Is headache always necessary for the diagnosis of rebound intracranial hypertension in patients successfully treated for cerebrospinal fluid venous fistula?

A young female in her 20s with Ehlers–Danlos syndrome initially presented to a community hospital with several months of orthostatic headache and postural tinnitus. The headache was occipital-predominant, pressure-like in quality, reliably precipitated by upright posture within 20–30 min, and relieved by recumbency. Her symptoms intensified later in the day and evenings. The tinnitus was similarly posture-dependent, described as a low-pitched pulsatile “fluid-like” sound. She had no prior history of migraine or obesity. She had a neurological assessment that revealed no evidence of intracranial hypertension. She underwent two nontargeted lumbar epidural blood patches (EBPs) of undocumented volume, with only temporary and partial benefit, before presenting 4 months later to the emergency department with severe orthostatic headache and nausea prompting transfer to our institution, a tertiary neurovascular center. She had two brain magnetic resonance imaging (MRI) scans over the course of her disease before the transfer in which both revealed no pachymeningeal enhancement, subdural collections, venous engorgement, or brain sagging (Bern score 0). Spinal MRI with a cerebrospinal fluid (CSF) leak protocol showed no spinal longitudinal epidural collections. On transfer to our institution, a decubitus computed tomography myelography was performed that identified two cerebrospinal fluid-venous fistulas (CSFVFs) at right T10 and T12. She underwent transvenous embolization, resulting in complete symptom resolution. After an asymptomatic interval of 2 months, her orthostatic headache and tinnitus recurred and progressively intensified requiring an admission at a community hospital where repeat brain and spine imaging demonstrated no relevant findings of brain sag (Bern score of 0) or spinal longitudinal epidural collections. A low volume (20 cc) nontargeted blood patch was also performed with no symptomatic relief. This again prompted transfer to our institution where a repeat decubitus computed tomography myelography demonstrated low probability of recurrent CSFVF at the previously treated right T12 level and a new CSFVF at the left T12 level, although findings were not discrete enough for re-embolization. A multidisciplinary decision was made to pursue an empiric high-volume EBP as a diagnostic and therapeutic modality. She underwent a 50 mL EBP at the level of suspicion, T11/12. The procedure was performed under procedural sedation with iv midazolam 1 mg and fentanyl 75 μg with continuous cardiac monitoring and frequent symptom checks specifically for back pressure, radicular symptoms, and paresthesia to gauge the amount of volume injected into the epidural space during the procedure aiming to minimize side effects and complications. On post-procedural day 1, the patient developed new-onset intractable, nonprojectile nausea and vomiting while supine. She denied headaches, fever, photophobia, neck stiffness, or visual changes, and specialized neurological assessment was essentially unremarkable. A multidisciplinary review (including neuroradiology, anesthesiology, and neurology) judged that the presentation could not be fully explained by rebound intracranial hypertension (RIH) in the absence of recumbent headache, and the symptoms were therefore considered indeterminate, with the possibility of representing a post sedation side effect. She was managed with iv ondansetron 8 mg every 8 h as needed while hospitalized, which provided partial relief. She was discharged at 48 h with instructions to continue lying flat for an additional 72 h. On telephone follow-up 5 days later, the patient reported complete resolution of her orthostatic headache and tinnitus within 4 days of the EBP, but she continued to experience nausea and vomiting without headache, worse when supine and earlier in the day. She denied visual changes, diplopia, or additional neurological symptoms. In view of the temporal association with EBP, reversal of the spontaneous intracranial hypotension (SIH) symptom profile, and persistence of symptoms, RIH was considered the most likely diagnosis at that time. She commenced acetazolamide 250 mg twice daily, and on repeat telephone follow-up after 2 days she reported complete resolution of her symptoms. She was advised to continue the same dose for an additional 7 days, with follow-up arranged thereafter. She did not report any adverse side effects. The course of medical treatment with acetazolamide was concluded by day 7. At 1- and 3-month follow-up, the patient continues to report no new symptoms, headaches, nausea, or vomiting. Her presumed baseline CSF leak headache had completely resolved and has been symptom free for (3 months) up to the point of writing this report. The day-1 onset of nausea after procedural sedation with midazolam and fentanyl could represent a transient sedation-related adverse effect that is generally recognized in clinical practice.1 However, the clinical profile was not typical for post-sedation sequelae: symptoms demonstrated clear postural and diurnal dependence and persisted beyond the immediate recovery period despite antiemetic therapy. The brisk, durable response to acetazolamide further argues against a purely sedation-related phenomenon. The timing after leak closure, reversal of the presenting phenotype (orthostatic headache and tinnitus resolved, gastrointestinal symptoms emerged), and the positional character of nausea favored RIH. Although the absence of a recumbent headache on day 1 argued against the classic presentation, it does not exclude a high-pressure shift, and the rapid, durable improvement on acetazolamide supports this diagnosis. Nausea may accompany ongoing SIH, so this remained a consideration early in the course. Nevertheless, the complete resolution of orthostatic headache and tinnitus within 4 days of the targeted EBP strongly suggested leak closure, and subsequent symptoms were dominated by positional nausea/vomiting without cephalalgia—features that are less consistent with persistent hypotension. SIH arises from spinal CSF leaks,2 with a clinical hallmark of orthostatic headache, although associated symptoms include tinnitus, diplopia, and dizziness.3 Diagnostic imaging may demonstrate brain sagging, pachymeningeal enhancement, venous engorgement, or subdural collections. These features are the basis of Bern score that is used to stratify patient's probability of having a spinal leak. A recently publish meta-analysis reported that nearly 18% of patients with a low Bern score were found to have a CSF-venous fistula.4 Treatment options include EBP, fibrin sealant, and targeted repair of CSFVFs through surgical ligation or transvenous embolization.5 Although often effective, these interventions can precipitate RIH, which is reported in roughly a quarter of treated patients with SIH in single-center series published in 2021 (overall ~27%, ranging ~21%–29% by treatment modality).6 RIH typically presents with frontal or peri-orbital headache, often recumbency-predominant, accompanied by nausea, vomiting, or visual symptoms.7 The mechanism of EBP is thought to involve both a transient mass effect, which increases CSF pressure and the formation of a gelatinous clot that seals the dural tear. Generally, patients are advised to avoid intense activity and Valsalva maneuvers in the days following the procedure; furthermore, severe nausea and vomiting, through repeated Valsalva efforts, may compromise clot stability and risk patch failure as the clot requires time for stabilization and maturity.8 A favorable response to EBP is usually defined by clear improvement or resolution of orthostatic headache within a few days of the procedure. This case illustrates an atypical RIH presentation after successful treatment of underlying CSFVF, highlighting diagnostic and management considerations for trainees. The primary goal of an EBP is to seal the CSF leak and relieve orthostatic headache. In this patient, that goal was achieved, as her headache and tinnitus resolved completely. The diagnostic challenge arose following her EBP when she developed persistent nausea and vomiting without headache. Nausea as an isolated presentation can be challenging to interpret. On one hand, nausea is often associated with a favorable response in SIH. Several studies have sought to identify clinical features that predict such success. In a retrospective series of 118 patients, it was found that nausea and diffuse headache were significantly associated with a good response to EBP, whereas male sex, focal headache, photophobia, and phonophobia were predictors of poorer response.9 On the other hand, forceful vomiting increases intra-abdominal and intrathoracic pressure, which can in turn raise intraspinal pressure and risk dislodging the blood clot formed by the EBP. However, when postural headache has resolved, isolated nausea and vomiting may instead reflect rebound intracranial hypertension. At the same time, vomiting itself is undesirable after EBP, as repeated Valsalva maneuvers can increase intraspinal pressure and risk dislodging the fragile clot that seals the leak.8, 10 For this reason, controlling nausea and preventing vomiting are essential, not only for patient comfort, but also to protect the success of the patch. Most published reports of RIH have been headache-centric. A previously published study defined RIH chiefly as a new headache emerging after leak closure, typically frontal or peri-orbital and worse when supine, and warned that this pattern can be misread as persistent SIH unless the change in phenotype is recognized.7 To systematize recognition, Schievink and colleagues6 proposed pragmatic features that also revolve around headache: a reverse orthostatic headache distinct from the presenting SIH headache, resolution of that headache with oral acetazolamide, and exclusion of other headache causes. This historical focus on headache helps explain why nonheadache–dominant presentations (e.g., isolated nausea and vomiting) may be under-appreciated in clinical practice. Nausea and vomiting have not been reported in isolation as a presenting symptom for RIH although they have been well-recognized as associated features of rebound intracranial hypertension, consistently reported in clinical series.7, 10, 11 Additionally, a subset of SIH patients may have developed CSF leak in the setting of chronic idiopathic intracranial hypertension (IIH), such that treating the leak may unmask symptoms of IIH.12 It has been established that IIH can manifest in patients who lack headaches.13 Furthermore, nausea and vomiting in isolation have been reported as a presentation of IIH.14 Nausea and vomiting in the setting of RIH may be explained by the rise in intracranial pressure driven by changes in venous outflow and craniospinal elastance after leak closure.15 The generation of nausea in the setting of rebound intracranial hypertension can be also understood through the central emetic pathways that converge on the area postrema (AP) as a suggested plausible mechanism. The AP is a circumventricular organ located at the base of the fourth ventricle with a permeable blood–brain barrier, allowing it to detect circulating signals and regulate nausea and vomiting through its connections with the nucleus tractus solitarius and dorsal vagal complex.16 Injury or stimulation of this region produces profound nausea and vomiting, as seen in AP syndrome, underscoring its central role in emetic regulation.16 Beyond the AP, vestibular inputs play a key role in modulating nausea. Experimental work has demonstrated that gastrointestinal emetic stimuli can alter vestibular signal processing, highlighting how integration between vestibular and visceral pathways contributes to nausea and motion sickness.17 In our case, the complete resolution of orthostatic headache and tinnitus argued against patch failure. The emergence of new, positional nausea and vomiting without cephalalgia suggested a reversal of the symptom profile, favoring RIH. The patient's rapid improvement on acetazolamide supported the suspected diagnosis as the best explanation for the patient's symptoms. Rayyan Kinsara: Conceptualization; investigation; writing – original draft; writing – review and editing; visualization; validation; methodology. Anish Kapadia: Conceptualization; methodology; validation; visualization; writing – review and editing; supervision. Howard Meng: Supervision; conceptualization; investigation; visualization; validation; methodology; writing – review and editing. Rayyan Kinsara, Anish Kapadia, and Howard Meng declare no conflicts of interest related to this work. Written informed consent was obtained from the patient for publication of this case report and any accompanying images.