- Top of page
- PATIENTS AND METHODS
In an attempt to evaluate the role of thrombopoietin (TPO) in the pathobiology of aplastic anaemia (AA), we have examined TPO levels in sera from 54 AA patients and 119 healthy controls. A total of 92 samples were collected from AA patients: 43 samples were harvested at diagnosis, 23 samples in the cytopenic period after treatment, and 26 samples when patients were in partial (n = 10) or complete remission (n = 16) following immunosuppressive treatment. TPO serum levels were assessed by a sandwich-antibody ELISA that utilized a polyclonal rabbit antiserum for both capture and signal. Serum samples from normal donors revealed a mean TPO level of 95.3 ± 54.0 pg/ml (standard deviation). Mean TPO levels in AA sera collected at diagnosis and before onset of treatment were 2728 ± 1074 pg/ml (P < 0.001 compared to normal controls; mean platelet count at that time: 27 × 109/l). TPO serum levels of AA patients in partial or complete remission after immunosuppressive treatment were significantly lower than TPO levels at diagnosis (P < 0.001). However, despite normal platelet counts (mean 167 × 109/l), TPO levels remained significantly elevated in complete remission (mean TPO 1009 ± 590 pg/ml, P < 0.001 compared to normal controls). There was a significant inverse correlation between serum TPO levels and platelet counts in AA patients who were not transfused for at least 2 weeks prior to sample collection (coefficient of correlation (r) = −0.70, P < 0.0001).
In summary, TPO levels were highly elevated in sera of patients with AA. Thus there is no evidence to suggest an impaired TPO response contributing to thrombocytopenia in AA. Thrombopoietin did not return to normal levels in remission, indicating a persisting haemopoietic defect in remission of AA. We hypothesize that elevated levels of TPO may be required to maintain normal or near normal platelet counts in remission of AA.
Acquired aplastic anaemia (AA) is characterized by peripheral blood bicytopenia or pancytopenia and bone marrow hypocellularity ( Young, 1995). In most cases the underlying cause remains unknown. Several mechanisms have been implicated in the pathogenesis of this disease, including an intrinsic defect of haemopoietic stem cells, lymphocyte-mediated suppression of haemopoiesis, dysfunction of the bone marrow microenvironment, and overproduction of inhibitory cytokines such as tumour necrosis factor (TNF) or interferon-γ (IFN-γ) ( Young, 1995).
There is no evidence to suggest that AA is caused by a deficiency of any known haemopoietic growth factor. Serum levels of positively acting haemopoietic growth factors such as granulocyte-macrophage colony stimulating factor (GM-CSF), granulocyte-CSF (G-CSF), erythropoietin (EPO) and flt3 ligand are consistently increased in the majority of AA patients ( Watari et al, 1989 ; Schrezenmeier et al, 1993 , 1994; Kojima et al, 1996 ; Wodnar Filipowicz et al, 1996 ), except for reduced serum levels of stem cell factor (SCF) ( Wodnar Filipowicz et al, 1993 ; Nimer et al, 1994 ). In vitro, production of G-CSF, GM-CSF and interleukin-6 (IL-6) by bone marrow stroma cells from AA patients is increased ( Kojima et al, 1992 ), whereas the mRNA expression of these factors and of IL-1 and SCF appears to be normal or increased ( Hirayama et al, 1993 ; Stark et al, 1993 ). It has been suggested that the ability to release high amounts of burst stimulating activity or granulocyte colony-stimulating activity may be associated with favourable treatment outcome in AA patients ( Nissen et al, 1985 , 1989; Schrezenmeier et al, 1993 ).
Recently, thrombopoietin (TPO), the ligand of the c-mpl receptor, has been molecularly identified and cloned ( Kaushansky, 1995). In-vitro and in-vivo studies indicated that TPO is the primary regulator of the megakaryocyte lineage. To evaluate the role of TPO in the pathobiology of aplastic anaemia, we have examined TPO serum levels in healthy controls and patients with AA. The present study was designed to answer the following questions: (a) are serum levels of endogenous TPO elevated in AA?, (b) is there a relationship between platelet count and endogenous TPO serum levels?, (c) do TPO serum levels correlate with clinical features of AA patients including duration and severity of aplasia and response to immunosuppressive treatment?
- Top of page
- PATIENTS AND METHODS
We have measured serum TPO levels in 54 patients with acquired AA by a sensitive ELISA method. It has been previously reported that thrombopoietic activity is elevated in serum and urine of patients with AA ( Yang et al, 1986 ). However, it was unclear which factor(s) account for this bioactivity. Beside TPO, IL-6 and IL-11 were major candidate molecules in this regard. We found a 25-fold increase of the TPO serum concentration in AA patients as compared to healthy controls. These results are in agreement with previous studies that have reported on markedly elevated TPO levels in AA patients ( Emmons et al, 1996 ; Ichikawa et al, 1996 ; Kosugi et al, 1996 ; Kunishima et al, 1996 ; Marsh et al, 1996 ; Tahara et al, 1996 ; Usuki et al, 1996 ; Kojima et al, 1997 ). Most of these studies reported on small patient groups that did not allow detailed analysis. Since we have measured TPO levels in a large cohort of AA patients at various stages of their disease, we were able to perform a statistical analysis of correlation between TPO values and platelet counts and other clinical features of AA patients. There were variations in the extent of TPO elevation, but a significant inverse correlation between platelet count and TPO serum concentration in untransfused AA patients. This is in contrast to the report by Marsh et al (1996 ) who could not find a correlation between serum TPO levels and platelet counts. This contradiction might be explained by a difference in the patient groups. Only 8/31 patients reported by Marsh et al (1996 ) had severe or very severe AA and only 4/31 patients were newly diagnosed and untreated. In contrast, most of our patients (37/54) had severe or very severe AA and in the majority of our patients (43/54) TPO was measured before treatment and in 26 patients TPO was measured also in remission after treatment.
One reason for carrying out this study was to estimate whether patients with AA might benefit from treatment with TPO. There is no evidence to suggest a deficiency of TPO which might contribute to the thrombocytopenia in AA. This does not exclude the possibility of benefit from treatment with exogenous c-mpl ligands. For example, G-CSF treatment can induce at least a granulocyte response in a proportion of AA patients ( Kojima et al, 1991 ; Kojima & Matsuyama, 1994) despite significant elevation of endogenous serum G-CSF levels ( Kojima et al, 1996 ). Our data suggest that exogenous mpl-ligands might have to be administered at high pharmacological doses sufficient to induce a significant increase above the already elevated serum TPO levels.
In AA, serum levels of some growth factors correlate with response to immunosuppressive treatment. It has been demonstrated that G-CSF and GM-CSF concentrations of responders to immunosuppressive treatment were significantly higher than those seen in non-responders ( Takahashi et al, 1991 ; Kojima et al, 1996 ). Nissen et al (1989 ) have reported that complete recovery of marrow function after treatment with ATG was associated with high level release of an endogenous granulocyte colony stimulating activity. High serum levels of SCF were associated with favourable clinical outcome, whereas SCF serum levels in patients who died in the course of the disease were considerably decreased ( Wodnar Filipowicz et al, 1993 ). In contrast to these results, TPO levels at diagnosis do not correlate with response to immunosuppression.
We noted that TPO serum levels in AA patients declined in response to haemopoietic recovery. Despite normal or near normal platelet counts in remission patients, TPO did not return to normal levels in remission patients. In the same way, EPO levels in remission patients were still higher with respect to their haematocrit compared with normal controls ( Schrezenmeier et al, 1994 ). These observations indicate a persisting defect of haemopoiesis in remission of AA. The incidence of CFU-E and BFU-E ( Kern et al, 1977 ), as well as the incidence of very primitive progenitor cells in the bone marrow, is still reduced in AA patients who recover from aplasia ( Schrezenmeier et al, 1996 ). The increased TPO serum levels might reflect the reduced haemopoietic potential that persists in remission patients. We hypothesize that elevated levels of TPO are required to maintain normal platelet counts in remission.
The mechanisms that regulate TPO levels are still a matter of debate ( Kaushansky, 1995). It has been proposed that TPO gene expression is constant, and TPO levels are regulated by the binding and metabolism by platelets and megakaryocytes ( Kuter & Rosenberg, 1995; Cohen Solal et al, 1996 ; Stoffel et al, 1996 ). Alternatively, TPO levels could be regulated, at least in part, at the level of transcriptional activation ( McCarty et al, 1995 ; Sungaran et al, 1997 ). Our results in AA patients at diagnosis are compatible with both models. However, persistent elevation of TPO serum levels in remission patients may argue against a simple model of regulation via platelet mpl-receptor mediated uptake and consumption. To explain elevated TPO levels in remission patients one has to assume either increased TPO production or a reduced megakaryocyte mass despite relatively normal platelet counts or reduced consumption by abnormal c-mpl expression or structure.
In summary, our results have shown an elevated TPO serum levels and an inverse correlation between TPO levels and platelet counts in AA patients. A surprising finding is the persistent elevation of TPO serum levels in remission patients. Further studies are necessary to evaluate whether exogenous c-mpl ligands will be of therapeutic value in AA.