Anna Tamburini

The kinetics of reduction of minimal residual disease impacts on duration of response and survival of patients with acute myeloid leukemia

We assessed by multiparametric flow cytometry the levels of minimal residual disease (MRD) in 100 adult patients with acute myelogenous leukemia (AML) achieving complete remission after intensive chemotherapy. The aim of the study was to determine the optimal threshold, in terms of residual leukemic cells, and the time point of choice, that is, post-induction (post-Ind) or post-consolidation (post-Cons), able to better predict outcome. By applying the maximally selected log-rank statistics, the threshold discriminating MRD− from MRD+ cases was set at 3.5 × 10−4 residual leukemic cells, a level that allowed the identification of distinct subgroups of patients, both at post-Ind and post-Cons time points. Post-Cons MRD− patients had a superior outcome in terms of relapse rate, overall survival (OS) and relapse-free survival (RFS) (P<0.001, for all comparisons), regardless of the MRD status after induction. In particular, patients entering MRD negativity only after consolidation showed the same outcome as those achieving early negativity after induction. Multivariate analysis, including karyotype, age, MDR1 phenotype, post-Ind and post-Cons MRD levels, indicated that the post-Cons MRD status independently affected relapse rate, OS and RFS (P<0.001, for all comparisons). In conclusion: (1) the threshold of 3.5 × 10−4 is valid in discriminating risk categories in adult AML and (2) post-Cons MRD assessment is critical to predict disease outcome.

Prognostic relevance of the expression of Tdt and CD7 in 335 cases of acute myeloid leukemia

We have analyzed the expression of Tdt and CD7 in 335 cases of unequivocal acute myeloid leukemia (AML). Tdt was expressed in 80 (25%) of 321 evaluable cases. Twenty-six of 77 (34%) Tdt+ patients assessable for response, entered complete remission (CR) vs 121 of 209 (58%) Tdt− cases (P < 0.001). cd7 was expressed in 102 of 332 (30%) evaluable cases; 37 of 93 assessable (40%) cd7+ patients attained a CR as compared to 114/204 (56%) CD7− (P = 0.013). Duration of survival was significantly shorter for patients with CD7+ or Tdt+ AML (P = 0.006 and 0.001, respectively). In a multivariate analysis, Tdt was found to significantly adverse achievement of CR (P = 0.018), while CD7 affected duration of CR (P = 0.037). Overall the expression of either Tdt or CD7 correlated with a relatively high expression of CD34 (P < 0.001), gp-170 (P = 0.003), lymphoid antigens (LyAg) (P < 0.001), t(9;22) or anomalies of chromosome 5/7 (P < 0.001). finally, we pooled the patients into four phenotypic classes, according to the presence of tdt, cd7 or both: [tdt−CD7−], [Tdt+CD7−], [Tdt−CD7+] and [Tdt+CD7+]. The category [Tdt+CD7+] was characterized by a more unfavorable outcome as suggested by a lower rate of CR (P < 0.001) and a shorter duration of survival as compared to cases [tdt−CD7−], [Tdt+CD7−] and [Tdt−CD7+] (P = 0.002). This figure is consistent with the frequent convergence in the subset [Tdt+CD7+] of GP-170 positivity (P = 0.003), translocation t(9;22), anomalies of chromosome 5 and/or 7 (P < 0.001) and signs of lineage infidelity (deviant expression of lymphoid antigens) (P < 0.001). we conclude that the expression of tdt or cd7 is associated with an unfavorable outcome and that the combination of both defines a clinical subset with a poorer prognosis due to the significantly higher association with mdr phenotype, and ‘poor prognostic’ chromosomal abnormalities.

P-glycoprotein and terminal transferase expression identify prognostic subsets within cytogenetic risk classes in acute myeloid leukemia

Clinical and biological features were assessed in 204 consecutive de novo adult acute myeloid leukemia (AML) patients who received intensive chemotherapy regimens. Multiparameter flow cytometric assays both of the multidrug resistance (MDR-1)-associated P-glycoprotein (PGP) using the UIC2 monoclonal antibody (MoAb), and of terminal transferase (TdT) were performed. Cytogenetic findings were obtained from 196 patients with high resolution banding. At onset, UIC2 and TdT positivities were detected in 58.5% and 24% of cases, respectively. There were strict correlations either between UIC2 negativity and FAB M3 or between TdT and FAB M0-M1 (P = 0.001 and < 0.0001, respectively). On the other hand, age was significantly associated with cytogenetic risk classes (P < 0.0001). CD34 positivity was highly correlated with TdT expression (P < 0.0001). Moreover, CD7 and CD11b were significantly represented in UIC2+ subset (P < 0.0001). Rhodamine 123 (Rh 123) efflux was significantly higher in 75 UIC2 positive patients compared to 65 UIC2 negative ones (P < 0.001). As regards to cytogenetics, TdT positivity was strongly related either to t(9;22) or single/associated anomalies of chromosome 7; on the other hand, most or all cases with t(8;21) or t(15;17) were UIC2 or TdT negative, respectively. The rate of first complete remission (CR) differed both between UIC2+ and UIC2- cases and between TdT+ and TdT- ones (40% versus 72%, P < 0.001; and 36% versus 61%, P = 0.001, respectively). The survival rates (Kaplan-Meier method) were significantly shorter either in UIC2+ or in TdT+ patients (P = 0.005 and = 0.011, respectively). UIC2 and TdT negative cases showed longer remission duration (P = 0.03 and = 0.22, respectively). The additional effect of UIC2 and TdT on prognosis allowed us to identify two subsets of patients, the first [UIC2- TdT-] at better and the second [UIC2+ TdT+] at worse clinical outcome compared to single UIC2 and TdT cases, concerning CR (P < 0.001), survival (P < 0.0001) and CR duration (P = 0.007). The combinations [UIC2+ TdT-] and [UIC2- TdT+] showed an intermediate clinical course. A strong difference was found between poor risk and intermediate/favorable risk cytogenetic classes with regard to CR rate (P < 0.0001), overall survival and CR duration (P < 0.001). Nevertheless, within the poor risk class, UIC2 positivity was able to identify patients at worst prognosis with regard to CR (P = 0.005), survival (P = 0.02) and CR duration (P = 0.015). On the other hand, UIC2 and TdT negativity allowed us to distinguish patients with longer survival (P = 0.012 and = 0.04, respectively) and CR duration (P = 0.04 and = 0.025, respectively) within the intermediate/favorable risk class. The independent prognostic value of UIC2, TdT and cytogenetic risk classes was confirmed in multivariate analysis. These results suggest that PGP and TdT expressions, together with cytogenetic findings, may represent a basic predictor of chemotherapeutic failure in AML.

P-glycoprotein and BCL-2 levels predict outcome in adult acute lymphoblastic leukaemia

Summary. Concurrent resistance mechanisms, such as P‐glycoprotein (PGP) and bcl‐2, may contribute to a worse outcome in adult acute lymphoblastic leukaemia (ALL). Between 1990 and 2000, we analysed PGP and bcl‐2 by flow cytometry, using two anti‐PGP (C219 and JSB‐1) monoclonal antibodies (mAbs) and an anti‐bcl‐2 mAb in 115 de novo adult ALL patients. Both a longer overall survival (OS) and longer disease‐free survival (DFS) were observed in PGP‐negative patients (23%vs 0% at 3 years, P = 0·011 and 29%vs 0% at 2 years, P = 0·006 for C219 respectively; 42%vs 0% at 1·5 years, P = 0·004 and 53%vs 0% at 8·5 months, P = 0·00006 for JSB‐1 respectively). Bcl‐2 positivity was associated with a significantly higher complete remission rate (90%vs 66%, P = 0·01). Moreover, in 69 patients not presenting with either t(9;22) or B‐mature immunophenotype, PGP negativity (JSB‐1) maintained its significant favourable prognostic impact with regard to OS (41%vs 0% at 1·5 years, P = 0·009) and DFS (83%vs 0% at 6 months, P = 0·0005). Importantly, within a subset of 62 patients with normal (n = 31) or unknown (n = 31) karyotype, PGP (JSB‐1)‐negative patients showed both a significantly longer OS and DFS (63%vs 0% at 1·4 years, P = 0·018 and 84%vs 0% at 6 months, P = 0·001 respectively). In multivariate analysis, JSB‐1 (P = 0·008) and cytogenetics (P = 0·02) were found to be independent prognostic factors with regard to DFS. Therefore, in adult ALL, PGP and bcl‐2 represent sensitive indicators of clinical outcome, and potential targets of novel molecules aimed at overcoming chemoresistance and recurrent relapses.

Enumeration of CD34+ hematopoietic progenitor cells for clinical transplantation: comparison of three different methods.

Three different methods for determination of CD34+ cells in G-CSF-mobilized peripheral blood were compared. The methods were: the Milan/Mulhouse protocol, the ISHAGE guidelines for CD34+ cells enumeration and our own protocol. The procedure we have adopted is essentially a Milan/Mulhouse protocol-derived methodology combined with a multiparametric approach using the PAINT-A-GATE software analysis program. The samples were collected from 70 patients affected by acute leukemia, non-Hodgkin’s lymphoma, Hodgkin’s lymphoma, myeloma and breast cancer who were scheduled to receive autologous PBSC transplantation. PBSC collection was performed following mobilization with subcutaneous G-CSF at 5–10 μg/kg/day. A minimum target of 2 × 106/kg CD34+ cells was considered an acceptable harvest to ensure a safe transplant. On average, three aphereses per patient were performed and a total of 204 apheresis samples were analyzed. Regression analysis of the percentage and absolute number of CD34+ cells, as calculated with each method, achieved an excellent correlation in spite of methodological differences. In fact, both CD34+dim and CD34+CD45− events were included in our gating strategy. In the setting of a triple staining associating CD34, CD38 and CD45, we identified a variable fraction of CD34+CD38+CD45− cells which would be otherwise undetected due to its CD45 negativity. To this end, we used a new technology referred to as laser-scanning cytometry (LSC) which allowed the isolation and morphological identification of CD34+CD45− cells. By comparing CD34+CD45+ and CD34+CD45− cells, we found that they share a common morphology, thus confirming the hypothesis that the latter are to be considered for CD34+ cell calculation. The median number of CD34+ cells/kg, as calculated by the three methods, was: 4.79 × 106/kg (range 1–570) for the Milan/Mulhouse protocol, 3.9 × 106/kg (range 0.8–498) for the ISHAGE one, and 5.17 × 106/kg (range 2–599) for our protocol. The median time to ANC and PLT engraftment was 11 (range 9–24) and 20 (range 10–70) days, respectively. Our protocol achieved the best correlation between CD34+ cells/kg and time to ANC/PLT recovery according to the Spearman’s rank test (r = −40 and P < 0.015 for anc, r= −46 and P = 0.005 for PLT). We conclude that (1) CD45 does not appear the ideal partner of HPCA-2 for determination of hematopoietic progenitors in mobilized peripheral blood; and (2) for clinical application, a single staining with 8G12 appears simple, reliable and feasible when rigorous procedures for sample preparation and acquisition are followed and an adequate software for multiparametric analysis is available.

CD90/Thy-1 is preferentially expressed on blast cells of high risk acute myeloid leukaemias*

Different transformation mechanisms have been proposed for elderly acute myeloid leukaemia (AML) and secondary AML (sAML) when compared with de novo AML or AML of younger patients. However, little is known regarding differences in the immunophenotypic profile of blast cells in these diseases. We systematically analysed, by flow cytometry, 148 patients affected by de novo (100 cases) or sAML (48 cases). By defining a cut‐off level of 20% of CD34+ cells co‐expressing CD90, the frequency of CD90+ cases was higher in sAML (40%) versusde novo AML (6%, P < 0·001), elderly AML (>60 years) (24%) versus AML of younger patients (10%, P = 0·010) and poor‐ versus good‐risk karyotypes (according to the Medical Research Council classification, P < 0·001). The correlation between CD90 expression, sAML and unfavourable karyotypes was confirmed by analysing the subset of CD34+ AML cases alone (91/148). Consistently, univariate analysis showed that expression of CD90 was statistically relevant in predicting a shorter survival in CD90+ AML patients (P = 0·042). Our results, demonstrating CD90 expression in AML with unfavourable clinical and biological features, suggest an origin of these diseases from a CD90‐expressing haemopoietic progenitor and indicate the use of CD90 as an additional marker of prognostic value in AML.

P-glycoprotein expression in de novo acute myeloid leukemia

Detection of the multidrug resistance P-glycoprotein (PGP) phenotype was performed at the time of diagnosis in 223 patients with acute myeloid leukemia (AML) by flow cytometry using C219 Monoclonal Antibody (MoAb). On the other hand, JSB1 MoAb was tested in 173 of these samples. At onset, PGP was detected in 57.4% of cases with C219 and 75.9% of cases with JSB1. There was no correlation between PGP expression and sex, age, marrow blast percentage or extramedullary disease. On the contrary, strict correlations were noted either between C219 negativity and FAB M3 subtype or between C219 positivity and FAB M5 group (P = 0.003). Significant correlation was found between PGP phenotype and CD7, as 143 of 223 samples had similar patterns of staining with C219 (P < 0.0001). Finally, there was a close relationship between C219 and JSB1 positivity: all the C219+ cases were positive for JSB1 (P < 0.0001). Concerning the karyotype, most patients with monosomy or del (7) were MDR positive; on the other hand, most patients with t(8;21) or t(15;17) were MDR negative. Rh123 accumulation studies showed a significant decrease of mean fluorescence intensities both in C219 and in JSB1 positive cases in comparison with PGP negative ones (P < 0.001). A significant decrease of remission induction rates (CR) was highlighted both between C219+ and C219- and between JSB1+ and JSB1- cases (32.1% v 62.1% and 32.6% v 73.8%, respectively, with P < 0.0001). The overall survival and the remission duration (CCR) were significantly shorter both in C219+ and in JSB1+ patients with no relationship to age. Furthermore, a higher rate of early relapses was noted among MDR+ when compared with MDR- patients both for C219+ and JSB1+ cases. The combination (C219- JSB1+) identified a subset of patients with an intermediate prognosis. On multivariate analysis, C219 and JSB1 were confirmed to be independent prognostic factors for achievement of CR, overall survival and CCR. In conclusion, the assessment of MDR phenotype by flow cytometry is a crucial prognostic factor of treatment outcome in AML.

Clinical Relevance of Minimal Residual Disease Detection in Adult Acute Myeloid Leukemia

We have used flow cytometry to quantify minimal residual disease (MRD) in 63 patients with acute myeloid leukemia (AML). No significant correlation was found between the level of MRD after induction and disease outcome. After consolidation, a threshold of 3.5 x 10(-4) residual leukemic cells divided the 57 evaluable patients into two distinct groups: the MRDCons(+) and the MRDCons(-) group, with a relapse rate of 81% (22/27) and 27% (8/30), respectively (p = 0.000035). Although not correlated with prognosis, the level of MRD after induction course affected the degree of cytoreduction achieved with consolidation. In fact, the patients who entered a MRDCons(-) status had a median number of leukemic residual cells of 1.8 x 10(-4) after induction; at the same stage, the bone marrow of patients who were in a MRDCons(+) condition harbored a median level of 1.7 x 10(-3) malignant residual cells (p = 0.00073). The MRDCons(+) status also correlated significantly with poor/intermediate risk cytogenetics, MDR1 phenotype, short duration of overall survival, and relapse-free survival (p = 0.024, 0.021, 0.00001, and 0.00001, respectively). In multivariate analysis, the MRDCons(+) status was associated with a high probability of relapse (p < 0.00026) and short duration of relapse free survival (p = 0.008). Stem cell transplantation did not seem to alter the prognostic impact of high levels of MRD after consolidation: within the MRDCons(+) group, the relapse rate after transplant was 78%. Thus, a MRD > or = 3.5 x 10(-4) leukemic cells at the end of consolidation strongly predicts relapse, and is significantly associated with MDR1-positive phenotype and intermediate/unfavorable cytogenetics.

A phase-II trial of all trans retinoic acid and low-dose cytosine arabinoside for the treatment of high-risk myelodysplastic syndromes

trans retinoic acid (45 mg/m2) and s.c. low-dose cytosine arabinoside (LDARAc) given at the dose of 20 mg twice per day. The courses were repeated monthly until response or progression; in the case of response, the therapy was administered until relapse. Morphologic diagnoses were refractory anemia with excess blasts (RAEB) in nine, RAEB in transformation (RAEB-t) in nine, and chronic myelomonocytic leukemia (CMMoL) in four patients; in all cases, bone-marrow blast infiltration was greater than 10% (median 20%, range 12–30%). When the international prognostic scoring system was applied, all the cases qualified as intermediate/high-risk categories. Nineteen patients were males and three were females; the median age was 69 years (range 25–90 years); three patients had previously been treated with conventional chemotherapy, and one of them had also undergone autologous bone-marrow transplantation. The criteria of response were defined as follows: (1) complete response: normalization of blood counts and bone-marrow blasts ( <5%), and (2) partial response: decrease in bone-marrow blast infiltration by 50%, and two of the following parameters – improvement in hemoglobin level by 1.5 g/dl or decrease by 50% in transfusional requirement, increase by 50% in absolute neutrophil count, and increase by 50% in platelet count. Overall, 7 (32%) of 22 patients achieved a response, with 5 (23%) being classified as complete responders and 2 (9%) as partial responders. Fifteen (68%) patients did not achieve any response, and 14 died of progressive disease or infectious disease. The overall median survival was 8 months (range 1–27 months), whereas the median survival of responders was 16 months (range 8–27 months); the median duration of response was 11 months (range 2–21 months). Moderate to severe hematological toxicity and infections were the most common side effects. In conclusion, it seems that the association of ATRA and LDARA-C may be effective in approximately 30% of HRMDS patients. Optimizing this approach might be pursued by selecting, on a biological basis, those cases more likely to respond or by incorporating other differentiating agents or growth factors.

Multidimensional Flow Cytometry for Detection of Minimal Residual Disease in Acute Myeloid Leukemia

The term minimal residual disease (MRD) describes the situation in which, after chemotherapy for acute leukemia (AL), a morphologically normal bone marrow (BM) can still harbor a relevant amount of residual malignant cells. Several techniques are now amenable to investigate MRD, and all together they have designated a new era in which a re-definition of the current criteria of complete remission (CR) is required. Depending upon the measured level of MRD we can distinguish a variety of clinical situations ranging from a potentially cured disease to short-term remission. In the context of this spectrum of conditions there would be room for different therapeutic strategies ranging from no further therapy to pre-emptive therapy to treat early relapses (immunologic and/or molecular relapses). This review will focus on the state of art of MRD detection in acute myeloid leukemia (AML) using multidimensional flow cytometry (MFC), and will cover the laboratory and clinical aspects of this approach.