Osmotic Demyelination Syndrome (Central Pontine Myelinolysis)
An iatrogenic complication of rapid correction of hyponatremia (or osmotic): central pons (central pontine myelinolysis) and/or extrapontine demyelination. Typically occurs 2-6 days (range 1-14 days) AFTER Na correction, with a biphasic course: initial temporary improvement with hyponatremia correction, followed by encephalopathy, dysarthria, dysphagia, quadriparesis, pseudobulbar state, locked-in, seizures.
PREVENTION is essential: keep Na increase <8 mmol/L in the first 24 hours (high risk <6), <18 mmol/L in 48 hours; frequently monitor Na. If overcorrection is detected/anticipated, early relowering: with desmopressin + 5% dextrose, return Na to a safe limit.
In established ODS, supportive care + rehabilitation; some patients partially/totally recover over months/years.
Rapid management
- Two questions: (1) Was Na correction excessive (first 24h >8, high risk >6 mmol/L; or >18/48h)?
- If yes, IMMEDIATELY assess relowering (DDAVP + D5W) and stop active Na increase. (2) Is there new neurological deterioration 2-6 days after hyponatremia correction?
- If yes, consider ODS, request MRI (may be normal early, does not exclude), initiate support + rehabilitation.
| Drug | Adult | Pediatric | Route | Duration |
|---|---|---|---|---|
| Desmopressin (DDAVP) - stops and reverses overcorrection Stops renal free water excretion (aquaresis), prevents further rise of Na; lowers Na in a controlled manner with D5W. | 2-4 mcg IV/SC, every 6-8 hours (proactive 'clamp' or reactive) | 0.5-1 mcg IV/SC (under expert supervision; depending on Na response) | IV/SC | Continue until Na returns to target/safe limit; stop at Na ~125 mEq/L |
| 5% Dextrose (D5W) - lowers sodium (relowering) Administer together with DDAVP. Goal: Reduce 24h Na increase to <8 (high risk <6), total <18 in 48h. Discontinue both agents at Na 125. | 3 mL/kg/h IV; lowers Na by approximately 1 mmol/L/h | 3 mL/kg/h IV (weight-based); titrate according to Na response | IV | Continue until Na reaches the lower limit; then stop D5W, continue DDAVP |
| Stop active Na elevation (hypertonic/isotonic saline) Once hypovolemia/hypokalemia resolves, ADH is suppressed and unexpected rapid aquaresis (Na spike) occurs; early recognition of this transition is critical. | Reassess ongoing %3/%0.9 saline and K replacement; eliminate the source of overcorrection | Same principle; review all administered fluids and electrolytes | - | Until the relowering target is reached |
| Potassium replacement (hypokalemia cofactor) - use caution Hypokalemia independently increases ODS risk; K correction can contribute to overcorrection by increasing Na load. | Correct K deficit; K administration also increases effective Na - include in total correction calculation | Standard pediatric K replacement; monitor its effect on Na | IV/PO | Until K returns to normal |
| Supportive care + rehabilitation (established ODS) No specific evidence-based pharmacological treatment for established ODS; steroid/IVIG/plasmapheresis is not routinely recommended (CONFIRM). | Airway/aspiration protection, dysphagia management (NG/PEG), VTE prophylaxis, seizure treatment, physical + speech-swallowing rehabilitation, nutritional support | Same principles; with pediatric dosing and support | - | Weeks to months; recovery may be late and partial |
Alternative
- Relowering (Sterns approach): Once target is reached, stop D5W - continue DDAVP; if needed, add controlled %3 NaCl after 24-48h; discontinue both agents at Na 125 mEq/L.
- Prevention DDAVP 'clamp' strategy: In high-risk cases, ddAVP + controlled %3 NaCl from the start makes Na rise predictable (reduces risk of overcorrection).
- In established/severe ODS, steroid, IVIG, plasmapheresis, or TRH reported only at case level; evidence inadequate, not routinely recommended (CONFIRM).
Red flags
- Overcorrection: Na increase >8 (high risk >6) mmol/L in first 24h or >18 mmol/L in 48h
- Biphasic deterioration: new encephalopathy, dysarthria, dysphagia, quadriparesis, pseudobulbar pattern, locked-in syndrome, or seizure 2-6 days (1-14 days) after hyponatremia correction
- High risk: Na <=105, chronic/severe hyponatremia (<120), alcoholism, malnutrition, advanced liver disease/transplant, hypokalemia
- Unexpected rapid Na increase (aquaretic) with sudden urine output when hypovolemia or hypokalemia is corrected
- Early MRI may be normal; negative imaging DOES NOT EXCLUDE ODS - diagnosis with clinical + rapid-correction history, repeat MRI in 1-2 weeks
Disposition and follow-up
- Red zone
- Excessive correction detected (>8/24h; >6 in high risk) or early/established ODS findings: immediately evaluate relowering (DDAVP + D5W), stop active Na elevation, monitor Na very frequently (every 1-2 hours), ICU + neurology and nephrology consultation.
- Yellow zone
- High-risk patient under active Na correction who has not yet reached nervous system threshold: monitor Na every 2-4 hours, track urine output/osmolality, have DDAVP ready or consider clamp, monitored ward.
- Green zone
- Acute phase passed, stable established ODS: inpatient/outpatient rehabilitation, dysphagia-aspiration management, VTE prophylaxis, outpatient neurology follow-up; recovery may take months-years, do not assume prognosis overly pessimistic early on.
Clinical notes and doses
Adult
Desmopressin (DDAVP)
FIXED dose: 2-4 mcg IV/SC, every 6-8 hours. Not multiplied by weight, no mcg/kg calculation.
In both reactive relowering and proactive 'clamp' strategies, adult dose is fixed; if calculated based on kg, would result in a lethal dose of 140-280 mcg for 70 kg. Unit is mcg (not mg): 2-4 mcg = 0.002-0.004 mg. Response monitored via urine output and urine osmolality; stop when Na reaches ~125 mEq/L. Source: CJASN 2020 Therapeutic Relowering (PMC7015090); Sterns RH, CJASN 2023 (PMC10843202).
Potassium replacement - latent Na load
Each mEq given raises K up to the effective Na: KAT the K replacement to the total 24-hour correction calculation.
Hypokalemia is both an independent risk factor for ODS and a hidden source of overcorrection: K enters the intracellular compartment while Na exits, so K replacement raises serum Na like hypertonic saline. The classic trap: only the saline velocity is calculated and K is ignored and the 24-hour limit is exceeded without realizing it. The unit is mEq (NOT mg). In addition, when hypovolemia/hypokalemia resolves, ADH is suppressed and Na jumps with unexpected rapid aquarasia. Source: Sterns RH, CJASN 2023 (PMC10843202); ERBP/ESICM/ESE 2014.
Classic traps - ODS
Early MRI may be NORMAL and does NOT EXCLUDE ODS; diagnosis is made with clinical signs + history of rapid correction.
Three classic traps: (1) Imaging trap - MRI findings may be delayed by days to weeks; ODS CANNOT BE RULED OUT with an early negative MRI, repeat in 1-2 weeks. (2) Biphasic course trap - when hyponatremia corrects, the patient initially IMPROVES; new encephalopathy/dysarthria/dysphagia/quadriparesis 2-6 days (1-14 days) later is ODS, do not discharge thinking 'they got better'. (3) Prognosis trap - despite a locked-in state, a significant portion of patients partially or fully recover over months to years; do not give an overly pessimistic prognosis early that would justify withdrawal of care. There is NO proven pharmacological treatment for established ODS; steroids/IVIG/plasmapheresis are not routinely recommended. Source: PMC6804268; BMC Neurol. 2024 (doi:10.1186/s12883-024-03934-3); Sterns RH, CJASN 2023.
Pediatric
Pediatric relowering - dose not confirmed
No verified source was found for pediatric DDAVP relowering dose: calculator INTENTIONALLY not included.
The same principles apply in children (stop overcorrection, maintain safe limit), but the verified DDAVP dose for pediatric relowering is NOT PROVIDED in this protocol calculator - dose should be determined in consultation with pediatric nephrology/intensive care. WARNING: the 0.3 mcg/kg DDAVP dose in labels applies to hemophilia/von Willebrand and procedural bleeding indications; it is NOT USED in hyponatremia relowering, do not confuse these two indications. D5W 3 mL/kg/hour is weight-based and the above rate calculator gives weight multiplication, but pediatric sodium deficits and fluid load must be individualized. Source: left intentionally blank due to dose uncertainty; CJASN 2020 only provides adult data.
Related calculators
Featured external sources
- CJASN hyponatremia treatment guidance 2023.Correction limits and high risk.
- Therapeutic relowering evidence review.Relowering with DDAVP and D5W.
Detailed assessment
Clinical framework
Overview
Osmotic demyelination syndrome (ODS) is an iatrogenic demyelinating disease that occurs due to the rapid correction of a brain that has osmotically adapted to chronic hyponatremia. Lesions are classically located in the central pons (central pontine myelinolysis) and appear on magnetic resonance imaging (MRI) T2/FLAIR sequences as symmetric triangular or trident-shaped hyperintensities in the middle of the pons; in this typical appearance, the ventrolateral pons and the pontine part of the corticospinal tracts are not affected.
However, in some cases, extrapontine areas such as the basal ganglia, thalamus, cerebellum, and subcortical white matter are also involved; in fact, in some patients, extrapontine lesions can appear before pontine lesions. The mechanism is as follows: in chronic hyponatremia, brain cells expel idiogenic osmoles (organic osmolytes) to relatively dilute the intracellular environment and limit brain edema; when sodium is rapidly increased, cells cannot rapidly reclaim these osmoles, and an abrupt osmotic gradient forms between the serum and intracellular space, leading to oligodendrocyte apoptosis and myelin loss.
The most common trigger is the overly rapid correction of hyponatremia; rarely, other rapid osmotic changes such as hypernatremia or hyperglycemia can produce a similar picture. The typical course is biphasic: as hyponatremia corrects, the patient initially shows clinical improvement, then within 2-6 days (reported range 1-14 days), secondary neurological deterioration begins, including lethargy, confusion, dysarthria, dysphagia, quadriparesis, pseudobulbar state, and in severe cases, progressing to locked-in syndrome.
Diffusion-weighted imaging (DWI) may show changes even 24 hours after symptom onset, whereas conventional T2/FLAIR findings may remain normal up to 10 days after symptom onset; therefore, an early negative MRI does not rule out the diagnosis. The incidence of ODS after liver transplantation is significantly higher compared to the general population and has been reported between 0.8% and 1.4%; major risk factors include severe pre-transplant hyponatremia (Na <125 mEq/L), the magnitude of Na change before and after transplantation, high positive intraoperative fluid balance, and postoperative hemorrhagic complications; in patients who develop ODS after transplantation, the rate of death or permanent disability is significantly higher compared to non-transplanted ODS patients (reported series approximately 77% and 45%, respectively).
Historically, the general incidence of ODS has been reported in 0.4-0.56% of neurological service hospitalizations and approximately 0.06% of all internal medicine-neurology hospitalizations; as MRI usage became widespread, the incidence reported increased over time due to detection of oligosymptomatic and asymptomatic cases. In a nationwide study conducted in Sweden, the incidence of ODS increased from 0.271 per million person-years in the 1997-2001 period to 0.945 per million person-years in the 2007-2011 period; this increase is largely attributed to improved diagnostic awareness and access to MRI.
In contrast, in a recent multicenter cohort in Toronto examining 22,858 hyponatremic hospitalizations across five academic hospitals, ODS was found to be extremely rare despite significant numbers of patients experiencing overcorrection of >8 mmol/L within 24 hours (reported rate below one per thousand); this finding suggests that overcorrection is necessary but not sufficient alone for ODS, and additional risk factors (severe baseline hyponatremia, alcoholism, malnutrition, liver disease, hypokalemia) generally need to be present simultaneously.
Differential diagnosis and assessment
Key considerations
- Wernicke encephalopathy is related to thiamine deficiency and can coexist with ODS in alcoholic or malnourished patients; the triad of ophthalmoplegia-ataxia-confusion and rapid response to thiamine treatment are distinctive.
- Ischemic stroke or basilar artery thrombosis can cause acute-onset pontine syndrome and a locked-in state; however, basilar perforator artery infarcts usually involve the surface of the pons and are asymmetric, whereas the classic trident lesion of ODS is symmetric and centrally located.
- Other demyelinating diseases such as acute disseminated encephalomyelitis (ADEM) and multiple sclerosis can involve the pons, but these lesions are rarely symmetric and typically also involve the calloseptal interface; in ODS, a recent history of rapid Na correction is present.
- The most important differential diagnosis of extrapontine ODS with basal ganglia or cortical involvement is metabolic diseases; Wilson's disease and mitochondrial disorders involve the basal ganglia, but pontine involvement is much less common than in ODS.
- Hypoglycemia, hepatic encephalopathy, and metabolic encephalopathy due to sepsis can produce a similar confusion-lethargy picture; they are differentiated by blood glucose, ammonia, and infection markers.
- Posterior reversible encephalopathy syndrome (PRES) usually develops in the context of acute hypertension or immunosuppressive drug use and typically shows posterior parieto-occipital vasogenic edema on imaging; it does not overlap with the central pontine location of ODS, although there may be clinical overlap.
- Substance/drug toxicity (e.g., lithium, methotrexate, or opioid overdose) can cause similar encephalopathy and dysarthria; it is differentiated by toxicology history and screening tests.
- Ongoing, unresolved hyponatremic encephalopathy may be confused with the second deterioration phase in the biphasic course of ODS; the distinguishing point is that ODS occurs following delayed deterioration AFTER Na correction, while hyponatremic encephalopathy continues while Na is low.
- COVID-19-related late-onset central pontine myelinolysis constitutes a separate triggering group in reported cases, in which infection-related hyponatremia and cytokine storm may contribute, and it is recognized with the same clinical-imaging criteria.
Clinical decision reasoning
Why it matters
The most effective treatment for ODS is PREVENTION: because the brain osmotically adapts in chronic hyponatremia, Na increase is limited to <8 mmol/L in the first 24 hours (<6 in high-risk cases) and <18 mmol/L in 48 hours; the targeted daily increase is 4-8 mmol/L (Sterns RH, CJASN 2023; ERBP/ESICM/ESE Guideline, Eur J Endocrinol 2014; Verbalis Expert Panel, Am J Med 2013). These three sources are largely consistent with each other, and these thresholds are currently the most widely accepted.
Even in severe hyponatremia (seizure, coma, herniation risk), the initial goal is to rapidly reduce brain edema by raising Na approximately 4-6 mmol/L within a few hours, then to limit the rate according to the daily limits above; this approach is common to all guidelines and is referred to as the 'fast start, then stop' principle. If overcorrection is detected or anticipated, early relowering - desmopressin 2-4 mcg IV/SC every 6-8 hours and 5% dextrose (D5W) 3 mL/kg/hour (reduces Na by approximately 1 mmol/L/hour) - is used to safely bring Na back within safe limits; this strategy has been found effective in reversing overcorrection and preventing ODS in retrospective series and case reports, but there is NO randomized controlled trial, and the evidence level is low (the CJASN 2020 review explicitly points out this gap).
The benefit of relowering in established ODS is uncertain; at this stage, treatment is primarily supportive care and rehabilitation, because demyelination has already occurred and there is no evidence that lowering Na reverses the progression. Immunomodulatory treatments such as steroids, intravenous immunoglobulin (IVIG), and plasmapheresis have only been tried in isolated case reports in established ODS, and evidence for routine use is insufficient; routinely recommending these treatments would mean applying an old assumption (that ODS is an inflammatory process) to practice without sufficient evidence and SHOULD BE AVOIDED.
Diagnosis is primarily based on clinical assessment + a history of rapid correction; in MRI, symmetric trident-shaped hyperintensity in the central pons on T2/FLAIR sequences, hypointensity on T1, and early-stage diffusion restriction are expected, but conventional MRI findings can remain normal up to 10 days from symptom onset, so early normal imaging does NOT RULE OUT the diagnosis, and imaging should be repeated 1-2 weeks later in clinically suspicious cases. Regarding prognosis, the old teaching (that ODS is universally devastating and usually fatal) is no longer valid; in recent series, most patients survive, and a significant portion show substantial improvement, so giving an excessively pessimistic prognosis early and limiting care is a mistake.
There is a real point of debate among guidelines: while the Sterns and ERBP/ESICM/ESE school advocates strict maintenance of the classic <8/24h-<18/48h limit, the multicenter cohort covering 22,858 hyponatremic admissions in Toronto (NEJM Evidence 2024) and the subsequent systematic review/meta-analysis (Kidney Medicine 2024) showed that although overcorrection occurred in a significant number of patients, ODS remained extremely rare (overall incidence in the meta-analysis 0.23%; 0.73% in rapidly corrected cases, 0.10% in non-rapidly corrected cases). This finding has generated a nephrology perspective that the classic thresholds may be overly cautious and that overcorrection alone does not pose a significant risk for ODS but does so when combined with additional risk factors (severe baseline hyponatremia, alcoholism, malnutrition, liver disease, hypokalemia); however, none of the current international guidelines have yet relaxed the thresholds, and the conservative <8/<6-<18 limits continue to be the clinical standard—therefore, the thresholds in the protocol have been maintained unchanged, but it is important for the clinician to understand that these thresholds are not absolute but a cautious safety margin.
Complications and safety
Critical complications
- Aspiration pneumonia develops due to dysphagia and bulbar muscle weakness; it is prevented by early swallowing assessment and, if necessary, switching to a nutritional route via nasogastric or percutaneous endoscopic gastrostomy (PEG).
- Deep vein thrombosis and pulmonary embolism develop due to prolonged immobilization; they are reduced with pharmacological VTE prophylaxis and early mobilization.
- Seizures can be seen with cortex or subcortical white matter involvement; they are confirmed with electroencephalography and managed with standard antiepileptic treatment.
- Locked-in syndrome develops with severe bilateral corticospinal and corticobulbar tract involvement; the patient is conscious but may have lost almost all motor function, so communication should be established through eye movements and early decisions to discontinue treatment should not be made because meaningful motor recovery has been reported in some cases.
- Iatrogenic hypernatremia or excessively rapid relowering can develop during DDAVP and D5W treatment if Na is lowered beyond the target; therefore, Na should also be monitored every 1-2 hours during relowering, and when Na reaches 125 mEq/L, both agents should be discontinued.
- Nutritional deficiency and muscle atrophy develop during prolonged intensive care and rehabilitation; they are limited with early enteral nutrition support and physical therapy.
- Pressure sores (decubitus ulcers) develop due to prolonged immobility; they are prevented with regular position changes and skin care.
- Secondary infections (urinary system infection, hospital-acquired pneumonia) develop due to prolonged hospital stay and use of invasive devices (urinary catheter, mechanical ventilation); they are reduced with routine infection control measures.
Special patient populations
Special considerations
- ODS in pregnancy has been reported most frequently in the context of hyperemesis gravidarum when severe hyponatremia and hypokalemia are corrected rapidly together; in one case, Na increased by 17 mmol/L in the first 24 hours and extrapontine ODS developed, therefore electrolyte replacement in pregnant patients with hyperemesis should also be done with the same limits (<8 mmol/L in 24 hours).
- In the postpartum period, an unexpected rapid Na increase due to sudden free water loss along with temporary diabetes insipidus has been reported; in postpartum patients, sudden increases in urine output should be closely monitored.
- In liver transplant patients, the incidence of ODS is significantly higher compared to the general population (0.8-1.4%), and major risk factors include pre-transplant Na <125 mEq/L, large perioperative Na change, high positive fluid balance, and postoperative bleeding; in these patients, Na should be monitored very frequently intraoperatively and postoperatively, and a Na management plan should be prepared with the transplant team in advance.
- In patients with advanced liver disease (not transplanted), chronic hyponatremia is the most common electrolyte disorder, and these patients are included in the standard high-risk group due to low protein synthesis and more fragile osmotic balance; Na correction should be done with the strictest limits (<6 mmol/L/24 hours).
- In patients with alcohol use disorder and malnutrition, both the risk of ODS and Wernicke's encephalopathy coexist; these patients should be given thiamine before or simultaneously with Na correction, otherwise the two conditions may be confused with each other.
- Separate validated Na threshold values or DDAVP relowering doses for pediatric ODS could not be found in the literature; the recovery mechanism in children is considered similar to adults, but dose and threshold decisions should be individualized together with pediatric nephrology/intensive care. In reported pediatric case series, unlike adults, isolated extrapontine involvement is seen more frequently than in adults; of the compiled 106 pediatric cases, approximately one-quarter had only extrapontine involvement, almost half had only pontine involvement, and the remaining quarter had involvement in both areas, so seeing only basal ganglia/thalamus involvement without pontine lesions in a child should not exclude the diagnosis of ODS.
- In patients with hypokalemia (regardless of the source group), it should be remembered that potassium replacement will also increase serum Na; in these patients, when calculating the total correction rate, the amount of K replacement must be included in the Na calculation, otherwise the 24-hour limit may be exceeded unknowingly.
Follow-up and reassessment
Follow-up plan
During the active correction period, monitor Na in severe cases every 1-2 hours, then every 4-6 hours; follow urine output and urine osmolality, because sudden aquaresis (unexpected increase in urine output) indicates overcorrection. If the target is exceeded or anticipated to be exceeded, reverse Na with DDAVP and D5W; when Na reaches 125 mEq/L, stop both agents and verify by measuring Na every 1-2 hours that relowering has not been excessive.
If ODS develops clinically or is suspected, the initial MRI may be normal, so rather than excluding the diagnosis, repeat imaging in 1-2 weeks; meanwhile, start early management of aspiration/dysphagia, VTE prophylaxis, seizure monitoring, nutritional support, and physical-speech-swallow rehabilitation. Before discharge, ensure safe swallowing evaluation, adequate respiratory and airway protection, and that a rehabilitation plan and outpatient neurology-nephrology appointments are arranged.
Clearly explain to the patient and relatives that in case of newly developed speech disturbance, swallowing difficulty, weakness, or changes in consciousness, they should return to the emergency department. In current series, the majority of patients survive, and a significant portion show marked or complete recovery over months to years; therefore, it is important to avoid giving an excessively pessimistic prognosis early, but patients with poor prognostic indicators such as Na <120 mEq/L, concurrent hypokalemia, and low Glasgow Coma Scale should have closer monitoring planned.