Resections in eloquent areas entail the risk of altering the cognitive capacities of patients (Sanai and Berger, 2008). Awake surgery is performed to preserve neuropsychological functions (Ojemann et al., 1989; Ojemann, 1993; Duffau et al., 2003, 2005).

Assessment of cognitive functioning is highly recommended as the neuropsychological profile is an important outcome (Ng et al., 2020). Neuropsychological deficits could arise because of tumor growth or surgery and may affect the quality of life. For instance, Ng et al. (2020) compared pre- and postsurgery cognitive assessments in a consecutive series of 47 patients with (incidental) low-grade glioma (LGG) who underwent awake surgery. Of note, 72.3% of the patients presented stable cognitive functioning, 14.9% of the patients improved cognitive functioning, and 12.8% of the patients showed slight impairments.

Sparing neuropsychological functions during awake surgery is crucial. However, only language or motor functions are routinely mapped. A review on the neuropsychological tests used during awake brain surgery (Ruis, 2018) reported that cognitive functions (i.e., memory, calculation, and emotions) were assessed in a minority of cases.

To test the neuropsychological functions during awake surgery, direct electrical stimulation (DES) mapping is performed (Berger and Ojemann, 1992; Berger, 1994, 1996; Duffau et al., 2000, 2005), while the patient carries out a cognitive task. DES mapping minimizes the risk of permanent postoperative deficits while maximizing the quality of resection, especially in infiltrating tumors such as LGGs where the larger the resection, the longer the survival of a patient (Duffau et al., 2003; Keles et al., 2004; Sanai and Berger, 2008, 2011; Sanai et al., 2010; Ius et al., 2012). The surgeon determines where DES evokes an interference response and imports this spatial information into a 3D MRI image to reproduce a somewhat personalized non-resectability map to guide surgery at cortical and subcortical levels. DES mapping is repeated intra-resections at intervals by pausing excision.

The most frequently used task for surgery in the left Sylvian and peri-Sylvian areas is naming (or counting; Duffau et al., 2000, 2002, 2003; Brennan et al., 2007; Tonn, 2007; Chang et al., 2017). DES mapping at cortical or subcortical portions of the brain induces an interference with task performance. For example, if the patient is carrying out a counting task, DES in Broca’s area may elicit speech arrest. As another example, during a naming task, DES mapping delivered in temporal areas may trigger anomia or phonological and semantic paraphasia. Sometimes, semantic processing can be tested by means of the Pyramids and Palm Trees Test (Talacchi et al., 2013).

Other tests for the left temporal lobe include repetition (Moritz-Gasser and Duffau, 2013; Sierpowska et al., 2017; Leonard et al., 2019). DES to the left peri-Sylvian network causes interference with word repetition with phonological errors, perceptual deficits, and speech arrest (Leonard et al., 2019). Dissociations between naming and pseudoword repetition can be observed: DES to the arcuate fasciculus causes interference with pseudoword repetition (Sierpowska et al., 2017), while DES to the posterior part of the superior and middle temporal gyri induces anomias with preserved word repetition (Moritz-Gasser and Duffau, 2013). These data suggest that it is worth using a multiple testing approach with more than one test.

There are occasional reports of DES during a reading task (Roux et al., 2004, 2012; Gil-Robles et al., 2013; Motomura et al., 2014; Zemmoura et al., 2015; Hirshorn et al., 2016). A study used a somewhat different approach and administered the reading test during resection and not during DES (Tomasino et al., 2020). A decrease in performance (vs. presurgery data) was found in 19/49 patients, and their x, y, and z brain coordinates, where resection causes a decrement, were distributed on the posterior inferior and middle temporal gyrus to the temporoparietal areas and the precentral gyrus (Tomasino et al., 2020).

There are reports on DES during writing (Roux et al., 2003, 2009, 2014; Lubrano et al., 2004; Motomura et al., 2014) and less data on a picture description task to test syntax (Chang et al., 2018).

In addition to DES mapping, surgeons may obtain further information from neuropsychological monitoring providing feedback on the cognitive status of patients during resection. This (Skrap et al., 2016) can be performed using real-time neuropsychological testing (RTNT). The RTNT is listed among the neuropsychological techniques (Tomasino and Rumiati, 2020) available for neuropsychologists as clinical tools. Contrary to DES, the RTNT is performed without pausing excision, as it is administered from the beginning of resection until hemostasis. Upon the request of a surgeon, the RTNT can be temporarily paused to perform DES. An RTNT protocol consists of a battery of neuropsychological tasks tapping the area under resection, e.g., for insular lobe, see Tomasino et al. (2018). It is individually selected for each patient based on the lesion localization, presurgical functional magnetic resonance imaging (fMRI)/diffusion tensor imaging (DTI) results, and presurgical neuropsychological status. Each RTNT task includes 10–15 items (e.g., pictures, sentences, words, and numbers to be repeated) and is repeated several times, by changing items, in order to test several cognitive functions in 1–2 min. Thus, the RTNT can provide continuous feedback on cognitive functions in real time in order to monitor the consequence of each surgical act. A decrease in performance might warrant pausing resection, after searching for confirmation of the functionality of the anatomical site using DES. If RTNT runs are performed correctly, the surgeon resumes resection. Finally, a temporary decrease in performance, which could be related to mechanical manipulation or fatigue of a patient, deserves in-depth investigation. During this time, a “stop-and-go” approach (Sala et al., 2007) is used by the surgeon.

In this case report, we presented our RTNT protocol for resections in the left temporal lobe and describe our multimodal approach combining neuropsychology, fMRI, DTI, and RTNT plus DES. By using the RTNT approach, we aimed to obtain a complete overview of the neuropsychological functions of a patient supported by the craniotomy site and to determine whether—and if so, how—the neuropsychological status evolves during resection. Second, we expected to confirm that this method can be applied in a short resection time.

We presented the RTNT protocol for resections in the left temporal lobe and a case report. We chose to present a single case rather than our experience with the entire series of temporal patients in order to detail the entire RTNT protocol and performance, the procedure, fMRI images, and DTI results.

Our experience suggests that DES remains the gold standard in awake surgery; nonetheless, the surgeon can use the RTNT technique to increase the amount and type of available information about the cognitive status of the patient.

Studies investigating the postoperative cognitive status of patients often report deficits in functions that were not directly assessed during DES, since DES is mainly matched to object naming or counting tasks. In this specific patient, too, DES was performed during object naming and showed a positive speech arrest site. The RTNT added considerable information. This RTNT protocol for temporal resections includes tests that are most commonly used in awake surgery (e.g., see introduction mentioning naming, reading, repetition). In addition, it includes monitoring of short-term memory and working memory, phonological processing, lexical decision, and verbal comprehension (Token test) as well as verb naming.

The same tests were then assessed postsurgery. This is possible since RTNT takes advantage of resection time as the time of test presentation. In this way, surgery time is not extended; rather it is optimized, thanks to the online feedback provided to the surgeon, which enables him to proceed in a more confident manner, when RTNT demonstrates that the patient is responding proficiently to testing.

In the case report illustrated, our patient showed slight decrements in accuracy for pseudoword repetition, short-term memory and working memory, phonological processing, and verbal comprehension (Token test). His postsurgery testing performance was within the normal range. Based on our experience with RTNT, we believe that the lower performance on short-term memory and working memory may be due to fatigue, since immediate postsurgery performance was perfect. The patient was otherwise highly collaborative. In addition, his initial level of vigilance was high while in the final part of the RTNT he was tired and fell asleep. One disadvantage of the RTNT is that it presupposes levels of collaboration and flexibility, which can affect neuropsychological scores such as, in this case example, short-term memory and working memory. It is likely that the slight decrease in the Token test, too, was due to the decrease in short-term memory performance (Pisoni et al., 2019). The warning errors that we considered with much attention during surgery are phonological errors. The patient made some phonologic errors, both in production and in perception. Such errors occurred toward the end of resection.

Our multimodal approach combined RTNT, fMRI, and DTI together with this method. Combining this approach with DES presents some advantages in terms of extent of resection. In a previous study (Ius et al., 2019) using the multimodal approach, we reported a median extent of resection of 98%, compared with a median extent of resection of 83% with DES alone.

Preoperative mapping with fMRI and DTI provides essential, yet insufficient, information. In our opinion, it provides insight during surgery, but only neuropsychological testing can provide accurate levels of information about the cognitive status of the patient. The important aspect is the consistency between instruments and tests. Tongue movement, object naming, verb naming, word reading, and pseudoword reading were used in fMRI. The same tests were used intra-surgery. For example, a mistake made in the presurgery evaluation (e.g., patient saying “to marry” in response to a figure showing a shining ring shine instead of the verb “to shine”) and repeated during the intra-surgery examination as well is considered the baseline of the patient. The response will not be considered an error as it would happen in a formal neuropsychological assessment. In our example, some lexical decision errors were made both presurgery and during RTNT.

The postoperative assessment provides the information such as whether the patient activates or not the same areas as presurgically, whether performance on neuropsychological tests is the same as presurgically, as in our case report. Interestingly, the fMRI assessment also tells us if there is any plasticity or rearrangement pre- or postsurgery in glioma patients (Cargnelutti et al., 2020), showing, for example, that the postsurgery fMRI map for object naming shows an activation in Broca’s area that was not observed presurgery.

Postsurgery DTI assessment and fiber reconstruction indicates whether white matter pathways are preserved, and the results are interpreted in the light of the intra-surgery RTNT performance and according to which white matter pathway was detected during resection. In our case example, two data emerge. First, FA of the left IFOF was significantly lower pre-operatively compared with controls, while FA of the right IFOF was higher than controls both pre- and postoperatively, a sign of plasticity or pre-operative rearrangement. FA of the SLF was lower than controls presurgery and in the normal range postsurgery. Based on pre-operative data, follow-up values should never be viewed as absolute and isolated values but should be compared with pre-operative values. In addition, we are aware that we should be cautious in interpreting the DTI results above, as DTI measurements are sensitive to the presence of “free water” (Weninger et al., 2020). The presence of the tumor itself, or, postsurgery, the presence of the edema could temporarily modify the DTI measurements.

Taken together, the RTNT is highly informative, especially if combined with multimodality data. It does not require further effort. In neurosurgeries where DES is performed, the RTNT does not require additional materials or devices other than the same PC screen used to present DES stimuli. Finally, the RTNT provides essential information that can be used online, during surgery, for clinical aims to provide the surgeon with useful feedback. In addition, RTNT data can be used offline, for research purposes, by listening to patient performance recordings to detect error types and run error analysis or to correlate RTNT data with MRI data according to a multimodal approach.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by CEUR fvg. The patients/participants provided their written informed consent to participate in this study.

MS and BT designed the research. BT, IG, MM, and TI performed the research. BT and MM analyzed the data. BT, MS, and IG wrote the manuscript. All authors edited the manuscript and revised the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.