Limbic encephalitis: Experience of a moroccan center

Abstract Objectives Histologically defined as an inflammation—degeneration of limbic structures, limbic encephalitis (LE) is a rare disease and often difficult to diagnose particularly in institutions with limited access to laboratory tests such as antineuronal antibodies or HSV‐PCR, and functional imaging. We aimed to describe the demographic, clinical, paraclinical, and etiological features of LE, as well as its medium‐term prognosis in Moroccan patients. Materials and Methods We collected retrospectively all patients diagnosed with LE in the Department of Neurology of the University Hospital Hassan II of Fez (Morocco) between September 2008 and December 2016. We analyzed their demographic features, clinical manifestations, magnetic resonance imaging and laboratory findings, etiologies, and medium‐term prognoses. Results We included 22 men and 9 women aged 14–76 years (mean age: 45.8 years). In 64.5% of cases, the onset of symptoms was acute. The clinical manifestations included generalized status epilepticus (16.1%), confusional syndrome (29%), epileptic seizures (38.7%), psychiatric disorders (48.4%), and memory disorders (45.2%). The nonlimbic symptoms were nuchal stiffness (22.6%), headaches (9.7%), fever (61.3%), vesicular rash (3.2%), and language disorders (6.5%). The different etiologies found were herpes simplex virus (6.5%), syphilis (16.1%), tuberculosis (3.2%), varicella (3.2%), paraneoplastic autoimmune LE (22.6%), anti‐NMDA‐R LE (6.5%), and sarcoidosis (3.2%). We found 12 cases (38.7%) of LE without definite etiology and with an incomplete diagnostic workup. The medium‐term clinical course includes a complete remission in 45.2% of cases and partial remission in 45.1% of cases. The different sequelae were temporal lobe epilepsy (9.7%), anterograde amnesia (16.1%), and severe cognitive impairment (19.4%). The mortality rate was 9.7% (3 patients). Conclusion Our study shows a wide diversity of etiologies of LE in Morocco with essentially an acute mode of onset of symptoms.


| INTRODUC TI ON
Histologically defined as an inflammation-degeneration of limbic structures (Brierley, Corsellis, Hierons, & Nevin, 1960;Corsellis, Goldberg, & Norton, 1968), limbic encephalitis (LE) is a rare disease and often difficult to diagnose particularly in institutions with limited access to functional imaging and laboratory tests such as herpes simplex virus polymerase chain reaction (HSV-PCR) and antineuronal antibodies. The clinical manifestations of LE are diverse, and the most common are short-term memory disorders, psychiatric disorders, confusional state, and temporal lobe epilepsy, which have an acute or subacute onset (Anderson, & Barber, 2008;Corsellis et al., 1968;Geisler et al., 2013;Kerling, Blümcke, & Stefan, 2008). The diagnosis of LE based on both clinical manifestations that suggest the involvement or dysfunction of the limbic system, magnetic resonance imaging (MRI) or functional imaging [SPECT (single-photon emission computed tomography) or FDG-PET (fluorodeoxyglucosepositron emission tomography)] findings and laboratory findings (Asztely, & Kumlien, 2012). Etiologies of LE are multiple, and the most common are infectious causes and autoimmune encephalitis (Asztely, & Kumlien, 2012;Fujimoto et al., 2001).
The aim of our study was to describe the demographic, clinical, paraclinical, and etiological features of LE, as well as its mediumterm prognosis in Moroccan patients.

| ME THODS
We collected retrospectively all patients diagnosed with LE in the

Department of Neurology of the Hassan II University Teaching
Hospital of Fez (Morocco) between September 2008 and December 2016. We collected all the data from the medical records of the patients from the archives of our Department. All our patients had a suggestive clinical picture of acute or subacute encephalitis associating to various degrees the following symptoms: short-term memory disorders, confusional state, epileptic seizures, and behavior disorders. We selected only patients with positive brain MRI (signal abnormalities in the limbic structures on T2-weighted fluid-attenuated inversion recovery [FLAIR] imaging and T2-weighted images) due to limited access in Morocco to functional imaging and laboratory tests such as HSV-PCR and antineuronal antibodies that could help to establish a diagnosis of LE with negative brain MRI. We analyzed demographic features, clinical manifestations, MRI and laboratory findings, etiologies, and medium-term prognosis of all patients in-  antiamphiphysin,  The signal abnormalities revealed by brain MRI (Figures 1, 2, 3,4) and the EEG findings are presented in Table 2.

| RE SULTS
Routine CSF examination was normal in 10 patients (32.3%) and revealed lymphocytic meningitis in 67.7% of cases with an average of white blood cells of 54/µl (range 19-140), an average of proteinorachia of 0.56 g/L (range 0.23-1.37) and an average of glycorrhachia of 0.54 g/L (range 0.34-0.71). Serological tests for syphilis in the CSF had been performed in 30 patients and were positive in five cases.
The patient presenting a vesicular rash that occurred 7 days before its admission to our Department (the sole patient who had not benefited from serological tests for syphilis), diagnosis of chickenpox was suspected. The serological tests revealed high plasma levels of IgM specific for VZV (varicella-zoster virus). The CSF examination in this patient revealed a white blood cell count of 54/µl, with glucose of 0.67 g/L and protein of 0.42 g/L. Although VZV-PCR has not been performed in the CSF, diagnosis of varicella LE was considered. We The measurement of antineuronal antibodies was made only in one of these last four patients. The thyroid function tests were normal in these four patients. Overall, we considered the diagnosis of paraneoplastic autoimmune LE in seven patients and anti-NMDA-R LE in two patients.
In one patient, we made the diagnosis of sarcoidosis LE whose observation was previously published (Toudou-Daouda, Assadeck, & Efared, 2017). The diagnosis of sarcoidosis LE has been confirmed by the histological examination of nasal biopsy that showed a granulomatous inflammation made of confluent granulomas with multinucleated giant and epithelioid cells surrounded by a rim of lymphocytes without caseous necrosis. In this patient, the thoracic-abdominopelvic CT scan revealed mediastinal and hilar lymphadenopathies without parenchymal lung lesions. In addition, the dosage of angiotensin-converting enzyme revealed high plasma levels. The MRI abnormalities in this patient were hyperintensities in right temporo-insular region on FLAIR sequences and T2-weighted images and on gadolinium-enhanced T1-weighted images, an enhancement and nodular leptomeningeal thickening in the basilar perimesencephalic cistern extended to the right temporal lobe, hypothalamus, and third ventricle. We also made the diagnosis of tuberculous LE in one patient whose observation was previously published (Toudou Daouda, Obenda, TA B L E 1 Demographic, clinical, and etiological features, as well as the medium-term prognosis of the 31 patients  as well as thoracic-abdominopelvic CT scan. Only six of the 12 patients had benefited from HSV-PCR, which was negative.
Serological tests for syphilis in the CSF were negative in the 12 patients. The thyroid function tests were performed in only two of the 12 patients and were negative.  and a predominance of syphilis among the infectious causes (5/9

| D ISCUSS I ON
cases; 55.6%). In addition, we found a hospital frequency of LE of 0.81%.
The clinical manifestations of LE in our study were polymorphic, but very typical of those reported in the literature: acute or subacute onset of anterograde memory disorders, psychiatric disorders, confusional state, temporal lobe, or generalized epilepsy (Anderson, & Barber, 2008;Corsellis et al., 1968;Geisler et al., 2013;Kerling et al., 2008). Epileptic seizures were the most common clinical manifestations in our study (54.8%), which is consistent with the literature data (Aupy et al., 2013;Irani, Bien, & Lang, 2011;Irani, Michell, et al., 2011;Vincent et al., 2004). In the literature, the clinical picture is essentially subacute (Alamowitch et al., 1997;Corsellis et al., 1968;Gultekin et al., 2000;Lawn, Westmoreland, Kiely, Lennon, & Vernino, 2003;Vedeler et al., 2006), whereas in our study, the F I G U R E 3 It is a 41-year-old man without no past medical history, admitted for generalized epilepticus status. Brain MRI shows on T2weighted images (a) and FLAIR sequences (b) hyperintensity in the right temporo-insular without enhancement on gadolinium-enhanced T1-weighted images (c). Diagnosis of syphilitic LE was made after the biological investigations It is a 65-year-old man without no past medical history, admitted for the febrile confusional syndrome. Brain MRI shows on T2-weighted images (a) and FLAIR sequences (b) bilateral hyperintensity in temporo-insular regions predominant on the right. Diagnosis of herpetic LE was made after the biological investigations onset was acute (1-7 days) in 64.5% of cases. This could be explained by the fact that infectious causes are more common in our study because these etiologies usually have an acute onset of symptoms (Geisler et al., 2013;Hirai et al., 2005). The autoimmune LE most often has a subacute onset of symptoms.
Functional cerebral imaging (SPECT or FDG-PET) is the most sensitive radiological examination in the diagnosis of LE allowing to highlight signs of hypermetabolism in the temporal and extratemporal regions (Franck et al., 1987;Provenzale, Barboriak, & Coleman, 1998;Fakhoury, Abou-Khalil, Kessler, 1999;Na, Hahm, Park, & Kim, 2001;Kassubek et al., 2006). This examination was not performed in our study because it is not available in our hospital, and its practice was not indispensable because all patients had positive brain MRI.
EEG is a frequently performed examination in cases of suspected LE, but its practice is not required to make the diagnosis of LE. The EEG abnormalities commonly reported are diffuse slowing without epileptiform activity, slowing of the background rhythm, epileptiform activity in the temporal or frontotemporal regions, periodic lateralized epileptiform discharges in the temporal regions (Kerling et al., 2008;Anderson, & Barber, 2008;Asztely, & Kumlien, 2012;Franck et al., 1987;Fakhoury, Abou-Khalil, Kessler, 1999;Scheid, Lincke, Voltz, Von Cramon, & Sabri, 2004;Ances et al., 2005). In our study, the main EEG abnormalities included the slow waves in the temporal regions, slowing of the background rhythm, spikes, and waves in the frontotemporal regions.
We found in this study 12 cases (38.7%) of LE without definite etiology, but with an incomplete diagnostic workup. We think that among these patients, some of them could have probably diagnosis of herpetic LE because of the acute onset of symptoms and the clinical improvement under antiherpetic treatment, and knowing that only six of these patients had benefited from HSV-PCR in CSF.
On the other hand, these patients may simply correspond to patients who spontaneously improve and who would have had a favorable evolution without antiherpetic treatment.
Of the infectious causes of LE, HSV is the most common infectious agent (Asztely, & Kumlien, 2012). Other infectious causes are also reported such as syphilis, VZV, and more rarely tuberculosis.
The case of tuberculous LE reported in our study is the second case reported in the literature to our knowledge after that reported by Sonkaya et al. (2014).
The paraneoplastic autoimmune LE seems to be the most common causes of LE in our study (7/31 cases; 22.6%). The diagnosis of paraneoplastic autoimmune LE was made in our patients according to the diagnostic criteria for autoimmune LE of Graus et al. (2016) which are (a) subacute onset (rapid progression of <3 months) of working memory deficits, seizures, or psychiatric symptoms suggesting involvement of the limbic system, (b) bilateral brain abnormalities on T2-weighted FLAIR MRI highly restricted to the medial temporal lobes, (c) CSF pleocytosis (white blood cell count of more than five cells per mm 3 ) or EEG with epileptic or slow-wave activity involving the temporal lobes, and (d) reasonable exclusion of alternative causes.
When the four criteria mentioned above are met, the positivity of antineuronal antibodies is not indispensable to consider LE as having a definite autoimmune origin. In addition, studies have shown that autoimmune LE can occur without detectable autoantibodies (Graus et al., 2008;Najjar, Pearlman, Zagzag, & Devinsky, 2011). However, the detection of antineuronal antibodies may be important to determine the immunological type of LE, to guide to an associated tumor, and to evaluate the prognosis which might differ according to the type of antibody (Graus et al., 2016;Höftberger et al., 2015;Malter et al., 2014).
On the other hand, if one of the first three criteria above mentioned is not met, the diagnosis of autoimmune LE can also be made when the antineuronal antibodies are positive (Graus et al., 2016).
Anti-NMDA-R encephalitis affects mainly women, and the clinical manifestations included abnormal behavior, memory disorders, speech disorders, epileptic seizures, abnormal movements (orofacial, limb, or trunk dyskinesias), loss of consciousness or autonomic dysfunction, central hypoventilation, and cerebellar ataxia or hemiparesis (Titulaer et al., 2013). Our two patients with anti-NMDA-R LE were women. The clinical manifestations were characterized by temporal lobe epilepsy, abnormal behavior, memory disorders and confusional syndrome in one patient, and epileptic seizures associated with abnormal behavior and language disorders in the other patient.
One patient had anterograde amnesia as sequelae, while the other patient recovered completely. A study showed that the patients with anti-NMDA-R LE have a better prognosis when a tumor is discovered and resected than the patients without any tumor found (Florance et al., 2009). The complete remission in our patient with anti-NMDA-R LE may simply correspond to a spontaneously favorable evolution, and the immunomodulatory treatment would not have been necessary to permit de complete remission as previously described in the literature (Iizuka et al., 2008). Our study has some limitations. First, the diagnostic tools of LE are poorly developed in our hospital in particular and in Morocco in general: (a) lack of functional imaging (SPECT and FDG-PET) not only for the positive diagnosis of LE but also for research of associated cancer (FDG-PET/CT full body), (b) limited access to systemic immunological tests, antineuronal antibodies, and HSV-PCR. Secondarily, our study had not included patients with negative brain MRI because of the diagnostic criteria for autoimmune LE of Graus et al. (2016) are not met in these patients. In addition, there is no evidence of a particular infectious cause especially herpes infection. Thus, it is possible that a good part of the patients with autoimmune LE with negative brain MRI or patients with benign herpetic LE who had not benefited from HSV-PCR in the CSF has not included in this study.
Thirdly, the retrospective nature of this study explains why some details were not provided especially some clinical data and treatments.

ACK N OWLED G M ENTS
None.

CO N FLI C T O F I NTE R E S T S
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

AUTH O R S' CO NTR I B UTI O N S
AFA and AS conceived and designed the study, collected and analyzed the data. MTD wrote the first draft of the manuscript in its entirety and participated to the interpretation of the MRI and EEG.
MFB and ZS participated in the design and conception of the study, interpreted MRI and EEG, and had critically revised the manuscript. All authors have seen and approved the final version of the manuscript.