Tefferi, Ayalew

Abstract Background: In childhood acute lymphoblastic leukemia (ALL), a rapid decline of circulating blasts in response to induction chemotherapy or prednisone is one of the most important prognostic factors for achieving remission and for relapse-free survival (RFS). However, in AML parameters of chemosensitivity have been restricted to the assessment of residual bone marrow (BM) blasts during aplasia. We hypothesized that time to peripheral blood (PB) blast clearance would have prognostic relevance with respect to RFS in AML. Methods: From 1994 to 2006 outcomes of 86 adults (47 males; median age 52 y) with previously untreated AML (non- acute promyelocytic leukemia) achieving CR and receiving high-dose cytarabine (HDAC)-based consolidation have been included in this retrospective analysis. The median and range of white blood cell count (WBC), platelet count and PB blast percent were 5.3 × 109/L (0.9–267), 59 ×109/L (5–361) 22.5% (0–97), respectively. A 100 cell manual slide differential count was done in all with PB blasts unless the WBC was less than 0.5 ×109/L. Karyotype was classified as favorable, intermediate and unfavorable in 10, 63 and 12 cases, respectively and failed in one. Induction consisted of idarubicin at 12 mg/m2/ d (n=70) or daunorubicin at 45 mg/m2/d (n=16) on days 1–3 with infusional cytarabine at 100 mg/m2/ d on days 1–7 and repeated in those with persistent BM blasts on day 14 (n=17). The first consolidation was the same as that used to achieve CR. Thereafter, 3 courses of HDAC at 3g/m2 (1.5 g/m2 if ≥ 60 y) q12 h on days 1, 3 and 5 were planned. Results: At the time of analyses, 43% had died, primarily of relapse (97%). Median overall survival (OS) and RFS were 30.2 m (5.5–133) and 14 m (2–131.5), respectively. The median time to PB blast clearance was five days (range: 2–10). For this analysis, only those with PB blasts at initiation of induction (n=73) were included. We defined the day of PB blast clearance as the first day after commencing induction that PB blasts were absent. Separation according to blast clearance on or before day 5 resulted in the most balanced distribution and strongest significant difference between each subgroup of 45 and 28 patients, with significantly different rates of relapse of 33% and 79%, respectively (p<0.0001). We also defined three “blast risk groups” as good, intermediate and poor, according to PB blast clearance on or before day 3, on days 4 or 5, or on day 6 or beyond, respectively. This provided 3 well-balanced groups (good, intermediate and poor) of 16, 30 and 27 patients with significantly different and escalating relapse rates of 12.5%, 47% and 78%, respectively (Fig 1). Univariate analysis of several variables identified PB blast clearance day (p<0.0001), number of inductions to CR (p =0.0043) and cytogenetic risk group (p=0.028) as being significantly associated with RFS. On multivariable analysis, only the first sustained its significance. The results were similar for OS. Conclusion: In adults who achieve CR after induction chemotherapy for non-APL AML, early PB blast clearance predicts superior RFS and OS. Figure Figure

Abstract Background: Previous cytogenetic studies in polycythemia vera (PV) have included a relatively small number of patients (“n” ranging 10–64). In the current study (n=137), we describe cytogenetic findings at presentation and examine their relationship to clinical and laboratory features, including bone marrow JAK2V617F allele burden. Methods: The study consisted of a consecutive group of patients with PV who fulfilled the World Health Organization (WHO) diagnostic criteria and in whom bone marrow biopsy and cytogenetic studies were performed at diagnosis. Results I: cytogenetic details At diagnosis: A total of 137 patients (median age, 64 years; 49% females) were studied at diagnosis and had adequate metaphases for interpretation. Cytogenetics were normal in 117 patients (85%) and displayed either a sole -Y abnormality in 5 patients (7% of the male patients), and other chromosomal abnormalities in 15 (11%). The latter included trisomy 8 in five patients, trisomy 9 in three patients, two patients each with del(13q), del(20q), and abnormalities of chromosome 1, and one patient each with del(3)(p13p21), dup(13)(q12q14), and del(11)(q21). At follow-up: Repeat cytogenetic studies while still in the chronic phase of the disease were performed in 19 patients at a median of 60 months (range, 8–198) from diagnosis. Of these, 4 had aquired new cytogenetic clones including 3 with normal cytogenetics at time of initial PV diagnosis. The new abnormalities included del(20q), del(5q), del(1p), chromosome 1 abnormality, and inv(3)(q21q26.2). At time of disease transformation: Leukemic transformation was documented in 3 patients of whom cytogenetic information at the time was available in 2 patients; both patients had normal results at time of initial PV diagnosis and complex cytogenetic abnormalities at time of leukemic transformation. In contrast, among 6 patients with available cytogenetic information at time of fibrotic transformation, the results were unchanged from those obtained at time of diagnosis in 5 patients. ii) Correlation between cytogenetics at diagnosis and JAK2V617F allele burden: Allele-specific, quantitative PCR analysis for JAK2V617F was performed in 71 patients using genomic DNA from archived bone marrow obtained at the time of the initial cytogenetic studies. JAK2V617F mutation was detected in 64 of the 71 (90%) patients; median mutant allele burden was 16% (range 3–80%) without significant difference among the different cytogenetic groups: normal vs. –Y vs. other cytogenetic abnormalities (p=0.72). iii) Clinical correlates and prognostic relevance of cytogenetic findings at diagnosis: Among several parameters studied for significant correlations with cytogenetic findings at diagnosis, an association was evident only for age (p=0.02); all –Y abnormalities (n=5) as well as 13 of the 15 (87%) other cytogenetic abnormalities occurred in patients ≥ 60 years of age. Stated another way, the incidence of abnormal cytogenetics (other than -Y) was 4% for patients younger than age 60 years and 15% otherwise. The presence of abnormal cytogenetics at diagnosis had no significant impact on either overall or leukemia-free survival. Conclusions: Abnormal cytogenetic findings at diagnosis are infrequent in PV, especially in patients below age 60 years. Furthermore, their clinical relevance is limited and there is not significant correlation with bone marrow JAK2V617F allele burden.

Abstract Background: Pruritus in polycythemia vera (PV) has been associated with a high JAK2V617F allele burden (Tefferi et al. Cancer. 2006;106:631, Vannucchi et al, Leukemia2007, in press). However, there is limited information regarding clinical correlates of pruritus per se in PV. Accordingly, we conducted a large (n=418) retrospective study in PV to accurately assess the prevalence and severity of pruritus as well as its relationship with presenting clinical features, bone marrow JAK2V617F allele burden, and prognosis. Methods: The study cohort consisted of a consecutive group of patients with PV who fulfilled the World Health Organization (WHO) diagnostic criteria and in whom information regarding the presence or absence of pruritus at diagnosis was fully documented. Pruritus was graded as being mild, moderate, or severe based on careful review of patient history. Results: Prevalence and severity of pruritus: A total of 418 patients fulfilled the above stipulated criteria for study inclusion (median age of 60 years; 56% males; median follow-up of 73 months). Of these, 131 (31%) experienced pruritus at time of initial diagnosis: mild in 97 patients (74%), moderate in 22 (17%), and severe in 12 (9%). Most patients with mild pruritus (n=92) received no specific therapy for pruritus whereas all 12 patients with severe pruritus required treatment; all 7 patients who received treatment with paroxetine responded. Pruritus and JAK2V617F allele burden: Quantitative measurement of JAK2V617F was performed in 64 patients using genomic DNA from archived bone marrow obtained at the time of PV diagnosis. Mutational frequency was 97%; median mutant allele burden was 25.4% (range 0.2 - 93.3%). Among the 62 V617F-positive patients, 24 (39%) had pruritus; the proportion of patients in the upper and lower quartile allele burden ranges were 50% and 0% in the presence of pruritus and 18% and 34% in its absence (p=0.002). Clinical correlates of pruritus: Pruritus was more prevalent in non-smokers (35% vs. 19%; p=0.004) and non-diabetics (33% vs. 16%; p=0.04) and was associated with a lower rate of arterial thrombosis at diagnosis (8% vs. 17%; p=0.01) as well as during follow-up (16% vs. 30%; p=0.003). These significant associations remained intact during multivariable analysis. Pruritus did not have an impact on the rate of leukemic or fibrotic transformation. Conclusions: Pruritus in PV is independently associated with a lower risk of arterial thrombosis despite high JAK2V617F allele burden and is also more prevalent in non-smokers. These observations suggest that pruritus might be a surrogate for an underlying platelet pathology with functional relevance.

Abstract Background: We sought to validate the International Prognostic Scoring System (IPSS) for myelodysplastic syndromes (MDS) in a well-characterized group of patients with refractory anemia with ringed sideroblasts (RARS), as well as to assess the validity of novel and easily obtained prognostic factors such as complete blood count (CBC) parameters. For instance, Bowles and colleagues (Br J Haematol2006; 135: 198) recently reported that a low platelet mass (i.e., mean platelet volume (MPV) x platelet count <0.60 mL/L) is a marker for poor survival (median, 5 months) in patients with MDS, and also asserted the MPV and platelet count do not correlate so the MPV provides independent prognostic information. Because platelet parameters have been of particular interest in RARS, especially with respect to the provisional entity RARS with thrombocytosis (RARS-T), we assessed the prognostic relevance of MPV and platelet count/mass in this MDS subtype. Methods: We analyzed all pts diagnosed with RARS by FAB criteria at Mayo Clinic between 1994 and 2002, a period chosen because it was the interval when the MPV was routinely reported along with CBCs at our center. We excluded pts who were evaluated at Mayo Clinic >6 months after initial diagnosis or were being treated for other malignancies (e.g., lymphoma or myeloma). Medical records, blood parameters, bone marrow findings, and cytogenetic results were reviewed. We performed time-to-event analysis using Kaplan-Meier estimates and Cox proportional hazards modelling, with Chi-Square statistics to assess significance (i.e., type I error rate <0.05). Results: A total of 127 pts (79 males; median age 73 years, range 51-90) were identified, with 78 verified events during the follow-up period. The median survival was 3.9 years. The correlation between MPV and platelet count was poor (Pearson correlation coefficient, -0.18). Platelet count was low in 26 pts, within the normal range in 93 pts, and elevated in 8 pts (but >600 ×109/L in only 2). In univariate analysis, lower Hb, lower MCV (but not MPV), higher total white count, lower platelet count, abnormal cytogenetics, and higher IPSS score each predicted poorer survival. In multivariate analysis using a proportional hazards model, only the platelet count (hazard ratio 0.93 for each increase in count by 25, 95% confidence interval (CI) 0.88-0.98, p=0.0055) and IPSS score (hazard ratio 7.7 for each increase in IPSS score by 1, 95% CI 4.1-14.4, p<0.0001) remained independently predictive of survival. Relatively few patients with RARS have either low platelet mass (<0.60 mL/L, 8%) or intermediate platelet mass (0.60-1.20 ml/L, 12%) as defined by Bowles et al. While the platelet mass was indeed predictive of survival (p<0.0001), this was due entirely to the platelet count, and there was no additional information from the MPV. Conclusion: The IPSS is an excellent predictor of survival in pts with RARS, and the quantitative platelet count adds independent information. The MPV has no additional effect on survival in patients with RARS.

Abstract Background: Most experts agree that iron chelation therapy is unlikely to benefit myelodysplastic syndrome (MDS) patients with a bone marrow (BM) blast percentage of ≥ 5% because of the associated short life expectancy. In contrast, using a serum ferritin of 1,000 ng/mL as a surrogate for iron overload, a recent study suggested a negative impact of iron overload on overall survival in MDS patients with < 5% BM blasts including those with refractory anemia with ringed sideroblasts (RARS) (Malcovati et al. JCO2005:23:7594). In the current retrospective study, we examined the validity of this observation in a large group of RARS patients seen at a single institution. Methods: The diagnosis of RARS was based on the French-American-British cooperative group criteria. Serum ferritin levels obtained both at diagnosis and during follow-up as well as total number of packed red blood cells transfused were recorded. Standard statistical methods were used for survival and other analyses. Results: A total of 126 RARS patients (median age 73 years, range 44-90; 67% males) were seen at our institution over the last several years. At diagnosis, median (range) values were 9.4 g/dL (5.7-13.4) for hemoglobin (Hgb), 2.8 × 109/L (0.3-13.7) for absolute neutrophil count (ANC) and 214× 109/L (22-819) for platelet count; 38% of the patients had received red blood cell (RBC) transfusions at the time of initial diagnosis. International Prognostic Scoring System (IPSS) risk distributions in evaluable patients were 66% for low, 28% for intermediate-1 and 6% for intermediate-2 risk. Median follow-up was 36 months and during this time 83 patients (66%) had died and leukemic transformation was documented antemortem in 8 patients (6%). As expected, IPSS was highly predictive of survival outcome (p<0.0001). In addition, history of RBC transfusions at diagnosis (p=0.001) but not the total number of RBC transfusions received during the entire disease course (p=0.17) carried an independent prognostic value for inferior survival. There were no significant correlations between overall survival and serum ferritin level at either diagnosis (median 567 ng/mL, range 16-3,475; p=0.24) or the maximum value during follow-up (median 1,108 ng/mL; range 238-43,500; p=0.72). Similarly, Kaplan-Meier plots of 77 evaluable patients stratified by serum ferritin levels of < or ≥1000 ng/mL at diagnosis or 107 evaluable patients stratified by maximal serum ferritin levels of < 1000, 1000-5000, or > 5000 ng/mL during follow-up revealed similar curves (Figure). Conclusions: The current study suggests no significant association between transfusional hemosiderosis and survival even in good risk patients with MDS. The study also undermines the utility of serum ferritin as a surrogate for assessing the value of therapeutic iron chelation. Figure Figure Figure Figure

Abstract The infused autograft absolute lymphocyte count (A-ALC) is an independent prognostic factor for survival after autologous hematopoietic stem cell transplant (AHSCT) in non-Hodgkin’s lymphoma (NHL). Previous studies have shown that A-ALC directly correlates with the peripheral blood absolute lymphocyte count (PC-ALC) at the time of apheresis collection. However, factors affecting the PC-ALC at apheresis remain undefined. We hypothesized that one possible factor impacting PC-ALC may be the time interval from last chemotherapy to stem cell collection (TILC). Data from 160 patients who underwent AHSCT for treatment of relapsed NHL at Mayo Clinic between 1993 and 2001 were collected and analyzed. The primary end point of this study was correlation between TILC and PC-ALC, and our analysis revealed a strong correlation (r = 0.67, p < 0.0001). Further analysis revealed higher PC-ALC numbers in patients with TILC ≥ 60 days versus TILC < 60 days (median of 7.44 x 109/L in the ≥ 60 day group versus 3.87 x 109/L in the < 60 day group, p < 0.0001). Both the median overall survival (figure below) and the progression-free survival were longer in the TILC ≥ 60 days group versus the TILC < 60 days group (76 versus 21 months, p < 0.0037; 76 versus 11 months, p < 0.018, respectively). These findings were independent of other prognostic indicators for relapsed NHL patients undergoing APHSCT. The data supports our hypothesis that TILC affects PC-ALC and survival post-AHSCT in NHL. Figure Figure

Abstract BACKGROUND : Leukemic transformation (LT) eventually occurs in 10–30% of patients with myelofibrosis with myeloid metaplasia (MMM). Hoever, there is limited data regarding bone marrow histological and cytogenetic changes that accompany the process. METHODS : The study population consisted of 67 consecutive patients with MMM in whom the diagnosis of LT was documented at our institution by bone marrow examination. Serial comparisons of both histological and cytogenetic changes were performed to monitor longitudinal changes accompanying disease progression. RESULTS : I. Bone marrow histologic changes during LT : All 67 patients displayed acute myeloid leukemia (AML) and all FAB subtypes except M3 were represented (most common were M7 at 25.4% and M0 at 22.4%). Median bone marrow myeloblast percentage was 43% (range 20–95) and cellularity 80% (range 5–95). However, only 37.5% and 30.3% had marked increases (> grade 2) in either reticulin fibrosis or osteosclerosis, respectively. Serial bone marrow comparisons (available in 33 patients at a median interval of 24.5 months (range; 2.1–110.4) after the diagnosis of MMM) demonstrated that LT was occasionally a focal process, but was always accompanied by a marked increase in myeloblasts. However, overall bone marrow cellularity changed by >15% in only 25% of the cases, none of the LT was associated with a > grade 1 change in reticulin fibrosis, and only 12% had an increase in osteosclerosis. On the other hand, LT was frequently associated with morphologic changes in megakaryocytes from large, round and clumped at the time of diagnosis of MMM to small, condensed and non-clumped at the time of LT. II. Karyotypic changes of MMM to LT: Karyotypic analysis was available in 32 patients (48%) at the time of MMM (abnormal in 63%) and 56 patients (84%) at the time of LT (abnormal in 91%). At LT the majority of patients (n=34; 60.7%) displayed poor risk karyotypes with either complex karyotypes or single abnormalities of adverse prognostic significance. In contrast, only two patients (3.6%) manifested prognostically favorable karyotypes with a translocation or inversion involving one of the core binding factors (t(8;21)(q22;q22);Inv(16)(q22)). Twenty-eight patients had chromosome analysis performed both at diagnosis of MMM and at LT. At diagnosis of MMM, 20 patients had either normal chromosomes or a single chromosome anomaly while only 7 displayed a complex karyotype with 3 or more chromosome anomalies. However, at LT all twenty-eight patients had abnormal karyotypes, including 17 patients with markedly complex karyotypes and only 9 with simple clonal anomalies. The recurrent numeric and structural anomalies observed at transformation to AML most frequently involved chromosomes 5, 6, 7, 8, 17, 19 and 21. Additionally, amongst the 26/28 (93%) of patients whose karyotype changed from MMM to LT 11 patients (43%) displayed evolution of an existing clone, and 15 (57%) developed one or more additional abnormal clones. CONCLUSIONS: The progression of MMM to LT is characterized by the emergence of new cytogenetic abnormalities as well as morphological changes in megakaryocytes. Unfortunately, the karyotypic profile that accompanies clonal evolution in MMM is often of the poor-risk category that undermines the value of aggressive chemotherapy in such patients. However, such therapy might be of value to the occasional patient with good-risk cytogenetic lesions.

Abstract Introduction: The core-binding factor subunit RUNX1/CBFA2/AML1 (Runt-related transcription factor 1, Core-binding factor alpha 2 subunit, Oncogene AML1) is critical for generation of hematopoietic stem cells during embryogenesis and for normal megakaryopoiesis in adults. RUNX1/CBFA2/AML1 has long been recognized as an oncogene, and is frequently involved in leukemia-associated translocations that create aberrant fusion proteins (e.g., AML1-ETO). Recently, acquired somatic point mutations in the RUNX1/CBFA2/AML1 gene have been described in a subset of patients with myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). These point mutations are associated with the loss of normal RUNX1 trans-activation potential and/or altered binding to the non-DNA-binding subunit of core-binding factor, CBFB. Given the importance of core-binding factor in megakaryocytic differentiation and platelet production and the central role of megakaryocytes in the pathophysiology of myelofibrosis with myeloid metaplasia (MMM), and in light of the myelofibrotic phenotype of GATA-1low mice (GATA-1 and RUNX1 are co-expressed and co-operate in megakaryocytic differentiation), we hypothesized that RUNX1 gene mutations might be common in MMM. In addition, we wanted to know whether patients with MDS-associated acquired alpha thalassemia (i.e., ATMDS), a special MDS subgroup with a very high incidence of point mutations in the X-linked ATRX gene, have an especially high incidence of RUNX1 mutations that might account for the more severe hematopoietic defect found in these patients compared with boys with inherited ATR-X syndrome (detailed in Steensma et al, Blood2004;103:2019). Methods and Results: We analyzed samples from 26 well-characterized patients with MMM and from 52 with MDS (11 RA, 3 RARS, 8 RCMD, 3 RAEB-1, 7 RAEB-2, 2 MDS in transition to AML-M6, 14 atypical/unclassifable MDS, 4 MDS subtype unknown - including 18 with ATMDS) for RUNX1 point mutations by denaturing high-performance liquid chromatography (DHPLC) followed by subcloning and sequencing of samples exhibiting heteroduplex peaks. We found 5 RUNX1 mutations in MDS patients (9.6%), all of whom had RAEB-2 (4 patients) or a history of treated AML (1 patient), but detected no mutations in MMM patients. All patients with RUNX1 mutations progressed to overt leukemia within 1 year. ATMDS patients did not have an increased risk of RUNX1 point mutations (2/18 patients with mutations, 11.1%) when compared with MDS without thalassemia (3/34 patients, 8.8%; p=0.58). Additionally, 1 MDS patient had a clonally restricted mutation in GATA-1 (c.1066 C>T) (also screened by DHPLC), but this is not predicted to change the GATA-1 amino acid sequence and is likely a random event reflecting the genomic instability characteristic of MDS. Conclusion: RUNX1 point mutations are common in high-risk MDS, but not in MMM, and that ATMDS patients do not have a risk of RUNX1 mutations in excess of that expected for MDS in general.

In this phase 1 trial, inhibition of granulocyte-macrophage colony-stimulating factor (GM-CSF) was associated with clinically meaningful responses in 5 of 15 patients with relapsed or refractory chronic myelomonocytic leukemia (CMML). Preliminary data suggest that this approach may be tractable in CMML bearing activating NRAS mutations.