Haematologica Journal of Hematology volume 85 – number 6 – June 2000

Haematologica 2000; 85:561-563 Editorial and Views

vation and platelet reactivity in humans,3 and these A global strategy for prevention may substantially increase the risk of thromboembolic and detection of blood doping complications especially in individuals with a genetic predisposition to thrombophilia.4 Although a minor- with erythropoietin and related ity of athletes abusing rHuEpo will eventually develop drugs a thromboembolic disease, the unlucky ones might die because of this, or experience serious handicaps for the rest of their life. Administration of rHuEpo The definition blood doping was introduced by the also involves an increase in the systolic blood pres- media in the 1970s to describe the use of blood trans- sure during submaximal exercise.5 A large portion of fusion to increase red cell mass artificially, and in turn the professional cyclists whose data have been exam- to enhance both maximal oxygen uptake and perfor- ined in recent investigations by Italian magistrates mance in endurance sports. Since the late 1980s show a degree of iron overload comparable to that of blood doping is no longer achieved by autologous patients with genetic hemochromatosis, with ferritin transfusion but instead through administration of levels often in excess of 1,000 ng/mL. These individu- recombinant human erythropoietin (rHuEpo). als were clearly given intravenous iron together with Magnitude of blood doping and prospects for rHuEpo. Although intravenous iron is primarily taken up by the reticuloendothelial cells, it is later redistrib- the next years uted to parenchymal cells. Therefore, this type of iron Clearly we cannot adopt the principles of evidence- overload will eventually produce organ damage com- based medicine in examining this issue, simply parable to that occurring in genetic hemochromato- because there is no conventional data source. Most of sis, including the risk of developing hepatic carcino- the available information derives from articles in ma. Finally, preliminary observations1 suggest that the newspapers and, more recently, from police investi- abuse of rHuEpo might involve a risk of post-treat- gations. According to data collected by CONI (Comi- ment blunted endogenous erythropoietin production, tato Olimpico Nazionale Italiano), [Italian National Olympic 1 including severe anemia. In particular, these individ- Committee], blood doping with rHuEpo is particular- uals would be unable to develop an adequate ery- ly common in professional cycling and cross-country thropoietic response to stress conditions. More gen- skiing. This is largely expected since these are typical erally, we still do not know the effects of long-term endurance sports. Nevertheless, several observations by CONI indicate that the abuse of rHuEpo is extend- treatment with hematopoietic grwoth factors, but ing to other disciplines. It is believed, in fact, that observations in animals suggest that there may be a blood doping can help reduce physiologic strain dur- risk of development of myeloproliferative disorders. ing exercise and accelerate recovery after training. It Over 500,000 patients throughout the world are should also be noted that blood doping is no longer now receiving rHuEpo for the treatment of anemia of a problem restricted to career athletes, since it now renal failure and deriving great benefit from such involves also amateurs and young athletes. treatment in terms of both quality of life and prolon- Prospects for the next years are discouraging. In gation of survival. Interestingly, the potential adverse effects on normal hemostasis are useful in renal fact, the major pharmaceutical companies are cur- 3 rently developing long-acting, modified Epo mole- patients to prevent hemorrhagic complications. In the last years, in addition, rHuEpo has been approved cules. One weekly injection of a long-acting stimula- 6 tor of erythropoiesis would be the ideal procedure for for other indications, including prevention of ane- dishonest athletes. mia in surgical patients or in patients undergoing plat- inum-based chemotherapy, treatment of anemia of Medical risks of blood doping with rHuEpo and prematurity, of anemia induced by zidovudine thera- vital importance of this drug for thousands of py in HIV-infected patients, and of anemia induced by patients chemotherapy for non-myeloid malignancies. One of There is speculation that blood doping with rHuE- the many adverse effects of rHuEpo abuse is that a po may be involved in the death of professional vitally important drug – that can prolong survival of cyclists from the Netherlands in the early 1990s.2 At thousands of patients – is nowadays reported by the that time, rHuEpo abuse was largely uncontrolled and media as a doping drug. It must be clearly stated that, Hct values in excess of 60% were presumably achieved. although blood doping is very common, the large These polycythemic conditions compounded by dehy- majority of rHuEpo preparations are used for patients dration during exercise readily predisposed athletes who benefit from them. Misinformation may be no to thromboembolic complications. Nowadays rHuE- less harmful than doping itself. It must also be clear- po abuse is undoubtedly more finely tuned. Howev- ly stated, however, that pharmaceutical companies, er, the medical risks associated with blood doping are just because of the vital importance of rHuEpo for so still considerable. many patients, should be more active in adopting Erythropoietin markedly enhances endothelial acti- measures to discourage erythropoietin abuse in sport.

Haematologica vol. 85(6):June 2000 562 editorials and views

What is currently being done to prevent blood soluble transferrin receptor and serum erythropoietin. doping It should be noted that these parameters can be deter- The use of rHuEpo is officially prohibited by the mined automatically in a few minutes, so that this International Olympic Committee (IOC) and other major approach could be adopted to test athletes before a sporting organizations. In 1989, the IOC Medical race. The above parameters were then utilized to devel- Commission introduced the new doping class of pep- op mathematical models aimed at discriminating tide hormones and analogues, which includes rHuE- between athletes given rHuEpo and those given place- po, human chorionic gonadotrophin and related com- bo. Interestingly the ON-model repeatedly identified pounds, adrenocorticotrophic hormone, human 94-100% of rHuEpo group members during the final growth hormone, and all the releasing factors of these 2 weeks of the rHuEpo administration phase (one hormones. false positive from a possible 189), while the OFF- However, there is no reliable method for routinely model repeatedly identified 67-72% of recent users detecting doping with rHuEpo at present since the with no false positives. recombinant molecule cannot be readily differentiat- The many sophists who have a seat in sports orga- ed from the endogenous hormone. Even if reliable nizations state that the use of indirect methods for tests could be developed quickly, rHuEpo will be detecting rHuEpo abuse is not formally correct. I invite replaced by modified molecules in the next years, so them to reflect on the following points: a) rHuEpo is that current efforts might become useless. officially prohibited; b) rHuEpo cannot be analytical- To dissuade the abuse of rHuEpo, some sports have ly detected; c) indirect methods are not allowed. If imposed upper limits on hematocrit and hemoglobin point c) is true, the only conclusion is that point a) is [50% and 17 g/dL, respectively, in males, International no longer true in reality. As hematologists we have Cycling Union [UCI]) or hemoglobin [18.5 g/dL in defined criteria and pathways for differential diagno- males, International Ski Federation (FIS)]. This strategy sis of erythrocytosis.9 Based on these algorithms we has many pitfalls, which have been extensively dis- routinely make a diagnosis of polycythemia vera (PV), cussed elsewhere and include: large natural variation in spite of the fact that we still lack a marker of this between individuals, postural effects on hematocrit, myeloproliferative disorder.10,11 Diagnosing PV involves risk of false positivity and ease of manipulation a lot of medical responsibility12 but an indirect diag- through interventions such as saline infusion. In par- nostic method is nevertheless accepted by the scientific ticular the upper limit of 18.5 g/dL adopted by the FIS community. Why should we not be allowed to diag- is difficult to understand: in a prospective study that we nose erythrocytosis due to previous administration of rHuE- are conducting on elite soccer players in Italy, no Hb po or related drugs using a rational approach? level greater than 17 g/dL has so far been observed. The indirect method reported in this issue of Hae- In my opinion, the adoption of upper limits might matologica can be further improved and the Aus- paradoxically generate more blood doping. In fact, it tralian researchers are currently conducting several induces cleverly manipulated uses of rHuEpo with the studies in Australia and worldwide. Clearly, the first aim of approaching the target hematocrit or hemo- objective is to abolish false positives as far as is possi- globin without exceeding it.1 This is the only explana- ble. The refined approach, however, should be part of tion I can provide for the elevated – although below global strategy for preventing and detecting blood 50% – hematocrit values frequently found in some pro- doping. Such a strategy must include definition of the fessional endurance sport athletes. One should con- individual hematologic profile, the so-called hemato- sider, in fact, that hematocrit levels greater than 47% logic passport. are found in only 1-2 out of 100 elite soccer players (see below The hematologic passport). Marginally elevat- The hematologic passport ed levels are even more suspicious if the same individ- CONI and FIGC (Federazione Italiana Giuoco Calcio) uals show considerable lower values during non-com- [Italian Soccer Federation] have recently launched a cam- petition periods: the reverse, in fact, should be expect- paign called Io non rischio la salute! [I take care of my ed. A not negligible adverse effect of adopting upper health!]. This involves regular hematologic investiga- limits is that they generate aberrant beliefs in athletes: tions (2-3 times a year) including blood counts, retic- doping is no longer taking rHuEpo but instead having ulocyte count, serum ferritin and soluble transferrin hematocrit levels greater than the upper limit. In oth- receptor measurements. These determinations must er words, abusing rHuEpo and having hematocrit val- be performed by laboratories participating in quality ues below 50% is felt by some athletes as a fully nor- control programs to keep analytical errors as low as mal behavior. possible. Sequential evaluation of the above parame- ters allows definition of the individual hematological Indirect methods for detecting erythropoietin profile. abuse in athletes A top-level Italian soccer team has undergone such In the last few years, a number of studies have inves- sequential studies over the last two years. Hematocrit tigated indirect methods to detect of rHuEpo abuse by values ranged from 37.2% to 47.3% (with less than 2% means of parameters indicative of accelerated ery- of values being ≥ 47%), and hemoglobin levels ranged thropoiesis. In this issue of Haematologica, Parisotto from 12.8 to 16.5 g/dL. Interestingly, mild-field play- and coworkers7 describe a novel method for the detec- ers showed the lowest hematocrit values (37-40%) and tion of rHuEpo abuse in athletes utilizing markers of hemoglobin (between 13 and 14 g/dL), confirming altered erythropoiesis.8 A number of parameters that the so-called sports anemia is a spurious anemia, proved to be closely related to rHuEpo administra- caused by an expanded plasma volume that dilutes red tion: Hct , reticulocyte Hct, percentage of macrocytes, blood cells as a beneficial adaptation to aerobic exer-

Haematologica vol. 85(6):June 2000 editorials and views 563 cise. Considering blood cell counts associated with refined, this approach will likely be insufficient by itself. normal values for reticulocytes, serum ferritin and sol- A careful definition of one individual’s hematologic uble transferrin receptor, the within-subject biological profile should form the basis for a more successful coefficients of variation ranged from 1.2% to 4.7% for application of any indirect method. Introducing into a hematocrit, and from 1.6% to 5.0% for hemoglobin. powerful predictive model a comparison with basal This study, which will be reported elsewhere, indicates hematologic data will improve the method’s sensitivi- that it is possible to define the individual hematologic ty and reduce the risk of false positivity. Should ongo- profile: a physiologic individual range for hematocrit ing studies in French laboratories provide us with a test and hemoglobin can be defined as the mean value ± that readily differentiates between exogenous and 10% of the mean. We can discuss for years about the endogenous Epo, also this test could become part of best size of variation (5%, 7.5%, 10%): a major factor a global strategy to prevent and detect blood doping. of variability here is produced by between-laboratory Mario Cazzola, M.D. differences, and the figure of 10% represents a better guarantee for the athlete at present. University of Pavia School of Medicine, Division of Hematology, The CONI campaign I take care of my health! not only IRCCS Policlinico S. Matteo, 27100 Pavia, Italy involves regular hematologic evaluations performed by athletes themselves, but also random controls Disclosure about potential conflicts of interest decided by the central authority. Increases in hemat- Professor Mario Cazzola is a member of the FIGC [Italian ocrit and hemoglobin > 10% of the mean value are Soccer Federation] Scientific Committee. In the last two years considered potentially harmful to the athlete’s health. he has been collaborating with CONI in the developing the For instance, if the mean hematocrit value (calculat- campaign I take care of my health! and is currently following ed on previous sequential determinations) is 42% and this campaign and conducting studies to improve it further. the present value is 47%, the increase is equal to 11.9% On behalf of the FIGC, Professor Mario Cazzola has collabo- and therefore non-physiologic. Any athlete showing rated with major Italian soccer teams by analyzing the hema- one of the following: a) increases in both hematocrit tologic data of their players in order to define the concept of the and hemoglobin > 10%; b) an increase in hematocrit individual hematologic profile. or in hemoglobin > 10% plus an abnormal reticulo- cyte count, or serum ferritin, or soluble transferrin receptor levels, is stopped to prevent damage to References his/her health. He or she will start to compete again once hematologic values return within normal ranges. 1. Commissione Scientifica Antidoping del CONI. Prog- We are fully aware that the approach of the CONI etto "Io non rischio la salute!". CONI, Rome, 2000. campaign I take care of my health! is not an anti-doping 2. Adamson JW, Vapnek D. Recombinant human ery- procedure, for several reasons, but mainly because this thropoietin to improve athletic performance. N Engl J campaign has been designed primarily to take care of Med 1991:324: 698-9. the athletes’ health. For instance, the inclusion of 3. Stohlawetz PJ, Dzirlo L, Hergovich N, et al. Effects of serum ferritin within the laboratory battery has already erythropoietin on platelet reactivity and throm- allowed us to identify several individuals with genetic bopoiesis in humans. Blood 2000;95:2983-9. 4. Zoller B, Garcia de Frutos P, Hillarp A, Dahlback B. hemochromatosis. Thrombophilia as a multigenic disease. Haematolog- Conclusions ica 1999; 84:59-70. 5. Berglund B, Ekblom B. Effect of recombinant human As physicians, one of our major duties is to prevent erythropoietin treatment on blood pressure and some diseases, and we have sworn this with the Hippocrat- haematological parameters in healthy men. J Intern ic Oath. Since blood doping exposes athletes to sev- Med 1991; 229:125-30. eral medical risks, we must be against blood doping, 6. Cazzola M, Mercuriali F, Brugnara C. Use of recom- and more generally against any form of doping. Blood binant human erythropoietin outside the setting of doping is not an abstract, intellectual challenge on uremia. Blood 1997; 89:4248-67. how to circumvent sports regulations licitly, but a 7. Parisotto R, Gore CJ, Emslie KR, et al. A novel method betrayal of the Hippocratic Oath for the physicians for the detection of recombinant human erythropoi- who are involved in it. Sport is intended to improve etin abuse in athletes utilizing markers of altered ery- thropoiesis. Haematologica 2000; 85:564-72. people’s health, doping worsens it. 8. Lacombe C, Mayeux P. Biology of erythropoietin. As hematologists, over the next years we could face Haematologica 1998; 83:724-32. problems related to blood doping with increasing fre- 9. Cazzola M, Beguin Y. New tools for clinical evaluation quency: atypical cases of iron overload, erythrocytosis of erythron function in man. Br J Haematol 1992; of unknown origin, unexplained anemias, atypical 80:278-84. thomboembolic complications, and so on. 10. Barosi G. Pathogenesis of polycythemia vera: do we Those of us who are involved in sports medicine have have the right pieces to the puzzle? Haematologica a hard task trying to prevent and detect blood doping. 1998; 83:97-8. There is no question that hematologists must play a 11. Fernandez-Luna JL, Silva M, Richard C, Sanz C, Beni- to A. Pathogenesis of polycythemia vera. Haemato- central role in this task force. To reach the objective, a logica 1998; 83:150-8. global strategy will be required. Although indirect 12. Barbui T, Finazzi G. Treatment of polycythemia vera. methods for the detection of rHuEpo abuse will be Haematologica 1998; 83:143-9.

Haematologica vol. 85(6):June 2000