The ability to process, prepare, and produce music in its instrumental and vocal forms is a complex function that demands the coordination of multiple domains from both cerebral hemispheres [1]. As described by Dalla Bella and Berkowska [2]; these neural network entails severe cortical networks including the auditory cortex, motor cortex, and basal ganglia. The output utilizes faculties critical to speech and language, including pitch, prosody, melody, rhythm, and syntax [3]. In addition, functional neuroimaging demonstrates the substantial overlap between music and language [4].

Awake craniotomy (AC) has proven to be very helpful in preserving eloquent speech and motor cortical regions while performing cortical and subcortical dissections. While motor and speech stimulation testing is standard and widely studied in the medical literature, music mapping is still not widely performed and remains a controversial topic, as it is not known if this later approach will add meaningful information beyond standard mapping during surgery. Clinical experience teaches that musical function and language/speech function are domains that, despite the overlap, can also be selectively impaired [56]. There have been instances of amusia without aphasia, as in the case of the resection of right Heschl gyrus gliomas [6]. Conversely, there have been instances of aphasia without amusia, such as a case of a talented French musician that has been reported [5]. Furthermore, refinement in functional neuroimaging demonstrates that even within regions of broad overlap between music and speech/language, distinct activation patterns may still be extrapolated [7].

It may be reasonable to include musical mapping during AC for brain tumor resection in patients who require their musical skills for personal or professional reasons. Recently, Kappen et al. [8] performed a systematic review of patients performing music during AC which showed that mapping music is feasible. In this case report, we explore our experience with AC for brain tumor resection in a patient that played an instrument (guitar) and sang simultaneously during surgery. In these cases, music processing mapping is not fully standardized or understood yet, as we will show from the data within this review. We did not perform music processing mapping and we intend to give our insights about the use of standard mapping in these cases in comparison.

A 34-year-old male professional guitarist and singer with a past medical history of depression and 1 year of impaired word-finding and mild cognitive changes presented to the emergency department after a generalized seizure. MRI revealed a non-enhancing mass predominantly located within the left temporal lobe with extension into the insula (Fig. 1). Given the patient’s young age as well as the size, eloquent location and concern for an underlying infiltrative glial tumor, the patient opted for surgical resection via an awake left pterional craniotomy with brain mapping. Before undergoing the procedure, surgical planning included functional magnetic resonance imaging (fMRI) (Fig. 2), using the blood-oxygen-level dependent (BOLD) technique, and diffusion tensor imaging (DTI) (Fig. 3). During fMRI, the patient performed our standard language and tongue motor protocol, demonstrating activation of the left hemisphere and adjacent regions to the tumor in the posterior superior temporal gyrus (pSTG) and inferior motor homunculus in the precentral gyrus.

Resection was planned using the asleep-awake-asleep method with intraoperative speech mapping and electrocorticography. One gram of levetiracetam was administered intraoperatively. Following the craniotomy, the patient was awoken, and the cortical grid was placed to localize the central sulcus. Phase reversal was negative, indicating the motor cortex was posterior to the resection area. With bipolar stimulation, a negative map for speech arrest was obtained and corticectomy was performed. The speech was preserved with continuous testing during the surgery, including singing, as was motor function with guitar playing (Fig. 4, Supplementary Video 1 in the online-only Data Supplement). A near-total resection was obtained and the final pathology was reported as WHO Grade III anaplastic astrocytoma (IDH1 mutant).

Postoperatively, the patient surprisingly displayed neologisms, phonemic and semantic paraphasia. The Western Aphasia Battery showed transcortical sensory aphasia. This aphasia was attributed to postoperative changes and swelling from the resection (Fig. 3). His aphasia began to improve by the second postoperative day and returned to his baseline, and at the 2-month follow-up, he had further improvement from baseline with reported improvement in music and singing performance. He completed 30 sessions of 59.4 Gy of fractionated external beam radiation with concomitant temozolomide. The patient has since returned to professional musical activity and there have been no signs of tumor progression now 18 months postoperation.

Awake mapping studies have been built out of a diversity of cortical areas where stimulation specifically impacts musical function while sparing non-music motor/language tasks. This is important for cases where it is important to preserve musicality. Music itself is a complex task that involves several components (e.g., pitch, melody, score reading) located in different regions of the brain [9]. The complexity of music processing and variability among patients underscores the potential benefit of awake music mapping for those patients for whom its preservation is critical.

To date, 20 patients have been reported in the literature of AC for brain tumor resection to map different forms of musical function, not including ours. Furthermore, there is only 1 review of the literature, which encased craniotomies for all causes [8]. In this review, we specifically focus on brain tumor resections to analyze the benefits of this approach in this patient population. In Table 1, we summarize the demographic and clinical characteristics of patients who underwent AC with standard musical tasks for intraoperative assessment. Only 4 patients were assessed intraoperatively while they played a musical instrument (Table 2).

Only 3 patients underwent standardized music testing (15%), while 8 patients were evaluated by a trained third party (40%), and 9 were self-evaluated (45%); 20 cases reported preoperative assessment. Intraoperative testing of musical abilities included singing on 11 patients (52.4%), humming on 3 patients (14.3%), melody recognition tasks on 2 patients (9.5%), score reading on 7 patients (33.33%), and instrument playing on 4 patients (19.04%). Ten studies reported the size of tumor resection. Of these, a gross total resection was achieved in 6 cases (60%), near-total resection in 3 patients (30%), and subtotal resection in only 1 (10%). Outcomes of musical ability postoperatively were reported in 15 cases; 12 of them (80%) had a musical ability that was comparable to preoperative assessment, and in 3 cases (20%), these skills were improved including our case. Only our patient (6.7%) had transcortical sensory aphasia after surgery with complete recovery during follow-up, among all cases with reported outcomes (15 cases). Only 1 patient was reported to have passed away at the end of follow-up [10].

Overall, outcomes following awake mapping are favorable. In all cases, postoperative assessments demonstrated a postoperative interval where the musical function was comparable to the preoperative baseline or superior. We found in our review that music mapping in some of the patients were located in areas close to or within regions that are critical for language and motor tasks. For instance, our review shows that in 4 patients (19%) the pSTG was activated during the musical task; in 6 patients (28.6%) the precentral gyrus was activated, while in 1 patient (4.8%) the supplementary motor area (SMA), and 4 patients (19%) the inferior frontal gyrus (IFG) both were activated [41112131415161718]. pSTG and IFG gyrus are the locations of Wernicke’s and Broca’s area, respectively; while precentral gyrus and SMA are motor areas. These results suggest either a close proximity or overlap of both music with motor and language functions. In 5 patients (23.8%), other areas non-related to motor and speech were related to music production (middle temporal gyrus and postcentral gyrus) [118]. These results point out that music involves a heterogeneous number of brain areas, which can overlap with others related to motor and speech.

In our review, 3 cases (14.3%) including ours, were not specifically mapped for music processing, instead, they focused on the assessment of speech and motor tasks [1315]. Good outcomes in 3 patients could be explained by the overlap mentioned above, hence evaluation of speech could protect the music faculty. An alternative for future cases like these is performing intraoperative functional testing with consistent rotation between language and music assessment. Furthermore, only 4 patients (19%) in the literature have undergone AC while playing an instrument intraoperatively for brain tumor resection including ours (Table 2) [101519]. In addition, we combined singing and guitar strumming, a method of assessment, which has not been done before. Playing an instrument involves multiple skills including coordination, spatial intelligence, musical notation, and rhythm. Hence, we would like to propose performing this type of assessment in future cases when feasible and indicated. With more cases added to the literature, it will be possible to determine if this assessment is the best approach for these types of cases within this patient population.

Despite not undergoing music processing mapping, our case had a postoperative self-assessment with an improvement in musical skills with the guitar and singing. This phenomenon has been attributed in previous reports to the alleviation of brain edema and compression after tumor resection [4]. In addition, neural plasticity and recruitment of adjacent zones play an important role according to image studies [420]. This leads to the question if specific music mapping is necessary for having better outcomes on AC for brain tumor resection. According to the limited experience provided by the literature, we can assume that standard AC for motor and speech functions can be safely done in these cases where it is fundamental to preserve musical abilities. Moreover, in these cases playing an instrument intraoperatively could be enough to preserve the functionality of the patient without requiring complex music mapping. Importantly, we provide evidence that “negative mapping” may be sufficient for music preservation when operating in cerebrum dense with musical function as an excellent functional outcome can be obtained without positively stimulating eloquent musical sites. This strategy of negative mapping for music aligns with the current evolution in language mapping that de-emphasizes positive site control and instead focuses on identifying negative areas that allow a more precisely tailored surgery [21]. These approaches should be addressed when there are more cases in the literature, enough to provide statistical evidence about what is the best procedure in these cases. Furthermore, the role of preoperative imaging (fMRI and DTI) in these cases needs to be elucidated. Similar to language testing, it remains an outstanding question whether high-quality noninvasive functional imaging could obviate the need for intraoperative testing.

Another caveat of AC for brain tumor resection in musicians that need to be addressed is the lack of a standardized assessment of musical capabilities, unlike speech. Only 3 cases in our review were evaluated with a musical scale the Montreal Battery of Evaluation of Amusia (MBEA), whereas the majority were subjectively assessed by themselves or a third party [22]. Objective assessments eliminate the risk of bias and could help reduce morbidity after AC, as suggested by Papatzales et al. [22], during AC for language mapping. A potential limitation in our case was our reliance on the patient’s subjective assessment of their musical function pre- and postoperatively. On the one hand, we may have missed subtle benefit or injury from surgery if measured by objective markers; on the other hand, if the goal of surgery is preserving a function subjectively important to the patient, it is not clear that objective musical function test are needed pre- and post-surgery so long as the patient retains the ability to appraise their own function and is pleased with the outcome. We recommend that the degree of pre- and postoperative testing to assess impact on musical function should be decided on a case-by-case basis but advocate consideration should be given intraoperatively as to the role of a trained musicologist intraoperatively similar to trained administrators of speech testing using a standardized test such as the MBEA.

In this study, we present a case of awake cortical mapping in which we performed standard mapping for brain tumor resection in a professional musician. Our approach did not provide worse outcomes than musical mapping and helped to preserve music function in these patients with outstanding results. The literature review showed that there are few cases of intraoperatively instrument playing during AC for brain tumor resection, to date. In addition, other aspects of the technique need to be addressed. More experience is needed to understand and standardize this procedure. More cases should be added to the literature to carry on prospective studies with higher evidence in the future. This will allow the neurosurgeons to better understand if music mapping is necessary and provides better outcomes for patients in these selected cases.