Research
www. AJOG.org
ONCOLOGY
Deoxyribonucleic acid ploidy in endometrial carcinoma: a reproducible and valid prognostic marker in a routine diagnostic setting Elisabeth Wik, MD; Jone Trovik, MD; Ole E. Iversen, MD, PhD; Ingeborg B. Engelsen, MD, PhD; Ingunn M. Stefansson, MD, PhD; Liv C. Vestrheim, MD; Hans K. Haugland, MD, PhD; Lars A. Akslen, MD, PhD; Helga B. Salvesen, MD, PhD OBJECTIVE: The objective of the study was to investigate the prognos-
tic impact of deoxyribonucleic acid (DNA) ploidy in endometrial carcinoma in a routine diagnostic series as compared with a research series. STUDY DESIGN: We studied a population-based series of 363 endo-
metrial carcinomas prospectively collected, with long and complete follow-up. The prognostic value of DNA ploidy was investigated in a routine diagnostic series (n ⫽ 262) and compared with the results from a previous research series (n ⫽ 101). RESULTS: The proportion of DNA aneuploid tumors was 21% in the
research series and 25% in the routine diagnostic series (P ⫽ NS).
In both series, DNA aneuploidy was significantly correlated to higher age at diagnosis, nonendometrioid subtype, and high histologic grade. Patients with DNA aneuploid tumors had significantly poorer survival, adjusted for established clinicopathologic prognostic factors. CONCLUSION: DNA ploidy estimation in endometrial carcinoma adds
independent prognostic information in a routine diagnostic setting. Key words: deoxyribonucleic acid ploidy, endometrial carcinoma, prognostic marker, survival, validation study
Cite this article as: Wik E, Trovik J, Iversen OE, et al. Deoxyribonucleic acid ploidy in endometrial carcinoma: a reproducible and valid prognostic marker in a routine diagnostic setting. Am J Obstet Gynecol 2009;201:603.e1-7.
E
ndometrial cancer is the most common gynecologic malignancy in developed countries.1 The incidence has been increasing over the last decades; in Norway, a 50% increase in age-adjusted incidence from 10.5–15.8 per 100,000 person-years has been observed from 1970 –2005.1 Several clinicopathologic factors have been evaluated in relation to course of disease. Age at diagnosis, stage of disease, histologic subtype, histologic grade, and hormone receptor status in tumors have
all been found to be reliable prognostic factors, and these are used today in the decision making regarding treatment.2 Deoxyribonucleic acid (DNA) ploidy has in several studies been found to have prognostic impact,3 and some centers have used this marker as an integrated part of the treatment algorithm for these patients.4-6 Our institution was among the first in exploring the value of DNA ploidy prospectively among endometrial carcinoma patients.7 Based on our finding of DNA ploidy as an independent
From the Institute of Clinical Medicine (Drs Wik, Trovik, Iversen, and Salvesen) and the Section of Pathology, The Gade Institute (Drs Stefansson and Akslen), University of Bergen, Bergen, Norway; and the Departments of Obstetrics and Gynecology (Drs Trovik, Iversen, Engelsen, Vestrheim, and Salvesen) and Pathology (Drs Stefansson, Haugland, and Akslen), Haukeland University Hospital, Bergen, Norway. Presented at a meeting of gynecologic oncologists within the Norwegian Forum for Oncology, Bergen, Norway, Nov. 21, 2008.
prognostic marker, we applied DNA ploidy as a routine test for patients with endometrial cancer. However, the data were not used in clinical decision making. Before clinical implementation of a prognostic marker, it is necessary to examine its significance prospectively, and it might be of importance to examine the marker in a routine diagnostic setting.8 Similar requirements for implementing new markers for treatment stratifications as for implementation of new treatment strategies has been suggested. To our knowledge, no previous studies have validated the prognostic impact of DNA ploidy analysis in a routine diagnostic series of endometrial carcinomas, and we therefore examined the prognostic value of DNA ploidy from routine assessment in 262 patients.
Received Feb. 26, 2009; revised May 20, 2009; accepted July 14, 2009. Reprints: Elisabeth Wik, MD, Institute of Clinical Medicine, Section for Gynecology and Obstetrics, University of Bergen, Haukeland University Hospital, 5021 Bergen, Norway.
[email protected]. This study was supported in part by Helse Vest and a legacy of Harald Andersen to the Norwegian Cancer Society. 0002-9378/$36.00 • © 2009 Published by Mosby, Inc. • doi: 10.1016/j.ajog.2009.07.029
M ATERIALS AND M ETHODS We performed prospective DNA ploidy analyses on fresh, ethanol-fixed tumor tissue from hysterectomy specimens of 262 patients with endometrial carcinomas (routine diagnostic series) and com-
DECEMBER 2009 American Journal of Obstetrics & Gynecology
603.e1
Research
Oncology
pared the findings with previous analyses of 101 cases of endometrial carcinoma also collected prospectively (research series).7,9 Patients were all treated at the Department of Obstetrics and Gynecology, Section of Gynecological Cancer, Haukeland University Hospital (Bergen, Norway). This is a referral hospital for patients in Hordaland County; the area is demographically well defined, with about 450,000 inhabitants, representing approximately 10% of the Norwegian population and with a similar incidence rate and prognosis as the total population of endometrial cancers.10 The clinicopathologic variables included age at diagnosis, International Federation of Gynecology and Obstetrics (FIGO) stage according to the 1988 criteria,11 histologic subtype,12 histologic grade,12 treatment, recurrence, and disease-specific survival and were recorded for all the patients. The prognostic value of DNA ploidy was studied in a research series (n ⫽ 101), 1 researcher performing all parts of the processing and analytical work and in which the DNA analyses were done as part of a research project focusing on the DNA ploidy analysis. We compared these data with the prognostic value of DNA ploidy in the routine diagnostic series (n ⫽ 262), in which DNA ploidy was performed as part of the routine diagnostics at the Department of Pathology, Haukeland University Hospital. The results from the DNA ploidy analyses did not influence treatment decisions.
The research series For the period 1981–1990, DNA ploidy was related to a panel of clinicopathologic variables and survival as previously reported.7,9 The median follow-up period for the survivors was 18.5 years (range, 13.2–23.2). The last date of follow-up was June 30, 2004. The histopathologic diagnoses were revised by 2 experienced pathologists (I.M.S. and L.A.A.). The treatment protocol for the period was abdominal hysterectomy, with bilateral salpingo-oophorectomy as primary treatment. The pelvic and paraaortic lymph nodes were palpated 603.e2
www.AJOG.org and biopsied only if considered suspect, as previously reported.13
The routine diagnostic series DNA ploidy was determined in ethanolfixed endometrial carcinoma tissue from hysterectomy specimens collected prospectively from 262 patients treated in the period 2001–2007 at the same institution. Patients were followed up from the time of primary surgery until November 2008 or until death. The median follow-up period for the survivors was 2.25 years (range, 0.1–7.0 years). Routine histopathologic reports were used intentionally in line with our approach to study the value of DNA ploidy in a routine diagnostic setting. Two patients had not yet had their first follow-up examination at the day of closure of the study and were thus not included in the survival analyses. Among the 37 patients who died during the follow-up period, 28 patients died from endometrial carcinoma, whereas 9 died because of other causes. The surgical treatment protocol for this time period was abdominal hysterectomy, with bilateral salpingo-oophorectomy as primary treatment. The primary surgery did also include pelvic lymph node sampling as staging procedure. Adjuvant therapy was recommended for patients with FIGO stage greater than IIA and high-risk FIGO stages I and IIA patients, defined as nonendometrioid tumors or deeply infiltrating endometrioid grade 3 tumors. Thus, 12.6% received radiation therapy as primary adjuvant treatment, mainly as external radiation (11.8 %), although a few received brachytherapy (0.8%). In addition, 9.2% of the patients received chemotherapy as adjuvant treatment. Pelvic lymphadenectomy was performed in 81% of the cases. The median number of glands removed was 12 (range, 1–50). Positive lymph nodes were found in 6.5% of the patients. Among patients considered cured by primary treatment, 52 recurrences (20%) occurred, of which 15 received radiotherapy, 11 chemotherapy, 6 hormonal therapy, and 4 received a combination of chemo-radiation. Eight patients received other treatment.
American Journal of Obstetrics & Gynecology DECEMBER 2009
DNA ploidy analyses The flow-cytometric method for DNA ploidy analysis in the research series has been published previously.7,14 The tissue for DNA analysis was taken fresh in the surgical operating room from a lesion in the hysterectomy specimen judged macroscopically to be representative for the malignant tumor. Tissue was taken from the same area to confirm by histology that representative tumor tissue had been collected. Tumor tissue was preserved by first adding 1 mL phosphatebuffered saline to the tissue and thereafter ethanol. The analysis was performed using a FacsCalibur flow cytometer (BD Biosciences, San Jose, CA). When only 1 cell population could be identified, the tumor was classified as DNA diploid. In cases with 2 cell populations, the DNA index (DI) was calculated. Cases with DI greater than 1.0 were classified as DNA aneuploid. Because there were no DNA tetraploid or polyploid cases in our study population, we used the term DNA diploid vs aneuploid to describe the DNA ploidy distribution. Use of statistics Data were analyzed using SPSS (Statistical Package of Social Sciences, version 15.0; SPSS, Inc, Chicago, IL). P ⬍ .05 was considered statistically significant. Groups were compared using Pearson 2 test or Fisher exact test when appropriate for categorical variables. Kappa statistics were used to assess reproducibility of the method. Univariate survival analyses of time to death because of endometrial carcinoma (disease-specific survival) were performed using the Kaplan-Meier (product-limit) method. Entry date was the date of primary surgery. Patients who died from other causes were censored at the date of death. Differences in survival were calculated using the Mantel-Cox (log-rank) test. The variables were visually examined by a log-minus-log plot to check the assumptions about proportionality before incorporation into multivariate proportional hazards regression models (Cox analyses). Unadjusted and adjusted hazard ratios were calculated as measures of effect. All P values are 2-sided.
Oncology
www.AJOG.org To evaluate the strength of agreement in the interpretation of the DNA histogram (DNA aneuploid/DNA diploid) in the routine diagnostic series, 2 of the main observers (A and B) carried out repeated assessments of the DNA histograms of 60 patients (50 DNA diploid and 10 DNA aneuploid).
Research
TABLE 1
Characteristics of the 2 populations: the research series and the routine diagnostic series Routine diagnostic series, n (%)
P valuea .9
Variable
Category
Research series, n (%)
Patient age
⬍65
50 (50)
127 (48)
ⱖ65
51 (50)
135 (52)
Pre/perimenopause
19 (19)
34 (13)
...........................................................................................................................................................................
..............................................................................................................................................................................................................................................
R ESULTS There were no statistically significant differences between data in the research series as compared with the routine diagnostic series with respect to the panel of clinicopathologic variables listed in Table 1. There was a significantly higher proportion of aneuploid tumors among the nonendometrioid tumors in the routine diagnostic series as compared with the research series that had been subjected to histopathologic revision (P ⫽ .0001). There were no other significant differences in distribution of clinicopathologic variables when comparing the subgroups of patients with aneuploid tumors in the 2 series. The proportion of DNA aneuploid tumors was 25% (25/101) in the research series as compared with 21% (55/262) in the routine diagnostic series (P ⫽ .5). In the research series, DNA ploidy was significantly correlated to patient age, menopausal status, histologic subtype, and grade. Similarly, age, histologic subtype, and grade, as well as depth of myometrial invasion, were significantly correlated to DNA ploidy in the routine diagnostic series (Table 2). Ploidy influenced survival significantly in both patient series, with a 90% 5-year disease-specific survival for patients with DNA diploid tumors compared with 71% for patients with DNA aneuploid tumors in the research series (P ⫽ .01). In the routine diagnostic series, the 5-year disease-specific survival was 88% for patients with DNA diploid tumors compared with 70% for patients with DNA aneuploid tumors (P ⫽ .001) (Figure 1). When the importance of ploidy was investigated in Cox regression models, an independent prognostic impact was found for ploidy in the research series, as previously reported.7 In the routine di-
Menopausal status
.2
...........................................................................................................................................................................
Postmenopause
82 (81)
228 (87)
Endometrioid
92 (91)
219 (84)
Nonendometrioid
9 (9)
43 (16)
Grades 1 and 2
77 (76)
174 (67)
..............................................................................................................................................................................................................................................
Histologic subtype
.09
...........................................................................................................................................................................
.............................................................................................................................................................................................................................................. b
Histologic grade
.1
...........................................................................................................................................................................
Grade 3
24 (24)
84 (33)
I/II
88 (87)
218 (83)
III/IV
13 (13)
44 (17)
⬍50%
52 (67)
129 (68)
..............................................................................................................................................................................................................................................
FIGO stage
.4
...........................................................................................................................................................................
.............................................................................................................................................................................................................................................. c,d
Myometrial invasion
.9
...........................................................................................................................................................................
ⱖ50%
26 (33)
60 (32)
Diploid
76 (75)
207 (79)
Aneuploid
25 (25)
55 (21)
..............................................................................................................................................................................................................................................
DNA ploidy
.5
...........................................................................................................................................................................
..............................................................................................................................................................................................................................................
FIGO, International Federation of Gynecology and Obstetrics. a
P value calculated by 2 or Fisher exact test according to what was appropriate; b Missing data for histologic grade in 4 cases in the routine diagnostic series; c Only stage I is included for myometrial invasion; d Missing data for myometrial invasion in 3 cases in the research series.
Wik. DNA ploidy in endometrial carcinoma. Am J Obstet Gynecol 2009.
agnostic series, we found significant prognostic impact for age, histologic subtype, histologic grade, FIGO stage, and DNA ploidy in the univariate analyses. These significant variables were included in a multivariate analysis. We found the independent prognostic impact of ploidy to be consistent with the findings in the research series, with a hazard ratio of 2.6 (95% confidence interval [CI], 1.1– 6.0; P ⫽ .02) adjusted for age, FIGO stage, histologic subtype, and grade (Table 3). We performed survival analyses, excluding the cases of carcinosarcomas (n ⫽ 10) from the nonendometrioid group of the routine diagnostic series and found the same survival pattern as when these cases were included in the analyses. The prognostic impact of DNA ploidy was confirmed in analyses comprising the 269 stage I tumors from a merged dataset from both series (P ⫽ .01) (Figure 2). In a separate Cox regression analysis of the routine diagnostic series com-
prising only stage I cases (n ⫽ 176), including DNA ploidy, myometrial invasion, histologic subtype grade, and age, only myometrial infiltration showed significant prognostic impact of hazard ratio (HR) 6.3 (95% CI, 1.2–33), followed by DNA ploidy showing borderline significance (P ⫽ .1; Table 3). There was significantly better survival among patients within the low-risk FIGO stages I (endometrioid, FIGO stages IA/B, grades 1/2) as compared with the intermediate- and high-risk FIGO stage I group (endometrioid, FIGO stage IC, grade 3/nonendometrioid) (P ⬍ .001). Because of the few events in FIGO stage I, larger datasets are needed to investigate the effect of DNA ploidy on prognosis within the subgroups of high/intermediate/low risk. Tests for reproducibility for interpretation of DNA histograms gave a kappa value of 1.0 for observer A’s results compared with routine result from the Department of Pathology. Comparing observer B’s results with the routine report,
DECEMBER 2009 American Journal of Obstetrics & Gynecology
603.e3
Research
Oncology
www.AJOG.org
TABLE 2
DNA ploidy in relation to clinicopathologic prognostic markers in endometrial carcinoma (all subtypes)
Variablea Patient age, y
c
DNA ploidy
DNA ploidy
Research series (n ⴝ 101)
Routine diagnostic series (n ⴝ 262)
1980–1990
2001–2007
Categories
Diploid, n (%)
Aneuploid, n (%)
⬍65
43 (86)
7 (14)
ⱖ65
33 (65)
18 (35)
Pre/perimenopausal
18 (95)
1 (5)
P valueb
Diploid, n (%)
Aneuploid, n (%)
.01
107 (84)
20 (16)
100 (74)
35 (26)
27 (79)
7 (21)
180 (79)
48 (21)
183 (84)
36 (16)
24 (56)
19 (44)
71 (93)
5 (7)
P valueb .04
.......................................................................................................................................................................................................................................................................................................
................................................................................................................................................................................................................................................................................................................................................................................
Menopausal status
.03
.95
.......................................................................................................................................................................................................................................................................................................
Postmenopausal
58 (71)
24 (29)
Endometrioid
74 (80)
18 (20)
2 (22)
7 (78)
Grade 1
18 (95)
1 (5)
Grade 2
47 (80)
11 (19)
79 (81)
19 (19)
Grade 3
11 (48)
13 (54)
55 (66)
29 (34)
I/II
69 (78)
19 (22)
171 (78)
47 (22)
7 (54)
6 (46)
36 (82)
8 (18)
⬍50%
43 (83)
9 (17)
107 (83)
22 (17)
ⱖ50%
18 (69)
8 (31)
39 (65)
21 (35)
................................................................................................................................................................................................................................................................................................................................................................................ d
Histologic subtype
⬍ .001
⬍ .001
.......................................................................................................................................................................................................................................................................................................
Nonendometrioid
................................................................................................................................................................................................................................................................................................................................................................................ e e
Histologic grade
ⱕ.001
⬍ .001
....................................................................................................................................................................................................................................................................................................... .......................................................................................................................................................................................................................................................................................................
................................................................................................................................................................................................................................................................................................................................................................................ f
FIGO stage
.06
.6
.......................................................................................................................................................................................................................................................................................................
III/IV
................................................................................................................................................................................................................................................................................................................................................................................ g,h
Myometrial invasion
.2
.009
.......................................................................................................................................................................................................................................................................................................
................................................................................................................................................................................................................................................................................................................................................................................
FIGO, International Federation of Gynecology and Obstetrics. a
Missing data for histologic grade in 4 cases in the routine diagnostic series and for myometrial invasion in 3 cases in the research series; b 2 test, Pearson 2-sided P value; c Median used as cutpoint for age; d Histopathologic subtyping and grading were based on histopathologic revision in the research series and routine reports in the routine diagnostic series. Adenoacanthoma/adenosquamous are merged with the endometrioid group. Nonendometrioid tumors included in the research series clear cell carcinoma (n ⫽ 6), serous papillary carcinoma (n ⫽ 3), whereas in the routine diagnostic series, there are clear cell carcinoma (n ⫽ 10), serous papillary carcinoma (n ⫽ 17), carcinosarcoma (n ⫽ 10), and undifferentiated carcinoma (n ⫽ 6); e Grades 1 and 2 vs grade 3; f According to the FIGO 1988 criteria; g Stage IIIC in the research series (n ⫽ 1; 1.0%), in the routine diagnostic series (n ⫽ 17; 6.5%); h Data only for FIGO stage I.
Wik. DNA ploidy in endometrial carcinoma. Am J Obstet Gynecol 2009.
the kappa value was 0.95. For the interobserver evaluation, the kappa value was 0.95. The kappa values thus show a very good agreement between the observers in this analysis.
C OMMENT DNA content in tumor cells, evaluated by DNA ploidy analysis, has been found to be a significant prognostic factor in several cancer types,15,16 including endometrial cancer.17 For this tumor type, DNA aneuploidy has been correlated to an aggressive phenotype, including poor prognosis in a substantial number of retrospective studies,3,17-21 although some studies have not confirmed independent prognostic impact.22-25 Differences may be related to the method applied for DNA ploidy estimates; sample size; patient population studied; other prognostic markers taken into account, especially histologic subtype and surgical FIGO stage; extent of lymphadenec603.e4
tomy; and varying measures of outcome.3,21,26-29 Few studies have investigated the prognostic impact of DNA ploidy in multivariate analysis, including all the traditional histopathologic variables, such as age, histologic subtype, histologic grade, and surgical FIGO stage with pelvic lymph node sampling. To the best of our knowledge, this is the first prospective study validating the prognostic impact of DNA ploidy in a routine diagnostic series of endometrial carcinoma. We examined the role of DNA ploidy analyses in a relatively large patient series, with complete recording of clinicopathologic information and follow-up data. DNA ploidy analyses, measured prospectively, were an independent prognostic factor for diseasespecific survival for patients with endometrial carcinoma. DNA ploidy measurements have been used in the stratification in which patients received adjuvant treatment of
American Journal of Obstetrics & Gynecology DECEMBER 2009
early-stage endometrial cancer in other studies.4-6 Hogberg et al4 included DNA ploidy in a risk-defining variable and demonstrated a significant association between survival (both overall and disease specific) and DNA ploidy. One study showed that patients with DNA aneuploid tumors treated with adjuvant radiotherapy had the same risk of relapse as untreated patients with DNA diploid tumors.5 For patients with stage I endometrial cancer, the benefit of adjuvant treatment is uncertain.6,30 Improved identification of patients in risk groups for disease recurrence may be useful for determining the patients with need for adjuvant therapy and thus spare low-risk groups from overtreatment.2 Our findings of an independent prognostic impact of DNA ploidy in endometrial cancer is in agreement with other recent studies.6,31 Susini et al31 examined the prognostic value among 174 patients
Oncology
www.AJOG.org
Research
FIGURE 1
Prognostic value of DNA ploidy in a research series and routine diagnostic series
A, Prognostic impact of DNA ploidy in a research series (n ⫽ 101). B, Prognostic value of DNA ploidy in a routine diagnostic series (n ⫽ 262). Both survival curves are estimated by the Kaplan-Meier method and are based on estimation of DNA ploidy in fresh, ethanol-fixed tumor tissue from hysterectomy specimens. For each category the number of cases is given, followed by the number of endometrial cancer (all subtypes) deaths. Wik. DNA ploidy in endometrial carcinoma. Am J Obstet Gynecol 2009.
with endometrial carcinoma and found that DNA aneuploidy was significantly associated with poor prognosis. However, they found a better separation be-
tween the groups with respect to survival, as compared with our study. This might be explained by the fact that in our study, 81% of the patients underwent
lymph node staging and that including information on lymph node status might influence the impact of other variables in the Cox analyses.
TABLE 3
Routine diagnostic series: prognostic value of DNA ploidy in relation to traditional prognostic markers in endometrial carcinoma (all subtypes) Variable c
Patient age at diagnosis, y
Cases, n (%)
Unadjusted HRa
95% CI
P valueb
256
1.0
1.0–1.1
.04
Adjusted HR
95% CI
1.0
0.982–1.1
P valueb .3
................................................................................................................................................................................................................................................................................................................................................................................
⬍ .001
Histologic subtype
.2
.......................................................................................................................................................................................................................................................................................................................................................................
Endometrioid
216 (84)
1.0
40 (16)
6.4
1.0
.......................................................................................................................................................................................................................................................................................................................................................................
Nonendometrioid
3.0–13.8
2.3
0.7–7.2
................................................................................................................................................................................................................................................................................................................................................................................
⬍ .001
Histologic grade
.3
.......................................................................................................................................................................................................................................................................................................................................................................
Grades 1 and 2
174 (68)
1.0
82 (32)
4.8
1.0
.......................................................................................................................................................................................................................................................................................................................................................................
Grade 3
2.2–10.6
1.8
0.6–6.0
................................................................................................................................................................................................................................................................................................................................................................................
⬍ .001
FIGO
⬍ .001
.......................................................................................................................................................................................................................................................................................................................................................................
I/II
213 (83)
1.0
43 (17)
7.7
1.0
....................................................................................................................................................................................................................................................................................................................................................................... d
III/IV
3.6–16.7
6.9
3.1–15.5
................................................................................................................................................................................................................................................................................................................................................................................
DNA ploidy
.006
.02
.......................................................................................................................................................................................................................................................................................................................................................................
Diploid
203 (79)
1.0
53 (21)
2.9
1.0
....................................................................................................................................................................................................................................................................................................................................................................... e
Aneuploid
1.4–6.3
2.6
1.2–6.0
................................................................................................................................................................................................................................................................................................................................................................................
The HR for DNA ploidy 3.0 (95% CI, 1.3– 69) was adjusted for lymph node status and the other variables listed in the table. CI, confidence interval; FIGO, International Federation of Gynecology and Obstetrics; HR, hazard ratio. a
Unadjusted and adjusted HRs are based on cases with data available for all variables in the multivariate analysis (n ⫽ 256); b Likelihood ratio test; c Continuous variable with HR given per year; d Including 17 patients with FIGO stage IIIC; e HR for DNA ploidy 4.5 (95% CI, 0.7–27) adjusted for histologic subtype, grade, age, and myometrial invasion; only FIGO stage I tumors were included (n ⫽ 176).
Wik. DNA ploidy in endometrial carcinoma. Am J Obstet Gynecol 2009.
DECEMBER 2009 American Journal of Obstetrics & Gynecology
603.e5
Research
Oncology
www.AJOG.org
FIGURE 2
Kaplan-Meier survival analysis
Kaplan-Meier survival analysis, comprising the 269 FIGO stage I endometrial carcinoma patients from both datasets (research series and routine diagnostic series). For each category the number of cases is given, followed by the number of endometrial cancer (all subtypes) deaths. FIGO, International Federation of Gynecology and Obstetrics. Wik. DNA ploidy in endometrial carcinoma. Am J Obstet Gynecol 2009.
The effect of lymphadenectomy for disease-specific survival is debated.30,32-33 Only pelvic lymphadenectomy was performed as standard surgical treatment in the routine diagnostic series. The proportion of skip metastasis, defined as metastasis in aortic glands without metastasis in pelvic lymph glands, has been reported to be about 1%.32 The present study was not designed to evaluate the role of aortic lymphadenectomy or a survival effect of lymphadenectomy in general. Still, in line with previous literature,2 we found a worse prognosis for patients with positive pelvic lymph nodes compared with those with negative nodes (P ⬍ .001). Pelvic lymphadenectomy will move patients from FIGO I to FIGO IIIC. Despite this fact, we could not detect a significant change in survival within each FIGO stage for the research and routine diagnostic series. Similarly, for DNA ploidy there was no change in prognostic impact within each FIGO stage (data not 603.e6
shown). This may well be due to relatively small subgroups in the datasets. In the study by Mangili et al,6 multivariate survival analyses showed that DNA ploidy is an independent prognostic factor for the subgroup of FIGO stage I endometrial carcinomas. This appears to be in line with our results from the subgroup univariate analysis including FIGO stage I cancers only. Further subgroup analyses according to risk groups within FIGO stage I would be interesting but would require larger datasets for sufficient power. Acceptable reproducibility is important for a method’s usefulness in a clinical setting.34 In the present study, we found a very good intra- and interobserver reproducibility for evaluation of the flow cytometric histograms. Falsenegative results for detecting DNA aneuploidy may be due to the high amount of stromal tissue and tumour heterogeneity.17,35 Method of extraction, the number and proportion of tumour cells ana-
American Journal of Obstetrics & Gynecology DECEMBER 2009
lyzed, and the program for debris correction may also influence the results.17 Despite these potential pitfalls when applying the method, our study indicates that DNA ploidy measurements by flow cytometry are robust enough to detect prognostic impact independent of the traditional clinicopathologic prognostic markers when implemented in a routine diagnostic setting. Patients with aneuploid tumors appear to have a better survival the first 2 years after diagnosis in the routine diagnostic series as compared with the research series (Figure 1). This pattern was not seen for stage I tumors (data not shown). This difference in curve shape might be explained by the fact that patients in the routine diagnostic series received chemotherapy more often as compared with patients in the research series. However, this study was not designed to evaluate the role of DNA ploidy as a predictive factor for treatment response in endometrial carcinomas. DNA ploidy in curettage tumor material has previously been shown to have prognostic impact.36 We have not assessed ploidy in curettage tumor material in the present study, but it would be interesting to compare DNA ploidy in tumor tissue from curettage with hysterectomy specimens in future studies. In conclusion, we find that DNA ploidy analysis is a robust method, with very good reproducibility in a routine diagnostic series as compared with a research series. We also validate its ability to detect independent, prognostically relevant information for surgically staged endometrial carcinoma patients f in a routine diagnostic setting. ACKNOWLEDGMENTS We thank Gerd Lillian Hallseth, Bendik Nordanger, Britt Edvardsen, Raymond Lygre, Åsne Mokleiv, and Erlend Njølstad for excellent technical assistance. The research has been approved by the Norwegian Data Inspectorate (961478-2), Norwegian Social Sciences Data Services (15501), and the local ethical committee (REKIII 052.01). Women gave informed consent.
Oncology
www.AJOG.org REFERENCES 1. Cancer Registry of Norway. Cancer in Norway 2005. Special edition. Oslo, Norway: Institute of Population-Based Cancer Research; 2005. 2. Amant F, Moerman P, Neven P, Timmerman D, Van Limbergen E, Vergote I. Endometrial cancer. Lancet 2005;366:491-505. 3. Salvesen HB, Akslen LA. Molecular pathogenesis and prognostic factors in endometrial carcinoma. Apmis 2002;110:673-89. 4. Hogberg T, Fredstorp-Lidebring M, Alm P, et al. A prospective population-based management program including primary surgery and postoperative risk assessment by means of DNA ploidy and histopathology. Adjuvant radiotherapy is not necessary for the majority of patients with FIGO stage I-II endometrial cancer. Int J Gynecol Cancer 2004;14:437-50. 5. Lim P, Aquino-Parsons CF, Wong F, et al. Low-risk endometrial carcinoma: assessment of a treatment policy based on tumor ploidy and identification of additional prognostic indicators. Gynecol Oncol 1999;73:191-5. 6. Mangili G, Montoli S, De Marzi P, et al. The role of DNA ploidy in postoperative management of stage I endometrial cancer. Ann Oncol 2008;19:1278-83. 7. Iversen OE. Flow cytometric deoxyribonucleic acid index: a prognostic factor in endometrial carcinoma. Am J Obstet Gynecol 1986; 155:770-6. 8. Hall PA. DNA ploidy analysis in histopathology. DNA ploidy studies in pathology—a critical appraisal. Histopathology 2004;44:614-20. 9. Salvesen HB, Iversen OE, Akselen LA. Identification of high-risk patients by assessment of nuclear Ki-67 expression in a prospective study of endometrial carcinomas. Clin Cancer Res 1998;4:2779-85. 10. The Cancer Registry of Norway. Cancer in Norway, 1996, 1st ed. Oslo, Norway: Institute for Epidemiological Cancer Research; 1999. 11. Announcement. FIGO stage, 1988 revision. Gynecol Oncol 1989;35:125-7. 12. Silverberg SG, Kurman RJ, Nogales F, et al. Tumors of the uterine corpus. In: Tavassoli FA, Devilee P, eds. Tumours of the breast and female genital organs. World Health Organization classification of tumours: pathology and genetics, 5th ed. Lyons, France: IACR Press, Inc; 2003:217-58. 13. Engelsen IB, Stefansson I, Akslen LA, Salvesen HB. Pathologic expression of p53 or
p16 in preoperative curettage specimens identifies high-risk endometrial carcinomas. Am J Obstet Gynecol 2006;195:979-86. 14. Iversen OE, Utaaker E, Skaarland E. DNA ploidy and steroid receptors as predictors of disease course in patients with endometrial carcinoma. Acta Obstet Gynecol Scand 1988; 67:531-7. 15. Barlogie B, Raber MN, Schumann J, et al. Flow cytometry in clinical cancer research. Cancer Res 1983;43:3982-97. 16. Friedlander ML, Hedley DW, Taylor IW. Clinical and biological significance of aneuploidy in human tumours. J Clin Pathol 1984;37:961-74. 17. Fox H. Ploidy in gynaecological cancers. Histopathology 2005;46:121-9. 18. Kodama S, Kase H, Tanaka K, Matsui K. Multivariate analysis of prognostic factors in patients with endometrial cancer. Int J Gynaecol Obstet 1996;53:23-30. 19. Lukes AS, Kohler MF, Pieper CF, et al. Multivariable analysis of DNA ploidy, p53, and HER2/neu as prognostic factors in endometrial cancer. Cancer 1994;73:2380-5. 20. Nordstrom B, Strang P, Bergstrom R, Nilsson S, Tribukait B. A comparison of proliferation markers and their prognostic value for women with endometrial carcinoma. Ki-67, proliferating cell nuclear antigen, and flow cytometric S-phase fraction. Cancer 1996;78:1942-51. 21. Terada K, Mattson D, Goo D, Shimizu D. DNA aneuploidy is associated with increased mortality for stage I endometrial cancer. Gynecol Oncol 2004;95:483-7. 22. Konski A, Domenico D, Tyrkus M, et al. Prognostic characteristics of surgical stage I endometrial adenocarcinoma. Int J Radiat Oncol Biol Phys 1996;35:935-40. 23. Mariani A, Sebo TJ, Webb MJ, et al. Molecular and histopathologic predictors of distant failure in endometrial cancer. Cancer Detect Prev 2003;27:434-41. 24. Orbo A, Rydningen M, Straume B, Lysne S. Significance of morphometric, DNA cytometric features, and other prognostic markers on survival of endometrial cancer patients in northern Norway. Int J Gynecol Cancer 2002;12:49-56. 25. Sorbe B, Risberg B, Frankendal B. DNA ploidy, morphometry, and nuclear grade as prognostic factors in endometrial carcinoma. Gynecol Oncol 1990;38:22-7. 26. Braly PS. Flow cytometry as a prognostic factor in endometrial cancer—what does it add? Gynecol Oncol 1995;58:145-7.
Research
27. Lindahl B, Alm P, Fernø M, et al. Prognostic value of flow cytometrical DNA measurements in stage I-II endometrial carcinoma: correlations with steroid receptor concentration, tumor myometrial invasion, and degree of differentiation. Anticancer Res 1987;7:791-7. 28. Nordstrøm B, Strang P, Lindgren A, Bergstrøm R, Tribukait B. Carcinoma of the endometrium: do the nuclear grade and DNA ploidy provide more prognostic information than do the FIGO and WHO classifications? Int J Gynecol Pathol 1996;15:191-201. 29. Pradhan M, Abeler VM, Danielsen HE, Trope CG, Risberg BÅ. Image cytometry DNA ploidy correlates with histological subtypes in endometrial carcinomas. Mod Pathol 2006; 19:1227-35. 30. Creutzberg CL, van Putten WL, Koper PC, et al. Surgery and postoperative radiotherapy versus surgery alone for patients with stage-1 endometrial carcinoma: multicentre randomised trial. PORTEC Study Group. Post operative radiation therapy in endometrial carcinoma. Lancet 2000;355:1404-11. 31. Susini T, Amunni G, Molino C, et al. Tenyear results of a prospective study on the prognostic role of ploidy in endometrial carcinoma: dNA aneuploidy identifies high-risk cases among the so-called ‘low-risk’ patients with well and moderately differentiated tumors. Cancer 2007;109:882-90. 32. Cragun JM, Havrilesky LJ, Calingaert B, et al. Retrospective analysis of selective lymphadenectomy in apparent early-stage endometrial cancer. J Clin Oncol 2005;23:3668-75. 33. ASTEC Study Group, Kitchener H, Swart AM, Qian Q, Amos C, Parmar MK. Efficacy of systematic pelvic lymphadenectomy in endometrial cancer (MRC ASTEC trial): a randomised study. Lancet 2009;373:125-36. 34. Van den Bruel A, Cleemput I, Aertgeerts B, Ramaekers D, Buntinx F. The evaluation of diagnostic tests: evidence on technical and diagnostic accuracy, impact on patient outcome and cost-effectiveness is needed. J Clin Epidemiol 2007;60:1116-22. 35. Herman CJ. Cytometric DNA analysis in the management of cancer. Clinical and laboratory considerations. Cancer 1992;69:1553-6. 36. Podratz KC, Wilson TO, Gaffey TA, Cha SS, Katzmann JA. Deoxyribonucleic acid analysis facilitates the pretreatment identification of high-risk endometrial cancer patients. Am J Obstet Gynecol 1993;168:1206-13.
DECEMBER 2009 American Journal of Obstetrics & Gynecology
603.e7
