A 65-year-old right-handed retired woman presented with a depressive relapse. Her medical history revealed hypertension, hyperlipidemia, and rheumatoid arthritis in clinical remission for 6 years (DAS 28 score: 1.74), treated weekly with methotrexate [15 mg once daily (od)]. She reported neither a history of addictive behavior nor any neurological medical condition. She suffered from a TRD of 25 years (recurrent resistant episodes, with resistance to several types of antidepressant treatment, including resistance to electroconvulsive therapy), without any event of suicidal behavior or hypomanic/manic episode. She was treated with fluoxetine (40 mg od) and clomipramine (75 mg od) before relapsing. At presentation, she reported sadness, loss of interest, psychomotor retardation, thoughts of worthlessness, hyporexia, and difficulty concentrating. No agitation, irritability, insomnia, confusional state, or suicidal thoughts were reported. She was diagnosed with a severe TRD according to the Diagnostic and Statistical Manual of Mental Disorders IV-TR.

Considering her TRD and clomipramine-induced orthostatic hypotension, treatment was progressively switched to mianserin, sertraline, and diazepam. LF-rTMS was initiated 2 weeks after the onset of the new pharmacological treatment [at an intermediary dosage, mianserin (10 mg od), sertraline (50 mg od), and diazepam (2 mg od)], over the rDLPFC [Magstim Rapid² stimulator (Whitland, UK, www.magstim.com), using a 70-mm diameter figure-of-eight coil, at 110% of resting motor threshold, with continuous patterns at 1 Hz (1,000 pulses/day) which followed international guidelines for clinical rTMS uses (1)].

The cortical target was defined as the scalp location 6 cm rostral and in a parasagittal plane from the hotspot of the abductor pollicis brevis muscle in the primary motor cortex (M1). Pretreatment assessments including standard blood tests, brain computerized tomography scanner, and electroencephalography (EEG) were normal.

After the eighth rTMS session, the patient presented significant behavioral changes, including disinhibition, loss of manners, decrease in personal grooming, perseverative behaviors, hyperorality with sweet cravings, and impulsive actions (the patient ran away from the hospital to buy large amounts of chocolate at the supermarket). Aspontaneity and economy of speech without aphasia were also observed. She did not exhibit elevated mood, tachypsychia, logorrhea, or insomnia, excluding a diagnosis of hypomanic/manic episode. Moreover, the patient did not report any delusion or hallucination. On a neurological evaluation, the patient was alert and oriented, without any motor or sensitive deficit. However, the patient exhibited signs of right frontal lobe dysfunction including distractibility, environmental dependency, and left-sided visuospatial neglect (9).

Considering this acute frontal syndrome, we conducted an extensive checkup in order to rule out any acute neurological etiology. Epilepsy was ruled out with an EEG conducted during the acute symptomatic phase, revealing no ictal activity, notably in the frontal and temporal areas (even after provocation with hyperventilation test and flicker-fusion test). A second control EEG was conducted 7 days after the first EEG, revealing no abnormalities. Viral and bacterial encephalitis were ruled out by normal blood tests and cerebral spinal fluid (CSF) analysis [normal cell counts including no pleocytosis (0 cells), normal CSF protein level (0.33 g/L), normal CSF glucose level (3.5 mmol/L), negative HSV1/2-PCR, negative 14-3-3 protein]. General metabolic disturbances were ruled out by normal laboratory tests (normal glycemia, calcemia, and natremia). In order to rule out the hypothesis of an acute ischemic stroke, a 1.5 T brain magnetic resonance imaging (MRI) was performed on day 1. Right DLPFC hyperintensity on diffusion-weighted imaging was reported with a significant decreased apparent diffusion coefficient (ADC) as compared to the homotopic region [difference of 0.039 × 10⁻³ mm²/s, showing 0.723 in the rDLPFC versus 0.762 in the left DLPFC (×10⁻³ mm²/s), see Figure 1]. Nonetheless, these findings lacked an association with a high signal in FLAIR sequence more than 24 h after the onset of symptoms, and ADC level in our patient was higher than ADC level seen in patients suffering from an acute ischemic stroke (ADC mean in acute ischemic stroke = 0.533 × 10⁻³ mm²/s, SD = 0.157, 95% confidence intervals: 0.501–0.563) (10). The hypothesis of an acute ischemic stroke was then ruled out. In the context of rheumatoid arthritis, it was important to rule out the hypothesis of a cerebral vasculitis. First, considering rheumatoid arthritis, our patient had been in clinical remission for 6 years. There was no systemic sign of inflammatory activity, on a clinical and paraclinical level (including fever, fatigue, arthralgia, edema, cardiopulmonary function, renal and hepatic function, peripheral neuropathy, ophthalmic abnormalities). Second, blood tests were unremarkable {no inflammatory syndrome [normal C-reactive protein, normal white blood cell count, normal albuminemia (38 g/L)], normal serum protein electrophoresis}, and there was no inflammatory process revealed by CSF analysis (as already described). Furthermore, ADC level was not consistent with ADC level seen in acute ischemic stroke of any kind, including acute stroke induced by cerebral vasculitis (10). As the patient was treated with methotrexate, we aimed to address its imputability in the occurrence of such acute frontal syndrome. Methotrexate intake has always been well tolerated in our patient for 6 years. Our patient did not show any hypersensitivity reaction to methotrexate. She was supplemented with B vitamins, and laboratory tests revealed no vitamin deficiency. Finally, the period of time between the last intake of methotrexate and the acute phase had to be taken into account. Nonetheless, the reversibility of frontal symptoms was not correlated with a modification of this specific treatment. All put together, there was no argument in favor of a methotrexate-induced encephalitis.

After having ruled out these hypotheses, we tried to specify the frontal lobe dysfunction by conducting a neuropsychological evaluation in combination with other brain imaging techniques. Neuropsychological evaluation revealed deficits in executive and working memory tasks, without any instrumental impairment (see Table S1 in Supplementary Material). A brain 99mTc-hexamethylpropyleneamine oxime single-photon emission computerized tomography (99mTc-HMPAO SPECT) performed 13 days after the onset of symptoms revealed a significant decrease in regional cerebral blood flow (rCBF) within the rDLPFC (see Figure 2).

Repetitive TMS and sertraline were discontinued and the patient’s frontal lobe syndrome started to decline after 10 days off treatment. A control MRI performed on day 9 showed normal ADC levels in the rDLPFC [no significant difference, showing 0.803 in the rDLPFC versus 0.785 in the left DLPFC (×10⁻³ mm²/s), see Figure 1]. At the 3-month follow-up examination, the patient was stabilized on a regimen of mianserin 30 mg od in a complete recovery state, with a regression of the frontal dysfunction and a normalization of all neuropsychological functions. A 10-month control 99mTc-HMPAO SPECT revealed a complete cancelation of the rDLPFC rCBF decrease (see Figure 2).