Between 1989 and 2021, a total of 18 patients ≤ 21 years of age underwent resection of a meningioma at our department. Table 1 shows the baseline characteristics of the cohort with a median age of 15.5 years and a male:female ratio of 10:8. One patient with a large skull base meningioma was 21 years and 2 months of age at the time of surgery; however, onset of symptoms was 7 months before surgery, so the patient was included. The most frequent locations of meningiomas were the skull base (33%) and convexity (27%), followed by intraventricular and spinal tumors, whereas meningiomas rarely originated from other locations, such as the falx, parenchyma or optic nerve sheath. Three patients had a history of prior radiation exposure at the site of the tumor, one located at the falx, one at the convexity and one spinal at the C4 level. Multiple meningiomas were present in three patients, one of which was diagnosed with neurofibromatosis II (NF II).

The most common symptom at time of diagnosis were epileptic seizures (33%), which was clearly associated with convexity location (p = 0.025) and 4 patients (22%) presented with cranial nerve palsy. Other clinical findings were headache, signs of increased intracranial pressure (ICP) and absence of neurological symptoms as listed in Table 1. The median KPS at presentation was 80 (20–100).

Data on treatment and outcome are shown in Table 2. All 18 patients underwent resection of a meningioma, in most of the patients (67%) GTR (i.e., Simpson grade I or II) resection was documented in the operating report and Simpson grade IV in the remaining 6 cases (33%); however, 9/18 patients (50%) underwent a second resection due to tumor relapse and 1 patient (5%) had a third resection. Two patients with relapse did not undergo a second operation, one patient rejected further surgery and one was treated with gamma knife radiation only. In half of the cases of skull base meningiomas, Simpson grade IV only was achieved compared to grade I in all cases of convexity meningiomas. In cases of subtotal resection, adherence to major blood vessels and/or brain stem was documented in the operative report.

In total 8 patients (44%) experienced surgery-related complications, including 4 cases of shunt-dependent hydrocephalus, 1 wound infection, 1 postoperative hemorrhage and 5 cases of new neurological deficits after surgery. Of note, all 4 cases of postoperative shunt-dependent hydrocephalus occurred in patients with skull base meningiomas and in 1 patient with an intraventricular meningioma. Furthermore, 4 of 5 cases of new neurological symptoms were assigned to skull base meningiomas and 1 case to an optic nerve sheath meningioma. One case of hemorrhage was documented in a spinal meningioma. One patient with a skull base meningioma had to undergo revision surgery due to wound infection. There were no complications in patients with convexity meningiomas. The difference in incidence of complications between skull base and other location was significant (odds ratio [OR] 15.0, 95% confidence interval [CI] 1.2–185.2, p = 0.043, see Table 3).

Of the patients 2 (11%) received radiotherapy, 3 (17%) received gamma knife treatment, and 2 patients (11%) underwent chemotherapy in addition to surgery including bevacizumab, gemcitabine, interferon and trabectedine. One patient (5%) underwent implantation of iodine-125 seeds. The indications for such adjuvant treatments were limited to inoperable remnant or relapse of an atypical or anaplastic meningioma.

The median overall time of follow-up (FU) was 123 months (range 26–335 months) and no patients were lost to FU. The overall median KPS at time of the last FU was 85 (0–100). In total 4 patients died according to a death register comparison. All of the 4 patients died due to relapse of a tumor WHO grade 2 or 3. Of note, clinical outcome was clearly dependent on tumor location: As depicted in Table 2 none of the patients with a skull base meningioma improved in KPS score during the observation period, compared to all of the patients with a convexity meningioma. Conversely, the overall KPS of patients with skull base meningioma substantially declined during follow-up. The median KPS score at time of last FU in skull base meningioma patients was 15 (0–90) compared to a median score of 100 (0–100) in patients with tumors in other locations (p = 0.009). This difference was not seen preoperatively, when patients in the skull base group had a median KPS score of 85 (70–90) compared to 80 (20–100) in the other patients (p = 0.553), see Table 3. In total, 6 of 18 patients (33%) eventually had an unfavorable outcome (= KPS < 80); however, 3 of these patients initially had a favorable outcome but then deteriorated slowly and gradually during follow-up, starting after a median time of 145 months (80–309 months) after surgery due to an inoperable tumor relapse. One patient with unfavorable outcome had NF II and suffered preoperatively from multiple cranial nerve palsies due to large bilateral vestibular schwannomas. Hence, the patient was not able to improve in KPS after resection of the meningioma. Furthermore, unfavorable outcome in 2 patients (KPS 30 and 70) resulted from complications leading to new neurological symptoms after surgery.

In total, 6 patients (33%) suffered from seizures before surgery, 4 of whom were patients with convexity meningioma. All patients with convexity meningioma were seizure-free (Wieser class I) at time of last FU. In total, 3 patients suffered from seizures at time of last FU, including 2 cases (WHO 2 and 3) who developed de novo seizures during follow-up and one case of a recurrent intraparenchymal meningioma which was classified Wieser class IV.

The WHO grades and specific histological diagnoses are listed in Table 2. In total, 56% were classified as WHO 1, 39% WHO 2 and 5% WHO 3 as per last histological diagnosis. Immunostaining revealed SMARCE1 mutations in both cases of clear cell meningiomas. When the two subgroups skull base and convexity meningioma were compared, the histological subtype was associated with tumor location (p = 0.012). Thus, tumors located in the skull base were primarily either classified as meningotheliomatous (WHO 1) or clear cell meningiomas (WHO 2). In contrast, histological findings of convexity meningiomas revealed either a fibroblastic (WHO 1) or atypical (WHO 2) subtype. Furthermore, as shown in Table 2, clinical outcome in all patients with convexity meningiomas was favorable, regardless of the WHO grade (p = 0.628). In contrast, clinical outcome was unfavorable in patients with skull base meningioma of WHO grade 2 or 3 but favorable in those with WHO grade 1 (p = 0.037).

Of note, in one case, which was diagnosed with a meningotheliomatous subtype (WHO 1) after first and second surgeries, transformation into an anaplastic meningioma (WHO 3) had been observed 23 years after initial surgery, when a biopsy was done.

The median time from surgery to last radiological FU was 109 months (11–335 months) and no patient was lost to radiological FU. In total, in 11 of 18 patients (61%) tumor relapse was seen during follow-up after the initial resection. Hence, in these cases, either growth of a remnant or recurrence after gross total resection (GTR) were documented. As shown in Table 3, relapse occurred significantly more often in cases of skull base location (p = 0.038) compared to other locations and significantly less often in cases of convexity location (p = 0.047) compared to other locations. Furthermore, relapse among the whole cohort was significantly associated with subtotal resection, i.e., Simpson grade IV (p = 0.038) but not with higher WHO grades (p = 0.709). Figure 1 shows the overall relapse-free survival with a median time to tumor recurrence of 79.0 months (39.9–118.1 months). Actual rates of relapse-free survival according to Simpson grade, WHO grade and tumor location are depicted in Figs. 2, 3 and 4, respectively, showing a trend but no significant difference in the former (p = 0.062) and no association in the latter two (p = 0.507 and p = 0.187).

Graph showing relapse-free survival of the whole cohort with an actuarial median time to relapse of 79 months (55–103 months)

Graph showing relapse-free survival according to extent of resection. A trend could be observed towards earlier relapse in patients who underwent subtotal resection (median time to relapse 69 months vs. 145 months, log-rank: 0.062)

Graph showing relapse-free survival according to WHO grade. No difference was observed between the two groups (78 vs. 118 months, log-rank: 0.507)

Graph showing relapse-free survival in patients with convexity meningiomas compared to other locations. Median time to relapse was 78 months (63–93 months) in patients with locations other than convexity. No median time to relapse is given for convexity meningiomas. According to log-rank test, no significant difference was seen (p = 0.187)

As shown in Table 2, half of the patients had to undergo repeat surgery due to relapse. After last surgery, at time of last follow-up, relapse was seen in three of those patients (33%). In all of these cases gross total resection was not possible due to infiltrative tumor growth affecting structures such as cavernous sinus, brain stem and the middle cerebral artery. In two of the patients, further surgery was abandoned due to poor clinical status; however, one patient received adjuvant gamma knife treatment after second surgery. The third patient who had an intraparenchymal meningioma, refused to undergo repeat surgery for personal reasons.