatment toxicities QOL Locoregional control and Metastatic progression OS, PFS ADL

Abbreviations: ADL, activities of daily living; BMI, body-mass index; CCI, Charlson Comorbidity Index; CGA, comprehensive geriatric assessment; CIRS-G, Cumulative Illness Rating Scale for Geriatrics; CRT, chemoradiation therapy; CTCAE, Common Terminology Criteria for Adverse Events; ECOG PS, Eastern Cooperative Oncology Group Performance Status; EORTC QLQ, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire; fx, fractions; GDS, Geriatric Depression Scale; HP, hypopharynx; IADL, instrumental activities of daily living; IC, induction chemotherapy; LX, larynx; MMSE, Mini-Mental Status Examination; MNA, Mini-Nutritional Assessment; NP, nasopharynx; OC, oral cavity; OP, oropharynx; OS, overall survival; PFS, progression-free survival; QOL, quality of life; RT, radiation therapy; SCCHN, squamous cell carcinoma of the head and neck; TUG, Timed Up and Go; UP, unknown primary.

Other older adult-specific HNSCC trials evaluating CRT with novel therapies, such as immunotherapeutic agents, are in various stages of development among U.S. and European cooperative groups, but the exact status of GA and/or geriatric-oriented outcomes in these study concepts and protocols remains to be seen. However, the GA may be beneficial for more than just treatment decision-making. When linked to geriatric-based interventions in non-cancer populations, the GA-guided medical decision-making has been shown to decrease rates of mortality, nursing home admissions, hospitalizations, and assists older adults to maintain better physical and mental function.^34,35 Unfortunately, in patients with cancer, there are limited data to support GA-based interventions in busy oncology clinics.³⁶ That being said, some of these interventions are simple in nature and are used by many clinicians treating HNSCC and other cancer types already. A panel of experts in Geriatric Oncology, using a Delphi consensus technique, published recommendations on GA-guided interventions among older adults with cancer.³⁷ These interventions include: referrals to physical therapy or home safety inspections for difficulties with physical function or gait instability; consultations with nutritionists for patients with significant weight loss prior to initiation of therapy (already performed in many oncology centers); social work interventions for patients with poor social support or transportation issues; psychiatric referrals for patients with anxiety or depression; and neuropsychological consults for those with evidence of dementia or cognitive decline.³⁷ Many of these interventions would take little time on the part of the oncology team and could significantly impact patient-oriented outcomes. Yet, these interventions need to be studied in older adults with cancer, specifically in those with HNSCC cancer in a prospective manner. In doing so, it may be that the knowledge gaps that exist between younger and older HNSCC patients will begin to decrease. Based on these findings, we propose the following priorities for further clinical research regarding the role of the GA in older adults with LA-HNSCC:

• Compare existing GAs used in the U.S. with other tools such as G8 in a prospective fashion and in association with cancer-specific as well as geriatric-oriented outcomes

• Examine the impact of GA findings on the utilization of healthcare resources during CRT for older vs. younger, frail vs. fit patients undergoing CRT/combined modality therapy

• Evaluate the competing non-cancer risks associated with older adults identified by GA and their impact upon outcomes beyond survival.

• Evaluate the relationship between the GA and geriatric-specific outcomes:

  • Neurocognitive and psychological changes during and after CRT/combined modality therapy

  • Musculoskeletal health (e.g., sarcopenia) and related changes during and after CRT/combined modality and its impact on functional independence vs. disability

  • Swallowing and nutritional changes with a focus on enteral feeding utilization and dependence during and after CRT/combined modality therapy

WHAT ARE THE THERAPEUTIC CHALLENGES AND CONSIDERATIONS FOR OLDER ADULTS WITH LA- HNSCC?

The majority of patients with HNSCC present with LA disease (i.e., Stages III-IVB).³⁸ Although LA-HNSCC is often curable, it requires combined modality therapy, including a combination of surgery, radiation, and/or chemotherapy, to maximize the chance of disease eradication while concurrently attempting to minimize toxicities. Each of these modalities is uniquely difficult to deliver in older adults. Older patients are more likely to have cardiovascular and pulmonary comorbidities, making some poor surgical candidates due to anesthesia risk. Older patient may have higher risk of aspiration and less robust nutritional status, complicating radiation delivery. Older adults may have lower baseline hematologic, renal, and hepatic reserve, potentially increasing vulnerability to chemotherapy toxicity. However, assuming a patient is medically fit to undergo surgery and/or radiation, the main therapeutic decision for oncologists is typically related to utilization of therapeutic intensification strategies that have been shown to improve survival by decreasing the risk of locoregional recurrence in HNSCC, including concomitant chemotherapy, cetuximab, and altered fractionation of RT dosing.^1,2,6,39,40 None of these strategies have shown any survival benefit in subgroups of patients older than 70 years.⁵ Although it is possible that these intensification strategies are less efficacious at eradicating tumors in older adults due to biologic and micro-environmental differences in head and neck cancers arising in this patient population, the most likely explanation is that the salutary tumor control effects of these intense regimens are offset by treatment-related toxicities and competing non-cancer causes of mortality.^41,42 Older adults with cancer such as HNSCC are also likely to have concomitant comorbidities and age-related pharmacodynamics that impact cancer drug metabolism.^4,43,44 Older age has been demonstrated to strongly predict increased late-term treatment-related toxicities in HNSCC,³ including aspiration pneumonia, dysphagia, and G-tube dependence following CRT,⁴⁵ as well as independently predict for non-cancer associated mortality following CRT.⁴⁶

However, it is important to frame these findings within a “gero-centric” context. The results from the MACH-NC and MARCH meta-analyses showing patients older than 70 experienced no benefit from either CRT or altered fractionation, respectively, were comprised of trials initiated 20 to 50 years ago. As a result, these trials enrolled predominantly patients with HPV-negative HNSCC, who were likely life-long tobacco and alcohol users and, in turn, had a higher comorbidity burden. These patients have been reported to be at higher risk of late toxicities following CRT.⁴⁷ Even the somewhat more recent IMCL-9815 trial, demonstrating no benefit of adding cetuximab to RT alone in patients older than 65 years, was initiated nearly two decades ago and enrolled predominantly HPV-negative HNSCC.^5,48 In the U.S. over the past decade, the incidence of tobacco- and alcohol-related HNSCC has been steadily decreasing among patients age 65 and older, while the incidence of oropharyngeal squamous cell (OPSCC), which is often HPV-related,⁴⁹ is increasing at least as quickly as what is being observed in younger patients.⁸ Thus, older adults with HNSCC in modern practice may represent a very different population than those enrolled on historical trials, with more HPV-related tumors and lower comorbidity burden. Additionally, the trials comprising the MACH-NC and MARCH meta-analyses universally utilized older non-conformal radiotherapy techniques. Intensity-modulated radiotherapy (IMRT), the current standard of care for HNSCC treatment, has been shown to decrease toxicity in comparison to less conformal radiotherapy platforms in multiple phase III trials.^50–52 Thus, IMRT may potentially lessen adverse sequelae in older adults undergoing multimodality therapy. Given the radical changes in the epidemiology of HNSCC in developed countries, the shift in causative risk factors, and improvements in RT delivery, it is possible that the therapeutic ratio for treatment intensification in older adults with HNSCC is somewhat more favorable in the modern era.^53,54

Despite these advances, even now, the toxicity of combined modality therapy such as concurrent CRT remains a significant challenge for older adults with HNSCC. For example, a recent SEER-Medicare study showed that from 2000 to 2009, 62% of patients age 66 or older receiving CRT required either an emergency room visit or hospitalization during their treatment, compared to 46% of patients receiving RT alone.⁵⁵ Furthermore, patients receiving CRT were three times as likely to have gastrointestinal or hematologic toxicity, and twice as likely to have dehydration, fever, or oral complications as patients receiving RT alone. Among these patients, feeding tubes were placed in 74% of those receiving CRT at any time after diagnosis, and were still used for nutrition in 15% at 1 year after treatment, compared to 50% and 6%, respectively, of patients receiving RT alone. These results highlight the need for improved supportive care and more evidence-based evaluation and treatment strategies for older patients with LA-HNSCC.

In addition to these studies, there are multiple non-comparative retrospective studies that have suggested that CRT can be feasible in older patients, albeit with higher toxicity in some cases.^7,56 Taken together, these findings underscore the fact that older adults with cancer comprise an extraordinarily heterogeneous group of patients, made of a range of very fit to very frail individuals with disparate performance status, functional status, and comorbidity burden. In fact, a recent study from the National Cancer Data Base including patients diagnosed between 1998 and 2011 demonstrated that for patients aged 71 and older, there was improved survival with concomitant CRT in those aged 71 to 80 with low comorbidity scores, but not in those with multiple comorbidities or over the age of 80. Of note, these results were very different from a SEER-Medicare study that found CRT was associated with increased risk of death in patients aged 66 years or older, with no benefit for any subgroup of patients.⁵⁷ These disparate results could have been in part due to the fact that the SEER-Medicare study included a cohort of patients from a somewhat earlier era than the National Cancer Data Base study.

Despite the lack of high level evidence supporting the use of therapeutic intensification approaches in older patients and conflicting retrospective database studies, the use of CRT for patients aged 65 and older with HNSCC has increased from 49% in 2000 to 82% in 2009 in the United States.⁵⁵ This was predominantly driven by increased use of cetuximab, which accounted for approximately 50% of concomitant systemic therapy regimens in 2009.⁵⁸ However, the rising use of cetuximab in the older adult population should be scrutinized closely. As noted previously, the only trial showing a benefit with cetuximab in combination with RT demonstrated no improvement in survival in patients 65 or older.⁵ Furthermore, recent phase II randomized comparisons of EGFR inhibitors and cisplatin have shown similar or worse radiation-related toxicity and strong trends towards poorer locoregional control with cetuximab.^59,60 The main benefit of cetuximab is decreased renal, auditory, and hematologic toxicity inherent to cisplatin, which can be of critical importance in older adults. However, given the uncertainties surrounding the efficacy of cetuximab, its widespread use among older adults, its high cost relative to other available systemic agents, and the fact that its toxicity profile is potentially less favorable based on these more recent studies than had initially been reported, the role of cetuximab specifically for older patients with LA-HNSCC warrants more robust, prospective evaluation.

Given the increasing incidence of OPSCC among older adults, presumably driven by HPV-associated pathogenesis, it is possible that the de-intensification approaches being investigated in low-risk HPV-related LA-OPSCC may be particularly beneficial for older adults and worthy of further investigation.^61–63 However, de-intensification in older adults with LA-OPSCC must be approached with great caution, given that historically this group has had vastly inferior disease-specific survival outcomes compared to younger adults even when controlling for other factors.^8,64 Other strategies for improving outcomes in older patients with HNSCC include improving patient selection for intense multimodality therapy (i.e., the role of the GA); intense and early supportive care measures including rehabilitation; and investigating and implementing both evaluative and therapeutic algorithms specific to this unique population, including regimens using RT alone for select individuals. Based on the review of these findings, we propose the following priorities for therapeutic studies in older adults with LA-HNSCC:

• Examine and develop geriatrics-based assessment tools to help predict which older adult patients benefit from CRT/combined modality therapy

• Determine the optimal systemic therapy (e.g. high dose cisplatin, weekly cisplatin, carboplatin doublets, carboplatin alone, cetuximab alone, etc.) for fit older patients

• Examine the benefit of cetuximab-based CRT in comparison to RT alone or other treatment paradigms in patients deemed ineligible for cisplatin

• Evaluate and determine optimal de-intensification strategies for older adults with HPV-associated OPSCC

• Systematically evaluate the role of combined modality immunotherapy in older patients with HNSCC

• Increase the evidence base and facilitate the implementation of supportive care strategies that could improve the therapeutic ratio in older adults receiving CRT/combined modality therapy

WHAT ARE SOME KEY SUPPORTIVE CARE ISSUES FACING OLDER ADULTS WITH LA-HNSCC?

Although many complex layers of short- and late-term toxicities of CRT for LA-HNSCC warranting interdisciplinary attention and intervention exist, we sought to emphasize several issues that may be more salient in our older adults with LA-HNSCC. Firstly, fatigue is one of the most common symptoms experienced by patients undergoing chemotherapy or radiation therapy for head and neck cancer.^65,66 Moderate to severe levels of fatigue occur in up to 60% of patients during RT and even greater levels of severity in those receiving concomitant chemotherapy.⁶⁶ RT-related fatigue peaks at the completion of treatment and may never return to baseline in some patients.⁶⁵ Thus, fatigue can be a long-term issue for many LA-HNSCC patients. Additionally, fatigue may be highly correlated with overall QOL during LA-HNSCC treatment.^67,68 Despite the multifactorial nature of fatigue (impacted by other factors such as anemia, poor nutritional intake, and depression), it can be driven by RT itself, even when treating anatomic sites that are less likely to confer additional treatment-related toxicities.^69,70 Although the etiology of RT-related fatigue is unknown, data support an association between pro-inflammatory cytokines including TNF-α, IL-1β, and IL-6 activity and post-treatment fatigue in breast, prostate, and head and neck cancer.^71–74 These pro-inflammatory cytokines are of particular interest in older adults with LA-HNSCC as they have also demonstrated an association with disease and disability,⁷⁵ and serum levels of these biomarkers and their associated symptoms decrease with exercise in older adults.^76,77

Fatigue experienced by older adult with LA-HNSCC can be compounded by sarcopenia, another common treatment effect. Sarcopenia is usually defined as the loss of muscle mass due to aging-related changes in physiology and metabolism.⁷⁸ Cachexia is weight loss including both muscle and adipose tissue due to pathologic pro-catabolic, inflammatory conditions such as cancer and other chronic illnesses, which can further exacerbate sarcopenia, thereby making it difficult to tease them apart clinically.⁷⁹ Sarcopenia coupled with advanced age, comorbidity, and fatigue commonly seen in older adults with cancer can induce a frailty phenotype.⁸⁰ Frailty, generally defined as an acquired loss of physiologic reserve, has been described as a geriatric syndrome that overlaps with sarcopenia and these other factors, which have been intimately associated with a higher risk for adverse outcomes including falls, disability, hospitalization, and mortality.^81,82

Early studies on the impact of sarcopenia/cachexia on cancer outcomes have emanated from the surgical oncology realm. Baseline sarcopenia has been linked to worse survival and worse post-surgical outcomes in cancer patients undergoing cancer surgery for various gastrointestinal cancers.^83–85 More recently, it has been linked to higher chemotherapy toxicity rates and hospitalizations during treatment for patients with metastatic breast and pancreatic cancer.^86,87 This treatment-related frailty has been associated with worse physical performance measures over the course of treatment and higher risk for falls in this patient population. Exercise-based interventions may offset some of the impact of this decline.⁸⁸

In patients with LA-HNSCC, sarcopenia (as defined by CT imaging-based criteria) is more prevalent, with approximately 35–60% and 80–85% of patients having it at baseline or developing it by the end of definitive RT/CRT, respectively.⁸⁹ Sarcopenia in this population has been associated with lower locoregional control and cancer-specific survival.⁸⁹ As a result, sarcopenia can potentially have a significant impact upon HNSCC survivors’ quality of life, function, and adherence to therapy. Thus, sarcopenia, coupled with the cancer- and treatment-related mucositis, dysphagia, pain, and overall impaired nutrition, often requires enteral nutritional support among patients with LA-HNSCC, particularly LA-OPSCC because of the anatomic factors conferred by these tumors. It is not uncommon for adults with LA-HNSCC to lose 5–10% of their pre-treatment body weight after CRT/combined modality therapy.⁹⁰ However, loss of specific muscle mass may not directly correlate with total body weight loss or need for enteral nutritional support for patients with LA-OPSCC.⁹⁰ Mucositis severity and thus need for nutritional support have also been associated with non-therapeutic factors such as degree of weight loss.⁹⁰

Age-related dysphagia, further magnifies these impairments and can be a potential source for long-term swallowing issues, nutritional compromise, and potential aspiration risk. Speech and swallowing as assessed by self-report measures appear to return to baseline levels more commonly and more rapidly in younger versus older patients with HNSCC receiving definitive RT.⁹¹ Additionally, higher rates of enteral feeding tube use with longer-term follow-up following CRT for LA-HNSCC are observed among adults age ≥70 years.^92,93 There are limited data evaluating and comparing methods of enteral nutritional support [i.e., nasogastric (NG) tube versus percutaneous gastrostomy (PEG) tube placement] in terms of relative safety and efficacy in mitigating more significant weight loss or long-term dysphagia in patients with HNSCC regardless of age.^94–96 A prospective trial evaluating NG vs PEG tube placement in patients undergoing RT/CRT (median age = 60 years) showed less weight loss with PEG placement but higher duration of use/dependence, higher Grade 3 dysphagia at 6 months post-treatment, and higher costs and complication rates.⁹⁷. Reactive gastrostomy tube placement (i.e., only when significant weight loss and/or nutritional compromise becomes manifest during the course of treatment) versus prophylactic placement (i.e., placement prior to treatment initiation) remains controversial.^98–100 Therefore, the optimal timing for initiating enteral feeding support and the route of delivery remain unclear for older patients with LA-HNSCC undergoing CRT who are more susceptible to nutritional and swallowing issues. Due to these intertwined challenges in supportive care facing these patients, prospective studies evaluating the relationships among age, swallowing, nutrition, and sarcopenia and longer-term functional outcomes among younger vs. older patients with LA-HNSCC are greatly needed. Among these physical health issues, the role of the caregiver will require closer examination as managing feeding tubes, for example, can be a source of great distress for both HNSCC patients and their caregivers.^101,102

In addition to the physical health issues affecting older adults with HNSCC, cancer and its treatment can also adversely impact neurocognitive (e.g., memory, executive function, attention, processing speed)^103–105 and psychosocial (e.g., depression, anxiety, self-esteem, body image)^106–108 function. Research on the neurocognitive and psychosocial impact of cancer in patients with HNSCC is more limited than in other cancer types. Yet, research suggests that patients with HNSCC also suffer declines in neurocognitive and psychosocial function, particularly over the course of treatment. Prior to treatment initiation, approximately one-third of patients with HNSCC may suffer impairments in verbal learning, executive function, verbal memory, and processing speed.^109,110 Cancer treatment may exacerbate these problems as the negative impact of RT-based therapy on neurocognitive function has been well documented.¹¹¹ Patients experience declines in short-term memory, language abilities, and verbal fluency after RT-based therapy with higher RT doses associated with greater cognitive decline.^{103,112–114} These changes have significant implications for patients’ ability to engage in daily activities, such as their ability to drive.¹¹⁵ To date, studies of neurocognitive function in patients with HNSCC have been conducted in relatively young patients (average age <60 years). Thus, the additional impact of aging on neurocognitive function in patients with HNSCC is not well known.

Patients with HNSCC also experience impaired psychosocial well-being. Prior to initiating radiation therapy, 10–20% of patients report moderate to severe sadness, drowsiness, pain, distress, disturbed sleep, and fatigue with higher rates in patients who received previous treatment.¹¹⁶ In an analysis of 3,533 older adults with HNSCC, 10.6% were diagnosed with depression in the two years prior to cancer diagnosis and an additional 8.9% were diagnosed in the year following diagnosis.¹¹⁷ Over the course of RT, patients report increases in depression and anxiety,^118–122 pain,¹¹⁸ and reductions in sleep quality¹²² and QOL.^123,124 In a sample of HNSCC patients age ≥50 years, a diagnosis of depression was associated with 28% increase in the likelihood of hospital admission following an emergency department visit for men and a 31% increase for women,¹²⁵ suggesting that psychosocial distress may have implications for medical treatment and healthcare utilization.

HNSCC patients also report experiencing interpersonal stressors, uncertainty, interference of their disease and treatment on life activities,¹²⁶ and body image concerns.^127,128 Additionally, in the patient discussed in our case, the fact that he also acts as a caregiver likely will cause increased distress. In caregivers of HNSCC patients, distress and burden increase during their loved ones’ treatment,^119,129 and health and QOL decrease¹³⁰ due to worry, fatigue, sleep, dealing with the patient, and work.¹¹⁹ These stressors cannot be ignored in light of evidence that HNSCC patients are at increased risk for suicide relative to the general population.¹³¹ Despite the proportion of HNSCC patients who are older adults, few studies on the neurocognitive and psychosocial sequelae of treatment focus specifically on older adults who may already be at risk for impairment in these areas. While older adults are included in the samples of current studies, the unique impact of HNSCC and its treatment on the neurocognitive and psychosocial function of older adults has not been widely considered.

Based on this review, we suggest the following priorities for supportive care and survivorship research for older patients with LA-HNCC undergoing CRT/combined modality therapy:

• Using both patient-reported outcomes and objective-based measures, evaluate the differences in the acute and late-term treatment toxicities experienced by older vs. younger patients by outcomes other than survival: physical health, functional outcomes, quality of life, neurocognition, psychological health, and caregiver support

• Examine and develop more evidence-based exercise, nutrition, and other strategies and interventions to mitigate fatigue, sarcopenia, and nutritional compromise during active treatment and during the recovery process.

• Evaluate the impact of acute and late-term treatment toxicities on patient caregivers in order to inform interventions tailored to their specific needs and prevent caregiver burn-out

CONCLUSION

As our population ages, the proportion of older adults with HNSCC will continue to grow. Many of the patients we see in the clinic would either not have been eligible for or would have been poorly accrued to the studies laying the groundwork for what are now considered “standard-of-care” treatment approaches for LA-HNSCC. Thus, we cannot assume that this evidence base can be applied to our older adults with LA-HNSCC. As highlighted above, significant progress has been made with respect to developing full and screening GAs to predict tolerance to therapy in older patients with cancer. Yet, there are significant knowledge gaps for HNSCC specific populations and GA based interventions. Questions still exist regarding how to incorporate GA into clinical practice and the ideal multimodality therapy for older or frail patients with HNSCC. Additionally, geriatric specific supportive care research is lacking. It is imperative that we improve our knowledge base for assessment, treatment decision-making, and supportive care for these vulnerable patients, many of whom will become long-term survivors. It is the goal of the CARG-HNC group to bring together like-minded researchers, clinicians, and patient advocates to develop studies to better address these knowledge gaps. In this era of precision medicine, just as research is evolving to guide treatment decision-making for targeted therapy based on the molecular profile of an individual’s cancer, we need to expand the definition of personalized medicine to include an individual older adult’s baseline functional status to tailor their treatment and supportive care planning.