Predicting peritumoral edema development after gamma knife radiosurgery of meningiomas using machine learning methods: a multicenter study

OBJECTIVES

Edema is a complication of gamma knife radiosurgery (GKS) in meningioma patients that leads to a variety of consequences. The aim of this study is to construct radiomics-based machine learning models to predict post-GKS edema development.

METHODS

In total, 445 meningioma patients who underwent GKS in our institution were enrolled and partitioned into training and internal validation datasets (8:2). A total of 150 cases from multicenter data were included as the external validation dataset. In each case, 1132 radiomics features were extracted from each pre-treatment MRI sequence (contrast-enhanced T1WI, T2WI, and ADC maps). Nine clinical features and eight semantic features were also generated. Nineteen random survival forest (RSF) and nineteen neural network (DeepSurv) models with different combinations of radiomics, clinical, and semantic features were developed with the training dataset, and evaluated with internal and external validation. A nomogram was derived from the model achieving the highest C-index in external validation.

RESULTS

All the models were successfully validated on both validation datasets. The RSF model incorporating clinical, semantic, and ADC radiomics features achieved the best performance with a C-index of 0.861 (95% CI: 0.748-0.975) in internal validation, and 0.780 (95% CI: 0.673-0.887) in external validation. It stratifies high-risk and low-risk cases effectively. The nomogram based on the predicted risks provided personalized prediction with a C-index of 0.962 (95%CI: 0.951-0.973) and satisfactory calibration.

CONCLUSION

This RSF model with a nomogram could represent a non-invasive and cost-effective tool to predict post-GKS edema risk, thus facilitating personalized decision-making in meningioma treatment.

CLINICAL RELEVANCE STATEMENT

The RSF model with a nomogram built in this study represents a handy, non-invasive, and cost-effective tool for meningioma patients to assist in better counselling on the risks, appropriate individual treatment decisions, and customized follow-up plans.

KEY POINTS

• Machine learning models were built to predict post-GKS edema in meningioma. The random survival forest model with clinical, semantic, and ADC radiomics features achieved excellent performance. • The nomogram based on the predicted risks provides personalized prediction with a C-index of 0.962 (95%CI: 0.951-0.973) and satisfactory calibration and shows the potential to assist in better counselling, appropriate treatment decisions, and customized follow-up plans. • Given the excellent performance and convenient acquisition of the conventional sequence, we envision that this non-invasive and cost-effective tool will facilitate personalized medicine in meningioma treatment.

Meningioma involving the superior sagittal sinus: long-term outcome after robotic radiosurgery in primary and recurrent situation

Objective

Treatment for meningiomas involving the superior sagittal sinus (SSS) is challenging and proved to be associated with higher risks compared to other brain locations. Therapeutical strategies may be either microsurgical (sub-)total resection or adjuvant radiation, or a combination of both. Thrombosis or SSS occlusion following resection or radiosurgery needs to be further elucidated to assess whether single or combined treatment is superior. We here present tumor control and side effect data of robotic radiosurgery (RRS) in combination with or without microsurgery.

Methods

From our prospective database, we identified 137 patients with WHO grade I meningioma involving the SSS consecutively treated between 2005 and 2020. Treatment decisions were interdisciplinary. Patients underwent RRS as initial/solitary treatment (group 1), as adjuvant treatment after subtotal resection (group 2), or due to recurrent tumor growth after preceding microsurgery (group 3). Positive tumor response was assessed by MRI and defined as reduction of more than 50% of volume. Study endpoints were time to recurrence (TTR), time to RRS, risk factors for decreased survival, and side effects. Overall and specific recurrence rates for treatment groups were analyzed. Side effect data included therapy-related morbidity during follow-up (FU).

Results

A total of 137 patients (median age, 58.3 years) with SSS meningiomas WHO grade I were analyzed: 51 patients (37.2%) in group 1, 15 patients (11.0%) in group 2, and 71 patients (51.8%) in group 3. Positive MR (morphological response) to therapy was achieved in 50 patients (36.4%), no response was observed in 25 patients (18.2%), and radiological tumor progression was detected in 8 patients (5.8%). Overall 5-year probability of tumor recurrence was 15.8% (median TTR, 41.6 months). Five-year probabilities of recurrence were 0%, 8.3.%, and 21.5% for groups 1-3 (p = 0.06). In multivariate analysis, tumor volume was significantly associated with extent of SSS occlusion (p = 0.026) and sex (p = 0.011). Tumor volume significantly correlated with TTR (p = 0.0046). Acute sinus venous thrombosis or venous congestion-associated bleedings did not occur in any of the groups.

Conclusion

RRS for grade I meningiomas with SSS involvement represents a good option as first-line treatment, occasionally also in recurrent and adjuvant scenarios as part of a multimodal treatment strategy.

Analysis of the common complications and recurrence-related factors of superior parasagittal sinus meningioma

Objectives

Parasagittal meningioma resection is prone to postoperative complications and tumor recurrence because the tumor invades the superior sagittal sinus. This study aimed to clarify the incidence of perioperative complications and the recurrence of superior sagittal paranasal meningiomas and explored potential predictors in this context.

Methods

The study retrospectively reviewed the clinical, imaging, and follow-up data of parasagittal meningiomas among patients who underwent microsurgical resection in the authors' institution from January 2008 to December 2017. Univariate and multivariate logistic regression analyses were conducted to explore independent predictors of perioperative complications and tumor recurrence.

Results

A total of 212 parasagittal meningioma patients were included in this study. The incidence of perioperative complications was 23.6% (50/212), and perioperative death occurred in 6 (2.8%) patients. In univariate and multivariate logistic regression analyses of perioperative complications, peritumoral edema ≥1 cm (odds ratio [OR] 2.163, 95% confidence interval [CI] 1.057-4.428, P = 0.035) and the Sindou invasion Class V-VI(OR0.018, 95% CI 1.248-11.064, P = 0.018) were independent predictors. After an average of 83 (39-154) months of clinical follow up among the living 206 patients, 22 (10.7%) patients showed tumor recurrence. In univariate and multivariate logistic regression analyses of tumor recurrence, the Sindou invasion Class III-IV (OR 5.539, 95%CI 1.469-20.884, P = 0.011) and the Sindou invasion Class V-VI (OR 9.144, 95%CI 2.215-37.757, P = 0.002) were independent predictors.

Conclusions

Peritumoral edema ≥1 cm and the Sindou invasion Class V-VI were the independent predictors of perioperative complications, and the Sindou invasion Class III-IV and the Sindou invasion Class V-VI were the independent predictors of tumor recurrence. The part of the parasagittal meningioma involving the sinus wall should be actively removed to the largest degree possible to reduce the recurrence rate.

Long-Term Outcomes After Linac Radiosurgery for Benign Meningiomas

AIMS

Although several studies on outcomes following stereotactic radiosurgery (SRS) for benign meningiomas have been reported, Linac-based SRS outcomes have not been as widely evaluated. The aim of this retrospective institutional single-centre study was to determine long-term outcomes of Linac-based SRS for benign intracranial meningiomas.

MATERIALS AND METHODS

From July 1996 to May 2011, 60 patients with 69 benign meningiomas were included. All patients were treated with single-fraction Linac-based SRS with four to five non-coplanar arcs, dynamic or not. The marginal dose prescribed for the periphery was 16 Gy. Prognostic factors associated with local control, progression-free survival (PFS) and overall survival were tested.

RESULTS

The median follow-up was 128 months. No patient was lost to follow-up. The values observed at 1, 5 and 10 years were, respectively, 100%, 98.4% and 92.6% for local control, 94.9%, 93.2% and 78% for PFS and 100%, 94.7% and 92.7% for overall survival. In univariate analysis, local control after SRS was significantly higher for skull base and parasagittal meningiomas compared with convexity meningiomas (P = 0.031). Multivariate analyses showed significantly longer PFS when the minimum dose delivered to the tumour was greater than 10 Gy (P = 0.0082). No grade 5 toxicity was reported.

CONCLUSION

Our long-term results from a large sample size of benign meningiomas treated with Linac-based SRS confirmed excellent local control (>90%) and good safety, which is in line with published studies on Gamma Knife surgery. Above all, we showed significantly poorer PFS if the minimum dose to the tumour was under 10 Gy.

Linear Accelerator-Based Radiosurgery of Grade I Intracranial Meningiomas

Objective

To determine the local control rate and complication rate in the treatment of grade I intracranial meningiomas.

Methods

A retrospective study was performed of patients with grade I meningioma who received radiosurgery with a dedicated linear accelerator from January 2002 to August 2012 with a minimum follow-up of 2 years. We performed descriptive statistics, logistic regression, and progression-free survival analysis through a Kaplan-Meier curve.

Results

Seventy-five patients with 78 grade I meningiomas received radiosurgery, 39 underwent surgery plus adjuvant radiosurgery, and 36 only radiosurgery. The follow-up median time was 68 months (range, 35–120 months). The tumor control rate was 93%, the 5-year progression-free survival was 92% (95% confidence interval, 77%–98%). Acute toxicity was 2.6%, and grade 1–2 late toxicity was 26.6%. Postradiosurgery edema was the main late morbidity. Age >55 years was the only significant factor for attaining a response >75%. The background of surgery before radiosurgery was the only significant prognostic factor for showing edema (odds ratio 5.78 [95% confidence interval, 2.14–15.64]).

Conclusions

The local control rate attained in our series is similar to that reported in other series worldwide; the acute toxicity rate was low and late toxicity was moderate.

Radiation-Induced Edema After Single-Fraction or Multifraction Stereotactic Radiosurgery for Meningioma: A Critical Review

PURPOSE

Potential dosimetric and clinicopathologic predictors of radiation-induced brain edema after single-fraction or multifraction stereotactic radiosurgery (SRS) for non-base of skull (non-BOS) meningiomas are summarized based on a systematic review of the published literature.

METHODS AND MATERIALS

Reviewed studies (PubMed indexed from 1998 through 2017) included all or some non-BOS meningioma patients, reported risks of edema after SRS, and correlated dosimetric and/or nondosimetric measures with the magnitude of risk.

RESULTS

Twenty-six studies reporting risks of edema after SRS for meningioma are reviewed. The treatment techniques as well as distribution of tumor locations, target dosing, and target volume varied across studies. Among 13 studies that included only non-BOS tumors or separately grouped non-BOS tumors, symptomatic edema occurred in 5% to 43% of patients and any edema occurred in 28% to 50%. The reported average time to onset of edema ranged from approximately 3 to 9 months in most studies. Factors reported to significantly correlate with increased risks of edema and/or symptomatic edema after SRS for meningioma include the following: greater tumor margin and/or maximum dose, greater tumor size and/or volume, non-BOS (particularly parasagittal) location, no prior resection for meningioma, and presence of pretreatment edema. Nevertheless, the extent and significance of these factors were inconsistent across studies. Potentially important dosimetric factors, such as volume of brain or tissue receiving single-fraction doses > 10 to 12 Gy, are not well studied.

CONCLUSIONS

The variability in risks of edema and in factors impacting those risks is likely a result of differences across studies in the clinicopathologic characteristics of the patient populations, as well as differences in treatment modalities and SRS planning and delivery parameters. More studies on pooled populations, grouped by potential prognostic factors such as tumor location and prior therapy, are needed to better understand dosimetric and nondosimetric factors predictive of edema risk after SRS for meningioma.

Multimodal treatment of parasagittal meningiomas: a single-center experience

OBJECTIVE

Treatment of parasagittal meningiomas is still considered a challenge in modern microsurgery. The use of microsurgical resection, radiosurgery, or a microsurgery-radiosurgery combination treatment strategy is often debated. The aim of this study was to evaluate the treatment of parasagittal meningioma and provide evidence that a multimodal approach reduces complication rates and achieves good tumor control rates.

METHODS

The authors retrospectively reviewed long-term follow-up data on 117 patients who had been treated for parasagittal meningiomas at their institution between 1993 and 2013. Treatment included microsurgery, Gamma Knife radiosurgery (GKRS), and radiotherapy.

RESULTS

The median tumor volume prior to the first microsurgical resection was largest in the microsurgery-radiosurgery combination treatment group. Invasion of the superior sagittal sinus was significantly associated with a Simpson Grade IV resection and subsequent radiosurgery treatment. The Simpson resection grade did not influence time to progression or recurrence in benign meningioma cases. Complete sinus occlusion was followed by microsurgical resection of the occluded sinus, by tumor resection without resection of the sinus, or by GKRS. Histopathology revealed WHO Grade I tumors in most patients. However, a high percentage (33%) of atypical or malignant meningiomas were diagnosed after the last microsurgical resection. The time to recurrence or progression after microsurgery was significantly longer in patients with WHO Grade I meningiomas than in those with Grade II or III meningiomas. At follow-up, tumor control rates after GKRS were 91% for presumed meningioma, 85% for benign meningioma, 71% for atypical meningioma, and 38% for malignant meningioma.

CONCLUSIONS

A multimodal treatment approach to parasagittal meningiomas reduces the rate of complications. Thus, microsurgery, radiotherapy, and radiosurgery are complementary treatment options. Gamma Knife radiosurgery is safe and effective in patients with meningiomas invading the superior sagittal sinus. The procedure can be part of a multimodal treatment plan or administered as a single treatment in well-selected patients.