Insights on cognitive reorganization after hemispherectomy in Rasmussen’s encephalitis. A narrative review

1 Introduction

Rasmussen’s encephalitis (RE) is a neurological disorder that affects one hemisphere, causing progressive unilateral atrophy, pharmacoresistant epilepsy, contralateral hemiparesis, and cognitive impairment (Bien 2005; Cay-Martinez et al. 2020; Varadkar et al. 2014). This uncommon condition predominantly impacts children, with approximately 80 % of cases occurring before the age of 10, and presents an incidence rate of 2.4 cases per 10⁷ individuals under 18 (Bien et al. 2013). This pathology advances through three successive stages (prodromal, acute, residual), during which the frequency of epileptic seizures increases, and neurological and cognitive disturbances intensify and become chronic (Bien 2005). Indeed, if these patients experience normal cognitive development until the onset of seizures, they subsequently show decline in language, memory, attention, and reasoning due to both brain atrophy and epileptic seizures (Caplan et al. 1996; Granata and Andermann 2013; Hertz-Pannier et al. 2002; Mitra et al. 2016; Rudebeck et al. 2018; Silva et al. 2020; Vargha-Khadem et al. 1991). As epileptic seizures are not controlled by antiepileptic medication in about 90 % of patients, surgery remains the recommended treatment option (Bien and Schramm 2009; Granata et al. 2014; Guan et al. 2014; Kim et al. 2018). The typical surgery consists of hemispherectomy, the anatomical removal of the affected hemisphere. However, surgical procedures progressed, and what is preferred currently is the hemispherotomy, a functional hemispheric disconnection. This later procedure is safer, and approximately 80 % of patients becomes free of seizures (Alotaibi et al. 2021; Delalande et al. 2007; Lettori et al. 2008; Marras et al. 2010). Given that the topic of this review will not distinguish between hemispherectomy and hemispherotomy, to keep the clarity of this review, we will use a unique term, hemispherectomy. This term, abbreviated HEM hereafter, encompasses either anatomical hemispherectomy or hemispherotomy.

When examining cognitive outcomes post-HEM, it is essential to consider the distinctive features of RE. Unlike other causes, RE manifests as a progressive condition that typically emerges during childhood and can have a late-onset presentation. It is characterized by normal cerebral and cognitive development until the onset of the pathology. The etiology and characteristics of each condition play a crucial role in post-HEM plasticity (Nahum and Liégeois 2020). Therefore, while studies encompassing patients with diverse etiologies are valuable and informative, it is equally important to specifically focus on investigating Rasmussen’s patients. Doing so allows for a deeper understanding of their unique developmental trajectories and plasticity. It facilitates a more comprehensive comprehension of this rare pathology.

Following HEM, RE patients show various cognitive outcomes; some studies report continued impairment, while others indicate more favorable cognitive outcomes. In general, intellectual functioning may develop after HEM but is frequently compromised, generating cognitive slowdown (Battaglia et al. 2006; Lee et al. 2014; McGovern et al. 2019; Schmeiser et al. 2017; van Schooneveld et al. 2011). Moreover, cognitive functions are unequally affected by the HEM, and significant interindividual variability has been reported (Guan et al. 2017; Hertz-Pannier et al. 2002; Liu et al. 2022; Rudebeck et al. 2018; Terra-Bustamante et al. 2009; Vining et al. 1997). Although language impairment and the decline of cognitive functioning were the most frequently reported (see Table A.1), the impairment of other functions may occur, such as visuospatial (Grosmaitre et al. 2015; Pinabiaux et al. 2022), memory (Tavares et al. 2020; Wilson et al. 2017), executive functions (Borne et al. 2022; Save-Pédebos et al. 2016; van Schooneveld et al. 2016) as well as social and theory of mind processing (Borne et al. 2022; Kliemann et al. 2021).

The cognitive profiles associated with these functions are challenging to attribute solely to a specific impaired hemisphere. The complexity inherent in our understanding of HEM neuroplasticity stems from the intricate interplay of mechanisms associated with both the pathology and the surgical intervention. Indeed, the decision-making process for HEM is highly complex. This necessitates thoroughly analyzing the benefit–risk ratio between allowing the disease to spread or making the surgical intervention The primary objective of HEM is to cease epileptic seizures and expect that cognitive deficit will be less important than the patient shall experience with the natural evolution of RE. The potential exacerbation of cognitive impairments postsurgery (e.g., language loss or aphasia, while language functions were preserved preoperatively) constitutes, however, a significant risk that must be evaluated. When the anticipated risk of cognitive decline post-HEM is important, the decision to undertake the surgery may be critically re-evaluated and potentially deferred. Consequently, it is crucial to assess and preempt this risk accurately to inform surgical decision-making processes (Hartman and Cross 2014; Silva et al. 2020). Furthermore, these mechanisms of neuroplasticity interact with factors related to intercognitive development and individual variability. While a substantial body of literature has concentrated on language reorganization following HEM, there is a compelling need to extend the inquiry to encompass other cognitive domains. It is essential to consider systematically the role of modulatory factors that contribute to this intricate process. Overall, it has been shown that factors such as the age at seizure onset, age at surgery, side of surgery, duration of epilepsy, and postoperative follow-up (Basheer et al. 2007; Christodoulou et al. 2021; Cloppenborg et al. 2022; Curtiss et al. 2001; Kadish et al. 2019; Qu et al. 2020) affect how the brain reorganizes and, therefore, the cognitive development and recovery. A schematic illustration of these factors is presented in Figure 1.

Figure 1:

Modulatory factors in posthemispherectomy cognitive outcome in Rasmussen’s encephalitis. Note. The complexity of cognitive outcomes following hemispherectomy (HEM) for Rasmussen’s encephalitis is shaped by a myriad of clinical and demographic factors, resulting in a spectrum of recovery trajectories. Timely surgical intervention and incorporating prehabilitation and rehabilitation programs can positively impact cognitive function and improve outcomes. By recognizing and addressing the multidimensional nature of cognitive recovery, clinicians can optimize patient outcomes and pave the way for a more comprehensive understanding of cognitive reorganization mechanisms. ASO, age of seizure onset; ASM, antiseizure medication.

This narrative review aims to comprehensively synthesize and enhance our understanding of brain reorganization after HEM in patients with RE. It aims to extend beyond the language domain and address other less-explored cognitive functions. The objective is to shed light on the mechanisms and factors that explain the cognitive development variability between patients and the postsurgical functional reorganization. Addressing these issues is crucial from a clinical perspective to enhance patient care and support and implement effective prehabilitation and rehabilitation programs. These programs may be personalized based on an individual’s cognitive profile and patterns of brain functional reorganization.

To provide this review, a literature search was conducted using electronic databases (PubMed, PsycINFO, and Google Scholar) with the keywords “hemispherectomy or hemispherotomy,” “Rasmussen’s encephalitis,” “cognitive outcome,” “cognitive skills,” “language,” “memory,” “executive functions,” “neuroplasticity,” and “reorganization/recovery.” Additional relevant articles were identified by reviewing the references of the selected studies. The inclusion criteria included articles reporting postoperatively cognitive or behavioral outcomes and including RE patients in their cohort. Articles written in English and published between 1990 and 2022 in peer-reviewed journals were included. Articles were excluded if RE patients were not included in their cohort, did not involve hemispherectomy/hemispherotomy, or only had a neuroimaging approach without cognitive scores (i.e., neuropsychological assessment or behavioral testing). Moreover, we did not include articles in which global cognitive outcome was described for a large cohort of HEM patients, encompassing multiple etiologies, without specifying outcomes for RE patients specifically. Appendix Table A.1 summarizes the publications selected, which serve as the basis of this review. We present according to the year of publication, the article reference, assessment method, number of patients (with mean age at assessment in year), number of RE patients included (with mean age at HEM in year), age of seizure onset for RE patients (year), follow-up duration for all population (year), cognitive functions assessed, and the main results on cognitive outcome.

A total of 43 studies were selected, including one literature review. Among the 42 selected experimental studies, the participants varied from large cohorts of more than 100 patients (including patients with RE and other etiologies) to in-depth case studies of RE patients. In all, about 450 patients with RE were included in these studies. Patients were evaluated in late childhood (mean 13.9 years), several years after HEM (mean 5.0 years postoperatively). The results were discussed considering language outcome, recovery beyond language, and factors which modulate cognitive outcome. By the end of each section, we will summarize the main findings.

2 Hemispheric specialization of language after left or right HEM in RE

After HEM, cognitive functions are supported by the only unimpaired contralateral hemisphere. This implies a drastic reorganization of cognitive functions, which are, in neurotypical situations, sustained preferentially by one hemisphere. In the majority of individuals, language is typically lateralized in the left hemisphere (Baciu and Perrone-Bertolotti 2015; Tzourio-Mazoyer and Courtin 2013). However, the hemispheric lateralization of language functions varies depending on the specific linguistic domains under consideration. Typically, syntax (rules governing word arrangement in sentences) and phonology (awareness and manipulation of phonemes and sounds within a language) exhibit a pronounced left-hemispheric specialization, engaging specific regions of the left hemisphere. In contrast, semantics (word meanings) and prosody (rhythmic and intonational aspects of speech) demonstrate a more bilateral representation, engaging networks distributed across both hemispheres (Perrone-Bertolotti et al. 2016; Tzourio-Mazoyer et al. 2017; Vigneau et al. 2011). Meanwhile, the right hemisphere is more involved in visuospatial and emotional processing (Everts et al. 2009; Gainotti 2019; Hervé et al. 2013; Lochy et al. 2019). Certain cognitive functions, including memory and executive functions, exhibit less strict lateralization, involving networks that are more distributed across both hemispheres (Gratton et al. 2018; Jeong et al. 2015). Consequently, the variability in cognitive recovery and the reorganization of neurocognitive networks following HEM is contingent upon the specific cognitive domains under scrutiny and the affected side of the pathology. Notably, as a result of the early left hemispheric lateralization for language and the potential risks of aphasia, the majority of studies on patients with RE have concentrated on evaluating language (Borne et al. 2022; Mahmoudzadeh et al. 2013; Ramantani et al. 2023; Weiss-Croft and Baldeweg 2015). For more clarity, we designate the left hemisphere as the default specialized side for language unless there is a clear description of specialization and lateralization in the right hemisphere.

In one of the pioneering studies on language recovery after left HEM for RE, Boatman et al. (1999) conducted longitudinal pre- and postoperative follow-ups in six patients. These individuals initially exhibited aphasia and mutism resulting from the disconnection of the specialized hemisphere for language. However, within a few months postsurgery, the patients regained the ability to discriminate phonemes. A year later, their comprehension of words and sentences recovered, although naming performance remained impaired (Boatman et al. 1999). These findings suggest that the right hemisphere may take on language functions in young RE patients. This aligns with other research (Bulteau et al. 2015) that reported favorable long-term language outcomes in six patients who underwent left HEM despite variations in performance across different language processes. In this study, patients exhibited relatively preserved lexical and semantic abilities, while verbal fluency, phonological, and syntactic abilities were more impaired several years after left HEM. Notably, the right hemisphere weakly underpins complex grammatical and syntactic functions after left HEM (Bulteau et al. 2015; de Bode et al. 2015).

A recent study by Borne et al. (2022) further supports positive verbal outcomes following left language specialized HEM in RE patients. In this study, the progression of language performance before and after HEM was examined across three patients with distinct cognitive recovery profiles, reflecting the variability documented in existing literature. Despite variations in language profiles, particularly concerning phonological and syntactic abilities, all three patients regained skills enabling simple conversation and demonstrated a verbal IQ surpassing preoperative levels. This underscores the absence of permanent aphasia following left HEM. Indeed, the transfer of language functions to the right hemisphere has been demonstrated in late childhood, after the consolidation of hemispheric specialization (Hertz-Pannier et al. 2002; Telfeian et al. 2002; Vargha-Khadem et al. 1991). However, as reported in the different studies described in this review, the right hemisphere may not optimally support all linguistic operations. Patients who undergo HEM of the left hemisphere specialized for language may preserve specific communication abilities, such as the capacity to engage in conversations and comprehend language, akin to patients with specific language disorders. Research indicates that scores on tasks measuring receptive and expressive language skills for individuals with left RE are notably below the norm and lower than those who have undergone right HEM (Nahum and Liégeois 2020). Despite these differences, sentence comprehension appears to be rapidly recovered after HEM. Receptive language functions tend to recover faster than expressive language functions, with patients being more proficient in understanding words and sentences than producing them in the initial months after the surgery (Boatman et al. 1999; Bulteau et al. 2015; Hertz-Pannier et al. 2002). Similarly, semantic functions tend to recover sooner and better than syntactic and phonological ones (Bulteau et al. 2015; de Bode et al. 2015). Interestingly, language development in the isolated right hemisphere after left surgery follows a similar development pattern for the grammatical skills than that observed in healthy individuals (Curtiss and de Bode 2003). Indeed, the authors illustrated that the grammatical skills and language production of patients post-HEM followed developmental stages similar to those of neurologically intact children matched for expressive language development. Notably, the language profile observed post-HEM, marked by stronger receptive than expressive abilities, difficulties with grammatical and phonological aspects, yet the ability to engage in simple conversations, has also been noted in multilingual patients across all languages (Trudeau et al. 2003).

When comparing language abilities following left and right HEM, research indicates nevertheless superior performance in patients who underwent right HEM for vocabulary (Liégeois et al. 2008b; Pulsifer et al. 2004), expressive (Curtiss et al. 2001; Terra-Bustamante et al. 2009) and receptive (Stark and McGregor 1997) language skills, reading (Christodoulou et al. 2021), and general cognitive abilities (Korkman et al. 2005; Liu et al. 2022). It is important to note that, due to the left hemispheric specialization for language, literature has described specific language profiles less extensively after right HEM than left HEM. Nevertheless, some delays, albeit less pronounced, have also been reported, underscoring the impact of RE and HEM on language performances after disconnection of the nondominant hemisphere (Caplan et al. 1996). Generally, patients with right HEM for RE exhibit scores in the low-average range of the norm for both expressive and receptive language abilities (Liégeois, Cross, et al. 2008a,b; Pulsifer et al. 2004). These findings are supported by a recent literature review that statistically highlighted language scores in patients with right HEM for RE, especially vocabulary, significantly lower than those observed in the general population (Nahum and Liégeois 2020). In this review, the authors reported that patients with right HEM may also present scores below the norm for oral comprehension and syntactic judgment tasks. However, these impairments observed after right HEM are less pronounced than after left HEM. In the case of right HEM, it appears that such language impairments are more pronounced in early childhood (Vargha-Khadem et al. 1991) and tend to normalize during follow-up (Guan et al. 2014; Liu et al. 2022; Save-Pédebos et al. 2016; Terra-Bustamante et al. 2009). Altogether, these results demonstrate that after right HEM, patients with RE exhibit language abilities ranging from mild impairments to low-average or normative performances, emphasizing that language development may also be affected after RE and surgery in the nondominant hemisphere for language. On the contrary, right RE HEM patients are more likely to experience significant disorders concerning language functions typically lateralized in the right hemisphere, notably pragmatic language (Save-Pédebos et al. 2016).

At a neuroimaging level, research examining language recovery and functional reorganization in the right hemisphere indicates that language tasks engage perisylvian regions traditionally specialized in language processing in the left hemisphere (Bulteau et al. 2017; Hertz-Pannier et al. 2002; Liégeois et al. 2008a; Voets et al. 2006). Therefore, networks involved in language tasks appear to be similar between hemispheres, with both hemispheres capable of supporting cognitive functions by recruiting specialized networks and regions in a mirrored fashion. Nevertheless, the transition of language functions to the right hemisphere can occur before HEM (Bien 2005). For instance, Loddenkemper et al. (2003) documented cases where patients initially displayed left hemisphere dominance for language but, following RE, exhibited a transfer to the right hemisphere even later in childhood or adolescence, with right hemisphere activations observed during language tasks. This preoperative transfer may facilitate the preservation of language abilities postoperatively. However, this transfer is not consistently observed before surgery. In cases where it has not occurred during RE development, evidence emerges after HEM, with brain activations for language functions observed in the right hemisphere (Bulteau et al. 2017; Hertz-Pannier et al. 2002). Moreover, additional activations have been observed after surgery on the specialized left hemisphere in RE patients, suggesting a more diffuse mobilization at the functional level after HEM and a greater cerebral cost of cognitive recovery (Bulteau et al. 2017; Danelli et al. 2013). In summary, these findings imply that, despite the impact of left HEM on language abilities, the brain can undergo adaptation and compensation for the loss of specialized regions, enabling the retention of certain language abilities.

The differences in language outcome after left and right HEM are consistent with the differentiated roles of the two hemispheres in language functions observed in healthy individuals (refer to, for instance, Tzourio-Mazoyer et al. 2017; Vigneau et al. 2011) and in patients with focal epilepsy (Baciu and Perrone-Bertolotti 2015). However, recent research has additionally suggested that after childhood HEM, the left or right isolated hemisphere can recognize words and faces with a relatively high accuracy of approximately 80 %, despite not being specialized for processing these specific stimuli (Granovetter et al. 2022).

Main points

1. Language can develop in a single hemisphere after HEM, highlighting neurocognitive plasticity, even if deficits typically persist.

2. Following left HEM, language may be sustained by the right hemisphere, even in late childhood with declined brain plasticity and after left lateralization, but generally remains significantly compromised.

3. After right HEM, while language performance is better than after left HEM, delays can still be observed.

3 Recovery beyond language: cognitive scaffolding after HEM

Cognition encompasses a multitude of functions, including language, but also memory, executive functions, and social cognition among others. Notably, these functions are intricately interwoven during development and across the lifespan (Bressler and Menon 2010). According to the mutualism model of intelligence, cognition emerges from dynamic mutual interactions among cognitive functions (Kan et al. 2019; Van Der Maas et al. 2006). In this perspective of cognitive scaffolding, the development of each function is intertwined with the development of others, fostering interactive cognitive development that relies on functions emerging early in the process. For instance, during the development, there is a reciprocal relationship between language abilities and nonverbal reasoning skills, as evidenced in both typical development and among children with language disorders (Griffiths et al. 2022; Kievit et al. 2019). Therefore, studying cognitive development and reorganization necessitates a broader perspective, adopting the standpoint of cognitive scaffolding, and thus considering complex interactions between different domains that allow cognition to emerge and develop. In this section, we will first review principal findings regarding cognitive outcomes other than language after HEM for RE and discuss results that prompt us to consider specific cognitive interactions in post-HEM recovery.

The exploration of cognitive functions beyond language in patients with RE has been limited (see Table A.1). Studies conducted with long follow-ups have reported favorable recovery of memory abilities in RE patients (Thomas et al. 2010). Nevertheless, these findings illustrate that recovery may vary depending on memory functions, with verbal memory better recovered after right than left HEM. At the same time, nonverbal modalities are not affected by the side of surgery (Wilson et al. 2017). A recent study by Tavares et al. (2020) highlighted the scarcity of data on memory abilities after HEM. However, the authors found relatively stable memory performance 1 year after surgery without significant decline.

Interestingly, short-term verbal memory has been demonstrated as the best predictor of language outcome (Liégeois et al. 2008b), highlighting an overlap between language and working memory abilities in patients after HEM. Executive functions, on the other hand, may remain low but stable after a HEM (van Schooneveld et al. 2016). Specifically, more executive deficits have been found in patients who underwent early left HEM, before the age of 18 months and in those who underwent late right HEM after the age of 18 months (Save-Pédebos et al. 2016), indicating the early role of the left hemisphere in language and the preferential involvement of the right hemisphere in executive functioning. Moreover, deficits resulting from HEM significantly affect patients’ daily and social lives, often manifesting as behavioral issues (Moosa et al. 2013). Patients may experience deficits in face processing and social cognition, including emotional recognition and theory of mind, particularly after the right surgery (Pinabiaux et al. 2022). Additionally, recent evidence suggests that some patients may experience a reorganization of the theory of mind networks in the isolated hemisphere, indicating potential recovery of social cognition functions after HEM (Kliemann et al. 2021). However, patients who undergo epilepsy surgery typically experience lasting impairments in social cognition and functioning, with few clear improvements (Braams et al. 2019). In this review, the authors suggest that such deficits may be related to the epileptic condition during critical periods of development and to persistent impairments in other cognitive domains, such as language, memory, and executive functions, linked to social cognition functions like the theory of mind.

This standpoint aligns with the mutualism approach of cognition and cognitive scaffolding. For instance, executive functions mature slowly in life (Diamond 2013), and bidirectional relationships between language and executive skills during early development have been demonstrated, both in healthy and clinical populations (Akbar et al. 2013; Gooch et al. 2016; Kaushanskaya et al. 2017; Marini et al. 2020; Morgan et al. 2021; Romeo et al. 2022; Weiland et al. 2014). Specifically, verbal abilities constitute a tangible predictive factor in developing early childhood executive functions, with language as a precursor of executive development (Fuhs and Day 2011; Kuhn et al. 2014). In the context of post-HEM development, studies about cognitive recovery outside language are scarce, and those investigating the interaction between different cognitive functions are even rarer. Still, some results highlight a specific cognitive scaffolding after HEM.

After left HEM, language recovery and the mobilization of cerebral networks in the nonspecialized healthy hemisphere are assumed (Hoffman et al. 2016). However, this process may impact the development of nonverbal functions, as suggested by the “crowding hypothesis” according to which recovery of language functions after left cerebral injuries is characterized by the recovery of language functions but significant impairments of visuospatial ones (Powell et al. 2012; Teuber et al. 1974). These profiles have been reported in the literature after HEM. In the study by Bulteau et al. (2015), five out of six patients who underwent left HEM for RE exhibited a discrepancy during postoperative follow-up characterized by progress in verbal functions but a loss of abilities in nonverbal ones. Recently, in a study including 40 patients with right and left HEM, crowding mechanisms have been highlighted by reporting visuospatial impairments with no difference between right and left HEM (Pinabiaux et al. 2022). Again, these results illustrate a cognitive scaffolding during postoperative recovery, with the recovery of certain functions being detrimental to others, especially the recovery of language in the right nondominant hemisphere being at the expense of visuospatial performance initially specialized in this hemisphere (Silva et al. 2020). Extensive case studies have investigated cognitive profiles in RE patients and showed that the preservation and favorable recovery of verbal functions are usually associated with fewer abilities in the nonverbal domain after left HEM (Borne et al. 2022; Grosmaitre et al. 2015). Furthermore, after HEM, the reorganization of language functions in the right hemisphere has been demonstrated as the strongest predictor of low visuospatial performance (Danguecan and Smith 2019; Lidzba et al. 2006), suggesting a functional interaction and hierarchy of cognitive functions. This discrepancy between verbal and nonverbal functions remains evident even in adulthood after a HEM performed years ago during childhood (Cummine et al. 2009; Ogden 1989). Altogether, these findings underline that regardless of which hemisphere undergoes HEM, language seems to be preferentially preserved, even at the expense of other functions, as cognitive networks are restructured to prioritize verbal functions, including language, verbal intellectual functioning, and memory, during postoperative recovery (Bulteau et al. 2015, 2017; Shurtleff et al. 2021). Language recovery in the isolated hemisphere highlights its adaptative importance in development and centrality in the cognitive landscape.

These findings emphasize a specific orchestration of cognitive reorganization after HEM following a period of postnatal normal cognitive development as it is the peculiar situation of the patient who experienced RE and HEM. It suggests that the reorganization impacts different cognitive domains in relation to each other, as well as the critical periods in the neurodevelopment within a broader perspective of cognitive scaffolding. However, these results are still limited, and research focusing on the reorganization of cognition as a whole and the interactions between cognitive domains in development after HEM are too scarce. Studies that no longer focus on the outcome of a function in isolation but rather on the entirety of cognitive functions and their interactions are truly necessary to understand the mechanisms involved in the development and reorganization of the entire cognitive landscape after HEM for RE.

Main points

1. Beyond the preferential recovery of certain functions (i.e., language), the entirety of cognitive functioning is impacted after HEM for RE.

2. While memory remains stable, face processing, social cognition and executive functions have different developmental trajectories in post-HEM outcome.

3. The cognitive outcomes of patients should be studied within a broader perspective of cognitive scaffolding.

4 Modulation of cognitive outcomes after HEM in RE

Together, the results reported in the literature highlight a highly variable cognitive recovery after left or right HEM, depending notably on the considered functions and their hemispheric lateralization. However, even when considering patients who underwent HEM with the same surgical procedure on the same hemisphere, after evaluating the same cognitive functions, markedly different trajectories of evolution and recovery can be identified (Borne et al. 2022). Indeed, as mentioned in the introduction to this review, there is significant interindividual variability within the population of patients undergoing HEM for RE. Each patient has an individual clinical history, and cognitive reorganization is thus partially conditioned by it. Specifically, clinical predictors, including age at seizure onset (ASO), epilepsy duration, and age at the time of hemispherectomy, are generally considered determinants of functional and cognitive outcomes post-HEM (see Figure 1). In this section, we will review the various clinical and cognitive modulating factors that can impact post-HEM cognitive reorganization. However, it is important to note that these factors interact, resulting in complex predictive patterns.

5 Discussion

HEM in patients with RE is an extreme situation of brain plasticity. This surgical intervention can result in the sustained preservation of cognitive functions in the isolated hemisphere, showcasing favorable cognitive recovery characterized by stability and potential improvements in performance despite persistent deficits. Significantly, the nonspecialized right hemisphere can support language functions beyond the typical sensitive period for language acquisition, highlighting the brain’s remarkable plastic potential and offering insights into hemispheric specialization, critical periods, and cognitive reorganization at a fundamental level. Importantly, this plasticity remains existent even in late childhood.

The recovery pattern, with verbal functions taking precedence, underscores the central role of language in human brain function and behavior. Altogether, the findings highlight the intricate interplay between various cognitive domains during development, supported by interactive networks and progressively leading to diversified and integrative cognition.

Recent findings advocate for the adoption of a more integrative approach to study cognitive mechanisms within the context of an isolated hemisphere (Yeo et al. 2011). Considering the involvement of brain networks rather than isolated regions offers a more comprehensive and suitable perspective for understanding neurocognitive functioning. This effectively allows for a better comprehension of the dynamic and modulable nature of networks, how they organize in healthy functioning and reorganize in pathological conditions, underpinning complex behaviors (Cohen 2018; Herbet and Duffau 2020). Indeed, pathological conditions, such as epilepsy, impact network organization on a larger scale, underscoring the importance of an interactive perspective in exploring cognitive recovery (Banjac et al. 2021; Larivière et al. 2020; Roger et al. 2020). Following HEM, several studies have highlighted a general reorganization of networks within the healthy hemisphere. In a recent study utilizing resting-state fMRI, Kliemann et al. identified a typical organization of canonical rest networks in the healthy hemisphere of posthemispherectomy patients, with increased connectivity and integration among these networks in patients compared to control participants. The enhanced connectivity between functional networks in these patients could be associated with better cognitive recovery at the cognitive level (Ibrahim et al. 2015; Ivanova et al. 2017; Kliemann et al. 2019). Post-HEM increases in brain connectivity within the healthy hemisphere have also been demonstrated anatomically at the level of the white matter (Gaubatz et al. 2020). However, these studies remain scarce, and neuroimaging methods should be used in relation with cognitive methods.

This narrative review specifically focuses on cognitive outcomes following HEM for RE. While the primary objective of surgery is seizure control, a spectrum of cognitive outcomes is evident, with individuals undergoing left HEM displaying partial or even favorable recovery of language functions in the right isolated hemisphere (Boatman et al. 1999; Bulteau et al. 2015). Unique recovery trajectories, shaped by the interplay of clinical and demographic factors (Liu et al. 2022), may result in various cognitive phenotypes. Those with shorter epilepsy durations before HEM, deeper cognitive reserves, and sufficient plastic potential tend to exhibit the most favorable outcomes. However, comprehensive cross-cognitive studies evaluating all cognitive domains post-HEM are scarce.

Experimental studies are warranted to confirm and refine the definition of distinct cognitive phenotypes in this population and elucidate their relationships with clinical and cognitive factors. Garcia-Ramos et al. (2021), Kellermann et al. (2015), and Schraegle and Babajani-Feremi (2022) highlight the importance of network reconfiguration for cognitive recovery in epilepsy patients, emphasizing the need for an interactive perspective in investigating cognitive recovery. A comprehensive approach considering all cognitive domains and their interactions during development is essential for fully understanding the brain’s potential for reorganization (Ferguson 2022).

Moreover, incorporating individual clinical variables can lead to more accurate long-term outcome predictions and underscore the importance of promptly performing the surgical intervention following the onset of seizures. This information is critical for clinicians making surgical decisions and for patients and their families to assess the potential for long-term recovery after surgery (Hartman and Cross 2014; Ozanne et al. 2016). Thus, this review emphasizes the importance of thoroughly understanding clinical and demographic factors determining cognitive outcomes post-HEM for RE. This knowledge can guide personalized surgical decisions and informed rehabilitation planning, potentially enhancing cognitive functions through strategies like prehabilitation (Sousa et al. 2021).

6 Conclusions

In summary, the study of patients with Rasmussen’s encephalitis after hemispherectomy is a great opportunity to unravel the neuroplasticity potential for cognitive functions in one isolated hemisphere. This review underscores the significant plasticity of cognitive functions, showcasing their ability to restructure and partially develop in the nonspecialized remaining hemisphere, with language as a noteworthy example. Nevertheless, while cognitive recovery is possible, it is also constrained and contingent on various factors. Both comprehensive multicenter studies and detailed case analyses are imperative for a more detailed understanding of Rasmussen’s encephalitis patients’ cognitive phenotypes and developmental trajectories posthemispherectomy. Approaching this exploration with an integrative and multimodal perspective will unveil intricate connections between cognitive phenotypes, neuroimaging, and clinical characteristics, culminating in an integrative model of unihemispheric reorganization. Such a model promises a thorough comprehension of the mechanisms governing cognitive reorganization in Rasmussen’s encephalitis patients and stands to enhance clinical decision-making for optimal patient outcomes. This field of research has the potential to significantly improve our comprehension of the brain’s plasticity, ultimately leading to improved lives for patients with Rasmussen’s encephalitis. To the best of our knowledge, this review stands as the sole exploration of the cognitive outcomes of individuals with Rasmussen’s encephalitis after hemispherectomy.