The boy who was a state secret

In the autumn of 1912, at a hunting lodge at Spała in Russian Poland, an eight-year-old boy lay screaming in a room the whole of Europe was watching and nobody was allowed to explain.

He had knocked himself getting into a boat some weeks earlier — some accounts say climbing out of a bath — and seemed to recover. Then a jolting carriage ride with his mother restarted it. Over the following days his left groin and upper thigh swelled, the swelling tracked up into his abdomen, and he drew the leg up towards his chest and would not let anyone straighten it. He ran a high fever. He became delirious. The pain came in waves roughly every quarter of an hour and he barely slept for days. Two surgeons came down from St Petersburg, considered operating, and decided that opening him would kill him faster. The last rites were administered.

Public bulletins were issued on the heir's condition for the first time in the reign, and the newspapers settled into a death watch. The bulletins never named the illness. They could not: an heir to the Russian throne could not be publicly known to be chronically ill, and so the most consequential medical fact in the empire was managed as a secret for the whole of the boy's short life.

He recovered. It took weeks, and he was left with the leg fixed in flexion and unable to walk unaided for about a year. The imperial train took him home slowly, with the driver instructed not to brake hard.

What he actually had

For most of the twentieth century the answer was “haemophilia”, which everyone took to mean haemophilia A, because most haemophilia is A. In 2009 Rogaev and colleagues sequenced DNA from the Romanov remains — including the second Yekaterinburg grave found in 2007, which held the boy and one of his sisters — and answered it properly.

Nobody ever measured his factor IX, because no such assay existed in his lifetime and none can be done on bone. Severity is inferred from the predicted null effect of the splice defect and from how he actually behaved. It is worth saying that plainly rather than writing that he had a level under one per cent, which is a number nobody has.

The bleed with nowhere to show

Now read the Spała description again as a clinician. Groin and upper thigh swelling, extending into the abdomen. The hip held flexed and resisting extension. Days of fever. No external blood loss at all. Weeks to resolve, then a flexion contracture.

That is an iliopsoas haemorrhage — a large retroperitoneal bleed. No contemporary source uses the word; the diagnosis is a retrospective reading of lay descriptions by people who know what the lesion looks like. But it fits so exactly that it is worth using, because the lesion has lost none of its capacity to catch people out.

• The blood goes somewhere with no wallsThe retroperitoneum is a large potential space with nothing to tamponade against. Litres can accumulate with nothing to see and nothing to feel. Circulatory collapse can arrive before any swelling does.
• It impersonates appendicitisRight iliac fossa pain, a positive psoas sign, low-grade fever from resorption, a child who will not extend the hip. People with haemophilia have been taken to theatre for a hot appendix that was an iliopsoas bleed — a laparotomy in an uncorrected coagulopathy.
• It is not a hip bleed, and the difference is examinableIn an iliopsoas haemorrhage rotation of the hip is relatively preserved and relatively comfortable, while extension is not tolerated. In a haemarthrosis of the hip, rotation is restricted and exquisite in every direction.
• The femoral nerve is inside the compartmentIt runs within the iliacus fascia. Compression gives quadriceps weakness, a lost knee jerk and numbness over the anteromedial thigh and medial leg. Recovery is often incomplete, which is one plausible reading of why a boy who survived this bleed could not walk properly for a year.
• Factor first, scanner secondReplace before imaging, not after. Then confirm and size it, replace to near-normal levels and keep going for weeks rather than days because the rebleed rate is high, rest then mobilise under supervision, and do not drain or aspirate it.

And the thing that makes the 1912 story worth telling at all: none of that existed. There was no factor concentrate, no plasma, no cryoprecipitate, no tranexamic acid, no imaging that would have shown the lesion, and no operation that would not have killed him. The entire treatment was morphine, warmth, stillness and waiting to see whether the haematoma would reabsorb before the child died. Every item on that list above is a century old or less. It is a useful thing to remember on a bad night.

The Ras man

You cannot tell this story without him, and the medical literature has generally told it badly.

During the worst of the crisis, the Empress had Anna Vyrubova telegraph Grigori Rasputin, who was then at home in Siberia, more than two thousand kilometres away. A reply came back. As reported — and it is reported, in memoirs, rather than preserved as a document, so the wording should be held loosely:

Within a day or two the boy began to improve. Alexandra's faith in Rasputin became absolute, and the political consequences of that ran all the way to 1917. The episode is the hinge of his whole hold over the family, and Massie, whose 1967 account is where most people have met it, called it one of the most mysterious episodes of the entire Rasputin legend.

Three things are worth saying about it, in descending order of confidence.

First, and least romantic: haematomas resolve. If the patient survives the acute phase, that is what they do. The telegram arrived somewhere around the seventh or eighth day of an eleven-day illness — which is to say, at about the point when a survivor would be expected to turn. Rasputin had a habit of intervening late in episodes and predicting recovery, which is a textbook way to build a reputation for miracles out of base rates. This explanation requires nothing that is not already known.

Second: the advice in the telegram was not bad advice. Keeping the doctors away from a child with a huge retroperitoneal haematoma means less handling, less examination, less moving him about. That is, as it happens, correct management. Douglas Smith, his most serious modern biographer, argues that the operative effect was the calming of a frantic mother and a terrified child, and there is a real clinical observation underneath that — agitation and pain make bleeding worse, and calm, analgesia and immobility help. It is worth holding on to as contributory. It is not worth believing that serenity arrests an established retroperitoneal bleed in a child with no measurable factor IX.

Why this sits in a fortnight on the school-age child

Because Alexei was eight, and because severe haemophilia in a school-age child is a particular condition with a particular shape.

The bleeding that defines this age group is recurrent haemarthrosis — ankles first in childhood, then knees and elbows, the classic index joints. Children learn to recognise an aura: a tingling, a warmth, a stiffness in the joint that arrives hours before anything is visible. That matters clinically, because the child telling you their ankle feels “funny” is giving you the window in which factor works best, and the adult who waits for a swollen joint has missed it. Untreated, repeated bleeds drive synovial hypertrophy, then a target joint, then arthropathy and fixed deformity — which is the road Alexei was already some distance down.

The other thing to know for an emergency department is that intracranial haemorrhage is the leading haemorrhagic killer in haemophilia, that a substantial share of events — between a third and a half — happen without any identified trauma at all, and that the incidence is highest in the very young. Any head injury in a child with haemophilia is factor first, scan second, and a normal-looking child is not reassurance.

And there is a quieter reason this piece belongs to this fortnight. The theme running through the school-age block is the child who can tell you what is wrong and has learned not to — because they are being brave, because it might mean a needle, or because a parent is in the room and they are protecting them. Alexei is that child taken to its limit: a boy who could not run, could not be allowed to fall, was carried in public by a sailor assigned to catch him, and whose whole diagnosis was a thing the adults around him were forbidden to say out loud. Families still do versions of this. Not from reasons of state, but from shame, or fear of blame, or a wish to protect a child from their own chart.

What would happen now

Factor IX has turned out to be much easier to extend than factor VIII, because it is not tethered to the clearance of von Willebrand factor. Modern extended half-life products — albumin fusion, Fc fusion, glycoPEGylation — give terminal half-lives in the range of three and a half to four days and are dosed in practice somewhere between weekly and fortnightly, holding troughs well clear of the severe range. That converts severe haemophilia B into something that behaves like mild disease, and it is the real comparator against which everything newer has to be judged.

Then there is gene therapy, which is where the story gets genuinely interesting and not in the way the press releases intended.

• Etranacogene dezaparvovec (Hemgenix)A single infusion of an AAV5 vector carrying the hyperactive Padua variant of factor IX. FDA-approved November 2022 — the first gene therapy for haemophilia B. The five-year HOPE-B results published at the end of 2025 are strong: mean factor IX activity around 36 IU/dL with no meaningful fall over five years, bleeding rates down by roughly two-thirds, and 94 per cent of the trial cohort off prophylaxis and never restarted.
• Fidanacogene elaparvovec (Beqvez)Approved by the FDA in April 2024 and in Europe that July, on positive pivotal data. Pfizer discontinued it in February 2025 and withdrew the European authorisation in May, citing limited interest. Essentially no commercial patient was ever treated with it.

That second entry is the more instructive one. A one-time treatment for a lifelong disease, approved by two major regulators on good evidence, priced at around three and a half million US dollars, was taken off the market inside a year having treated almost nobody — not for any safety reason, but because patients doing well on a fortnightly injection were unenthusiastic about an irreversible single shot with unknown twenty-year durability, and payers were unenthusiastic about the price. Regulatory approval and clinical adoption turn out to be very different things.

In Australia the position is that Hemgenix holds provisional TGA registration from March 2024, and that MSAC did not support public funding in August 2024 — citing low-certainty evidence, wide variation between patients with no way to predict who will respond, a screening assay for anti-AAV5 antibodies that was unvalidated and performed only in the United States, and a price that did not stack up. Australian patients with haemophilia B are managed on factor IX concentrate funded through the national blood arrangements. And in March 2026 CSL announced a temporary global stockout of Hemgenix, which it attributes to the difficulty of manufacturing gene therapies rather than to anything about the product.

So the honest summary, in 2026, is that the disease which helped bring down a dynasty is now technically curable with one infusion, that almost nobody in this country is getting it, and that the unglamorous injection every week or two is what actually changed these children's lives.