Methaemoglobinaemia

Meet the patient

A 62-year-old man turns slate-grey ten minutes after a benzocaine throat spray for upper gastrointestinal endoscopy. He is breathless and confused. The nurse puts him on 100 percent oxygen and his saturation, stuck at 84 percent, does not move. The arterial blood gas shows a PaO2 of 90 mmHg and a calculated saturation of 98 percent.^[1][3]

The monitor looks reassuring and lies. This is methaemoglobinaemia until co-oximetry proves otherwise. Two questions run the resus bay: why is the cyanosis not responding to oxygen? and is it safe to give methylene blue? Everything below answers them — and the second question turns on a single enzyme assay, G6PD, because in its absence the antidote causes haemolysis.^[1][4]

The signature triad — and the monitor that lies

Methaemoglobinaemia tests three habits at once: your reflex to trust the monitor, your grasp of one elegant antidote, and your awareness of the one contraindication that turns the antidote into a poison. The pathology is deceptively simple — the haem iron is oxidised from ferrous Fe2+ to ferric Fe3+, and ferric iron cannot bind oxygen. Worse, by stabilising the relaxed conformation of the neighbouring subunits it shifts the oxygen-haemoglobin dissociation curve to the left, so the oxygen that remains bound is held more tightly and released less readily to the tissues.^[2]

A small amount of methaemoglobin forms constantly by spontaneous auto-oxidation, and an efficient red-cell reducing system holds the steady state under 1 percent; disease is defined as a level over 1.5 percent. The bedside syndrome to recognise is sudden cyanosis after a procedure, a drug, or an exposure that does not improve with oxygen, with chocolate-brown blood and a saturation gap — the pulse oximeter near 85 percent while the arterial PaO2 is normal.^[1]

Acquired versus congenital — and severity by percentage

Enlarge

Classify on mechanism first, then on severity, because together they set the treatment. Acquired oxidant poisoning is over 99 percent of cases and reverses with methylene blue; congenital disease is rare, lifelong, and largely managed without the antidote.^[1]

^[1]

The second axis is severity by percentage, and it sets who gets the antidote. Mild disease, under 10 percent, is often asymptomatic and managed by withdrawing the oxidant and giving oxygen. Moderate disease brings slate-grey cyanosis, exertional dyspnoea, headache and fatigue. Severe disease adds dyspnoea at rest, syncope, confusion, chest pain and ischaemic ECG changes; very high levels bring seizures, coma, lactic acidosis, arrhythmia and shock — occupational data put concentrations around 80 percent in the life-threatening range. The threshold to treat is symptomatic disease or a level over 30 percent.^[1][5][13]

How common, and who is susceptible

Acquired methaemoglobinaemia is uncommon but under-recognised, because the monitor looks reassuring. The largest modern series — Ash-Bernal, 138 cases over two academic centres in 28 months — found dapsone the most common cause at 42 percent of cases, with five of the most severely affected patients having used a topical benzocaine spray; one fatality and three near-fatalities were directly attributable, and almost all patients — 94 percent — were anaemic.^[4]

^{[1] [1]}

The susceptibility factors are the ones examiners probe. Infants are the classic susceptible group: they drink more water per kilogram of body weight, have lower NADH-cytochrome b5 reductase activity, and carry a higher proportion of fetal haemoglobin, which oxidises to methaemoglobin more readily — and nitrate-contaminated well water used for formula is the best-studied exposure behind infantile blue baby syndrome. G6PD deficiency impairs NADPH-dependent reduction and raises the risk of oxidative haemolysis, and it makes methylene blue ineffective. Anaemia and cardiopulmonary disease lower the reserve for a given percentage — in the Ash-Bernal series almost all of the patients, 94 percent, were anaemic — and an anaemic patient may be profoundly hypoxic without visible cyanosis because cyanosis depends on the absolute amount of reduced haemoglobin rather than the percentage.^[1][5][4]

The oxidants — OXIDANTS

The unifying theme is the oxidant, and the sources cluster into one mnemonic. Name the oxidant from the history and you have named the cause.^[1]

^[1]

The detail worth carrying into the viva: dapsone was the single most common cause in the Ash-Bernal series — 42 percent of all 138 acquired cases — and methaemoglobinaemia can rebound or persist after an apparently successful dose, because dapsone and its metabolites keep generating oxidant load; cimetidine is used to block the cytochrome P450 generation of the toxic hydroxylamine metabolite. Benzocaine is the classic iatrogenic cause: in that same series, five of the patients with the most severely elevated levels followed a topical 20 percent benzocaine spray, with mean peak levels around 43.8 percent, and benzocaine given during transoesophageal echo dominates hospital case clusters; the FDA warns against benzocaine in children under two years after severe infant cases. Smoke inhalation can combine carbon monoxide, cyanide and methaemoglobinaemia at once — which changes both monitoring and antidote choice.^[4][6][8]

The redox balance — and the methylene-blue hinge

Enlarge

The whole syndrome flows from one number: the steady-state balance between haemoglobin auto-oxidation and the red-cell reducing systems. Two enzymatic routes hold methaemoglobin under 1 percent, and the antidote works through the normally dormant one.^[1]

The major pathway — NADH-cytochrome b5 reductase, also called diaphorase I — performs about 95 percent of daily methaemoglobin reduction. NADH from glycolysis reduces cytochrome b5, which reduces ferric methaemoglobin back to functional ferrous haemoglobin. This is the pathway genetically deficient in congenital cytochrome b5 reductase deficiency. The minor pathway — NADPH-methaemoglobin reductase, diaphorase II — uses NADPH from the hexose-monophosphate shunt but is normally dormant because it lacks an endogenous electron carrier.^[1]

Methylene blue supplies that missing carrier, and that single fact is the whole of antidote therapy. It is reduced to leukomethylene blue by NADPH-methaemoglobin reductase, and leukomethylene blue non-enzymatically reduces ferric methaemoglobin back to ferrous haemoglobin. This is the molecular basis of methylene-blue therapy — and the reason it is completely dependent on G6PD: no G6PD means no NADPH, no reduction, and methylene blue itself, an oxidant dye, then causes haemolysis. The non-enzymatic routes — ascorbic acid and reduced glutathione — are small and slow, but they are the rationale for ascorbic acid as the alternative antidote in G6PD deficiency and congenital disease.^[2]

Why the ferric iron is doubly harmful

Ferric iron cannot bind oxygen, so each methaemoglobin subunit is a lost carrying site — a functional anaemia. Worse, by stabilising the relaxed state of the neighbouring ferrous subunits it shifts the dissociation curve to the left, so the oxygen that is bound is held more tightly and released less readily to the tissues. The patient suffers both a loss of carrying capacity and impaired unloading — the molecular basis of the tissue hypoxia, lactic acidosis and end-organ injury at high levels.^[1]

The two monitor paradoxes

The pulse-oximetry paradox. A standard pulse oximeter uses two wavelengths calibrated only for oxy- and deoxyhaemoglobin. Methaemoglobin absorbs strongly at both, so as it rises the saturation is driven toward about 85 percent and becomes insensitive to the true arterial oxygenation — giving the saturation gap. Giving more oxygen does not move the number.^[3]

The arterial-blood-gas paradox. A standard blood-gas machine does not measure saturation — it calculates it from PaO2, pH and temperature on the assumption that all haemoglobin is normal, so it reports a normal calculated saturation despite the true saturation being low. Only co-oximetry, with four or more wavelengths that directly measure oxy-, deoxy-, carboxy- and methaemoglobin, reveals the truth. The lung is intact, so PaO2 is normal.^[2]

Why methylene blue can fail or harm

^{[1] [1]}

One more trap lives in the drug chart, not the blood. Methylene blue carries monoamine-oxidase-A-inhibiting properties, and serotonergic toxicity is documented when it is given to patients on serotonin reuptake inhibitors — at doses as low as well under a full antidotal dose. Check the serotonergic history before the syringe.^[1][14]

The mimics — and the discriminator that ends each

The unifying problem is cyanosis that does not respond to oxygen, and the differential splits into the dyshaemoglobins and everything else. The discriminator is the blood colour and the co-oximetry.^[1][6]

^{[1] [1]}

Two distinctions deserve emphasis. Methaemoglobin drives the pulse oximeter toward 85 percent; carboxyhaemoglobin is misread as oxyhaemoglobin so the oximeter reads falsely high near 100 percent — a clean discriminator. Methaemoglobin and sulphaemoglobin are both measured by co-oximetry and produced by the same oxidants, but sulphaemoglobin is irreducible and lasts the red-cell lifespan. Distinguishing acquired from congenital at the bedside: congenital means lifelong cyanosis from infancy, a family history, a methaemoglobin of only 10 to 30 percent with chronic compensation, and no precipitant.^[2][5]

The history is the investigation

Build the assessment around the cardinal triad and a hunt for the precipitant. Establish the exposure — prescription drugs (dapsone, sulphonamides, primaquine, phenazopyridine, nitroprusside, nitroglycerin, benzocaine or prilocaine), occupational or recreational exposure (aniline dyes, nitrites or poppers, nitrobenzene, chlorates, naphthalene), recent procedures (endoscopy, bronchoscopy, dental work, intubation, transoesophageal echo with benzocaine), the water source in an infant, and smoke inhalation. Note the timing of onset relative to the exposure, the symptoms, the comorbidity — anaemia, cardiac, respiratory, G6PD status, pregnancy — and a family history for suspected congenital disease.^[1]

Three bedside manoeuvres earn marks. Draw blood into a heparinised tube and compare it with a control: methaemoglobin blood is chocolate-brown and does not turn bright red on shaking in air — the filter-paper drop that stays brown is a classic. Apply 100 percent oxygen: the cyanosis of methaemoglobin does not improve, unlike true hypoxaemic cyanosis. Recognise the saturation gap — a pulse oximeter near 85 percent with a normal PaO2 and a normal calculated saturation. In the unstable patient run ABCDE: intubate if the GCS is under 8, give high-flow oxygen, establish IV access, and monitor continuously.^[1][3]

^[1]

Co-oximetry — the only test that measures it directly

Enlarge

The gold standard is co-oximetry on arterial or venous blood, reporting methaemoglobin as a percentage of total haemoglobin; over 1.5 percent is abnormal. A venous sample is acceptable and avoids arterial puncture. The first-line panel adds a venous or arterial blood gas to confirm the metabolic picture — severe methaemoglobin causes a lactic acidosis with low bicarbonate, while PaO2 stays normal — a full blood count, reticulocytes and film for anaemia and Heinz bodies or bite cells of oxidative haemolysis, a haemolysis screen of LDH, haptoglobin and bilirubin, electrolytes, lactate, glucose, troponin and an ECG for end-organ injury, and a drug or oxidant screen.^[2]

Check G6PD ideally before methylene blue, but never delay treatment in a severely symptomatic patient. The caveat is that G6PD levels are falsely elevated during acute haemolysis because the most deficient cells have already lysed — recheck two to three months later. When the picture is congenital — lifelong cyanosis, family history, a methaemoglobin of 10 to 30 percent in an otherwise well patient with no precipitant — send a cytochrome b5 reductase enzyme assay and haemoglobin electrophoresis or globin gene sequencing for haemoglobin M.^[1][5]

Some older co-oximeters cannot distinguish sulphaemoglobin from methaemoglobin, very high methaemoglobin can saturate the assay and report falsely low, and frozen or mishandled samples artefactually raise the level. Interpret the percentage against the total haemoglobin: an anaemic patient may be profoundly hypoxic without cyanosis because the absolute methaemoglobin in g/dL is what determines the colour.^[3]

Resuscitation — stop the oxidant, give oxygen, then the antidote

Enlarge

^{[1] [7]}

Resuscitation and the first specific measures are simultaneous: stop the oxidant and give oxygen while you prepare the antidote. Secure the airway and intubate if the GCS is under 8 or deterioration is rapid, give high-flow 100 percent oxygen through a non-rebreather mask to maximise the oxygen carried by the remaining functional haemoglobin and the dissolved oxygen, establish IV access, monitor continuously for the arrhythmia of high-level disease, give isotonic crystalloid for hypotension, and check bedside glucose. Stop and remove the oxidant — discontinue dapsone, withhold benzocaine, remove the patient from the chemical source, change the infant's water source. Confirm with co-oximetry and check G6PD where practical, but never delay methylene blue in a severely symptomatic patient.^[1][3]

^{[1] [4]}

Methylene blue — the first-line antidote

Know the dose, the mechanism, and the contraindications verbatim. The initial dose is 1 to 2 mg/kg of a 1 percent solution IV, administered over 5 minutes; reduction of methaemoglobin is usually complete within about an hour — cyanosis visibly fades. If methaemoglobinaemia persists, a second dose may be given, not exceeding a total of 5 to 7 mg/kg, because larger amounts of this oxidant dye can themselves generate methaemoglobin and cause haemolysis.^[1][7]

^{[1] [1]}

When the antidote cannot be used — the alternatives

For G6PD deficiency, refractory disease, or contraindication, the ladder steps to four alternatives. Know which to reach for and why.^[1]

^{[1] [1]}

Source-specific measures that change the plan

Three precipitants each need a tailored add-on. Dapsone, well water, and smoke inhalation behave differently and the extra step matters.^[4]

For dapsone-induced methaemoglobinaemia, expect recurrence: the drug and its metabolites keep generating oxidant load after an apparently successful dose, and rebound weeks into treatment is documented — give repeated methylene blue as needed, add cimetidine to inhibit the cytochrome P450 generation of the toxic hydroxylamine metabolite, and monitor the level serially. For infantile well-water nitrate poisoning, stop the exposure — use bottled or tested water for formula, because well water above the nitrate limit is the best-studied cause, and have the water tested before the baby goes home; treat with oxygen, supportive care and methylene blue if symptomatic. For smoke inhalation, where carbon monoxide, cyanide and methaemoglobin coexist, treat carbon monoxide with 100 percent oxygen, cyanide with hydroxocobalamin — avoid inducing methaemoglobinaemia with nitrite antidotes in a hypoxaemic patient — and methaemoglobin with methylene blue, sequencing by the dominant toxin.^{[4][6][8][10]}

Disposition follows severity. Severe disease — coma, seizures, severe acidosis, arrhythmia or haemodynamic instability — needs intensive care, with benzodiazepines for seizures, ACLS protocols for arrhythmia, fluids and vasopressors for shock, and transfusion for anaemia or haemolysis; remember that in the Ash-Bernal series almost all patients were anaemic and the one death followed massive benzocaine exposure. An asymptomatic patient whose oxidant has been stopped is observed with serial levels and discharged when the level is falling and symptoms have gone. Congenital forms are managed as an outpatient with oral ascorbic acid — with or without riboflavin in case reports — reassurance that chronic cyanosis is well tolerated, and genetic counselling.^{[1][4][5][11]}

The preventable harm — pitfalls that recur

The recurring failures trace to a short list, and most are preventable. Trusting the monitor and missing the saturation gap delays co-oximetry — the cyanosed patient with a reassuring pulse oximetry and a normal calculated saturation is the classic miss. Giving methylene blue without checking G6PD in a susceptible patient causes oxidative haemolysis, the antidote turning into the toxin. Giving methylene blue to a patient on serotonergic drugs precipitates serotonin syndrome.^[1]

The dosing pitfalls matter: cumulative methylene blue beyond the 5 to 7 mg/kg total is itself oxidant — in the presence of haemolysis high doses can paradoxically generate methaemoglobin; using methylene blue for sulphaemoglobinaemia is futile because it is irreducible; and missing the recurrent dapsone case because the cimetidine was forgotten sends the level back up. Finally, treating an anaemic patient on percentage alone misses severe disease — in one series 94 percent of patients were anaemic — so judge the hypoxia, not just the number, and transfuse to restore functional haemoglobin.^[1][4]

Evidence and regional practice

The anchoring references shape modern practice. The Wright 1999 review (Annals of Emergency Medicine) remains the canonical teaching reference for aetiology, pharmacology and management. The Ash-Bernal 2004 series (Medicine) of 138 cases established dapsone and topical or local anaesthetics as the leading drug causes. The Iolascon 2021 recommendations (American Journal of Hematology) give the current diagnostic and treatment framework, including the conservative threshold for mild asymptomatic cases and the G6PD and serotonergic caveats.^[1][4][5]

In the United States the FDA warns against benzocaine products in children under two years of age after severe infant cases, and requires methaemoglobinaemia warnings on benzocaine packaging; the same agency warns that methylene blue can precipitate serotonergic toxicity in patients on serotonin reuptake inhibitors. Methylene blue 1 to 2 mg/kg IV remains first-line for symptomatic acquired disease, with the G6PD check made where practical — but never at the cost of delaying treatment in a severely symptomatic patient.^[1][9]

^[1]

Ward-round test — four stems

Stem 1 — slate-grey ten minutes after benzocaine (answer)ShowHide

A 62-year-old turns slate-grey ten minutes after a benzocaine throat spray for endoscopy. He is breathless and confused; on 100 percent oxygen his saturation stays at 84 percent, the PaO2 is 90 mmHg, and the calculated saturation is 98 percent. What is the diagnosis, the confirmatory test, and the treatment? Model: This is acquired methaemoglobinaemia from benzocaine — the oxygen-refractory cyanosis with a normal PaO2 and a saturation gap is the signature. Confirm with co-oximetry, the only test that measures methaemoglobin directly. Stop the benzocaine, give high-flow 100 percent oxygen, and give methylene blue 1 to 2 mg/kg of the 1 percent solution IV over 5 minutes — reduction is usually complete within an hour and the cyanosis visibly fades; a second dose not exceeding a 5 to 7 mg/kg total can be given if the level remains high. Check G6PD where practical before the dose, but do not delay treatment in a severely symptomatic patient, and check the drug chart for serotonergic agents because methylene blue carries monoamine-oxidase-inhibiting properties.^[1][7]

Stem 2 — cyanosis in a G6PD-deficient patient on dapsone (answer)ShowHide

A 34-year-old on dapsone for dermatitis herpetiformis becomes cyanosed and dyspnoeic over days; the methaemoglobin is 38 percent and his G6PD assay is low. Why is methylene blue contraindicated, and what do you give instead? Model: Methylene blue depends on the NADPH-methaemoglobin reductase shunt, which needs NADPH from G6PD — in G6PD deficiency there is no NADPH, so the antidote fails, and methylene blue itself is an oxidant dye that then causes oxidative haemolysis. Give high-flow oxygen and stop the dapsone; treat with ascorbic acid for slow non-enzymatic reduction — the mainstay when methylene blue cannot be used — add N-acetylcysteine as an antioxidant adjunct, and arrange exchange transfusion for life-threatening levels or severe haemolysis. Expect recurrence: dapsone and its metabolites keep generating oxidant load, so monitor the level serially and repeat treatment as needed.^[1][4]

Stem 3 — the blue baby on well water (answer)ShowHide

A four-month-old rural infant fed formula made with well water is slate-grey, lethargic and feeding poorly; the methaemoglobin is 28 percent. What is the cause, and what three things do you do — and what do you not do? Model: This is infantile well-water nitrate methaemoglobinaemia — infants drink more water per kilogram, have lower cytochrome b5 reductase activity and a higher proportion of fetal haemoglobin that oxidises readily, and gut flora can reduce nitrate from contaminated well water to the more potent nitrite. Stop the exposure: switch to bottled or tested water for the formula — and do NOT boil the well water, because boiling concentrates rather than removes nitrate. Give oxygen and supportive care and treat with methylene blue if symptomatic. Test the well water before the baby goes home, because well water above the nitrate limit is the best-studied cause of infantile methaemoglobinaemia.^[1][10]

Stem 4 — the smoke-inhalation patient who is cyanosed (answer)ShowHide

A fire survivor is cyanosed despite high-flow oxygen; co-oximetry shows methaemoglobin 22 percent and carboxyhaemoglobin 28 percent. How do you sequence the treatment, and what cyanide antidote do you avoid? Model: This is combined carbon monoxide, cyanide and methaemoglobinaemia from combustion. Treat carbon monoxide with high-flow 100 percent oxygen. Treat the methaemoglobin with methylene blue given the symptomatic level. For cyanide give hydroxocobalamin — the first-line antidote that does not itself induce methaemoglobinaemia — and AVOID sodium nitrite, because deliberately generating more methaemoglobin in a patient who cannot oxygenate worsens the picture; thiosulphate is slower and largely reserved for non-smoke exposures. Sequence by the dominant toxin, intubate and ventilate early, and monitor all three dyshaemoglobins on co-oximetry.^[1][6][12]

References14ShowHide

1. [1]Wright RO, Lewander WJ, Woolf AD. Methemoglobinemia: etiology, pharmacology, and clinical management Ann Emerg Med, 1999.PMID 10533013
2. [2]Skold A, Cosco DL, Klein R. Methemoglobinemia: pathogenesis, diagnosis, and management South Med J, 2011.PMID 22024786
3. [3]Cortazzo JA, Lichtman AD. Methemoglobinemia: a review and recommendations for management J Cardiothorac Vasc Anesth, 2014.PMID 23953868
4. [4]Ash-Bernal R, Wise R, Wright SM. Acquired methemoglobinemia: a retrospective series of 138 cases at 2 teaching hospitals Medicine (Baltimore), 2004.PMID 15342970
5. [5]Iolascon A, Bianchi P, Andolfo I, Russo R. Recommendations for diagnosis and treatment of methemoglobinemia Am J Hematol, 2021.PMID 34467556
6. [6]Borron SW, Bebarta VS. Asphyxiants Emerg Med Clin North Am, 2015.PMID 25455664
7. [7]Ramírez Rivera J, Garayúa JE Methemoglobinemia: life-threatening hazard of multiple drug ingestions Vet Hum Toxicol, 2006.PMID 19606800
8. [8]Vallurupalli S, Manchanda S Risk of acquired methemoglobinemia with different topical anesthetics during endoscopic procedures South Med J, 2011.PMID 22915889
9. [9]Walsh KL, Silva CR, Lee PH Occult Methemoglobinemia in a Medically-Complex 7-Year-Old Child and the Opportunity for Pharmacist Intervention in Pediatric Emergency Medicine Pediatr Emerg Care, 2024.PMID 39659856
10. [10]Fossen Johnson S Methemoglobinemia: Infants at risk J Community Hosp Intern Med Perspect, 2019.PMID 30956100
11. [11]Aldeeb M, Khalil IA, Yassin MA Congenital Methemoglobinemia in a 33-Year-Old Patient: A Case Report on a Rare Presentation and a Review of the Literature Am J Case Rep, 2023.PMID 37041911
12. [12]Kiernan EA, Carpenter JE, Dunkley CA, et al. Elevated methemoglobin levels in patients treated with hydroxocobalamin: a case series and in-vitro analysis Clin Toxicol (Phila), 2022.PMID 35549585
13. [13]Bradberry SM Occupational methaemoglobinaemia. Mechanisms of production, features, diagnosis and management including the use of methylene blue QJM, 2003.PMID 14579544
14. [14]Charbonneau A Toxicité sérotoninergique résultant d'une interaction médicamenteuse entre le bleu de méthylène et les inhibiteurs de la recapture de la sérotonine. Ann Pharmacother, 2013.PMID 23950608