Ewing Sarcomas With p53 Mutation or p16/p14ARF Homozygous Deletion: A Highly Lethal Subset Associated With Poor Chemoresponse
http://www.100md.com
《临床肿瘤学》
the Departments of Pathology, Pediatrics, Surgery (Orthopaedics), and Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY
ABSTRACT
PATIENTS AND METHODS: We studied 60 patients with ES (stage: localized in 54, metastatic in six). All cases were confirmed to contain the EWS-FLI1 (29 type 1, 12 type 2, 14 other types) or EWS-ERG fusions (five cases). Homozygous deletion of p16/p14ARF, and p53 mutations were determined by fluorescent in situ hybridization and Affymetrix (Santa Clara, CA) p53 GeneChip microarray hybridization, respectively.
RESULTS: Eight cases (13.3%) contained point mutations of p53, and eight cases (13.3%) showed p16/p14ARF deletion, including one case with both alterations. Among 32 cases with data on histologic chemoresponse, all 10 with alterations in p53 or p16/p14ARF showed a poor chemoresponse (P = .03). Variables predicting poorer overall survival included p53 mutation alone (P < .001), either p53 or p16/p14ARF alteration (P < .001), and stage (P < .01). In multivariate analysis, alterations of p53 and/or p16/p14ARF as a single variable, was the most adverse prognostic factor (P < .001), followed by stage (P = .04). In a multivariate analysis with alterations of p53 and p16/p14ARF as separate variables, both were significant (P < .001 and P = .03, respectively). Six cases with p16/p14ARF deletion were also studied for co-deletion of the contiguous methylthioadenosine phosphorylase gene, and this was detected in four cases.
CONCLUSION: Alterations in p53 or p16/p14ARF are found in a fourth of ES cases and define a subset with highly aggressive behavior and poor chemoresponse.
INTRODUCTION
At the molecular level, ES/PNET is characterized by chromosomal translocations that fuse EWS, located at 22q12, and a gene of the ETS family of transcription factors. In more than 95% of cases, the gene fusion is EWS-FLI1 (90% to 95%) due to the classic t(11;22)(q24;q12), or EWS-ERG (5% to 10%) due to a variant translocation with the ERG gene at 21q22.4-6 These gene fusions are presumed to be the initiating oncogenic event in ES/PNET, and seem to play a critical role in the proliferation and tumorigenesis of ES/PNET cells.7,8 Very rare cases of ES/PNET show fusions of EWS to other ETS family genes (such as ETV1, E1AF, and FEV), or TLS-ERG fusion.9
At least some of the clinical heterogeneity in ES/PNET may correlate with EWS-FLI1 fusion transcript structure.5 We and others have found that the survival of patients whose tumor contains the most common type of EWS-FLI1 fusion (type 1: EWS exon 7 fused to FLI1 exon 6) is significantly better than that of patients with other EWS-FLI1 fusion types.10-12 Although the differences in EWS-FLI1 fusion structure are paralleled by differences in transactivation and proliferative rate,13,14 the clinical differences in retrospective series have been moderate. The variant EWS-ERG fusion has been found to be associated with clinical phenotypes indistinguishable from EWS-FLI1–positive ES/PNET.15
ES/PNET is also heterogeneous for the occurrence of genetic alterations involving certain critical regulators of cell cycle progression and apoptosis, in particular p16/p14ARF and p53. As reviewed in Table 1, homozygous deletion of p16/p14ARF and p53 alteration have been detected in approximately 20% and 10% of ES/PNET tumor samples, respectively. Although these are small subsets, both alterations are prognostically unfavorable, as reported by our group and others.23-25,27 However, these studies have had several limitations—many of them were small, few studies looked at both alterations, and some did not use optimal methods for the detection of these alterations.
The cellular effects of inactivating alterations of p16/p14ARF and p53 may overlap because of the roles of these proteins in the response to inappropriate mitogenic signals and the regulation of G1/S phase progression by the RB protein. The p53 protein mediates G1/S phase arrest or apoptosis (or senescence) in response to DNA damage or inappropriate mitogenic signals.28 P53, a transcription factor, blocks G1/S progression primarily through its downstream target gene, p21WAF. P53 itself is negatively modulated by binding to MDM2. The p16 and p14ARF proteins are encoded by the CDKN2A gene, located at 9p21 and closely flanked by the MTAP gene. MTAP, a gene 100 kb telomeric to p16/p14ARF at 9p21, encodes methylthioadenosine phosphorylase, an enzyme essential in the salvage of adenine and methionine. The CDKN2A gene encodes p16 and p14ARF by overlapping transcripts that use different reading frames.29 In the present article, we use the term p16/p14ARF gene synonymously with CDKN2A to clearly reflect this dual biology. The p16 protein inhibits cyclin D1-dependent kinases CDK4 and CDK6 and thus prevents phosphorylation of the RB protein, thereby blocking G1/S progression. In contrast to p16, p14ARF acts mainly through the p53 pathway by inhibiting the p53 antagonist MDM2.
To clarify the interrelationships between fusion gene structure, p53 mutation, and p16/p14ARF deletion and their relative prognostic impact, we characterized these genetic alterations in 60 patients with ES/PNET and correlated molecular data with clinical parameters and overall survival. In addition, the subset of tumors with homozygous deletion of p16/p14ARF was examined for co-deletion of the adjacent MTAP gene, frequently co-deleted with p16/p14ARF in other cancers30-32 because the loss of this gene renders tumors susceptible to novel MTAP-directed agents.33
PATIENTS AND METHODS
There were 30 males and 30 females. Mean age at diagnosis was 23.7 years (median, 17 years; range, 1 to 72 years). The primary tumors were skeletal in origin in 38 patients and extraskeletal in 22. The higher proportion of extraskeletal primaries and cases in older adults may reflect the impact of confirmatory molecular diagnostic data in establishing this diagnosis in nonclassic settings.34,35 The primary tumor locations were axial in 32 patients and peripheral in 28 (lesions arising in limb girdles were considered peripheral).
All patients had a chest computed tomography and radionuclide bone scan as part of their pretreatment extent of disease evaluation. At presentation, the tumor was localized in 54 patients, whereas six (10%) had distant metastases. The source of tissue used for molecular analyses was metastatic in 16 cases and nonmetastatic (primary or local recurrence) in 44 cases. The procurement of metastatic samples did not bias the molecular data compared with nonmetastatic samples (see Results), but may be responsible for poorer than expected chemoresponse and overall survival figures in the present series (see Results).
Patients With Repeat or Serial Analyses
Twenty cases were included in three prior series,11,24,25 but their inclusion in the present study was based entirely on the above three criteria, and not on their previously determined p16/p14ARF or p53 status. To include these 20 cases in the present series, p16/p14ARF and p53 status were restudied using the different methods described below, with equivalent results in all but one case (case ES-129 in Wei et al,25 case 12 in the present series) that had been negative for p16/p14ARF deletion by Southern blotting, but showed homozygous deletion by fluorescence in situ hybridization (FISH) in the present study. This finding is not unexpected given the greater risk of false-negative results due to admixed non-neoplastic cells in assays based on Southern blotting compared with FISH.
Treatment and Response to Chemotherapy
Surgery was the preferred method for achieving local control, almost always after preoperative chemotherapy. Incompletely resected or unresected tumors, including most axial tumors, received radiotherapy with few exceptions. Neoadjuvant and/or adjuvant chemotherapy was administered according to the second- or third-generation protocols consisting of vincristine, doxorubicin, and cyclophosphamide, plus additional agents including etoposide or ifosfamide. The standard protocol at MSKCC during this period (1991 to 2001) was the P6 protocol.1 Hematoxylin and eosin slides of postchemotherapy tumor resections, available in 32 of the 60 patients, were evaluated for histological response to chemotherapy using the Huvos grading scheme (grade 1 = tumor necrosis ≤ 50%; grade 2 = 51% to 90% tumor necrosis; grade 3 = only scattered foci of viable tumor cells [91% to 99% tumor necrosis]; grade 4 = complete tumor necrosis [100%]).36 For statistical analysis, cases showing grade 3 or grade 4 responses were categorized as good responders whereas poor responders consisted of grade 1 or grade 2 cases. The survival status as of February 15, 2003 was determined.
Fusion Transcript Detection
RNA was extracted from frozen tumors and analyzed by reverse-transcriptase polymerase chain reaction (RT-PCR) for EWS-FLI1 and EWS-ERG, performed as previously described.10
FISH
FISH for deletions involving p16/p14ARF and MTAP was performed on tumor imprints of frozen tumors as described previously.32 The probes consisted of approximately 100-kb fragments of human genomic DNA from the p16/p14ARF (clones P1-1063) or MTAP (P1-1069) regions (both gifts of Dr Alex Kamb), with each used in combination with a chromosome 9 centromere probe (CEP-9; Vysis Inc, Downers Grove, IL). Signal counts were recorded in more than 100 nuclei. Homozygous deletion was defined as 20% or more nuclei lacking both signals for the locus specific probe (p16/p14ARF or MTAP) and showing at least one signal for the CEP-9 probe.
p53 Mutation Analysis
We first screened for p53 missense mutations by immunohistochemistry (IHC) with antibody DO7 (DAKO, Carpinteria, CA; 1:500) as described24 on tumor specimens from 49 patients with available paraffin blocks. Nuclear overexpression of p53 by IHC has high sensitivity but poor specificity for p53 missense mutations.37,38 Therefore, the IHC screening was followed by p53 GeneChip analysis (Affymetrix, Santa Clara, CA) on DNA extracted from frozen tissue in cases with ≥ 10% immunoreactivity (n = 17) or with missing IHC data (n = 11). Mutations with GeneChip report scores ≥ 13 were accepted as genuine based on previous studies showing very high specificity for scores in the 13 to 36 range.39-41 Cases with reproducible borderline scores (5 to 12) for specific base changes were subjected to confirmatory direct sequencing. Based on previous studies, this combined IHC/p53 GeneChip approach is expected to have high detection sensitivity and specificity for p53 missense mutation.41
Statistical Analysis
Statistical computations were performed using S-plus software (2000; Mathsoft Inc, Cambridge, MA). Survival time was from diagnosis to last follow-up (or death). Survival rate was estimated by Kaplan-Meier methodology. The relationship between survival and other variables was investigated using the log-rank test for categorical variables and a score test based on the Cox proportional hazards model for continuous variables. Age was analyzed as a continuous variable. A multivariate model was fitted using Cox regression with variables found to be significant at the univariate level. Associations among categorical and ordinal variables were checked by Fisher's exact test.
RESULTS
P16/P14ARF Deletion and MTAP Co-Deletion
Homozygous and hemizygous deletions of the p16/p14ARF gene were detected in eight (13.3%; Table 1; Fig 1A through D) and six (10%) cases, respectively, and the former group consisted of one metastatic and seven localized cases. FISH analysis showed no deletion of p16/p14ARF in the remaining 46 (76.7%) cases. Because the survival of patients with tumors containing loss of only one copy of p16/p14ARF paralleled that of cases lacking any evidence of deletion (see Univariate Survival Analysis subsection), cases with hemizygous deletion of p16/p14ARF were grouped with nondeleted cases for all subsequent analyses. Of the eight cases with homozygous deletion of the p16/p14ARF gene,six were also studied for co-deletion of MTAP. Homozygous or hemizygous MTAP deletions were observed in four and one cases, respectively, as detected by probe P1-1069 (Fig 1E and F).
p53 Mutation
By IHC, 17 cases showed p53 immunoreactivity in ≥ 10% tumor nuclei. These cases, together with another 11 cases lacking IHC data, were subjected to Affymetrix p53 GeneChip analysis to detect point mutations. This approach identified p53 point mutations in eight (13.3%) of 60 cases (Table 1), including six patients with localized tumors and two with metastatic disease. Affymetrix p53 GeneChip scores for specific point mutations ranged from 14 to 37 in six cases (Table 3). In two cases with borderline scores (ie, 7 and 10), direct sequencing confirmed the base change detected by chip analysis (results not shown). Among the eight cases showing p53 mutation, there was only one case (case 2 in Table 3) that also displayed homozygous deletion of p16/p14ARF. Taken together, the total number of cases with either p53 missense mutation or p16/p14ARF homozygous deletion was 15 (25%; Table 1). Table 3 summarizes the clinical and molecular data on these 15 cases.
Correlations Between Parameters
Among the 32 cases available for evaluation of histological response to chemotherapy, there were nine good, and 23 poor responders. Good chemotherapeutic effect was correlated with the absence of either p53 or p16/p14ARF alterations (P = .03, Fisher's exact test; Table 4).
There was also a trend for p53 mutation alone (P = .14) and the combination of either p53 or p16/p14ARF alteration (P = .13) to show a relationship to stage, reflecting the finding that three of six tumors from patients with metastatic disease at diagnosis contained one of these two alterations, compared with only 11 of the remaining 54 patients. The association of p16/p14ARF status alone and stage did not approach statistical significance.
There were no significant associations among the other variables. Specifically, there was no significant association between EWS-FLI1 fusion type (type 1 v others) and alterations in p53 and p16/p14ARF in the 55 EWS-FLI1 cases. EWS-FLI1 fusion type was also not related to chemotherapy response. The seemingly mutually exclusive relationship between p53 mutation and p16/p14ARF deletion was also not statistically significant. In addition, neither the EWS-FLI1 type nor the status of the p16/p14ARF gene was associated with stage or whether the tumor was axial or peripheral, or skeletal or extraskeletal.
Because the source of the tissues used for molecular analyses was metastatic in 16 cases, we examined the possibility of a relationship between sample source and abnormalities of p53 or p16/p14ARF, but there was none: five of 16 metastatic and 10 of 44 nonmetastatic samples showed p53 or p16/p14ARF alteration (P = .52). This supports the notion that these genetic changes generally occur in the primary tumor at a preclinical time point.
Univariate Survival Analyses
The median follow-up periods among all 60 patients and all 34 survivors were 29 months (range, 6 to 118 months) and 43.5 months (range, 6 to 118 months), respectively. The median overall survival time for all 60 patients was 99 months. At last follow-up, 25 patients were alive with no evidence of disease, nine were alive with disease, 24 were dead of disease, and the remaining two patients died of other causes. Five of six patients presenting with distant metastases died of their disease by 17 months, with a median survival of 12 months in contrast to 99 months in the localized subgroup (Fig 2A). All eight patients showing p53 mutation died before 21 months, with a median survival of 11 months, compared with 99 months in the nonmutated subset (Fig 2B).
By univariate analysis (Table 5), stage (Fig 2A, P < .001; risk ratio [RR] = 2.2; 95% CI, 1.4 to 3.7), p53 mutation (Fig 2B, P < .0001; RR = 3.7; 95% CI, 2.1 to 4.6), and either or both of a p53 or p16 alteration (Fig 2C, P < .0001; RR = 2.5; 95% CI, 1.6 to 3.8), were strong negative predictors of overall survival, whereas p16/p14ARF deletion alone (Fig 2D, P = .06; RR = 1.6; 95% CI, 0.96 to 2.6) only reached marginal statistical significance. Since the six cases with evidence of hemizygous deletion of p16/p14ARF showed no difference in survival compared with the nondeleted cases (not shown), we combined these two groups for the survival analysis.
No other factors were associated with survival, including age, sex, location, and origin of the primary tumor. As in previous studies,10,11 among cases with EWS-FLI1 (n = 55), patients with type 1 fusion showed slightly better survival, but this did not reach statistical significance (Fig 2E, P = .4; RR = 1.5; 95% CI, 0.6 to 3.4).
Multivariate Survival Analyses
In the multivariate analysis (Table 6), the presence of p53 mutation (P < .001; RR = 3.9; 95% CI, 1.9 to 8.0) remained the most significant independent prognostic factor, and p16/p14ARF deletion emerged as the second most important independent variable (P = .03; RR = 1.8; 95% CI, 1.04 to 2.9), while stage lost its statistical significance. Nevertheless, when p53 and p16 were considered together as a single parameter, the presence of either or both of these two genetic alterations (P < .001; RR = 2.3; 95% CI, 1.4 to 3.5) was the strongest negative factor, and stage (P = .04; RR = 2.3; 95% CI, 1.03 to 3.03) became the second most important adverse factor for overall survival.
DISCUSSION
Homozygous deletion of p16/p14ARF (n = 8, 13%) was as frequent as p53 mutation in this series, a somewhat lower prevalence than most previous studies (Table 1). Possible reasons for this difference may include variability in methodologies or in criteria to define positive cases, and small sample sizes in previous studies. By univariate analysis, p16/p14ARF deletion alone had only marginal value as a negative factor. However, in the multivariate analysis, p16/p14ARF homozygous deletion emerged as the second most significant factor after p53 mutation. The lack of significance of stage in the latter multivariate analysis should be interpreted with caution given the low number of metastatic samples (n = 6) in the present study. Finally, when p16/p14ARF deletion and p53 mutation were combined as a single factor, it was the most crucial determinant of overall survival followed by stage in both univariate and multivariate analysis. The negative impact of p16/p14ARF loss on survival in ES/PNET is consistent with several previous smaller studies.26,27,42
Some groups have found an association between p16/p14ARF or p53 alterations and advanced stage,23,42 while others have not.24,25 In the current series, the presence of p53 mutation or p16/p14ARF deletion was weakly, but not significantly, related to distant metastases at presentation, but the small number of metastatic cases (n = 6; 10%) limited the power of this analysis. Regardless of this possible association, it is of more practical clinical importance that most p53 mutations or p16/p14ARF deletions are detected in patients with localized disease at diagnosis—22% in the present series.
Our finding that the six cases with hemizygous deletion of p16/p14ARF showed no difference in survival compared with the nondeleted cases suggests that loss of one copy of p16/p14ARF combined with inactivation of the remaining copy by mutation or epigenetic mechanisms is unlikely to be significant in ES/PNET. Indeed, Kovar et al17 found only one tumor in 27 to show mutation of p16/p14ARF with loss of heterozygosity. In another study, the presence of p16 promoter hypermethylation had no impact on survival of ES/PNET patients.27 Finally, further evidence indicating that p16 promoter hypermethylation is rare in ES/PNET is that the prevalence of loss of p16 protein expression in both tumors and cell lines closely approximates that of p16/p14ARF homozygous deletion.42,43 In aggregate, these data support homozygous deletion as the principal mode of p16/p14ARF inactivation in ES/PNET.
The finding of combined p16/p14ARF deletion and p53 mutation in only a single case of a total of 15 with either or both alterations might suggest that alterations in these two genes are mutually exclusive. However, this apparent reciprocal relationship p16/p14ARF deletion and p53 mutation did not approach statistical significance. This, along with the report of another ES/PNET case with both alterations26 and of several ES/PNET cell lines with both p16/p14ARF deletion and p53 mutation,17 reinforces the impression that the infrequent coincidence of these two alterations in primary ES/PNET stems mainly from their low prevalence. Other modes of inactivation of the p53 and Rb pathways, through MDM2, CCND1, or Rb alterations, are rare in ES/PNET.16,44
The striking impact of p53 and p16/p14ARF alterations on prognosis points to profound biologic differences inconsistent with a functionally equivalent perturbation of these pathways in the remaining 75% of cases. The strong clinical impact of these alterations in ES/PNET is reminiscent of the marked impact of impaired apoptosis due to p14ARF or p53 loss on lymphomagenesis in Em-myc transgenic mice, a model for P53 pathway alterations in human tumors with translocation oncogenes.45,46 Oncogenic or inappropriate proliferative stimuli are known to elicit an apoptotic or senescence response dependent on the P53/p14ARF pathway.47,48 By analogy, it is tempting to hypothesize that the oncogenicity of EWS-FLI1 may be modulated or tempered by the apoptotic function of an intact P53/p14ARF pathway. Indeed, data from three groups support an important relationship between EWS-FLI1 and the P53 pathway.
Kovar et al found that reintroduction of normal p53 in p53-null ES/PNET cell lines triggers extensive apoptosis.49,50 It should be noted that essentially all ES/PNET cell lines contain p53 alterations or p16/p14ARF deletions, suggesting selection pressure for these genetic alterations to permit in vitro growth.8,51 These data suggest that reintroduction of functional p53 in ES/PNET cell lines promptly leads EWS-FLI1 to trigger a p53-dependent apoptotic program because ES/PNET cells cannot overcome an apoptotic response to in vitro adherent culture conditions unless the p53 pathway is nonfunctional. Deneen et al showed that EWS-FLI1 induces apoptosis and growth arrest in normal mouse embryonic fibroblasts (MEF).52 However, in p16/p14ARF–null or p53-deficient MEFs, apoptosis and growth arrest in response to EWS-FLI1 were reduced.52 Finally, Lessnick et al53 found that in primary human fibroblasts immortalized with telomerase cDNA, expression of EWS-FLI triggered growth arrest associated with transcriptional upregulation of p53. While these data suggest that specific response mechanisms to EWS-FLI1 may differ in different experimental settings, taken together, they support a central role for the p53 pathway in the cellular response to EWS-FLI1.
A similar pattern of infrequent but highly clinically significant p53 or p16/p14ARF alterations has been reported in several other developmental and/or translocation-associated cancers.54 For instance, the outcome of Wilms' tumors,55,56 neuroblastomas,57 follicular lymphomas,58,59 and myxoid liposarcomas60 is dramatically worsened by p53/p14ARF pathway alterations. This is in contrast to most carcinomas and most sarcomas lacking specific translocations, where p53 alterations usually have only moderate or marginal clinical significance.61-63
MTAP is co-deleted in 75% to 90% of tumors, with p16/p14ARF homozygous deletion32 reflecting the close genetic linkage between these two genes. Indeed, four of our six p16/p14ARF–deleted cases (66%) had also lost both copies of MTAP. Codeletion of the MTAP gene with p16/p14ARF has been observed in a variety of cancers.30,32,64-66 Because cancer cells that lack MTAP are unable to salvage adenine from methylthioadenosine, they become dependent on the de novo synthesis pathway of purine metabolism. Thus, MTAP-deleted cancer cells are highly and specifically sensitive to inhibition of the de novo synthesis pathway (eg, by L-alanosine, also known as SDX-102), presenting an opportunity for tailored chemotherapy.33,64 Although only a small proportion of patients with ES/PNET would be candidates for such MTAP-directed chemotherapeutic approaches, it is a subset of ES/PNET that includes approximately a third of the most lethal cases.
We and others have previously shown that the survival of patients with localized disease whose ES/PNET bear the type 1 EWS-FLI1 fusion is better than those with tumors containing other EWS-FLI1 fusion types.10-12 These series have ranged from 55 to 83 informative cases.12 In the present series, among the 55 EWS-FLI1 cases (including six metastatic), EWS-FLI1 fusion type was not a statistically significant prognostic factor, though the survival trends did parallel the findings in prior larger series in that patients with the type 1 fusion seemed to have better outcomes (Fig 2E). The prognostic impact of EWS-FLI1 fusion type appears lesser than that of p53 or p16/p14ARF alterations and may not be demonstrable unless a larger sample size is studied.
A poor histologic response to preoperative chemotherapy has been found to be a highly significant adverse factor in studies of 74 and 118 patients with ES/PNET, respectively.36,67 The limited number of cases in the present study with available chemotherapy response data (n = 32) made the series unsuitable for re-evaluation of this question. However, a significant correlation was observed between good chemotherapeutic effect and the absence of genetic alterations in p53 or p16/p14ARF. Previous studies relating p53 alterations (by IHC) to chemoresponse in ES/PNET reached conflicting conclusions.23,24 P53 mutations have been shown to be associated with resistance to chemotherapy and radiotherapy in diverse tumor types, including sarcomas.68,69 The drug resistance caused by p53 dysfunction can be exerted through several underlying mechanisms.68,69 In addition, p14ARF deletions promote chemoresistance in murine Eμ-myc lymphomas by disabling p53 function with consequent apoptotic defects,45 further emphasizing the links between p16/p14ARF and p53 in oncogenic transformation and chemoresistance.
There had so far been no strong prognostic factor available for prechemotherapy risk stratification of patients presenting with localized ES/PNET. The present retrospective study suggests that alterations in p53 and p16/p14ARF, found in approximately a fourth of all ES/PNET, are strong negative predictors of overall survival, thereby providing for the first time a robust molecular marker of clinical outcome in ES/PNET that could be used to assign certain patients with localized disease to high-risk regimens before initial chemotherapy. Definitive confirmation of these results will require a prospective analysis of these molecular prognostic markers in a larger series of patients with ES/PNET, with longer follow-up.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank the following for providing clinical follow-up data: Ronald Jaffe, MD, Children's Hospital of Pittsburgh; Alex Aledo, MD, New York Presbyterian Hospital; Charles F. Timmons, MD, PhD, Children's Medical Center of Dallas.
NOTES
Supported by American Cancer Society Project Grant 99-216 (M.L.) and the Ewing Sarcoma Research Fund (M.L.).
H.-Y.H. was a visiting fellow sponsored by Chang Gung Memorial Hospital, Kaohsiung Medical Center, Kaohsiung, Taiwan.
H.-Y.H. and P.B.I. contributed equally to this work. P.B.I. is presently in the Department of Pathology, New York University Medical Center, 560 First Ave, New York, NY 10016.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Kolb EA, Kushner BH, Gorlick R, et al: Long-term event-free survival after intensive chemotherapy for Ewing's family of tumors in children and young adults. J Clin Oncol 21:3423-3430, 2003
2. Shankar AG, Ashley S, Craft AW, et al: Outcome after relapse in an unselected cohort of children and adolescents with Ewing sarcoma. Med Pediatr Oncol 40:141-147, 2003
3. Meyers PA, Krailo MD, Ladanyi M, et al: High-dose melphalan, etoposide, total-body irradiation, and autologous stem-cell reconstitution as consolidation therapy for high-risk Ewing's sarcoma does not improve prognosis. J Clin Oncol 19:2812-2820, 2001
4. Delattre O, Zucman J, Plougastel B, et al: Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours. Nature 359:162-165, 1992
5. Zucman J, Melot T, Desmaze C, et al: Combinatorial generation of variable fusion proteins in the Ewing family of tumours. EMBO J 12:4481-4487, 1993
6. Sorensen PHB, Lessnick SL, Lopez-Terrada D, et al: A second Ewing's sarcoma translocation, t(21;22), fuses the EWS gene to another ets-family transcription factor, ERG. Nat Genet 6:146-151, 1994
7. Arvand A, Denny CT: Biology of EWS/ETS fusions in Ewing's family tumors. Oncogene 20:5747-5754, 2001
8. Ladanyi M: EWS-FLI1 and Ewing's sarcoma: Recent molecular data and new insights. Cancer Biol Ther 1:330-336, 2002
9. Shing DC, McMullan DJ, Roberts P, et al: FUS/ERG gene fusions in Ewing's tumors. Cancer Res 63:4568-4576, 2003
10. Zoubek A, Dockhorn-Dworniczak B, Delattre O, et al: Does expression of different EWS chimeric transcripts define clinically distinct risk groups of Ewing tumor patients J Clin Oncol 14:1245-1251, 1996
11. de Alava E, Kawai A, Healey JH, et al: EWS-FLI1 fusion transcript structure is an independent determinant of prognosis in Ewing's sarcoma. J Clin Oncol 16:1248-1255, 1998
12. Aryee DN, Sommergruber W, Muehlbacher K, et al: Variability in gene expression patterns of Ewing tumor cell lines differing in EWS-FLI1 fusion type. Lab Invest 80:1833-1844, 2000
13. Lin PP, Brody RI, Hamelin A, et al: Differential transactivation by alternative EWS-FLI1 fusion proteins correlates with clinical heterogeneity in Ewing's sarcoma. Cancer Res 59:1428-1432, 1999
14. de Alava E, Panizo A, Antonescu CR, et al: Association of EWS-FLI1 type 1 fusion with lower proliferative rate in Ewing's sarcoma. Am J Pathol 156:849-855, 2000
15. Ginsberg JP, de Alava E, Ladanyi M, et al: EWS-FLI1 and EWS-ERG gene fusions are associated with similar clinical phenotypes in Ewing's sarcoma. J Clin Oncol 17:1809-1814, 1999
16. Kovar H, Auinger A, Jug G, et al: Narrow spectrum of infrequent p53 mutations and absence of MDM2 amplification in Ewing tumours. Oncogene 8:2683-2690, 1993
17. Kovar H, Jug G, Aryee DNT, et al: Among genes involved in the RB dependent cell cycle regulatory cascade, the p16 tumor suppressor gene is frequently lost in the Ewing family of tumors. Oncogene 15:2225-2232, 1997
18. Komuro H, Hayashi Y, Kawamura M, et al: Mutations of the p53 gene are involved in Ewing's sarcomas but not in neuroblastomas. Cancer Res 53:5284-5288, 1993
19. Hamelin R, Zucman J, Melot T, et al: P53 mutations in human tumors with chimeric EWS/FLI-1 genes. Int J Cancer 57:336-340, 1994
20. Mangham D, Cannon A, Li X, et al: p53 overexpression in Ewing's sarcoma/primitive neuroectodermal tumour is an uncommon event. J Clin Pathol: Mol Pathol 48:M79-M82, 1995
21. Patino-Garcia A, Sierrasesumaga L: Analysis of the p16INK4 and TP53 tumor suppressor genes in bone sarcoma pediatric patients. Cancer Genet Cytogenet 98:50-55, 1997
22. Radig K, Schneider-Stock R, Rose I, et al: p53 and ras mutations in Ewing's sarcoma. Pathol Res Pract 194:157-162, 1998
23. Abudu A, Mangham DC, Reynolds GM, et al: Overexpression of p53 protein in primary Ewing's sarcoma of bone: Relationship to tumour stage, response and prognosis. Br J Cancer 79:1185-1189, 1999
24. de Alava E, Antonescu CR, Panizo A, et al: Prognostic impact of P53 status in Ewing's sarcoma. Cancer 89:783-792, 2000
25. Wei G, Antonescu CR, de Alava E, et al: Prognostic impact of INK4A deletion in Ewing's sarcoma. Cancer 89:793-799, 2000
26. Tsuchiya T, Sekine K, Hinohara S, et al: Analysis of the p16INK4, p14ARF, p15, TP53, and MDM2 genes and their prognostic implications in osteosarcoma and Ewing sarcoma. Cancer Genet Cytogenet 120:91-98, 2000
27. Lopez-Guerrero JA, Pellin A, Noguera R, et al: Molecular analysis of the 9p21 locus and p53 genes in Ewing family tumors. Lab Invest 81:803-814, 2001
28. Sherr CJ, McCormick F: The RB and p53 pathways in cancer. Cancer Cell 2:103-112, 2002
29. Sherr CJ: The INK4a/ARF network in tumour suppression. Nat Rev Mol Cell Biol 2:731-737, 2001
30. Chen ZH, Zhang H, Savarese TM: Gene deletion chemoselectivity: Codeletion of the genes for p16(INK4), methylthioadenosine phosphorylase, alpha- and beta-interferons in human pancreatic cell carcinoma lines and its implications for chemotherapy. Cancer Res 56:1083-1090, 1996
31. Hori Y, Hori H, Yamada Y, et al: The methylthioadenosine phosphorylase gene is frequently co-deleted with the p16INK4a gene in acute type adult T-cell leukemia. Int J Cancer 75:51-56, 1998
32. Illei PB, Rusch VW, Zakowski MF, et al: Homozygous deletion of CDKN2A and co-deletion of the methylthioadenosine phosphorylase gene in the majority of pleural mesotheliomas. Clin Cancer Res 9:2108-2113, 2003
33. Batova A, Diccianni MB, Omura-Minamisawa M, et al: Use of alanosine as a methylthioadenosine phosphorylase-selective therapy for T-cell acute lymphoblastic leukemia in vitro. Cancer Res 59:1492-1497, 1999
34. Ladanyi M, Bridge JA: Contribution of molecular genetic data to the classification of sarcomas. Hum Pathol 31:532-538, 2000
35. Lawlor ER, Mathers JA, Bainbridge T, et al: Peripheral Primitive Neuroectodermal Tumors in Adults: Documentation by Molecular Analysis. J Clin Oncol 16:1150-1157, 1998
36. Wunder JS, Paulian G, Huvos AG, et al: The histological response to chemotherapy as a predictor of the oncological outcome of operative treatment of Ewing sarcoma. J Bone Joint Surg 80:1020-1033, 1998
37. Baas IO, Mulder JW, Offerhaus GJ, et al: An evaluation of six antibodies for immunohistochemistry of mutant p53 gene product in archival colorectal neoplasms. J Pathol 172:5-12, 1994
38. Hsia CC, Nakashima Y, Thorgeirsson SS, et al: Correlation of immunohistochemical staining and mutations of p53 in human hepatocellular carcinoma. Oncol Rep 7:353-356, 2000
39. Ahrendt SA, Halachmi S, Chow JT, et al: Rapid p53 sequence analysis in primary lung cancer using an oligonucleotide probe array. Proc Natl Acad Sci U S A 96:7382-7387, 1999
40. Wen WH, Bernstein L, Lescallett J, et al: Comparison of TP53 mutations identified by oligonucleotide microarray and conventional DNA sequence analysis. Cancer Res 60:2716-2722, 2000
41. Wikman FP, Lu ML, Thykjaer T, et al: Evaluation of the performance of a p53 sequencing microarray chip using 140 previously sequenced bladder tumor samples. Clin Chem 46:1555-1561, 2000
42. Maitra A, Roberts H, Weinberg AG, et al: Aberrant expression of tumor suppressor proteins in the Ewing family of tumors. Arch Pathol Lab Med 125:1207-1212, 2001
43. Dauphinot L, De Oliveira C, Melot T, et al: Analysis of the expression of cell cycle regulators in Ewing cell lines: EWS-FLI-1 modulates p57KIP2and c-Myc expression. Oncogene 20:3258-3265, 2001
44. Ladanyi M, Lewis R, Jhanwar SC, et al: MDM2 and CDK4 gene amplification in Ewing's sarcoma. J Pathol 175:211-217, 1995
45. Schmitt CA, McCurrach ME, de Stanchina E, et al: INK4a/ARF mutations accelerate lymphomagenesis and promote chemoresistance by disabling p53. Genes Dev 13:2670-2677, 1999
46. Eischen CM, Weber JD, Roussel MF, et al: Disruption of the ARF-Mdm2-p53 tumor suppressor pathway in Myc-induced lymphomagenesis. Genes Dev 13:2658-2669, 1999
47. Lowe SW: Activation of p53 by oncogenes. Endocr Relat Cancer 6:45-48, 1999
48. Schmitt CA, Fridman JS, Yang M, et al: A senescence program controlled by p53 and p16INK4a contributes to the outcome of cancer therapy. Cell 109:335-346, 2002
49. Kovar H, Jug G, Printz D, et al: Characterization of distinct consecutive phases in non-genotoxic p53-induced apoptosis of Ewing tumor cells and the rate-limiting role of caspase 8. Oncogene 19:4096-4107, 2000
50. Kovar H, Pospisilova S, Jug G, et al: Response of Ewing tumor cells to forced and activated p53 expression. Oncogene 22:3193-3204, 2003
51. Kovar H, Auinger A, Jug G, et al: The concept of tumor cell survival factors to explain discrepant p53 mutation frequencies in primary Ewing tumors versus cell lines. Proc Am Assoc Cancer Res 36:2, 1995 (abstr 2)
52. Deneen B, Denny CT: Loss of p16 pathways stabilizes EWS/FLI1 expression and complements EWS/FLI1 mediated transformation. Oncogene 20:6731-6741, 2001
53. Lessnick SL, Dacwag CS, Golub TR: The Ewing's sarcoma oncoprotein EWS/FLI induces a p53-dependent growth arrest in primary human fibroblasts. Cancer Cell 1:393-401, 2002
54. Peller S, Rotter V: TP53 in hematological cancer: Low incidence of mutations with significant clinical relevance. Hum Mutat 21:277-284, 2003
55. Bardeesy N, Falkoff D, Petruzzi MJ, et al: Anaplastic Wilms' tumour, a subtype displaying poor prognosis, harbours p53 gene mutations. Nat Genet 7:91-97, 1994
56. Bardeesy N, Beckwith JB, Pelletier J: Clonal expansion and attenuated apoptosis in Wilms' tumors are associated with p53 gene mutations. Cancer Res 55:215-219, 1995
57. Keshelava N, Zuo JJ, Waidyaratne NS, et al: p53 mutations and loss of p53 function confer multidrug resistance in neuroblastoma. Med Pediatr Oncol 35:563-568, 2000
58. Sander CA, Yano T, Clark HM, et al: p53 mutation is associated with progression in follicular lymphomas. Blood 82:1994-2004, 1993
59. Elenitoba-Johnson KS, Gascoyne RD, Lim MS, et al: Homozygous deletions at chromosome 9p21 involving p16 and p15 are associated with histologic progression in follicle center lymphoma. Blood 91:4677-4685, 1998
60. Antonescu CR, Tschernyavsky SJ, Decuseara R, et al: Prognostic impact of P53 status, TLS-CHOP fusion transcript structure, and histological grade in myxoid liposarcoma: A molecular and clinicopathologic study of 82 cases. Clin Cancer Res 7:3977-3987, 2001
61. Wurl P, Taubert H, Meye A, et al: Prognostic value of immunohistochemistry for p53 in primary soft-tissue sarcomas: A multivariate analysis of five antibodies. J Cancer Res Clin Oncol 123:502-508, 1997
62. Petersen S, Thames HD, Nieder C, et al: The results of colorectal cancer treatment by p53 status: Treatment-specific overview. Dis Colon Rectum 44:322-333, 2001
63. Liu MC, Gelmann EP: P53 gene mutations: Case study of a clinical marker for solid tumors. Semin Oncol 29:246-257, 2002
64. Harasawa H, Yamada Y, Kudoh M, et al: Chemotherapy targeting methylthioadenosine phosphorylase (MTAP) deficiency in adult T cell leukemia (ATL). Leukemia 16:1799-1807, 2002
65. M'soka TJ, Nishioka J, Taga A, et al: Detection of methylthioadenosine phosphorylase (MTAP) and p16 gene deletion in T cell acute lymphoblastic leukemia by real-time quantitative PCR assay. Leukemia 14:935-940, 2000
66. Zhang H, Chen ZH, Savarese TM: Codeletion of the genes for p16INK4, methylthioadenosine phosphorylase, interferon-alpha1, interferon-beta1, and other 9p21 markers in human malignant cell lines. Cancer Genet Cytogenet 86:22-28, 1996
67. Picci P, Bohling T, Bacci G, et al: Chemotherapy-induced tumor necrosis as a prognostic factor in localized Ewing's sarcoma of the extremities. J Clin Oncol 15:1553-1559, 1997
68. Ferreira CG, Tolis C, Giaccone G: p53 and chemosensitivity. Ann Oncol 10:1011-1021, 1999
69. Pirollo KF, Bouker KB, Chang EH: Does p53 status influence tumor response to anticancer therapies Anticancer Drugs 11:419-432, 2000(Hsuan-Ying Huang, Peter B)
ABSTRACT
PATIENTS AND METHODS: We studied 60 patients with ES (stage: localized in 54, metastatic in six). All cases were confirmed to contain the EWS-FLI1 (29 type 1, 12 type 2, 14 other types) or EWS-ERG fusions (five cases). Homozygous deletion of p16/p14ARF, and p53 mutations were determined by fluorescent in situ hybridization and Affymetrix (Santa Clara, CA) p53 GeneChip microarray hybridization, respectively.
RESULTS: Eight cases (13.3%) contained point mutations of p53, and eight cases (13.3%) showed p16/p14ARF deletion, including one case with both alterations. Among 32 cases with data on histologic chemoresponse, all 10 with alterations in p53 or p16/p14ARF showed a poor chemoresponse (P = .03). Variables predicting poorer overall survival included p53 mutation alone (P < .001), either p53 or p16/p14ARF alteration (P < .001), and stage (P < .01). In multivariate analysis, alterations of p53 and/or p16/p14ARF as a single variable, was the most adverse prognostic factor (P < .001), followed by stage (P = .04). In a multivariate analysis with alterations of p53 and p16/p14ARF as separate variables, both were significant (P < .001 and P = .03, respectively). Six cases with p16/p14ARF deletion were also studied for co-deletion of the contiguous methylthioadenosine phosphorylase gene, and this was detected in four cases.
CONCLUSION: Alterations in p53 or p16/p14ARF are found in a fourth of ES cases and define a subset with highly aggressive behavior and poor chemoresponse.
INTRODUCTION
At the molecular level, ES/PNET is characterized by chromosomal translocations that fuse EWS, located at 22q12, and a gene of the ETS family of transcription factors. In more than 95% of cases, the gene fusion is EWS-FLI1 (90% to 95%) due to the classic t(11;22)(q24;q12), or EWS-ERG (5% to 10%) due to a variant translocation with the ERG gene at 21q22.4-6 These gene fusions are presumed to be the initiating oncogenic event in ES/PNET, and seem to play a critical role in the proliferation and tumorigenesis of ES/PNET cells.7,8 Very rare cases of ES/PNET show fusions of EWS to other ETS family genes (such as ETV1, E1AF, and FEV), or TLS-ERG fusion.9
At least some of the clinical heterogeneity in ES/PNET may correlate with EWS-FLI1 fusion transcript structure.5 We and others have found that the survival of patients whose tumor contains the most common type of EWS-FLI1 fusion (type 1: EWS exon 7 fused to FLI1 exon 6) is significantly better than that of patients with other EWS-FLI1 fusion types.10-12 Although the differences in EWS-FLI1 fusion structure are paralleled by differences in transactivation and proliferative rate,13,14 the clinical differences in retrospective series have been moderate. The variant EWS-ERG fusion has been found to be associated with clinical phenotypes indistinguishable from EWS-FLI1–positive ES/PNET.15
ES/PNET is also heterogeneous for the occurrence of genetic alterations involving certain critical regulators of cell cycle progression and apoptosis, in particular p16/p14ARF and p53. As reviewed in Table 1, homozygous deletion of p16/p14ARF and p53 alteration have been detected in approximately 20% and 10% of ES/PNET tumor samples, respectively. Although these are small subsets, both alterations are prognostically unfavorable, as reported by our group and others.23-25,27 However, these studies have had several limitations—many of them were small, few studies looked at both alterations, and some did not use optimal methods for the detection of these alterations.
The cellular effects of inactivating alterations of p16/p14ARF and p53 may overlap because of the roles of these proteins in the response to inappropriate mitogenic signals and the regulation of G1/S phase progression by the RB protein. The p53 protein mediates G1/S phase arrest or apoptosis (or senescence) in response to DNA damage or inappropriate mitogenic signals.28 P53, a transcription factor, blocks G1/S progression primarily through its downstream target gene, p21WAF. P53 itself is negatively modulated by binding to MDM2. The p16 and p14ARF proteins are encoded by the CDKN2A gene, located at 9p21 and closely flanked by the MTAP gene. MTAP, a gene 100 kb telomeric to p16/p14ARF at 9p21, encodes methylthioadenosine phosphorylase, an enzyme essential in the salvage of adenine and methionine. The CDKN2A gene encodes p16 and p14ARF by overlapping transcripts that use different reading frames.29 In the present article, we use the term p16/p14ARF gene synonymously with CDKN2A to clearly reflect this dual biology. The p16 protein inhibits cyclin D1-dependent kinases CDK4 and CDK6 and thus prevents phosphorylation of the RB protein, thereby blocking G1/S progression. In contrast to p16, p14ARF acts mainly through the p53 pathway by inhibiting the p53 antagonist MDM2.
To clarify the interrelationships between fusion gene structure, p53 mutation, and p16/p14ARF deletion and their relative prognostic impact, we characterized these genetic alterations in 60 patients with ES/PNET and correlated molecular data with clinical parameters and overall survival. In addition, the subset of tumors with homozygous deletion of p16/p14ARF was examined for co-deletion of the adjacent MTAP gene, frequently co-deleted with p16/p14ARF in other cancers30-32 because the loss of this gene renders tumors susceptible to novel MTAP-directed agents.33
PATIENTS AND METHODS
There were 30 males and 30 females. Mean age at diagnosis was 23.7 years (median, 17 years; range, 1 to 72 years). The primary tumors were skeletal in origin in 38 patients and extraskeletal in 22. The higher proportion of extraskeletal primaries and cases in older adults may reflect the impact of confirmatory molecular diagnostic data in establishing this diagnosis in nonclassic settings.34,35 The primary tumor locations were axial in 32 patients and peripheral in 28 (lesions arising in limb girdles were considered peripheral).
All patients had a chest computed tomography and radionuclide bone scan as part of their pretreatment extent of disease evaluation. At presentation, the tumor was localized in 54 patients, whereas six (10%) had distant metastases. The source of tissue used for molecular analyses was metastatic in 16 cases and nonmetastatic (primary or local recurrence) in 44 cases. The procurement of metastatic samples did not bias the molecular data compared with nonmetastatic samples (see Results), but may be responsible for poorer than expected chemoresponse and overall survival figures in the present series (see Results).
Patients With Repeat or Serial Analyses
Twenty cases were included in three prior series,11,24,25 but their inclusion in the present study was based entirely on the above three criteria, and not on their previously determined p16/p14ARF or p53 status. To include these 20 cases in the present series, p16/p14ARF and p53 status were restudied using the different methods described below, with equivalent results in all but one case (case ES-129 in Wei et al,25 case 12 in the present series) that had been negative for p16/p14ARF deletion by Southern blotting, but showed homozygous deletion by fluorescence in situ hybridization (FISH) in the present study. This finding is not unexpected given the greater risk of false-negative results due to admixed non-neoplastic cells in assays based on Southern blotting compared with FISH.
Treatment and Response to Chemotherapy
Surgery was the preferred method for achieving local control, almost always after preoperative chemotherapy. Incompletely resected or unresected tumors, including most axial tumors, received radiotherapy with few exceptions. Neoadjuvant and/or adjuvant chemotherapy was administered according to the second- or third-generation protocols consisting of vincristine, doxorubicin, and cyclophosphamide, plus additional agents including etoposide or ifosfamide. The standard protocol at MSKCC during this period (1991 to 2001) was the P6 protocol.1 Hematoxylin and eosin slides of postchemotherapy tumor resections, available in 32 of the 60 patients, were evaluated for histological response to chemotherapy using the Huvos grading scheme (grade 1 = tumor necrosis ≤ 50%; grade 2 = 51% to 90% tumor necrosis; grade 3 = only scattered foci of viable tumor cells [91% to 99% tumor necrosis]; grade 4 = complete tumor necrosis [100%]).36 For statistical analysis, cases showing grade 3 or grade 4 responses were categorized as good responders whereas poor responders consisted of grade 1 or grade 2 cases. The survival status as of February 15, 2003 was determined.
Fusion Transcript Detection
RNA was extracted from frozen tumors and analyzed by reverse-transcriptase polymerase chain reaction (RT-PCR) for EWS-FLI1 and EWS-ERG, performed as previously described.10
FISH
FISH for deletions involving p16/p14ARF and MTAP was performed on tumor imprints of frozen tumors as described previously.32 The probes consisted of approximately 100-kb fragments of human genomic DNA from the p16/p14ARF (clones P1-1063) or MTAP (P1-1069) regions (both gifts of Dr Alex Kamb), with each used in combination with a chromosome 9 centromere probe (CEP-9; Vysis Inc, Downers Grove, IL). Signal counts were recorded in more than 100 nuclei. Homozygous deletion was defined as 20% or more nuclei lacking both signals for the locus specific probe (p16/p14ARF or MTAP) and showing at least one signal for the CEP-9 probe.
p53 Mutation Analysis
We first screened for p53 missense mutations by immunohistochemistry (IHC) with antibody DO7 (DAKO, Carpinteria, CA; 1:500) as described24 on tumor specimens from 49 patients with available paraffin blocks. Nuclear overexpression of p53 by IHC has high sensitivity but poor specificity for p53 missense mutations.37,38 Therefore, the IHC screening was followed by p53 GeneChip analysis (Affymetrix, Santa Clara, CA) on DNA extracted from frozen tissue in cases with ≥ 10% immunoreactivity (n = 17) or with missing IHC data (n = 11). Mutations with GeneChip report scores ≥ 13 were accepted as genuine based on previous studies showing very high specificity for scores in the 13 to 36 range.39-41 Cases with reproducible borderline scores (5 to 12) for specific base changes were subjected to confirmatory direct sequencing. Based on previous studies, this combined IHC/p53 GeneChip approach is expected to have high detection sensitivity and specificity for p53 missense mutation.41
Statistical Analysis
Statistical computations were performed using S-plus software (2000; Mathsoft Inc, Cambridge, MA). Survival time was from diagnosis to last follow-up (or death). Survival rate was estimated by Kaplan-Meier methodology. The relationship between survival and other variables was investigated using the log-rank test for categorical variables and a score test based on the Cox proportional hazards model for continuous variables. Age was analyzed as a continuous variable. A multivariate model was fitted using Cox regression with variables found to be significant at the univariate level. Associations among categorical and ordinal variables were checked by Fisher's exact test.
RESULTS
P16/P14ARF Deletion and MTAP Co-Deletion
Homozygous and hemizygous deletions of the p16/p14ARF gene were detected in eight (13.3%; Table 1; Fig 1A through D) and six (10%) cases, respectively, and the former group consisted of one metastatic and seven localized cases. FISH analysis showed no deletion of p16/p14ARF in the remaining 46 (76.7%) cases. Because the survival of patients with tumors containing loss of only one copy of p16/p14ARF paralleled that of cases lacking any evidence of deletion (see Univariate Survival Analysis subsection), cases with hemizygous deletion of p16/p14ARF were grouped with nondeleted cases for all subsequent analyses. Of the eight cases with homozygous deletion of the p16/p14ARF gene,six were also studied for co-deletion of MTAP. Homozygous or hemizygous MTAP deletions were observed in four and one cases, respectively, as detected by probe P1-1069 (Fig 1E and F).
p53 Mutation
By IHC, 17 cases showed p53 immunoreactivity in ≥ 10% tumor nuclei. These cases, together with another 11 cases lacking IHC data, were subjected to Affymetrix p53 GeneChip analysis to detect point mutations. This approach identified p53 point mutations in eight (13.3%) of 60 cases (Table 1), including six patients with localized tumors and two with metastatic disease. Affymetrix p53 GeneChip scores for specific point mutations ranged from 14 to 37 in six cases (Table 3). In two cases with borderline scores (ie, 7 and 10), direct sequencing confirmed the base change detected by chip analysis (results not shown). Among the eight cases showing p53 mutation, there was only one case (case 2 in Table 3) that also displayed homozygous deletion of p16/p14ARF. Taken together, the total number of cases with either p53 missense mutation or p16/p14ARF homozygous deletion was 15 (25%; Table 1). Table 3 summarizes the clinical and molecular data on these 15 cases.
Correlations Between Parameters
Among the 32 cases available for evaluation of histological response to chemotherapy, there were nine good, and 23 poor responders. Good chemotherapeutic effect was correlated with the absence of either p53 or p16/p14ARF alterations (P = .03, Fisher's exact test; Table 4).
There was also a trend for p53 mutation alone (P = .14) and the combination of either p53 or p16/p14ARF alteration (P = .13) to show a relationship to stage, reflecting the finding that three of six tumors from patients with metastatic disease at diagnosis contained one of these two alterations, compared with only 11 of the remaining 54 patients. The association of p16/p14ARF status alone and stage did not approach statistical significance.
There were no significant associations among the other variables. Specifically, there was no significant association between EWS-FLI1 fusion type (type 1 v others) and alterations in p53 and p16/p14ARF in the 55 EWS-FLI1 cases. EWS-FLI1 fusion type was also not related to chemotherapy response. The seemingly mutually exclusive relationship between p53 mutation and p16/p14ARF deletion was also not statistically significant. In addition, neither the EWS-FLI1 type nor the status of the p16/p14ARF gene was associated with stage or whether the tumor was axial or peripheral, or skeletal or extraskeletal.
Because the source of the tissues used for molecular analyses was metastatic in 16 cases, we examined the possibility of a relationship between sample source and abnormalities of p53 or p16/p14ARF, but there was none: five of 16 metastatic and 10 of 44 nonmetastatic samples showed p53 or p16/p14ARF alteration (P = .52). This supports the notion that these genetic changes generally occur in the primary tumor at a preclinical time point.
Univariate Survival Analyses
The median follow-up periods among all 60 patients and all 34 survivors were 29 months (range, 6 to 118 months) and 43.5 months (range, 6 to 118 months), respectively. The median overall survival time for all 60 patients was 99 months. At last follow-up, 25 patients were alive with no evidence of disease, nine were alive with disease, 24 were dead of disease, and the remaining two patients died of other causes. Five of six patients presenting with distant metastases died of their disease by 17 months, with a median survival of 12 months in contrast to 99 months in the localized subgroup (Fig 2A). All eight patients showing p53 mutation died before 21 months, with a median survival of 11 months, compared with 99 months in the nonmutated subset (Fig 2B).
By univariate analysis (Table 5), stage (Fig 2A, P < .001; risk ratio [RR] = 2.2; 95% CI, 1.4 to 3.7), p53 mutation (Fig 2B, P < .0001; RR = 3.7; 95% CI, 2.1 to 4.6), and either or both of a p53 or p16 alteration (Fig 2C, P < .0001; RR = 2.5; 95% CI, 1.6 to 3.8), were strong negative predictors of overall survival, whereas p16/p14ARF deletion alone (Fig 2D, P = .06; RR = 1.6; 95% CI, 0.96 to 2.6) only reached marginal statistical significance. Since the six cases with evidence of hemizygous deletion of p16/p14ARF showed no difference in survival compared with the nondeleted cases (not shown), we combined these two groups for the survival analysis.
No other factors were associated with survival, including age, sex, location, and origin of the primary tumor. As in previous studies,10,11 among cases with EWS-FLI1 (n = 55), patients with type 1 fusion showed slightly better survival, but this did not reach statistical significance (Fig 2E, P = .4; RR = 1.5; 95% CI, 0.6 to 3.4).
Multivariate Survival Analyses
In the multivariate analysis (Table 6), the presence of p53 mutation (P < .001; RR = 3.9; 95% CI, 1.9 to 8.0) remained the most significant independent prognostic factor, and p16/p14ARF deletion emerged as the second most important independent variable (P = .03; RR = 1.8; 95% CI, 1.04 to 2.9), while stage lost its statistical significance. Nevertheless, when p53 and p16 were considered together as a single parameter, the presence of either or both of these two genetic alterations (P < .001; RR = 2.3; 95% CI, 1.4 to 3.5) was the strongest negative factor, and stage (P = .04; RR = 2.3; 95% CI, 1.03 to 3.03) became the second most important adverse factor for overall survival.
DISCUSSION
Homozygous deletion of p16/p14ARF (n = 8, 13%) was as frequent as p53 mutation in this series, a somewhat lower prevalence than most previous studies (Table 1). Possible reasons for this difference may include variability in methodologies or in criteria to define positive cases, and small sample sizes in previous studies. By univariate analysis, p16/p14ARF deletion alone had only marginal value as a negative factor. However, in the multivariate analysis, p16/p14ARF homozygous deletion emerged as the second most significant factor after p53 mutation. The lack of significance of stage in the latter multivariate analysis should be interpreted with caution given the low number of metastatic samples (n = 6) in the present study. Finally, when p16/p14ARF deletion and p53 mutation were combined as a single factor, it was the most crucial determinant of overall survival followed by stage in both univariate and multivariate analysis. The negative impact of p16/p14ARF loss on survival in ES/PNET is consistent with several previous smaller studies.26,27,42
Some groups have found an association between p16/p14ARF or p53 alterations and advanced stage,23,42 while others have not.24,25 In the current series, the presence of p53 mutation or p16/p14ARF deletion was weakly, but not significantly, related to distant metastases at presentation, but the small number of metastatic cases (n = 6; 10%) limited the power of this analysis. Regardless of this possible association, it is of more practical clinical importance that most p53 mutations or p16/p14ARF deletions are detected in patients with localized disease at diagnosis—22% in the present series.
Our finding that the six cases with hemizygous deletion of p16/p14ARF showed no difference in survival compared with the nondeleted cases suggests that loss of one copy of p16/p14ARF combined with inactivation of the remaining copy by mutation or epigenetic mechanisms is unlikely to be significant in ES/PNET. Indeed, Kovar et al17 found only one tumor in 27 to show mutation of p16/p14ARF with loss of heterozygosity. In another study, the presence of p16 promoter hypermethylation had no impact on survival of ES/PNET patients.27 Finally, further evidence indicating that p16 promoter hypermethylation is rare in ES/PNET is that the prevalence of loss of p16 protein expression in both tumors and cell lines closely approximates that of p16/p14ARF homozygous deletion.42,43 In aggregate, these data support homozygous deletion as the principal mode of p16/p14ARF inactivation in ES/PNET.
The finding of combined p16/p14ARF deletion and p53 mutation in only a single case of a total of 15 with either or both alterations might suggest that alterations in these two genes are mutually exclusive. However, this apparent reciprocal relationship p16/p14ARF deletion and p53 mutation did not approach statistical significance. This, along with the report of another ES/PNET case with both alterations26 and of several ES/PNET cell lines with both p16/p14ARF deletion and p53 mutation,17 reinforces the impression that the infrequent coincidence of these two alterations in primary ES/PNET stems mainly from their low prevalence. Other modes of inactivation of the p53 and Rb pathways, through MDM2, CCND1, or Rb alterations, are rare in ES/PNET.16,44
The striking impact of p53 and p16/p14ARF alterations on prognosis points to profound biologic differences inconsistent with a functionally equivalent perturbation of these pathways in the remaining 75% of cases. The strong clinical impact of these alterations in ES/PNET is reminiscent of the marked impact of impaired apoptosis due to p14ARF or p53 loss on lymphomagenesis in Em-myc transgenic mice, a model for P53 pathway alterations in human tumors with translocation oncogenes.45,46 Oncogenic or inappropriate proliferative stimuli are known to elicit an apoptotic or senescence response dependent on the P53/p14ARF pathway.47,48 By analogy, it is tempting to hypothesize that the oncogenicity of EWS-FLI1 may be modulated or tempered by the apoptotic function of an intact P53/p14ARF pathway. Indeed, data from three groups support an important relationship between EWS-FLI1 and the P53 pathway.
Kovar et al found that reintroduction of normal p53 in p53-null ES/PNET cell lines triggers extensive apoptosis.49,50 It should be noted that essentially all ES/PNET cell lines contain p53 alterations or p16/p14ARF deletions, suggesting selection pressure for these genetic alterations to permit in vitro growth.8,51 These data suggest that reintroduction of functional p53 in ES/PNET cell lines promptly leads EWS-FLI1 to trigger a p53-dependent apoptotic program because ES/PNET cells cannot overcome an apoptotic response to in vitro adherent culture conditions unless the p53 pathway is nonfunctional. Deneen et al showed that EWS-FLI1 induces apoptosis and growth arrest in normal mouse embryonic fibroblasts (MEF).52 However, in p16/p14ARF–null or p53-deficient MEFs, apoptosis and growth arrest in response to EWS-FLI1 were reduced.52 Finally, Lessnick et al53 found that in primary human fibroblasts immortalized with telomerase cDNA, expression of EWS-FLI triggered growth arrest associated with transcriptional upregulation of p53. While these data suggest that specific response mechanisms to EWS-FLI1 may differ in different experimental settings, taken together, they support a central role for the p53 pathway in the cellular response to EWS-FLI1.
A similar pattern of infrequent but highly clinically significant p53 or p16/p14ARF alterations has been reported in several other developmental and/or translocation-associated cancers.54 For instance, the outcome of Wilms' tumors,55,56 neuroblastomas,57 follicular lymphomas,58,59 and myxoid liposarcomas60 is dramatically worsened by p53/p14ARF pathway alterations. This is in contrast to most carcinomas and most sarcomas lacking specific translocations, where p53 alterations usually have only moderate or marginal clinical significance.61-63
MTAP is co-deleted in 75% to 90% of tumors, with p16/p14ARF homozygous deletion32 reflecting the close genetic linkage between these two genes. Indeed, four of our six p16/p14ARF–deleted cases (66%) had also lost both copies of MTAP. Codeletion of the MTAP gene with p16/p14ARF has been observed in a variety of cancers.30,32,64-66 Because cancer cells that lack MTAP are unable to salvage adenine from methylthioadenosine, they become dependent on the de novo synthesis pathway of purine metabolism. Thus, MTAP-deleted cancer cells are highly and specifically sensitive to inhibition of the de novo synthesis pathway (eg, by L-alanosine, also known as SDX-102), presenting an opportunity for tailored chemotherapy.33,64 Although only a small proportion of patients with ES/PNET would be candidates for such MTAP-directed chemotherapeutic approaches, it is a subset of ES/PNET that includes approximately a third of the most lethal cases.
We and others have previously shown that the survival of patients with localized disease whose ES/PNET bear the type 1 EWS-FLI1 fusion is better than those with tumors containing other EWS-FLI1 fusion types.10-12 These series have ranged from 55 to 83 informative cases.12 In the present series, among the 55 EWS-FLI1 cases (including six metastatic), EWS-FLI1 fusion type was not a statistically significant prognostic factor, though the survival trends did parallel the findings in prior larger series in that patients with the type 1 fusion seemed to have better outcomes (Fig 2E). The prognostic impact of EWS-FLI1 fusion type appears lesser than that of p53 or p16/p14ARF alterations and may not be demonstrable unless a larger sample size is studied.
A poor histologic response to preoperative chemotherapy has been found to be a highly significant adverse factor in studies of 74 and 118 patients with ES/PNET, respectively.36,67 The limited number of cases in the present study with available chemotherapy response data (n = 32) made the series unsuitable for re-evaluation of this question. However, a significant correlation was observed between good chemotherapeutic effect and the absence of genetic alterations in p53 or p16/p14ARF. Previous studies relating p53 alterations (by IHC) to chemoresponse in ES/PNET reached conflicting conclusions.23,24 P53 mutations have been shown to be associated with resistance to chemotherapy and radiotherapy in diverse tumor types, including sarcomas.68,69 The drug resistance caused by p53 dysfunction can be exerted through several underlying mechanisms.68,69 In addition, p14ARF deletions promote chemoresistance in murine Eμ-myc lymphomas by disabling p53 function with consequent apoptotic defects,45 further emphasizing the links between p16/p14ARF and p53 in oncogenic transformation and chemoresistance.
There had so far been no strong prognostic factor available for prechemotherapy risk stratification of patients presenting with localized ES/PNET. The present retrospective study suggests that alterations in p53 and p16/p14ARF, found in approximately a fourth of all ES/PNET, are strong negative predictors of overall survival, thereby providing for the first time a robust molecular marker of clinical outcome in ES/PNET that could be used to assign certain patients with localized disease to high-risk regimens before initial chemotherapy. Definitive confirmation of these results will require a prospective analysis of these molecular prognostic markers in a larger series of patients with ES/PNET, with longer follow-up.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank the following for providing clinical follow-up data: Ronald Jaffe, MD, Children's Hospital of Pittsburgh; Alex Aledo, MD, New York Presbyterian Hospital; Charles F. Timmons, MD, PhD, Children's Medical Center of Dallas.
NOTES
Supported by American Cancer Society Project Grant 99-216 (M.L.) and the Ewing Sarcoma Research Fund (M.L.).
H.-Y.H. was a visiting fellow sponsored by Chang Gung Memorial Hospital, Kaohsiung Medical Center, Kaohsiung, Taiwan.
H.-Y.H. and P.B.I. contributed equally to this work. P.B.I. is presently in the Department of Pathology, New York University Medical Center, 560 First Ave, New York, NY 10016.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Kolb EA, Kushner BH, Gorlick R, et al: Long-term event-free survival after intensive chemotherapy for Ewing's family of tumors in children and young adults. J Clin Oncol 21:3423-3430, 2003
2. Shankar AG, Ashley S, Craft AW, et al: Outcome after relapse in an unselected cohort of children and adolescents with Ewing sarcoma. Med Pediatr Oncol 40:141-147, 2003
3. Meyers PA, Krailo MD, Ladanyi M, et al: High-dose melphalan, etoposide, total-body irradiation, and autologous stem-cell reconstitution as consolidation therapy for high-risk Ewing's sarcoma does not improve prognosis. J Clin Oncol 19:2812-2820, 2001
4. Delattre O, Zucman J, Plougastel B, et al: Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours. Nature 359:162-165, 1992
5. Zucman J, Melot T, Desmaze C, et al: Combinatorial generation of variable fusion proteins in the Ewing family of tumours. EMBO J 12:4481-4487, 1993
6. Sorensen PHB, Lessnick SL, Lopez-Terrada D, et al: A second Ewing's sarcoma translocation, t(21;22), fuses the EWS gene to another ets-family transcription factor, ERG. Nat Genet 6:146-151, 1994
7. Arvand A, Denny CT: Biology of EWS/ETS fusions in Ewing's family tumors. Oncogene 20:5747-5754, 2001
8. Ladanyi M: EWS-FLI1 and Ewing's sarcoma: Recent molecular data and new insights. Cancer Biol Ther 1:330-336, 2002
9. Shing DC, McMullan DJ, Roberts P, et al: FUS/ERG gene fusions in Ewing's tumors. Cancer Res 63:4568-4576, 2003
10. Zoubek A, Dockhorn-Dworniczak B, Delattre O, et al: Does expression of different EWS chimeric transcripts define clinically distinct risk groups of Ewing tumor patients J Clin Oncol 14:1245-1251, 1996
11. de Alava E, Kawai A, Healey JH, et al: EWS-FLI1 fusion transcript structure is an independent determinant of prognosis in Ewing's sarcoma. J Clin Oncol 16:1248-1255, 1998
12. Aryee DN, Sommergruber W, Muehlbacher K, et al: Variability in gene expression patterns of Ewing tumor cell lines differing in EWS-FLI1 fusion type. Lab Invest 80:1833-1844, 2000
13. Lin PP, Brody RI, Hamelin A, et al: Differential transactivation by alternative EWS-FLI1 fusion proteins correlates with clinical heterogeneity in Ewing's sarcoma. Cancer Res 59:1428-1432, 1999
14. de Alava E, Panizo A, Antonescu CR, et al: Association of EWS-FLI1 type 1 fusion with lower proliferative rate in Ewing's sarcoma. Am J Pathol 156:849-855, 2000
15. Ginsberg JP, de Alava E, Ladanyi M, et al: EWS-FLI1 and EWS-ERG gene fusions are associated with similar clinical phenotypes in Ewing's sarcoma. J Clin Oncol 17:1809-1814, 1999
16. Kovar H, Auinger A, Jug G, et al: Narrow spectrum of infrequent p53 mutations and absence of MDM2 amplification in Ewing tumours. Oncogene 8:2683-2690, 1993
17. Kovar H, Jug G, Aryee DNT, et al: Among genes involved in the RB dependent cell cycle regulatory cascade, the p16 tumor suppressor gene is frequently lost in the Ewing family of tumors. Oncogene 15:2225-2232, 1997
18. Komuro H, Hayashi Y, Kawamura M, et al: Mutations of the p53 gene are involved in Ewing's sarcomas but not in neuroblastomas. Cancer Res 53:5284-5288, 1993
19. Hamelin R, Zucman J, Melot T, et al: P53 mutations in human tumors with chimeric EWS/FLI-1 genes. Int J Cancer 57:336-340, 1994
20. Mangham D, Cannon A, Li X, et al: p53 overexpression in Ewing's sarcoma/primitive neuroectodermal tumour is an uncommon event. J Clin Pathol: Mol Pathol 48:M79-M82, 1995
21. Patino-Garcia A, Sierrasesumaga L: Analysis of the p16INK4 and TP53 tumor suppressor genes in bone sarcoma pediatric patients. Cancer Genet Cytogenet 98:50-55, 1997
22. Radig K, Schneider-Stock R, Rose I, et al: p53 and ras mutations in Ewing's sarcoma. Pathol Res Pract 194:157-162, 1998
23. Abudu A, Mangham DC, Reynolds GM, et al: Overexpression of p53 protein in primary Ewing's sarcoma of bone: Relationship to tumour stage, response and prognosis. Br J Cancer 79:1185-1189, 1999
24. de Alava E, Antonescu CR, Panizo A, et al: Prognostic impact of P53 status in Ewing's sarcoma. Cancer 89:783-792, 2000
25. Wei G, Antonescu CR, de Alava E, et al: Prognostic impact of INK4A deletion in Ewing's sarcoma. Cancer 89:793-799, 2000
26. Tsuchiya T, Sekine K, Hinohara S, et al: Analysis of the p16INK4, p14ARF, p15, TP53, and MDM2 genes and their prognostic implications in osteosarcoma and Ewing sarcoma. Cancer Genet Cytogenet 120:91-98, 2000
27. Lopez-Guerrero JA, Pellin A, Noguera R, et al: Molecular analysis of the 9p21 locus and p53 genes in Ewing family tumors. Lab Invest 81:803-814, 2001
28. Sherr CJ, McCormick F: The RB and p53 pathways in cancer. Cancer Cell 2:103-112, 2002
29. Sherr CJ: The INK4a/ARF network in tumour suppression. Nat Rev Mol Cell Biol 2:731-737, 2001
30. Chen ZH, Zhang H, Savarese TM: Gene deletion chemoselectivity: Codeletion of the genes for p16(INK4), methylthioadenosine phosphorylase, alpha- and beta-interferons in human pancreatic cell carcinoma lines and its implications for chemotherapy. Cancer Res 56:1083-1090, 1996
31. Hori Y, Hori H, Yamada Y, et al: The methylthioadenosine phosphorylase gene is frequently co-deleted with the p16INK4a gene in acute type adult T-cell leukemia. Int J Cancer 75:51-56, 1998
32. Illei PB, Rusch VW, Zakowski MF, et al: Homozygous deletion of CDKN2A and co-deletion of the methylthioadenosine phosphorylase gene in the majority of pleural mesotheliomas. Clin Cancer Res 9:2108-2113, 2003
33. Batova A, Diccianni MB, Omura-Minamisawa M, et al: Use of alanosine as a methylthioadenosine phosphorylase-selective therapy for T-cell acute lymphoblastic leukemia in vitro. Cancer Res 59:1492-1497, 1999
34. Ladanyi M, Bridge JA: Contribution of molecular genetic data to the classification of sarcomas. Hum Pathol 31:532-538, 2000
35. Lawlor ER, Mathers JA, Bainbridge T, et al: Peripheral Primitive Neuroectodermal Tumors in Adults: Documentation by Molecular Analysis. J Clin Oncol 16:1150-1157, 1998
36. Wunder JS, Paulian G, Huvos AG, et al: The histological response to chemotherapy as a predictor of the oncological outcome of operative treatment of Ewing sarcoma. J Bone Joint Surg 80:1020-1033, 1998
37. Baas IO, Mulder JW, Offerhaus GJ, et al: An evaluation of six antibodies for immunohistochemistry of mutant p53 gene product in archival colorectal neoplasms. J Pathol 172:5-12, 1994
38. Hsia CC, Nakashima Y, Thorgeirsson SS, et al: Correlation of immunohistochemical staining and mutations of p53 in human hepatocellular carcinoma. Oncol Rep 7:353-356, 2000
39. Ahrendt SA, Halachmi S, Chow JT, et al: Rapid p53 sequence analysis in primary lung cancer using an oligonucleotide probe array. Proc Natl Acad Sci U S A 96:7382-7387, 1999
40. Wen WH, Bernstein L, Lescallett J, et al: Comparison of TP53 mutations identified by oligonucleotide microarray and conventional DNA sequence analysis. Cancer Res 60:2716-2722, 2000
41. Wikman FP, Lu ML, Thykjaer T, et al: Evaluation of the performance of a p53 sequencing microarray chip using 140 previously sequenced bladder tumor samples. Clin Chem 46:1555-1561, 2000
42. Maitra A, Roberts H, Weinberg AG, et al: Aberrant expression of tumor suppressor proteins in the Ewing family of tumors. Arch Pathol Lab Med 125:1207-1212, 2001
43. Dauphinot L, De Oliveira C, Melot T, et al: Analysis of the expression of cell cycle regulators in Ewing cell lines: EWS-FLI-1 modulates p57KIP2and c-Myc expression. Oncogene 20:3258-3265, 2001
44. Ladanyi M, Lewis R, Jhanwar SC, et al: MDM2 and CDK4 gene amplification in Ewing's sarcoma. J Pathol 175:211-217, 1995
45. Schmitt CA, McCurrach ME, de Stanchina E, et al: INK4a/ARF mutations accelerate lymphomagenesis and promote chemoresistance by disabling p53. Genes Dev 13:2670-2677, 1999
46. Eischen CM, Weber JD, Roussel MF, et al: Disruption of the ARF-Mdm2-p53 tumor suppressor pathway in Myc-induced lymphomagenesis. Genes Dev 13:2658-2669, 1999
47. Lowe SW: Activation of p53 by oncogenes. Endocr Relat Cancer 6:45-48, 1999
48. Schmitt CA, Fridman JS, Yang M, et al: A senescence program controlled by p53 and p16INK4a contributes to the outcome of cancer therapy. Cell 109:335-346, 2002
49. Kovar H, Jug G, Printz D, et al: Characterization of distinct consecutive phases in non-genotoxic p53-induced apoptosis of Ewing tumor cells and the rate-limiting role of caspase 8. Oncogene 19:4096-4107, 2000
50. Kovar H, Pospisilova S, Jug G, et al: Response of Ewing tumor cells to forced and activated p53 expression. Oncogene 22:3193-3204, 2003
51. Kovar H, Auinger A, Jug G, et al: The concept of tumor cell survival factors to explain discrepant p53 mutation frequencies in primary Ewing tumors versus cell lines. Proc Am Assoc Cancer Res 36:2, 1995 (abstr 2)
52. Deneen B, Denny CT: Loss of p16 pathways stabilizes EWS/FLI1 expression and complements EWS/FLI1 mediated transformation. Oncogene 20:6731-6741, 2001
53. Lessnick SL, Dacwag CS, Golub TR: The Ewing's sarcoma oncoprotein EWS/FLI induces a p53-dependent growth arrest in primary human fibroblasts. Cancer Cell 1:393-401, 2002
54. Peller S, Rotter V: TP53 in hematological cancer: Low incidence of mutations with significant clinical relevance. Hum Mutat 21:277-284, 2003
55. Bardeesy N, Falkoff D, Petruzzi MJ, et al: Anaplastic Wilms' tumour, a subtype displaying poor prognosis, harbours p53 gene mutations. Nat Genet 7:91-97, 1994
56. Bardeesy N, Beckwith JB, Pelletier J: Clonal expansion and attenuated apoptosis in Wilms' tumors are associated with p53 gene mutations. Cancer Res 55:215-219, 1995
57. Keshelava N, Zuo JJ, Waidyaratne NS, et al: p53 mutations and loss of p53 function confer multidrug resistance in neuroblastoma. Med Pediatr Oncol 35:563-568, 2000
58. Sander CA, Yano T, Clark HM, et al: p53 mutation is associated with progression in follicular lymphomas. Blood 82:1994-2004, 1993
59. Elenitoba-Johnson KS, Gascoyne RD, Lim MS, et al: Homozygous deletions at chromosome 9p21 involving p16 and p15 are associated with histologic progression in follicle center lymphoma. Blood 91:4677-4685, 1998
60. Antonescu CR, Tschernyavsky SJ, Decuseara R, et al: Prognostic impact of P53 status, TLS-CHOP fusion transcript structure, and histological grade in myxoid liposarcoma: A molecular and clinicopathologic study of 82 cases. Clin Cancer Res 7:3977-3987, 2001
61. Wurl P, Taubert H, Meye A, et al: Prognostic value of immunohistochemistry for p53 in primary soft-tissue sarcomas: A multivariate analysis of five antibodies. J Cancer Res Clin Oncol 123:502-508, 1997
62. Petersen S, Thames HD, Nieder C, et al: The results of colorectal cancer treatment by p53 status: Treatment-specific overview. Dis Colon Rectum 44:322-333, 2001
63. Liu MC, Gelmann EP: P53 gene mutations: Case study of a clinical marker for solid tumors. Semin Oncol 29:246-257, 2002
64. Harasawa H, Yamada Y, Kudoh M, et al: Chemotherapy targeting methylthioadenosine phosphorylase (MTAP) deficiency in adult T cell leukemia (ATL). Leukemia 16:1799-1807, 2002
65. M'soka TJ, Nishioka J, Taga A, et al: Detection of methylthioadenosine phosphorylase (MTAP) and p16 gene deletion in T cell acute lymphoblastic leukemia by real-time quantitative PCR assay. Leukemia 14:935-940, 2000
66. Zhang H, Chen ZH, Savarese TM: Codeletion of the genes for p16INK4, methylthioadenosine phosphorylase, interferon-alpha1, interferon-beta1, and other 9p21 markers in human malignant cell lines. Cancer Genet Cytogenet 86:22-28, 1996
67. Picci P, Bohling T, Bacci G, et al: Chemotherapy-induced tumor necrosis as a prognostic factor in localized Ewing's sarcoma of the extremities. J Clin Oncol 15:1553-1559, 1997
68. Ferreira CG, Tolis C, Giaccone G: p53 and chemosensitivity. Ann Oncol 10:1011-1021, 1999
69. Pirollo KF, Bouker KB, Chang EH: Does p53 status influence tumor response to anticancer therapies Anticancer Drugs 11:419-432, 2000(Hsuan-Ying Huang, Peter B)