Brain Tumor Management in the Geriatric Patient

This chapter includes an accompanying lecture presentation that has been prepared by the author: .

Key Concepts

Life expectancy continues to rise exponentially. The concept of frailty, manifesting either phenotypically or in a deficit summation–based model, has merged as a tool helping to gauge overall health status and risk of adverse events in aging patients.
Meningioma is the second most common primary brain tumor. When considering treatment plans, clinicians must choose among gross total resection (GTR) or subtotal resection (STR), with associated risks; stereotactic radiosurgery (SRS); or a combined adaptive hybrid surgery approach, because progression rates after STR with no SRS are high.
SRS, less influenced by patient age, is gaining popularity, offering high rates of growth control with low incidence of neurological deficits as compared with other treatment modalities for meningiomas.
Several scoring systems have been constructed in an attempt to offer risk stratification in meningiomas. Five systems presented in this chapter show correlation with mortality and other outcome parameters.
The most comprehensive and most validated, the Geriatric Scoring System (GSS), validated for microsurgical resection and SRS, offers the opportunity for intervention, with potentially modifiable parameters.
The GSS, introduced in 2010 and based on a cohort of 250 patients, was validated in independent cohorts of elderly patients undergoing microsurgical resection and also in a separate independent cohort of patients who were treated with single-session SRS for intracranial meningioma, by several groups.
The incidence of glioblastoma multiforme (GBM) has been on the rise in recent decades, with the majority occurring in those >70 years. Advanced age and poor performance status are negative prognosticators. Genetic tumor markers serving as prognosticators and predictors, guiding treatment, differ in incidence and usefulness in elderly patients with GBM.
The optimal management of elderly patients with GBM is unclear because elderly participants have been consistently excluded from large controlled trials. Elderly patients with GBM are a heterogeneous population. Further insight into the molecular drivers of GBM in elderly patients coupled with an increased focus on a comprehensive geriatric assessment is mandated.
Elderly patients with a good functional status may benefit from an aggressive approach mirroring that in younger patients, whereas frail patients may be considered for less aggressive approaches such as hypofractionated radiotherapy or single-agent temozolomide.

Introduction

The average life expectancy has increased by as much as 30 years in developed countries during the last century, and continues to rise exponentially. Of the population, the subgroup of elderly people (≥60 years) has increased at the fastest pace ever (3.7% per year in 2010–15), and an increase of 2.9% per year until 2050 has been projected. The United Nations (UN) defines “aged people” as those >60 years, whereas the World Health Organization (WHO) sets the bar at ≥65 years. For the sake of discussion, we refer to elderly or “aging” patients in this chapter as those individuals aged ≥65 years. The “aging,” “aged” (≥70 years), and “super-aged” (≥80 years) societies have aging rates of 7% to 14%, 14% to 21%, and ≥21%, respectively. This growing geriatric population presents a growing challenge for physicians of all types, including neurosurgeons. Elderly patients are more vulnerable to stressors (attributable to either acute or chronic conditions) because of age-dependent decreased physiologic reserves. This in turn, raises the likelihood of adverse clinical consequences of these stressors. This elevated risk of adverse outcomes complicates treatment decisions, more so with invasive large-scale surgical interventions such as neurosurgery.

The concept of frailty has emerged in recent years as a clinical tool helping to gauge overall health status and risk of adverse events in aging patients. ^, Frailty, a measure of functional and physiologic vulnerability, exhibits a linear dose-response relationship with poor survival. Age, functional status, provider impression, or a combination of these factors have all been reported as measures to classify patients as frail. ^, Comorbid disease burden for frailty can be assessed with the Charlson Comorbidity In (CCI). Sheehan et al. and Michaud et al. have reported a positive association between body mass index (BMI) and frailty. Kolakshyapati et al. reported a correlation between frailty and both BMI and serum albumin. Isobe et al. reported a positive correlation between age, preoperative BMI, and serum albumin, as well as a skull base tumor location, as risk factors for deterioration of functional capacity (based on the Karnofsky performance scale [KPS]) at discharge and perioperative intracranial complications in a cohort of 265 elderly patients undergoing surgical resection of a meningioma. The literature on frailty among malignant glioma patients focuses on radiotherapy or chemotherapy rather than neurosurgical operative procedures. ^,

The pathophysiology underling the concept of frailty can be outlined using one of two commonly applied models. In the phenotype model, frailty manifests as “declines in lean body mass, strength, endurance, balance, walking performance, and low activity.” Accordingly, general reserves and stamina-related scores are used to assess the patient. In the deficit model (based on the Canadian Study of Health and Aging [CSHA]), the scope of frailty results from summing impairments and clinical deficits from a large range of symptoms and signs (capturing physiologic functioning, activities of daily living, mood disorders, and so on). ^, The Canadian Study of Health and Aging Modified Frailty Index (CSHA-mFI), validated for use with retrospective data, was shown to predict postoperative complications, morbidity, and morbidity after neurosurgical procedures (cranial or spinal). ^, The CSHA-mFI is constructed of 11 variables, with one point given for each variable present: (1) difficulty with activities of daily living; (2) history of diabetes mellitus; (3) lung or respiratory disease; (4) congestive heart failure; (5) myocardial infarction; (6) other cardiac disease; (7) arterial hypertension; (8) clouding, delirium, or cognitive impairment; (9) history of transient ischemic attack; (10) history of stroke; and (11) peripheral vascular disease.

We present a short discussion focusing on unique features of the elderly neurosurgical patient that merit specific clinical considerations. Two common neoplastic pathologic conditions will be presented and discussed in depth: benign meningioma and glioblastoma multiforme (GBM).

Meningioma

Introduction and Natural History in Elderly People

Meningioma, thought to originate from arachnoid cap cells, is the second most common primary brain tumor (even when postmortem incidental lesions are excluded) . The risk of developing a meningioma increases with age, dramatically so after the age of 65 years. Meningioma accounts for 13% to 33.8% of primary intracranial tumors, more so if postmortem findings are included. The increasing overall life expectancy, coupled with improved and available modern imaging techniques, results in neurosurgeons more frequently diagnosing incidental or asymptomatic intracranial meningioma in elderly patients. Rengachary and Suskind reported a 4.6% incidence of incidental meningiomas found at autopsy in patients >80 years old, not known prior to autopsy. Meningioma in the elderly patient has been defined by some authors as a separate clinical entity, accounting for the different cellular proliferation, vascularity, and intratumoral hormonal profiles noted. Understanding the natural history of meningioma discovered by chance (incidental finding) is crucial for appropriate decision making, and especially so in frail and elderly patients.

Tumor natural growth rates vary with age and reports. Nakamura et al. reported a series of 47 asymptomatic patients followed clinically and radiographically. In this series, 66% of lesions exhibited a growth rate of <1 cm ³ per year. A mean absolute annual growth rate of 0.796 cm ³ per year (0.03–2.62 cm ³ per year) and a mean relative annual growth rate of 14.6% (0.48%–72.8%) were reported. Chamoun et al. reported a moderate correlation between age and growth rates, with younger patients featuring a higher growth rate and shorter doubling time. Iwai et al. and Lee et al. reviewed the outcome of untreated meningiomas, reporting a 38% prevalence of lesions with growth rates higher than 4 mm per year when patients were followed for at least 24 months. Yano et al. reviewed a cohort of 603 asymptomatic meningiomas, reporting no tumor growth in 63% and a 6% prevalence of patients symptomatic for >5 years.

Studies use different formulas for calculating meningioma annual growth rates. Some authors have determined the lesion’s initial and final volumes during the follow-up period, assuming an exponential growth, whereas others have performed serial volumetric measurements. Oya et al. reported a ≥2-mm increase in maximum diameter in 44% ( n = 120) of lesions during a mean follow-up period of 45 months. Hashiba et al. reported volumetric follow-up in 70 patients. No tumor growth was noted in 37% ( n = 26). Of the growing lesions ( n = 44), 36.4% ( n = 16) showed an exponential pattern, 34.1% ( n = 15) showed a linear pattern, and the remaining 29.5% ( n = 13) did not fit any pattern. Thus, meningiomas may initially grow exponentially or linearly but then undergo a growth rate change. Such changes can possibly be due to changes in tumor-related available blood supply, calcifications, or acquisition of new mutations that inhibit or promote growth. Therefore, careful surveillance is recommended in these patients because some lesions remain clinically silent (and may not mandate intervention), whereas others warrant treatment. ^, Despite epidemiologic data, predicting a specific meningioma’s growth pattern and clinical behavior over time is virtually impossible.

Microsurgical resection has classically been referred to as the primary treatment option for intracranial meningiomas. Despite continuous improvement and advances in both operative techniques and equipment with related decrease in surgical morbidity reported in recent decades, a complete and safe resection (with minimal surgical morbidity and preservation of neurological function) is not always feasible. Perioperative morbidity in elderly patients undergoing a complete resection of meningioma has been reported to range from 11.8% to 52%. ^, The incidence rates of temporary and permanent cranial nerve deficits were reported to reach 44% and 56%, respectively. Surgical mortality in elderly patients has been reported to be as high as 45%. Some authors have noted that a skull-based lesion confers an even worse prognosis in the elderly patient population. ^, ^, On the other hand, incomplete resection of a lesion can lead to lower local tumor control rates, tumor progression or recurrence, morbidity, and mortality. ,

Most authors agree that a growing or symptomatic lesion merits consideration for a more aggressive approach. Still, when considering a treatment plan, one should factor in the patient’s life expectancy, expectations, frailty status, potential surgical risk, complications, and neurological sequelae. As a consequence, clinicians must at times choose between a complete gross total resection (GTR) with its substantial risk of morbidity and a subtotal resection (STR). Progression rates after partial removal of a meningioma with no stereotactic radiosurgery (SRS) or radiotherapy have been reported to be as high as 70%. Different studies have been conducted in attempts to determine treatment guidelines, risk stratification scoring systems, and roadmaps, as discussed in the following sections.

Surgical Resection

Which elderly patient will benefit from surgical resection of a meningioma (referring only to those lesions in which resection is technically feasible)? What patient profile or subpopulation will benefit from an aggressive approach in terms of overall physical and functional status? Different studies have been conducted in attempt to better define such patients and to set treatment guidelines. ^, ^, ^, Godfrey and Caird reported no difference in clinical presentation, progression of neurological deficits, or cognitive impairment in a cohort of 111 elderly patients compared with younger patients. They also reported marked clinical and neurological improvement with low operative morbidity and mortality on surgical intervention. Similarly, Drummond et al. in a review from Brigham and Women’s Hospital, reported no difference in postoperative survival between age groups after resection of a meningioma. Further validations were offered by Roser et al., who reported a cohort of 43 patients aged >70 years, analyzed and matched in a retrospective study to a cohort of 89 younger patients in terms of tumor size, histologic type, symptoms, recurrence, and so on. The authors reported no significant difference in functional outcome (KPS score) in different American Society of Anesthesiologists (ASA) grades. Similar incidence rates of postoperative deterioration of neurological function were noted (cranial nerves palsies; visual, sensory, or motor deficit) ( P = .035). Of note, the incidence of postoperative infections (meningitis, pneumonia, surgical wound infections) was significantly higher in the elderly group ( P < .05). Brokinkel et al. compared long-term prognosis and survival after meningioma surgery in elderly and younger patients in a cohort of 500 patients. The elderly subgroup (age ≥65 years, n = 162) was compared with the younger subgroup (age <65 years, n = 338), and no difference was reported. The authors concluded that elderly patients with surgically treated meningioma do not have impaired survival compared with the age-matched general population.

Ikawa et al. published a comprehensive review of evidence regarding surgery in elderly patients with meningioma, discussing 24 papers and a total of 9987 patients, segregated based on the definition of elderly age, evaluated during 1978–2013, and with follow-up durations ranging from 2 to 25 years. Cohort size in the different studies reviewed ranged from 21 to 5717, with 15 studies having small cohorts (<100 patients), 7 having medium-sized cohorts (101–258 patients), and 2 having large cohorts (>2000 patients). Four age categories were defined: >60, >65, >70, and >80 years. The largest age category (7 papers, 6607 patients, 66.5% female) was the group >65 years. All studies were retrospective. The overall average age was 75.5 years (60–92 years). The overall 1-year and 5-year mortality rates ranged from 0% to 16.7% and from 7% to 27%, respectively (comparable mortality rates with unselected cohorts).

Reviewing the different age categories, Ikawa et al. noted that for the patients >65 years (the largest group), >70 years (3131 patients), and >80 years (162 patients), the mean skull base–related location rates were 26.2%, 49.9%, and 35% respectively. The rates of mean tumor size >4 cm were 56.2%, 79.5%, and 48.7%, respectively. The mean incidence rates of no neurological deficit at presentation (pretreatment) were 24%, 45.7%, and 27.1%, respectively. The mean incidence rates of pretreatment KPS > 80 were 50.1%, 90%, and 51.6%, respectively. The mean frequencies of ASA III and IV were 40.7% and 3.6% for the group >65 years, 41.5% and 7.4% for the group >70 years, and 45.1% and 9.3% for the group >80 years, respectively. The mean posttreatment in-hospital 1-month, 3-month, 1-year, and 5-year mortality rates were 2.7%, 3.9%, 5.5%, 5.8%, and 12.9%, respectively, for the group >65 years; 2.0%, 4.3%, 5.0%, 8.7%, and 16.8%, respectively, for the group >70 years; and 2%, 5.4%, 6.9%, 13.8%, and 25.7%, respectively, for the group >80 years. The surgical and overall mortality rates were 7.0% and 1.2%, respectively, for the group >65 years; 2.0% and 7.9%, respectively, for the group >70 years group; and 7.0% and 1.2%, respectively, for the group >80 years.

Ikawa et al. reported recurrence rates ranging from 0% to 24.1% during follow-up durations of 2.8 to 15.6 years. Tumor local control after resection seems to be similar between elderly and younger patients. Still, elderly meningioma patients (all elderly subgroups as defined by Ikawa et al. and discussed previously) showed a higher frequency of atypical and anaplastic (WHO grade II or III) pathology. This histopathologic diagnosis served as the most serious challenge (based on the Central Brain Tumor Registry of the United States [CBTRUS]). This nonbenign histologic type is known to be influenced prognostically by age—that is, older age is an accepted negative prognosticator in patients with atypical meningioma, posing a significant risk for recurrence and shorter survival.

Stereotactic Radiosurgery

SRS is a well-established approach harnessed in past decades for treating up-front or as a complementary tool (e.g., adaptive hybrid surgery) for inaccessible, recurrent, or incompletely resected meningiomas. SRS has also been used in large inoperable lesions close to neurovascular critical structures. ^, Although recent operative results show improved morbidity and mortality rates in skull base surgery in elderly patients, SRS has been shown to be as effective as microsurgical resection in terms of local control. ^, The safety, efficacy, and patient-friendly design of SRS have been demonstrated in the control of benign tumors (more so for lesions <3 cm or <10 mL in volume), having distinct margins and sufficient distance from neurovascular critical structures to permit safe delivery of an effective target dose. The overall reported 5- and 10-year actuarial tumor control rates after SRS for benign WHO grade I lesions are 91% and 88%, respectively. ^, SRS is most frequently used in patients who have an adverse profile related to the lesion (e.g., location) and/or to the patient (e.g., overall health status, patient age, comorbidities).

Unlike with microsurgical series, numerous authors reviewing large independent SRS meningioma cohorts have repeatedly reported that age has not influenced prognosis in patients treated with SRS. ^, Fokas et al. reports 5-year rates of local control, overall survival (OS), and cause-specific survival to be 92.9%, 88.7%, and 97.2%, respectively. Over a decade of follow-up, these parameters appeared to change minimally with 5.4% and 14.6% decrease in local control and cause-specific survival (respectively) and no change in OS. The same group reported another cohort of 121 patients, treated with stereotactic-based radiotherapy and followed for a median time of 40 months (range, 12–124 months). Local control at 3 and 5 years was reported as 98.3% and 94.7%, respectively, and not influenced by age. Kondziolka et al. reported that long-term tumor control rates were sustained when reviewed ≥10 years after SRS. Sheehan et al. reported the incidence of perilesional edema and adverse radiation effects (AREs) around parasagittal or parafalcine meningiomas following SRS in a retrospective multicenter cohort of 212 patients through participating centers in the International Gamma Knife Research Foundation. Median age was 60 years (range, 18–90 years). Tumor location, tumor volume, venous sinus invasion, margin, and maximal dose were found to be significantly related to post-SRS edema in multivariate analysis. Age was not found to influence outcome.

Cohen-Inbar et al. reported in 2016 on the presentation, treatment, and long-term outcome of patients with benign skull base meningiomas after SRS in a cohort of 135 patients. The median age was 54 years (range, 19–80 years). The median follow-up was 102.5 months (range, 60.1–235.4 months). Tumor volume control was 88.1%. The 5-, 10-, and 15-year actuarial progression-free survival (PFS) rates were 100%, 95.4%, and 68.8%, respectively. Age was not found to influence outcome. Similarly, Cohen-Inbar et al. reported in 2018 on the therapeutic effect of SRS parameters, timing and volumetric changes, and long-term outcome parameters in a cohort of patients with benign WHO grade I parasellar meningiomas treated with SRS. A cohort of 189 patients, median age 54 years (range, 19–88 years) was analyzed. The median follow-up was 71 months (6–29 months). The lesion volume was computed by segmenting the tumor (slice by slice) with numerical integration using the trapezoidal rule. Tumor volume control was noted in 91.5% ( n = 173) and progression was documented in 8.5% ( n = 16), equally divided among in-field recurrences (4.2%, n = 8) and out-of-field recurrences (4.2%, n = 8). Prescription margin dose >16 Gy was noted to significantly influence PFS. Age was not found to influence outcome. Thus, SRS is a durable and minimally invasive treatment option for meningiomas of the skull base and other intracranial locations. SRS offers high rates of growth control with low incidence of neurological deficits. This treatment modality seems to be less influenced by patient age as compared with other treatment options.

Risk Stratification

What validated tools are available for predicting a favorable outcome for elderly patients harboring meningiomas treated with microsurgical resection, SRS, or both? The term favorable outcome serves as a composite outcome parameter, encompassing both tumor volume control and preservation of functional and neurological outcome. Several large studies have suggested that increasing age serves as a negative prognosticator in patients undergoing microsurgical resection of a meningioma, ^, ^, whereas other reports claim that clinical and radiologic features , ^, ^, ^, and functional status ^, ^, , ^, play a larger role as risk factors. Female sex (although being an unchangeable risk factor) has been associated with better prognosis. ^, ^, Of note, there are also risks associated with the conservative (observation, “wait and see”) approach in elderly patients that merit notice, as the potential possibility to improve health by waiting is less feasible in this patient age group. Arienta et al. reported that tumor-related mortality increases among patients managed conservatively compared with those who underwent resection.

Several scoring systems have been constructed in an attempt to offer risk stratification for surgical resection in elderly patients with meningiomas. Five grading systems are presented ^, , ^, ^, and are summarized in Tables 39.1 and 39.2 . All grading systems reviewed here ^, ^, ^, ^, showed correlations with mortality and other outcome parameters. The Clinical Radiological Grading System (CRGS), GSS, CLASS (comorbidities, location, age, size, and signs and symptoms) scale, and CCI do not consider patient sex, which is an unchangeable parameter and therefore less valuable for preoperative risk stratification modification. The SKALE (sex, KPS score, ASA grade, location, edema) grading system does not incorporate tumor size or preoperative neurological deficits and defines location subjectively (i.e., critical versus not critical; anatomic location). The CRGS, CLASS scale, SKALE grading system, GSS, and CCI all consider comorbidities, whereas the CCI does not incorporate any of the tumor’s radiologic features. A short description of the different grading systems, focusing on the more comprehensive and most validated—the GSS (see Table 39.1 ), validated for both microsurgical resection and SRS—is presented.

TABLE 39.1

The Geriatric Scoring System

The Geriatric Scoring System (GSS)				Outcome Parameters ( P values for GSS Score >16)
Parameter	1 point	2 points	3 points	Surgical Series
Size ^a	>5 cm (>62.5 cm ³ )	3–5 cm (13.5–62.5 cm ³ )	<3 cm (<13.5 cm ³ )		Survival 3 mo/overall survival	P = .001, P < .000 P < .0001, P = .003 ,
Neurological deficit	Progressive	Stable severe	None, minor		Barthel Index at 5 yr	Linear correlation , PC = 0.775
KPS	≤50	60–80	90–100		KPS at 5 yr	P = .004, P < .0001
Location	F, PS, FM	T, PF, JF	C, IV, SW, TS, CS, ON		GCS at 5 yr	P = .0001 PC = 0.627
Peritumoral edema	Severe	Mild	None		Recurrence: <1 yr, >1 yr	P = .002, P = .017 P = .007, P = .008
				Stereotactic Radiosurgery (SRS)
Diabetes mellitus	NC	MC	None		Post-SRS NP/NI	P < .0001
Essential hypertension	NC	MC	None		Post-SRS KPSi	P < .0001
Pulmonary disease	Severe	Mild	None		TVC	P = .028

C, Convexity; CS, cavernous sinus; F, falcine; FM, foramen magnum; GCS, Glasgow Coma Scale; IV, intraventricular; JF, jugular foramen; KPS, Karnofsky performance scale; KPSi, improvement in KPS at last follow-up; MD, medically controlled; NC, not controlled; NI, neurological improvement; NP, neurological preservation; ON, optic nerve; PC, Pearson correlation; PF, posterior fossa; PS, parasagittal; SW, sphenoid wing; T, tentorial; TS, tuberculum sellae; TVC, tumor volume control.

a Size expressed in maximal diameter cm, and converted to volume equivalent.

TABLE 39.2

Comparison of Meningioma Grading Systems

Score Component	CLASS	SKALE	CRGS	CCI	GSS
Tumor size	+	−	+	−	+
Tumor location	+	+	+	−	+
Peritumoral edema	−	+	+	−	+
Neurological condition	+	−	+	−	+
KPS	−	+	+	−	+
Sex	−	+	−	−	−
ASA grade	+	+	−	−	−
Comorbidities
Diabetes mellitus	−	−	−	+	+
Essential hypertension	−	−	−	+	+
Pulmonary disease	−	−	−	+	+
General/other disease	+	−	+	+	+

ASA, American Society of Anesthesiologists; CCI, Charlson Comorbidity Index; CLASS, comorbidities, location, age, size, and signs and symptoms; CRGS, Clinical Radiological Grading System; GSS, Geriatric Scoring System; KPS, Karnofsky performance scale; SKALE, sex, Karnofsky performance scale score, American Society of Anesthesiologists grade, location, edema.

CLASS Algorithm

Published in 2009 by Lee and Sade, the CLASS algorithm (i.e., comorbidities, location, age, size, and signs and symptoms) was aimed at balancing the risks and benefits of surgical resection. Comorbidities (C) were defined according to ASA definitions; location (L) was defined by the senior author as low, moderate, or high risk. These parameters and age were assigned a score ranging from −2 to 0. Additional factors defined were size (S) (with score ranging from 0 to 2 with increased volume) and signs and symptoms (S) (with score ranging from 0 to 2 with increasing severity; see Table 39.2 ). ^, The CLASS algorithm was not validated independently by other groups, possibly in part because of the subjective definitions of size, location, and comorbidities (defined with a general nonspecific scale). Such definitions do not offer reproducibility or a registry for improving a patient’s score by better controlling a chronic illness. Thus the CLASS algorithm provides no benefit in terms of risk stratification or preoperative optimization of patient status.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here