Asians Relatively Protected From Venous Thromboembolism
Stein PD, Kayal F, Olson RE, Milford ED. Pulmonary thromboembolism in Asians/Pacific Islanders in the United States: analysis of data from the National Hospital Discharge Survey and the United States Bureau of the Census. Am J Med 2004; 116:435-442.
By Andrew Schafer, M.D.
There have been intriguing observations of significant differences in the relative rates of venous thromboembolism (VTE) in various racial/ethnic groups. In this paper, Stein et al have assessed the rate of diagnosis and incidence in hospitalized patients of pulmonary embolism (PE), deep venous thrombosis (DVT) and combined VTE based on data from the National Hospital Discharge Survey from 1990 through 1999. They also obtained population estimates and data on deaths from pulmonary embolism from the United States Bureau of Census for the same period. The results showed that age-adjusted rates of diagnosis of DVT, PE, and VTE were markedly lower in Asians-Pacific Islanders than in African-Americans and Whites (see Figure). The age-adjusted mortality rates from PE (per 100,000 population per year) were: 3.4 for Whites, 6.9 for African-Americans, and only 1.0 for "others" (including Asians-Pacific Islanders). The authors concluded that clinical assessment of the prior probability of VTE at the bedside should be adjusted based on these ethnic differences.
This paper extends a number of other previous studies that have reported the fascinating observation that VTE occurs much less commonly among Asians-Pacific Islanders (and also Hispanics) than among Caucasians and African-Americans. For example, in an earlier study of an ethnically diverse population in California, the annual incidence of VTE (per 100,000) was found to be: 23 among Caucasians, 29 among African-Americans, 14 among Hispanics, and only 6 among Asians-Pacific Islanders (Ann Intern Med 1998; 128:737). What accounts for this strikingly lower relative rate of clinically diagnosed VTE among Asians, which has been repeatedly noted in many different settings? Inherited factors are likely to be important. Asians may have a genetically more favorable balance of antithrombotic-to-prothrombotic factors. This is supported by the observation that Asians appear to be more sensitive to warfarin anticoagulation than Whites (Whether or not warfarin sensitivity is related to thromboembolic risk has yet to be ascertained). The genetic thrombophilia, factor V Leiden, occurs in about 5 percent of European and North American Whites, but only 0.5 percent of Asians. (This factor alone cannot explain, however, the lower incidence of VTE in Hispanics than in African-Americans as the mutation is present in about 2 percent of the former and <1 percent of the latter populations.) Genetically lower levels of prothrombotic coagulation factor levels may also contribute.
The effect of race and ethnicity on level of hypercoagulability may not be the only genetic factor accounting for differences in VTE incidence. As the intrinsic mechanical function of lower extremity veins has also been recently reported to have a genetic basis, it would be interesting to study this potential risk factor for VTE in different racial and ethnic groups. Finally, of course, the potentially critical role of environmental factors (e.g. diet, smoking) should be evaluated. The impressive differences in relative rates of VTE in different populations present a goldmine of opportunities for research by genetic and molecular epidemiologists to unravel gene-gene and gene-environmental interactions as well as to identify novel candidate genes that are involved in the pathogenesis of venous thromboembolism.
Return to Top
Galectin-1: An Important Weapon for Cancer Immune Evasion
Rubinstein N, Alvarez M, Zwirner NW, Toscano MA, Ilarregui JM, Bravo A, Mordoh J, Fainboim L, Podhajcer OL, Rabinovich GA. Targeted inhibition of galectin-1 gene expression in tumor cells results in heightened T cell-mediated rejection: A potential mechanism of tumor-immune privilege. Cancer Cell 2004 Mar;5(3):241-51.
By Peter Lee, M.D.
Through novel immune assays developed over the past decade, such as peptide-MHC tetramers, ELISPOT, cytokine flow cytometry, and real-time PCR, endogenous anti-cancer T cell responses have been detected in patients with leukemia, melanoma, breast, colorectal, head, and neck cancers. Despite the existence of such anti-cancer T cells, most of these patients have persistent and/or progressive cancer growth. This constitutes an important paradox in tumor immunology - the persistence of cancer in spite of demonstrable anti-tumor T cell responses. Implicit in this is that successful cancer cells acquire mechanisms to evade immune destruction. This paper by Rubinstein et al. sheds light on this important issue. The study focuses on galectin-1 (Gal-1), which is a negative regulator of T cell activation and survival. The authors showed that Gal-1 is expressed at increased levels in a number of human and murine tumor cell lines, and that secretion of Gal-1 by these cells directly correlated with T cell death in a dose-dependent manner. Importantly, inhibition of Gal-1 expression prevented tumor growth in immunocompetent mice, but had no effect on tumor growth in immunodeficient mice. They demonstrated that this protective effect is mediated via generation of anti-tumor T cells in response to Gal-1-negative tumor cells. In contrast, mice injected with Gal-1-positive tumor cells had little or no anti-tumor T cells due to their apoptosis. Furthermore, mice previously challenged with Gal-1-negative tumor cells resisted subsequent challenge with Gal-1-positive tumor cells. Together, these data strongly suggest that Gal-1 plays a pivotal role in promoting escape from T cell-dependent immunity, thus conferring immune evasion to tumor cells.
Increased expression of Gal-1 and other members of the galectin family (Gal-3 in particular) has been recently demonstrated in a number of human leukemias and lymphomas. Galectin expression has been thought to confer apoptosis resistance to these tumor cells. This study suggests that Gal-1 expression may have the dual effect of also protecting tumor cells from immune attack. As leukemia and lymphoma cells exist within the same compartments as normal immune cells - and express novel proteins (immunoglobulin in B cells and TCR in T cells) which may serve as unique 'cancer antigens' - it seems logical that they should be targets for the host immune response. Expression of molecules which promote immune evasion may thus be an important event in successful leukemogenesis. Remarkably, expression of a single protein, Gal-1, may convert a tumor cell from being immune sensitive to being immune resistant. If so, drugs which block the expression of Gal-1 by tumor cells or block the effects of Gal-1 on responding T cells may constitute novel cancer therapies.
Return to Top
Designer Stromal Cells for Hematopoietic Support Versus Gene Therapy
Bensidhoum M, Chapel A, Francois S, et al. Homing of in vitro expanded Stro-1- or Stro-1+ human mesenchymal stem cells into the NOD/SCID mouse and their role in supporting human CD34 cell engraftment. Blood 2004;103:3313-19.
By Stephen Emerson, M.D., Ph.D.
Non-blood-forming bone marrow stromal or mesenchymal stem cells (MSC) are fascinating cells that are loosely defined by their adherence to plastic. This mixed population of endothelial cells, osteoblasts, fibroblasts, and smooth muscle cells creates the hematopoietic stem cell (HSC) niche, providing the key attachment factors and hematopoietic growth factors that support HSC survival and differentiation. Over the past decade, hematologists have sought to exploit the normal physiologic role of stromal cells in bone marrow transplantation, either to improve the pace and quality of HSC engraftment, or as a cellular vector for gene delivery. Which stromal cell subset contributes either of these abilities is unknown, thus inhibiting the development of stromal (MSC) based therapies. Morad Bensidhoum, Manuel Lopez, and their colleagues have now found that the cell surface protein Stro-1, first discovered by Simmons and Torok-Storb in 1991, identifies a subpopulation of stromal cells which is particularly adept at engrafting in multiple organs after intravenous infusion, although these cells do not themselves augment hematopoietic recovery when combined with umbilical cord blood stem cells. They find that Stro-1- stromal cells, in contrast, do not themselves engraft so well, but are able to accelerate and improve the engraftment of cotransplanted HSCs.
Stromal cells were isolated from bone marrow aspirates by culturing dilute
cell suspensions on plastic tissue culture dishes. After one week, the nonadherent
cells were washed away, and the remaining adherent "MSCs" were separated
into Stro-1+ and Stro-1- cells using anti-Stro-1 Ab. These isolated Stro-1+
and Stro-1- cells were then transplanted into irradiated immunodeficient mice,
either with or without UCB CD34+ HSCs. Their first impressive finding was that
Stro-1- MSCs (including perhaps osteoblasts and endothelial cells) dramatically
increased both the rate and tripled the overall level of human HSC engraftment
in these mice. On the other hand, the Stro-1+ cells had no ability to improve
HSC engraftment.
The simplest explanation for this disparity would have been if Stro-1- cells
are able to engraft along with the HSCs, but the Stro-1+ cells are trapped,
lost, or never engrafted. But this turned out not to be the case. Using sensitive
DNA-based detection techniques, the authors found that both populations were
equally trapped in the lungs, but that Stro-1+ cells engrafted at far higher
rates than Stro-1 cells in all other target organs, including the bone marrow.
These results strongly suggest that the difference in HSC supportive ability
of these populations reflects a real functional difference, not simply a quantitative
or non-specific effect.
Even though Stro-1+ cells may not support HSC engraftment, they could still
be quite useful. The authors show that Stro-1+ MSCs can successfully express
a retrovirally transduced and encoded protein, at least 12 weeks after transplantation.
Thus purified Stro-1+ cells could be very useful for delivering deficient proteins
in inherited disorders such as hemoglobinopathies and glycogen storage diseases.
Overall, the strong message is that each subset of stromal cells is likely
to have unique abilities that can be exploited in graft engineering for safer,
more effective, and more comprehensive therapies.
Return to Top
More Partner Swapping in Sickle Cell Disease
He Z and Russell JE. Antisickling Effects of an Endogenous Human A-Like Globin. Nature Medicine 2004; 10:365-367.
By Nancy Andrews, M.D., Ph.D.
Although its molecular pathology is well understood, sickle cell disease continues to be devastating for many patients. The fundamental problem is a glutamic acid to valine substitution within the β globin polypeptide (βs), which promotes the polymerization of deoxy-α2βs2 hemoglobin tetramers within cells. Resulting hemoglobin fibers distort erythrocyte morphology, producing misshapen, fragile cells that are highly susceptible to hemolysis.
There have been three general approaches to modify the molecular pathology of sickle cell disease. Exploiting observations that increased expression of the fetal version of the β chain (γ globin) ameliorates sickle cell disease in some patients, investigators have sought agents that interfere with the normal genetic switch to shut off γ globin expression. Others have taken a more aggressive approach, using bone marrow transplantation when feasible. Most recently, attempts have been made to hone gene therapy techniques with the goal of forcing erythroid precursors to express normal β chain. Now, in this paper, He and Russell argue that a fourth approach may have been overlooked - reactivation of the embryonic ζ(zeta) globin gene, which is normally shut off early in development.
The investigators started with the observation that a small amount of hemoglobin ζ2βs2 can intervene to destabilize α2βs2 polymers. ζ globin is a close cousin of a globin, differing in a few key amino acids. They took advantage of sickle cell mice engineered to express only human globin proteins, and asked whether the presence of some ζ2βs2 hemoglobin made them better. The sickle cell mice expressing only α2βs2 had several classic features of sickle cell disease including hemolysis, shortened erythrocyte lifespan, reticulocytosis, and defective renal concentration. All of these improved significantly when 36 percent of the circulating hemoglobin was ζ2βs2.
It looks like it helps, but can the ζ gene be re-activated in patients with sickle cell disease? The authors of this paper argue that this may be feasible. They note that ζ is reactivated in some hematological disorders, and that genetic control elements that normally turn it off can be inactivated by small mutations. They propose a combinatorial approach - reactivating some ζ expression and some γ expression to take advantage of the benefits of both.
Return to Top
Autologous Stem Cell Transplantation in the Initial Therapy of Aggressive Lymphoma
Milpied N, Deconinck E, Gaillard F, et al. Initial treatment of aggressive lymphoma with high-dose chemotherapy and autologous stem-cell support. N Engl J Med 2004; 350: 1287-95.
By Michael Williams, M.D.
The role of high-dose chemotherapy and autologous stem cell transplantation (ASCT) as a component of front-line treatment for aggressive non-Hodgkin lymphomas remains controversial. This important issue is addressed in this trial, in which the authors randomized 197 patients ages 15-60 to eight cycles of standard CHOP chemotherapy (cyclophosphamide, doxorubicin, vincristine, prednisone) versus two cycles of CEEP (cyclophosphamide, epirubicin, vindesine, prednisone) plus one cycle each of high-dose methotrexate/cytarabine and BEAM (BCNU, etoposide, cytarabine, melphalan) followed by ASCT. Virtually all patients in each group were in the low-intermediate or high-intermediate risk groups of the clinically-based International Prognostic Index (IPI); IPI high-risk patients were excluded from this trial. Their results showed a benefit for event-free survival (EFS) for the ASCT group by intention-to-treat analysis (Figure 1); overall survival (OS) did not differ significantly at five years, although this may be explained in part by the use of rescue therapy including ASCT for relapsing patients in the CHOP group. When outcomes were assessed by IPI risk groups, a significant improvement in both EFS and OS was observed for the subset of IPI high-intermediate risk patients (Figure 2), whereas there was no significant difference for the two treatment approaches for IPI low- or low-intermediate-risk patients.
The benefit observed for
ASCT in this study is encouraging, although it remains unclear whether the
differences would be less by utilizing the current standard therapy, CHOP plus
rituximab, as the control arm. The
important U.S. and Canadian Intergroup S9704 Phase III trial of CHOP versus CHOP
plus ASCT is in progress for patients 15-66 years of age with IPI
high-intermediate- and high-risk aggressive NHL. This trial was amended in 2003 by adopting CHOP-R into both
the control and ASCT arms. Taken
together, several (but not all) trials over the past decade have suggested a
benefit for high-dose therapy and ASCT, especially in IPI higher-risk patients.
The challenge for laboratory and clinical investigators will be to define
more reliably the markers of risk and response, distinguishing those aggressive
NHL patients most likely to be cured by CHOP-R or equivalent induction therapy
from those destined to fail and who thus might benefit from dose-intensive and
ASCT treatment strategies. Evolving
approaches to such prognostic stratification include the use of cDNA microarray
and immunophenotypic analysis1,2, gene expression analysis3,
and the ability to achieve a negative PET scan early in the course of
chemotherapy treatment4.
References:
- Hans et al. Blood 2004; 103:275.
- The Hematologist, January/February/
March 2004; Vol 1, Issue 1, p. 12.
- Lossos et al. NEJM 2004; 350:1828.
- Spaepen et al. Ann Oncol 2003; 13:1356.
Return to Top
Eosinophils and Asthma: the Case Thickens
Cho JY, Miller M, Baek KJ, Han JW, Nayar J, Lee SY, McElwain K, McElwain S, Friedman S, Broide DH. Inhibition of airway remodeling in IL-5-deficient mice. J Clin Invest 2004 Feb; 113:551-560.
By Stephen Emerson, M.D., Ph.D.
In this paper in the Journal of Clinical Investigation, the case that eosinophils
are of primary importance in allergic asthma was strengthened. In addition
to the acute pathologic manifestations of asthma, patients with the disorder
progressively develop airway wall thickening, smooth muscle hypertrophy, and
fibrosis. Increasing evidence points to an important role for transforming
growth factor (TGF)β1 in this process. Eosinophils are a potent source
of TGFb1, prompting Cho and colleagues to reexamine the role of eosinophil-derived
cytokines in the chronic phase of allergic asthma. By studying long-term airway
effects in a repetitive administration model of ovalbumin-induced allergic
airways disease in normal and interleukin-5 (IL-5) null mice, these investigators
established that IL-5, eosinophils, and TGFβ1 were involved in the collagen
deposition, bronchial smooth muscle hypertrophy, and epithelial mucus expression
characteristic of this experimental model, one that closely mimics the chronic
phase of airways remodeling found in patients with allergic asthma.
Asthma is a chronic and recurring inflammatory airway condition, usually but
not always induced by allergic stimuli, and characterized pathologically by
prominent tissue accumulation of eosinophils and other leukocytes, edema, and
bronchospasm in the acute setting. After repeated attacks, peribronchial fibrosis,
smooth muscle hypertrophy, and the deposition of collagen and other elements
of a fibrotic response inevitably appear. For over a century the blood and
tissue eosinophil has been considered both friend and foe, mediating the elimination
of tissue helminthic and other invasive infectious disorders, but also found
at sites of allergic reactions and asthma. Our understanding of the generation
and trafficking of eosinophils has grown greatly over the past 20 years. The
cloning of IL-5 in 1987 led to its immediate recognition as a primary regulator
of the production, functional priming, and trafficking of blood and tissue
eosinophils1. More recently, the identification of the chemokine eotaxin, a
potent eosinophil chemoattractant, in experimental animal models of allergic
asthma2 reinforced the hypothesis that eosinophils play a central role in the
human disorder, and provided a mechanism for their recruitment. Additional
evidence supporting a role of these two Th2 lymphocyte-derived molecules in
asthmatic bronchitis was provided by studies in an ovalbumin murine model of
reactive airways disease; most, but not all, studies of antibody mediated inhibition
of IL-5 reduced airway hyperresponsiveness, and the additional elimination
of eotaxin further reduced this response. However, studies with a neutralizing
IL-5 antibody failed to significantly affect bronchospasm in patients enrolled
in a recent clinical trial, calling into question the role of the cytokine
and eosinophils in human allergic airways disease.
Thus, these investigators have moved the link between eosinophils and the
chronic bronchial wall thickening characteristic of human allergic asthma one
notch tighter. Although inhibition of eosinophils in a human clinical trial
failed to favorably affect the clinical disease, it is likely that inadequate
numbers of patients, inadequate suppression of airway eosinophils, and the
relatively advanced stage at which any intervention must take place in a clinical
trial of well established asthma all conspired to obscure any effects of anti-eosinophil
therapy, forcing a reexamination of the potential of anti-IL-5 and/or eotaxin
therapeutic approaches to this common and insidious disease.
References:
-
Campbell HD, Tucker WQJ, Hort Y, Martinson ME, Mayo G, Clutterbuck EJ, Sanderson CJ, Young IG: Molecular cloning, nucleotide sequence, and expression of the gene encoding human eosinophil differentiation factor (interleukin 5). Proc Natl Acad Sci USA 84:6629-6633, 1987.
-
Jose PJ, Griffiths-Johnson DS, Collins PD, Walsh DT, Moqbel R, Totty NF, Truong O, Hsuan JJ, Williams TJ: Eotaxin: A potent eosinophil chemoattractant cytokine detected in a guinea pig model of allergic airways inflammation. J Exp Med 179:881-887, 1994.
Return to Top
IS Intravenous Iron Really Better?
Auerbach M, Ballard H, Trout JR, et al. Intravenous iron optimizes the response to human erythropoietin in cancer patients with chemotherapy-related anemia: A multi-center, open-label, randomized trial. J Clin Oncol 2004; 22:1301-1307.
By J. Douglas Rizzo, M.D.
Iron is essential to erythropoiesis. Erythropoietin usage has increased for chemotherapy-induced anemia, sparking interest in the role of iron supplementation for patients who demonstrate poor response. Less allergenic intravenous (IV) iron preparations have been proposed as a better mechanism to circumvent the limitations in oral iron absorption, transport, and utilization for this form of anemia. This study involved a randomized trial to address this important question. 157 patients with chemotherapy-related anemia were randomized to treatment with weekly epoetin (40,000 units subcutaneously) together with no iron replacement, oral ferrous sulfate twice daily, or one of two IV iron dextran treatment strategies. Efficacy, as determined by rise in hemoglobin (Hgb) or hematopoietic response, was reported to be statistically significantly higher in the IV iron replacement groups compared to no iron or oral ferrous sulfate groups. Mean differences in Hgb improvement between the oral and IV iron groups were about 10 g/L, with 68 percent of patients in the IV iron groups achieving a Hgb level >120 g/L compared to 36 percent of patients receiving oral iron and 25 percent of patients receiving no iron. The authors conclude that IV iron augments erythropoiesis better than oral iron.
Some limitations compromise the validity of the conclusions. Randomization was not stratified for disease, disease status, or amount or intensity of prior or current chemo- and radiation therapy, and differences between groups in these factors were not described, formally tested, or adjusted in the analysis. Length of anemia therapy was different in the treatment groups, as was time to response assessment. If, despite the limitations in study design, one accepts the superior erythropoietic response to IV iron, these conclusions may stimulate dialogue on other relevant issues. First, the entity of functional iron deficiency (FID) requires further study and more precise definition. The best laboratory parameters for guidance of iron therapy in functional iron deficiency require elucidation. Use of traditional measures, including serum ferritin and iron saturation, are limited by their status as acute phase reactants. Data regarding usefulness of reticulocyte Hgb content and serum soluble transferrin receptors are incomplete and conflicting. Will hepcidin expression become a useful diagnostic tool for anemia of chronic disease or FID? With more precise definition of FID and laboratory parameters to follow therapy, patients could be stratified by iron requirements and iron "accessibility" to oral versus IV supplementation based on need. Second, would a more bioavailable oral iron supplement with fewer side effects, heme-iron polypeptide, achieve the same results as IV iron more conveniently? Third, though no clinical data have demonstrated development of iron excess with IV iron supplementation in this setting, a prudent approach to its use (and study for long-term consequences) would be necessary if it becomes widely adopted. If the administration rate or frequency exceeds utilization for erythropoiesis, hemochromatosis may develop. Finally, although IV iron may offer promise to patients whose response to erythropoietic stimulation is inadequate, cost-effectiveness remains to be determined by comparing all relevant treatment approaches.
Return to Top
Defending Against the Attack in PNH
Hillmen P, Hall C, Marsh J, Elebute M, Bombara M, Petro B, Cullen M, Richards S, Rollins S, Mojcik C, Rother R. Effect of Eculizumab on Hemolysis and Transfusion Requirements in Patients with Paroxysmal Nocturnal Hemoglobinuria. N Engl J Med 2004; 350: 552-9.
By Lilli Petruzzelli, M.D., Ph.D.
Paroxysmal nocturnal hemoglobinuria (PNH) is associated with an array of clinical sequelae including intravascular hemolysis, thrombosis, hemoglobinuria, and bone marrow failure. Loss of expression of CD55 and CD59 on the cell surface, because of the inability to synthesize critical glycosylphosphatidylinositol (GPI) linkages, results from mutations in the PIG-A gene and is associated with complement-mediated lysis of erythrocytes. CD55 inhibits C3 convertase and CD59 inhibits the assembly of the membrane attack complex C5b-C9 by interacting with C8 and C9. In this report, the investigators utilized Eculizumab, an antibody that blocks the formation of C5b, to reduce hemolysis. Eleven patients with PNH received 600 mg of Eculizumab weekly for four weeks followed by 900 mg every two weeks for 12 weeks. All patients tolerated Eculizumab, and noted a marked improvement in their overall well-being. Hemolytic activity was reduced as measured by a number of parameters: a decrease in LDH to near normal levels after one dose of antibody, an increase in the relative proportion of Type III PNH erythrocytes, a decrease in transfusion requirements, and a decrease in hemoglobinuria. Of note is that the hemoglobin levels and reticulocyte count were not significantly increased in these patients. The serum level of Eculizumab was maintained at or above 35 µg/ml in all but one patient; in the latter, the fall was associated with increased hemolysis and resolved when the level was above 35 µg/ml.
Despite a detailed understanding of the clinical and molecular events that are associated with PNH, treatment options for these patients are limited. Bone marrow transplantation has remained the mainstay of therapy, and targeting the PIG-A gene to hematopoietic precursors remains in the future. Rather than targeting the molecular basis of this disease, these investigators have directed therapy at the debilitating hemolysis associated with PNH. Interestingly, although hemoglobin levels did not change significantly in this study, treatment with Eculizumab resulted in a marked reduction in the need for transfusions when compared to the pretreatment requirements. Although both CD55 and CD59 play a role in inhibiting complement-mediated erythrocyte lysis, the results support the model that it is the loss of the inhibitory effect of CD59 on the C5b-C9 complex that is critical in the hemolytic process of PNH.
The duration of response to therapy will be of great interest and the enrollment of the entire patient population into a long-term study is poised to address this issue. This may also enable us to learn whether modulation of the C5b-C9 complex is associated with a change in other events seen in PNH, such as thrombosis and hemorrhage; it is likely that a larger cohort of patients will be needed to address whether there is an effect on these other parameters. At the least, this treatment offers the promise of modulating hemolysis and reducing the risk of iron overload from transfusions, and may serve as a bridge to more aggressive therapy. The long-term benefits, tolerance, fate of the stem cell clone, and requirements for maintaining levels of the drug remain questions for the future.
Return to Top
AML and Microarrays: No Longer Keeping a Low Profile
Bullinger L, Dohner K, Bair E, Frohling S, Schlenk RF, Tibshirani R, Dohner H, Pollack JR. Use of gene-expression profiling to identify prognostic subclasses in adult acute myeloid leukemia. N Engl J Med. 2004; 350:1605-1616.
Valk PJ, Verhaak RG, Beijen MA, Erpelinck CA, Barjesteh van Waalwijk van Doorn-Khosrovani S, Boer JM, Beverloo HB, Moorhouse MJ, van der Spek PJ, Lowenberg B, Delwel R. Prognostically useful gene-expression profiles in acute myeloid leukemia. N Engl J Med. 2004; 350:1617-1628.
By Robert Lowsky, M.D., FRCPC
Should patients with AML achieving a complete remission following induction therapy receive consolidation chemotherapy, or consider autologous or allogeneic transplantation? Current decision algorithms are predominantly based on patient age, history of an antecedent hematologic disorder, and the presence or absence of cytogenetic findings. Yet, despite these prognosticators, considerable heterogeneity in outcomes remain, thereby confounding treatment decisions and recommendations. The April 16, 2004, issue of the New
England Journal of Medicine includes two articles that report the use of genomics (microarray gene expression profiling) in AML to comprehensively cluster this disease into new distinct subgroups containing unique molecular signatures, each with predictable clinical outcomes.
In one report, Bullinger et al. isolated blasts from 116 patients with AML and evaluated 26,000 genes for their expression levels using microarrays. Approximately 6000 genes were informative, enabling identification of new subgroups of AML. Computer-enhanced statistical methods revealed 133 highly predictive genes which, depending on expression profiles, correlated with either good or poor outcome. In a companion paper, Valk et al. analyzed blasts from 285 AML patients and evaluated expression profiles among 13,000 unique genes, 21 percent being informative. Sixteen subgroups of AML were identified on the basis of molecular signatures, and each was correlated with clinical outcomes allowing identification of poor, intermediate, and favorable risk cluster groups. Taken together, these papers highlight that gene expression profiling can uncover novel subclasses of AML and predict clinical outcomes that cannot be otherwise identified by conventional cytogenetics or other clinical variables.
These studies bring the promise of a new classification scheme for AML that will ultimately refine predictions of clinical outcomes, improve treatment decisions, and promote discoveries of targeted therapies. Yet, before exportation to clinical practice can be realized, exciting challenges remain. Reproducibility, accessibility, and standardization are not simple matters to resolve and require working groups and consensus conferences. Bullinger et al. and Valk et al. used different microarray platforms with different reference messenger RNA and different statistical software packages. Although concordance in some of their findings was observed, fundamental differences were also seen. For example, cases with t(8;21) and inv(16) were each found to separate into multiple clusters with distinct molecular signatures in one study but not in the other. Moreover, ambitious efforts to compare gene expression profiles between the hierarchy of leukemic stem cells and their non-malignant counterparts will aid the development of novel targeted therapies that disrupt pathways of self-renewal and arrest leukemia at its source. It remains unclear how long before the power of genome-wide analysis teases apart the molecular basis of leukemia. However, one thing is certain: the gene expression signatures that underlie the malignant transformation and progression of the various subtypes of AML will be exposed and AML will no longer be able to keep a low profile.
Return to Top
Return to Table of Contents
|
