Tumour prothymosin alpha content, a potential prognostic marker for primary breast cancer

In a previous report we suggested that the estimation of prothymosin α (PTA) levels in primary breast tumours might be used to identify breast cancer patients at high risk for distant metastasis (Dominguez F et al (1993) Eur J Cancer29A: 893–897). Here the role of tumour PTA levels as predictor was investigated with respect to both disease-free survival (DFS) and survival. Tumours were obtained from a series of 210 consecutive female patients with ductal carcinoma who underwent surgery at the Hospital Xeral de Galicia (Santiago de Compostela, Spain). Characteristics including PTA tumour levels, number of positive axillary nodes, patient's age at surgery and tumour histological grade were significantly associated with DFS and survival, as determined by univariate analysis. Patients with tumours with low or moderate PTA levels demonstrated a statistically decreased rate of tumour recurrence and a statistically significant increased overall survival in comparison with those whose tumours had high PTA levels. Patient's relative risk of dying was 2.1 times greater for tumours with high PTA levels than for those tumours with low or moderate PTA levels. In conclusion, these data support the hypothesis that tumour high PTA levels is associated with a worse outcome. © 2000 Cancer Research Campaign

. Immunohistochemical studies have shown that PTA is expressed in proliferating but not in quiescent cells in all tissues studied so far (Conteas et al, 1990;Fraga et al, 1993;Gallego et al, 1992;Garcia et al, 1994;Roson et al, 1990aRoson et al, , 1990bRoson et al, , 1993. PTA is expressed in various human tumours of different origins (reviewed in Dominguez et al, 1997), supporting the idea that PTA expression is required for tumour growth. Certainly, when PTA expression is blocked the proliferation of myeloma cells is inhibited (Sburlati et al, 1991). Non-transformed NIH 3T3 cells shows a similar behaviour , indicating that PTA must play an essential role in normal proliferation that is preserved in tumorous cells. Moreover, when PTA is overexpressed the G1 phase of the cell cycle is shortened  as it also happens with other factors that play a major role in cell proliferation. This finding indicates that PTA must be a limiting factor necessary for the progression of the cell cycle. On the other hand, the analysis of the intracellular signalling pathways governing PTA expression point to the fact that PTA expression is an event occurring downstream in several mitogenic pathways. PTA expression seems to be under the direct control of MYC (Eilers et al, 1991); however, other reports did not confirm this finding (Mol et al, 1995). Vareli et al (1996) have found that the transcription factor E2F activates a reporter gene under the PTA promoter, indicating that E2F could directly control PTA expression. In mammary tumours arising in c-myc, c-neu and v-ras transgenic mice we found that there is a differential regulation of PTA with respect to other putative c-myc target genes, indicating that regulation of PTA in these tumours is complex and depends on more than a single activated oncogene (Loidi et al, 1999).
PTA physiological function remains unknown. It is a nuclear protein (Gomez and Segade, 1988;Roson et al, 1990a) that is present throughout the cell cycle (Zalvide et al, 1992); it has been shown that it binds histones in vitro (Papamarcaki and Tsolas, 1994;Diaz et al, 1996) and has been proposed to affect the chromatin state . Recently, we have found that PTA modulates the activity of calcium/calmodulindependent kinases (CaM-kinase) II and III in a cell cycle-dependent manner (Vega et al, 1998). These enzymes play various roles during the cell cycle. CaM-kinase III is necessary for the phosphorylation of EF-2 during mitosis (Ryazanov, 1987;Celis et al, 1990) and CaM-kinase II affects the G 1 progression as well as the transition from G 2 to mitosis (Lu and Means, 1993).
We had previously shown in 71 patients with invasive ductal carcinomas, that PTA levels as assayed by a radioimmunoassay that detects thymosin α 1 , the NH 2 -terminal fragment of PTA, were significantly greater in tumour samples than in normal breast tissue. PTA levels were correlated with: (a) the number of positive axillary lymph nodes, and (b) the percentage of tumour cells in the S phase plus in the G2/M phase as assessed by flow cytometry. Of special relevance was to find that PTA levels might be used to identify patients at high risk for distant metastasis . Recently, others independently confirmed these findings (Tsitsilonis et al, 1998). In the present report, the role of tumour PTA content as predictor was investigated with respect to both disease-free survival and overall survival.
Tumour prothymosin alpha content, a potential prognostic marker for primary breast cancer

Patients
High PTA levels in tumours have been suggested to be a predictor of a poor prognosis . A prospective study was designed to test this hypothesis. Tumours were obtained from a series of 210 consecutive female patients with invasive ductal carcinoma who underwent definitive surgery at the Hospital General de Galicia (Santiago de Compostela, Spain) since 1989. Patients with a second malignant neoplasm were excluded.

Specimen analysis
Each tissue block was sectioned for histological analysis as previously described . Nodal involvement was histologically determined. Tumour grading was done according to the Bloom and Richardson method as described by Freedman et al (1979). Ploidy analysis was done by DNA flow cytometry. Tumour samples were processed as previously described .

PTA radioimmunoassay
PTA radioimmunoassay (RIA) was done as previously published . Briefly, slices (about 200 mg) of tumour tissues obtained during surgery were homogenized with a Polytron homogenizer in phosphate-buffered saline (PBS)-EDTA, centrifuged at 14 000 g for 15 min and the supernatant analysed for PTA and protein. The radiolabelled ligand was I 125 -Tyr 0thymosin α 1 and synthetic thymosin α 1 was employed to standardize the assay. Tissue homogenates were prepared at room temperature in order to convert tissue PTA to thymosin α 1 ; therefore, the results are expressed as thymosin α 1 equivalents .

Patient outcome end points
The role of tumour PTA levels as predictor was investigated with respect to both disease-free survival (DFS) and survival; as secondary end point, we also examined distant disease-free survival (DDFS). DFS was defined as the time from surgery prior to recurrent breast cancer or death from any cause. Survival was defined as the time from surgery to death from any cause. DDFS was defined as the time from surgery to tumour recurrence at distant sites.

Statistical analysis
Logistic regression analysis, Kaplan-Meier survival analysis and the Cox regression model were used. All data were analysed with the use of the SPSS statistical package (SPSS Inc., Chicago, IL USA). Table 1 summarizes the clinical and biological data for all 210 patients enroled in this study. Patients were divided in three groups according to tumour PTA content as follows: low level, PTA levels ranging from 0 to 20 ng mg -1 of protein (34 patients); moderate level, PTA levels ranging from 20 to 100 ng mg -1 of protein (122 patients); and high level, PTA levels greater than 100 ng mg -1 of protein (54 patients). A significant Pearson correlation was found between PTA levels and number of positive lymph nodes (twosided P < 0.05), histological grade (two-sided P < 0.01), DNA ploidy (two-sided P < 0.01) and patient's age at surgery (P = 0.05). By means of a logistic regression model relating PTA levels (high vs low and moderate) with all co-variates, we found that the histological grade was significantly associated with PTA high levels (P < 0.001). On the other hand, PTA low levels (low vs moderate and high) were significantly associated with the absence of positive nodes (P < 0.03) and with the histological grade (P < 0.03). Table 2 shows estimates of the number of patients at risk and the hazard rate for the first 5 years. Figure 1 shows Kaplan-Meier plots for overall survival, DFS and DDFS for PTA levels. Patients with low levels of PTA showed a superior behaviour to patients with moderate or high PTA levels on the basis of each end point. Characteristics including PTA tumour levels, number of positive axillary nodes, patient's age at surgery and tumour histological grade were significantly associated with DFS and survival, as determined by univariate analysis (Table 3). The same variables plus tumour size were significantly associated with overall survival (Table 3). The multivariate analysis of DFS and survival is shown in Table 4. The effect of PTA levels, number of positive lymph nodes and the histological grade is significant using the multivariate model while tumour size and the age of the patient loses the significance found in the univariate analysis. As seen in Table 4 the risk of patients with tumours with high PTA levels relative to tumours with low to moderate PTA levels is high for both DFS and survival. In Figure 2, the data is stratified according to other covariates of clinical relevance. The risk of patients with tumours with high PTA levels relative to tumours with low or moderate PTA levels is apparent among patients with   negative axillary lymph nodes. This is also true for other covariates shown in Figure 2 but the number of patients in each stratum is low; therefore, present data should be further validated. The central hypothesis of this investigation was to see if patients with tumours with high proliferating activity, as seen by estimating tumour PTA levels, have a worse outcome. Bonferronis joint multivariate analysis was used to test the effect of high PTA levels on patient's outcome using as primary end point survival, DFS and DDFS and the following co-variates: tumour size (> 2 cm vs ≤2 cm), positive lymph nodes (none vs ≥ 1), histological grade (poorly differentiated vs well and moderately differentiated), and age at surgery (> 50 years vs ≤ 50 years). The hypothesis was significant for the three end points, survival (P = 0.000), DFS (P = 0.04149) and DDFS (P = 0.0001).

DISCUSSION
The goal of the current study was to determine if tumours with high PTA levels were associated with a worse outcome. We have analysed the overall survival of patients with breast tumours that underwent definitive surgery since the beginning of this study in 1989. PTA levels were assayed by a radioimmunoassay that detects thymosin α 1 , the NH 2 -terminal fragment of PTA . We present new evidences indicating that estimation of PTA levels in the tumour is a potential prognostic marker for primary breast cancer. A statistically significant association between PTA levels and other relevant clinical factors as the number of positive lymph nodes, the histological grade, DNA ploidy and the age at surgery was observed. Patients with tumours with low or moderate PTA levels demonstrated a statistically significant decreased rate of tumour recurrence and a statistically significant increased overall survival in comparison with those patients whose tumours had high PTA levels. Multivariate Cox  proportional analyses showed that those patients with high PTA levels have 2.1 times the risk of dying than patients with low or moderate PTA levels. Moreover, patients with high PTA levels had a greater risk (1.42 times) of regional or distant tumour recurrence than patients with low or moderate PTA levels. However, these numbers must be taken with caution since, as always, the model will fit the sample from which it is estimated better than it will fit the population from which the sample is selected. Bonferronis joint multivariate analysis was used to test the effect of high PTA levels on patient's outcome using as primary end point survival, DFS and DDFS. PTA levels had significant effect on the three end points. The fact that the three end points were significantly affected provides a stronger support to the hypothesis that patients with tumours with high PTA levels have a worse outcome. These findings are not surprising since cell proliferation indices have been shown to be prognosis-related (Freedman et al, 1979;Sigurdsson et al, 1990). By means of immunohistochemical techniques we found that PTA and the proliferating cell nuclear antigen (PCNA) had a similar pattern of expression both in normal tissues  and hepatocellular carcinomas . PTA is expressed in the late G1, S, G2 and M phases of the cell cycle (Zalvide et al, 1992). Interestingly, the fraction of cells in S and G2 phases is more closely associated with the clinical outcome than the S phase fraction alone (Witzig et al, 1994).
Recently, Rudolph et al, using the proliferation marker Ki-S2 concluded that the prognosis is best indicated by the percentage of cells in S through M phases of the cell cycle (Rudolph et al, 1999).
In agreement with this we found in breast tumours that PTA expression was correlated with the percentage of cells in S and G2/M phases fraction but not with the S phase alone . Here we presented the estimation of tumour proliferation rate as assessed using a RIA for PTA. PTA RIA provides an objective measure of tumour proliferation activity. A major trend in the biologic research in breast cancer is focused in factors that predict benefit from therapy. At present, it is not possible to identify those patients who are cured without adjuvant chemotherapy or those patients who will relapse in spite of adjuvant chemotherapy, with resistant disease (Powles, 1997). Tumour PTA content was investigated as predictor with respect to DFS, DDFS and overall survival. Patients received surgical, radiation and other adjuvant therapies following an established pattern, irrespectively of tumour PTA content. Patients with tumours with high PTA levels had a worst outcome irrespective of the adjuvant therapy received. On the other hand, tumours with low levels of PTA had a good prognosis. Further studies with a larger series of patients are now needed to establish the role of tumour PTA as a predictor of the therapeutic response. However, it is interesting to point out that PTA antisense oligonucleotides that blocks PTA expression promoted apoptosis in HL60 cells (Rodriguez et al, 1999). If this finding is confirmed it is tempting to speculate that high PTA levels could have a protective effect against apoptosis induced by chemotherapeutic agents.
In summary, this paper reaffirms two major points in breast cancer research: first, the importance that tumour's proliferation rates have in patient's outcome and, second, the benefit of using quantitative assays to estimate proliferation rates. We propose that PTA content could be used as a predictor of the potential malignancy of breast tumours that might help to identify patients at high risk of dying. Our study represents a preliminary attempt to correlate tumour PTA levels and overall mortality in breast cancers.