Introduction of Epo into clinical practice – the UK contribution
Christopher G Winearls
The cause of, and solution to, the problem of renal anaemia had been known since 1960 when Hegstrom working for Belding Scribner in the Seattle haemodialysis unit described their experience and wrote presciently “Erythropoietin, were it available would greatly reduce transfusion requirements.” Hegstrom became a family doctor and the problem was given to Joe Eschbach who worked on it for the rest of his career.
UK nephrologists worked on the subject and understood it rather well. Dr Raman Gokal did excellent work on iron metabolism with Sir David Weatherall in Oxford. Dr P Mary Cotes at the CRC in Northwick Park Hospital developed an accurate assay for erythropoietin (epo). Prof Harry Lee in Portsmouth and Dr John Curtis at Charing Cross tried the effect of cobalt and showed some benefit in a few patients.
Epo could not be purified from animal kidneys because it was not stored after synthesis.
Various British nephrologists working on aluminium toxicity noticed that a hypochromic anaemia was a feature. Aluminium inhibits the haem synthase enzyme. Treatment with desferrioxamine improved the anaemia but the mechanism was only later understood. (Altmann P et al. .Lancet. 1988 May 7;1(8593):1012) .
In 1983 a small California based biotech company decided to try to clone the epo gene so they could make a recombinant protein for clinical use. This had been done for other endocrine deficiency states notably insulin for diabetes. Bigger commercial players had eschewed this challenge believing the market for epo to be too small. They were not in business to create drugs for orphan diseases. The Amgen scientists were successful.
They approached Joe Eschbach and Dr John Adamson a haematologist in Seattle to do a pilot study. But Mary Cotes had wind of the cloning and ambushed Dr Noel Stebbing, an Amgen clinical scientist, during his visit to Prof David Tyrell at the CRC in Northwick Park discussing interferons to treat the common cold. She persuaded Stebbing to let her and a team of UK haematologists test the recombinant protein, r-HuEpo as it was called (most often abbreviated to Epo). She had the assay for epo so could do the pharmacokinetics. Prof Martin Pippard would perform the RBC survival and ferrokinetic studies, and Dr Cecil Reid would study the effect on the bone marrow. What she did not have was a nephrologist or any patients with renal anaemia. She phoned Prof Keith Peters at the Royal Postgraduate Medical School (RPMS) Hammersmith Hospital. He remembered that his senior registrar when asked at interview what problem in nephrology he would like to see solved had said, “renal anaemia.” He sent the senior registrar a memorandum:
Amgen agreed to Mary Cotes designing a pilot study with extensive investigations so that if r-HuEpo did not work they might know why. She made contact and planning began.
She measured the epo levels in the Hammersmith haemodialysis patients – there were thirty five.
The mean epo levels were in the normal range and a few were quite high but nothing like as high as those seen in patients with equivalent anaemia but no renal impairment, which were >1000mIu/ml. So how should the r-HuEpo be dosed? Should we aim for >1000mu/ml?
There was a year of planning – all day Saturday meetings attended by the Amgen scientists and regulatory affairs people. All was going well with Amgen when out of the blue they announced that they would no longer be running the study. They had come to an arrangement with Johnson and Johnson (J&J) who would have the worldwide rights to develop and market ‘Epogen’ outside the USA. J &J sent a senior doctor to speak to us; he was a nice orthopaedic surgeon who was “unfamiliar” with renal anaemia. They decided to stick with the Cotes team. There was ethics to be considered and patient selection. Here there was a problem – Hammersmith had a very small unit and the patients had to be fit and willing to undergo the rigorous testing. I contacted Dr Desmond Oliver who ran the Oxford Renal Unit where I had been an SHO and later registrar. In Oxford there were more than 200 patients on HD and many were fit, employed and on home HD. He was delighted to help and recruited quickly. The r-huEpo arrived with an Amgen scientist, Michael Downing who insisted that he see it into the pharmacy with a high-low thermometer installed. There had been a fridge failure in Seattle resulting in a whole batch being discarded.
The second patient responded very slowly and then had status epilepticus as her blood pressure increased as her haematocrit reached 25%. She was transfusion dependent and was used to a haematocrit around 15%. All hell broke loose and Amgen not J & J sent someone over straight away to crawl over all the notes and records. I was told that a Seattle patient had had the same side effect. I took advice from Colin Dollery (Professor of Clinical Pharmacology at RPMS) who calmed me down and came up with the simple explanation that her anaemic vasodilatation had been reversed by the rise in haemoglobin. The second patient responded beautifully:
There were no blood pressure problems in Oxford where the patients did longer hours and had immaculate blood pressures. They did have two fistulas thrombose – perhaps because platelet function is improved by a higher haematocrit. The fistulas involved had anatomical problems and this was enough to cause them to clot. The haematocrits and the viscosity would both have been higher at the end of a dialysis session. We soon had convincing efficacy data which I presented internally. Keith Peters told me to publish immediately. Mary Cotes was not keen as she wanted a nice JCI paper with all the pharmacokinetics, ferrokinetics and marrow function included. Keith Peters prevailed. I typed the paper on my BBC computer over a weekend. The Editor of the Lancet, Ian Munro was phoned and he said he would publish in two weeks. He did:#
(Later Mary Cotes published her paper describing the pathophysiology of renal anaemia and its response to r-huEpo in the Quarterly Journal of Medicine. Erythropoiesis was reduced but increased in response to Epo to a degree that compensated for the shortened red survival that was not affected by treatment.)
After the pilot study, J & J through their European company, Ortho Biotech, undertook Phase 3 trials in transfusion dependent subjects. The UK participated. One unit had to leave the study as the PI decided that trial protocols were for lesser mortals. There were more seizures in the UK and these were described in detail by Dr Mair Edmunds in Leicester:
Administration of recombinant erythropoietin (r-HuEPO) is an effective treatment for the anaemia of chronic renal failure, but in some patients it has been accompanied by elevated blood pressure. This study focuses on seven patients with end-stage renal failure, managed on haemodialysis, who developed probable hypertensive encephalopathy with seizures during treatment with r-HuEPO. All made a full recovery. The events were not clearly related to the haemoglobin concentrations achieved, and four patients have subsequently been restarted on r-HuEPO therapy at a reduced dose, resulting in a slower increase in haemoglobin with no recurrence of episodes of severe hypertension. Close attention needs to be paid to blood pressure in patients commencing erythropoietin therapy, and it seems prudent to aim for a gradual increase in haemoglobin concentration to allow the circulation to adapt to changes in oxygen delivery and haematocrit. (Edmunds ME et al. NDT 1989; 4: 10651065–1069).
r-HuEpo was recommended for licensing by the CPMP ( Committee for Proprietary Medicinal Products) in Brussels in 1989 and available for UK patients that year as ‘Eprex’. The price was set at ~£10/1000units which at a dose of 6000-12000 units a week was going to add £3-6k to the annual cost of treating a patient on HD. The cost was already £12-15k. This set alarm bells ringing in Regional Health Authorities and the Department of Health. Epoetin-Beta, another recombinant erythropoietin had been acquired from Genetics Institute by Boehringer Mannheim (later Roche). It had been trialled in Cardiff by Iain MacDougall working with Prof John Williams. It was also soon licensed and the price was set exactly the same as Eprex.
The Renal Association was asked to make recommendations on the criteria for Epo use. We recommended that patients who were: iron overloaded; transfusion- dependent; HLA sensitised transplant candidates and individuals whose employment was limited by anaemia should be treated to achieve Hb 12 g/dl. This was tough on those symptomatic patients who had Hb <12g/dl but not in these categories. Money was found but was limited. Units were told to prescribe Epo inside this budget. A Health Minister was challenged when she boasted about the budget but failed to admit that a slice would be clawed back in VAT. HD Units could not avoid this so we came up with the wheeze of asking GPs to prescribe at our direction on FP10s (the prescription forms for community pharmacies) for home HD patients. The GP drug budget was not capped. Some units asked the GPs to prescribe for unit patients who brought in the vials in chill bags to be administered during dialysis. This apparently broke the rules as the hospital staff actually administered it! It was soon shown that subcutaneous (sc) administration was more cost effective as lower doses could be used. Gradually the limits were breached, and in the 2020s 90% of HD patients and 80% of PD patients are now receiving Epo.
The pharmacokinetic data showed high peak concentrations and an 8-hour half-life of the iv injected epo. This was unphysiological and led to the change to sc administration and the development of long-acting forms of Epo – either pegylated epoetin-beta (Micera) or a genetically modified epoetin that had more carbohydrate moieties – darbepoetin alfa (Aranesp). Much of the early testing of Aranesp was done by Iain MacDougall’s group at Kings College Hospital. This new molecule developed by Amgen was not epoetin-alfa so Amgen were free to compete with the J&J ESA that they had ceded in 1988. This was contested. The NHS was able to go out to tender, the cartel was broken and prices came down. When the patents expired the biosimilars appeared and made anaemia treatment cheaper still.
Epo was soon taken for granted but a number of problems were unresolved. The first was the choice of appropriate target haemoglobin levels. In the United States HD patients with cardiac disease receiving Epo aiming for a normal haematocrit had worse outcomes. Would achieving a Hb of 12g/dl improve cardiac outcomes in CKD patients? There were two trials, one led by Iain MacDougall in non-diabetics, and TREAT, a US trial asking the same questions in those with diabetes. The answer was that achieving Hb 12g/dl did not improve outcomes. There was a slight excess of strokes, but there was no proven harm.
In the mid-1990s a complication emerged that we, when we were performing the pilot studies, had feared – the development of neutralising antibodies. We had found no antibodies in the 1980s, but suddenly cases of PRCA (Pure Red Cell Aplasia) were being diagnosed and neutralising antibodies found. Although initially denied by J & J it was largely product specific problem – only epoetin-alfa formulated by J & J and given sc from prefilled syringes. When the formulation was changed the incidence dropped but it still occurs. An Epo-mimetic was tested in the patients with Epo resistance caused by the neutralising antibodies by Ian MacDougall. This was a neat and effective solution but the drug had to be withdrawn when the patients developed antibodies to it. (MacDougall I et al. N Engl J Med 2009; 361:1848-1855).
The other controversy concerned iron supplementation. In Oxford, Dr Martin Pippard was working on iron loading anaemias, while Dr Raman Gokal studied iron metabolism and overload in HD patients, as an MRC training fellow in Oxford supervised by Professor Sir David Weatherall and Dr Sheila Callender. This was in the late 1970s, well before Epo. Patients were given high doses of iron dextran and became significantly iron overloaded. Gokal showed several important things. First that the ferritin normal range was different in HD patients – a ferritin of <100ug/L indicated minimal bone marrow iron stores. The second was that iron deficient HD patients can increase iron absorption. The pilot studies showed that once the marrow was stimulated with Epo there was need for a lot of iron. The iron loaded patients mobilised it but the iron replete patients could not keep up with demand by increasing absorption and required regular supplementation. The PIVOTAL Study eventually in 2018 showed that a proactive iron supplementation policy is safe and may even reduce ESA requirements. The PIVOTAL study was another brainchild of Iain MacDougall who drove it to completion (MacDougall I et al.New Engl J Med 2019; 380: 447-458).
Although Epo was a major boon – relieving patients of the effects of severe anaemia and the need for regular transfusion – there were still concerns about the safety of anaemia correction which was thought to cause an excess of cardiovascular events. The evidence for this, mostly from US trials, was weak. Certainly, aggressive treatment with high dose iron and Epo in HD patients with heart disease aiming for a normal haematocrit did not improve prognosis and was probably deleterious. Curiously those subjects achieving a normal haematocrit did appear to benefit. NICE weighed up the evidence and made a series of reasonable recommendations which have allowed CKD patients whether on dialysis or not to be treated with confidence. Results reported by the UK Renal Registry now show near uniform performance in UK kidney units.
An alternative to Epo is now available thanks to the exploitation of the discovery of the oxygen sensing mechanism by among others Sir Peter Ratcliffe. Drugs that inhibit prolyl hydroxylase prolong the action of HIF (hypoxia-inducible factor) on residual renal and hepatic Epo-producing cells so treated patients produce sufficient endogenous Epo to maintain normal erythropoiesis. This work upset the previous dogma that the explanation of the failure to produce sufficient Epo was a lack of Epo producing cells. There were sufficient cells but they did not respond to the anaemia. These orally active prolyl hydroxylase inhibitors (PHIs) are now licensed but not yet widely used. There are unsubstantiated claims that they will on their own improve iron delivery and directly suppress hepcidin which by suppressing iron uptake is part of the cause of the anaemia of inflammation.
Last Updated on June 18, 2024 by neilturn