Endometrial cancer is the most common gynecologic malignancy in the United States.1 The majority of endometrial cancers possess type I histology highlighted by low-grade tumors and early-stage disease.2,3 Therefore, type I endometrial cancer is often curable with hysterectomy-based surgical treatment, generally with a good prognosis.2,4 Most cases of endometrial cancer are diagnosed by endometrial biopsy before surgery that enables both patients and care providers to have time to prepare for surgery, which is called the “wait time” for surgical staging. Wait time for surgical staging may reflect the spectrum of geographic and socioeconomic barriers, delays for medical clearance or optimization by specialists, or patient delays.5
The effect of wait time for surgical staging on survival outcome of endometrial cancer is controversial. A recent population-based study concluded that longer wait time for surgical staging was associated with worse survival outcomes in uterine cancer.6 Others have found that type I endometrial cancers commonly progress very slowly in nature, and wait time for surgical staging in patients with endometrial cancer is not associated with survival outcomes.7–9 Therefore, whether wait time for surgical staging affects tumor progression is unclear.
The aims of our study were to 1) examine the correlation of wait time for surgical staging and survival outcomes of patients with type I endometrial cancer and 2) evaluate the patterns and significance of tumor grade change between the endometrial biopsy and hysterectomy.
MATERIALS AND METHODS
After institutional review board approval was obtained at the University of Southern California, an institutional database for endometrial cancer was searched.10 This electronic database collects consecutive gynecologic malignancies that are diagnosed and treated at Los Angeles County+University of Southern California Medical Center. This study followed the STROBE (STrengthening the Reporting of OBservational studies in Epidemiology) guideline for a retrospective cohort study. Eligibility criteria included women with grade 1 and 2 endometrioid adenocarcinoma of the endometrium diagnosed by endometrial biopsy before surgical staging: those patients subsequently underwent hysterectomy-based surgical staging between January 2000 and December 2013. Exclusion criteria were: neoadjuvant chemotherapy or hormonal treatment before hysterectomy, no preoperative biopsy demonstrating endometrial cancer, surgery, biopsy, or both at an outside hospital, lost to follow-up after endometrial biopsy, and mixed or nonendometrioid histology types on the endometrial biopsy.
Among eligible cases, the following information was abstracted from the medical record: 1) clinical demographics at the time of endometrial biopsy for endometrial cancer diagnosis including age, ethnicity, body mass index (BMI, calculated as weight (kg)/[height (m)]2), and medical comorbidities including hypertension, diabetes mellitus, and hypercholesterolemia; 2) tumor characteristics including tumor marker CA 125 level, grade, and histology from the endometrial biopsy as well as hysterectomy specimens and the International Federation of Gynecology and Obstetrics (FIGO) stage; and 3) survival data including disease-free survival and overall survival.
Type I endometrial cancer was defined as grade 1 and 2 endometrioid adenocarcinoma.11,12 Wait time for surgical staging was defined as the time interval between the date of endometrial biopsy and the date of hysterectomy. Tumor grade was based on the FIGO system for endometrial cancer: 5% or less solid component for grade 1, 6–50% solid component for grade 2, and more than 50% solid component for grade 3.13 Grade change was defined as a change of FIGO grade between the diagnostic endometrial biopsy and hysterectomy. The grade changes were classified as: grade 1 to 1, grade 1 to 2, grade 1 to 3, grade 2 to 1, grade 2 to 2, and grade 2 to 3. All the histopathology slides from the endometrial biopsy and hysterectomy specimens were reviewed by pathologists with special expertise in gynecologic pathology at the time of cancer diagnosis and treatment. Disease-free survival was defined as the time interval between endometrial biopsy and the date of the first recurrence or last follow-up date. Overall survival was defined as the time interval between endometrial biopsy and the date of death resulting from endometrial cancer of last follow-up date. Data entry (deidentified) was performed by one of the coinvestigators (K.M., M.A.C., J.G., and K.E.T.), and the principal investigator examined all the medical records of collected data for accuracy, consistency, and quality (K.M.).
To maximize the power of this study to examine the effects of wait time on survival outcome, three classifications of wait time for surgical staging based on its clinical relevance were tested in the analysis: 1) monthly grouping (1–28 days, 29–56 days, 57–84 days, and 85 days or more), 2) validation pattern of day intervals based on prior publication (1–14 days, 15–42 days, 43–84 days, and 85 days or more),6 and 3) quartile pattern based on median value (1–25%ile, 26–50%ile, 51–75%ile, and 76–100%ile). Among the three methods of wait time classification, grouping based on day intervals as reported in a prior publication6 was the most predictive for recurrence (area under the curve [AUC] 0.674, 95% confidence interval [CI] 0.571–0.776) when compared with other methods (monthly grouping, AUC 0.663, 95% CI 0.556–0.770; and quartile pattern, AUC 0.662, 95% CI 0.566–0.759). Similar results were seen in predicting death event in that grouping based on day intervals per prior publication6 was the most sensitive method (AUC 0.691, 95% CI 0.570–0.811) followed by quartile pattern (AUC 0.674, 95% CI 0.567–0.791) and monthly grouping (AUC 0.671, 95% CI 0.541–0.801). Therefore, wait time was classified based on the prior publication (1–14 days, 15–42 days, 43–84 days, and 85 days or more) for our further analysis in the study.
Because the previous large-scale population-based study on this topic6 was performed in a population with universal access to health care and substantially different demographical and clinical characteristics, sample size was not calculated based on these results. Instead, sample size in this study was calculated based on the ability to detect at least a 0.20 correlation between wait time for surgical staging and both primary outcomes (disease-free survival and overall survival). With an α of 0.05 and 80% power, at least 153 patients were estimated to be required to detect this difference.
The primary purpose of this study was to examine the effects of wait time on survival outcomes of endometrial cancer. A secondary purpose was to examine the effects of grade change between endometrial biopsy and hysterectomy on wait time and survival outcomes. Continuous variables were assessed for normality by Kolmogorov-Smirnov test expressed as mean (±standard deviation) or median (range) as appropriate. Statistical significances of continuous variables were examined by Student's t test or Mann-Whitney U test as appropriate. Pearson's correlation coefficient was used to examine nonnormally distributed continuous variables. Categorical or ordinal variables were examined by Fisher exact test or χ2 test as appropriate expressed its magnitude of statistical significance with odds ratio (OR) and 95% CI. Receiver operator characteristic curve analysis was performed to determine the most sensitive method for wait time classification to predict survival event (recurrence and death) as comparing the values for AUC. Survival analysis was examined by Cox's proportional hazard regression model for multivariable analysis expressed its magnitude of statistical significance with hazard ratio (HR) and 95% CI. In the multivariable model, a priori each covariate was selected for inclusion in any final model based on clinical relevance and effect in endometrial cancer: age (younger than 60 compared with 60 years or older), ethnicity (Hispanic compared with non-Hispanic), BMI (less than 30 compared with 30 or higher), hypertension (yes compared with no), diabetes mellitus (yes compared with no), and hypercholesterolemia (yes compared with no), CA 125 level (less than 35 compared with 35 international units/L or more), stage (I–II compared with III–IV), grade change between endometrial biopsy and hysterectomy (grade 1 to 1, 1 to 2, 1 to 3, 2 to 1, 2 to 2, and 2–3), and wait time for surgical staging (1–14, 15–42, 43–84, and 85 days or more). Kaplan-Meier method was used to plot survival curves. All tests were two-tailed, and P values of <.05 were considered statistically significant in this study. SPSS 12.0 was used for statistical analysis.
There were 738 women with endometrial cancer who underwent hysterectomy during the study period (Fig. 1). Of those, 206 (27.9%) patients were excluded as a result of: neoadjuvant chemotherapy (n=54), surgery or biopsy at outside hospital (n=51), progestin-based hormonal therapy before surgery (n=21), noncancer diagnosis on preoperative endometrial biopsy (complex atypical hyperplasia, n=58, and atrophic endometrium, n=5), and no preoperative endometrial biopsy (n=17). The remaining 532 patients had no treatment before surgical staging and had a histologic diagnosis of endometrial cancer by endometria biopsy before hysterectomy. Of these, 97 (18.2%) patients were excluded as a result of grade 3 endometrioid, serous, or clear cell type histology (n=91); grade 1 mixed histology (n=4); and grade 1 mucinous histology (n=2). The remaining 435 patients with grade 1 and 2 endometrioid-type endometrial cancer diagnosed on preoperative endometrial biopsy who underwent subsequent hysterectomy-based surgical staging without any neoadjuvant chemotherapy or hormonal treatment comprised the study group.
Patient demographics are shown in Table 1. The mean age was 52.2 years, and the majority of patients were Hispanic (70.6%), obese (BMI 30 or higher, 70.6%), had early-stage disease (stage I–II, 87.1%), and had grade 1 tumors on endometrial biopsy (73.8%). The median wait time for surgical staging was 57 days (range 1–177 days) in this cohort. The majority of wait times for surgical staging was between 43 and 84 days (57.7%). Approximately one sixth of women had a wait time for surgical staging more than 85 days or more (16.8%). Grade change between endometrial biopsy and hysterectomy was seen in 128 (29.4%) of 435 patients (Fig. 1). Recurrence and death resulting from endometrial cancer were seen in 25 (5.7%) and 17 (3.9%) patients, respectively. Median follow-up time of the cohort was 28.8 months. In bivariate analysis using Pearson's correlation, a significant inverse association was found between wait time for surgical staging and disease-free survival of −0.120 (P=.012) and between wait time and overall survival of −0.148 (P=.002). Additional analyses showed that a quadric curve was the best overall fit for these data (see Appendix 1, available online at http://links.lww.com/AOG/A600).
Contributing factors associated with wait time for surgical staging in patients with endometrial cancer were examined (Table 2). Among the variables examined in this study, Hispanic race (median wait time, Hispanic compared with non-Hispanic, 60 compared with 51 days, P=.001, Mann-Whitney U test) and obesity (BMI 30 or higher compared with less than 30, 59 compared with 53 days, P=.013) were significantly associated with longer wait time for surgical staging. On the contrary, elevated CA 125 level (35 international units/L or more compared with less than 35 international units/L, 48 compared with 59.5 days, P<.001), advanced-stage disease (stage III–IV compared with I–II, 51.5 compared with 58 days, P=.026), and grade 2 tumors on endometrial biopsy (grade 2 compared with 1, 50.5 compared with 59 days, P=.012) were significantly associated with shorter wait time for surgical staging. None of the medical comorbidities were associated with wait time for surgical staging (all, P>.05). There was no statistical correlation between wait time for surgical staging and grade change between endometrial biopsy and hysterectomy (all, P>.05; Table 2).
Survival analysis was performed (Table 3). The 5-year disease-free survival rates based on the duration of wait time for surgical staging were 56.5% for 1–14 days, 86.5% for 15–42 days, 91.2% for 43–84 days, and 100% for 85 days or more (Fig. 2A). The 5-year overall survival rates based on the duration of wait time for surgical staging were 62.5% for 1–14 days, 93.6% for 15–42 days, 95.2% for 43–84 days, and 100% for 85 days or more (Fig. 2B). When wait time for surgical staging was examined by Cox's proportional hazard regression test controlling for other covariates (age, ethnicity, BMI, medical comorbidities, CA 125 level, cancer stage, and interval grade change), wait time for surgical staging was not associated with disease-free survival (15–42 compared with 1–14 days, HR 1.19, 95% CI 0.20–6.91, P=.85; 43–84 compared with 1–14 days, HR 0.70, 95% CI 0.13–3.82, P=.68; 85 or more compared with 1–14 days, HR not available, P=.97) and overall survival (15–42 compared with 1–14 days, HR 1.16, 95% CI 0.14–9.40, P=.89; 43–84 compared with 1–14 days, HR 0.64 95% CI 0.08–5.16, P=.68; and 85 or more compared with 1–14 days, HR not available, P=.98; Table 3). Advanced-stage disease at the time of hysterectomy was associated with decreased disease-free survival and overall survival on multivariable analysis (both, P<.001).
Significance of grade change between endometrial biopsy and hysterectomy was examined for effect on survival outcome. There were 36 (8.3%) women whose tumors were upgraded to grade 3 in the hysterectomy specimen including 15 (4.7%) women who had grade 1 tumors and 21 (18.4%) women who had grade 2 tumors on the endometrial biopsy (Fig. 1). Of these 36 women, 22 (61.1%) had an upgrade to pure grade 3 endometrioid cancer, 13 (36.1%) had an upgrade to a mixed endometrioid and serous or clear cell type, and one (2.8%) was upgraded to pure serous cancer. Grade in the endometrial biopsy did not correlate to these grade 3 tumor distributions in the hysterectomy specimens (P=.64, χ2 test). Among grade 1 tumors in the endometrial biopsy, patients who had grade 3 histology in the hysterectomy had significantly poorer disease-free survival (5-year rates, 76.9% compared with 94.6%, HR 11.1, 95% CI 2.18–56.5, P<.001) (see Appendix 2, part A, available online at http://links.lww.com/AOG/A600) and overall survival (5-year rates, 82.1% compared with 98.5%, HR 31.5, 95% CI 2.27–436, P=.01) (see Appendix 2, part B, available online at http://links.lww.com/AOG/A600) than those who had grade 1 disease in the hysterectomy group on multivariable analysis controlling for age, ethnicity, BMI, comorbidity, CA 125 level, stage, and wait time for surgical staging (Table 3). In contrast, women with grade 2 tumors in the endometrial biopsy who had grade 3 histology in the hysterectomy specimen did not have worse survival outcomes when compared with those who had grade 1 tumors in the hysterectomy specimen (disease-free survival, P=.57 [see Appendix 2, part C, available online at http://links.lww.com/AOG/A600]; and overall survival, P=.12 [Appendix 2, part D, available online at http://links.lww.com/AOG/A600]).
Because change from grade 1 in the endometrial biopsy to grade 3 in the hysterectomy was associated with a worse survival outcome, risk factors for this grade change were examined (Table 4). Among the tested variables, nonobesity (frequency of grade change from 1 to 3, BMI less than 30 compared with 30 or higher, 8.8% compared with 3.0%, OR 3.07, 95% CI 1.08–8.73, P=.039, Fisher exact test) and advanced-stage disease (stage III–IV compared with I–II, 16.7% compared with 3.7%, OR 5.20, 95% CI 1.52–17.8, P=.019) were significantly associated with the grade change from 1 to 3. When 321 patients with grade 1 tumors on endometrial biopsy were examined based on the combination patterns of nonobesity and advanced-stage disease, the number of risk factors was significantly associated with grade change from 1 to 3: no risk factor, 2.3%; nonobesity alone 7.2% (OR 3.26, 95% CI 0.97–11.0, P=.057); advanced-stage disease alone, 12.5% (OR 5.97, 95% CI 1.06–33.6, P=.043); and both risk factors, 25.0% (OR 13.9, 95% CI 2.24–86.8, P=.005).
This study found that wait time for hysterectomy-based surgery was not negatively associated with survival outcome of patients with endometrial cancer when endometrial biopsy showed type I histology. Moreover, although rare, grade change from a low-grade tumor in the endometrial biopsy to a high-grade tumor in the hysterectomy specimen was associated with a worse survival outcome.
Wait time for surgery is a crucial consideration in the context of treatment of patients with cancer and seems to vary in its effect depending on the specific gynecologic malignancy evaluated.14,15 Although several previous studies have examined the effect of wait times for surgical staging in endometrial cancer, the results have been inconsistent. The majority of reports fail to demonstrate a significant effect of treatment delay on survival, although the sample sizes have been small (n=7–182) with mixed histology.7–9 A recent population-based study examining 9,417 patients with uterine cancer demonstrated an inverse relationship between wait time for surgery and survival outcome (5-year overall survival rates, 1–14, 15–42, 43–84, and 85 days or more, 71.1%, 81.8%, 79.5%, and 71.9%, respectively).6 However, a criticism of this study is that the study population included 2,127 (22.6%) patients with high-grade tumors such as sarcomas and nonendometrioid histologies representing very heterogeneous tumor biologies as compared with grade 1 and 2 endometrioid type.16 In addition, this population-based study did not exclude the subset of patients who received neoadjuvant chemotherapy or hormonal therapy. In response to this criticism, the same investigators reported a secondary analysis of 3,381 patients with only endometrioid histology who received no chemotherapy or radiotherapy, which still detected a difference in survival among patients waiting 85 days or more (5-year overall survival rates, 90.2%, 92.5%, 92.4%, and 85.0%) (Elit LM, Pond G, Seow HY. Treatment delay in endometrial cancer [letter-reply]. J Clin Oncol 2014;32:2114). Their analysis, however, still included patients with grade 3 endometrioid histology and therefore cannot be directly compared with these data. In contrast, the current study restricted the analysis to include only type I histology, which ultimately provides more clinically relevant information to clinicians.
In this study, 8.3% of women had a final diagnosis of a grade 3 tumor in the hysterectomy specimen (grade 1 to 3, 3.4%; and grade 2 to 3, 4.8%). These statistics are similar to what was reported in the previous study that reported the frequency of grade change from 1 to 3 in 2% although the study did not provide survival significance of upgrade like this study.17 The likely cause of this grade change between the endometrial biopsy and the hysterectomy is a sampling bias at the time of endometrial biopsy in that an underlying high-grade tumor was missed at the time of biopsy. Tumor dedifferentiation during the wait time is much less likely, because this is a rare event. Indeed, these results showed that there was no correlation between wait time for surgical staging and grade change, which supports the theory (Table 2). In addition, recent genomic analysis of endometrial cancer reveals that low-grade tumors have characteristics distinct from high-grade tumors.18 In this setting, prediction of women with low-grade tumors on endometrial biopsy who might harbor high-grade tumors will help triage the patients who would benefit from expedited surgical staging. That is, if a patient with type I endometrial cancer is nonobese and there is a suspicion for metastatic disease, additional endometrial tissue sampling or molecular testing for biomarkers may be a potential option to identify an underlying high-grade tumor.
A major strength of this study is the focus on both wait time for surgical staging and histology grade in both the biopsy and the hysterectomy specimens. Furthermore, all eligible patients were included in this analysis, thereby reducing the likelihood of selection bias. However, because of the retrospective nature of this study, potential confounding factors may have been missed. For instance, results showed that wait time for surgical staging was shorter in advanced-stage and longer in early-stage disease. Although such patients with advanced-stage disease may have had a high CA 125 level that prompted earlier surgery, the exact indication or reason for delay could not clearly be obtained from the medical records. In addition, no other data source was used to verify death events. Finally, although there was sufficient power to detect an overall relationship between wait time and outcomes, the sample size may not have been sufficient for these results to persist in subanalyses or after correcting for confounding factors. Nevertheless, it is unlikely that a larger sample size would have reversed these findings leading to a different clinical conclusion.
In summary, there was no evidence that longer wait times lead to a poorer prognosis among women with type I endometrial cancer. Follow-up studies on many of the findings here are clearly warranted.
1. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin 2014;64:9–29.
2. Santin AD, Bellone S, O'Brien TJ, Pecorelli S, Cannon MJ, Roman JJ. Current treatment options for endometrial cancer. Expert Rev Anticancer Ther 2004;4:679–89.
3. Bokhman JV. Two pathogenetic types of endometrial carcinoma. Gynecol Oncol 1983;15:10–7.
4. Wright JD, Barrena Medel NI, Sehouli J, Fujiwara K, Herzog TJ. Contemporary management of endometrial cancer. Lancet 2012;379:1352–60.
5. Ward KK, Shah NR, Saenz CC, McHale MT, Alvarez EA, Plaxe SC. Cardiovascular disease is the leading cause of death among endometrial cancer patients. Gynecol Oncol 2012;126:176–9.
6. Elit LM, O'Leary EM, Pond GR, Seow HY. Impact of wait times on survival for women with uterine cancer. J Clin Oncol 2014;32:27–33.
7. Pirog EC, Heller DS, Westhoff C. Endometrial adenocarcinoma—lack of correlation between treatment delay and tumor stage. Gynecol Oncol 1997;67:303–8.
8. Menczer J, Krissi H, Chetrit A, Gaynor J, Lerner L, Ben-Baruch G, et al.. The effect of diagnosis and treatment delay on prognostic factors and survival in endometrial carcinoma. Am J Obstet Gynecol 1995;173:774–8.
9. Levy T, Golan A, Menczer J. Endometrial endometrioid carcinoma: a glimpse at the natural course. Am J Obstet Gynecol 2006;195:454–7.
10. Matsuo K, Yessaian AA, Lin YG, Pham HQ, Muderspach LI, Liebman HA, et al.. Predictive model of venous thromboembolism in endometrial cancer. Gynecol Oncol 2013;128:544–51.
11. Doll A, Abal M, Rigau M, Monge M, Gonzalez M, Demajo S, et al.. Novel molecular profiles of endometrial cancer—new light through old windows. J Steroid Biochem Mol Biol 2008;108:221–9.
12. Voss MA, Ganesan R, Ludeman L, McCarthy K, Gornall R, Schaller G, et al.. Should grade 3 endometrioid endometrial carcinoma be considered a type 2 cancer—a clinical and pathological evaluation. Gynecol Oncol 2012;124:15–20.
13. Rose PG. Endometrial carcinoma. N Engl J Med 1996;335:640–9.
14. Nagle CM, Francis JE, Nelson AE, Zorbas H, Luxford K, de Fazio A, et al.. Reducing time to diagnosis does not improve outcomes for women with symptomatic ovarian cancer: a report from the Australian Ovarian Cancer Study Group. J Clin Oncol 2011;29:2253–8.
15. Matsuo K, Whitman SA, Blake EA, Conturie CL, Ciccone MA, Jung CE, et al.. Feto-maternal outcome of pregnancy complicated by vulvar cancer: a systematic review of literature. Eur J Obstet Gynecol Reprod Biol 2014;179:216–23.
16. Menczer J. Treatment delay in endometrial cancer. J Clin Oncol 2014;32:2113–4.
17. Helpman L, Kupets R, Covens A, Saad RS, Khalifa MA, Ismiil N, et al.. Assessment of endometrial sampling as a predictor of final surgical pathology in endometrial cancer. Br J Cancer 2014;110:609–15.
18. Cancer Genome Atlas Research Network, Kandoth C, Schultz N, Cherniack AD, Akbani R, Liu Y, et al.. Integrated genomic characterization of endometrial carcinoma. Nature 2013;497:67–73.