Early motor function after local treatment of brain metastases in the motor cortex region with stereotactic radiotherapy/radiosurgery or microsurgical resection: a retrospective study of two consecutive cohorts

Two prospectively documented institutional databases, 1) the register for intraoperative neuromonitoring of the Department for Neurosurgery, University of Bonn Medical Center, and 2) the database for SRT/SRS of the Department for Radiotherapy, the MediClin Robert Janker Clinic (Bonn, Germany), were retrospectively searched for cases of centrally-located cerebral metastases, details of the selection process are presented in the flow chart (Fig. 1).

We performed a retrospective analysis of 2061 cases of micro-surgically operated patients under electrophysiological control, as well as 428 cases of stereotactic radiotherapy patients during the period of January 2008 to September 2012. The particular form of treatment was decided in a weekly multidisciplinary neuro-oncology tumor board with permanent participation of senior consultants from: the neurosurgery, radiotherapy, neuroradiology, neurooncology/neurology and was conformed to current guidlines for the treatment of cerebral metastases and the latest scientific evidence. In the case of centrally located metastases, the prognosis was initially estimated based on known scores (for example, SIR: [20, 21], and the less invasive option of stereotactic irradiation was generally used primarily in patients with a survival prognosis of less than half a year. Stereotactic biopsies were not used in this series and in generally only exceptionally in suspected metastases. In the case of an unclear primary or a missing histology, a resection was recommended for histological evaluation. Otherwise the patients were offered both therapy options. After clarification on the advantages and disadvantages of both treatment options, the patients could usually choose which therapy they prefer.

For metastases with a volume of ≤4.5 cm³ and a 10 Gy volume of ≤10 cm³, a classic one session radiosurgery with a marginal dose of 20 Gy was used. In case of metastases >4.5 cm³ or a 10 Gy volume > 10 cm³, the hypofractionated variant was used (see Table 2).

MRS was performed over a free craniotomy under general anesthesia with intravenous analgosedation. Intraoperative electrophysiology was used to identify the senso-motoric cortex. MRS was then monitored by somatosensory evoked potential and motoric evoked potentials.

Cerebral metastases located within either the precentral gyrus of the motor cortex, or adjacent to the precentral gyrus in the fronto-dorsal area or the postcentral area, were included for analysis. The motor strip was identified on magnetic resonance imaging (MRI) by using standard anatomical cues, including presence of an “omega sign” in the central sulcus, termination of the superior frontal sulcus into the prefrontal sulcus, and the paracentral lobule located directly anterior to the pars marginalis. In addition, intraoperative and electrophysiologically identified cases using the phase reversal technique were included [22, 23].

Patients with no follow-up after treatment or with other cerebral metastasis that could affect the motor pathway like basal ganglia, brain stem, and spinal metastases were excluded from this study (see Fig. 1).

Motor function was prospectively analyzed and documented both pre- and post-procedure by each department at 1 day before and over 1–3 weeks intervals after either MSR or SRT/SRS. Motor function was graded for all patients according to the BMRC scale [11], as: normal (5/5); mild weakness (4/5); moderate weakness with the strength to overcome gravity (3/5); weakness with not enough strength to overcome gravity, but with notable movement (2/5); severe weakness with not visually but palpable movement (1/5); and absolute weakness or plegia (0/5). Hemiparesis was defined as a grade lower than 5/5 independent of the underlying reason caused by each technique (edema, hemorrhage, empyema, or collateral damage).

Adverse events were retrospectively evaluated and taken in account up to 3 months after the initial therapy, by documentation in patient records or reported by practitioners afterwards. We divided the adverse events into local complications directly associated with the treatment such as bleeding, infection, radio-necrosis, and systemic adverse events not directly associated with the local treatment such as pneumonia, thrombosis, and pulmonary embolism.

Univariate and multivariate statistical analyses were performed to determine which patient and tumor-related factors influenced functional outcome. Factors evaluated included tumor volume, patient age, tumor localization, if the tumor directly involved the motor cortex, KPI, dexamethasone daily dose, and preoperative motor deficit. We evaluated the motor function outcome in patients with and without deficits prior to each treatment. As a surrogate parameter for clinically-relevant edema, we examined dexamethasone treatment during follow-up.

These statistical comparisons were analyzed using Fisher’s exact test and Student’s t-test. Multivariate analyses were performed as binary logistic regression analysis with IBM SPSS Statistics version 20.0 (IBM Corp., North Castle, NY, USA). A p-value of p ≤ 0.05 was considered statistically significant.

For treatment with MSR, there seems to be a significant difference regarding the recovery of motor deficits caused by central metastases as compared with SRT/SRS. However, this recovery may also depend on other factors, such as the exact neuroanatomical localization of the metastases. From the standpoint of neurosurgery, it certainly impacts recovery if a metastasis is located either directly in the precentral gyrus or only adjacent to it. In our study, a localization of the metastasis adjacent to the precentral gyrus revealed only a statistical trend toward better improvement of the motor deficit in the multivariate analysis, but not in the univariate analysis.

The electrophysiologically controlled, intra-operative identification of the pyramidal tract is standard practice and was carried out consistently in our study. The pre-operative identification of the pyramidal tract by navigated transcranial magnetic stimulation (nTMS) [24], is increasingly utilized over the last 5 years, which implies that this technique was not available during the treatment period of our study (2008–2012). If such pre-operative methods can demonstrate that a metastasis grows directly in the pyramidal tract, such critical cases could be excluded early from operation. Alternatively, a calculated hemiparesis risk could be discussed with the patient. Moreover, this could show a reduced quality of life and therefore be declined as a treatment for these patients. These critical patients could then choose SRT/SRS as an alternative treatment.

Due to the invasive nature of MSR, our results demonstrate a higher rate of local complications such as bleeding, infection, or infarction. We believe that the precise preoperative selection of patients with a disposition for local complications is a crucial factor to improve surgical outcomes. This screening should include the exclusion of adverse blood clotting and a comprehensive evaluation of the patient’s immune system, which is supported by the neurosurgical literature [25‐27]. The reduction of the local complication rate remains vital, as the start of most adjuvant therapy depends on rapid and complete wound healing and recovery post-surgery. In addition, by further improving the MSR technique and intraoperative monitoring, microsurgery should improve in both general and motor deficit outcomes.

A topic we do not assess in our study is local tumor recurrence for each treatment option. Conform to the guidelines for cerebral metastases for all patients a postoperative adjuvant radiation therapy was recommended and performed depending on the postoperative KPI. However, post-op radiation therapy was usually performed with a delay of 2–4 weeks after MSR and we did not assess these data because we focused on the early functional outcome. Recent publications cautioned not to underestimate local recurrence after microsurgery, especially if a supramarginal resection cannot be administered [28, 29]. This is particularly important in the critical central region, since resection must be conservative in this area of the brain. Microsurgery of cerebral metastases in the Rolandic area may therefore be inferior to local radiotherapy, regarding local tumor recurrence. This inferiority of MSR in comparison to SRT/SRS without adjuvant therapy has been previously described [7]. By future improvements in adjuvant systemic therapy, the progression-free survival in these cases may be less dependent on whether local therapy utilizes radiotherapy or microsurgery.

Due to early detection by routine MRI staging, as well as the MRI screening during follow-up, the majority of central metastases may be small and asymptomatic upon discovery. Especially for these types of metastases, SRT/SRS seems to be the optimal choice according to our data. However, treatment of central metastases by SRT/SRS also has some disadvantages. First, SRT/SRS failed to improve motor function in patients with pre-existing motor deficits in our study. Second, and possibly related to the first, SRT/SRS has a significantly higher dose and longer duration of edema therapy, as revealed by administration of dexamethasone (SRT/SRS vs. MSR = 9.5 mg vs. 4.5 mg, p = 0.001). The higher and prolonged dexamethasone therapy in SRT/SRS patients may have implications beyond local impact by also having a negative systemic influence. Dexamethasone was individually administered and adapted to the neurological symptomatics (in particular the motoric function) of the patients. The highest daily dose of 24 mg was reduced regularly for about 1–2 weeks. However, the more intense dexamethasone therapy in SRT/SRS patients may be one explanation for the positive correlation between the higher rate of systemic complications with SRT/SRS observed in our study. Therefore, an improvement in edema therapy by reducing dexamethasone dose and the adoption of new molecular anti-edematous treatment approaches, such as anti-VEGF therapy [30, 31], may improve outcomes for stereotactic radiotherapy of central metastases in the future. Another explanation for the higher prevalence of systemic complications in SRT/SRS patients could be a higher proportion of ill and/or and elderly patients than found in the MSR cohort. In our study, however, there were no significant differences concerning age and KPI, with only a weak correlation between systemic complications and age in the multivariate analysis. Another key factor for improvement in the treatment outcomes of stereotactic radiotherapy could be the optimization of radiotherapy fractionation, single fraction versus hypofractionation, to reduce the rate of radiogenic edema and necrosis. The tumor volume dependence of radio-necrosis has been proven in stereotactic radiosurgery [32]. Hypofractionated stereotactic techniques are increasingly used as an alternative option in cases with a large treatment volume, in an effort to protect the surrounding normal tissue [33].

Our study had some limitations. The results of this study should be interpreted with caution, due to its retrospective design. We are aware of the differences regarding tumor size and motor deficits before treatment and made efforts to taken them into account in the statistical analyses. We considered only the restricted collective with motor deficits before treatment concerning the calculation of improvement of motor deficits. In addition, we controlled all our univariate results by multivariate regression analyses. Moreover, there is a patient selection bias due to the guidelines for SRS, which usually restricts treatment to tumor sizes smaller than 2.5 cm in diameter, or 4.5 cm³ in volume. Despite this selections bias, we consider our results a valuable contribution, since they emphasize an often overlooked outcome parameter for the treatment of cerebral metastases, which is very important for patient quality of life: the motor function and a method to quantify and evaluate this important neurological function.

As cancer becomes more a chronic disease [1‐3], central cerebral metastases may become more common. As a result, neurological functional outcome may gain clinical importance in the future, in addition to oncological outcomes [4‐6]. New imaging techniques may improve the detection and the resolution of metastases and the surrounding structures. This is important for the neuronavigation in microsurgery and for the target volume defined for radiotherapy. Moreover, these diagnostic techniques might be able to give molecular properties, which can be taken into consideration for future treatment decisions. Furthermore, new bio-molecular drugs [30, 31] and bio-molecular materials [34], may improve the overall oncological outcomes, as well as control the side effects of both microsurgery [34‐36] and stereotactic radiotherapy [37] in the future.