Introduction

Meningiomas (i.e. primary tumors arising from the dural coverings of the brain) are the most common primary non-glial brain tumors, accounting for 13–26% of all primary brain tumors. Meningiomas have an annual incidence of approximately 6 per 100,000 population [1]. Histological grading of meningiomas is based upon the current World Health Organization (WHO) classification [2]. About 90% of the meningiomas are benign, which corresponds to WHO grade I. Atypical meningiomas (WHO grade II) make up 5–7%, and anaplastic variants (WHO grade III) arise in 1–3% of the cases [3].

Surgical excision of the tumor and its dural base is the usual initial treatment, particularly in easily accessible tumors localized on the outer brain surface or at the cerebral falx between both hemispheres (so-called convexity tumors). Radiotherapy has clinical benefits when tumor histology reveals atypia or anaplasia (WHO grade II and III). Radiotherapy also reduces the risk of local recurrence of grade I tumors after a subtotal resection. However, the clinical value and optimal timing of radiotherapy after (in-) complete surgery or after recurrence remains a matter of debate [47].

The controversy on the choice and timing of radiotherapy can be attributed to a lack of knowledge regarding its side effects. The few available data indicate that about 30% of meningioma patients experience severe long-term cerebral sequelae, mostly neurocognitive deficits [810]. It is unclear, however, whether this is due to pre-treatment brain damage by the tumor itself, to surgery, to radiotherapy, or to a combination of these. For other types of primary brain tumors it is known that neurocognitive deficits in patients can be attributed to a combination of these factors [11, 12]. Complications of treatment, especially when resulting in neurocognitive dysfunction, have a great impact on patients and their proxies. Unfortunately, little is known about the effect of different treatment options on the frequency and severity of neurocognitive dysfunction in meningioma patients. More information on this issue will lead to improved understanding of the complaints of meningioma patients, and offer support in the choice and timing of treatment.

The present study aimed to determine the effect of additional radiotherapy on neurocognitive functioning. We hypothesized that additional radiotherapy in meningioma patients will have a cumulative negative effect on neurocognitive functioning and health-related quality of life (HRQOL) compared with surgery only. To put neurocognitive functioning and HRQOL of meningioma patients who underwent surgery only in perspective, we compared their results with normative data from healthy general population samples.

Patients and methods

Patients

This study is part of a two-center, retrospective investigation into the prevalence and severity of neurocognitive problems in meningioma patients. For the present study we interviewed all adult (>18 years) patients with WHO grade I intracranial meningiomas, who were treated either with surgery only (RT−), or with surgery and adjuvant external beam conformal radiotherapy (RT+) at two tertiary referral centers for brain tumor patients in Amsterdam (i.e. the Academic Medical Center and the VU University Medical Center) from 1999 through 2005. In these centers, the decision to treat meningioma patients with radiotherapy after surgery is always made on basis of localization and size of the tumor. In some patients the tumor could only be resected partially because of the localization of the tumor, mainly skull-base meningiomas. In these patients and in patients with a recurrence after one or more surgical resections, additional radiotherapy was given. Patients must have undergone their last treatment at least 1 year previously. Exclusion criteria were: treatment by stereotactic radiotherapy, or the presence of two or more of the following conditions: cerebrovascular pathology, presence of other tumors of the nervous system, congenital malformations of the nervous system, multiple sclerosis, Parkinson’s disease, organic psychosis (other than dementia), and schizophrenia. Also excluded were patients with optic nerve meningiomas. Patients had to have sufficient command of the Dutch language to be able to carry out the neurocognitive tests. The medical ethics committees of both medical centers approved the study protocol. Eligibility was checked by medical chart review and, if necessary, with the general practitioner.

A total 89 patients with intracranial WHO grade I meningiomas were recruited, of whom 61 (69%) underwent surgery, 4 received radiotherapy only (4%), 21 had surgery and adjuvant conformal external beam radiotherapy (24%) and 3 received neither surgical treatment nor radiotherapy (3%). We invited patients by letter. Informed consent procedures preceded patients’ agreement to participate. Eventually, 11 eligible patients declined to participate; of these patients, 6 underwent surgery, 3 underwent surgery and subsequent radiotherapy, and 2 had not received therapy. The main reason for refusal was that participation was too burdensome. In total 94% of the meningioma patients who underwent RT− and 83% of the RT+ patients were tested at home; the remaining patients were tested in the hospital. Clinical data obtained from medical chart review at entry, included tumor characteristics [histology, location (convexity, tentorium/falx, skull base, orbit), size, hyperostosis, and edema]. The preoperative tumor volume was estimated by assuming an ellipsoid of the orthogonal tumor diameters x, y, and z on CT-scan and/or MRI:

$$ \begin{aligned} {\text{Tumor}}\;{\text{volume}} & = 4/3\pi \ast (1/2x\ast 1/2y\ast 1/2z) \\ {\text{Tumor}}\;{\text{area}} & = \pi \ast (1/2x\ast 1/2y) \\ \end{aligned} $$

Before neuropsychological testing, the patients completed a questionnaire regarding sociodemographic data (including age, sex, and educational level) and a questionnaire on HRQOL, and epilepsy and its treatment [13].

For this study, we selected all patients from the database who underwent surgery and adjuvant radiotherapy (RT+) and matched these patients with the same number of patients from the database who underwent surgery only. Patients were matched for age, sex, and educational level.

Healthy controls

In addition to RT− and RT+ patients, normative data of healthy controls were used as an additional anchor to interpret the results. Healthy controls were drawn from a large, cross-sectional study of the biological and psychological determinants of neurocognitive aging, the Maastricht Aging study [14]. We matched this control group with RT− patients with respect to age, sex, and educational level. Educational level was assessed by a Dutch scoring system consisting of an eight-point scale, ranging from unfinished primary education (level 1) to university education (level 8). In order to compare HRQOL outcomes, healthy controls matched for age, sex, and educational level were drawn from a nationwide study that aimed to translate, validate, and generate normative data on the Short-Form Health Survey (SF-36) for use among Dutch-speaking residents of the Netherlands [15].

Study measures

Health-related quality of life

We assessed patients’ overall degree of physical function with the Karnofsky Performance Status (KPS) scale, which is frequently used in clinical cancer research. Scores range from 0 (lowest score) to 100 (highest level) [16]. The ability to perform daily activities was assessed with the Barthel Activities of Daily Living index [17]. The index consists of ten items (assessing continence of bowel and bladder, grooming, toilet use, feeding, transfer, mobility, dressing, climbing stairs, and bathing); higher scores indicate good functional independence.

Neurological functioning was scored with the neurological functioning scale developed by Order et al. [18]. Scores for this scale range from 1 to 4, with higher scores indicating intact neurological functioning. For self-reported HRQOL we used the MOS Short-Form Health Survey (SF-36) [19]. The SF-36 is composed of 36 items, organized into eight multi-item scales assessing physical functioning (PF), role limitation caused by physical health problems (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role limitation caused by emotional problems (RE), and mental health (MH). Rough scores are converted linearly to 0–100 scales, with higher scores representing better levels of functioning. In addition, we calculated two higher-order compound scores, a physical component scale (PCS) and a mental component scale (MCS). The BCM-20 questionnaire was used to assess additional health problems associated specifically with meningioma and its treatment [20]. Of the 20 BCM items, 13 are organized into 5 subscales assessing future uncertainty, visual disorder, motor dysfunction, communication deficit, and emotional distress. The remaining seven items assess other disease symptoms and side-effects of treatment prevalent among patients with brain tumors, including headaches, seizures, drowsiness, hair loss, itching, weakness of the legs, and lack of bladder control. Since emotional status was already assessed by the SF-36, the 4-item emotional distress scale of the BCM-20 was not analyzed.

Neurocognitive functioning

Because of the different causes and severity of neurocognitive problems, we used a wide range of tests to assess neurocognitive functions. Neurocognitive functions refer to an individual’s ability to perceive, store, retrieve, and use sensory and perceptual information from the environment and past experience, and to such mental activities as planning and organizing. A battery of standard tests was used to assess neuropsychological status. The total time required to complete the battery was approximately 60 min. Appendix 1 provides detailed information on this test battery.

Statistical analysis

Statistical analyses were performed with SPSS (version 11.0). Chi-square tests were used to match the RT− patients with healthy controls for sex. Chi-square tests were also used to show differences in pathological features (e.g. meningothelial, transitional) between RT− and RT+ patients. RT+ meningioma patients were compared with RT− meningioma patients, and RT− meningioma patients were compared with healthy controls for HRQOL and neurocognitive functioning. Student’s t test was used for independent samples to determine whether neurocognitive function and HRQOL of meningioma RT+ patients differed from that of meningioma RT− patients. Student’s t test was also used to determine whether neurocognitive functioning and HRQOL of RT− meningioma patients differed from that of healthy controls. The level of significance was set at P < 0.05.

Results

Patients’ characteristics

Included were 18 patients who underwent surgery followed by radiotherapy (RT+) and these patients were matched with 18 patients who underwent surgery only (RT−). Table 1 presents the sociodemographic characteristics of the RT− and the RT+ patients. Regarding the RT+ patients, adjuvant radiotherapy consisted of conformal external beam fractionated radiotherapy; the radiation dose was 50.4–54.0 Gy in 1.8–2.0 Gy per fraction, five fractions per week, using 6–10 MeV photon beams. None of the 36 patients had clinical or radiological signs of tumor progression.

Table 1 Sociodemographic and clinical characteristics of the study patients

The near optimal levels for neurological functioning (Order scale) and daily living (Barthel index) did not differ significantly between the two groups, although RT+ meningioma patients were significantly more limited in their physical functioning (PCS) than RT− patients. In RT− meningioma patients, more tumors were localized in the convexity and less at the skull base than in RT+ meningioma patients. There were no significant differences in tumor volume, nor in pathology subgroups, between RT− and RT+ patients. Although, follow-up time was significantly different between RT− and RT+ patients, no differences were seen in time since last treatment.

Neurocognitive functioning

Data on neurocognitive functioning and HRQOL are given in Tables 2, 3 and 4. RT+ patients did not have a significantly impaired performance on the Line Bisection test compared to RT− patients, which excludes a major midline deviation (Table 4). In the tests for memory, especially AVLT total recall, AVLT max and AVLT delayed recall, RT− scored significantly worse compared with healthy controls. RT+ meningioma patients did not score differently from RT− meningioma patients on most tests measuring attention and executive functioning. RT+ meningioma patients were slightly faster on Stroop card I and II, but slower on Stroop card III. For RT− meningioma patients no clear differences were seen, except that Stroop card II took more time compared with healthy controls. RT+ patients performed worse on the Fluency test. RT+ patients took less time to accomplish the CST A and CST B test but took more time on the CST C test compared with RT− meningioma patients. RT− meningioma patients needed more time to complete for the CST tests compared with healthy controls.

Table 2 Scores on the HRQOL test of RT− meningioma patients, RT+ meningioma patients, and healthy controls
Table 3 Scores on the BCM-20 test of RT− meningioma patients, RT+ meningioma patients
Table 4 Scores on the neuropsychological tests of RT− meningioma patients, RT+ meningioma patients, and healthy controls

Health-related quality of life

Compared with RT− patients, RT+ meningioma patients scored less well on self-reported HRQOL. RT+ meningioma patients had significantly impaired physical functioning (PF), more role limitations caused by physical health problems (RP), lower Vitality (VT), and lower scores on the Physical Component Scale (PCS). These differences, however, disappeared after correction for the duration of disease. When RT− patients were compared with healthy controls, no significant differences were seen. Scores of the BCM-20 showed no significant differences between RT− and RT+ meningioma patients (Table 3).

Discussion

In this study, no significant differences were found in neurocognitive functioning between WHO grade I meningioma patients that underwent surgery only, and patients that received additional radiotherapy; however, even patients who were only treated surgically had a significantly lower neurocognitive functioning than healthy controls. The most profound neurocognitive disturbances were seen in memory tasks. Meningioma patients who were treated with surgery and radiotherapy had significantly lower HRQOL scores than meningioma patients who were treated surgically only, who had HRQOL scores comparable with healthy controls; these differences, however, disappeared after correction for the duration of disease.

Very few studies have been published on neurocognitive functioning of meningioma patients. Tucha et al. [21] examined neurocognitive functioning before and shortly after surgery for a frontal meningioma. Surgery improved neurocognitive functioning but, compared with healthy controls, significant postoperative neurocognitive deficits remained, particularly a lowered attention span and decreased executive functions. These latter data agree with our long-term results at least 1 year after treatment. Unlike differences in histology and biology of meningioma, some striking results are similar to those reported in previous studies in patients with other types of primary brain tumors [22, 23]. We found neurocognitive disturbances in our meningioma patients similar to those in glioma patients and, similarly, radiotherapy was not associated with poorer neurocognitive outcome.

Despite major neurocognitive deficits, we found no impaired HRQOL in patients who had surgery only. Discrepancies between HRQOL and neurocognitive functioning have been described for other patient groups [24, 25]. There may be limitations in the ability of patients with brain disease and resulting cognitive disturbance to appraise their own situation.

In contrast, we did find a decreased HRQOL in patients who had surgery plus radiotherapy, particularly in the physical component of the SF-36. It may be tempting to attribute the lower physical performance and quality of life to progressive radiation damage in these patients. However, the reverse is more plausible, namely that patients needing adjuvant radiotherapy had larger and more complex meningiomas that inherently cause more cerebral damage, perhaps even aggravated by more extensive surgery. Furthermore, we should keep in mind that the impending threat of tumor recurrence and heavier treatment imposes a psychological burden with resultant anxiety, depression, or fatigue, which can also negatively affect the patient’s neurocognitive function [22]. Patients who had surgery only had a much shorter disease history (on average 3.0 years) than patients who had surgery plus radiotherapy (on average 7.6 years), frequently after repeated earlier surgery. After correction for time after primary diagnosis (post hoc analysis), no significant differences were seen for HRQOL between RT− and RT+ meningioma patients. Most likely, the impaired HRQOL is therefore associated with a longer disease history.

Several limitations of this study should be mentioned. First of all, because we used a retrospective design, we lack a baseline pretreatment assessment. It is possible that the RT+ patients were functioning at a much higher HRQOL level before radiotherapy compared to the RT− patients. This would imply that the outcomes of post-treatment HRQOL should be seen in a different perspective. However, as the decision to add radiotherapy was solely based on surgical grounds, a selection bias on the basis of pre-existent HRQOL level is very unlikely. The anatomical distribution of meningiomas differs between both patient groups. As differences in localization might be associated with differences in vulnerability for the neurocognitive side effects of radiotherapy, e.g. higher risks for tumors involving eloquent brain areas, this should be taken into account when interpreting the neurocognitive scores of both groups. The historical cohort study design may have resulted in a selection bias and confounding bias. For example, patients with severely debilitating disease that precluded testing were excluded from analyses. Also, the number of patients is relatively small which might have influenced the results. On the other hand, this small number enabled to correct for tumor localization and use of anti-epileptic drugs, factors that are known to influence neurocognitive functioning [22].

In conclusion, the results of our study strongly suggest that the addition of radiotherapy has no significant detrimental impact on late neurocognitive functioning in meningioma patients. Our data also indicate that the negative effects on cognition are due to the tumor itself or to surgery. Further study with a prospective study design including baseline scores for QOL and neurocognitive functioning will be necessary to draw definite conclusions regarding the extent and causality of neurocognitive disturbances in meningioma patients.