PFS In Multivariate Analysis Research Articles

Prognostic Impact of Percentage CD34 Expression in Bone Marrow Trephines of Patients with Chronic Myeloid Leukaemia in Chronic Phase Treated with Imatinib.

CD34 is a surface glycophosphoprotein expressed on developmentally early lymphohematopoietic stem and progenitor cells, small-vessel endothelial cells, and embryonic fibroblasts. We studied the prognostic value of presence of CD34 expressing cells in the marrow trephine biopsies of 58 patients with chronic myeloid leukemia in chronic phase treated with imatinib (21 newly diagnosed and 37 IFN failures). Briefly, patients received 400 mg daily and the dose was then adjusted according to tolerance and response. Bone marrow examinations were performed before starting therapy with imatinib and then every 3–6 months. All patients achieved complete haematological remission and 48% achieved complete cytogenetic remission. Chronic, accelerated and blastic phases were defined as described by Kantarjian 1988. Paraffin embedded sections were stained with the monoclonal antibody QBEND10. A minimum of 500 nucleated cells were counted per trephine section. Results were expressed as percentage of CD34+ vs all nucleated marrow cells. The expression of CD34 by the endothelial cells was used as an internal control. The median CD34 percentage before the onset of the imatinib therapy was 0.8% (25 and 75 percentiles 0.8 and 2%). We performed univariate and multivariate analysis for PFS with variables defined at the time of starting imatinib. Sokal risk group (defined at diagnosis), presence or absence of additional cytogenetic abnormalities, percentage of blasts in the bone marrow aspirate and the percentage of CD34+ cells in the trephine (considered as a continuos variable) were significant in the univariate analysis. Only CD34 expression and Sokal risk group were significant in the multivariate model. Sokal risk group lost significance when the percentage of CD34 cells was categorized as <=2% and >2% (RR for PFS=11.7, 95 CI=1.9–70.4, p=0.007; RR for survival = 10.5, 95CI= 1.01–80.3, p=0.05). The 3 years PFS for patients with a CD34 percentage <= 2% before the onset of the imatinib therapy (n= 44) was 77% while for patients with a CD34 >2% (n=14) it was 36% (p=0.006), similarly the three year overall-survival was 80 and 45% respectively (p=0.009). The median value of CD34 in the first bone marrow examination performed after the start of imatinib (typically performed at 3–6 months) was 0.8 (25 and 75 percentiles 0.8 and 1.5%). The degree of reduction from the baseline level achieved in the first bone marrow examination had favourable prognostic significance (as a continuos variable) with a RR for PFS of 0.88, 95CI 0.53–0.93 (p= 0.006) and RR for overall survival of 0.93, 95CI 0.19–0.998 (p=0.05). During follow up, an increment of 2 percentage points over the pre therapy CD34 value in patients in otherwise complete haematological remission had an adverse prognostic implication with a RR for PFS of 10.2 95 CI 2.1–17, (p=0.006) and a RR for survival of 6.8 95 CI 1.01–101-(p=0.05). We conclude that the use of CD34 staining in bone marrow trephines is a valuable tool for both diagnosis and monitoring in patients with CML treated with imatinib.

Relationship of PS2 with response to tamoxifen therapy in patients with recurrent breast cancer.

PS2, an oestrogen-inducible protein, was measured in the cytosol of 230 primary tumours from patients who were subjected to first-line tamoxifen therapy for advanced disease without prior adjuvant therapy with tamoxifen. PS2 correlated positively with oestrogen receptor (ER, P < 0.01) and progesterone receptor content (PgR, P < 0.001), and with the length of progression-free survival (PFS, P = 0.05). Although not statistically significant, higher levels of PS2 (> or = 10 ng mg-1 protein) were also associated with increased probability of response to tamoxifen treatment and a longer total post-relapse survival (PRS). ER, PgR, menopausal status, site of disease and prior adjuvant chemotherapy were all associated with response to tamoxifen therapy and with PFS. In multivariate analysis for PFS, low levels of ER and PgR, visceral metastasis, a disease-free interval of less than 1 year and prior adjuvant chemotherapy were all significantly associated with an increased probability of a rapid disease progression after start of tamoxifen therapy. In the subset of 83 tumours with intermediate levels of ER and PgR (both > or = 10, but not both > or = 75 fmol mg-1 protein), PS2 was positively related with the length of PFS (P < 0.01) and PRS (P < 0.05). PS2 remained the strongest factor in multivariate analysis for PFS (P < 0.01) in this ER/PgR intermediate subgroup, but was not of predictive value in univariate or multivariate analysis for both PFS and PRS in tumours classified as ER/PgR low or high (> or = 75 fmol mg-1 protein). It is concluded that PS2 status may be used as a parameter, additional to ER and PgR, for better refinement of prediction of response to tamoxifen treatment in advanced breast cancer patients especially with intermediate ER/PgR levels in their primary tumour.