Background
Ovarian clear cell carcinoma (OCCC) is a rare subtype of epithelial ovarian malignancy (EOM), accounting for approximately 5% to 10% of all EOMs in North America.1–3 Advanced OCCC portends poor prognosis, which is often attributed to its inherent resistance to platinum-based chemotherapy.4–6 In contrast, stage IA, IB, or IC disease solely caused by capsular rupture has been found to confer excellent prognosis, and therefore the benefit of adjuvant therapy is questionable.1,6–9 Although debulking surgery followed by platinum-based chemotherapy is the most commonly used treatment for early OCCC, controversy remains regarding optimal number of chemotherapy cycles, most effective chemotherapy regimen, and role of radiotherapy (RT) in this patient population.10–12
RT has been explored as a salvage or adjuvant treatment option in OCCC. Retrospective studies have shown durable local control rates with salvage limited-field RT after first-line chemotherapy.11,13 A population-based study demonstrated superior disease-free survival with addition of whole abdominal RT (WART) to 3 cycles of adjuvant chemotherapy in patients with stage II OCCC or a select subgroup of patients with stage IC disease (ICo) characterized by either surface or cytologic positivity.8 Nagai et al14 demonstrated a survival advantage with adjuvant WART in 16 patients with stage IC–III OCCC compared with 12 patients from a historical cohort who received adjuvant platinum-based chemotherapy. Conversely, another retrospective study, combining data from 2 institutions, did not report any benefit with adjuvant WART in early OCCC.15
Although most of these studies used WART, significant concern remains regarding its practice,8,14,16 largely due to the technical challenges associated with its delivery and the substantial risk of RT-related toxicities, including bowel toxicities and secondary malignant neoplasms.17–19 Moreover, retrospective studies have shown that most patients with early OCCC experience a recurrence in the pelvis after surgery and adjuvant chemotherapy.1,13 As a result, there has been a gradual transition toward the use of pelvic nodal RT (PRT) instead of WART for these patients. Overall, the role of adjuvant RT and the preferred RT regimen for early OCCC are poorly defined. We reviewed our population database to compare the long-term outcome of patients with early OCCC treated with adjuvant chemotherapy alone versus those treated with chemoRT combination. Additionally, we evaluated and compared the oncologic outcome in patients treated with 2 adjuvant chemoRT regimens: chemotherapy + WART (chemo-WART) and chemotherapy + PRT (chemo-PRT).
Patients and Methods
In this Institutional Research Ethics Committee–approved study, retrospective chart review was conducted to collect and verify information for all patients diagnosed with stage I and II pure OCCC (using 1988 FIGO staging criteria) in all 6 provincial cancer centers of a Canadian province between January 1, 1984, and December 31, 2015. The provincial cancer registry and ovarian cancer outcome unit were used to identify patients. These centers follow the guidelines set by the provincial authority overseeing the delivery of cancer care in a province of 4.9 million people. Treatment policies are reviewed and issued by disease-specific tumor groups and are integrated into the provincially disseminated management protocols. These guidelines and protocols are available on the provincial cancer agency website. Policy and protocols are reviewed regularly as new data become available.20 The standard, uniformly applied surgery consisted of total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy, washings, and removal of suspicious nodes. Routine pelvic and/or para-aortic lymphadenectomy was not mandated. The provincial protocol did not recommend adjuvant treatment in stage IA or IB, or IC OCCC purely resulting from capsular rupture, and therefore these patients were excluded from the final analysis; only patients with stage ICo (characterized by either surface and/or cytologic positivity) or II disease were included for final analysis (see supplemental eFigure 1, available with this article at JNCCN.org). Patients were categorized into 2 groups based on cytologic finding: those with stage IC OCCC and positive peritoneal washing and/or positive ascitic cytology and those with stage IIC constituted group 1 (cytology positive), and others were categorized as group 2 (cytology negative).
Since 2000, standard provincial treatment in chemo-WART group consisted of 3 cycles of carboplatin (area under the curve [AUC], 5–6) and paclitaxel (175 mg/m2) every 3 to 4 weeks followed by abdominopelvic irradiation. RT was initiated 3 to 4 weeks after the third cycle of chemotherapy. Pelvic RT was administered to a midplane dose of 22.5 Gy in 10 fractions over 2 weeks. This was followed by WART to a midplane dose of 22.5 Gy in 22 fractions over 4.5 weeks. Prior to 2000, the chemo-WART protocol included chemotherapy with 3 cycles of cisplatin (75 mg/m2) and cyclophosphamide (600 mg/m2) every 4 weeks followed by abdominopelvic RT (same time/dose/fraction regimen) followed by another 3 cycles of the same chemotherapy. Since 2012, RT transitioned to PRT as a provincial policy. Before 2012, pelvic RT was offered at physician’s discretion. The protocol for the chemo-PRT group called for 3 cycles of carboplatin and paclitaxel (same dosage) every 3 to 4 weeks followed by external-beam RT to the pelvis to a total dose of 45 Gy in 25 fractions over 5 weeks. Standard treatment of patients in the chemotherapy group included 6 cycles of paclitaxel (175 mg/m2) and carboplatin (AUC, 5–6) combination every 3 to 4 weeks (since 2000) or 6 cycles of cisplatin (75 mg/m2) and cyclophosphamide (600 mg/m2) every 4 weeks prior to 2000.8 Further details of RT and posttreatment follow-up policy are summarized in supplemental eAppendix 1.
Statistics
The objectives of this study were to determine the adjusted treatment effects of adjuvant treatment (chemotherapy vs chemoRT) and RT regimen (chemo-WART vs chemo-PRT) on failure-free survival (FFS) and overall survival (OS) after adjustment for covariables. FFS was defined as time from surgery to the first of either disease failure, disease progression, or death caused by cancer. For FFS, patients who were event-free at the time of analysis were censored at their date of last follow-up with known disease status. For OS, patients were censored at the date they were last known to be alive. Median follow-up was estimated using reverse Kaplan-Meier method. Cox multivariable regression models were applied to estimate adjusted treatment effects (cause-specific hazard ratio [CSHR] with 95% CI). Stage, risk group, year of diagnosis, and age at diagnosis were included as covariables in the regression models. Patients were stratified into 3 groups (1984–1994, 1995–2004, and 2005–2014) based on year of diagnosis. Efron’s method was used for handling ties.21 Proportionality assumptions were tested.22 Sensitivity analyses were performed by applying inverse probability weighting (IPW)–adjusted Kaplan-Meier methods to estimate survival for the chemo-WART and chemo-PRT groups after adjustment for age at diagnosis, risk group, year of diagnosis, and stage at diagnosis as covariables. IPW-adjusted log-rank tests and Cox multivariable regressions were applied to compare the HR for FFS and OS between the 2 chemoRT regimens.23–25 Additionally, flexible parametric models for time-to-event data were applied to calculate the treatment effect on FFS and OS adjusted for covariables and time (supplemental eAppendix 2 and eTables 2–5).26
The competing risk models of Fine and Gray27 were applied to determine the cumulative incidence of treatment failure and cancer-specific mortality (CSM) among the treatment groups. Adjusted subdistribution HRs (SHR) were estimated using competing risk regression models. Non–cancer-related deaths (ie, not due to disease failure or treatment-related toxicities) were considered competing events for CSM. Deaths from any cause were considered competing events for failure. Statistical analyses were performed using R statistical version 3.5.2 (R Foundation for Statistical Computing).22–28
Results
A total of 403 patients with early OCCC were identified who had complete information for staging and risk stratification. Of them, 343 received adjuvant treatment with chemotherapy or chemoRT and 153 were eligible for final analysis. Overall, 41% (n=63) of eligible patients received adjuvant chemotherapy, whereas 59% (n=90) received adjuvant chemoRT, of whom 48 received WART and 42 received PRT (supplemental eFigures 2 and 3). Median age of the evaluable study population was 55.5 years (interquartile range [IQR], 48.4–66.6 years). Median ages of the chemotherapy, chemo-WART, and chemo-PRT groups were 58.1 (IQR, 48.3–64.3), 57 (IQR, 50.3–64.0), and 55.1 (IQR, 49.4–61.9) years, respectively. Table 1 summarizes the distribution of stage, year of diagnosis, and risk category in the treatment groups.
Patient Characteristics
Some deviation from protocol-defined treatment was noted in the treatment groups (supplemental eTable 1). In the chemo-PRT group, 3D conformal RT (3D-CRT) with 4-field beam arrangement was used in 34 patients, and intensity-modulated RT (IMRT) and volumetric modulated arc therapy (VMAT) were used in 4 and 3 patients, respectively. One patient received RT using parallel opposed (anterior and posterior) radiation portals. There was no significant difference (P=.87) in the proportion of patients who underwent pelvic or para-aortic lymphadenectomy among the 3 treatment groups. Median follow-up for the entire cohort was 9.1 years, whereas median follow-up for the chemo-WART, chemotherapy, and chemo-PRT groups was 16.2 (IQR, 10.1–24.2), 8.0 (IQR, 5.4–16.5), and 4.8 years (IQR, 2.7–5.7 years), respectively. At last-follow-up, 8 and 18 patients were alive without evidence of disease in the chemotherapy and chemoRT groups, respectively.
FFS at 5 years was 57.2% and 45.9% for the adjuvant chemoRT and chemotherapy groups, respectively (Figure 1A). Receipt of chemoRT was associated with a significant improvement in FFS (CSHR, 0.57; 95% CI, 0.34–0.94; P=.03) compared with chemotherapy alone (Table 2).
Kaplan-Meier curves showing FFS for (A) adjuvant chemotherapy and chemoRT groups and (B) chemo-WART and chemo-PRT groups.
Abbreviations: chemo, chemotherapy; chemoRT, chemoradiotherapy; FFS, failure-free survival; PRT, pelvic nodal radiotherapy; WART, whole abdominal radiotherapy.
Citation: Journal of the National Comprehensive Cancer Network 19, 2; 10.6004/jnccn.2020.7609
Kaplan-Meier curves showing FFS for (A) adjuvant chemotherapy and chemoRT groups and (B) chemo-WART and chemo-PRT groups.
Abbreviations: chemo, chemotherapy; chemoRT, chemoradiotherapy; FFS, failure-free survival; PRT, pelvic nodal radiotherapy; WART, whole abdominal radiotherapy.
Citation: Journal of the National Comprehensive Cancer Network 19, 2; 10.6004/jnccn.2020.7609
Kaplan-Meier curves showing FFS for (A) adjuvant chemotherapy and chemoRT groups and (B) chemo-WART and chemo-PRT groups.
Abbreviations: chemo, chemotherapy; chemoRT, chemoradiotherapy; FFS, failure-free survival; PRT, pelvic nodal radiotherapy; WART, whole abdominal radiotherapy.
Citation: Journal of the National Comprehensive Cancer Network 19, 2; 10.6004/jnccn.2020.7609
Cox Multivariable Regression Models for FFS and OS
Five-year FFS was 60.6% and 53.8% for chemo-WART and chemo-PRT groups, respectively (Figure 1B). No significant difference was noted in the HR for FFS between the chemo-WART and chemo-PRT groups (CSHR, 1.34; 95% CI, 0.40–4.44, P=.63). Patterns of failure in the treatment groups are summarized in Table 3. There was a significant difference between the 2 RT groups with respect to overall nodal failure (P=.007), with higher nodal failure rates observed in the chemo-PRT group; however, no difference was seen with respect to overall peritoneal (P=.76) or local failure (P=.45), respectively. Based on IPW-adjusted Kaplan-Meier method, 5-year FFS for the chemo-WART and chemo-PRT groups was 43.9% and 34.8%, respectively. No significant difference in FFS was found between the 2 RT regimens on IPW-adjusted log-rank test (P=.21) and multivariable regression (HR, 1.64; 95% CI, 0.67–3.99).
Sites of Treatment Failure
Cumulative incidence of failure at 5 years was 53.4% versus 41.1% for the adjuvant chemotherapy and chemoRT groups, respectively (supplemental eFigure 4). Receipt of adjuvant chemoRT was associated with reduced risk of cumulative incidence of failure (SHR, 0.51; 95% CI, 0.28–0.90; P=.02). However, no significant difference was found in the relative incidence of failure between chemo-WART and chemo-PRT groups (SHR, 1.32; 95% CI, 0.34–5.13; P=.69) (Table 4).
Competing Risk Regression for Cumulative Incidence of CSM and Failure
OS at 5 and 10 years was 65.8% and 54.1% for the chemoRT group and 65% and 39.2% for chemotherapy group, respectively (Figure 2A). We found no significant difference in the HR for OS between the chemotherapy and chemoRT groups (CSHR, 0.70; 95% CI, 0.43–1.13; P=.14).
Kaplan-Meier curves showing OS for (A) adjuvant chemotherapy and chemoRT groups and (B) chemo-WART and chemo-PRT groups.
Abbreviations: chemo, chemotherapy; chemoRT, chemoradiotherapy; OS, overall survival; PRT, pelvic nodal radiotherapy; WART, whole abdominal radiotherapy.
Citation: Journal of the National Comprehensive Cancer Network 19, 2; 10.6004/jnccn.2020.7609
Kaplan-Meier curves showing OS for (A) adjuvant chemotherapy and chemoRT groups and (B) chemo-WART and chemo-PRT groups.
Abbreviations: chemo, chemotherapy; chemoRT, chemoradiotherapy; OS, overall survival; PRT, pelvic nodal radiotherapy; WART, whole abdominal radiotherapy.
Citation: Journal of the National Comprehensive Cancer Network 19, 2; 10.6004/jnccn.2020.7609
Kaplan-Meier curves showing OS for (A) adjuvant chemotherapy and chemoRT groups and (B) chemo-WART and chemo-PRT groups.
Abbreviations: chemo, chemotherapy; chemoRT, chemoradiotherapy; OS, overall survival; PRT, pelvic nodal radiotherapy; WART, whole abdominal radiotherapy.
Citation: Journal of the National Comprehensive Cancer Network 19, 2; 10.6004/jnccn.2020.7609
For chemo-WART and chemo-PRT groups, 5-year OS was 67.4% and 67.9%, respectively (Figure 2B). No significant difference was noted in the HR for OS between the 2 groups (CSHR, 1.13; 95% CI, 0.37–3.46; P=.83) (Table 2). On IPW-adjusted analysis, 5-year OS was 46.3% and 43.4% for chemo-WART and chemo-PRT groups, respectively. We did not find any significant difference between the 2 groups on IPW-adjusted log-rank test (P=.24) or multivariable regression (HR, 1.74; 95% CI, 0.75–4.04).
Overall, 55 patients were noted to have cancer-specific deaths: 52 (94.5%) due to disease failure and 3 (5.5%) due to treatment-related toxicities that included radiation-induced small bowel obstruction (n=1) and radiation-induced liver disease (n=1) in the chemo-WART group and thromboembolic complication (n=1) in the chemo-PRT group. Cause of death could not be ascertained in 3 patients. Receipt of chemoRT was associated with a 54% reduced risk of CSM (SHR, 0.46; 95% CI, 0.24–0.89; P=.02) compared with chemotherapy. However, chemo-PRT (SHR, 0.72; 95% CI, 0.18–2.80; P=.63) was not associated with any significant difference in the risk of CSM compared with chemo-WART (Table 4).
Discussion
Our findings suggest long-term reduction in the risk of failure with receipt of adjuvant chemoRT. Compared with the previous study from our province, in which the study population included all patients with stage I and II OCCC with year of diagnosis restricted to 2008, we included all patients with early OCCC diagnosed in 1984 through 2015 and treated with adjuvant chemotherapy or chemoRT. The comparison of treatment was, however, restricted to patients with stage II disease and those with stage IC disease and presence of malignant cells in the ascites or peritoneal washings or tumor deposits on the surface of ovary (alone or in combination). Findings from the sensitivity analyses were consistent with the primary analyses. Improved FFS with adjuvant chemoRT could be attributed to superior local control and possible sterilization of microscopic disease in the pelvic and para-aortic lymph nodes with the addition of RT to adjuvant chemotherapy. Reduction in the risk of failure might have translated into improved CSM with receipt of chemoRT. This observation concurs with a recent retrospective study that revealed inferior survival in patients with OCCC who had disease progression after adjuvant treatment.29 We found a relatively higher number of para-aortic nodal failures in the chemo-PRT group compared with the chemo-WART group despite statistically similar rates of lymphadenectomy. Whether this could be reduced with the addition of para-aortic nodal RT to standard PRT target without increasing treatment-related morbidity should be investigated in additional studies. We did not find any statistically significant difference in OS, FFS, or CSM between the 2 RT regimens. A shorter follow-up and consequent lack of events in the chemo-PRT group in addition to the overall small patient population might have resulted in a corresponding loss of power in our study. This could be reflected in the relatively wide confidence intervals. The lack of OS difference among the groups could also be explained by the limited power.
Although the primary aim of this study was not to compare toxicities among the treatment groups, note was made of 2 deaths from treatment-related toxicities in the chemo-WART group and 1 in the chemotherapy group. This must be interpreted considering that there was no (WART group) or limited use of IMRT (PRT group) in our study. A randomized study compared 3D-CRT with IMRT in patients with locally advanced cervical cancer treated with PRT with concomitant chemotherapy. IMRT yielded reduced risk of acute and late gastrointestinal toxicities.30 Another nonrandomized phase II study evaluated the safety of IMRT-based consolidative WART in stage III ovarian cancer after surgery and platinum-based chemotherapy. Despite using a relatively higher dose (30 Gy in 20 fractions), no acute or late grade ≥3 renal, hepatic, or gastrointestinal toxicities were found.30 Overall, these findings suggest that use of IMRT or VMAT could potentially allow safe delivery of a higher dose of RT. Whether such dose escalation has the potential to enhance disease control without exacerbating treatment-induced morbidities in early OCCC needs further investigation.
We acknowledge several other limitations of this retrospective study. The findings are subject to bias considering evolving standards of staging, diagnostic evaluation, and treatment techniques in the domain of surgery, chemotherapy, and RT over the span of the study.31 All treatment modalities had remarkable advancements during the study period that might have implications on our observations.32,33 Similarly, there has been significant improvement in the accuracy and reliability of diagnostic imaging for ovarian malignancies. Our study period started in an era when there was a lack of contrast-enhanced CT scans, which are now the preferred imaging modality when there is a suspicion of recurrence. Furthermore, RT was not randomly allocated in this study. Therefore, one cannot rule out the influence of unforeseen confounding variables. Despite existence of standard uniform provincial guidelines, unexplained heterogeneity among the centers and treating physicians cannot be ruled out. This is pertinent for our study because a significant proportion of patients were treated before guidelines on peer review of clinical decision, target volume delineation, and plan evaluation were endorsed and implemented in the field of RT.34 Sharp dissection, a previously found predictor of failure and death, was not incorporated in the regression models. Further caution is needed in interpreting the results due to exclusion of patients who received no adjuvant treatment or those with missing information on histopathology or cytology. We performed electronic and paper chart review for all patients to update follow-up; ensure meticulous reporting of disease stage, death, and its cause; and record sites of treatment failure. Nonetheless, studies with such prolonged time span are prone to attrition bias due to loss of follow-up and migration of individual subjects in and out of the population database over time.
Previous retrospective studies have shown positive cytology to be an adverse prognostic factor.11,35,36 Our findings were congruent with these studies. The failure pattern in our study is also consistent with other series.6,7,13,14 However, our findings contrast with those of a recent study by Hogen et al,15 which did not show any advantage of adjuvant RT in patients with early OCCC. The discordance in findings could be ascribed to several factors, including differences in patient population, surgical protocols, or adjuvant treatment paradigm. To explain further, cytology was optional as part of surgical procedure in this study, which pooled patient data from 2 institutions. The comparison between adjuvant RT and no RT groups was confounded by higher use of comprehensive surgical staging in the no-RT group. The groups were heterogeneous with respect to the proportion of patients who received chemotherapy; 70% of patients in the no-RT group received adjuvant chemotherapy, whereas 84% in the RT group were treated with a chemoRT combination.
Conclusions
This population-based study showed that combination of adjuvant chemotherapy and RT was associated with encouraging long-term improvement in risk of failure and CSM, although this combination was not associated with any significant OS benefit. Our study did not find any statistically significant difference in the hazard for OS, FFS, and CSM between adjuvant PRT and WART in conjunction with chemotherapy. A randomized study should be considered to validate the findings of the current report.
References
- 1.↑
Chan JK, Teoh D, Hu JM, et al.. Do clear cell ovarian carcinomas have poorer prognosis compared to other epithelial cell types? A study of 1411 clear cell ovarian cancers. Gynecol Oncol 2008;109:370–376.
- 2.↑
Tan DSP, Kaye S. Ovarian clear cell adenocarcinoma: a continuing enigma. J Clin Pathol 2007;60:355–360.
- 3.↑
Anglesio MS, Carey MS, Köbel M, et al.. Clear cell carcinoma of the ovary: a report from the first Ovarian Clear Cell Symposium, June 24th, 2010. Gynecol Oncol 2011;121:407–415.
- 4.↑
Takano M, Sugiyama T, Yaegashi N, et al.. Low response rate of second-line chemotherapy for recurrent or refractory clear cell carcinoma of the ovary: a retrospective Japan Clear Cell Carcinoma Study. Int J Gynecol Cancer 2008;18:937–942.
- 5.↑
Pather S, Quinn MA. Clear-cell cancer of the ovary-is it chemosensitive? Int J Gynecol Cancer 2005;15:432–437.
- 6.↑
Sugiyama T, Kamura T, Kigawa J, et al.. Clinical characteristics of clear cell carcinoma of the ovary: a distinct histologic type with poor prognosis and resistance to platinum-based chemotherapy. Cancer 2000;88:2584–2589.
- 7.↑
Takano M, Kikuchi Y, Yaegashi N, et al.. Clear cell carcinoma of the ovary: a retrospective multicentre experience of 254 patients with complete surgical staging. Br J Cancer 2006;94:1369–1374.
- 8.↑
Hoskins PJ, Le N, Gilks B, et al.. Low-stage ovarian clear cell carcinoma: population-based outcomes in British Columbia, Canada, with evidence for a survival benefit as a result of irradiation. J Clin Oncol 2012;30:1656–1662.
- 9.↑
Chan JK, Tian C, Fleming GF, et al.. The potential benefit of 6 vs. 3 cycles of chemotherapy in subsets of women with early-stage high-risk epithelial ovarian cancer: an exploratory analysis of a Gynecologic Oncology Group study. Gynecol Oncol 2010;116:301–306.
- 10.↑
del Carmen MG, Birrer M, Schorge JO. Clear cell carcinoma of the ovary: a review of the literature. Gynecol Oncol 2012;126:481–490.
- 11.↑
Shu CA, Zhou Q, Jotwani AR, et al.. Ovarian clear cell carcinoma, outcomes by stage: the MSK experience. Gynecol Oncol 2015;139:236–241.
- 12.↑
Armstrong DK, Alvarez RD, Bakkum-Gamez JN, et al.. NCCN Guidelines Insights: Ovarian Cancer, Version 1.2019. J Natl Compr Canc Netw 2019;17:896–909.
- 13.↑
Macrie BD, Strauss JB, Helenowski IB, et al.. Patterns of recurrence and role of pelvic radiotherapy in ovarian clear cell adenocarcinoma. Int J Gynecol Cancer 2014;24:1597–1602.
- 14.↑
Nagai Y, Inamine M, Hirakawa M, et al.. Postoperative whole abdominal radiotherapy in clear cell adenocarcinoma of the ovary. Gynecol Oncol 2007;107:469–473.
- 15.↑
Hogen L, Thomas G, Bernardini M, et al.. The effect of adjuvant radiation on survival in early stage clear cell ovarian carcinoma. Gynecol Oncol 2016;143:258–263.
- 16.↑
Swenerton KD, Santos JL, Gilks CB, et al.. Histotype predicts the curative potential of radiotherapy: the example of ovarian cancers. Ann Oncol 2011;22:341–347.
- 17.↑
Engelen MJA, Snel BJ, Schaapveld M, et al.. Long-term morbidity of adjuvant whole abdominal radiotherapy (WART) or chemotherapy for early stage ovarian cancer. Eur J Cancer 2009;45:1193–1200.
- 18.↑
Whelan TJ, Dembo AJ, Bush RS, et al.. Complications of whole abdominal and pelvic radiotherapy following chemotherapy for advanced ovarian cancer. Int J Radiat Oncol Biol Phys 1992;22:853–858.
- 19.↑
Einhorn N, Tropé C, Ridderheim M, et al.. A systematic overview of radiation therapy effects in ovarian cancer. Acta Oncol 2003;42:562–566.
- 20.↑
Cancer Management Guidelines. Accessed May 11, 2020. Available at: http://www.bccancer.bc.ca/health-professionals/clinical-resources/cancer-management-guidelines
- 21.↑
Efron B. The efficiency of Cox’s likelihood function for censored data. J Am Stat Assoc 1977;72:557–565.
- 22.↑
Therneau TM, Grambsch PM. Modeling survival data: extending the Cox model. Accessed June 3, 2019. Available at: https://cran.r-project.org/web/packages/survival/citation.html
- 23.↑
Le Borgne F, Giraudeau B, Querard AH, et al.. Comparisons of the performance of different statistical tests for time-to-event analysis with confounding factors: practical illustrations in kidney transplantation. Stat Med 2016;35:1103–1116.
- 24.↑
Shu D, Yoshida K, Fireman BH, et al.. Inverse probability weighted Cox model in multi-site studies without sharing individual-level data. Stat Methods Med Res 2020;29:1668–1681.
- 25.↑
Cole SR, Hernán MA. Adjusted survival curves with inverse probability weights. Comput Methods Programs Biomed 2004;75:45–49.
- 26.↑
Royston P, Parmar MKB. Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application to prognostic modelling and estimation of treatment effects. Stat Med 2002;21:2175–2197.
- 27.↑
Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999;94:496–509.
- 28.↑
Gerds TA, Scheike TH, Andersen PK. Absolute risk regression for competing risks: interpretation, link functions, and prediction. Stat Med 2012;31:3921–3930.
- 29.↑
Hogen L, Vicus D, Ferguson SE, et al.. Patterns of recurrence and impact on survival in patients with clear cell ovarian carcinoma. Int J Gynecol Cancer 2019;29:1164–1169.
- 30.↑
Gandhi AK, Sharma DN, Rath GK, et al.. Early clinical outcomes and toxicity of intensity modulated versus conventional pelvic radiation therapy for locally advanced cervix carcinoma: a prospective randomized study. Int J Radiat Oncol Biol Phys 2013;87:542–548.
- 31.↑
Liu SL, Tinker A V. Omission of adjuvant therapy in stage I clear cell ovarian cancer: review of the BC Cancer experience. Gynecol Oncol Reports 2020;31:100533.
- 32.↑
Falzone L, Salomone S, Libra M. Evolution of cancer pharmacological treatments at the turn of the third millennium. Front Pharmacol 2018;9:1300.
- 33.↑
Arruebo M, Vilaboa N, Sáez-Gutierrez B, et al.. Assessment of the evolution of cancer treatment therapies. Cancers (Basel) 2011;3:3279–3330.
- 34.↑
Vijayakumar S, Duggar WN, Packianathan S, et al.. Chasing zero harm in radiation oncology: using pre-treatment peer review. Front Oncol 2019;9:302.
- 35.↑
Takano M, Sugiyama T, Yaegashi N, et al.. Less impact of adjuvant chemotherapy for stage I clear cell carcinoma of the ovary: a retrospective Japan Clear Cell Carcinoma Study. Int J Gynecol Cancer 2010;20:1506–1510. 10.1111/IGC.0b013e3181fcd089
- 36.↑
Ku FC, Wu RC, Yang LY, et al.. Clear cell carcinomas of the ovary have poorer outcomes compared with serous carcinomas: results from a single-center Taiwanese study. J Formos Med Assoc 2018;117:117–125.
