High serum level of endostatin in multiple myeloma at diagnosis but not in the plateau phase after treatment.

We investigated the serum concentration of endostatin in 84 patients with multiple myeloma (MM) and in 13 healthy controls. The level of measured anti-angiogenic agent was correlated with the phase and stage of the disease, and most importantly with clinical and laboratory parameters depicting the disease activity (haemoglobin, creatinine, albumins, calcium, M-component, C-reactive protein, beta2-microglobulin, lactate dehydrogenase, stage of bone disease) as well as serum levels of pro-angiogenic cytokines such as vascular endothelial growth factor, hepatocyte growth factor, fibroblast growth factor and transforming growth factor-beta. The median serum level of endostatin in MM patients was 58 ng/ml and was statistically significantly higher than in the control group (median, 40 ng/ml; p=0.015). MM patients in phase I (at diagnosis) had higher levels of endostatin (median, 69 ng/ml) than those in phase II (plateau phase after treatment) (median, 49 pg/ml; p=0.044). We did not find any statistical correlation between the level of endostatin and stage of MM according to the Durie and Salmon system. The serum concentration of endostatin in MM patients with a normal level of albumins was significantly higher than in others with hypoalbuminaemia (median, 62 ng/ml versus 39 ng/ml; p=0.033). Also, patients with a normal value of lactate dehydrogenase had a higher concentration of endostatin than those with values >425 U/l (median, 70 ng/ml versus 39 ng/ml; p=0.019). We did not show any statistical correlation between the concentration of endostatin and level of haemoglobin, creatinine, calcium, C-reactive protein, beta2-microglobulin and stage of bone disease. We failed to find positive or negative correlations between the level of endostatin and vascular endothelial growth factor, hepatocyte growth factor, fibroblast growth factor and transforming growth factor-beta. The concentration of endostatin did not influence the probability of survival in MM patients in our study. In conclusion, our data indicate that endostatin has a higher level in MM patients than in healthy controls. Highest values were stated in active phases of the disease (at presentation and in progression). Different clinical and laboratory parameters generally do not influence the concentration of endostatin (except albumins and lactate dehydrogenase).


Introduction
Angiogenesis is the key process in growth and metastatic spread of neoplasms. In normal tissues blood vessel growth is minimal and angiogenesis is controlled by a balance of angiogenic stimulators and inhibitors. This balance is perturbed in tumours to favour angiogenesis either by overproduction of angiogenesis inducers or by lack of inhibitors. 1 The shift in balance between the production of antiangiogenic and pro-angiogenic factors is a process called angiogenic switch and it marks the onset of neovascularization, which is probably a key step during tumour progression. 1,2 In earlier data it was suggested that the generation of anti-angiogenic compounds in the presence of primary tumour suppresses the growth of distant metastases. 3 Recently, several naturally occurring inhibitors of angiogenesis have been identified. The most potent of them appears to be endostatin, the 20 kDa Cterminal proteolytic fragment of the basement membrane component collagen XVIII, isolated by O'Reilly et al. 4 Endostatin is implicated in the regulation of physiological and pathological angiogenesis. Its precise mechanism of action is still unclear. Endostatin inhibits endothelial cell proliferation in vitro . 5,6 Different investigators have proposed the interference of endostatin with such pro-angiogenic cytokines as vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF) (basic fibroblast growth factor (bFGF)) pathways, 7 induction of endothelial cell apoptosis, 8,9 inhibition of matrix metalloproteases, 10 and inhibition of endothelial cell migration by modulating c-myc levels. 11 Endostatin possesses a prominent heparan sulphate binding domain, although it is not clear whether this domain is required for anti-angiogenic activity. 6,12 According to MacDonald et al., endostatin binds tropomyosin in vitro and to tropomyosinassociated microfilaments in a variety of endothelial cell types. 13 The authors postulate that the interaction of endostatin with tropomyosin results in the disruption of microfilament integrity, leading to the inhibition of cell motility, induction of apoptosis and inhibition of tumour growth. Endostatin levels liberated into culture media and administered exogenously are decreased by hypoxia, the principal cause of angiogenesis. 14 Recent successes in producing active human recombinant endostatin in yeast have allowed testing this protein in an animal model, 15,16 in vitro, 17 and in clinical studies. 18 Endostatin was shown to induce tumour stabilization after chemotherapy or anti-CD20 immunotherapy in a mouse model of human highgrade non-Hodgkin lymphoma. Tumour growth was prevented in endostatin-treated mice as long as the drug was administered. Enostatin administered on days 25Á/29 after tumour re-growth still induced significant tumour regression, whereas chemotherapy and anti-CD20 immunotherapy were not effective. 15 Fujii et al. demonstrated that endostatin inhibits multiple myeloma (MM) growth and tumour burden in the SCID-hu host system. It was reflected by a reduction in serum levels of human monoclonal immunoglobulins. Furthermore, mice bone density was preserved, as demonstrated by X-ray evaluation. 19 This fact may have therapeutical implications. 20 There is evidence that different diseases such as connective tissue diseases, solid tumours, sarcomas and even Down's syndrome play a role in the generation of endogenous anti-angiogenic proteins, including endostatin. 21 Á 26 Bertolini et al. first reported plasma levels of endostatin in heamatologic malignancies (non-Hodgkin lymphoma patients). The authors did not notice any difference between these levels in complete remission and progression of the disease, as well as no correlation being found between endostatin levels and event-free survival. 15 Kay et al. detected mRNA for endostatin in chronic lymphocytic leukemia (CLL) patients. 27 MM cells are known to secrete pro-angiogenic cytokines such as VEGF and FGF in vitro and in vivo . VEGF expression is correlated with the severity of the disease, and circulating VEGF is a prognostic factor for the survival of MM patients. 28 Increased angiogenesis in MM should result in high endostatin value. The role of this most potent inhibitor of angiogenesis is established on in vitro and animal models also in MM, 16,19 but its clinical significance is not elucidated at present.
Recent data demonstrate that MM cells can produce anti-angiogenic as well as pro-angiogenic factors. In the present study we compare endostatin serum level in different phases and stages of MM using the enzyme-linked immunosorbent assay and correlate its concentration with pro-angiogenic cytokines such as VEGF, hepatocyte growth factor (HGF), FGF and transforming growth factor-b (TGF-b).

Patients
The study group comprised 84 patients with MM and 13 healthy controls comparable for median age and gender distribution to the patient population. The patients' characteristics are presented in Table 1.
There were 43 women and 41 men with mean age 66.5 years (range, 32 Á/88 years) in the patient group. The control group comprised seven women and six men with mean age 65.9 years (range, 47 Á/80 years).  29 Renal failure (substage B) was diagnosed when the level of creatinine was /2 mg/ml, and was observed in 18 patients. Advanced bone disease was stated in 29 out of 77 patients with at least multiple lytic lesions found in X-ray examination. MM patients were treated with different chemotherapy protocols adjusted to their age and clinical status. One patient received highdose chemotherapy.

Laboratory examinations
All laboratory tests were performed on the day of blood sampling for endostatin assay. Each patient's blood sample was examined for the following parameters: complete heamogram, serum protein electrophoretogram, immunoglobulins, urea nitrogen, creatinine, calcium, lactate dehydrogenase (LDH), C-reactive protein (CRP), b2-microglobulin, VEGF, HGF, FGF and TGF-b. Serum b2-microglobulin levels were measured by means of a radioimmunoassay kit. Correction for renal function was not calculated. Levels of immunoglobulins, albumin and CRP were measured by nephelometry. Normal values of CRP were considered as below 5.0 mg/l. Bone marrow examination was performed in all cases, and X-ray survey of flat bones in 77 patients.

Cytokine determination
Venous blood samples were collected on the day of physical examination and laboratory tests. They were collected in pyrogen-free tubes, allowed to clot for 1 h at 48C and centrifuged at 2000 x g for 10 min. The obtained sera were aliquoted into separate vials and stored at Á/808C until assayed for endostatin and other cytokines. Sera from each patient were randomly coded and the measurement of cytokines was carried out without the knowledge of the patient's clinical status and laboratory data. The obtained sera were assayed for the concentration of endostatin (also for VEGF, HGF, bFGF and TGF-b1). The commercially available enzyme-linked immunosorbent assay kit produced by Oncogene Research Products (Boston, MA, USA) was used for endostatin detection, and Quantikine, R & D Systems Inc. (Minneapolis, MN, USA) products were used for the determination of the remaining cytokines. All manufacturers' instructions were followed. The adsorbance was read at 492 nm. The appropriate recombinant human cytokine was used to generate the standard curve for each assay.
The concentration of cytokines in the samples was determined by the interpolation from the standard curve. This procedure has been described precisely in our previous works. 30,31 The sensitivity limit for endostatin was 1.95 ng/ml, for VEGF was 5.0 pg/ml, for HGF was 40 pg/ml, for TGF-b1 was 7 pg/ml and for bFGF was 3 pg/ml.

Statistical analysis
Statistical analysis was performed using the procedures of Statistica for Windows package on the IBMcompatible computer.

Results
The serum concentration values of endostatin in MM patients and healthy controls are presented in Table 2. Serum levels of endostatin were detectable in all patients and in 12 out of 13 controls. The mean concentration in the MM group was 72 ng/ml (range, 11Á/402 ng/ml). The mean concentration of this inhibitor in controls was 38 ng/ml (range, 0.0 Á/81 ng/ml; p0/0.015). The mean level of endostatin was highest in phases I and III of MM (mean, 77 ng/ml and 80 ng/ml, respectively). The concentration of this agent in phase II was lower at 48 ng/ml, and there was a statistically significant difference between the level of endostatin in phases I and II of the disease (p 0/0.044).

The stage of MM according to Durie and Salmon
Endostatin in multiple myeloma classification as well as stage of bone disease did not influence the endostatin concentration ( Table 2). The mean concentration of endostatin in MM patients with a normal level of albumins ( /30 g/l) was 78 ng/ml, and was significantly higher than in those with hypoalbuminaemia (46 ng/ml; p0/0.033). Similar results were obtained in correlation of endo-statin with LDH. In patients with a normal concentration of LDH ( B/210 and /425 U/l), the mean level of this factor was significantly higher than in patients with elevated LDH values (77 pg/ml versus 71 pg/ml, p0/0.019). The level of endostatin was lowest in light chain disease patients (mean, 49 ng/ml) in comparison with IgG MM (mean, 75 ng/ml) and IgA MM (mean, 83 ng/ml). The light chain type did not influence the level of this inhibitor. Endostatin concentration did not correlate with other analysed laboratory parameters such as haemoglobin, creatinine, total protein, calcium, CRP and b2-microglobulin (Table 3).
In two patients with non-secretory MM, the endostatin concentrations were 35 and 68 ng/ml.
The endostatin serum level was correlated with some pro-angiogenic cytokines such as VEGF, HGF, FGF and TGF-b (Table 4). We did not find any statistically significant correlation between endostatin and these cytokines (Fig. 1).
The serum concentration of endostatin higher or lower than its median value did not influence the probability of survival of MM patients (Fig. 2).

Disscusion
The aim of our study was to evaluate the serum concentration of an anti-angiogenic agent (endostatin) in patients with MM, and to correlate it with disease activity and advancement, the most important clinical and laboratory parameters as well as known pro-angiogenic cytokines such as VEGF, HGF, FGF and TGF-b. The level of endostatin was detectable in all 84 analysed MM patients and in 12 out of 13 healthy cotrols. Most authors report the detection of endostatin in serum of the vast majority of patients with solid tumours, soft tissue sarcomas, autoimmune diseases and haematologic malignancies. 21 Á 23,26,27,32 In our study, the concentration of endostatin in MM patients was significantly higher than in healthy individuals in comparable age. Higher values of this agent were found in patients with MM before treatment (phase 1 of the disease) than after therapy (phase 2). We found no significant increase of endostatin in advanced stages */ IIIA and IIIB of the disease according to the Salmon and Durie system */ as well as in advanced stages of bone disease. The aggressiveness of MM in our data has no association with the level of endostatin. Higher values of that inhibitor were found in MM patients with a normal level of albumins in comparison with those with hypoalbuminaemia. Similarly, in cases of MM with normal values of LDH, endostatin had higher concentrations than in other patients with decreased or elevated levels of LDH. We found no correlation between the level of endostatin and the   15 These authors also stated lack of correlation between endostatin levels with the grade of tumour malignancy. They also found no difference in endostatin concentration between patients in complete remission and progressive disease. Our data suggest that the serum level of endostatin is lower in cases after therapy and highest before treatment and in progression of MM. Feldman et al. measured the level of endostatin in soft tissue sarcoma patients and stated that the high pre-operative serum level predicted poorer outcome. 21 Lai et al . showed that endostatin has prognostic value in predicting complete remission and overall survival in patients with acute myeloid leukaemia and myelodysplastic syndromes. 32 In our previous study we showed positive correlations between pro-angiogenic cytokines (VEGF and TGF-b, bFGF and TGF-b, VEGF and bFGF) and our data indicated a strong relationship between the serum levels of these cytokines and clinical course, as well as selected laboratory parameters of patients with MM. 28 In this study we failed to prove the correlation of the most potent anti-angiogenic factor, endostatin, with these most clinically relevant proangiogenic cytokines. Probably their interaction is not direct, and further research is needed to elucidate this problem.
We found no influence of endostatin level on the probability of survival in our patients with MM. Previously we stated that VEGF and FGF concentrations have an effect on MM patients' survival. 28 The Endostatin in multiple myeloma latter group was composed of 167 patients whereas the former was two times smaller, and probably these proportions might influence the obtained results, so further studies are needed in this field.
In conclusion, the present study is to our knowledge the first report analysing the serum concentration and clinical significance of endostatin in a large group of MM patients. Our data suggest that serum endostatin levels are elevated in MM patients. The elevated serum values of this agent are especially relevant in patients before treatment and this is caused by the high tumour mass at the beginning of the disease. After treatment the level of this inhibitor falls, and perhaps enables in that way the intensification of angiogenesis during progression of the disease. The plateau phase of MM seems then to be the best time to apply therapy with endostatin.