Thrombopoietin (TPO) is thought to be the primary physiologic regulator involved in the regulation of megakaryopoiesis as well as platelet production. After almost 40 years of research for a primary regulator of platelet production, TPO has recently been isolated and cloned (1–3). Secondary thrombocytosis has been reported in several cancer patients. Before the cloning of TPO, many investigators challenged isolation of the novel growth factor, TPO from tumor cells derived from cancer patients with thrombocytosis. Cytokines of serum or culture supernatant were measured and immunohistochemial data suggested that IL-1, IL-3, IL-6, IL-11, GM-CSF were the candidates for such thrombocytosis. Overproduction of the IL-1β in hepatoblastoma (4), IL-6 in malignant mesothelioma (5) and ovarian cancer (6) were reported to involve secondary thrombocytosis.

After the cloning of TPO, to date, three kinds of tumors have been reported to produce TPO, resulting in moderate to marked thrombocytosis (7–9). Immunohistochemistry as well as high TPO levels in both serum and/or culture supernatant confirmed these TPO-producing tumors. Furuhashi et al (7) reported a case of TPO-producing ovarian carcinoma. Komura et al (8) reported seven cases of TPO-producing hepatoblastomas.

In this issue, Ryu et al (9) reported a 66-year-old man with hepatocellular carcinoma (HCC) showing marked thrombocytosis.

It is not surprising that HCC produces a high amount of TPO resulting in secondary thrombocytosis, because TPO mRNA is expressed in several organs such as the liver, kidney, spleen, smooth muscle, brain, and small intestine (2), and the liver is the primary organ that produces TPO (2, 10, 11).

In this case several cytokines such as erythropoietin, IL-6, TGF-β1 and hepatocyte growth factor in addition to TPO were also increased in serum. Since immunohistochemical analysis of these cytokines was not performed in this patient, authors could not clarify whether HCC cells produced some of these cytokines or TPO alone.

Two possible mechanisms of thrombopoiesis have been postulated. One is the sponge theory; TPO serum levels are dependent on the rate of platelet/megakaryocyte TPO receptor (c-mpl)-mediated uptake and catabolism (12). The other is the feedback control at the level of TPO mRNA in bone marrow (13). However, the patient in this issue showed decreased platelet counts despite an increased serum TPO concentration in terminal stage. The possible causes of thrombocytopenia in cancer patients with high serum levels of TPO may be attributed to platelet destruction or ineffective production of platelets. The former is usually due to disseminated intravascular coagulation (DIC). The latter has various causes; 1) the production of anti-TPO antibody neutralizing TPO (14), 2) increased levels of cytokines inhibiting megakaryocytopoiesis, such as TGF-β1, 3) catabolism of c-mpl in platelets/megakaryocytes by a high amount of TPO, 4) damage of microenvironment in bone marrow due to tumor invasion or post-chemotherapies.

The regulation of thrombopoiesis may be more complicated in patients with TPO-producing tumors rather than the cases in the clinical trials with recombinant TPO in thrombocytopenic patients, because these tumors might produce other cytokines in addition to TPO. These cytokines include both stimulating and inhibiting megakaryocytopoiesis. A novel pathway of TPO regulation by IL-6 has recently been proposed in inflammatory thrombocytosis (15). In cancer patients, the cytokine network is involved in thrombopoiesis even though serum concentration of TPO is extremely high.