This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Li, Q. Articles by Issa, J.-P. J. Search for Related Content PubMed PubMed Citation Articles by Li, Q. Articles by Issa, J.-P. J.

Estrogen Receptor Methylation Is Associated with Improved Survival in Adult Acute Myeloid Leukemia¹

The Johns Hopkins Oncology Center, Baltimore, Maryland 21231 [Q. L., A. M., J-P. J. I.]; University of New Mexico School of Medicine, Albuquerque, New Mexico 87131 [C. L. W.]; Southwest Oncology Group Statistical Center, Seattle, Washington 98104; and Southwest Oncology Group Leukemia and Leukemia Programs, San Antonio, Texas 78245 [K. J. K., C. L. W., F. R. A., J. K. W.]

	ABSTRACT

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Estrogen receptor methylation (ERM) is a frequent molecular alteration in adult acute myeloid leukemia (AML). In this study, we sought to determine the clinical characteristics and prognostic significance of ERM in AML. ERM was determined for 268 patients who had leukemic blasts available for molecular analysis. ERM was measured by Southern blot analysis, and results were obtained for 261 patients (ages 17–69). ERM ranged from 0–99.1%, with a median of 25%. One hundred sixty patients (61%) had ERM values over 15% and were considered ERM+. In a subset of patients analyzed, ERM+ samples had markedly lower ER gene expression compared with ERM- samples. In multiple regression analyses of patient and disease characteristics at diagnosis, two factors had significant independent association with ERM: ERM decreased with increasing age (P = 0.0001) and was significantly lower in patients with French-American-British classification M4 or M5 (P = 0.0019). In regression analyses of outcome measures, ERM had no significant impact on complete remission rate after initial induction therapy. However, ERM+ patients had significantly better overall survival [OS; 18% at 6 years; 95% confidence interval (CI), 12–24% versus 9%; CI, 3–14% for ERM- patients; P = 0.022]. In multiple regression analyses, OS increased with increasing ERM (P = 0.0044). Similar results were seen for relapse-free survival (23% at 6 years; CI, 15–32% for ERM+ versus 10%; CI, 2–19% for ERM-), although the effect of ERM was not statistically significant (P = 0.15 in multiple regression analysis). Our results indicate that ERM at diagnosis may be a favorable prognostic factor for OS in adult AML.

	INTRODUCTION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Cytosine methylation within the normally unmethylated promoters of selected genes has recently emerged as one of the most constant molecular changes in neoplasia, in general (1 , 2) , and leukemias, in particular (3) . Such promoter methylation seems to be an alternate mechanism to coding region mutations in inactivating tumor-suppressor and other genes in neoplasia. The P16 gene, for example, has recently been shown to be a tumor-suppressor gene for which promoter methylation and transcriptional inactivation are relatively common in multiple types of neoplasms (4, 5, 6, 7, 8) . The ER³ gene is another locus for which aberrant methylation has been described in multiple tumors, including hematopoietic malignancies (3) .

The causes of aberrant methylation in cancer and the clinical determinants associated with such methylation have not been well defined. Observations in an animal model of lung cancer have suggested that, for the ER gene, methylation was partially dependent on the carcinogenic insult that induced tumor formation (9) . In this model, spontaneous and radiation-induced tumors were much more likely to be methylated than tumors induced by the tobacco-derived carcinogen, NNK. These data suggested the possibility that ERM may be associated with specific clinical characteristics in human tumors, as well. In addition, the high frequency of promoter methylation at selected loci in cancer has raised the possibility that, similar to other molecular defects such as mutations, loss of heterozygosity, aneuploidy, or karyotypic abnormalities, this readily detectable DNA change could also be used as a prognostic factor and to supplement conventional clinical staging (3) .

To explore these various possibilities, we have studied ERM in a well characterized and uniformly treated population of patients with AML. Previous data had indicated that ERM was very low (<10%) in normal peripheral blood or bone marrow-derived cells, but relatively common (70%) in patients with AML (10) . We now show that ERM in AML is associated with specific clinical parameters, such as age and FAB classification, and is an independent favorable prognostic factor for survival in this disease.

	MATERIALS AND METHODS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patients.
S8600 compared high-dose versus standard dose 1-ß-D-arabinofuranosylcytosine with daunorubicin in adult patients with previously untreated AML, including both de novo AML and secondary AML (11) . S8600 was largely limited to patients <65-years-old, although a small number of older patients participated in an early pilot phase of the study. The results of S8600 have been reported (11) . As part of the study protocol, pretreatment leukemic blasts derived from either peripheral blood or bone marrow were stored for future studies. For the present study, we selected all patients with centrally confirmed diagnoses of AML who had at least three vials of stored blasts available for analysis. Of 774 patients on S8600 with confirmed AML, 267 fulfilled this criterion. One additional patient who received the S8600 standard dose induction regimen on a different SWOG trial was also included in this study. In both trials, written informed consent in accordance with institutional and Department of Health and Human Services policies was obtained from all patients.

Measurement of ERM.
DNA was extracted from mononuclear cell fractions using standard techniques. As in previous studies, Southern blots were used to determine the methylation state of the ER CpG island (10) . This island contains a cluster of methylation-sensitive restriction enzymes. We used one of these enzymes, NotI, to study the methylation state of the island. NotI will digest DNA to completion if the two CG sites in its recognition sequence are unmethylated, but will not cut DNA if either of the two CG sites are methylated. Briefly, 5 µg of genomic DNA were digested with 50 units of EcoRI and 100 units of NotI for 16 h as specified by the manufacturer (New England Biolabs), run on a 1% agarose gel, transferred to a Zetaprobe nylon membrane (Bio-Rad), and probed with a 428-bp fragment of the first exon of the ER gene obtained by PCR amplification from normal genomic DNA using primers AACCTCGGGCTGTGCTCTTTTTCCAG (upper) and AGTAGCATCAGCGGGCTCGGAGACAC (lower). The Southern blots were then exposed on a phosphor screen for 2–3 days and developed using a Phosphorimager (Molecular Dynamics). For quantification of ERM, the relative density of the methylation band at 3.1 kb was measured using the ImageQuant software (Molecular Dynamics) and expressed as a percentage of the density of all bands (3.1 kb, 1.9 kb, and 1.2kb) in each lane. To rule out incomplete digestion with NotI, all of the blots were reprobed with a 5' fragment of the c-abl gene, which contains two NotI sites in its 5' CpG island. In all cases, an expected 5-kb band was the only band present, indicating complete digestion of the DNA with NotI.

RT-PCR.
Total RNA was prepared using standard techniques. Total RNA (6 µg ) was used to generate cDNA using random hexamers and M-MuLV reverse transcriptase enzyme, as recommended by the manufacturer (Boehringer Mannheim). About one-tenth of the cDNA was used as a template to amplify a 430-bp fragment specific to the ER gene (10) . Another one-thirtieth of cDNA product was used to amplify a 306-bp fragment specific to the GAPDH gene transcript as a control for RNA integrity. The primers used are: CGGAGTCAACGGATTTGGTCGTAT (upper) and AGCCTTCTCCATGGTGGTGAAGAC (lower) for GAPDH. All amplifications were performed at least twice, and all included a positive control consisting of mRNA from the ER-positive breast cancer cell line ZR75 and a negative control consisting of mRNA from an ER-negative colon cancer cell line, RKO. Reverse transcription (-) controls, where the reverse transcriptase enzyme was omitted, were also used for each sample.

Statistical Analysis.
Demographic, clinical, and outcome data for all patients were collected and evaluated according to standard procedures of the SWOG. Effects of patient and disease characteristics on ERM were examined using least squares regression analysis, two-sample t tests, and ANOVA. Preliminary examination indicated that the ERM values had a rather skewed distribution. Therefore, to more nearly satisfy the assumption of a normal (Gaussian) distribution under which least squares analysis is optimal, the ERM values were transformed to logits: logit(ERM) = log(ERM) - log(100 - ERM). Logits, which are often appropriate when the original data are proportions, proved to have a much more nearly Gaussian distribution. The effect of ERM on the probability of achieving CR was examined using Fisher’s exact test and logistic regression analysis (12) . OS was defined as the number of days from entry into the SWOG clinical trial until death from any cause, with observation censored at the date of last contact for patients last known to be alive. RFS was defined for patients who achieved CR as the number of days from establishment of CR until relapse or death from any cause, whichever occurred first, with observation censored at the date of last contact for patients last known to be alive without report of relapse. Distributions of OS and RFS were estimated by the method of Kaplan and Meier (13) . The effects of ERM on OS and RFS were examined using PHs regression analysis (14) . Because the skewness of the distribution of untransformed ERM values might inordinately increase the relative influence of the patients with extremely high ERM on the results of logistic or PHs regression analyses, parallel analyses were performed using the untransformed and logit-transformed ERM values. Statistical significance was represented by two-tailed Ps. OS and RFS results were based on data available on December 21, 1998.

	RESULTS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
ERM.
Two hundred sixty-eight patients had leukemic blasts available for analysis. DNA was extracted from the leukemic blasts, and ERM was quantitatively determined using Southern blot analysis (examples in Fig. 1 ). The DNA from seven patients was degraded and unsuitable for ERM determination, leaving 261 cases for analysis. Of these 261 patients, blasts were obtained from the peripheral blood in 152 cases and from the bone marrow in 103 cases. The specimen type was unknown for six patients. As shown in Fig. 2 , the distributions of ERM were somewhat skewed toward lower values, although they were approximately Gaussian after transformation to logits. In the entire group of 261 patients, ERM ranged from 0.3–99.1%, with a median of 24.7%. Because normal peripheral blood or bone marrow-derived cells have previously been shown to have up to 10% ERM (10) , we arbitrarily chose 15% as a cutoff for calling a sample ERM+. In this analysis, 160 patients were ERM+ (61%), which is consistent with our previous findings. Patient characteristics are described in Table 1 . Because S8600 was largely limited to younger patients, all but one were <65-years-old. Table 1 also summarizes the characteristics of 514 patients on S8600 with centrally confirmed diagnoses of AML who did not have ERM data, either because they had less than three vials of cryopreserved blasts (507 patients) or because their DNA was degraded (7 patients). The 261 patients with ERM data were similar in most respects to these 514 patients. They differed primarily in having relatively high WBC and peripheral blast counts, an artifact reflecting their selection on the basis of availability of stored cells: patients with low WBC and blast counts were less likely to have sufficient cells available for this study. FAB categories M6 and M0 were also somewhat less common among the 261 patients with ERM data.

View larger version (67K): [in this window] [in a new window] [Download PPT slide]   Fig. 1. Representative Southern blot of ERM in AML. DNA was digested with the restriction enzymes EcoRI and NotI (which is methylation-sensitive), electrophoresed on an agarose gel, transferred to a nylon membrane, and probed with a fragment of exon 1 of the ER gene. The band at 3.1 kb (arrow) indicates failure to digest with NotI, due to methylation of the NotI recognition site. The relative density of this band is indicated below each lane as percentage of ERM.

View larger version (39K): [in this window] [in a new window] [Download PPT slide]   Fig. 2. Frequency distributions of untransformed and logit-transformed ERM values, by type of specimen.

ERM and ER Gene Expression in AML.
To determine whether differences in ERM status corresponded to potential functional differences among the AML samples, we measured ER expression by RT-PCR in a subset of the cases. Fifteen cases (seven ERM- and eight ERM+) were selected based on the availability of additional stored samples. RNA was degraded in one ERM- sample, leaving 14 cases for analysis. ER expression was very low or absent in all eight ERM+ cases, but present at significant levels in five of six ERM- cases (Fig. 3) . These data suggest that ERM status may have functional significance in AML through differential expression of the ER gene.

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Fig. 3. Expression of ER in AML by ERM status. ER expression was determined by RT-PCR using primers that yield a fragment of 430 bp. Integrity of RNA was verified by RT-PCR using primers specific for the GAPDH gene (which yields a 306-bp fragment). All reactions were controlled using reverse transcription-negative samples where the reverse transcription enzyme was omitted (not shown). ER was expressed in ZR75 (positive control) and ERM- cases (AML1–6), but was very low or absent in RKO (negative control) and ERM+ cases (AML7–14). ERM values are indicated below each sample.

In multiple regression analysis (also including adjustment for specimen type), only two factors retained clearly significant independent associations with ERM: logit(ERM) decreased with increasing age (P = 0.0001) and was significantly lower in patients with M4 or M5 AML (P = 0.0001). The difference between blood and marrow specimens remained marginally significant in this multiple regression model (P = 0.026). The effects of these variables on ERM are demonstrated in Table 2 .

In multiple logistic regression analysis, after accounting for this effect of FAB classification, none of the other variables in Table 1 had a statistically significant independent association with CR. In particular, there was no significant effect of ERM, whether treated as a quantitative variable (P = 0.22 with untransformed ERM and 0.15 with logit of ERM) or dichotomized at 15% (P = 0.088). Because increasing age was associated with a large but nonsignificant effect on CR rate, multiple logistic regression analyses including both FAB and age were also performed; however, these analyses gave essentially the same results regarding the effect of ERM on response. Analyses of interactions produced no evidence that ERM had an effect on the CR rate, which varied according to age (P = 0.45) or ara-C induction dose (P = 0.47). Similar results were obtained in analyses using the logit of ERM.

ERM and OS.
Of the 261 patients with ERM data, 226 have died. Except for one patient lost to follow-up after 40 months, the remaining 35 patients have been under follow-up between 6.0 and 11.9 years (median, 7.7 years). As shown in Table 3 and Fig. 4 , OS was somewhat poorer for ERM- patients (9% at 5 years; 95% CI, 3–14%) compared with ERM+ patients (18%; CI, 12–24%). On the basis of PHs regression analyses, this difference was marginally significant (two-tailed P = 0.022). The estimated hazard ratio (ERM+ relative to ERM-) was 0.73 (CI, 0.56–0.95), indicating that the average mortality rate was an estimated 27% lower on average for ERM+ patients. Similarly, with ERM treated as a quantitative (continuous) variable, the trend in OS was marginally significant (P = 0.024). There was no significant interaction between specimen type and the effect of ERM (P = 0.49). In the parallel analysis based on logit(ERM), the association between OS and ERM was somewhat more significant (P = 0.0042) and, again, the effect of ERM did not differ significantly between the blood and marrow groups (P = 0.51). OS of the ERM+ patients was similar to that of the 514 patients from S8600 without ERM data (Table 3) .

View larger version (20K): [in this window] [in a new window] [Download PPT slide]   Fig. 4. OS by ERM status. The solid line represents ERM-positive (15%) cases, whereas the dashed line represents ERM-negative (<15%) cases. Tickmarks indicate censored observations.

ERM and RFS.
Of the 144 patients who achieved CR, 103 have relapsed and 15 others died without report of relapse. As shown in Table 2 and Fig. 5 , RFS was somewhat poorer for ERM- patients (10% at 5 years; CI, 2–19%) compared with ERM+ patients (23%; CI, 15–32%). In simple PHs regression analyses, this difference was not statistically significant (two-tailed P = 0.060), with an estimated hazard ratio of 0.60 (CI, 0.48–1.01). Similarly, with ERM treated as a quantitative (continuous) variable, the trend in RFS was not significant (P = 0.087 based on untransformed ERM; P = 0.085 with logit of ERM). RFS of patients with ERM data was similar to that of those without ERM data.

View larger version (19K): [in this window] [in a new window] [Download PPT slide]   Fig. 5. RFS by ERM status. The solid line represents ERM-positive (15%) cases, whereas the dashed line represents ERM-negative (<15%) cases. Tickmarks indicate censored observations.

	DISCUSSION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Our results confirm that ERM is frequently present in AML at diagnosis and is associated with repressed ER gene transcription in vivo. These data also suggest potentially important interactions between this molecular marker, clinical characteristics, and OS.

Among the clinical parameters that affect ERM in AML, the inverse correlation between age and ERM stands out as the most intriguing. Previous studies had suggested that aging is associated with an increase in ERM in normal appearing colonic epithelium (15) and other tissues.⁴ Thus, we would have predicted that ERM should be more frequent in AML in the elderly. On the other hand, we were unable to detect a clear association between aging and ERM in normal hematopoietic cells (10) , and it is possible that the relationship between aging, ERM and neoplasia is not conserved between epithelial cells and hematopoietic cells. One possible explanation for the present findings then, may reside in the triggering factors that lead to AML. As discussed earlier, we have previously shown that, in an animal model of lung cancer, tumors induced by radiation exposure were associated with a higher rate of ERM than tumors induced by a tobacco-derived carcinogen (9) . Furthermore, in glioblastoma multiforme, there is a strong correlation between methylation at the ER locus and methylation at N33 (16) , a distinct gene on a different chromosome (17) , suggesting again that ERM may reflect a specific pathophysiological event that leads to hypermethylation at multiple loci. Thus, it is possible that the presence or absence of ERM in AML mark different pathways in the molecular pathophysiology of this disease. If this hypothesis is correct, the different rates of ERM among young and old patients with AML may, therefore, reflect differing exposures to carcinogenic insults. Future studies should clarify the nature and potential causes of this inverse association between age and ERM.

Perhaps most interesting is the independent association between ERM and a relatively favorable prognosis in AML. OS increased significantly with increasing ERM, even after adjustment for other significant prognostic factors, including age (P = 0.0032). Because both age and ERM were significantly associated with survival in this multiple regression analysis, ERM is not simply serving as a marker for age or age-related factors. Although the association of ERM with RFS was not statistically significant, the magnitude of the effect of ERM was similar for OS and RFS: the estimated hazard ratios (ERM+ relative to ERM-) were 0.71 (CI, 0.54–0.93) for OS and 0.67 (CI, 0.46–0.97) for RFS.

This, of course, is not the first time that a molecular abnormality has been associated with a good prognosis in acute leukemias. Indeed, several specific chromosomal translocations have previously been shown to be associated with a favorable prognosis in AML and acute lymphocytic leukemia (18 , 19) . Nevertheless, the mechanism of this association between ERM and survival remains obscure. We have shown that AML cases without ERM express relatively high levels of ER, and it is, therefore, possible that this expression itself is detrimental in AML. This possibility deserves further investigation because if ER expression is indeed responsible for this poorer outcome, than the use of ER antagonists such as Tamoxifen may be warranted in this subset of patients. Another possible explanation for the difference in outcome based on ERM status goes back to the hypothesis mentioned above that ERM might simply be a molecular marker for the carcinogenic event(s) that resulted in leukemia formation. Interestingly, a somewhat analogous situation has recently emerged in colorectal cancer, where tumors with mismatch repair deficiency appear to have more pronounced methylation defects, in association with a more favorable prognosis when compared with tumors proficient in mismatch repair (20) . There are also other situations wherein, within a tumor type, different pathophysiological pathways are associated with differences in survival. For example, ovarian neoplasms in patients with germline BRCA1 mutations have a better prognosis than sporadic ovarian cancers (21) .

There are two important limitations of our study that should be mentioned. First, because cytogenetics was not mandatory for patient registration on this clinical study initiated in 1986, we do not have reviewed high quality cytogenetic data on the majority of patients. Second, there is a relative lack of information regarding whether each patient developed AML after an antecedent hematological disorder and/or other factors linked to the development of secondary leukemia. This latter information may be particularly important because secondary leukemias tend to occur in older patients and are associated with a poor prognosis (22 , 23) . Although there is no a priori reason to suspect a particular link between ERM and specific chromosomal anomalies or secondary leukemias, this issue should be addressed in detail in future studies.

Whatever the mechanism of this association between ERM and better outcome in AML, our data has potentially important clinical implications. There are few clinical parameters of prognosis in AML that retain a predictive value in patients who achieve a CR (19 , 24) . These results, if confirmed in separate populations of patients, suggest the possibility that lack of ERM at diagnosis may mark a population of patients with such a poor outcome that alternate forms of treatment should be considered early in their management.

	ACKNOWLEDGMENTS

 
We thank Dr. Stephen B. Baylin for reviewing the manuscript and helpful comments.

	FOOTNOTES

 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ Supported by a Leukemia Society of America Translational Research Grant, and NIH Grants 5RO1CA43318, CA38926, CA12213, and U01 CA32102 (supporting the Southwest Oncology Group Leukemia and Leukemia Biology Programs). J-P. J. I. is a Kimmel Foundation Scholar.

² To whom requests for reprints should be addressed, at The Johns Hopkins Oncology Center, 424 North Bond Street, Baltimore, MD 21231. Phone: (410) 955-8506; Fax: (410) 614-9884; E-mail: jpissa{at}welchlink.welch.jhu.edu

³ The abbreviations used are: ER, estrogen receptor; ERM, ER methylation; AML, acute myeloid leukemia; SWOG, Southwest Oncology Group; S8600, SWOG study 8600; CR, complete response; OS, overall survival; PH, proportional hazard; RFS, relapse-free survival; RT-PCR, reverse transcription-PCR; CI, confidence interval; FAB, French-American-British; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

⁴ J-P. J. Issa et al., unpublished observations.

Received 8/14/98; revised 1/29/99; accepted 2/ 1/99.

	REFERENCES

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Baylin S. B., Herman J. G., Graff J. R., Vertino P. M., Issa J-P. J. Alterations in DNA methylation—a fundamental aspect of neoplasia. Adv. Cancer Res., 72: 141-196, 1998.[Medline]
2. Jones P. A. DNA methylation errors and cancer. Cancer Res., 56: 2463-2467, 1996.[Free Full Text]
3. Issa J-P. J., Baylin S. B., Herman J. G. DNA methylation changes in hematologic malignancies: biologic and clinical implications. Leukemia (Baltimore), 11: S7-S11, 1997.
4. Gonzalez-Zulueta M., Bender C. M., Yang A. S., Nguyen T., Beart R. W., Van Tornout J. M., Jones P. A. Methylation of the 5' CpG island of the p16/CDKN2 tumor suppressor gene in normal and transformed human tissues correlates with gene silencing. Cancer Res., 55: 4531-4535, 1995.[Abstract/Free Full Text]
5. Herman J. G., Merlo A., Mao L., Lapidus R. G., Issa J-P., Davidson N. E., Sidransky D., Baylin S. B. Inactivation of the CDKN2/p16/MTS1 gene is frequently associated with aberrant DNA methylation in all common human cancers. Cancer Res., 55: 4525-4530, 1995.[Abstract/Free Full Text]
6. Wong D. J., Barrett M. T., Stoger R., Emond M. J., Reid B. J. p16^INK4a promoter is hypermethylated at a high frequency in esophageal adenocarcinomas. Cancer Res., 57: 2619-2622, 1997.[Abstract/Free Full Text]
7. Martinez-Delgado B., Fernandez-Piqueras J., Garcia M. J., Arranz E., Gallego J., Rivas C., Robledo M., Benitez J. Hypermethylation of a 5' CpG island of p16 is a frequent event in non-Hodgkin’s lymphoma. Leukemia (Baltimore), 11: 425-428, 1997.[Medline]
8. Batova A., Diccianni M. B., Yu J. C., Nobori T., Link M. P., Pullen J., Yu A. L. Frequent and selective methylation of p15 and deletion of both p15 and p16 in T-cell acute lymphoblastic leukemia. Cancer Res., 57: 832-836, 1997.[Abstract/Free Full Text]
9. Issa J-P. J., Baylin S. B., Belinsky S. A. Methylation of the estrogen receptor CpG island in lung tumors is related to the specific type of carcinogen exposure. Cancer Res., 56: 3655-3658, 1996.[Abstract/Free Full Text]
10. Issa J-P. J., Zehnbauer B. A., Civin C. I., Collector M. I., Sharkis S. J., Davidson N. E., Kaufmann S. H., Baylin S. B. The estrogen receptor CpG island is methylated in most hematopoietic neoplasms. Cancer Res., 56: 973-977, 1996.[Abstract/Free Full Text]
11. Weick J. K., Kopecky K. J., Appelbaum F. R., Head D. R., Kingsbury L. L., Balcerzak S. P., Bickers J. N., Hynes H. E., Welborn J. L., Simon S. R., Grever M. A randomized investigation of high-dose versus standard-dose 1-ß-D-arabinofuranosylcytosine with daunorubicin in patients with previously untreated acute myeloid leukemia: a Southwest Oncology Group study. Blood, 88: 2841-2851, 1996.[Abstract/Free Full Text]
12. Cox D. R. Analysis of Binary Data Chapman and Hall London 1970.
13. Kaplan E. L., Meier P. Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc., 53: 457-481, 1958.
14. Cox D. R. Regression models and life-tables. J. R. Stat. Soc. B, 34: 187-220, 1972.
15. Issa J-P., Ottaviano Y. L., Celano P., Hamilton S. R., Davidson N. E., Baylin S. B. Methylation of the oestrogen receptor CpG island links aging and neoplasia in human colon. Nat. Genet., 7: 536-540, 1994.[Medline]
16. Li Q., Jedlicka A., Ahuja N., Gibbons M. C., Baylin S. B., Burger P. C., Issa J-P. J. Concordant methylation of the ER and N33 genes in glioblastoma multiforme. Oncogene, 16: 3197-3202, 1998.[Medline]
17. MacGrogan D., Levy A., Bova G. S., Isaacs W. A., Bookstein R. Structure and methylation-associated silencing of a gene within a homozygously deleted region of human chromosome band 8p22. Genomics, 35: 55-65, 1996.[Medline]
18. Rowley J. D., Aster J. C., Sklar J. The clinical applications of new DNA diagnostic technology on the management of cancer patients. J. Am. Med. Assoc., 270: 2331-2337, 1993.[Medline]
19. Buchner T., Heinecke A. The role of prognostic factors in acute myeloid leukemia. Leukemia (Baltimore), 10 (Suppl. 1): S28-S29, 1996.
20. Ahuja N., Mohan A. L., Li Q., Stolker J. M., Herman J. G., Hamilton S. R., Baylin S. B., Issa J. P. Association between CpG island methylation and microsatellite instability in colorectal cancer. Cancer Res., 57: 3370-3374, 1997.[Abstract/Free Full Text]
21. Rubin C., Benjamin I., Behbakht K., Takahashi H., Morgan M., LiVolsi V., Berchuck A., Muto M., Garber J., Weber B., Lynch H., Boyd J. Clinical and pathological features of ovarian cancer in women with germ-line mutations of Brca1. N. Engl. J. Med., 335: 1413-1416, 1996.[Abstract/Free Full Text]
22. Vogel V. G., Fisher R. E. Epidemiology and etiology of leukemia. Curr. Opin. Oncol., 5: 26-34, 1993.[Medline]
23. Smith M. A., McCaffrey R. P., Karp J. E. The secondary leukemias: challenges and research directions. J. Natl. Cancer Inst., 88: 407-418, 1996.[Abstract/Free Full Text]
24. Kantarjian H. M., O’Brien S., Anderlini P., Talpaz M. Treatment of myelogenous leukemia: current status and investigational options. Blood, 87: 3069-3081, 1996.[Free Full Text]

This article has been cited by other articles:

				J. C. Baker Jr., J. H. Ostrander, S. Lem, G. Broadwater, G. R. Bean, N. C. D'Amato, V. K. Goldenberg, C. Rowell, C. Ibarra-Drendall, T. Grant, et al. ESR1 Promoter Hypermethylation Does Not Predict Atypia in RPFNA nor Persistent Atypia after 12 Months Tamoxifen Chemoprevention Cancer Epidemiol. Biomarkers Prev., August 1, 2008; 17(8): 1884 - 1890. [Abstract] [Full Text] [PDF]

				L. Shen, Y. Kondo, S. Ahmed, Y. Boumber, K. Konishi, Y. Guo, X. Chen, J. N. Vilaythong, and J.-P. J. Issa Drug Sensitivity Prediction by CpG Island Methylation Profile in the NCI-60 Cancer Cell Line Panel Cancer Res., December 1, 2007; 67(23): 11335 - 11343. [Abstract] [Full Text] [PDF]

				W. Blum, R. B. Klisovic, B. Hackanson, Z. Liu, S. Liu, H. Devine, T. Vukosavljevic, L. Huynh, G. Lozanski, C. Kefauver, et al. Phase I Study of Decitabine Alone or in Combination With Valproic Acid in Acute Myeloid Leukemia J. Clin. Oncol., September 1, 2007; 25(25): 3884 - 3891. [Abstract] [Full Text] [PDF]

				K. Mrozek, G. Marcucci, P. Paschka, S. P. Whitman, and C. D. Bloomfield Clinical relevance of mutations and gene-expression changes in adult acute myeloid leukemia with normal cytogenetics: are we ready for a prognostically prioritized molecular classification? Blood, January 15, 2007; 109(2): 431 - 448. [Abstract] [Full Text] [PDF]

				J.-P. J. Issa, G. Garcia-Manero, F. J. Giles, R. Mannari, D. Thomas, S. Faderl, E. Bayar, J. Lyons, C. S. Rosenfeld, J. Cortes, et al. Phase 1 study of low-dose prolonged exposure schedules of the hypomethylating agent 5-aza-2'-deoxycytidine (decitabine) in hematopoietic malignancies Blood, March 1, 2004; 103(5): 1635 - 1640. [Abstract] [Full Text] [PDF]

				J.-P. J. Issa Methylation and Prognosis: Of Molecular Clocks and Hypermethylator Phenotypes Clin. Cancer Res., August 1, 2003; 9(8): 2879 - 2881. [Full Text] [PDF]

				M. V. Brock, M. Gou, Y. Akiyama, A. Muller, T.-T. Wu, E. Montgomery, M. Deasel, P. Germonpre, L. Rubinson, R. F. Heitmiller, et al. Prognostic Importance of Promoter Hypermethylation of Multiple Genes in Esophageal Adenocarcinoma Clin. Cancer Res., August 1, 2003; 9(8): 2912 - 2919. [Abstract] [Full Text] [PDF]

				G. Garcia-Manero, J. Daniel, T. L. Smith, S. M. Kornblau, M.-S. Lee, H. M. Kantarjian, and J.-P. J. Issa DNA Methylation of Multiple Promoter-associated CpG Islands in Adult Acute Lymphocytic Leukemia Clin. Cancer Res., July 1, 2002; 8(7): 2217 - 2224. [Abstract] [Full Text] [PDF]

				G. Garcia-Manero, C. Bueso-Ramos, J. Daniel, J. Williamson, H. M. Kantarjian, and J.-P. J. Issa DNA Methylation Patterns at Relapse in Adult Acute Lymphocytic Leukemia Clin. Cancer Res., June 1, 2002; 8(6): 1897 - 1903. [Abstract] [Full Text] [PDF]

				M. Toyota, K. J. Kopecky, M.-O. Toyota, K.-W. Jair, C. L. Willman, and J.-P. J. Issa Methylation profiling in acute myeloid leukemia Blood, May 1, 2001; 97(9): 2823 - 2829. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Li, Q. Articles by Issa, J.-P. J. Search for Related Content PubMed PubMed Citation Articles by Li, Q. Articles by Issa, J.-P. J.