Posterior Reversible Encephalopathy Syndrome (PRES) following gastrointestinal sepsis in a child: a case report

Abstract

Background

Posterior Reversible Encephalopathy Syndrome (PRES) is a rare neurological entity in the paediatric population, whose diagnosis relies primarily on brain MRI. Although at-risk populations are well established, cases occurring in the context of sepsis of gastrointestinal origin are rare, particularly in paediatric intensive care settings in sub-Saharan Africa. Case presentationA 12-year-old girl with no medical history was admitted for peritonitis secondary to ileal perforation, managed after a two-week delay. The initial presentation combined severe sepsis, grade 2 arterial hypertension, acute kidney injury, severe hypoglycaemia and metabolic acidosis. Postoperatively, two haemodialysis sessions corrected renal function and resolved the choreiform movements and bruxism, whereas arterial hypertension and agitation persisted. On the tenth postoperative day, the onset of refractory seizures and bilateral cortical blindness prompted brain MRI, which revealed bilateral symmetrical parieto-occipital vasogenic oedema — involving the medial occipital lobes and the calcarine cortex — with cerebellar involvement and haemorrhagic stigmata on susceptibility-weighted sequences, confirming the diagnosis of PRES. Intravenous nicardipine and levetiracetam were initiated. The outcome was favourable, with complete neurological recovery and full visual restitution after 65 days.

Conclusion

This case illustrates the diagnostic complexity of PRES in paediatric intensive care within a resource-limited setting. It suggests that sepsis of gastrointestinal origin may contribute to the occurrence of PRES both directly and by initiating associated dysfunctions — arterial hypertension, acute kidney injury, hypovolaemia and uraemic encephalopathy — acting as contributory cofactors. The diagnosis of PRES was retained as predominant on the basis of brain imaging and clinical reversibility under treatment. Finally, this case underscores that the timeliness of diagnosis and management of the triggering factor — peritonitis — is an essential determinant of neurological outcome.

Key points

• PRES should be considered in any acute neurological presentation occurring in a septic context, including in paediatric patients.

• Sepsis of gastrointestinal origin, although rarely reported, may act as a triggering factor for PRES, both directly through the systemic inflammatory response and indirectly by initiating associated dysfunctions — arterial hypertension, acute kidney injury, hypovolaemia and uraemic encephalopathy.

• The persistence of neurological signs despite correction of metabolic disorders and renal function should prompt reconsideration of a metabolic encephalopathy diagnosis and lead to brain MRI.

• Involvement of the calcarine cortex and medial occipital lobes (structures classically spared in PRES) accounts for cortical blindness and may be associated with prolonged visual recovery.

• Timeliness of diagnosis and management of the triggering factor is an essential determinant of neurological outcome, including in resource-limited settings where access to MRI remains constrained.

Introduction

Posterior Reversible Encephalopathy Syndrome (PRES) is a neurological entity with heterogeneous clinical manifestations [1, 2]. In the absence of specific clinical signs, its diagnosis relies primarily on brain MRI, which typically demonstrates white-matter vasogenic oedema predominating in the occipital and, at times, parietal lobes [1]. Although initially described in adults, PRES is also encountered in the paediatric population [3–7].

The populations at risk of developing this radioclinical syndrome are well characterised [1]. More recently, sepsis has been identified as an independent risk factor, with a limited number of cases reported in the literature [8, 9]. While paediatric cases of PRES associated with sepsis have been described, those arising in the context of sepsis of gastrointestinal origin remain rare [3, 5]. In sub-Saharan Africa, and particularly in Côte d’Ivoire, this entity does not appear to have been documented in paediatric intensive care, likely owing to diagnostic constraints [10].

We report here a case of PRES revealed by seizures and blindness in an adolescent girl admitted for acute peritonitis with delayed management.

Case presentation

A 12-year-old adolescent girl with no notable medical history, living in a rural area, was admitted to the paediatric emergency department for abdominal pain associated with unquantified prehospital fever and deterioration of her general condition, evolving over approximately two weeks. The working diagnosis was peritonitis with signs of dehydration. While awaiting surgery, she experienced two seizures treated with two 5-mg doses of midazolam, prompting her admission to the Intensive Care Unit (ICU).

On admission, clinical examination revealed arterial hypertension (160/90 mmHg), tachycardia (92 bpm) and mild jaundice. Consciousness was impaired (Glasgow Coma Scale 10 — E2V2M6), with intermittent agitation and choreiform movements. There was no sensorimotor deficit and no meningeal syndrome. Respiratory status was stable (SpO₂ 96% on room air). Laboratory investigations revealed severe hypoglycaemia (0.3 g/L), impaired renal function, signs of hepatocellular insufficiency and severe metabolic acidosis (Table 1). Initial management consisted of emergency glucose administration, rehydration, dual antibiotic therapy (piperacillin–tazobactam and amikacin), blood transfusion and vitamin K1 supplementation. Exploratory laparotomy, performed approximately six hours after ICU admission, revealed a single ileal perforation without mesenteric lymphadenitis. A segmental ileal resection with diverting stoma was performed.

Table 1.

Summary of the kinetics of biological parameters during the clinical course of our patient

Parameters (units)	Admission	D2	D10 (Post-RRT)	D20 in ICU	Normal values
WBC (cells/mm³)	24,800	30,400	17,800	9,910	[4,000–10,000]
Hb (g/dL)	11.8	12.3	8.9	10.2
Hct (%)	34	36.5	26.9	34.7	[36–46]
Platelets (cells/mm³)	121,000	180,000	367,000	257,000	[150,000–400,000]
Natremia (mEq/L)	130	129	136	143	[135–145]
Kalemia (mEq/L)	5.1	5.4	3.79	3.57	[3.5–5.0]
Chloremia (mEq/L)	91	86	99	105	[95–105]
Calcemia (mg/L)	87	—	—	78	[88–104]
Magnesemia (mg/L)	19	—	—	14	[17–24]
CRP (mg/mL)	234	98	46	—	< 6
PCT (ng/mL)	65.88	—	12	—	< 0.5
ASAT (IU/L)	108	89	—	—	< 40
ALAT (IU/L)	67	31	—	—	< 40
Urea (g/L)	2.28	3.38	0.49	0.38	[0.15–0.45]
Creatinine (mg/L)	57	80	28	12	[7–13]
pH	7.12	7.34	7.40	—	[7.38–7.42]
PaO₂ (mmHg)	95	101	102	—	[80–100]
PaCO₂ (mmHg)	27.5	25	32	—	[35–45]
HCO₃ (mmol/L)	10.4	13.5	22	—	[22–26]
Lactate (mmol/L)	5.6	3.8	0.7	—	≤ 2

Abbreviations: RRT Renal Replacement Therapy, Hb Hemoglobin, Hct Hematocrit, CRP C-Reactive Protein, PCT Procalcitonin, WBC White Blood Cells, D Day, ICU Intensive Care Unit

In the immediate postoperative period, the patient remained agitated, with persistent choreiform movements and intermittent bruxism. These neurological disturbances, together with impaired renal function and oedema, raised the hypothesis of uraemic encephalopathy. Two haemodialysis sessions were performed, improving consciousness (Glasgow 15) and renal function (Table 1) and resolving the choreiform movements and bruxism. However, agitation and arterial hypertension persisted (systolic blood pressure 140–160 mmHg), without notable improvement despite renal replacement therapy.

On postoperative day 10, the course was marked by the recurrence of midazolam-refractory seizures and the onset of bilateral blindness. Seizures were controlled by the addition of levetiracetam. Brain MRI demonstrated cortical white-matter hyperintensity without restriction of the apparent diffusion coefficient (ADC), consistent with vasogenic oedema. The lesions were diffuse, bilateral, symmetrical, with a posterior parieto-occipital predominance and associated cerebellar involvement. They showed no enhancement after gadolinium administration. Haemorrhagic stigmata were also noted as rare hypointensities on susceptibility-weighted sequences (Fig. 1). PRES was diagnosed, and antihypertensive therapy with intravenous nicardipine was initiated, targeting a systolic blood pressure below 140 mmHg. The patient spent 17 days in intensive care before transfer to paediatric surgery, where intestinal continuity was subsequently restored. The outcome was favourable, without seizure recurrence. Full visual recovery was achieved after approximately 65 days.

Fig. 1.

Brain MRI at postoperative day 10. a Diffusion-weighted sequence showing vasogenic oedema as hyperintense areas (yellow arrows) without true ADC restriction. b T2/FLAIR sequence showing occipital hyperintensity (green arrows). c Susceptibility-weighted (SWAN) sequence showing hypointense foci (blue arrows) in the bilateral occipital and cerebellar regions, consistent with haemorrhagic stigmata

Figure 2 outlines the sequence of bioclinical and therapeutic events, which can be summarised in four phases: an initial phase (Day 0) marked by sepsis with multi-organ dysfunction and preoperative seizures; an early postoperative phase (Days 2–9) characterised by persistent arterial hypertension and neurological disturbances despite correction of acute kidney injury and the other associated metabolic disorders; the emergence of blindness on Day 10 together with refractory seizures, which prompted brain MRI and led to the diagnosis of PRES; and finally a phase of progressive recovery from Day 17 onward, with complete visual restitution by Day 65.

Fig. 2.

Chronologie des évènements bio cliniques et thérapeutiques

Discussion

We report a case of Posterior Reversible Encephalopathy Syndrome (PRES) occurring in an adolescent girl following acute peritonitis due to ileal perforation, revealed by status epilepticus associated with transient cortical blindness. To our knowledge, this is one of the few paediatric cases of PRES secondary to sepsis of gastrointestinal origin reported in sub-Saharan Africa. The rarity of this association warrants both a comparison with the existing literature (Table 2) and a discussion of the underlying pathophysiological mechanisms.

Table 2.

Comparative bioclinical and outcome characteristics of pediatric PRES cases secondary to sepsis of gastrointestinal origin

Parameters	Our case	Saley [3]	Rafee [5]
Age	12 years	5 years	11 years
Sex	Female	Female	Male
Medical history	None	None	None
Gastrointestinal lesion	Ileal perforation	Perforated appendicitis	Perforated appendicitis
Identified organisms	Not specified	E. coli, B. fragilis	E. coli, P. aeruginosa, S. constellatus, Bacteroides
Delay before admission	≈ 2 weeks	Prompt	7 days
Hypertension at admission	Grade 2	Grade 2	Grade 2
Organ dysfunction	Acute kidney injury	Pulmonary edema	None
Seizures	Yes	Yes	Yes
Visual disturbances	Yes (Cortical blindness)	No	No
Hemorrhage on MRI	Yes	No	No
Antiepileptic treatment	Benzodiazepine, Levetiracetam	Benzodiazepine, Fosphenytoin	Lorazepam
Antihypertensive treatment	IV Nicardipine	None	Oral Isradipine
Time to recovery	65 days	25 days	—

Pathophysiology

Whatever the aetiology of PRES, the formation of vasogenic oedema through disruption of the blood–brain barrier (BBB) represents the final common pathway of all potential pathophysiological mechanisms. Classically, PRES results from severe arterial hypertension exceeding the limits of cerebral autoregulation, leading to passive cerebral hyperperfusion that damages the BBB and generates vasogenic oedema. In the setting of severe infection such as peritonitis, pro-inflammatory cytokines (TNF-α, IL-1β, IL-6) exert direct toxicity on the cerebral microvascular endothelium, impairing BBB integrity; this mechanism accounts for the rare cases of PRES occurring in normotensive septic patients [8, 9].

Nevertheless, other contributory factors may have directly participated in the occurrence of PRES in our patient. The PRES observed here appears to result from the convergence of several factors — arterial hypertension, sepsis-related systemic inflammation, uraemic encephalopathy and hypovolaemia. In our view, ileal perforation peritonitis therefore constitutes the triggering factor of this pathophysiological cascade rather than the direct cause of PRES. The arterial hypertension observed in our patient, which resulted from activation of the renin–angiotensin–aldosterone system (RAAS) and the release of endogenous catecholamines in response to septic stress, remained the predominant mechanism, potentiated by acute kidney injury, uraemic encephalopathy and hypovolaemia. It is presumably this interplay of factors, amplified by the two-week delay in surgical management, that accounts for the exceptional severity of the presentation.

Microbiology

The absence of microbiological documentation of the peritoneal fluid constitutes a notable limitation of our report. Nonetheless, given the high prevalence of Salmonella typhi as an aetiological agent of ileal perforation in our setting, a typhoid origin remains the most likely hypothesis, although it could not be formally confirmed.

Timing of onset

Determining the exact onset of PRES was challenging in our patient, owing to confounding factors such as severe hypoglycaemia and uraemic encephalopathy that could account for the earliest neurological signs. In retrospect, however, the manifestations most consistent with PRES — seizures persisting alongside bilateral blindness despite correction of hypoglycaemia and completion of haemodialysis — were already present during the early postoperative window (Days 2–9), before radiological confirmation on Day 10. Taking this window as the true point of clinical onset, our case is notable for an unusually early presentation compared with previous paediatric reports [5, 11], plausibly related to the estimated two-week delay in initial management.

Clinical features

Paediatric PRES most often occurs in children without prior medical history, in a septic context, with neurological manifestations dominated by seizures. These constitute the principal mode of presentation and frequently progress to status epilepticus, particularly in young patients [12]. Consciousness may remain preserved, as in our case. The occurrence of bilateral cortical blindness represents the most striking feature of our observation. Reported in 20–39% of PRES cases [1], it remains rare in paediatrics and, in our patient, resulted from bilateral vasogenic oedema of the occipital cortex, confirmed on MRI, without involvement of the anterior optic pathways.

Neuroimaging

MRI interpretation was performed by the radiology team of our university hospital, which includes an expert neuroradiologist. The diagnosis was based on clinical–radiological correlation [1]. The absence of external neuroradiological review may constitute a limitation (evaluation bias) of our report. From an anatomical standpoint, the bilateral cortical blindness is explained by the involvement of the medial occipital lobes and the calcarine cortex observed in our patient, the latter being the site of the primary visual cortex. Furthermore, the presence of haemorrhagic stigmata on susceptibility-weighted sequences (SWAN) is not uncommon in PRES (15–20%) [1] and is recognised as a marker of severity associated with prolonged recovery. These haemorrhagic foci did not influence our therapeutic strategy but were probably correlated with the prognosis — the prolonged visual recovery (65 days), markedly longer than the 20–30 days usually reported [3, 5]. Cerebellar involvement is considerably less frequent than parieto-occipital localisation; however, it may be observed in nearly half of patients (53%) in some series [13] and is more frequent in children [11]. Its presence in our patient, like the haemorrhagic stigmata, likewise reflects the severity of the presentation and is consistent with the prolonged visual recovery observed [12].

Follow-up MRI after clinical recovery could not be performed owing to the economic constraints of our setting (absence of health insurance). In the absence of control imaging, the complete clinical resolution of all neurological manifestations — cessation of seizures and full visual recovery — constitutes in itself a recognised marker of the reversibility of PRES [1, 8].

Differential diagnosis

In our patient, several confounding factors other than PRES could have accounted for the seizures. Hypoglycaemia, the first hypothesis raised on admission, was corrected as an emergency by glucose administration. The recurrence of seizures beyond this correction ruled out a hypoglycaemic origin of the encephalopathy. Hepatic encephalopathy was deemed unlikely, given the rapid decline in transaminases and the absence of asterixis or hepatic coma. Uraemic encephalopathy was ruled out given the persistence of neurological signs despite restoration of renal function after two haemodialysis sessions, and the absence of the lentiform fork sign on MRI — pathognomonic of uraemic PRES. The concurrent existence of sepsis-associated encephalopathy — a diagnosis of exclusion — could not, however, be formally ruled out. Nonetheless, in accordance with established diagnostic criteria [1], PRES was retained on the basis of a threefold convergence: clinical (persistent arterial hypertension, bilateral cortical blindness, refractory seizures), radiological (characteristic MRI lesions) and evolutive (improvement followed by complete reversibility of signs).

Therapeutic aspects

Management of our patient rested on seizure control, blood pressure management and treatment of the underlying sepsis. Antiepileptic treatment was conducted according to the usual protocols for status epilepticus [14]. It consisted of first-line therapy with a benzodiazepine (midazolam), followed by the introduction of a long-acting antiepileptic agent (levetiracetam). The latter was chosen for its favourable tolerability profile: cardiorespiratory safety and the absence of significant sedative effect, two properties particularly suited to the neurological monitoring of a patient in intensive care. Intravenous nicardipine was selected as the antihypertensive of choice, on account of its availability in our setting, with the aim of a gradual reduction in systolic blood pressure [15]. However, its positive chronotropic effect constitutes a limitation in certain haemodynamic situations, and labetalol may represent a more appropriate alternative where available. Of note, in some reports of septic PRES, antihypertensive therapy was not required [3], underscoring the importance of tailoring the blood-pressure strategy to each patient’s individual profile.

Outcome

In our observation, the outcome was favourable, with complete neurological recovery despite a prolonged visual restitution of 65 days. This unusual delay may be explained by two factors specific to our case: the delay in initial management — estimated at approximately two weeks — and the presence of haemorrhagic stigmata on MRI, whose association with a more severe form and prolonged recovery has been suggested. The absence of seizure recurrence and the full visual recovery confirm the reversible nature of PRES in our case, despite these features of severity

Limitations

Our observation has several limitations that should be highlighted. From a diagnostic standpoint, the absence of microbiological documentation of the peritoneal fluid precluded formal confirmation of the typhoid origin of the peritonitis, which was inferred on epidemiological grounds. Furthermore, the absence of a dedicated ADC map among the archived images limits the formal illustration of the absence of diffusion restriction, which was nonetheless confirmed at the initial reading. Regarding pathophysiology, given the intricate combination of multiple factors liable to contribute to the neurological presentation, sepsis of gastrointestinal origin could not be isolated as a sole causal factor, but rather as the triggering element of a complex pathophysiological cascade. Finally, the absence of follow-up MRI after clinical recovery precluded radiological documentation of lesion reversibility, which rested solely on the complete clinical resolution of the neurological manifestations.

These limitations, largely related to the constraints inherent to a resource-limited setting, do not, however, detract from the illustrative and educational value of this observation

Conclusions

This case illustrates the diagnostic complexity of PRES in paediatric intensive care within a resource-limited setting. It suggests that sepsis of gastrointestinal origin may contribute to the occurrence of PRES both directly and by initiating associated dysfunctions — arterial hypertension, acute kidney injury, hypovolaemia and uraemic encephalopathy — acting as contributory cofactors. The diagnosis of PRES was retained as predominant on the basis of brain imaging and clinical reversibility under treatment. Finally, this case underscores that the timeliness of diagnosis and management of the triggering factor — peritonitis — is an essential determinant of neurological outcome.

Author contributions

SSS : Conceptualization, Data curation, Writing – original draft, Writing – review & editing. KKLL : Data curation, Clinical management, Figures and tables design, Writing – review & editing. KKI : Data curation, Clinical management, Figures and tables design, Writing – review & editing. KKA : Data curation, Clinical management, Writing – review. ASCE : Investigation, Formal analysis, Figures and tables design, Writing – review & editing. BN : Supervision, Validation, Writing – review & editing. All authors read and approved the final manuscript.

Funding

No funding was received for this study.

Data availability

All data supporting the findings of this case report are included in this published article. Clinical, biological and radiological data are presented in the text, tables and figures.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Written informed consent was obtained from the patient’s legal guardian for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.

Competing interests

The authors declare no competing interests.