Methaemoglobinaemia: When Haem Iron Rusts
A patient turns blue, but the oxygen is flowing and the lungs are clear. You crank the oxygen to maximum and the colour doesn't budge; the pulse oximeter sits frozen near 85% no matter what you do; and when blood is drawn it looks not dark red but chocolate-brown. This is not a lung problem or a heart problem. It is a chemistry problem inside the red cell: an oxidising drug has rusted the iron of haemoglobin, and the molecule that carries oxygen has quietly stopped doing its one job. The fix is not more oxygen — it is a dye that runs the chemistry backwards.
A 3-week-old infant is rushed in blue around the lips and fingertips after a day of poor feeding and diarrhoea. The oxygen saturation on the monitor reads 84% and will not rise, yet the baby is breathing comfortably and the chest is clear. High-flow oxygen through a mask changes nothing — the cyanosis is stubborn, slate-grey, wrong. A blood gas is drawn, and the sample in the syringe is the colour of chocolate. The measured PaO2 comes back completely normal, in open contradiction to the pulse oximeter. A co-oximeter finally names the culprit: methaemoglobin at 38%. Somewhere in this child's short life — a topical anaesthetic, a well-water nitrate, an oxidant drug — the iron in haemoglobin has been rusted from the form that carries oxygen to the form that cannot. The saturating oxygen the baby is breathing has nowhere to bind.
The iron that must stay ferrous
Haemoglobin's whole trick rests on the oxidation state of a single atom. Each of haemoglobin's four haem groups holds one iron atom, and that iron must sit in its ferrous form, Fe2+, to bind and release oxygen. Fe2+ is the working form — it grips an oxygen molecule loosely enough to pick it up in the lungs and let it go in the tissues. But iron is easily oxidised. Strip one more electron and it becomes ferric, Fe3+ — the same reaction that turns iron nails to rust. Ferric haem, called methaemoglobin, physically cannot bind oxygen at all. The body knows this happens spontaneously at a slow rate and keeps a repair crew on constant duty; normal blood carries less than 1–2% metHb at any moment. Toxicity is simply that repair crew being overwhelmed by an oxidant faster than it can keep up.
Two hits, not one: why it mimics hypoxia twice over
Methaemoglobinaemia starves the tissues by two mechanisms at once. The first is obvious: every haem locked in the Fe3+ state is a seat that oxygen can never occupy, so the blood's oxygen-carrying capacity falls — a functional anaemia. But there is a second, subtler injury. Haemoglobin is a cooperative molecule: its four subunits talk to each other, and when some of them are jammed as metHb, the remaining normal Fe2+ subunits are pulled into a high-affinity state. That shifts the oxygen–haemoglobin dissociation curve to the left — the good haem now clings to its oxygen too tightly and releases less of it to the tissues. So the patient loses on both ends: less oxygen is carried, and what little is carried is handed over reluctantly. This is why a metHb level of, say, 30% causes far more distress than simply losing 30% of the blood's carrying capacity would predict.
Think of haemoglobin as a fleet of delivery vans, each carrying oxygen from the lungs to the tissues. Oxidation doesn't damage the fuel or the roads — it welds some of the van doors shut. Those vans (the metHb) still drive around, but they can never pick up a package. Worse, the welding jams the doors of the healthy vans too, so even they open grudgingly and drop off less at each stop. The garage has a mechanic — the enzyme that reduces iron back to Fe2+ — quietly un-welding doors all day. Poisoning is an oxidant welding doors shut faster than one mechanic can free them. Methylene blue is a second mechanic hired on the spot, working an alternate, much faster tool.
The culprits: oxidant drugs and chemicals
Almost every case is acquired — a drug or chemical delivering an oxidant load. The classic offender is dapsone, used in leprosy, dermatitis herpetiformis and Pneumocystis prophylaxis; its slow oxidant metabolites make dapsone-induced methaemoglobinaemia notoriously prolonged, often needing repeated antidote. The local anaesthetics are a favourite exam trap: benzocaine (throat and teething sprays) and prilocaine — including the prilocaine in EMLA cream, which is exactly why EMLA carries dosing limits in infants. Then the nitrates and nitrites: contaminated well water, industrial and recreational "poppers," and topical or inhaled nitrites. Add the sulfonamide antibiotics, aniline dyes and related industrial chemicals, and rasburicase, the recombinant urate oxidase used in tumour lysis prophylaxis, whose reaction generates hydrogen peroxide. Neonates are uniquely vulnerable: they have lower levels of the reducing enzyme, and fetal haemoglobin is more easily oxidised — the reason "blue baby" from well-water nitrates is a paediatric classic.
- Haem iron works only as ferrous Fe2+; oxidation to ferric Fe3+ makes methaemoglobin, which cannot bind oxygen.
- Two hits: reduced carrying capacity (functional anaemia) plus a leftward shift that starves the tissues further.
- It is almost always acquired — an oxidant drug or chemical overwhelming the red cell's reducing enzyme.
- Classic causes: dapsone (slow, prolonged), benzocaine/prilocaine (EMLA), nitrates/nitrites, sulfonamides, aniline dyes, rasburicase.
- Neonates are especially vulnerable — less reducing enzyme and more easily oxidised fetal haemoglobin.
The tell-tale discordance at the bedside
Methaemoglobinaemia announces itself as a set of contradictions. The first clue is cyanosis that will not correct with oxygen. A patient with a lung problem pinks up on supplemental oxygen; here the slate-grey colour is fixed, because the defect is in the blood, not the alveoli. The second clue is the blood itself — drawn dark chocolate-brown rather than dark red, and it stays brown on exposure to air instead of brightening. The third, and most exam-worthy, is the pulse oximeter: it reads a falsely low but strangely fixed value, tending to plateau around 85% and refusing to track the true saturation. This happens because the two-wavelength oximeter is fooled by metHb, which absorbs light at both wavelengths and drags every reading toward roughly 85%. The clincher is the arterial blood gas: the measured PaO2 is normal, because the oxygen dissolved in plasma is unaffected — the lungs are doing their job perfectly. A normal PaO2 with a low, fixed SpO2 and brown blood is methaemoglobinaemia until proven otherwise; the definitive test is co-oximetry, which measures the metHb fraction directly.
Severity tracks the metHb fraction fairly predictably. Below about 15% patients are often only dusky or mildly cyanotic. Around 20–30% brings headache, fatigue, dizziness and breathlessness. Above 50% the picture darkens into a metabolic acidosis, arrhythmias, seizures, coma; levels approaching 70% can be fatal. Two modifiers matter. Anaemic patients decompensate at lower metHb fractions, because they have less good haemoglobin to spare in the first place. And co-existing conditions that already limit oxygen delivery stack on top. The same clinical family — a normal PaO2 with genuine tissue hypoxia — links this chapter to the Cyanide and Carbon monoxide chapters, where the oxygen is present but the cell still cannot use it.
The pulse oximeter is not just unreliable here — it is actively lying, and knowing exactly how it lies is the diagnosis. Because metHb pulls the two-wavelength reading toward ~85%, a patient with severe poisoning and a true saturation far lower can still show 85% on the monitor, while a mildly affected patient shows the same 85%. The number is decoupled from reality and pinned. So never let a "not-that-bad" oximeter reading reassure you in a cyanotic patient who won't respond to oxygen. Trust the co-oximeter and the chocolate-brown blood, not the SpO2.
The antidote: methylene blue runs the chemistry backwards
Management begins with the obvious and ends with an elegant piece of redox biochemistry. First, stop the offending agent and give high-flow oxygen — not because oxygen can reload the poisoned haem, but to maximise what the healthy haem and dissolved plasma oxygen can deliver while you treat. The specific antidote is methylene blue, indicated for symptomatic patients or a metHb level roughly above 20–30%. Its mechanism is a beautiful shortcut. Normally the red cell's minor NADPH-dependent methaemoglobin reductase does little; methylene blue supercharges it. NADPH-methaemoglobin reductase reduces methylene blue to leucomethylene blue, and leucomethylene blue is a potent electron donor that reduces the ferric Fe3+ of metHb back to ferrous Fe2+ — regenerating functional haemoglobin. In effect the antidote borrows the reducing power of the pentose-phosphate pathway and channels it, through the dye, onto the rusted iron. Improvement is often visibly rapid. For the slow, self-renewing oxidant load of dapsone, expect the metHb to rebound and repeat dosing to be needed.
The critical caveat: G6PD deficiency
Here is the link that examiners love, because it turns the antidote against the patient. Methylene blue only works if the red cell can reduce it to leucomethylene blue, and that reduction is powered by NADPH — which comes almost entirely from the pentose-phosphate pathway, whose gatekeeper enzyme is glucose-6-phosphate dehydrogenase (G6PD). In G6PD deficiency there is little NADPH, so methylene blue is not reduced and simply doesn't work as an antidote. Worse, methylene blue is itself an oxidant, so in a G6PD-deficient patient it can precipitate acute haemolysis — the very cells you are trying to rescue burst instead. This is one continuous thread with the Haematology section: haem iron oxidation states, the pentose-phosphate pathway, NADPH, and G6PD deficiency are the same biochemistry seen from the toxicology side. So screen for or suspect G6PD deficiency, especially in males of African, Mediterranean, Middle Eastern or Southeast Asian descent, and if methylene blue is contraindicated or fails, reach for alternatives: high-dose ascorbic acid (vitamin C), and for severe or refractory cases, exchange transfusion (and, in the right setting, hyperbaric oxygen). The same logic explains a prescribing rule from the oncology wards — rasburicase is contraindicated in G6PD deficiency, because its hydrogen-peroxide by-product both drives methaemoglobinaemia and triggers haemolysis in exactly these patients.
Methylene blue — first-line for significant or symptomatic methaemoglobinaemia; reduced by NADPH to leucomethylene blue, which reduces Fe3+ back to Fe2+. Ascorbic acid (vitamin C) — a slower chemical reducer, useful when methylene blue is contraindicated (G6PD deficiency) or unavailable. Exchange transfusion — physically removes methaemoglobin and the offending agent; reserved for severe, refractory cases and neonates. And a mirror-image use worth remembering: in cyanide poisoning, nitrites are given deliberately to induce a controlled methaemoglobinaemia — the Fe3+ of metHb scavenges cyanide as cyanmethaemoglobin, pulling it off the mitochondrial enzyme, a trick explored fully in the Cyanide chapter.
- First steps: stop the oxidant, give high-flow oxygen, and measure metHb by co-oximetry.
- Methylene blue is first-line: NADPH reduces it to leucomethylene blue, which reduces Fe3+ back to Fe2+.
- In G6PD deficiency methylene blue is ineffective (no NADPH) and can itself cause haemolysis.
- Alternatives when methylene blue fails or is contraindicated: ascorbic acid; exchange transfusion for severe cases.
- Dapsone causes prolonged, rebounding methaemoglobinaemia — anticipate repeat antidote dosing.
- Rasburicase is contraindicated in G6PD deficiency for the same redox reason.
- Chasing a low pulse-oximeter reading with ever more oxygen. Oxygen cannot reload ferric haem; the SpO2 is fixed near 85% and lying. Measure metHb and give the antidote.
- Giving methylene blue reflexively without considering G6PD deficiency — where it doesn't work and can trigger haemolysis, turning the antidote into a second toxin.
- Being falsely reassured by a normal PaO2. The dissolved oxygen is normal, but the haemoglobin cannot carry it — a normal PaO2 with cyanosis and brown blood is the diagnosis, not the exclusion.
A cyanotic patient on dapsone has an SpO2 fixed at 85% despite high-flow oxygen; the blood is chocolate-brown and the PaO2 is normal. Co-oximetry confirms a metHb of 34%. He is known to be G6PD-deficient. What is the most appropriate management?
- Oxidant drugs/toxins convert haem iron from ferrous Fe2+ to ferric Fe3+ (methaemoglobin), which cannot carry oxygen and shifts the remaining normal haem's curve leftward — a functional anaemia with tissue hypoxia.
- Classic causes: dapsone (slow/prolonged), benzocaine/prilocaine (EMLA), nitrates/nitrites, sulfonamides, aniline dyes, rasburicase; neonates are especially vulnerable.
- The signature is discordance: cyanosis unresponsive to oxygen, chocolate-brown blood, a normal PaO2, and an SpO2 falsely stuck near 85% — co-oximetry is diagnostic.
- Antidote is methylene blue (NADPH → leucomethylene blue → reduces Fe3+ to Fe2+) — but in G6PD deficiency it is ineffective and can cause haemolysis; use ascorbic acid or exchange transfusion instead, and rasburicase is contraindicated in G6PD deficiency.
- Goldfrank's Toxicologic Emergencies — Methemoglobin inducers and methylene blue.
- Rang & Dale's Pharmacology — Haemoglobin, oxygen carriage and drug-induced methaemoglobinaemia.
- Katzung Basic & Clinical Pharmacology — Antidotes: methylene blue; oxidant haemolysis and G6PD deficiency.
- British National Formulary (BNF) — Methylthioninium chloride (methylene blue); dapsone; rasburicase (G6PD contraindication).
- UpToDate / TOXBASE — Methemoglobinemia: clinical features, diagnosis and management.
- Ash-Bernal R, Wise R, Wright SM. Acquired methemoglobinemia: a retrospective series of 138 cases. Medicine (Baltimore).
