Irreversible TrxR1 inhibitors block STAT3 activity and induce cancer cell death.

Because of its key role in cancer development and progression, STAT3 has become an attractive target for developing new cancer therapeutics. While several STAT3 inhibitors have progressed to advanced stages of development, their underlying biology and mechanisms of action are often more complex than would be expected from specific binding to STAT3. Here, we have identified and optimized a series of compounds that block STAT3-dependent luciferase expression with nanomolar potency. Unexpectedly, our lead compounds did not bind to cellular STAT3 but to another prominent anticancer drug target, TrxR1. We further identified that TrxR1 inhibition induced Prx2 and STAT3 oxidation, which subsequently blocked STAT3-dependent transcription. Moreover, previously identified inhibitors of STAT3 were also found to inhibit TrxR1, and likewise, established TrxR1 inhibitors block STAT3-dependent transcriptional activity. These results provide new insights into the complexities of STAT3 redox regulation while highlighting a novel mechanism to block aberrant STAT3 signaling in cancer cells.

IL-6 activated JAK/STAT3 pathway and sensitivity to Hsp90 inhibitors in multiple myeloma.

Modern anti-cancer treatment involves targeted therapy that aims at inactivating particular oncoproteins or signaling pathways in a cancer-type-specific manner. A number of potent targeted therapies affecting oncogenic kinases or receptor tyrosine kinases have revolutionized anti-cancer treatment. These drugs inactivate signaling pathways that cancer cells depend on and therefore inhibit their proliferation and survival. Molecular chaperones of the Hsp90 family (heat shock protein 90) support the integrity, folding and function of many proteins involved in proliferation, survival, DNA damage and repair. Hsp90 proteins are thus required to maintain activity of a large variety of oncogenic proteins, including members of the JAK/STAT and the PI3K pathways. Accordingly cancer cells rely on Hsp90 proteins and their expression is often elevated in malignant cells. In line with this, inhibitors of Hsp90 (Hsp90-Is) have demonstrated potent antitumor activity in preclinical studies. While Hsp90-Is can be considered as targeted therapy, their broad effects on multiple signaling pathways make it difficult to predict the therapeutic outcome. Multiple myeloma (MM) is one of the tumor types with elevated Hsp90 levels. Hsp90-Is demonstrated promising activity in preclinical studies of MM and in several clinical trials. However, large variability in response questioned the use of Hsp90-Is as single drugs in the treatment of myeloma. A critical factor in targeted therapies, including Hsp90-Is, is identification of susceptible subgroups of patients. Predictive biomarkers in each particular tumor type are important in order to use anti-cancer drugs in a rational way. Interestingly, levels of Hsp90 expression has not proven to be decisive for treatment response and hence stratification of myeloma patients. Others and we have recently found that MM cells with an IL-6-activated JAK/STAT3 pathway are particularly sensitive to Hsp90-Is. In this review we will discuss these findings, both in terms of molecular mechanisms and applications for selection of MM patients amenable to Hsp90-I treatment in an individually targeted treatment strategy.

GCN5 acetylates and regulates the stability of the oncoprotein E2A-PBX1 in acute lymphoblastic leukemia.

The t(1;19) translocation in pediatric pre-B-cell acute lymphoblastic leukemia (ALL) fuses the genes, which encode the transcriptional activator E2A and homeobox pre-B-cell leukemia transcription factor 1 (PBX1), resulting in expression of the chimeric transcription factor E2A-PBX1. E2A-PBX1 can promote cell transformation both in vitro and in vivo; however, the mechanisms by which E2A-PBX1 contributes to malignancy merit further investigation. In the current work we report, for the first time, a physical and functional interaction between the SPT3-TAFII31-GCN5L acetylase (STAGA) complex and E2A-PBX1. STAGA, and its acetyltransferase subunit GCN5, directly interacted with the E2A portion of E2A-PBX1. GCN5 acetylated E2A-PBX1 and increased the stability of E2A-PBX1 protein in cells. Moreover, the GCN5 inhibitor α-methylene-γ-butyrolactone 3 (MB-3) decreased E2A-PBX1 acetylation and E2A-PBX1 protein levels in leukemic cells, indicating that GCN5 inhibitors have potential value as therapeutic agents for ALL. In addition, we show that the E3 ubiquitin ligase HDM2 potentiates the degradation of E2A-PBX1. We suggest that dynamic regulation of E2A-PBX1 protein levels in vivo has a fundamental role in ALL.

Targeting of distinct signaling cascades and cancer-associated fibroblasts define the efficacy of Sorafenib against prostate cancer cells.

Sorafenib, a multi-tyrosine kinase inhibitor, kills more effectively the non-metastatic prostate cancer cell line 22Rv1 than the highly metastatic prostate cancer cell line PC3. In 22Rv1 cells, constitutively active STAT3 and ERK are targeted by sorafenib, contrasting with PC3 cells, in which these kinases are not active. Notably, overexpression of a constitutively active MEK construct in 22Rv1 cells stimulates the sustained phosphorylation of Bad and protects from sorafenib-induced cell death. In PC3 cells, Src and AKT are constitutively activated and targeted by sorafenib, leading to an increase in Bim protein levels. Overexpression of constitutively active AKT or knockdown of Bim protects PC3 cells from sorafenib-induced killing. In both PC3 and 22Rv1 cells, Mcl-1 depletion is required for the induction of cell death by sorafenib as transient overexpression of Mcl-1 is protective. Interestingly, co-culturing of primary cancer-associated fibroblasts (CAFs) with 22Rv1 or PC3 cells protected the cancer cells from sorafenib-induced cell death, and this protection was largely overcome by co-administration of the Bcl-2 antagonist, ABT737. In summary, the differential tyrosine kinase profile of prostate cancer cells defines the cytotoxic efficacy of sorafenib and this profile is modulated by CAFs to promote resistance. The combination of sorafenib with Bcl-2 antagonists, such as ABT737, may constitute a promising therapeutic strategy against prostate cancer.

MicroRNA-203 functions as a tumor suppressor in basal cell carcinoma.

Basal cell carcinoma (BCC) of the skin represents the most common malignancy in humans. MicroRNAs (miRNAs), small regulatory RNAs with pleiotropic function, are commonly misregulated in cancer. Here we identify miR-203, a miRNA abundantly and preferentially expressed in skin, to be downregulated in BCCs. We show that activation of the Hedgehog (HH) pathway, critically involved in the pathogenesis of BCCs, as well as the EGFR/MEK/ERK/c-JUN signaling pathway suppresses miR-203. We identify c-JUN, a key effector of the HH pathway, as a novel direct target for miR-203 in vivo. Further supporting the role of miR-203 as a tumor suppressor, in vivo delivery of miR-203 mimics in a BCC mouse model results in the reduction of tumor growth. Our results identify a regulatory circuit involving miR-203 and c-JUN, which provides functional control over basal cell proliferation and differentiation. We propose that miR-203 functions as a 'bona fide' tumor suppressor in BCC, whose suppressed expression contributes to oncogenic transformation via derepression of multiple stemness- and proliferation-related genes, and its overexpression could be of therapeutic value.

Glucocorticoid-induced cell death is mediated through reduced glucose metabolism in lymphoid leukemia cells.

Malignant cells are known to have increased glucose uptake and accelerated glucose metabolism. Using liquid chromatography and mass spectrometry, we found that treatment of acute lymphoblastic leukemia (ALL) cells with the glucocorticoid (GC) dexamethasone (Dex) resulted in profound inhibition of glycolysis. We thus demonstrate that Dex reduced glucose consumption, glucose utilization and glucose uptake by leukemic cells. Furthermore, Dex treatment decreased the levels of the plasma membrane-associated glucose transporter GLUT1, thus revealing the mechanism for the inhibition of glucose uptake. Inhibition of glucose uptake correlated with induction of cell death in ALL cell lines and in leukemic blasts from ALL patients cultured ex vivo. Addition of di-methyl succinate could partially overcome cell death induced by Dex in RS4;11 cells, thereby further supporting the notion that inhibition of glycolysis contributes to the induction of apoptosis. Finally, Dex killed RS4;11 cells significantly more efficiently when cultured in lower glucose concentrations suggesting that modulation of glucose levels might influence the effectiveness of GC treatment in ALL. In summary, our data show that GC treatment blocks glucose uptake by leukemic cells leading to inhibition of glycolysis and that these effects play an important role in the induction of cell death by these drugs.

Cell death induced by dexamethasone in lymphoid leukemia is mediated through initiation of autophagy.

Glucocorticoids are fundamental drugs used in the treatment of lymphoid malignancies with apoptotic cell death as the hitherto proposed mechanism of action. Recent studies, however, showed that an alternative mode of cell death, autophagy, is involved in the response to anticancer drugs. The specific role of autophagy and its relationship to apoptosis remains, nevertheless, controversial: it can either lead to cell survival or can function in cell death. We show that dexamethasone induced autophagy upstream of apoptosis in acute lymphoblastic leukemia cells. Inhibition of autophagy by siRNA-mediated repression of Beclin 1 expression inhibited apoptosis showing an important role of autophagy in dexamethasone-induced cell death. Dexamethasone treatment caused an upregulation of promyelocytic leukemia protein, PML, its complex formation with protein kinase B or Akt and a PML-dependent Akt dephosphorylation. Initiation of autophagy and the onset of apoptosis were both dependent on these events. PML knockout thymocytes were resistant to dexamethasone-induced death and upregulation of PML correlated with the ability of dexamethasone to kill primary leukemic cells. Our data reveal key mechanisms of dexamethasone-induced cell death that may inform the development of improved treatment protocols for lymphoid malignancies.

Cleavage of the IPS-1/Cardif/MAVS/VISA does not inhibit T cell-mediated elimination of hepatitis C virus non-structural 3/4A-expressing hepatocytes.

BACKGROUND: Hepatitis C virus (HCV) effectively establishes persistent infection in human livers. The non-structural (NS) 3/4A complex participates in this process by cleavage of interferon beta (IFN beta) promoter stimulator-1 (IPS-1; also termed Cardif/MAVS/VISA), which inhibits responses to double stranded (ds) RNA. However, it is not known whether this effect extends beyond innate responses. AIMS: To test if HCV NS3/4A affects innate and adaptive immune responses in vivo. METHODS: NS3 levels were semi-quantified in human liver biopsies, transfected cells, and in transgenic (Tg) mouse livers by western blot. The effect of NS3/4A on dsRNA-mediated signalling and on the integrity of IPS-1 was analysed using in vitro translation, transfected cells and Tg mice. Cytotoxic T cell (CTL)-mediated clearance of transient firefly luciferase (FLuc)- and/or NS3/4A-Tg hepatocytes was determined using in vivo imaging and western blot. RESULTS: NS3 protein levels were in a comparable range (0.1-49 microg/g tissue) in infected human livers and Tg mouse livers. Importantly, these levels of NS3/4A reduced murine innate responses to synthetic dsRNA in vivo, supporting the possibility that this occurs also in infected humans. The likely explanation for this was the NS3/4A-mediated cleavage of mouse IPS-1, albeit less efficiently than human IPS-1. Despite this, FLuc- and/or NS3/4A-expressing murine hepatocytes were effectively eliminated by hepatic CTLs, utilising the classical molecules for virus-infected cell lysis, including CD8, IFN gamma, perforin and FasL. CONCLUSIONS: Although HCV NS3/4A inhibits the innate immunity, this does not prevent CTL-mediated clearance of NS3/4A-expressing hepatocytes in vivo. Thus, other HCV proteins are most likely responsible for interfering with the adaptive immunity.

Children and adults with acute lymphoblastic leukaemia have similar gene expression profiles.

OBJECTIVES: To compare the gene expression pattern in children and adults with acute lymphoblastic leukaemia (ALL) in order to improve our understanding of the difference in disease biology and prognosis. METHODS: The gene expression profiles in diagnostic samples from 29 children and 15 adults with ALL were analysed using the oligonucleotide chip Hu95ver2a, produced by Affymetrix. RESULTS: Unsupervised hierarchical cluster analysis revealed that, in spite of differences in outcome, patients clustered irrespective of age, first by T-cell or B-precursor immunophenotype, and second by cytogenetic changes within the B-precursor group. The expression pattern analysis allowed the reclassification of some samples into the proper cytogenetic group. We also showed that separate clustering of samples with the BCR/ABL translocation could be explained by different breakpoint regions in the BCR. No significant difference in gene expression was observed between samples with and without CDKN2A deletion within the B-precursor group. Analysis of different age groups revealed a similarity in expression profiles when infants with the MLL translocation and adults over 40 yr of age were compared irrespective of karyotype. CONCLUSIONS: In spite of the difference in clinical outcome, the gene expression pattern in children and adults with ALL is very similar and is primarily dependent on immunophenotype and cytogenetic aberrations. However, when age groups are compared, the expression patterns of infants and adults over 40 show a remarkable similarity.

Deletion of the Ink4-locus (the p16ink4a, p14ARF and p15ink4b genes) predicts relapse in children with ALL treated according to the Nordic protocols NOPHO-86 and NOPHO-92.

Inactivation of the Ink4 gene locus locus on 9p comprising the tumour suppressor gene p16ink4a and its neighbours p14ARF and p15ink4b is common in childhood acute lymphoblastic leukaemia (ALL), but the prognostic significance is controversial. DNA from 230 patients was retrospectively analysed by Southern blotting, single strand conformation polymorphism (SSCP) and sequencing techniques. The results were correlated with clinical characteristics and outcome. One hundred and ninety-four fully analysed patients, similarly treated using the Nordic NOPHO-86 or the current NOPHO-92 protocols, were included in the outcome analysis. Deletions approached a minimally deleted region between the p16ink4a and p15ink4b genes, making the p14ARF gene the most commonly deleted coding sequence. Bi-allelic deletion was associated with high white blood cell count (WBC) (P < 0.001), T cell phenotype (P < 0.001) and mediastinal mass (P < 0.001). Patients with Ink4 locus bi-allelic deletions had an inferior pEFS (P < 0.01) and multivariate analysis indicated that bi-allelic deletion of the p16ink4a and the p14ARF genes was an independent prognostic risk factor (P < 0.05). Sub-group analysis revealed a pronounced impact of deletion status for high-risk patients, ie with high WBC. Deletion-status and clinical risk criteria (WBC) could thus be combined to further differentiate risk within the high-risk group. The analysis of the Ink4 locus adds independent prognostic information in childhood ALL treated by Nordic protocols and may help in selection of patients for alternative treatment.

Identification and characterization of two novel human mitochondrial elongation factor genes, hEFG2 and hEFG1, phylogenetically conserved through evolution.

Rapid progress in the sequencing of the genome of man and other species allows for the comparative analysis of their genetic structure and content. We have used a combined biochemical and computer-based approach to characterize a 146 kb human genomic bacterial artificial chromosome clone from chromosome 5q13 and discovered a novel human elongation-factor gene, hEFG2. The complete human EFG2 cDNA sequence is 3033 bp and contains 21 exons with conserved exon-intron splice junctions encompassing 45 kb of the genomic sequence with its 5'-end residing within a CpG island, characteristic of a housekeeping gene. The complete size of the hEFG2 cDNA was confirmed by Northern blot and reverse transcription/polymerase chain reaction analysis, which showed a single transcript of 3.2 kb ubiquitously expressed in various human tissues. The hEFG2 protein shows significant homology to several bacterial EF-G proteins, including that of Thermus thermophilus, and to the yeast Saccharomyces cerevisiae mitochondrial elongation factor-G ( MEF2). Multiple alignments reveal a novel gene family of mitochondrial EF-G proteins that can by divided into two subgroups, EF-G1 and EF-G2, in several eukaryotic species including S. pombe, Caenorhabditis elegans and Drosophila melanogaster. Using the information contained in the public databases, we also identified and cloned the complete coding sequence of the human EFG1 gene on chromosome 3q25. The cloning and characterization of these human mitochondrial elongation factor genes should permit us to address their role in the regulation of normal mitochondrial function and in various disease states.

Reactive oxygen species and mitochondria mediate the induction of apoptosis in human hepatoma HepG2 cells by the rodent peroxisome proliferator and hepatocarcinogen, perfluorooctanoic acid.

We have previously shown that one of the most potent rodent hepatocarcinogens, perfluorooctanoic acid (PFOA), induces apoptosis in human HepG2 cells in a dose- and time-dependent manner. In this study we have investigated the involvement of reactive oxygen species (ROS), mitochondria, and caspase-9 in PFOA-induced apoptosis. Treatment with 200 and 400 microM PFOA was found to cause a dramatic increase in the cellular content of superoxide anions and hydrogen peroxide after 3 h. Measurement of the mitochondrial transmembrane potential (Delta Psi(m)) after PFOA treatment showed a dissipation of Delta Psi(m) at 3 h. Caspase-9 activation was seen at 5 h after treatment with 200 microM PFOA. In order to evaluate the importance of these events in PFOA-induced apoptosis, cells were cotreated with PFOA and N-acetylcysteine (NAC), a precursor of glutathione, or Cyclosporin A (CsA), an inhibitor of mitochondrial permeability transition pore (MPT pore). NAC reduced Delta Psi(m) dissipation, caspase 9 activation, and apoptosis, indicating a role for PFOA-induced ROS. In addition, CsA also reduced Delta Psi(m) dissipation, caspase 9 activation, and apoptosis, indicating a role for PFOA-induced opening of the MPT pore. In summary, we have delineated a ROS and mitochondria-mediated pathway for induction of apoptosis by PFOA.

[Transcription map of the 13q14 region, frequently deleted in B-cell chronic lymphocytic leukemia patients]. / Transkriptsionnaia karta raiona 13q14, chasto utrachivaemogo u bol'nykh B-kletochnym khronicheskim limfoleikozom.

Deletions in the region located between the STS markers D13S1168 and D13S25 on chromosome 13 are the most frequent genomic changes in patients with B-cell chronic lymphocytic leukemia (B-CLL). After sequencing of this region, two novel candidate genes were identified: C13orf1 (chromosome 13 open reading frame 1) and PLCC (putative large CLL candidate). Analysis of the repeat distribution revealed two subregions differing in composition of repetitious DNA and gene organization. The interval D13S1168-D13S319 contains 131 Alu repeats accounting for 24.8% of its length, whereas the interval GCT16C05-D13S25, which is no more than 180 kb away from the former one is extremely poor in Alu repeats (4.1% of the total length). Both intervals contain almost the same amount of the LINE-type repeats L1 and L2 (20.3 and 21.24%, respectively). In the chromosomal region studied, 29 Alu repeats were found to belong to the evolutionary young subfamily Y, which is still capable of amplifying. A considerable proportion of repeats of this type with similar nucleotide sequences may contribute to the recombinational activity of the chromosomal region 13q14.3, which is responsible for its rearrangements in some tumors in humans.

Interferon alpha down-regulates telomerase reverse transcriptase and telomerase activity in human malignant and nonmalignant hematopoietic cells.

Recently, the derepressed expression of the catalytic subunit of telomerase, human telomerase reverse transcriptase (hTERT), the enzyme that elongates telomeres, has been implicated as an important step in the immortalization process. The exact regulation of hTERT expression, which is the rate-limiting factor for telomerase activity, is at present unclear. As transformed cells seem to be dependent on a constitutive telomerase activity, the availability of inhibitors would potentially be of great value in antineoplastic therapy. Interferons (IFNs) have been successfully used in the treatment of several forms of malignancies, but the underlying molecular mechanisms responsible for the antitumor activity are poorly defined. In this study we have investigated the effects of IFNs on hTERT expression and telomerase activity. We found that IFN-alpha rapidly (commonly within 4 hours) and significantly down-regulates the expression of hTERT and telomerase activity in a number of human malignant hematopoietic cell lines, primary leukemic cells from patients with acute leukemia as well as T-lymphocytes from healthy donors. This effect of IFN-alpha did not seem to depend on IFN-alpha-mediated cell growth arrest or alterations in c-myc expression. The finding that IFN induces a repression of hTERT and a decrease in telomerase activity suggests a novel mechanism that may play a significant role in the antitumor action of IFN. (Blood. 2000;96:4313-4318)

Mechanisms of interferon-induced cell cycle arrest.

The interferons (IFNs) are a group of cytokines, which in addition to their antiviral activity are capable of modulating a variety of cellular responses. One such prominent effect of IFNs is their potent antimitogenic action, which can be observed both on malignant and non-malignant cells of many different origins. IFNs are also used in the clinic, mainly in malignant and viral diseases, and their cell growth -inhibitory effect has been suggested to be of major importance in their antitumour and antiviral action. The aim of the present review is to provide insight into the molecular mechanisms by which IFNs modulate cell cycle progression in various cell types. With the recent progress in our understanding of how the cell cycle is regulated at the molecular level, it has become possible to delineate intracellular effectors of IFN in this respect. Understanding the antiproliferative effects of IFN may not only help in understanding its antineoplastic and antiviral activities, but may also provide an insight into cell cycle regulation in general and aid in making IFNs a more useful tool in treating disease.

How does interferon-alpha exert its antitumour activity in multiple myeloma?

The interferons (IFNs) have become accepted therapy in a range of hematological and non-hematological malignancies. One malignancy where IFN has demonstrated antitumour activity is multiple myeloma, where monotherapy with IFN-alpha yields a response rate of around 15%. It has also been suggested that myeloma patients may benefit from the addition of IFN-alpha in induction therapy and/or as maintenance treatment. The mechanism behind IFNs antitumour action in myeloma is, however, highly unclear. This probably means that current treatment regimens including IFN are far from optimal, since for example we do not know what factors that are responsible for resistance to IFN therapy in the individual patient, and furthermore we lack optimal tools to design new treatment strategies including IFN. The IFNs are capable of modulating a variety of cellular responses. One prominent effect being their cell growth inhibitory activity through induction of cell cycle arrest and apoptosis, which has also been suggested to be of major importance in IFNs antitumour action. In the present review possible antitumour mechanisms will be discussed, with the focus on ways that IFN may restrict myeloma cell growth.

Comparative sequence analysis of a region on human chromosome 13q14, frequently deleted in B-cell chronic lymphocytic leukemia, and its homologous region on mouse chromosome 14.

Previous studies have indicated the presence of a putative tumor suppressor gene on human chromosome 13q14, commonly deleted in patients with B-cell chronic lymphocytic leukemia (B-CLL). We have recently identified a minimally deleted region encompassing parts of two adjacent genes, termed LEU1 and LEU2 (leukemia-associated genes 1 and 2), and several additional transcripts. In addition, 50 kb centromeric to this region we have identified another gene, LEU5/RFP2. To elucidate further the complex genomic organization of this region, we have identified, mapped, and sequenced the homologous region in the mouse. Fluorescence in situ hybridization analysis demonstrated that the region maps to mouse chromosome 14. The overall organization and gene order in this region were found to be highly conserved in the mouse. Sequence comparison between the human deletion hotspot region and its homologous mouse region revealed a high degree of sequence conservation with an overall score of 74%. However, our data also show that in terms of transcribed sequences, only two of those, human LEU2 and LEU5/RFP2, are clearly conserved, strengthening the case for these genes as putative candidate B-CLL tumor suppressor genes.

Inverse correlation between Ink4-locus deletions and ICM-DNA hyperdiploidy in childhood acute lymphoblastic leukaemia, relation to clinical characteristics and outcome.

Malignant cells from 72 children with ALL were analysed to investigate the relationship between DNA-ploidy and deletion of the Ink4 locus containing the cell-cycle regulating genes p16ink4a, p15ink4b and p14ARF. Image-DNA cytometry (ICM) was used to assess DNA index (DI), and Southern hybridisation was carried out to detect deletions of the Ink4-locus in the leukaemic cells. A DNA content equal to or exceeding 1.16, indicating hyperdiploidy, was detected in 21/72 patients (29%), 1/72 (1.3%) showed DNA-hypodiploidy, and the remaining 50 patients (69%) had a DI within normal limits. Bi-allelic deletion of at least two of the coding sequences from the Ink4 locus was observed in 23/70 (33%) patients. Mono-allelic deletions within the locus were observed in 10/70 patients (14%), and 37/70 patients (53%) had normal signals for both sequences. Out of the 70 patients that could be analysed by both techniques only two had the combination of DNA hyperdiploidy and Ink4-locus bi-allelic deletion (p = 0.004). DNA hyperdiploidy was not associated with any specific clinical characteristics, but there was a trend for a better prognosis for patients with DNA hyperdiploidy (p = 0.09). Ink4-locus deletion was associated with T-cell phenotype and higher white blood cell counts at diagnosis and poor prognosis (p = 0.0015). Multivariate analysis confirmed that Ink4-locus deletion is an independent prognostic marker and a stronger determinant of outcome than DNA ploidy. When DNA ploidy and Ink4-locus deletions were combined, novel subgroups with significantly different outcome could be observed. A group with DNA index > or = 1.16 and no Ink4-locus bi-allelic deletions had an excellent prognosis (p-EFS 0.93 at 60 months), patients with Ink4-locus bi-allelic deletion and a DNA index < 1.16 fared worst (p-EFS 0.57) and patients with no Ink4 deletions and without hyperdiploidy had an intermediate outcome (p-EFS 0.79). The reason for the inverse correlation between DNA ploidy and Ink4 deletion and their combined impact on prognosis remains unclear, and possible reasons are discussed.