Frailty in Patients With IDH-Mutant Gliomas: Experience from a High-Volume Tumor Center

doi:10.21203/rs.3.rs-4087976/v1

Download PDF

Research Article

Frailty in Patients With IDH-Mutant Gliomas: Experience from a High-Volume Tumor Center

https://doi.org/10.21203/rs.3.rs-4087976/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 04 Jun, 2024

Read the published version in Journal of Neuro-Oncology →

You are reading this latest preprint version

Purpose

Gliomas are increasingly diagnosed in an aging population, with treatment outcomes influenced by factors like tumor genetics and patient frailty. This study focused on IDH-mutant gliomas and assessed how frailty affects 30-day readmission and overall survival (OS). We aimed to address a gap in understanding the impact of frailty on this specific glioma subtype.

Methods

136 patients with an IDH mutant glioma between 2007 and 2021 were identified at our institution. High frailty was classified by scores ≥ 1 on the 5-factor modified frailty index (mFI-5) and ≥ 3 on the Charlson Comorbidity Index (CCI). Patient and tumor characteristics including age, sex, race, Karnofsky Performance Status (KPS), Body Mass Index (BMI), tumor type and location, type of operation, and therapy course were recorded. Outcomes measured included 30-day readmission and overall survival (OS). Analysis was conducted utilizing logistic regression and Kaplan–Meier curves.

Results

Of the 136 patients, 52 (38%) had high frailty: 18 with CCI ≥ 3, 34 with mFI-5 ≥ 1. High frailty correlated with increased BMI (CCI: 30.2, mFI-5: 30.1 kg/m2), more neurological deficits (CCI: 61%, mFI-5: 56%), and older age at surgery (CCI: 63, mFI-5: 48 years). Hospital readmission within 30 days occurred in 8 (5.9%) patients. Logistic regression indicated no significant difference in 30-day readmission rates (CCI: p = 0.30, mFI-5: p = 0.62) or median OS between high and low frailty groups. However, patients treated at our institution with newly diagnosed tumors with high mFI-5 had a 6.79 times higher adjusted death hazard than those with low mFI-5 (p = .049).

Conclusion

Our analysis revealed that CCI and mFI-5 were not significantly associated with 30-day nor OS. However, in patients with non-recurrent tumors, there was a significant association of mFI-5 with OS. Further study of frailty with larger cohorts is warranted to enhance prognostication of outcome after neurosurgical treatment.

IDH-mutant Gliomas

Frailty Indices

Hospital Readmission

Overall Survival

Gliomas are a common diagnosis within neuro-oncology, and the risk of developing a glioma increases with age[1]. The average 5-year survival rate of patients ranges from 6–90%, depending on tumor genetics and patient age [2–3]. By the year 2050, the United States (US) is projected to have 83.7 million individuals (about twice the population of California) aged over 65; almost double that which was projected for the year 2012 [4]. Surgical resection, chemotherapy, and radiation treatment play important roles in glioma treatment [5–6]. As the US and world population ages, neurosurgeons may encounter increasing numbers of gliomas. Additionally, due to an aging population, there may be more glioma patients that need aggressive operations at older ages and with more medical comorbidities. Updated World Health Organization (WHO) glioma grading highlights Isocitrate Dehydrogenase 1 and 2 (IDH) mutations as the principal criterion for prognostication and genetic lineage of gliomas [7]. Even still, IDH-mutant gliomas represent a diverse range of tumors that have median survivals that range from 18 months to greater than 10 years [8–10].

While the concept of “frailty” and its impact upon medical/surgical care has been present for over 30 years, it has only recently been applied to prognostication of outcome after neurosurgical treatment [11–12]. “Frailty” has been defined as a patient’s ability to respond to a given stressor [13]. In the neuro-oncological literature, frailty has been used to predict surgical decision-making in geriatric patients with WHO grade IV glioma, 30-day readmission in patients undergoing cranial neuro-oncological procedures and increases in hospital charges during neuro-oncological hospitalizations [14–16]. In sum, neurosurgical studies about frailty in neuro-oncology have focused on frailty as an exposure variable in glioblastoma and other more common cranial tumor-types, where it has been associated with worsened outcomes. However, there are no studies that have studied the impact of frailty in patients with IDH-mutant glioma. This is due to the recent reclassification of WHO glioma grading, and lack of experience with this specific pathology. It is unclear whether increased frailty is associated with worsened outcomes in patients with IDH-mutant glioma (like in other intracranial tumor types), or if other tumor-related factors, such as tumor genetics or size/location, matter more.

The primary aim of this study was to examine the association between frailty (measured by two frailty metric scores) and 30-day readmission in patients undergoing biopsy or surgical resection of IDH-mutant glioma. Our secondary aim was to study the effect of frailty on overall survival in the same surgical population of patients with IDH-mutant glioma. We hypothesize that (1) higher level of frailty would be associated with an increased risk for 30-day readmission, and (2) that higher level of frailty would be associated with shorter overall survival.

Study Population

This retrospective cohort study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.

We identified patients from the Central Nervous System (CNS) Tumor Outcome Registry at Emory (CTORE), a prospectively maintained database of patient outcomes for CNS tumors treated Emory University Hospital and Emory University Hospital, Midtown. Both hospitals contributing patients to our database are large, tertiary/quaternary care, referral, academic institutions. In this study, we included patients 18 years or older with pathological diagnosis of IDH-mutant glioma who underwent a neurosurgical procedure at the above institutions between 01 January 2007 to 01 January 2021. The diagnosis of IDH-mutant glioma was made using multimodal neuropathological technique according to latest available WHO guidelines at the time of diagnosis [9–10, 17]

Our patient flow diagram is available in Supplemental Fig. 1. Our study size was obtained by collecting all available patients for our retrospective analysis.

Variables:

Main exposure:

Frailty was defined using the 5-factor modified frailty index (mFI-5), and the Charlson Comorbidity Index (CCI) [18–19]. Both metrics have been validated in multiple settings in surgical and neurosurgical literature, and across surgical disciplines. Specifically, a mFI-5 ≥ 1 or a CCI ≥ 3 have been associated with poorer outcomes [18, 20–23]. These scores were calculated using the preoperative comorbidity list noted in preoperative history and physical or clinic note. If the preoperative history documentation denoted any comorbidities present in the CCI or mFI-5, it was recorded in our database as a binary (i.e., yes/no) or leveled (high, medium, low) categorical variable specific to the CCI/mFI-5 metric. The mFI-5 is scored 0–5. If the patient has the comorbidity included in the index, they receive a “1” for the condition, and receive a “0” otherwise. Factors included in the mFI-5 include functional status (1 = requiring assistance with activities of daily living, 0 = not requiring assistance) and history of diabetes, chronic obstructive pulmonary disease, heart failure, or hypertension. The CCI is calculated from a sum score of 19-possible weighted conditions and is age-adjusted [19, 24]. Categories include: history of HIV/AIDS, metastatic solid tumor, liver disease, lymphoma, leukemia, any tumor, diabetes with end organ damage, renal disease, hemiplegia, diabetes, ulcer disease, connective tissue disease, chronic pulmonary disease, dementia, cerebrovascular disease, peripheral vascular disease, heart failure, myocardial infarction, and age (increasing for each decade ≥ 50 years). We categorized the patients into “higher” vs “lower” risk by the respective frailty indices, which correlates to mFI-5 ≥ 1 or a CCI ≥ 3.

Outcomes:

Our primary outcome was 30-day rehospitalization at Emory University Hospital or Emory University Hospital, Midtown. Our secondary outcome was defined as mortality date after surgery, or overall survival. The time to survival was calculated as time interval from date of surgery to outcome or censoring. Patients were censored if they were lost to follow-up. Their survival time was calculated as the interval between date of surgery and last communication. The last possible follow-up date was 9/1/2021. The data for outcome variables were obtained through chart review and confirming with patient/patient family phone calls. There was an attempt to contact each patient/patient family 3 times if they did not initially answer.

Follow-up was obtained through the electronic medical record, or, in cases longer than six months of missing follow-up data, phone calls to patients and/or patients’ designated healthcare advocates.

Other covariates of interest:

Patient charts were reviewed for patient demographics, including age at surgery (difference between initial surgery date and date of birth), biological sex, race (white, African American, Latinx, Asian, or other), body mass index (BMI, kg/m2), Karnofsky Performance Status (KPS), and preoperative comorbidities (see above). Preoperative neurological status was assessed by report of seizures, presence of neurological deficit. Postoperative neurological status was assessed by presence of a neurological deficit, seizures after surgery (during hospitalization), KPS at discharge, and Modified Rankin Scale (mRS) at discharge [25–26]. Postoperative complications included presence of hemorrhage, surgical site infection, length of stay, or other medical complications (including non-surgical site infection, deep vein thrombosis, or cardiopulmonary event). Data about surgical procedures included if the lesion was recurrent, i.e., operated upon at another institution prior to our care. In these cases, the date of the initial surgery was recorded. Other surgical data included whether the patient received a stereotactic needle biopsy, or craniotomy for resection, and date of surgery (and subsequent surgeries at one of our institutions). Post-hospitalization covariates included whether the patient received adjuvant chemotherapy or chemotherapy, discharge disposition (home, rehabilitation center, long-term acute care center).

Pre-operative imaging was obtained within 1 month of surgical intervention, and post-operative magnetic resonance imaging (MRI) was performed within 72 hours of surgery. These images along with the attending neuroradiologist documented interpretation were used to evaluate the location (lobe of brain) of glioma, size, contrast enhancement pattern, and extent of resection (EOR). EOR was dichotomized to gross total resection (GTR) where the entirety of tumor mass was resected, otherwise designated subtotal resection (STR).

All patients included in the study had genetic analysis completed to confirm histopathological diagnosis. Glioma tissue obtained through surgery is examined with immunohistochemistry, cytogenomic DNA copy number microarray (OncoScan® -Thermo Fisher Scientific), multiplex PCR (SNaPshot™ - Thermo Fisher Scientific) with MetaCore™ (Clarivate Analytics) enrichment for identification of associated molecular pathways.

Statistical analysis:

Statistical analyses were performed using R Statistical Software (version 4.1.1, R Foundation for Statistical Computing, Vienna, Austria) and SAS version 9.4 (SAS Institute, Cary, NC). We described the cohort characteristics by frailty status (i.e., high or low) dichotomized All Variables were assessed for normality; means were reported when variables were distributed normally, and medians/interquartile ranges were reported when otherwise. To assess the association between frailty and 30-day readmission, we conducted an unadjusted logistic regression. A directed acyclic graph (DAG—see Supplemental Fig. 2) was used to our multivariable-adjusted model. We assessed the association between frailty and 30-day readmission both for frailty and CCI indices. For the logistic regression models, we reported odds ratios (OR) and adjusted ORs (aOR) with 95% confidence intervals (CI).

Overall survival was assessed with cumulative Kaplan-Meier survival curves and Cox- proportional hazard models to evaluate whether the mFI-5 or CCI frailty exposure was associated with the rate of overall survival. We dichotomized survival curves by surgery type (biopsy vs. resection surgery), tumor genetic lineage (astrocytoma vs. oligodendroglioma), and primary tumor location (cortically-based/lobar vs. deep brain structure). Censorship was defined as above (Outcomes). Proportional hazard assumptions for covariates were assessed graphically, with goodness-of-fit tests, and time-dependent models. Similar to the process completed for assessment of variables to include in our logistic models, we used bivariate associations between frailty exposure and OS outcome as well as a DAG to inform our Cox-proportional hazards model (Supplemental Fig. 2). For the Cox-proportional hazards model, we reported hazard ratios (HR), adjusted HRs (aHR), and 95% CIs.

For logistic regression and Cox-proportional hazards procedures, we employed a complete case analysis. Missing data for each covariate of interest can be found in Table 1.

Table 1

Cohort characteristics, dichotomized by CCI 3 and mFI-5 ≥ 1.
		Charlson Comorbidity Index		5-Factor Modified Frailty Index
Variables, n (%)	Total Cohort	CCI = 0–2, N = 118¹	CCI ≥ 3, N = 18¹	mFI = 0, N = 102¹	mFI-5 ≥ 1, N = 34¹
Age at Surgery, Years	38 (31, 47)	37 (30, 43)	63 (56, 73)	36 (30, 43)	48 (39, 62)
Male Sex	84 (62%)	74 (63%)	10 (56%)	62 (61%)	22 (65%)
Race
White	111 (82%)	97 (82%)	14 (78%)	83 (81%)	28 (82%)
African American	14 (10%)	14 (12%)	0 (0%)	10 (9.8%)	4 (12%)
Latino	5 (3.7%)	4 (3.4%)	1 (5.6%)	5 (4.9%)	0 (0%)
Asian	1 (0.7%)	1 (0.8%)	0 (0%)	1 (1.0%)	0 (0%)
Other	3 (2.2%)	2 (1.7%)	1 (5.6%)	3 (2.9%)	0 (0%)
Not Reported	2 (1.5%)	0 (0%)	2 (11%)	0 (0%)	2 (5.9%)
Body Mass Index (kg/m²)	26.7 (23.7, 31.7)	26.3 (23.6, 30.3)	30.2 (24.4, 37.0)	26.3 (23.8, 29.8)	30.1 (23.4, 34.4)
Missing	10 (7%)	7 (6%)	3 (17%)	6 (6%)	4 (12%)
Preop Karnofsky Performance Status ≥ 70	124 (91%)	113 (96%)	11 (61%)	100 (98%)	24 (71%)
Preop Seizures	85 (62%)	73 (62%)	12 (67%)	63 (62%)	22 (65%)
Preop Neurological Deficit	45 (33%)	34 (29%)	11 (61%)	26 (25%)	19 (56%)
Tumor Location
Frontal	85 (62%)	74 (63%)	11 (61%)	65 (64%)	20 (59%)
Parietal	13 (9.6%)	11 (9.3%)	2 (11%)	8 (7.8%)	5 (15%)
Temporal	29 (21%)	26 (22%)	3 (17%)	24 (24%)	5 (15%)
Occipital	6 (4.4%)	5 (4.2%)	1 (5.6%)	4 (3.9%)	2 (5.9%)
Insula	2 (1.5%)	1 (0.8%)	1 (5.6%)	1 (1.0%)	1 (2.9%)
Cerebellum	1 (0.7%)	1 (0.8%)	0 (0%)	0 (0%)	1 (2.9%)
Left-Sided Tumor	75 (55%)	66 (56%)	9 (50%)	57 (56%)	18 (53%)
Tumor Primarily Centered Outside Deep Structures	99 (73%)	88 (75%)	11 (65%)	75 (74%)	24 (73%)
Unknown	1	0	1	0	1
Tumor Centered in Eloquent Location	61 (46%)	52 (44%)	9 (53%)	44 (43%)	17 (53%)
Unknown	2	1	1	0	2
Type of Surgery
Stereotactic Biopsy	47 (35%)	37 (31%)	10 (56%)	36 (35%)	11 (32%)
Open Biopsy	2 (1.5%)	1 (0.8%)	1 (5.6%)	1 (1.0%)	1 (2.9%)
Craniotomy for Resection	87 (64%)	80 (68%)	7 (39%)	65 (64%)	22 (65%)
Astrocytic Lineage (vs. Oligodendroglioma)	87 (64%)	80 (68%)	7 (39%)	68 (67%)	19 (56%)
Total # Copy Number Variations	9 (5, 18)	10 (5, 17)	8 (5, 31)	10 (5, 17)	8 (5, 26)
Unknown	43	38	5	31	12
Immediate Postoperative Neurological Deficit
None	112 (82%)	100 (85%)	12 (67%)	86 (84%)	26 (76%)
Motor	12 (8.8%)	6 (5.1%)	6 (33%)	5 (4.9%)	7 (21%)
Sensory	2 (1.5%)	2 (1.7%)	0 (0%)	2 (2.0%)	0 (0%)
Language	6 (4.4%)	6 (5.1%)	0 (0%)	6 (5.9%)	0 (0%)
Visual	2 (1.5%)	2 (1.7%)	0 (0%)	2 (2.0%)	0 (0%)
Other	2 (1.5%)	2 (1.7%)	0 (0%)	1 (1.0%)	1 (2.9%)
Postop Karnofsky Performance Status ≥ 70	124 (91%)	114 (97%)	10 (56%)	100 (98%)	24 (71%)
Adjuvant Temozolomide Use	106 (84%)	96 (85%)	10 (77%)	82 (84%)	24 (86%)
Unknown	10	5	5	4	6
Adjuvant Radiation Use	100 (74%)	90 (76%)	10 (56%)	78 (76%)	22 (65%)
Median Time to Death	36 (18, 77)	55 (25, 94)	13 (6, 24)	55 (25, 90)	14 (6, 32)
Unknown	113	99	14	85	28
30-Day Readmission	8 (5.9%)	7 (5.9%)	1 (5.6%)	5 (4.9%)	3 (8.8%)
¹Median (IQR); n (%) Abbreviations: CCI: Charlson Comorbidity Index; mFI-5: 5-factor modified frailty index; #: number

Ethical Considerations:

Our Institutional Review Board reviewed this study (IRB00117860 and STUDY00000332). Our review board approved the waiver of informed patient consent for this study.

The overall cohort characteristics are shown in Table 1. In total, there were 136 patients included in our analyses. Eighty-seven (64%) patients had an astrocytoma and 49 (36%) had an oligodendroma. Overall, 87 (64%) patients underwent maximal safe resection while 49% had a biopsy only. Eight (5.9%) patients were readmitted to the hospital in less than 30 days from the date of surgery. Looking at frailty, 18 (13%) of patients had a CCI greater than or equal to 3 (“high CCI”), while 34 (25%) of patients had a mFI-5 greater than or equal to 1 (“high mFI-5”) (Table 1). Patients with high frailty scores had high body mass index (BMI) (CCI: 30.2 kg/m² [Interquartile Range/IQR: 24.4–37.0 kg/m²]; mFI-5: 30.1 kg/m² [IQR: 23.4–34.4]), more preoperative neurological deficit (CCI: 11/18, 61%; mFI-5: 19/34, 56%), and older age at surgery (CCI: 63 years [IQR: 56–73 years]; mFI-5: 48 years [IQR: 39–62 years]). The type of surgery (biopsy vs. craniotomy for resection) as well as the total mutational burden of the tumor, measured by CN variation, was relatively equal among those with high and low frailty.

In our logistic regression for the outcome of odds of 30-day readmission, the crude odds of 30-day readmission in patients with high CCI was 0.93 (0.04–5.72) times that of those that had a low CCI (Table 2). In the adjusted logistic regression, the odds of 30-day readmission in patients with high CCI was 0.22 (0.01–3.24), adjusting for tumor location, BMI, type of surgery, and age (Table 2). The crude odds of 30-day readmission in patients with high mFI-5 was 1.88 (0.37–8.10) times that of those that had a low mFI-5. In the adjusted logistic regression, the odds of 30-day readmission in patients with high mFI-5 was 1.56 (0.24–8.96), adjusting for tumor location, BMI, type of surgery, and age (Table 2).

Table 2

Results from multivariable adjusted models for effect of frailty upon 30-day readmission.
		Odds of 30-day readmission, n = 136
Frailty Metric	N (%)	Crude OR (95%CI)	aOR (95% CI)	P-value
CCI 0–2	118 (87)	REF	REF
CCI ≥ 2	18 (13)	0.93 (0.04–5.72)	0.22 (0.01–3.24)^a	0.30
mFI-5 = 0	102 (75)	REF	REF
mFI-5 ≥ 1	34 (25)	1.88 (0.37–8.10)	1.56 (0.24–8.96)^b	0.62
Abbreviations: OR: odds ratio; aOR: adjusted odds ratio; 95% CI: 95% confidence interval; REF: reference. ^aModel adjusted for tumor location, BMI, type of surgery, and age, model n = 125 ^bModel adjusted for tumor location, BMI, type of surgery, and age, model n = 125

The median overall survival for the entire cohort was 54 months. When separated by frailty metrics (Table 3), we found that the median survival time of patients with high CCI (11.1 months [IQR: 1.1–58.0 months]) were lower than those with low CCI (55.3 months [IQR: 22.0–99.1 months]). Additionally, we found the median survival time of patients with high mFI-5 (26.3 months [IQR: 2.5–55.3 months]) were lower than those with low mFI-5 (63.1 months [IQR: 31.6–99.8 months]). In both comparisons, however, the IQRs crossed. Although the survival curves show some initial separation when comparing the frailty index groups, there is eventual overlap between high and low frailty curves (Fig. 1).

Table 3

Results from adjusted models for effect of frailty upon overall survival.
			Hazard ratio for rate of overall survival, n = 136
Frailty Metric	Death N (%)	Median months to death (IQR)	Crude HR (95%CI)	aHR (95% CI)	P-value
CCI 0–2	19 (83)	55.3 (22.0, 99.1)	REF	REF
CCI ≥ 3	4 (17)	11.1 (1.1, 58.0)	3.33 (1.12–9.97)	0.59 (0.05–6.37)^a	0.7
mFI-5 = 0	17 (74)	63.1 (31.6, 99.8)	REF	REF
mFI-5 ≥ 1	6 (16)	26.3 (2.5, 55.3)	2.14 (0.83–5.47)	1.15 (0.29–4.52)^b	0.8
Abbreviations: HR: hazard ratio; aHR: adjusted hazard ratio; 95% CI: 95% confidence interval; REF: reference. ^aModel adjusted for age, tumor location, BMI, and type of surgery, history of prior surgery; model n = 125 ^bModel adjusted for age, tumor location, BMI, and type of surgery, history of prior surgery; model n = 125

In our Cox-proportional hazards analysis (Table 3), the crude hazard rate of death in patients with high CCI was 3.33 (1.12–9.97) times the hazard rate of death in patients with low CCI. In the adjusted analysis, the adjusted hazard rate of death in patients with high CCI was 0.59 (0.05–6.37) times the hazard rate of death in patients with low CCI, adjusting for age, tumor location, BMI, history of prior surgery, and type of surgery. Using our other frailty metrics, the crude hazard rate of death in patients with high mFI-5 was 2.14 (0.83–5.47) times the hazard rate of death in patients with low mFI-5. In the adjusted analysis, the adjusted hazard rate of death in patients with high mFI-5 was 1.15 (0.29–4.52) times the hazard rate of death in patients with low mFI-5, adjusting for age, tumor location, BMI, history of prior surgery, and type of surgery. Interestingly, when we subdivided the cohort by patients who had new tumor diagnoses and first operated upon at our institution (i.e. excluding recurrent tumors), we found that the adjusted hazard ratio of death in patients with high mFI-5 was 6.79 (1.00–45.9) times that of the adjusted hazard ratio of death in patients with low mFI-5, when adjusting for age, tumor location, BMI, and surgery type (p = .049) (Table 4).

Table 4

Adjusted hazard ratio of death in patients with de novo tumors/new diagnosis (n = 99).
	Hazard ratio for rate of overall survival in patients with de novo/ new tumors, n = 99
Variable	HR¹	95% CI¹	P-value
mFI-5 = 0	REF	REF	REF
mFI-5 ≥ 1	6.79	1.00, 45.9	0.049
Age at Surgery	1.01	0.96, 1.07	0.6
Convexity Location	0.50	0.13, 1.99	0.3
BMI	0.81	0.69, 0.94	0.006
Open Craniotomy for Resection	3.22	0.84, 12.4	0.088
Abbreviations: ¹HR: Hazard Ratio; 95% CI: 95% confidence interval; REF: Reference.

We subdivided the cohort into patients that had biopsy vs. craniotomy for resection and into astrocytoma vs. oligodendroglioma lineage (Supplemental Fig. 3). This showed no significant separation of survival curves.

In this paper, we described the relation of the exposure of frailty (measured by two metrics) upon two outcomes; 30-day readmission and OS in patients with IDH-mutant glioma. We found that CCI and mFI-5 were not associated with 30-day readmission. We also found that CCI and mFI-5 were not associated with the OS. However, in patients that had their first surgery at our institution (not recurrent tumors), there was an association of one frailty measure (mFI-5) with OS.

While we hypothesized that frailty would be associated with 30-day readmission and OS in our cohort, there are multiple reasons why we may have not discovered such associations. While our single-institution experience of IDH-mutant glioma is relatively large, the overall small sample size (n = 136) and resulting under-powered statistical analyses did not allow for effective testing of true associations. In short, this study is potentially marred by type II error.

Additionally, the proportion of patients with high frailty in our cohort is small, with only 13.2% (n = 18) of patients having high CCI and 25% (n = 34) of patients having high mFI-5. Overall, patients with IDH-mutant glioma tend to be younger, and thus more healthy/less frail than other brain tumor cohorts (Table 1) [27]. Patients with metastatic tumors tend to be older and have more systemic disease than patients with IDH-mutant glioma we studied in our cohort and others [28, 29]. The event rate of 30-day readmission in the cohort was low as well (5.9%, n = 8). These limitations reduce the ability to detect associations between frailty and our primary outcome.

There are other factors that may explain the lack of association of frailty with outcome in patients with glioma. The type of surgery that patients receive has a significant association with readmission, PFS, and OS. While still somewhat controversial, it is generally accepted that patients with glioma that obtain a maximal safe surgical resection have increased time to recurrence and mortality [30–33]. However, the association of extent of resection and outcome is plagued by numerous confounders. For example, there is significant bias in the administration of attempted gross total resection; neurosurgeons will not attempt aggressive resection in tumors located in eloquent areas of the brain or in older patients or patients that have significant medical comorbidities [14, 32]. Additionally, tumor genetics play a larger role in outcome in patients with glioma compared to other lesions of the brain. Patients with high grade gliomas (HGG) have shorter OS and PFS than patients with low grade gliomas (LGG) [7–9, 27, 34]. Surgery type, extent of resection, and genetic factors are examples of parameters that play a role in outcome of patients with glioma and may disrupt associations of frailty and outcome in glioma cohorts.

Other groups have described the association of frailty with OS in patients with brain tumors. Youngerman et al. found that the modified frailty index was associated with 30-day readmission, mortality, medical complications, neurological complications, prolonged length of stay, and discharge to rehabilitation facility rather than home [15]. There are major differences in the IDH-mutant glioma cohort and the cohort used for the Youngerman et al. study. First, the data to form the cohort in the Youngerman et al. study were from the American College of Surgeons NSQIP database and had a much larger sample size (n = 9149 patients). This cohort was more adequately powered to discover the associations delineated above. Second, fewer than one-half of patients in the NSQIP database had glioma. Third, the cohort had a greater percentage of patients with frailty, with 48.5% having at least low frailty.

Cloney et al. and Khalafallah et al. also related frailty measures to outcome in patients with brain tumors [14, 35]. Cloney et al. studied the exposure of frailty within a cohort of 319 geriatric patients with HGG. They found that patients with more frailty were less likely to undergo surgical resection (vs. biopsy), had longer stay in hospital, and increased overall risk of complications. Differences in this cohort compared to our IDH-mutant cohort include the older age of patients, higher rates of frailty, and more homogenous tumor type (HGG). Khalafallah et al. described the relationship of frailty and outcome in 1692 patients with brain tumors. They found that increased frailty was related to 90-day mortality, in a dose-adjusted pattern. Key differences in this cohort to ours include increased sample size, low rate of glioma diagnosis (< 30%), and very low rate of outcome of mortality (3%).

In future studies with this cohort, we could strengthen our ability to test associations between frailty and outcome in IDH-mutant glioma by increasing sample size, testing different levels of the exposure (different cut off points for “high” vs. “low” frailty), and including sensitivity analyses. Additionally, we could implement Bayesian analytic strategies to analyze this low sample size database [36].

Looking forward, we aim to enhance the predictive power of our studies by employing machine learning algorithms for imaging analysis to forecast overall survival and determine tumor types. Moreover, high-dimensional analysis of microarray data will be pursued to discern whether specific gene mutations and mutational burdens correlate with overall survival and recurrence rates. These advancements in data analysis will hopefully provide a more nuanced understanding of the interplay between frailty and patient outcomes in IDH-mutant glioma.

Author Contribution

Conceptualization: DPB and NMS contributed to conceptualization. Data Curation: performed formal analysis and data curation. The Investigation was carried out by all authors. DPB, NMS, JC, AJ, EKN, JJO, KBH reviewed methodology DPB and NMS wrote the first version. All authors reviewed and edited this paper. All authors read and approved the final version of this manuscript. DPB and NMS are co-first authors of the present manuscript. KBH is last author of the present manuscript.

Dubrow R, Darefsky AS (2011) Demographic variation in incidence of adult glioma by subtype, United States, 1992–2007. BMC Cancer 11. 10.1186/1471-2407-11-325
Ostrom QT, Gittleman H, Truitt G, Boscia A, Kruchko C, Barnholtz-Sloan JS (2018) CBTRUS Statistical Report: Primary Brain and Other Central Nervous System Tumors Diagnosed in the United States in 2011–2015. Neuro Oncol 20(suppl4):iv1–iv86. 10.1093/neuonc/noy131
Ostrom QT, Cote DJ, Ascha M, Kruchko C, Barnholtz-Sloan JS (2018) Adult glioma incidence and survival by race or ethnicity in the United States from 2000 to 2014. JAMA Oncol 4(9):1254–1262. 10.1001/jamaoncol.2018.1789
Ortman JM, Velkoff Va, Hogan H (2014) An aging nation: The older population in the United States. Econ Stat Adm US Dep Commer 1964:1–28 census.gov
Stupp R, Mason WP, van den Bent MJ et al (2005) Radiotherapy plus Concomitant and Adjuvant Temozolomide for Glioblastoma. N Engl J Med 352(10):987–996. 10.1056/NEJMoa043330
Parsons DW, Jones S, Zhang X et al (2008) An integrated genomic analysis of human glioblastoma multiforme. Sci (80-) 321(5897):1807–1812. 10.1126/science.1164382
Louis DN, Perry A, Reifenberger G et al (2016) The 2016 World Health Organization Classification of Tumors of the Central Nervous System: a summary. Acta Neuropathol 131(6):803–820. 10.1007/s00401-016-1545-1
Eckel-Passow JE, Lachance DH, Molinaro AM et al (2015) Glioma groups based on 1p/19q, IDH, and TERT promoter mutations in tumors. N Engl J Med 372(26):2499–2508. 10.1056/NEJMoa1407279
Brat DJ, Verhaak RGW, Aldape K, Comprehensive (2015) June, Integrative Genomic Analysis of Diffuse Lower-Grade Gliomas.
https://doi.org/101056/NEJMoa1402121
. https://www.nejm.org/doi/10.1056/NEJMoa1402121
Louis DN, Perry A, Burger P et al (2014) International Society Of Neuropathology–Haarlem consensus guidelines for nervous system tumor classification and grading. Brain Pathol 24(5):429–435. 10.1111/bpa.12171
Vaupel JW, Manton KG, Stallard E (1979) The impact of heterogeneity in individual frailty on the dynamics of mortality. Demography 16(3):439–454
Karam J, Tsiouris A, Shepard A, Velanovich V, Rubinfeld I (2013) Simplified frailty index to predict adverse outcomes and mortality in vascular surgery patients. Ann Vasc Surg 27(7):904–908. 10.1016/j.avsg.2012.09.015
Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K (2013) Frailty in elderly people. Lancet 381(9868):752–762. https://doi.org/10.1016/S0140-6736(12)62167-9
Cloney M, D’Amico R, Lebovic J et al (2016) Frailty in Geriatric Glioblastoma Patients: A Predictor of Operative Morbidity and Outcome. World Neurosurg 89:362–367. 10.1016/j.wneu.2015.12.096
Youngerman BE, Neugut AI, Yang J, Hershman DL, Wright JD, Bruce JN (2018) The modified frailty index and 30-day adverse events in oncologic neurosurgery. J Neurooncol 136(1):197–206. 10.1007/s11060-017-2644-0
Huq S, Khalafallah AM, Jimenez AE et al (2020) Predicting Postoperative Outcomes in Brain Tumor Patients With a 5-Factor Modified Frailty Index. Neurosurgery 88(1):147–154. 10.1093/neuros/nyaa335
Brat DJ, Aldape K, Colman H et al (2020) cIMPACT-NOW update 5: recommended grading criteria and terminologies for IDH-mutant astrocytomas. Acta Neuropathol 139(3):603–608. 10.1007/s00401-020-02127-9
Subramaniam S, Aalberg JJ, Soriano RP, Divino CM (2018) New 5-factor modified frailty index using American College of Surgeons NSQIP data. J Am Coll Surg 226(2):173–181
Charlson ME, Pompei P, Ales KL, MacKenzie CR (1987) A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 40(5):373–383
Weaver DJ, Malik AT, Jain N, Yu E, Kim J, Khan SN (2019) The Modified 5-Item Frailty Index: A Concise and Useful Tool for Assessing the Impact of Frailty on Postoperative Morbidity Following Elective Posterior Lumbar Fusions. World Neurosurg 124:e626–e632. https://doi.org/10.1016/j.wneu.2018.12.168
Traven SA, Reeves RA, Sekar MG, Slone HS, Walton ZJ (2019) New 5-Factor Modified Frailty Index Predicts Morbidity and Mortality in Primary Hip and Knee Arthroplasty. J Arthroplasty 34(1):140–144. 10.1016/j.arth.2018.09.040
Robinson TN, Eiseman B, Wallace JI et al (2009) Redefining Geriatric Preoperative Assessment Using Frailty, Disability and Co-Morbidity. Ann Surg. ;250(3). https://journals.lww.com/annalsofsurgery/Fulltext/2009/09000/Redefining_Geriatric_Preoperative_Assessment_Using.13.aspx
Ening G, Osterheld F, Capper D, Schmieder K, Brenke C (2015) Charlson comorbidity index: an additional prognostic parameter for preoperative glioblastoma patient stratification. J Cancer Res Clin Oncol 141(6):1131–1137. 10.1007/s00432-014-1907-9
Hall WH, Ramachandran R, Narayan S, Jani AB, Vijayakumar S (2004) An electronic application for rapidly calculating Charlson comorbidity score. BMC Cancer 4:1–8. 10.1186/1471-2407-4-94
Schag CC, Heinrich RL, Ganz PA (1984) Karnofsky performance status revisited: reliability, validity, and guidelines. J Clin Oncol 2(3):187–193
Banks JL, Marotta CA (2007) Outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials: a literature review and synthesis. Stroke 38(3):1091–1096
Miller JJ, Gonzalez Castro LN, McBrayer S et al (2022) Isocitrate dehydrogenase (IDH) mutant gliomas: A Society for Neuro-Oncology (SNO) consensus review on diagnosis, management, and future directions. Neurooncology 25(1):4–25. 10.1093/neuonc/noac207
Miller KD, Ostrom QT, Kruchko C, Cancer et al (2021) J Clin 71(5):381–406. 10.3322/caac.21693
Tabouret E, Chinot O, Metellus P, Tallet A, Viens P, Gonçalves A Recent Trends in Epidemiology of Brain Metastases: An Overview. ANTICANCER RESEARCH. Published online 2012
Jiang B, Chaichana K, Veeravagu A, Chang SD, Black KL, Patil CG (2017) Biopsy versus resection for the management of low-grade gliomas. Cochrane Gynaecological, Neuro-oncology and Orphan Cancer Group. ed Cochrane Database Syst Reviews 2020(6). 10.1002/14651858.CD009319.pub3
Grabowski MM, Recinos PF, Nowacki AS et al (2014) Residual tumor volume versus extent of resection: predictors of survival after surgery for glioblastoma: Clinical article. J Neurosurg JNS 121(5):1115–1123. 10.3171/2014.7.JNS132449
Weller M, van den Bent M, Preusser M et al (2021) EANO guidelines on the diagnosis and treatment of diffuse gliomas of adulthood. Nat Rev Clin Oncol 18(3):170–186. 10.1038/s41571-020-00447-z
Khalafallah AM, Huq S, Jimenez AE, Brem H, Mukherjee D The 5-factor modified frailty index: an effective predictor of mortality in brain tumor patients. J Neurosurg Published online 2020:1–9. 10.3171/2020.5.jns20766
Yan H, Parsons DW, Jin G et al (2009) IDH1 and IDH2 Mutations in Gliomas. N Engl J Med 360(8):765–773. 10.1056/NEJMoa0808710
Khalafallah AM, Huq S, Jimenez AE, Brem H, Mukherjee D The 5-factor modified frailty index: an effective predictor of mortality in brain tumor patients. J Neurosurg Published online 2020:1–9. 10.3171/2020.5.jns20766
Lee SY, Song XY (2004) Evaluation of the Bayesian and Maximum Likelihood Approaches in Analyzing Structural Equation Models with Small Sample Sizes. Multivar Behav Res 39(4):653–686. 10.1207/s15327906mbr3904_4

No competing interests reported.

SupplementaryMaterial.docx

Download PDF

Journal Publication

published 04 Jun, 2024

Read the published version in Journal of Neuro-Oncology →

Editorial decision: Revision requested
25 Mar, 2024
Reviews received at journal
19 Mar, 2024
Reviewers agreed at journal
17 Mar, 2024
Reviewers agreed at journal
15 Mar, 2024
Reviewers invited by journal
13 Mar, 2024
Editor assigned by journal
13 Mar, 2024
Submission checks completed at journal
13 Mar, 2024
First submitted to journal
12 Mar, 2024

You are reading this latest preprint version

Frailty in Patients With IDH-Mutant Gliomas: Experience from a High-Volume Tumor Center

Status:

Journal Publication

Version 1

Abstract

Purpose

Methods

Results

Conclusion

Figures

Introduction

Methods

Study Population

Variables:

Main exposure:

Outcomes:

Other covariates of interest:

Statistical analysis:

Ethical Considerations:

Results

Discussion

Declarations

Author Contribution

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1