Epithelial ovarian cancer is the leading cause of death from gynaecologic malignancies in western countries [26
]. The tumor stage at time of diagnosis and the postoperative residual tumor mass are important prognostic factors and are unequivocally related to overall survival [26
]. Other prognostic factors are identified mostly in small series and are the source of controversial discussion in the relevant literature.
Serum TTR is traditionally a valid marker for nutritional status in general and in cancer patients it has gained considerable interest with regard to the use as an early diagnostic marker in ovarian cancer [21
]. As depletion of nutritional reserves and a subsequent significant weight loss can lead to an increased risk of morbidity, reduced chemotherapy response, and shorter survival in patients with cancer, TTR is a valid prognostic marker [28
]. Interestingly however, TTR and RBP levels of serum are affected not only by the nutritional status of the individual but are also reduced during the acute phase response associated with inflammation [29
]. Additionally to quantitative aspects, the TTR molecule in serum exists in numerous variants due either to genetic differences or because of modification on one readily accessible cystein within the molecule. The microheterogeneity is affected by different metabolic aspects such as oxidative stress or homocystein levels [20
]. Nothing is known however with regard to possible variation due to metabolic alterations in cancer.
The results of the present study confirm previous results for cancer patients in general and especially for patients with ovarian cancer, regarding the greatly reduced serum levels of TTR and RBP [31
]. Interestingly however, the intensity of the disease has no significant influence on serum levels, indicating that it is a general phenomenon possibly associated with cancer induced cachexia which is already present at early stages. For drawing a general conclusion, this group (stage I/II) was too small in sample size. On the other hand, when the differing inflammatory statuses were considered, obvious differences were observed between the cancer patients for serum levels of TTR and RBP. Using 40 mg/l as cut-off for C-reactive protein (CRP) TTR and RBP serum levels were reduced in those individuals with increased CRP values. This clearly supports observations showing that the inflammation status greatly reduces TTR and RBP serum levels as a consequence of a reduced synthesis of this negative acute-phase protein in the liver [5
Using immunological procedures we were recently able to show the presence of TTR in ascitic fluid from women with ovarian cancer, however no quantitative data, especially with regard to cancer stages, is as yet available [32
]. In accordance to our previous semi-quantitative study, TTR in ascitic fluid was more than 100-fold lower when compared to its serum levels. This ratio is much lower in comparison to the one observed for RBP (Tab. ). Based on the difference between their molecular masses, 55 kDa for the hetero-tetramer TTR and 21 kDa for RBP, one would expect a different ascites/serum ratio, as an inverse correlation exists between ascites/serum ratio and the mean of the molecular weight of various proteins [5
]. From this observation one could assume that RBP and TTR are not transferred individually but rather as the complex usually present in serum [33
]. In general, results support the hypothesis that the concentration of TTR and RBP in ascitic fluid is the result of a passive transfer from serum into the ascitic fluid. The accumulation of these and other serum constituents is mainly attributed to the increased capillary permeability caused by an increase in permeability-inducing factors such as the vascular endothelial growth factor (VEGF) [25
These observations and the fact that no obvious differences in microheterogeneity between TTR from serum and ascitic fluid can be observed, both with regard to the known modification at the Cys10 and the recently described truncated form, it can be assumed that all TTR in ascitic fluid originates through a passive transfer from serum. This is further supported by the observation that the tumor site itself does not express any immunoreactive TTR. It can not be excluded however, that the tumor site or components in the ascitic fluid may have proteolytic properties that possibly result in the unobserved modifications of TTR or other proteins.
With regard to the microheterogeneity of TTR in serum and ascitic fluid, the results support and confirm previous studies undertaken by us and others with regard to molecular variants of TTR in serum [17
]. As in these studies, TTR in serum and ascitic fluid was dominant in four variants. The 118 Da larger variant is the S-cysteinylated form of the native TTR whilst the signal at 14,062 Da can be attributed to the S-glutathionylated TTR form [30
]. As TTR contains only one cysteine residue (Cys10
), the adduct must result when the Cys10
residue forms a mixed disulfide with the amino acid cysteine, the dipeptide cysteinylglycine or the tripeptid glutathione. The shift in the mass spectrum of TTR variants towards the native TTR molecule mass in serum and ascites fluid after treatment with DTT indicates that the adducts are formed via the disulfide linkage at Cys10
. In addition to this, we confirmed in serum of healthy and affected women as well as in ascitic fluid the presence of a smaller immunoreactive form of TTR with a molecular mass of 12,830 Da, which was recently identified as a truncated form of TTR lacking the NH2
-terminal 10 amino acids [21
]. Its presence in both serum and ascitic fluid supports once again the idea of passive transfer from serum into ascitic fluid during its accumulation.