Thrombopoietin: Platelets on Demand

We are still without a trustworthy medicine which can always be relied upon to control purpura.

Sir William Osler, 1892

In many respects, this dilemma is as poignant today as it was when it was first penned by the famous physician. Severe thrombocytopenia complicates the course of many patients with naturally occurring or iatrogenic states of bone marrow failure. Because more patients with cancer are now undergoing stem cell transplantation and aggressive therapy, the transfusion of platelet concentrates now exceeds that of red blood cells. Unfortunately, many problems often arise in patients receiving platelet transfusions; these include a high incidence of adverse reactions, refractoriness to subsequent platelet transfusions, and high costs. Accelerated recovery of the patient's own platelet production would obviously be highly desirable. Unfortunately, until only recently, our understanding of thrombopoiesis lagged far behind our understanding of other blood cells.

Ever since James Homer Wright began to stain blood smears and noticed a small disc-shaped cytoplasmic fragment, which he termed a blood plate [1], physicians have sought to understand the regulation of platelet production. Pioneering work in the early 1960s and 1970s provided a hierarchical model of stem cell maturation and the in vitro assays necessary to characterize and purify some of the humoral regulators of this process. These events, which shaped hematopoietic research for three decades, have yielded impressive results. With the cloning of erythropoietin [2] and granulocyte colony-stimulating factor (G-CSF) [3] in the mid-1980s, the hormones primarily responsible for erythrocyte and neutrophil production became available for clinical use. However, notably absent from these physiologic and biotechnical success stories was an understanding of platelet production and the hormone that was thought to be the primary regulator of the process: thrombopoietin.

The term thrombopoietin was first used by Kelemen in 1958 to describe the humoral substance responsible for the rebound thrombocytosis that follows states of thrombocytopenia [4]. For the ensuing 36 years, belief in the existence of a distinct regulator of platelet production waxed and waned as the purification and cloning of this protein remained elusive. Many in the field believed that if modern technology couldn't clone it, it didn't exist.

Occasionally, findings in one area of research have a profound, catalytic influence on a seemingly unrelated field of study. Identification of the myeloproliferative leukemia virus had such an influence on the search for thrombopoietin. In 1990, the responsible viral oncogene was cloned; 2 years later, the corresponding cellular protooncogene, c-mpl, was identified [5]. The predicted amino acid sequence of the latter had all of the hallmarks of a cytokine receptor and contained two “cytokine receptor motifs,” which are 200-amino-acid domains specific for nearly all of the known hematopoietic cytokine receptors. Although its ligand was not known, several features of c-mpl biology suggested that it might encode the thrombopoietin receptor. Two years after it was described, five separate groups using three distinct strategies had cloned complementary DNA for the Mpl ligand [6]. Initial physiologic studies clearly showed that the Mpl ligand is identical to thrombopoietin. To paraphrase Descartes, “it has been cloned, therefore, it exists”!

Will thrombopoietin provide an answer to Osler's dilemma? There are many encouraging signs. During preclinical testing, full-length thrombopoietin or a truncated, pegylated derivative molecule termed megakaryocyte growth and development factor (MGDF) produced impressive results. Administration of either recombinant form of the hormone increased platelet production 10-fold in normal mice and monkeys, and administration to myelosuppressed animals improved nadir platelet counts and accelerated hematopoietic recovery by as much as 2 weeks. The report of Vadhan-Raj and colleagues in this issue [7] and the recent report by Basser and Fanucchi and their associates [8, 9] extend these promising results to the clinical arena.

The administration of recombinant human thrombopoietin [7] or MGDF [8, 9] before planned chemotherapy resulted in dose-related increases in peripheral blood platelet counts. The increases occurred within 4 to 6 days of the initiation of treatment and, remarkably, continued for as long as 11 days after administration of study drug was discontinued [7-9]. This prolonged response can partly be traced to the long plasma half-life of these proteins (20 to 30 hours), but it must also reflect the effects of thrombopoietin on the early stages of hematopoietic cell development. This hypothesis is supported by Vadhan-Raj and colleagues: In their study, as in previous in vitro and preclinical studies [10, 11], thrombopoietin increased the frequency of primitive marrow hematopoietic progenitors. The magnitude of the thrombopoietic response was also impressive. Dosing with a modest amount of thrombopoietin (2.4 µg/kg of body weight per day) for only a single day was associated with a sustained increase in peripheral blood platelet counts of more than 200%. Administering lower doses of MGDF (0.3 to 1 µg/kg per day) for as long as 10 days resulted in a 50% to 600% increase in platelet count [8]. In addition, administration of recombinant human thrombopoietin or MGDF was safe. No significant adverse effects could be related to the administration of either drug, despite particular attention to the potential for thrombotic complications. In vitro studies suggest that thrombopoietin can prime platelets to respond to subthreshold levels of platelet agonists; when carefully studied, however, the administration of MGDF was not associated with enhanced platelet responsiveness [12]. It should be noted, however, that the patients selected for the clinical studies thus far reported were carefully screened for preexisting cardiovascular disease. The safety of thrombopoietin and MGDF in unselected populations is still an unresolved issue.

In what clinical settings will thrombopoietin be useful, and how effective will it be? Two of the published reports detail the prechemotherapy phases of clinical trials of thrombopoietin in patients with cancer. The hormone will probably be useful for accelerating platelet recovery after myelosuppressive therapy for three reasons. First, endogenous thrombopoietin levels do not increase immediately after chemotherapy; this provides a window during which the exogenous hormone could immediately begin the process of megakaryocyte recovery before the patient mounts his or her own response. Second, platelet transfusions suppress blood levels of the hormone; administered thrombopoietin could overcome this suppressive effect. Third, preclinical trials using several different myelosuppressive protocols were uniformly successful in ameliorating thrombocytopenia: Nadir platelet counts improved, and the duration of thrombocytopenia decreased substantially. Thus, administration of thrombopoietin should maximally stimulate the megakaryocytic progenitors remaining after cytotoxic therapy to expand and contribute to enhanced platelet production. However, a cellular substrate of adequate quantity and quality must be present for any hematopoietic growth factor to work. This poses the greatest barrier to the success of thrombopoietin in other states of marrow failure, such as aplastic anemia and myelodysplastic syndromes, in which erythropoietin and G-CSF are also of less clinical usefulness. The modest therapeutic advantage reported by Fanucchi and colleagues [9] probably reflects this potential limitation. Only carefully controlled clinical trials done in various settings will allow us to rationally use this new therapy to advantage.

In addition to being used in patients with thrombocytopenia, thrombopoietin will probably also be given to normal donors of platelets for transfusion. The rationale for this application is clear. More than 7 million units of platelet concentrates are administered annually in the United States alone. Approximately 35% to 50% of these units are solicited from individual donors in whom the use of thrombopoietin could increase platelet yields by 200% to 400%. In this setting, too, we can find important precedents from the experience with other hemopoietins. Both erythropoietin and G-CSF have been administered to large numbers of normal persons, as part of autologous blood or stem cell collections or allogeneic granulocyte donation programs, without serious difficulties. It is likely that the use of thrombopoietin in healthy platelet donors will also meet with similar success. However, the use of any drug in normal persons must be measured by strict criteria: The therapeutic ratio (advantages:disadvantages) has a very small numerator!

Finally, the speed with which the modern biotechnology and pharmaceutical industries traverse the time from hormone discovery to clinical use deserves comment. The cloning of thrombopoietin occurred in February 1994 (as assessed by patent filing dates), the cloning was first reported in Nature in June 1994, clinical trials began in mid-1995, and the early results of these trials are now being reported. However, this impressive result is not likely to be repeated very often, at least for other biologicals. Erythropoietin, G-CSF, and thrombopoietin are lineage dominant hormones that have few other physiologic effects. In contrast, most other cytokines and interleukins display a multitude of activities, many of which mediate different aspects of the inflammatory response. As a consequence, the response to therapeutic administration of these substances is complex and partially indirect, often resulting in many undesirable effects. The promise and failure of such biologicals as interleukin-1, interleukin-2, interleukin-3, interleukin-6, and tumor necrosis factor bear witness to this.

Thrombopoietin has moved from theory to reality and from cloning to the clinic in just 3 short years. Its effects in preclinical trials have matched those of in vitro and in vivo physiologic studies, and the results of its administration before and after chemotherapy for cancer are encouraging. We now await more extensive clinical trials of the agent in patients with iatrogenic or natural states of marrow failure to see whether the hormone fulfills its promise to produce platelets on demand. If it does, it will certainly help to solve Osler's dilemma.

• Copyright ©2004 by the American College of Physicians