The use of erythropoietins in cancer treatment

teaser

Paul Cornes
BMBCh MRCPr
Consultant Clinical Oncologist
Bristol Haematology and Oncology Centre
University of Bristol
UK
E: [email protected]

Although the drugs used for treating chemotherapy-induced anaemia are generally effective and safe, uptake across Europe has been variable. What factors underlie these different rates of use?

Erythropoietins – commonly called “epoietins” or “EPOs” – are recombinant human proteins that are licensed for the treatment of cancer-�chemotherapy-induced anaemia. Three different epoietins are available for use in Europe.

Information about epoietins abounds: a search of English-language material using the Google internet search engine in January 2007 revealed 891,000 web pages of information. A considerable proportion of the information about these agents points to controversy over dosing, schedules, safety and cost-effectiveness. While epoietins are indubitably effective, and their safety when used within product licence has been reaffirmed in a recent review by the US Food and Drugs Administration (FDA), their uptake across Europe has been variable. The reasons for the different rates of use are multifactorial, and this article explores these issues.

Cancer anaemia
The human body contains 30 trillion red blood cells (RBCs), or about five million cells per cubic millimetre of blood. In humans, mature RBCs do not contain nuclei. Their cytoplasm is filled with an iron-containing protein called haemoglobin (Hb), the substance that gives the blood its red colour. The proportion of immature nucleated RBCs (reticulocytes) in the circulation can be used as a measure of unbalanced RBC production. In healthy people the reticulocyte count is 1.5% of RBC or less.

Human RBCs are formed in the bone marrow and have an average lifespan of up to 120 days. New cells are produced at the same rate that cells are destroyed. This occurs at a rate of about two million cells per second. The old red cells are removed from the body by the spleen and liver, where they are then broken down. The component iron is recycled by carrier molecules (such as transferrin) to the body’s insoluble iron store in the liver, or back to the bone marrow as soluble iron to be used once again in making new haemoglobin.

RBCs carry oxygen from the lungs to all the organs of the body. This is so important to humans that the body keeps the levels of RBCs under tight control. For example, people who live at high altitudes have relatively higher RBC levels to compensate for the relative lack of oxygen. Similarly, smokers, who inhale poisonous carbon monoxide that blocks the carrying of oxygen by Hb, also have relatively higher RBC levels.

Red cell production in the bone marrow is regulated by erythropoietin, a growth hormone produced in the kidney. Erythropoietin is a protein produced in many different structural variants, or “isoforms”.(1) Isoforms of human erythropoietin available from pharmaceutical companies include two pharmacological isoforms, EPO-alfa (Eprex(R); OrthoBiotech) and EPO-beta (Neorecormon(R); Roche), and one chemically modified isoform, darbepoetin-alfa (Aranesp(R); Amgen), in which the duration of activity is prolonged by two extra oligosaccharide side-chains. Readers should be aware of differences in spelling of the terms erythropoietin and epoietin, typically used in Europe versus the USA – the forms just cited are European, while the US spellings are erythropoetin and epoetin, respectively.

Anaemia is the presence of a low RBC count, measured as the haemoglobin (Hb) concentration of a blood sample in g/dl, or as a percentage of the blood volume occupied by red blood cells, referred to the “haematocrit” (Hct). Normal values vary by gender,(2) by age and by the altitude at which people live. As noted, the haemoglobin level is tightly controlled in the body, as it regulates the oxygen-carrying capacity of the blood. Too little Hb (anaemia) results in reduced oxygen delivery to the tissues, with progressive fatigue, reduced exercise tolerance and finally vital-organ failure. Too much Hb (polycythaemia) leads to increased blood viscosity and the risk of clots and strokes.

Anaemia is the most frequent haematological complication in patients with cancer. In the �European Cancer Anaemia Survey, 15,000 patients were followed over a six-month period; of these, 68% were found to be anaemic at some point.(3) Although cytotoxic cancer chemotherapy might be considered the major cause of anaemia (75% of patients were anaemic during treatment), the background rate of anaemia in cancer patients who were not treated by chemotherapy was also high, at 40%.

The causes of anaemia in cancer are multifactorial. Anaemia of chronic disease is a syndrome not limited to cancer. It is a recognised complication of inflammatory disorders such as rheumatoid arthritis, where there is excessive release of inflammatory cytokine molecules (interleukin-1, interferons, tumour necrosis factor) and the iron regulatory molecule hepcidin.(4)

Anaemia, by reducing available oxygen in tissues, impacts on almost every organ and body system. This produces a wide range of symptoms that can severely reduce the quality of life (QOL) of a cancer patient.(5) Symptoms include breathlessness, dizziness, headache, chest pain, demotivation, depression, cognitive dysfunction and fatigue.(6) Fatigue is the most debilitating symptom of anaemia, with complaints of chronic exhaustion, tiredness, weakness and lack of energy. In comparison with the everyday tiredness we all experience in our daily lives, the fatigue is unrelieved by rest or sleep.(7) When patients were asked which affected their lives the most – pain, nausea, fatigue or depression – 60% said fatigue, 22% nausea, 10% depression and 6% pain.(5)

Fatigue is now the most serious complication of cancer treatment when rated by cancer patients.(5,8�9) It is rated higher than anxiety, pain, hair loss and nausea.(10) One-third of cancer patients report it each day.(6) Although fatigue can normally be a consequence of a range of factors – pain, breathing difficulties, depression, poor diet, drug toxicity – in cancer patients it appears to have a direct relationship with the level of anaemia a patient has at any given time.(11) Indeed, it is the strength of this relationship that suggested that anaemia in cancer patients might be the dominant factor in causing fatigue,(11) and it raised the possibility that treating anaemia could offer cancer patients significant gains in quality of life.(12) Fatigue leads to deterioration in patients’ physical functioning and health-related quality of life.(13) Perhaps more importantly, anaemia can also interrupt patients’ treatment programmes. Patients who become severely anaemic may no longer be able to receive chemotherapy, which may not only further reduce their quality of life but also reduce their chances of survival.

Treatment options in cancer anaemia
Transfusions offer rapid correction of anaemia. One unit of blood transfused will typically raise adult Hb level by 1 g/dl. However, the effect is short-lived, as transfused blood has a short half-life in the circulation compared with the typical 120-day lifespan of a patient’s own red cells. Transfusion is therefore indicated when anaemia is life-threatening – which is often taken to be when Hb is at 7 g/dl or less. Readers will appreciate that the quality-of-life benefit of correcting anaemia is seen over much higher blood-count ranges, as the greatest correlation of quality of life (QOL) score gains occurred when anaemia was corrected back to the normal level of 12 g/dl.(11) This is the reason why epoietins offer a greater QOL benefit than a transfusion programme for mild or moderate anaemia, as is seen frequently in cancer patients,(14,15) and in which cases physicians seldom aim to correct anaemia to more than 10g/dl.

Randomised trials of epoietins or transfusions as treatment for mild or moderate anaemia in cancer patients clearly show that epoietins are a more effective treatment, with a greater benefit in both Hb levels and quality of life.

The European Cancer Anaemia Survey (ECAS) has shown that a wide range of treatments are in use across Europe.(16) The most frequent response to diagnosing anaemia was to ignore it! When asked to rank the importance of addressing the symptoms of cancer and its treatment, doctors seem to value the treatment of fatigue significantly less than do patients.(6)

Also, treatment rates varied according to cancer type.(17) For example, of patients with anaemia, only 26% of those with breast cancer received any anaemia treatment, in comparison with 43% of patients with gynaecological cancer. Access to epoietins varied by country of treatment in Europe, ranging from 66% of patients in Italy to 5% in the UK.(18)

Treatment guidelines for cancer anaemia
This wide variation in use of effective anaemia treatment persists despite the production of authoritative consensus guidelines.(19) All three suggest investigation and treatment of mild or moderate anaemia in cancer patients (Hb < 9-11 g/dl), in comparison with transfusion guidelines that limit blood transfusions to lower levels of Hb (< 7-8 g/dl).(19-22)

Issues that might have restricted the use of erythropoietins for cancer anaemia in Europe

Safety issues
There have been reports of an increased risk of deep-vein thrombosis (DVT), of pure red-cell aplasia and of cancer progression with epoietins.(23-25) These serious issues triggered a review by the US Food and Drug Administration’s Oncologic Drugs Advisory Committee.(26)

The excess risk of DVT appears small – in the range of that seen, for example, with tamoxifen hormone therapy (relative risk 1.67, 95% CI 1.35-2.06). There is no clear relationship between the risk of DVT and Hb level, but current treatment guidelines emphasise the monitoring of all cancer patients during treatment to diagnose DVT early, not letting haemoglobin responses get too great and not producing supranormal blood counts (polycythaemia).

Pure red-cell aplasia, with profound anaemia resistant to epoietin, appeared to follow in patients treated with a particular epoietin alfa product, Eprex(R) (in single-use syringes), which was manufactured and distributed outside the USA. It has been suggested that the use of polysorbate as a stabilising agent, and the presence of organic compounds rather than the erythropoietin molecule itself, triggered anti-EPO antibody development. Reformulation of epoietin alfa seems to have controlled this risk.(27)

Cancer growth promotion in clinical use was suggested by two randomised trials that showed a trend towards increased death and cancer progression rates.(28-30) Both trials used �erythropoietin outside their current licences, treating nonanaemic patients during cancer therapy with the aim of preventing anaemia or achieving supra normal Hb levels. Furthermore, both trials have been criticised for possible imbalance at randomisation. It has been suggested that the presence of epoietin receptors on cancer cells could mediate a growth-promoting effect.(31) Meta-analysis of randomised �trials within licensed indications did not support these findings, nor did the long clinical experience gained with these same agents in their other clinical use in the anaemia of renal failure. Consequently, the FDA has again approved erythropietin for the treatment of cancer anaemia.(32) Package inserts and guidelines again emphasise that the aim is to restore Hb to 12 g/dl, and that levels higher than 13 g/dl mandate a break in treatment or dose reduction of the erythropoietin. The overviews of trials suggest that, when such treatment is within these licensed indications, survival is unaffected but QOL is improved.(33)

A further off-licence trial, with an epoietin used for cancer anaemia not associated with active cytotoxic treatment, has been terminated early. The European Cancer Anaemia Survey showed that the background rate of anaemia in cancer patients who were not treated with chemotherapy was high, at 40%, so it seems logical that epoietins would be tested in this setting.(2) An FDA safety alert of 26 January 2007 has alerted us to a problem with a randomised trial of darbepoetin alfa against placebo in 989 anaemic cancer patients (Hb <11 g/dl) without concurrent cytotoxic therapy. At a median follow-up of 4.3 months the survival difference was 54% with placebo vs 51% with epoietin.(34) Survival was not the primary endpoint of the trial. The trial recruited 60% of patients with stage IV metastatic cancer, so unbalanced randomisation may be an explanation for the result. Again, this emphasises that unapproved use of these agents should be reserved for clinical trials, where independent data monitoring boards can maintain patient safety.

Effectiveness issues
Initial trials of epoietins in cancer anaemia were designed to reduce the need for transfusions and used the thrice-a-week dose strategy. With the results of these trials came a recognition of the benefits from treating anaemic cancer patients to return them to normal Hb levels as a way of �restoring QOL. The significant effect on QOL was then a �trigger for a series of trials designed to test whether epoietins might themselves be cancer drugs or act synergistically with cytotoxic cancer treatments to increase rates of cure, perhaps by taking Hb into high-normal or polycythaemic levels. This explains why the aims and endpoints of treatments seem to change over time. Furthermore, this frustrates attempts to analyse all epoietin trials together in a single overview study, and argues for a more sophisticated analysis.(32)

Epoietins will reduce transfusion requirements, but as the target Hb level for transfusions is 7-8 g/dl, and that for epoietin therapy is 12 g/dl, the absolute reduction of transfusion events will be small.

Trials to prevent cancer anaemia as a way of improving cancer response rates and cures often used unlicensed starting doses of epoietins in non-anaemic patients. This approach has proved ineffective or dangerous, and it is outside current licensed indications.

When epoietins are tested in randomised trials for current licensed indications for mild or moderate anaemia during cytotoxic treatment, then the majority of patients treated appear to respond. Response is usually measured by an improvement in Hb of at least 1 g/dl (which appears to deliver a clinically relevant improvement in QOL) or 2 g/dl, or by return to normal Hb levels (typically taken as Hb ≥ 12).

Trials have emphasised good practice in screening patients for other treatable causes of anaemia such as iron, vitamin B12 deficiency or active bleeding. Standard schedules(35-39) all appear to show responses in 69.4-74.2% of patients within four weeks of starting therapy, compared with controls, in a survey of more than 10,000 patients enrolled in 20 randomised trials.(35)

The discovery that the anaemia of chronic disease involves the iron regulatory molecule hepcidin has led to its description as “iron-reuptake anaemia”. Patients with adequate insoluble iron stores may not be able to mobilise the soluble iron and transfer it to the bone marrow to make RBCs at times of increased demand, such as when epoietins are prescribed.(40) In this situation, typical preparations of oral iron may not help, but soluble parenteral iron is required.

Auerbach and colleagues have shown in a randomised trial that intravenous iron increased the haemoglobin response to erythropoietin �compared with oral iron or no iron (68%, 36% and 25%, respectively).(41) This has been tested for solid tumours by the Amgen AIM3 trial group, comparing treatment of cancer anaemia with �darbepoetin or darbepoetin and intravenous iron in patients who had been screened to exclude formal iron deficiency. Further trials are likely to explore the optimum doses and schedules for epoietin and iron therapy combinations, but the overall result is that the clinical effectiveness of treating the anaemia of chronic disease in cancer is now very high, with the vast majority of patients improving. The availability of intravenous iron sucrose preparations has made this a straightforward outpatient procedure, as the latter preparations replace the iron dextran formulations that were associated with allergic-type reactions. Readers will need to be aware that although iron sucrose preparations are now often advised as standard therapy for cancer anaemia along with epoietins, available preparations may only hold a current licensed indication for use with epoietins in chronic renal disease.(42)

Readers are reminded that the aim of epoietin therapy with cancer anaemia is to correct fatigue, by way of anaemia treatment. Recent trials have suggested the use of validated patient performance scores as the most appropriate endpoint of a study. Scales such as the Functional Assessment of Cancer Therapy (FACT) or linear analogue scales of activity and energy are available for this use.(43)

Convenience issues
Epoietin alfa was the first available recombinant human epoietin. It was introduced to treat the anaemia associated with renal failure. The kidney is the organ where erythropoietin is made, and patients with renal failure suffer profound anaemia without treatment. Dialysis is given three times a week, so when epoietin was introduced to renal practice this same schedule was used. In cancer medicine, the production of erythropoietin is often unimpaired, but the response of the bone marrow to erythropoietin is blunted by the cancer and the inflammatory response to it. For this reason, higher doses of epoietins are used in oncology compared with nephrology. With greater experience, the frequent and individualised dosing schedules of renal medicine have been replaced by subcutaneous administration once weekly or longer, and fixed-dose schedules with click-stop or preloaded single-use syringes.(35-39) In comparison with epoietin subcutaneous injection, which can be self-administered by patients, blood transfusion is associated with hospital attendance and significant inconvenience.(44)

Cost issues
Epoietins have been seen as high-cost supportive care drugs that may be a luxury even for wealthy healthcare systems to provide. If they affect only QOL and not survival, then their benefits need to be seen in either improved QOL or in cost savings elsewhere in the care of patients.

Directly comparing a policy of transfusion support against epoietins for cancer anaemia is to misunderstand the different roles of the two treatments, and unsurprisingly the formal cost-benefit studies in this setting tend to favour treatment of the minority at severe anaemia levels with transfusions.(45) Transfusions are used for rapid correction of life-threatening severe anaemia (Hb < 7-8 g/dl), where quality of life is not the main issue. Epoietins are used to treat anaemia above the threshold levels of transfusion (9-12 g/dl), with the aim of improving QOL. The short-lived benefit of transfusion needs to be compared with the more durable effects of epoietins. For example, one study of patients anaemic during breast cancer chemotherapy showed that the benefits of epoietins with improved quality of life were seen up to six months after the treatment had been completed.(46)

Blood transfusion is often seen as a free resource, but this is not the case.(47,18) Collecting and testing donated blood has significant costs involved, which increase with new safety and purification standards.(48) The price that hospitals pay for blood has to cover not just acquisition cost but also costs for blood-bank handling, laboratory tests and blood administration. Delivering transfusions requires transport to hospital, medical and nursing staff, and transfusion chairs or hospital beds. Moreover, as single transfusions are ineffective for chronic anaemia correction, multiple hospital attendance may be required, with a quarter of transfusions requiring an overnight hospital admission.(49)

The expense of transfusion is thus spread across a wide range of hospital department budgets. In comparison, although the costs of epoietins are high and centred on the oncology pharmacy budget, the cost of their delivery is cheap and their use may save costs elsewhere in the healthcare and hospital system.

Blood transfusions have safety issues too, with associated risks of bloodborne infections, haemolytic reactions, volume overload, supply deficits for certain blood types, allergic-type reactions and iron overload.(50-53)

Failure to treat anaemia entails its own costs, with reduced QOL, absence from work, loss of wages, and increased morbidity or even mortality.(54) These consequences translate into increased �monetary costs, as caring for patients with anaemia has significant additional cost when compared with nonanaemic patients.(55-57)

An American study assessed the costs of cancer to an employer with more than 100,000 staff. Medical, pharmacy and disability claims data from 1996 to 1998 were used to identify cancer patients with and without anaemia.(54) The treatment of anaemia was associated with increased costs of approximately US$3,775 per year over the course of cancer treatment. Approximately US$1,353 per patient per year was spent on the direct diagnosis and treatment of anaemia (for example, lab tests, epoietin therapy and transfusions), accounting for 36% of the total US$3,775 increase in annual costs. Therefore, about 64% of the cost of anaemia in cancer patients is associated with conditions indirectly related to anaemia, such as costs related to the symptoms and side-effects of anaemia, including fatigue. Patients in the workforce who had cancer and anaemia also showed higher levels of absenteeism than other patients with cancer but no anaemia – the former were absent from work for an average of eight extra days over the first year, after anaemia diagnosis. The associated cost of work lost was US$1,117 a year per patient. Therefore, patients with anaemia were implicated in, on average, annual additional per-capita costs of US$4,892 � comprising treatment (US$3,775) and absenteeism (US$1,117).

It follows that a large proportion of the cost of anaemia in cancer patients could be avoided with early detection and more aggressive treatment of anaemia. Furthermore, the likely future rising costs of transfusions, hospital and homecare can be compared with the falling pharmacy costs of epoietins over recent years. Early analysis of the costs of epoietins to correct anaemia in renal failure show significant trends. The cost of a quality-adjusted life-year (QALY) in 1992 exceeded £100,000, but by 2003 this figure was estimated at £17,000 – well within the threshold for approval by the UK National Health Service.(58,59)

A measure of the clinical effectiveness of epoietins used in accordance with the EORTC and NCCN guidelines was provided by a randomised trial by Straus and colleagues. In this trial, 269 patients undergoing chemotherapy were randomised to start epoietins when Hb fell to 10-12 g/dl or when the Hb was <9 g/dl; a third group did not receive epoietins because baseline Hb was >12 g/dl.(60,61) Epoietins given by the guidelines at Hb = 10-12 g/dl halved the number of days that a patient spent in bed and almost halved the number of days with reduced activity.

When cost-benefit studies are performed to compare return to normal Hb levels or QOL as the chosen primary endpoint, then epoietins are the cheapest method of gaining them for anaemic cancer patients. Taking Hb level as endpoint, Cremieux et al estimated that effectiveness resulting from US$1 spent on standard care could be achieved with only US$0.81-worth of epoietin alfa care.(62) In a study with QOL as endpoint, epoietin beta was shown to be cost-effective compared with standard therapy including transfusions as required.(63) Mean costs of epoietin beta therapy for 12 weeks were US$13,700-15,900 vs US$6,400-8,800 for standard care in an open trial of 262 patients with cancer. When changes in QOL were taken into account for each group, using the FACT-Fatigue (FACT-F) scale, it was estimated that for every US$1 spent on standard care only US$0.14-0.19 was spent on epoietin beta therapy to obtain the same QOL improvement.

Physician information issues
The treatment of cancer anaemia is compounded by lack of recognition of the importance of fatigue as a symptom in cancer patients. Physicians focus routinely on issues of pain and nausea, while few appear to routinely assess fatigue. Stone and colleagues surveyed 358 patients with fatigue in cancer.(64) Only 14% were prescribed or recommended any treatment.

Of the 75 patients advised, 52% were advised rest – a treatment that has been shown to be either ineffective or detrimental. Even when physicians are aware of fatigue and its severity, they still seem to ignore treatment, even when as many as 12% of patients report that they would prefer death to struggling on.(7) This study showed that 61% of patients reported that their daily lives were more adversely affected by fatigue than by cancer-related pain, whereas only 37% of oncologists perceived this to be the situation.

Conclusions
The recent US FDA review has confirmed the safety of epoietins in current licensed indications. Used within these limits, they appear to be a safe and clinically effective treatment to relieve the fatigue associated with mild or moderate cancer anaemia.Epoietins will not replace blood transfusion, as severe anaemia still occurs in cancer medicine, which can be corrected rapidly and reliably by transfusion. Epoietins for cancer patients have been tested in trials with three main different areas of focus – transfusion avoidance, direct cancer treatment and improving quality of life. At the same time, doses and schedules of treatment have evolved. This means that overviews of all trials of these treatments can give confusing or contradictory results. European guidelines from the EORTC offer us evidence based guidelines for treatment.

Reimbursement seems to vary between EU member countries. The method of cost-effectiveness assessment, and the endpoint chosen (transfusion avoidance, survival or quality of life) all seem to explain why wide variations in the value of epoietins have been described.

References

1. Br J Haematol 1998;100:79-89.
2. http://mycourses.med.harvard.edu/ResCourses/_online_textbooks/PBD/nrrt.htm
3. Ann Oncol 2002;13 Suppl 5:69.
4. Int J Hematol 1999;70:7-12.
5. Oncologist 2000;5:353-60.
6. Semin Oncol 1998;25:43-6.
7. Semin Hematol 1997;34(3 Suppl 2):4-12.
8. Eur J Cancer 1998;34:1670-6.
9. Curr Opin Oncol 1999;11:244-9.
10. J Pain Symptom Manage 1998;16:298-306.
11. Cancer 2002;95:888-95.
12. Br J Cancer 2002;87(11):1341-53.
13. Eur J Cancer 2003;39:335-45.
14. Med Oncol 1998;15 Suppl 1:S47-9.
15. N Engl J Med 1990;14;322(24):1693-9.
16. Eur J Cancer 2004;40(15):2293-306.
17. Oncologist 2005;10(9):743-57.
18. BMJ 2002;325:655.
19. Arch Pathol Lab Med 1998;122:130-8.
20. Hematology (Am Soc Hematol Educ Program) [book] 2001:10-30.
21. US National Comprehensive Cancer Network. Jenkinstown PA, 2007. Homepage at: www.nccn.org
22. Eur J Cancer 2004; 40(15):2201-16.
23. Cancer 2003;98:1514-20.
24. Nephrol Dial Transplant 2003;18(5):865-9.
25. Lancet Oncol 2002;3:145-53.
26. http://www.fda.gov/ohrms/dockets/ac/04/briefing/4037b2.htm
27. Blood 2005;15;106(10):3343-7.
28. Lancet 2003;362:1255-60.
29. Lancet Oncol 2003;4:459-60.
30. J Clin Oncol 2005;1;23(25):5865-8.
31. Eur J Cancer 2007;43;3:510-9.
32. Cochrane Database Syst Rev 2006;19;3:CD003407.
33. J Clin Oncol 2005;23:6941-8.
34. http://www.fda.gov/medwatch/safety/2007/Aranesp_DHCP_012707.htm
35. Blood 2006;108(11):Abstract 3764.
36. http://www.orthobiotech.com/common/prescribing_information/PROCRIT/PDF/ProcritBooklet.pdf
37. http://www.rocheuk.com/ProductDB/Documents/rx/pil/NeoRecormon_PFS_PIL.pdf
38. Oncology 2005;69(Suppl. 2):8-16.
39. http://www.aranesp.com/pdf/aranesp_PI.pdf
40. Blood 108(11):Abstract 5516.
41. J Clin Oncol 2004;1;22(7):1301-7.
42. http://www.venofer.com/VenoferHCP/new_guidelines.asp
43. Semin Hematol 1997;34 3 Suppl 2:13-9.
44. Blood Purif 1990;8(5):268-71.
45. Br J Cancer 1998;78:781-7.
46. Clin Breast Cancer 2005;5(6):439-46.
47. Transfusion 2001;31:318-23.
48. Curr Med Res Opin 2003;19(7):643-50.
49. Br J Cancer 2000;82(1):93-7.
50. BMJ 1999;318:1435-6.
51. Hematology (Am Soc Hematol Educ Program) [Book] 2001;47-61.
52. BMJ 2002;325:400-1.
53. Transfus Clin Biol 2003;10:1-5.
54. Am J Manage Care 2000;6:1243-51.
55. Am J Med 2003;115:104-10.
56. Value Health 2005;8:149-56.
57. J Manage Care Pharm 2005;11:565-74.
58. Pharmacoeconomics 1992;1:346-56.
59. Eur J Health Econ 2003;4:115-21.
60. Blood 2003;102:497a.
61. Cancer 2006;107:1909-17.
62. Pharmacoeconomics</spa