Machine learning-based radiomic evaluation of treatment response prediction in glioblastoma.

AIM: To investigate machine learning based models combining clinical, radiomic, and molecular information to distinguish between early true progression (tPD) and pseudoprogression (psPD) in patients with glioblastoma. MATERIALS AND METHODS: A retrospective analysis was undertaken of 76 patients (46 tPD, 30 psPD) with early enhancing disease following chemoradiotherapy for glioblastoma. Outcome was determined on follow-up until 6 months post-chemoradiotherapy. Models comprised clinical characteristics, O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation status, and 307 quantitative imaging features extracted from enhancing disease and perilesional oedema masks on early post-chemoradiotherapy contrast-enhanced T1-weighted imaging, T2-weighted imaging (T2WI), and apparent diffusion coefficient (ADC) maps. Feature selection was performed within bootstrapped cross-validated recursive feature elimination with a random forest algorithm. Naive Bayes five-fold cross-validation was used to validate the final model. RESULTS: Top selected features included age, MGMT promoter methylation status, two shape-based features from the enhancing disease mask, three radiomic features from the enhancing disease mask on ADC, and one radiomic feature from the perilesional oedema mask on T2WI. The final model had an area under the receiver operating characteristics curve (AUC) of 0.80, sensitivity 78.2%, specificity 66.7%, and accuracy of 73.7%. CONCLUSION: Incorporating a machine learning-based approach using quantitative radiomic features from standard-of-care magnetic resonance imaging (MRI), in combination with clinical characteristics and MGMT promoter methylation status has a complementary effect and improves model performance for early prediction of glioblastoma treatment response.

Pituitary metastases: presentation and outcomes from a pituitary center over the last decade.

PURPOSE: Highlight and characterize manifestations, diagnostic/management approaches and outcomes in a contemporary cohort of patients with pituitary metastases (PM) from a large European pituitary center-over 10 years. METHODS: Retrospective review of PM cases between 1/2009 and 12/2018. Clinical, laboratory, imaging data at PM detection and during follow-up were analysed. RESULTS: 18 cases were identified (14 females; median age at diagnosis 61.5 years). Most common primary malignancies were lung (39%) and breast (32%). Most frequent presenting manifestation was visual dysfunction (50%). Gonadotrophin, ACTH, TSH deficiency were diagnosed in 85%, 67%, 46% of cases, respectively; diabetes insipidus (DI) was present in 17%. 33% of cases were detected during investigation for symptoms unrelated to PM. PM management included radiotherapy (44%), transsphenoidal surgery (17%), transsphenoidal surgery and radiotherapy (6%) or monitoring only (33%). One-year survival was 49% with median survival from PM detection 11 months (range 2-47). CONCLUSIONS: In our contemporary series, clinical presentation of PM has evolved; we found increased prevalence of anterior hypopituitarism, decreased rates of DI and longer survival compared with older literature. Increased availability of diagnostic imaging, improvements in screening and recognition of pituitary disease and longer survival of patients with metastatic cancer may be contributing factors.

Dexamethasone-related adrenal insufficiency in patients with brain and skull base tumours.

PURPOSE: This study aimed to evaluate the prevalence of glucocorticoid-induced adrenal insufficiency in a cohort of patients with brain and skull base tumours and to identify factors which may predict its occurrence. METHODS: Patients with brain or skull base tumours attending for a short synacthen test (SST) (adrenocorticotropin hormone (ACTH) stimulation test) at a single institution over a 3-year period were retrospectively identified. Baseline demographics and dexamethasone exposure were examined. Only patients with dexamethasone exposure were included in the final analysis looking at the primary end point of SST failure. Fisher's exact test, Student's t test, Mann-Whitney test and the Kendall's tau-b test were used to evaluate the influence of age, gender, diagnosis and mean pituitary radiation dose on the primary endpoint. Receiver operating characteristic (ROC) curves were generated to explore the impact of duration and total exposure to dexamethasone on likelihood of SST failure. RESULTS: Thirty-one of 51 patients with previous dexamethasone exposure failed their first SST (61%). No significant relationship was demonstrated between age, gender, diagnosis or mean pituitary radiation dose and SST failure. Duration of and total exposure to dexamethasone were significantly associated with SST failure (p = 0.001 and p = 0.007, respectively). ROC curves generated values of 78 days and 171 mg days to give a sensitivity of 94 and 97%, respectively, to detect SST failure. CONCLUSIONS: Duration of dexamethasone use and total exposure predict for adrenal insufficiency in patients with brain and skull base tumours. Values derived from this study may be useful to identify patients at higher risk of adrenal suppression who require empirical hydrocortisone pending formal testing of the hypothalamic-pituitary-adrenal axis.

A Randomised Phase II Trial of Hippocampal Sparing Versus Conventional Whole Brain Radiotherapy After Surgical Resection or Radiosurgery in Favourable Prognosis Patients With 1-10 Brain Metastases.

AIMS: To assess in patients with 1-10 brain metastases, each of which has been treated by neurosurgery or stereotactic radiosurgery, whether hippocampal sparing whole brain radiotherapy (HS-WBRT) better spares neurocognitive function (NCF) than standard WBRT. Further, to assess whether a phase III randomised trial of HS-WBRT would be feasible in the UK. MATERIALS AND METHODS: A multicentre, randomised, open label phase II trial was undertaken, randomising patients to 30Gy in 10 fractions of WBRT or HS-WBRT. The primary endpoint was decline in Total recall using Hopkins Verbal Learning Test Revised (HVLT-R) at 4 months post treatment. To assess this, we aimed to recruit 84 patients over 3 years. Secondary endpoints included further measures of NCF, quality of life, duration of functional independence, local control of treated metastases, development of new metastases, disease control within the hippocampal regions, overall survival, steroid and antiepileptic medication requirements, and toxicity. RESULTS: The trial closed prematurely due to slower than anticipated recruitment. From April 2016 to January 2018, 23 patients were randomised. Follow up was a median of 25 months. Fifteen patients (6 WBRT, 9 HS-WBRT) were assessed for the primary endpoint; of these, 1 in each arm experienced significant decline in the 4-month HVLT-R Total recall score (p = 0.8). Patients in the HS-WBRT arm experienced less insomnia (p < 0.01) and drowsiness (p < 0.01). There were no differences in other secondary endpoints. CONCLUSION: A phase III randomised trial of HS-WBRT was shown not to be feasible at this time in the UK. As most randomised trials of HS-WBRT reported to date share common endpoints, including NCF, an individual patient data meta-analysis should be undertaken.

UK 2022 Consensus on Normal Tissue Dose-Volume Constraints for Oligometastatic, Primary Lung and Hepatocellular Carcinoma Stereotactic Ablative Radiotherapy.

The use of stereotactic ablative radiotherapy (SABR) in the UK has expanded over the past decade, in part as the result of several UK clinical trials and a recent NHS England Commissioning through Evaluation programme. A UK SABR Consortium consensus for normal tissue constraints for SABR was published in 2017, based on the existing literature at the time. The published literature regarding SABR has increased in volume over the past 5 years and multiple UK centres are currently working to develop new SABR services. A review and update of the previous consensus is therefore appropriate and timely. It is hoped that this document will provide a useful resource to facilitate safe and consistent SABR practice.

Improvement in Dysphagia Outcomes Following Clinical Target Volume Reduction in the De-ESCALaTE Study.

AIMS: The De-ESCALaTE study showed an overall survival advantage for the administration of synchronous cisplatin chemotherapy with radiotherapy in low-risk oropharyngeal cancer when compared with synchronous cetuximab. During the trial, a radiotherapy quality assurance protocol amendment permitted centres to swap from the original radiotherapy contouring protocol (incorporating the whole oropharynx into the high-dose clinical target volume (CTV); anatomical protocol) to a protocol that incorporated the gross tumour volume with a 10 mm margin into the CTV (volumetric protocol). The purpose of this study was to examine both toxicity and tumour control related to this protocol amendment. MATERIALS AND METHODS: Overall survival and recurrence at 2 years were used to compare tumour control in the two contouring cohorts. For toxicity, the cohorts were compared by both the number of severe (grades 3-5) and all grades acute and late toxicities. In addition, quality of life and swallowing were compared using EORTC-C30 and MD Anderson Dysphagia Inventory, respectively. RESULTS: Of 327 patients included in this study, 185 were contoured according to the anatomical protocol and 142 by the volumetric protocol. The two cohorts were well balanced, with the exception of significantly more patients in the anatomical cohort undergoing prophylactic feeding tube insertion (P < 0.001). With a minimum of 2 years of follow-up there was no significant difference in overall survival or recurrence between the two contouring protocols. Similarly, there was no significant difference in the rate of reported severe or all grades acute or late toxicity and no sustained significant difference in quality of life. However, there was a significant difference in favour of volumetric contouring in several domains of the MD Anderson Dysphagia Inventory questionnaire at 1 year, which persisted to 2 years in the dysphagia functional (P = 0.002), dysphagia physical (P = 0.009) and dysphagia overall function (P = 0.008) domains. CONCLUSION: In the context of the unplanned post-hoc analysis of a randomised trial, measurable improvement in long-term dysphagia has been shown following a reduction in the CTV. Further reductions in the CTV should be subject to similar scrutiny within the confines of a prospective study.

Evaluation of Response to Stereotactic Radiosurgery in Brain Metastases Using Multiparametric Magnetic Resonance Imaging and a Review of the Literature.

AIMS: Following stereotactic radiosurgery (SRS), brain metastases initially increase in size in up to a third of cases, suggesting treatment failure. Current imaging using structural magnetic resonance imaging (MRI) cannot differentiate between tumour recurrence and SRS-induced changes, creating difficulties with patient management. Combining multiparametric MRI techniques, which assess tissue physiological and metabolic information, has shown promise in answering this clinical question. MATERIALS AND METHODS: Multiparametric MRI techniques, including spectroscopy, diffusion and perfusion imaging, were used for the differentiation of radiation-related changes and tumour recurrence after SRS for intracranial metastases in six cases. All patients presented with enlargement of the treated lesion, an increase in perilesional brain oedema and aggravation or appearance of neurological signs and symptoms from 7 to 29 weeks after primary treatment. RESULTS: Multiparametric imaging helped to differentiate features of tumour progression (n = 4) from radiation-related changes (n = 2). A low apparent diffusion coefficient (ADC) <1000 × 10-6 mm2/s, high relative cerebral blood volume (rCBV) ratio > 2.1, high choline:creatine (Cho:Cr) ratio > 1.8 suggested tumour recurrence. A high ADC > 1000 × 10-6 mm2/s, low rCBV ratio < 2.1, Cho:Cr ratio < 1.8 suggested SRS-induced radiation changes. Multiparametric MRI diagnosis was confirmed by histology or radiological and clinical follow-up. CONCLUSION: Multiparametric MRI was helpful in the early identification of radiation-related changes and tumour recurrence and may be useful for monitoring treatment changes in intracranial neoplasms after SRS treatment.

Chemoradiotherapy for rectal cancer: an updated analysis of factors affecting pathological response.

AIMS: With the aim of improving locoregional control, the use of preoperative chemoradiotherapy (CRT) for rectal cancer has increased. A pathological complete response (pCR) is often used as a surrogate marker for the efficacy of different CRT schedules. By analysing factors affecting pCR, this analysis aims to guide the development of future trials. MATERIALS AND METHODS: Searches of Medline, EMBASE and the electronic American Society of Clinical Oncology abstract databases were carried out to identify prospective phase II and phase III trials using preoperative CRT to treat rectal cancer. Trials were eligible for inclusion if they defined: the CRT drugs, the radiation dose and the pCR rate. Phase I patients were excluded from the analysis. A multivariate analysis examined the effect of the above variables on the pCR rate and in addition the use of neoadjuvant chemotherapy, the type of publication (peer reviewed vs abstract), the year of publication and whether the cancers were stated to be inoperable, fixed or threatening the circumferential resection margin were included. The method of analysis used was weighted linear modelling of the pCR rate. RESULTS: Sixty-four phase II and seven phase III trials were identified including a total of 4732 patients. Statistically significant factors associated with pCR were the use of two drugs, the method of fluoropyrimidine administration (with continuous intravenous 5-fluorouracil being the most effective) and a higher radiotherapy dose. Although the use of two drugs was associated with a higher rate of pCR, no single schedule seemed to be more effective. None of the other factors analysed significantly influenced pCR. CONCLUSIONS: A higher rate of pCR is seen in studies using two drugs, infusional 5-fluorouracil and a radiotherapy dose of 45 Gy and above.

Feasibility of Dose-escalated Hypofractionated Chemoradiation in Human Papilloma Virus-negative or Smoking-associated Oropharyngeal Cancer.

AIMS: Oropharyngeal squamous cell carcinoma (OPSCC) can be divided into favourable and poor prognostic groups by association with human papilloma virus (HPV) and smoking. This study prospectively investigated a dose-intensified schedule in poor/intermediate prognosis OPSCC. MATERIALS AND METHODS: Patients with p16/HPV-negative or p16-positive N2b OPSCC with a greater than 10 pack-year smoking history were eligible. Patients were planned to receive 64 Gy in 25 fractions with cisplatin. The primary end point was absence of grade 3 mucositis at 3 months. RESULTS: Fifteen patients were recruited over 14 months. All patients completed a minimum of 2 years of follow-up. All patients completed full-dose radiotherapy within a median treatment time of 32 days (31-35). Grade 3 mucositis was absent in all patients at 3 months. There was one grade 4 toxicity event due to cisplatin (hypokalaemia). Complete response rates at 3 months were 100% and 93% for local disease and lymph nodes, respectively. One patient developed metastatic disease and subsequently died. Overall survival at 2 years was 93% (95% confidence interval 61-99%). CONCLUSIONS: The schedule of 64 Gy in 25 fractions with concomitant chemotherapy is tolerable in patients with poor and intermediate prognosis OPSCC.

Randomised feasibility study to compare the use of Therabite® with wooden spatulas to relieve and prevent trismus in patients with cancer of the head and neck.

Our aim was to compare the efficacy of the Therabite® jaw motion rehabilitation system (Atos Medical) with that of wooden spatulas to relieve and prevent trismus in patients who have had radiotherapy for stage three and four oral and oropharyngeal cancer. Secondary aims were to assess the feasibility and the impact of exercise on health-related quality of life (QoL), and the use of health services after treatment. We designed a randomised, open-label, controlled, three-centre feasibility study to compare the effectiveness and cost of the Therabite® and wooden spatulas. We studied compliance with exercises and health-related QoL, assessed cost using three health economics measures, and conducted semistructured interviews with patients. Patients were randomised into two groups: the Therabite® group (n=37) and the wooden spatula group (n=34). All patients had some sense of jaw tightening before the study started. Mean mouth opening after six months increased in both groups, but the difference between the groups was not significant (p=0.39). Completion rates for the three economic measures were good. There was no significant difference between the two groups in frequency of contact with care services or in QoL. Exercises during and after radiotherapy can ameliorate trismus in patients with stage three and four oral and oropharygeal cancers, but differences between groups in efficacy, compliance, QoL, or use of hospital or community health services, were not significant.

Implementing Head and Neck Contouring Peer Review without Pathway Delay: The On-demand Approach.

AIMS: Peer review of contour volume is a priority in the radiotherapy treatment quality assurance process for head and neck cancer. It is essential that incorporation of peer review activity does not introduce additional delays. An on-demand peer review process was piloted to assess the feasibility and efficiency of this approach, as compared with a historic scheduled weekly approach. MATERIALS AND METHODS: Between November 2016 and April 2017 four head and neck clinicians in one centre took part in an on-demand peer review process. Cases were of radical or adjuvant intent of any histology and submitted on a voluntary basis. The outcome of contour peer review would be one of unchanged (UC), unchanged with variation or discretion noted (UV), minor change (M) or significant change (S). The time difference between the completion of the on-demand peer review was compared with the time difference to a hypothetical next Monday or Tuesday weekly peer review meeting. The time taken to review each case was also documented in the latter period of the pilot project. RESULTS: In total, 62 cases underwent peer review. Peer review on-demand provided dosimetrists with an average of an extra two working days available per case to meet treatment start dates. The proportion of cases with outcomes UC, UV, M and S were 45%, 16%, 26% and 13%, respectively. The mean peer review time spent per case was 17 min (12 cases). The main reason for S was discrepancy in imaging interpretation (4/8 cases). A lower proportion of oropharyngeal cases were submitted and had S outcomes. A higher proportion of complex cases, e.g. sinonasal/nasopharynx location or previous downstaging chemotherapy had S outcomes. The distribution of S outcomes appears to be similar regardless of clinician experience. The level of peer review activity among individuals differed by workload and job timetable. CONCLUSION: On-demand peer review of the head and neck contour volume is feasible, reduces delay to the start of dosimetry planning and bypasses the logistical barriers of weekly meetings. An audit of participation will be required to ensure successful implementation.

The Role of Positron Emission Tomography/Computed Tomography Imaging in Head and Neck Cancer after Radical Chemoradiotherapy: a Single Institution Experience.

AIMS: Positron emission tomography/computed tomography (PET/CT) is used to restage head and neck cancer 3 months after chemoradiotherapy. The purpose of this study was to determine the negative predictive value (NPV) of a scan reported as having no abnormal uptake and the positive predictive values (PPV) for different maximum standardised uptake value (SUVmax) thresholds. MATERIALS AND METHODS: Patients with squamous cell carcinoma of the oro-/hypopharynx/larynx (n = 206) were included. SUVmax and subsequent locoregional recurrence were documented. RESULTS: The median SUVmax was 11.2 (range 4-33)/4.6 (range 2-30), respectively, in patients with/without definite primary site recurrence (P = 0.004). The median SUVmax was 4.4 (range 2.6-15.6)/3.1 (range 2.1-4.6), respectively, in patients with/without definite nodal recurrence (P = 0.003). The NPV for a scan reported as having no abnormal uptake was 92%. The PPV for the SUVmax thresholds 4, 6 and 8, respectively, were 53, 65 and 92% (primary site) and 93, 100 and 100% (nodes). CONCLUSIONS: The NPV of PET/CT after chemoradiation is consistent with the literature and underlines the importance of PET/CT in restaging the primary site if salvage neck dissection is considered. The overall PPV of PET/CT remains low but is high for nodal SUVmax > 4. These data could be used to design risk-stratified follow-up schedules.

A prospective comparison of common toxicity criteria adverse events Version 3 and 4 in assessing oral mucositis for oral and oropharyngeal carcinoma.

BACKGROUND AND PURPOSE: Oral mucositis is an expected complication of radiotherapy in the management of carcinoma of the head and neck. The Common Terminology Criteria for Adverse Events (CTCAE) Version 3 (V3) and related systems based on mucosal appearance have been used in clinical trials historically. More recently, Version 4 (V4) which is based on patient symptoms has been employed. This study compares the use of V3 and V4 in the grading of mucositis in patients undergoing radiotherapy with or without concurrent systemic therapy for carcinoma of the oral cavity and oropharynx. METHODS: Oral mucositis was graded prospectively in patients receiving radiotherapy with or without concurrent systemic therapy using both V3 and V4. Grading was recorded during and after completion of therapy. RESULTS: Between November 2014 and November 2015, 555 measurements were taken from 73 patients. Mucositis scores were equal in both versions in 327 (59%) measurements. Significant differences between V3 and V4 were seen in patients receiving cetuximab-based concurrent therapy (p < 0.001) and beyond 8 weeks from the start of radiotherapy (p = 0.004). CONCLUSION: Differences in grading of mucositis scored by V3 and V4 are frequent. Relationships between biologically effective dose and rates of grade 3 mucositis have historically been based on mucosal appearances. It is not known whether the same relationships apply when mucositis is graded based on symptomatic grading systems. Both V3 and V4 should be used in clinical trials to improve understanding of mucositis and its relationship to quality of life and late mucosal toxicity.

Linear accelerator radiosurgery for arteriovenous malformations: Updated literature review.

Arteriovenous malformations (AVMs) are the leading causing of intra-cerebral haemorrhage. Stereotactic radiosurgery (SRS) is an established treatment for arteriovenous malformations (AVM) and commonly delivered using Gamma Knife within dedicated radiosurgery units. Linear accelerator (LINAC) SRS is increasingly available however debate remains over whether it offers an equivalent outcome. The aim of this project is to evaluate the outcomes using LINAC SRS for AVMs used within a UK neurosciences unit and review the literature to aid decision making across various SRS platforms. Results have shown comparability across platforms and strongly supports that an adapted LINAC based SRS facility within a dynamic regional neuro-oncology department delivers similar outcomes (in terms of obliteration and toxicity) to any other dedicated radio-surgical platform. Locally available facilities can facilitate discussion between options however throughput will inevitably be lower than centrally based dedicated national radiosurgery units.

Does Dose to an Oral Mucosa Organ at Risk Predict the Duration of Grade 3 Mucositis after Intensity-modulated Radiotherapy for Oropharyngeal Cancer?

Various methods have been described to delineate the oral mucosa organ at risk (OAR). This study examined whether dosimetric parameters derived from four different OARs correlated to the duration of acute grade 3 mucositis (G3M) in patients with oropharyngeal carcinoma. In total, 66 patients were included in this study. The duration of CTCAE version 3 G3M and the duration of strong opiate use were prospectively recorded, together with six patient factors. In addition, for each OAR the following dose parameters were derived: mean dose, V55, V50, V45, V40 and V30. No statistically significant correlation of version 3 G3M or duration of strong opiate use was noted with the tested parameters. However, a trend towards significance between duration of strong opiate use and pre-treatment weight was observed (P=0.053), reaching statistical significance on subsequent linear regression. This study failed to show a relationship between dosimetric parameters derived from four oral mucosa OARs and the duration of CTCAE version 3 G3M or duration of opiate use, potentially suggesting serial rather than parallel radiobiological phenomena. The utility of CTCAE version 4 G3M as an end point requires further investigation given its potential relationship to pre-treatment weight.

WITHDRAWN: Does Dose to an Oral Mucosa Organ at Risk Predict the Duration of Grade 3 Mucositis after Intensity-modulated Radiotherapy for Oropharyngeal Cancer?

This article has been withdrawn at the request of the author(s) and/or editor. The Publisher apologizes for any inconvenience this may cause. The full Elsevier Policy on Article Withdrawal can be found at http://www.elsevier.com/locate/withdrawalpolicy.

Neurocognitive Effects Following Cranial Irradiation for Brain Metastases.

About 90% of patients with brain metastases have impaired neurocognitive function at diagnosis and up to two-thirds will show further declines within 2-6 months of whole brain radiotherapy. Distinguishing treatment effects from progressive disease can be challenging because the prognosis remains poor in many patients. Omitting whole brain radiotherapy after local therapy in good prognosis patients improves verbal memory at 4 months, but the effect of higher intracranial recurrence and salvage therapy rates on neurocognitive function beyond this time point is unknown. Hippocampal-sparing whole brain radiotherapy and postoperative stereotactic radiosurgery are investigational techniques intended to reduce toxicity. Here we describe the changes that can occur and review technological, pharmacological and practical approaches used to mitigate their effect in clinical practice.

Linear accelerator stereotactic radiosurgery for vestibular schwannomas: a UK series.

AIMS: To evaluate non-auditory toxicity and local control after linear accelerator stereotactic radiosurgery (SRS) for the treatment of vestibular schwannomas. MATERIALS AND METHODS: The institutional policy was to use SRS for radiologically progressing vestibular schwannomas. Case notes and plans were retrospectively reviewed for all patients undergoing SRS for vestibular schwannomas between September 2002 and June 2012. All patients were surgically immobilised using a BrainLab stereotactic head frame. The treatment plan was generated using BrainLab software (BrainScan 5.03). The aim was to deliver 12 Gy to the surface of the target with no margin. Patients with a minimum of 12 months of follow-up were included for toxicity and local control assessment. Radiological progression was defined as growth on imaging beyond 2 years of follow-up. Overall local control was defined in line with other series as absence of surgical salvage. RESULTS: Ninety-nine patients were identified. Two patients were lost to follow-up. After a median follow-up interval of 2.4 years, the actuarial radiological progression-free survival at 3 years was 100% and overall local control was also 100%. However, two patients progressed radiologically at 3.3 and 4.5 years, respectively. Twenty-one of 97 (22%) evaluable patients suffered trigeminal toxicity and this was persistent in 8/97 (8%). Two of 97 (2%) suffered long-term facial nerve toxicity (one with associated radiological progression causing hemi-facial spasm alone). One of 97 (1%) required intervention for obstructive hydrocephalus. No statistically significant dosimetric relationship could be shown to cause trigeminal or facial nerve toxicity. However, 7/8 patients with persistent trigeminal nerve toxicity had tumours in contact with the trigeminal nerve. CONCLUSIONS: SRS delivering 12 Gy using a linear accelerator leads to high local control rates, but only prospective evaluation will fully establish short-term toxicity. In this study, persistent trigeminal toxicity occurred almost exclusively in patients whose tumour was in contact with the trigeminal nerve.