The 5-Factor Modied Frailty Index Predicts Complications and Worsening After Meningioma Surgery in Non-elderly Rather Than Elderly Patients: A Nationwide Registry Study

The simplied 5-factor modied frailty index (mFI-5) is a useful indicator of outcome for patients undergoing surgeries and considered as an important risk factor in elderly patients. However, its usefulness has not been validated based on age groups. We aimed to investigate the risk factors including the mFI-5 across age groups for complications and worse outcomes in meningioma surgery using data obtained from the nationwide database in Japan. We extracted data from the nationwide registry database in Japan between 2010 and 2015. Age (< 65, 65–74, and ≥ 75 years), sex, Barthel Index (BI), mFI-5 scores, and complications were evaluated. Multivariate logistic regression analyses identied risk factors across all age groups for worsening BI scores and complications after surgery. In 8,138 included cases, an mFI-5 score ≥ 2 items was a signicant risk factor for worsening BI scores in patients aged < 65 years (odds ratio: 2.00; 95% condence interval: 1.31-3.06), but not in patients aged 65-74 years and those aged ≥ 75 years. Similar results were noted for complications in patients aged < 65 years (2.40; 1.67–3.44), but not in patients aged 65-74 years and those aged ≥ 75 years. In conclusion, the mFI-5 scores can predict the risk of worsening outcome and complications in non-elderly patients aged < 65 years rather than in elderly patients aged ≥ 65 years. In meningioma surgeries, care must be taken when making decisions using the mFI-5 scores based on the patients’ age.


Introduction
The recent increase in average life expectancy and frequency of diagnostic neuroimaging has resulted in a globally increased rate of incidental meningioma detection in the elderly [14]. Chronological age is considered one of the most effective surgical indications for elderly patients with meningioma worldwide; however, surgical decision-making strategies for elderly patients should be carefully reviewed considering the increased frailty and decline in health that are associated with advanced age.
To assess frailty, the 11-factor modi ed frailty index (mFI-11) was derived from the Canadian Study of Health and Aging Frailty Index [25] by matching 11 comorbidity and de cit variables from the American College of Surgeons' National Surgery Quality and Improvement Project, which are well-validated health measures that have been applied within general medical and surgical datasets [7,23,30,33]. Recently, the simpli ed 5-factor modi ed frailty index (mFI-5) was introduced and validated within various elds of surgery, and it includes the following ve factors: the prevalence of functional dependence, history of diabetes mellitus, history of chronic obstructive pulmonary disease, congestive heart failure, and hypertension [13,26,28,31]. Frailty is observed more commonly among the elderly than non-elderly patients; therefore, the frailty index is generally considered an important risk factor in elderly patients.
However, it remains unknown whether the mFI-5 scores are associated with the same risk across each age group, and whether their predictive value is useful with regard to complications and outcomes of meningioma surgery across various age groups. Therefore, the aim of this study was to investigate risk factors including the mFI-5 across various age groups for complications and worse outcomes in meningioma surgery, with data extracted from the nationwide database in Japan.

Protocol approval and patient consent
The present study was approved by the local Institutional Review Boards of Hiroshima University (no. E-631) and Tokyo University (no. 3501- [1]). Due to the anonymous nature of the data in this study, the requirement for informed consent was waived.

Data source and selection of patients
The Japanese Diagnosis Procedure Combination (DPC) is a registry-based national database that includes abstract discharge data and administrative claims on inpatients in Japan. It has been described thoroughly elsewhere [10,11,17]. Both the sensitivity and speci city of the procedure exceeded 90% [32].
We included patients aged 18-95 years who were admitted to the hospital with a primary diagnosis of intracranial meningioma between July 1, 2010, and March 31, 2015. Diagnoses of meningioma (ICD-10 codes; D32) and the intracranial tumor removal procedure (medical fee code; K169) were identi ed. In total, 10,530 patients with meningiomas were identi ed, and we excluded cases with unknown location of the meningioma, multiple meningiomas, and no detection of BI assessments and body mass index (BMI). Consequently, 8,138 patients with meningiomas were included in this study (Fig. 1).
The database incorporates the coded variables as per the International Classi cation of Diseases and Related Health Problems, 10th Revision (ICD-10), and the Barthel Index (BI) at admission and at discharge were evaluated. We collected data on the patients' sex, age, BMI, medical history, internal oral medication on admission, location of the meningioma, BI score at admission and at discharge, and in-hospital complications. As part of the medical history, we included a review of diabetes mellitus (ICD-10 code: E14), hypertension (I10), cerebral infarction (I63), angina pectoris (I20), congestive heart failure (I50), and chronic obstructive pulmonary disease (J44). We reviewed the use of antiplatelet drugs (aspirin, ticlopidine hydrochloride, cilostazol, clopidogrel sulfate, clopidogrel-aspirin combination, and prasugrel hydrochloride), anticoagulation agents (warfarin, dabigatran etexilate, edoxaban tosilate hydrate, rivaroxaban, and apixaban), and statins (atorvastatin calcium hydrate, rosuvastatin calcium, and pitavastatin calcium hydrate). In addition, the patients for the presence of complications, including intracerebral hemorrhage (ICH; ICD-10 code: I61), subarachnoid hemorrhage (SAH; I60), cerebral infarction, congestive heart failure, and pneumonia (J18) was evaluated.
The BI was used to evaluate 10 activities of daily living (ADL) across two to four stages [2,29]. The BI was classi ed into three categories: 0-55, 60-80, and 85-100 points according to ADL, with higher scores indicating a higher level of independent functioning. One of the ve factors included in the mFI-5 was the functional status of requiring assistance with ADL. We de ned this dependent functional status as a BI score < 85. The scores of the mFI-5 were classi ed into three categories, namely, 0, 1 item, and ≥ 2 items. A worsening BI score indicated patients who demonstrated a decreased BI score ≥ 5 points at discharge compared to that at admission, and in-hospital mortality indicated death from any cause. Any complications included any stroke, congestive heart failure, pneumonia, a decreased BI score, and inhospital mortality.
The hospital data reviewed primarily assessed the case volume and type (academic or non-academic). Hospital case volume was de ned as the number of patients with meningiomas treated surgically at an individual facility during the study period, and this was categorized into three groups according to terciles of case volume, with an approximately equal number of patients in each of the three groups. Hospital case volume was categorized from 1 to 3, ordered from the lowest to the highest value. The anatomical locations of the meningiomas were classi ed as convexity, falx, parasagittal, lateral (sphenoidal ridge, middle fossa), midline (tuberculum sellae, olfactory groove), posterior fossa (foramen magnum, petrous, petroclival, petrotentorial, tentorial), or deep (ventricle, anterior clinoid, posterior clinoid, cavernous, orbital, falcotentorial). Patients were categorized into three age groups, based on the classi cation of the World Health Organization and the Japan Geriatrics Society, as follows: < 65 years (non-elderly), 65-74 years (pre-elderly), and ≥ 75 years (elderly).

Statistical analyses
All statistical analyses were performed using Stata (version 15; StataCorp, College Station, TX, USA). Categorical variables were compared using a Chi-square or Fisher's exact test. To compare continuous variables, we performed a t-test or Mann-Whitney U test. Multivariate logistic regression analyses were performed on the overall cohort to analyze the risk factors for worsening BI scores between admission and discharge, in-hospital mortality, and any complications. The same analyses were performed separately across the three categories of age groups, except for in-hospital mortality where the numbers of participants in each subgroup were too low for statistical analysis. For multivariate logistic regression analyses, independent variables were selected based on the existing literature [4,13,19,20,26,28,31], and no variable selection method was applied; odds ratios (ORs) and 95% con dence intervals (CIs) were calculated.

Results
During the study period, 8,138 eligible patients were surgically treated. The mean (interquartile range) age was 63.0 (range: 53.0-71.0). Table 1 shows the baseline characteristics of patients with surgically treated meningioma based on age group. The mFI-5 score was signi cantly increased with age. Elderly patients tended to undergo surgery in lower surgical volume and non-academic hospitals. The prevalence of ICH, SAH, cerebral infarction, congestive heart failure, and in-hospital mortality did not signi cantly increase with age. However, the prevalence of pneumonia, worsening BI scores, and any complications increased signi cantly with age. , oral medications on the admission of antiplatelets and anticoagulants, and tumor location (falx, parasagittal, midline, posterior fossa, and deep). Figure 2 (a, b, and c) depicts, in the form of forest plots for risk factors adjusted for other variables, the results of the multivariate logistic regression analysis for worsening BI scores, in-hospital mortality, and any complications in the cases of all ages.  Table 3 demonstrates the results of the multivariate logistic regression analyses for the worsening BI scores across all age groups. Age was a signi cant risk factor only in the pre-elderly and elderly groups.
An mFI-5 score ≥ 2 items was considered as a signi cant risk only in the non-elderly group (OR: 2.00; 95% CI: 1.31-3.06), but not in the pre-elderly (1.25; 0.85-1.84) and elderly (0.97; 0.67-1.40) groups. The administration of oral anticoagulant medications on admission and several locations of meningiomas in the non-elderly and pre-elderly groups were also identi ed as signi cant risk factors. Figure 3 shows the results of the multivariate logistic regression analysis for the worsening BI scores across all age groups, depicted as forest plots for risk factors adjusted for other variables.  Table 4 shows the results of the multivariate logistic regression analyses for any complications across all age groups. Age was a signi cant risk factor only in the pre-elderly and elderly groups. An mFI-5 score of

Discussion
In this study, worsening BI and mFI-5 scores and the proportion of patients with medical history were signi cantly increased with age. An mFI-5 score ≥ 2 items was a signi cant risk factor for worsening BI scores, in-hospital mortality, and any complications after surgery. In the analysis of all age groups, an mFI-5 score ≥ 2 items was a signi cant risk factor for worsening BI scores and any complications after surgery only in the non-elderly group, but not in the pre-elderly and elderly groups.

The rate of mortality and complications after surgery
The rates of in-hospital mortality and any complications after surgery in our study were 0.5-0.9% and 10.4-25.5%, respectively. These rates were similar to those observed in previous studies; the rates of mortality and complications after surgery were reported as 0.5-1.5%, and 6.8-14.8% respectively [5,8,15,27]. A recent systematic review reported that in elderly patients with meningiomas, the rate of one-year postoperative mortality and neurological complications after surgery ranged from 2.7-49.4% and from 0-16.7%, respectively [9]. mFI scores and age as predictors of surgical complications and worsening In both sexes, there was a signi cant non-linear association between age and the FI scores. [22] Overall, the FI scores were much stronger predictors of mortality than age [22]. In this study, the mFI-5 scores increased with age, similar to ndings from a previous report investigating the FI [12]. Aging is a heterogeneous process and chronologic age is not necessarily synonymous with an individual's health status. Even so, there is little evidence to determine whether there are differences in the frailty criteria in younger versus older adults. The relative risk of mortality in younger adults is associated more with frailty than age [21]. In the present study, the overall background of mFI-5 scores in the older groups were relatively higher; therefore, the mFI-5 may not be a signi cant predictor in advanced age groups. The mFI-5 is a useful predictor of complications and worsening after surgery, speci cally in non-elderly rather than elderly patients with meningiomas. While this may sound paradoxical at rst, we believe that it may be true. In meningioma surgeries, care must be taken when making decisions using the mFI-5 scores based on the patients' age. Furthermore, we can recommend preoperative tness recovery to improve frailty, especially in non-elderly patients, who have a greater potential to regain tness than the elderly [21].
Advanced age as a risk factor of surgical complication and worsening BI score We assessed the worsening BI scores between admission and discharge and found that the rate of worsening BI scores increased from 7.1-21.1% (mean = 11.0%) as age increased, which is comparable to the increase reported in previous studies (8.3-14.8%). [3,5,27] We found pneumonia and the worsening BI scores to be signi cantly correlated with advanced age in our study. The most common complications after surgery for meningioma were new focal neurological de cits and pneumonia [5,27]. Notably, these two complications are currently considered common and inevitable in advanced aged patients with meningiomas [27].
However, we found no signi cant association between advanced age and in-hospital mortality. According to a recent systematic review of elderly patients with meningioma surgery, the rate of in-hospital mortality, worsening of the postoperative performance status, neurological de cits, and general complications ranged widely based on the reports and were not necessarily associated with advanced-age [9].

Other risk factors in meningioma surgery
The administration of antiplatelet and anticoagulation drugs was found to be a risk factor for worse outcomes in previous neurosurgical studies [1]. In our study, these medications were also risk factors for worsening BI scores, in-hospital mortality, and any complications after surgery regardless of the patients' age. Antithrombotic drugs have been administered for the prevention of recurrence in cardiac and cerebral vascular events, necessitating the correct use of these drugs in the elderly. Minimum necessary use of antithrombotic medication is required, especially in the elderly during the perioperative period since antithrombotic drugs are more frequently administered to the elderly.
Several risk factors for mortality in meningioma surgery in the elderly have been identi ed, such as location [20], preoperative Karnofsky performance status [6], the BI score [4], other grading systems [24] and advanced age [19]. Although parasagittal and deep locations of the meningioma were signi cant risk factors for the worsening BI scores and any complications in both non-elderly and pre-elderly groups, they were no longer risk factors in the elderly group (Tables 3 and 4, Figs. 3 and 4). This phenomenon may be due to differences in aggressive treatment based on the location, delicate treatment management in the elderly, and the selective indication of surgery in elderly patients with meningiomas.
In this study, the in-hospital mortality in high volume hospitals was a signi cant inverse risk factor, due to the ability of high volume hospitals to rescue patients from major perioperative complications [16]. Academic hospitals were found to be signi cant risk factors for postoperative complications in pre-elderly patients. This may be likely because academic hospitals tend to receive the most serious and di cult cases.

Limitations
This study has some limitations. First, the DPC database does not include post-discharge data, meningioma size, pathological ndings, and the extent of resection and perioperative radiation therapy. Therefore, we could not assess the long-term status of recurrence or BI scores after discharge. These variables are usually not recorded in other administrative in-hospital databases. Second, this was a registry-based study, not a randomized controlled study. Therefore, we could not completely exclude bias.
However, we improved the integrity of this study by analyzing the location of meningiomas, as well as preoperative and postoperative BI scores. Third, some variables had wide CIs due to a limited number of cases. Fourth, our results may not be generalizable to other countries, which have different medical resources and systems, and they must be interpreted with caution considering that Japan has the highest proportion of elderly people worldwide.

Conclusions
Although advanced age could lead to postoperative functional decline and complications at discharge, the mFI-5 scores could predict the complication and worsening in the patients aged < 65 years, but not in the patients aged ≥ 65 years. Care must be taken when making decisions using the mFI-5 scores based on the patients' age.

Declarations
Funding: This work was supported by the Japan Society for the Promotion of Science, Grant-in-Aid for Scienti c Research (C) 17K10829.
Role of the Funding Source: The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of this report. The corresponding author has full access to all data in the study and has nal responsibility for the decision to submit for publication.

Con ict of interest:
All authors report no con ict of interest concerning the materials or methods used in this study or the ndings speci ed in this paper.

Data availability
The anonymized data for this study could be shared on the request of any quali ed investigator to the corresponding author. Only the results of primary data from the DPC could be made available for reasonable requests in accordance with the review board.

Code availability
Stata software (version 15; StataCorp, College Station, TX, USA) Contributors: All authors made contributions to the intellectual content of this paper, contributed to data interpretation, approved the nal manuscript, and agreed with submission to this journal. Fusao Ikawa: study design and concept; acquired funding; conducted research; collected, curated, managed, and analyzed data; quality control; statistical analysis; and drafted the manuscript. Nobuaki Michihata: collected, curated, and analyzed data; quality control; statistical analysis; and revised the manuscript. Soichi Oya, Toshikazu Hidaka, Shingo Matsuda, Iori Ozono, Kenji Ohata, Kiyoshi Saito, and Kazunari Yoshida: conceived and oversaw the study, as well as revised the manuscript. Kiyohide Fushimi, Hideo Yasunaga, Teiji Tominaga, and Kaoru Kurisu: conceived and oversaw the study, as well as assisted in collecting data.

Ethics approval
The present study was approved by the local Institutional Review Boards (no. E-631 and no. 3501- [1]).
Due to the anonymous nature of the data in this study, the requirement for informed consent was waived.

Consent to participate
Not applicable Consent for publication Not applicable  Forest plots of the risk factors adjusted for other variables for the worsening BI scores in the non-elderly, pre-elderly, and elderly Abbreviations: Conv, convexity; E, Lat, lateral; Para, parasagittal; PF, posterior fossa. *p < 0.05