Radiotherapy plus temozolomide in elderly patients with glioblastoma: a "real-life" report.

BACKGROUND: The optimization of the management for elderly glioblastoma patients is crucial given the demographics of aging in many countries. We report the outcomes for a "real-life" patient cohort (i.e. unselected) comprising consecutive glioblastoma patients aged 70 years or more, treated with different radiotherapy +/- temozolomide regimens. METHODS: From 2003 to 2016, 104 patients ≥ 70 years of age, consecutively treated by radiotherapy for glioblastoma, were included in this study. All patients were diagnosed with IDH-wild type glioblastoma according to pathological criteria. RESULTS: Our patient cohort comprised 51 female patients (49%) and 53 male. The median cohort age was 75 years (70-88), and the median Karnofsky performance status (KPS) was 70 (30-100). Five (5%) patients underwent macroscopic complete resection, 9 (9%) had partial resection, and 90 (86%), a stereotactic biopsy. The MGMT promoter was methylated in 33/73 cases (45%). Fifty-two (50%), 38 (36%), and 14 (14%) patients were categorized with RPA scores of III, IV, and I-II. Thirty-three (32%) patients received normofractionated radiotherapy (60 Gy, 30 sessions) with temozolomide (Stupp), 37 (35%) received hypofractionated radiotherapy (median dose 40 Gy, 15 sessions) with temozolomide (HFRT + TMZ), and 34 (33%) HFRT alone. Patients receiving only HFRT were significantly older, with lower KPSs. The median overall survival (OS; all patients) was 5.2 months. OS rates at 12, 18, and 24 months, were 19%, 12%, and 5%, respectively, with no statistical differences between patients receiving Stupp or HFRT + TMZ (P = 0.22). In contrast, patients receiving HFRT alone manifested a significantly shorter survival time (3.9 months vs. 5.9 months, P = 0.018). In multivariate analyses, the prognostic factors for OS were: i) the type of surgery (HR: 0.47 [0.26-0.86], P = 0.014), ii) RPA class (HR: 2.15 [1.17-3.95], P = 0.014), and iii) temozolomide use irrespective of radiotherapy schedule (HR: 0.54 [0.33-0.88], P < 0.02). MGMT promoter methylation was neither a prognostic nor a predictive factor. CONCLUSIONS: These outcomes agree with the literature in terms of optimal surgery and the use of HFRT as a standard treatment for elderly GBM patients. Our study emphasizes the potential benefit of using temozolomide with radiotherapy in a real-life cohort of elderly GBM patients, irrespective of their MGMT status.

Modelling human memory: connectionism and convolution.

The mathematical operation of convolution is used as an associative mechanism by several recent influential models of human memory. Convolution can be used to associate two vectors (representing items to be remembered) into a memory trace vector in one operation. An approximation to either of the input vectors can then be retrieved, using the other vector as a probe. Recent convolution-based memory models have accounted for a wide range of data. Connectionist models may have greater potential for providing developmental accounts, but the architectures that have been most widely used to account for developmental phenomena cannot perform one-trial learning and this has limited their use as models of human memory. We show that a connectionist-like architecture can learn, using a gradient-descent algorithm, to perform single-trial learning in a similar manner to convolution. The solution that the network finds leads to less variable retrieval than does convolution. Furthermore, the network can learn to carry out the convolution operation itself. This provides a link between connectionist and convolution approaches, and a basis for models with many of the attractions of both connectionist and convolution approaches.

Mathematical and connectionist models of human memory: a comparison.

Recent convolution-based models of human memory (e.g. Lewandowsky & Murdock, 1989), have accounted for a wide range of data. However such models require the relevant mathematical operations to be provided to the network. Connectionist models, in contrast, have generally addressed different data, and not all architectures are appropriate for modelling single-trial learning. Furthermore, they tend to exhibit catastrophic interference in multiple list learning. In this paper we compare the ability of convolution-based models and DARNET (Developmental Associative Recall NETwork), to account for human memory data. DARNET is a connectionist approach to human memory in which the system gradually learns to associate vectors, in one trial, into a memory trace vector. Either of the vectors can than be retrieved. It is shown that the new associative mechanism can be used to account for a wide range of relevant experimental data as successfully as can convolution-based models with the same higher-level architectures. Limitations of the models are also addressed.