This article was downloaded by: [OARE Consortium] On: 30 June 2015, At: 18:06 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Acta Oncologica Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/ionc20
Variations in adjuvant chemotherapy and survival in women with epithelial ovarian cancer – a populationbased study ab
b
b
c
Satyamurthy Anuradha , Peter J. Donovan , Penelope M. Webb , Alison H. Brand , Jeffrey d
e
f
g
h
Goh , Michael Friedlander , Martin K. Oehler , Michael Quinn , Christopher Steer & Susan J. Jordan a
b
The University of Queensland, School of Public Health, Brisbane, Queensland, Australia
b
The QIMR Berghofer Medical Research Institute, Gynaecological Cancers Group, Brisbane, Queensland, Australia c
Department of Gynaecological Oncology, Westmead Hospital, University of Sydney, Westmead, New South Wales, Australia d
Royal Brisbane and Women's Hospital, Brisbane, Queensland, Australia
e
Department of Medical Oncology, Prince of Wales Hospital, Sydney, New South Wales, Australia f
Department of Gynaecological Oncology, Royal Adelaide Hospital, Adelaide, South Australia, Australia g
Department of Obstetrics and Gynaecology University of Melbourne, Melbourne, Victoria, Australia h
Border Medical Oncology, Wodonga, Victoria, Australia Published online: 17 Jun 2015.
To cite this article: Satyamurthy Anuradha, Peter J. Donovan, Penelope M. Webb, Alison H. Brand, Jeffrey Goh, Michael Friedlander, Martin K. Oehler, Michael Quinn, Christopher Steer & Susan J. Jordan (2015): Variations in adjuvant chemotherapy and survival in women with epithelial ovarian cancer – a population-based study, Acta Oncologica To link to this article: http://dx.doi.org/10.3109/0284186X.2015.1054950
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
Downloaded by [OARE Consortium] at 18:06 30 June 2015
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Acta Oncologica, 2015; Early Online: 1–8
ORIGINAL ARTICLE
Variations in adjuvant chemotherapy and survival in women with epithelial ovarian cancer – a population-based study
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Satyamurthy Anuradha1,2, Peter J. Donovan2, Penelope M. Webb2, Alison H. Brand3, Jeffrey Goh4, Michael Friedlander5, Martin K. Oehler6, Michael Quinn7, Christopher Steer8 & Susan J. Jordan2 1The
University of Queensland, School of Public Health, Brisbane, Queensland, Australia, 2The QIMR Berghofer Medical Research Institute, Gynaecological Cancers Group, Brisbane, Queensland, Australia, 3Department of Gynaecological Oncology, Westmead Hospital, University of Sydney, Westmead, New South Wales, Australia, 4Royal Brisbane and Women’s Hospital, Brisbane, Queensland, Australia, 5Department of Medical Oncology, Prince of Wales Hospital, Sydney, New South Wales, Australia, 6Department of Gynaecological Oncology, Royal Adelaide Hospital, Adelaide, South Australia, Australia, 7Department of Obstetrics and Gynaecology University of Melbourne, Melbourne, Victoria, Australia and 8Border Medical Oncology, Wodonga, Victoria, Australia Abstract Background. To investigate whether variations in primary chemotherapy were associated with survival in a nationally complete cohort of Australian women with epithelial ovarian cancer (EOC). Material and methods. All 1192 women diagnosed with invasive EOC in Australia in 2005 were identified through state-based cancer registries. Medical record information including details of primary chemotherapy treatment was obtained and survival data updated in 2012. Those started on standard chemotherapy (carboplatin and paclitaxel given at three-weekly intervals) after primary cytoreductive surgery were included (n 351) and the relative dose intensity (RDI) was calculated. Time interval between surgery and start of chemotherapy was analysed in weeks. Hazard ratios [HR, 95% confidence interval (CI)] were calculated using multivariable Cox proportional hazards models. Results. Compared to women with RDI of 91–100%, those with RDI of 70% had significantly poor survival (HRadj 1.62, 95% CI 1.05–2.49). This association was stronger among women with advanced (FIGO stage III/IV) disease at diagnosis (HRadj 1.90, 95% CI 1.22–2.96). The interval between primary surgery and chemotherapy was not related to survival (HRadj 0.98, 95% CI 0.93–1.03 for every week of delay), at least up to a period of five weeks. Conclusion. Our results suggest that RDI of 70% or less was associated with poorer survival, particularly in women with advanced stage EOC. In contrast, the interval duration between primary surgery and chemotherapy was not related to survival, irrespective of disease stage or residual disease. These results provide some reassurance that, at least up until five weeks post-surgery, timing of chemotherapy commencement has a negligible effect on survival. Long-term survival for women with invasive epithelial ovarian cancer (EOC) is poor with five-year relative survival of 50% [1]. There is a concerted effort to improve outcomes through earlier diagnosis and better treatments but, until those longer term goals are realised, optimising use of existing evidencebased treatments may represent one achievable option for improving outcomes for women with ovarian cancer.
Standard treatment for women with EOC includes surgery to achieve maximal cytoreduction and combination platinum/taxane chemotherapy [2]. Variations in the timing of commencing chemotherapy and subsequent cycles as well as dose modifications are common because of individual patient/clinical circumstances and treatment-related side effects [3]. Such modifications occur much more commonly in the general population than in the more selected
Correspondence: S. Anuradha, 120 Fleming Road, Chapel Hill, Queensland 4069, Australia. Tel: 61 401 621977. E-mail: [email protected] (Received 14 March 2014; accepted 20 May 2015) ISSN 0284-186X print/ISSN 1651-226X online © 2015 Informa Healthcare DOI: 10.3109/0284186X.2015.1054950
Downloaded by [OARE Consortium] at 18:06 30 June 2015
2
S. Anuradha et al.
clinical trial populations where the benefits of these regimens were initially tested and, as a result, the effects of such variations on survival are not well established. Some studies have found no associations between survival and modifications in chemotherapy dosing and timing [4,5], while others reported that reduced chemotherapy relative dose intensity (RDI) was associated with poorer progression-free and/or overall survival in women with EOC [6,7]. In addition, the optimal interval between primary cytoreductive surgery and commencing chemotherapy has not been defined, with some studies [8,9] showing that this interval does not influence survival and others [10,11] reporting that delaying chemotherapy could compromise survival. The aim of our study was to clarify these important questions by investigating the association between common variations in chemotherapy timing and dosage and survival in an Australian populationbased national cohort of women diagnosed with EOC in 2005. Material and methods In 2009, Australian state-and territory-based cancer registries identified all women aged 18 years that were diagnosed with primary invasive EOC (including fallopian tube/primary peritoneal cancers) in Australia during 2005. Registration of all cancer diagnoses is mandated by Australian law. The cancer registries provided de-identified information on women’s age, grade and histological subtype of their cancer, and, if applicable, death details. The registries also provided research nurses with information to enable location of each woman’s medical record. From these, comprehensive clinical information including patient characteristics (age, comorbidities); cancer characteristics [International Federation of Gynaecology and Obstetrics (FIGO, 1988) stage and sub-stage, histological subtype, grade] and treatment (surgery, chemotherapy, other treatments) was abstracted [3,12]. We derived comorbidity scores and categorised women into four groups (score 0, 1, 2 or 3) based on the Charlson Comorbidity Index [13] which assigns points according to number and severity of diseases. The histological subtype of cancer was classified as serous, mucinous, endometrioid, clear cell, carcinosarcoma or other; and tumour grade as well differentiated (G1), moderately differentiated (G2) or poorly differentiated (G3). Where stage information was not recorded, we assumed a woman had advanced stage disease (stage III or IV) if she had macroscopic residual disease and/or poorly differentiated cancer with serous histological subtype (15 women in this analysis) as among those with stage information, most in these groups had advanced disease. Surgical details
collected included the date, type (primary/interval cytoreduction) and extent of surgery; amount of residual disease; and surgical complications (e.g. perioperative haemorrhage, pulmonary embolism). We categorised residual disease as present or absent. Chemotherapy information collected included: treatment dates; drugs and doses given; and reasons for delays, dose reductions or drug cessation. In 2005 standard treatment was considered to be combination carboplatin (5 or 6 AUC) and paclitaxel (175 mg/m2) given at three-weekly intervals [14]. For these analyses we included only women who had had primary cytoreductive surgery with adjuvant carboplatin and paclitaxel administered at threeweekly intervals. Women who died or had progressive disease before completing chemotherapy were excluded. We calculated RDI as the planned dose intensity divided by the actual dose intensity. Dose intensity for each drug (carboplatin and paclitaxel) was calculated by multiplying the initial dose of each drug by six and dividing this by 105 (five 3-week intervals over the six cycles; 5 21 days 105 days). We calculated the actual total dose intensity by summing up individual doses received and dividing by the total time taken to receive the doses. In cases where drugs were permanently discontinued for reason other than disease progression or death, remaining cycles were calculated with the standard duration and zero dose [15]. The RDI was averaged across the two drugs. We initially analysed RDI in increments of 10% using 91–100% RDI as the reference category. As our results suggested that survival declined with RDIs below the reference category, we also analysed the effect of a 5% decrement in RDI (from 100% to 70%). Analyses were repeated for RDI of the individual drugs and the combined total dose of the drugs using similar cut-points and reference categories. Women who received six cycles of combination chemotherapy were also categorised according to the cumulative number of days of delay over their course of chemotherapy (0–3 days; 4–21 days; 22–42 days; 42 days). The time interval between surgery and commencing chemotherapy was analysed in weeks both categorically (six categories: 7 days; 8–14 days; 15–21 days; 22–28 days; 29–35 days and 35 days) and continuously (per week). All analyses were repeated restricting to those with advanced (FIGO stage III/IV) disease at diagnosis (n 253, 72%) and, for analyses of the interval between surgery and chemotherapy, those who had residual disease. Survival data The outcome was all-cause mortality; over 95% of deaths were from ovarian cancer. Survival data were obtained from medical records and cancer registries
Chemotherapy and survival in ovarian cancer
in 2009 with updated data obtained from all but one cancer registry in 2012. Updated data for one-half of the women identified through this latter cancer registry were available up until October 2011 through linkage via another study [16] to the Australian National Death Index. Participants not known to have died were censored at the last date of data linkage. Survival time was calculated from date of diagnosis until death or censoring.
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Statistical analysis We calculated hazard ratios (HR) and 95% confidence intervals (CI) for the association between chemotherapy variations and survival using multivariable Cox proportional hazards models to adjust for potential confounders (age and comorbidity in RDI analyses; age, comorbidity, stage, grade, subtype, residual disease, extent of surgery, and surgical complications in analyses of interval between surgery and chemotherapy). Analyses were conducted using STATA v11.2 (Stata Corporation, College Station, TX, USA). The study was approved by the Human Research Ethics Committees at the QIMR Berghofer Medical Research Institute and all participating institutions. Results Overall, 1192 women were identified as being diagnosed with invasive EOC in Australia in 2005. Of these, 351 met inclusion criteria for the current analyses. Figure 1 shows reasons for exclusion. Overall, 334 women had either no cytoreductive surgery and/ or no chemotherapy. Mostly this was because of frailty or comorbidities, although 42 had low-grade/early
3
stage cancer and therefore chemotherapy was not indicated [3]. For 94 women who had primary cytoreductive surgery and adjuvant combination chemotherapy some dose information was missing and they were thus excluded from these analyses. However, there were no statistically significant differences in the characteristics of these women when compared to those included in the analyses (Table I). The average age of the included women was 57.7 years, the majority (72%) had advanced disease at diagnosis (FIGO stage III or IV) and most (67%) had serous cancers. Table II shows the associations between chemotherapy variations (RDI, total dose and delays) and survival. The median RDI was 92% (range 17– 116%). Compared to those whose RDI was 91– 100%, those whose RDI was 70% were just over 60% more likely to die over the observation period (HRadj 1.62, 95% CI 1.05–2.49). This association appeared somewhat stronger among women with advanced (FIGO stage III/IV) disease (HRadj 1.90, 95% CI 1.22–2.96). Figure 2 shows Kaplan-Meier survival curves by RDI of carboplatin and paclitaxel among women with advanced disease. No significant association was found among those with FIGO stage I or II disease (HRadj 0.51, 95% CI 0.06–4.16) although numbers in this group were small. Among those with RDI 70%, for 43% of women, the reason for the dose reduction or delay was recorded as ‘toxicity’ without specific details. Of those with information available, side effects included neutropenia/ anaemia (45%), gastrointestinal (14%), reaction to taxol (10%) and others (31%). We also considered the RDIs for each drug separately. The median RDI for carboplatin was 95% (range 17–133%) and 91% (range 10–109%) for
Figure 1. Flowchart showing inclusions and exclusions for women with ovarian cancer included in the study analysis (N 351).
4
S. Anuradha et al. Table I. Characteristics of women included in the analyses compared with those excluded due to missing chemotherapy dose data. Characteristic
Included N 351 n (%)
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Age (Mean in years) Comorbidity scoreb 0 1 2 3 Missing FIGO stage Early (I & II) Advanced (III & IV) Missing Histological subtype Serous Mucinous Endometroid Clear cell Carcinosarcoma Other epithelial Residual disease (cm) 0 0 Missing
57.7
No dose data N 94 n (%) 57.4 n (%)
p-Valuea 0.78
290 (83) 29 (8) 24 (7) 6 (2) 2
72 (79) 6 (7) 9 (10) 4 (4) 3
0.31
96 (28) 253 (72) 2
23 (25) 71 (75) 0
0.70
234 7 30 49 6 25
63 3 14 8 3 3
(67) (3) (15) (9) (3) (3)
0.17
34 (45) 41 (55) 19
0.97
(67) (2) (8) (14) (2) (7)
143 (45) 173 (55) 35
based on t-test for age and c2 statistic for other variables; bweighted score based on the Charlson Comorbidity Index. a p-Value
paclitaxel. The HRs were greater than one for those who received 70% RDI of either carboplatin or paclitaxel (HRadj 1.33, 95% CI 0.82–2.15 and HRadj 1.19, 95% CI 0.79–1.81, respectively) compared to those who received 91–100% but neither estimate reached statistical significance. Total dose received and treatment delays were also considered separately but the results (Table II) did not indicate that one element was more important than the other although these analyses were limited by small numbers. Time interval between surgery and chemotherapy The median interval between primary surgery and commencing chemotherapy was 17 days (range 2–178 days). Compared with women who commenced chemotherapy 8–14 days after surgery, the survival of those who commenced treatment either earlier or later (up to five weeks after surgery) than this was not significantly worse. There was no significant effect modification by disease stage or residual disease (Table III). Discussion This population-based study shows that among women with EOC who have primary cytoreductive surgery, adjuvant chemotherapy RDI of 70% was
associated with poorer survival, particularly in women with advanced disease at diagnosis. In contrast, we found that, up to five weeks, the interval duration between primary surgery and commencing chemotherapy, was not related to survival. While the importance of chemotherapy RDI for survival has been established for some other cancers including breast cancer and lymphoma [17,18], information on ovarian cancer is limited. Our finding of poorer survival among women with reduced RDI, particularly those with advanced cancer, is consistent with two previous studies which found that RDIs of less than 85% and 70%, respectively [6,7] were associated with worse overall or progression-free survival. An older study in women with advanced ovarian cancer (n 226) found no association between RDI and progression-free survival [4], however, the different drug regimens in that study preclude direct comparisons with our results. Another small study of women with advanced ovarian cancer (n 157) investigated associations between dose reductions and delays individually rather than calculating RDI and found these variations were not associated with overall survival [5]. Our analyses of these separate elements suggested that both may affect survival although estimates were not statistically significant. Both our analyses and those by Nagel et al. [5] were limited by small numbers. Similarly, our analyses of
Chemotherapy and survival in ovarian cancer
5
Table II. Hazard ratios for the association between survival and relative dose intensity (RDI), total chemotherapy dose, chemotherapy delays amongst women diagnosed with invasive epithelial ovarian cancer in Australia in 2005.
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Unadjusted HR (95% CI) Variable
N
Total RDI% 100b 91–100 81–90 71–80 70 Per 5% decrease in RDIc
13 175 84 28 51
Adjusted HR (95% CI)a N
All women N 351 1.65 (0.80–3.42) Ref 1.04 (0.71–1.51) 1.18 (0.64–2.16) 1.85 (1.23–2.78)
Unadjusted HR (95% CI)
1.50 (0.73–3.12) Ref 1.00 (0.68–1.45) 1.18 (0.64–2.16) 1.62 (1.05–2.49)
1.08 (1.01–1.16) 1.05 (0.98–1.13) % planned combined dose of carboplatin and paclitaxel 25 1.64 (0.97–2.78) 1.46 (0.86–2.48) 100d 91–100 230 Ref Ref 81–90 44 1.24 (0.79–1.95) 1.16 (0.74–1.83) 71–80 17 1.12 (0.55–2.30) 1.05 (0.51–2.16) 35 1.31 (0.79–2.16) 1.06 (0.61–1.83) 70 Per 5% decrease in % planned combined dose of carboplatin and paclitaxelc 1.04 (0.96–1.12) 1.00 (0.92–1.09) Duration of delay during the course of chemotherapy (days) 327e 0–3 175 Ref Ref 4–21 113 0.97 (0.69–1.38) 0.96 (0.68–1.36) 22–42 26 1.91 (1.18–3.08) 1.75 (1.08–2.85) 13 2.07 (1.04–4.12) 1.94 (0.97–3.86) 42 Per week of delay during the course of chemotherapy 1.14 (1.04–1.25) 1.12 (1.03–1.23)
13 120 62 22 36
24 162 34 12 21
244e 122 85 25 12
Adjusted HR (95% CI)a
Women with advanced disease N 253 1.14 (0.64–2.77) Ref 1.03 (0.70–1.54) 1.22 (0.64–2.19) 2.08 (1.35–3.19)
1.07 (0.51–2.23) Ref 1.00 (0.67–1.48) 1.16 (0.63–2.13) 1.90 (1.22–2.96)
1.09 (1.02–1.18)
1.08 (1.00–1.16)
1.19 (0.70–2.01) Ref 1.24 (0.78–1.96) 1.20 (0.56–2.58) 1.80 (1.06–3.06)
1.11 (0.65–1.88) Ref 1.21 (0.76–1.92) 1.10 (0.51–2.32) 1.57 (0.91–2.73)
1.09 (1.00–1.18)
1.06 (0.97–1.16)
Ref 0.91 (0.63–1.30) 1.26 (0.77–2.07) 1.50 (0.75–2.99)
Ref 0.89 (0.62– 1.28) 1.18 (0.71 –1.97) 1.44 (0.72 – 2.89)
1.07 (0.98–1.18)
1.06 (0.97–1.16)
aadjusted
for age, comorbidity; bThose in this category had RDI of 101–110 except one woman with RDI of 116%; cMeasures above 100% were excluded from this calculation; dThose in this category had a % planned dose of 101–110 except one woman with a dose of 116%; eOnly those who had at least 6 cycles of chemotherapy were included. All the bold values were significant at P 0.05
the RDIs of individual drugs were also limited by the small sample. We had expected that the RDI of carboplatin might have a stronger association with survival than that for paclitaxel but our results did not clearly show this. This issue should be further explored in a larger data set. Overall, our findings suggest that wherever possible, dose reductions and delays should be avoided and that measures to maintain RDI need to be
Figure 2. Kaplan-Meier survival curves for ovarian cancer by relative dose intensity (RDI %) of carboplatin and paclitaxel among women with advanced disease (N 253).
explored. A meta-analysis of randomised-controlled trials of granulocyte colony stimulating factor (G-CSF) use in adults receiving chemotherapy for solid tumours or lymphoma found that G-CSF support increase delivered dose intensity and reduced allcause mortality [19]. It is not clear how many ovarian cancer cases were included in this meta-analysis, and currently G-CSF is not approved in Australia for use in this patient group. Based on our findings, consideration should be given to conducting a trial specifically addressing the potential of growth factors to maintain RDI in women with EOC. Unfortunately, at present there are no effective interventions to prevent paclitaxel-related neurotoxicity which is a common cause for dose reductions, although some evidence suggests that serotonin-norepinephrine reuptake inhibitors (e.g. duloxetine) may help reduce associated neuropathic pain [20]. In contrast to our findings in relation to RDI, our results did not suggest that the duration of the interval between surgery and commencing chemotherapy was associated with survival. It has been suggested that surgery alters tumour biology resulting in accelerated repopulation of tumour cells [21]. Delay in commencing chemotherapy might there-
6
S. Anuradha et al. Table III. Hazard ratios for the association between time interval (TI) between surgery and initiation of chemotherapy and survival amongst women diagnosed with invasive epithelial ovarian cancer in Australia in 2005. Time interval (days)
Unadjusted HR (95% CI)
Adjusted HRa (95% CI)
1.21 (0.77–1.89) Ref 0.86 (0.54–1.37) 0.82 (0.50–1.35) 0.95 (0.51–1.79) 0.75 (0.48–1.19)
0.89 (0.55–1.43) Ref 0.95 (0.57–1.59) 0.80 (0.48–1.35) 0.93 (0.42–2.11) 1.02 (1.00–1.04)
0.98 (0.93–1.03)
1.00 (0.94–1.06)
173 36 49 30 26 11 21
0.84 (0.50–1.42) Ref 1.14 (0.68–1.92) 0.68 (0.36–1.27) 2.10 (1.03–4.29) 0.84 (0.45–1.57)
0.77 (0.44–1.35) Ref 1.06 (0.59–1.93) 0.71 (0.37–1.36) 1.30 (0.47–3.58) 0.81 (0.40–1.64)
143 11 37 24 24 12 35
1.39 (0.48–4.01) Ref 0.24 (0.05–1.10) 1.05 (0.43–2.63) 0.47 (0.10–2.13) 1.09 (0.47–2.51)
2.10 (0.68–6.51) Ref 0.49 (0.10–2.33) 0.61 (0.21–1.74) 0.62 (0.07–5.53) 2.36 (0.90–6.18)
253 75 44 38 37 15 44
0.96 (0.61–1.53) Ref 1.11 (0.70–1.78) 0.68 (0.39–1.17) 1.37 (0.71–2.64) 0.88 (0.55–1.41)
0.83 (0.50–1.35) Ref 1.02 (0.61–1.73) 0.67 (0.38–1.18) 1.11 (0.49–2.55) 1.03 (0.61–1.75)
N
All women N 351 7 8–14 15–21 22–28 29–35 35 Per week increase in TI
351 50 95 60 51 24 71
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Women with residual disease N 173 7 8–14 15–21 22–28 29–35 35 Women without residual disease N 143 7 8–14 15–21 22–28 29–35 35 Women with advanced disease N 253 7 8–14 15–21 22–28 29–35 35 a adjusted
for age, comorbidity, stage, grade, subtype, residual disease, extent of surgery, surgical complications.
fore negatively affect women’s survival. Alternatively, chemotherapy given too soon after surgery might adversely affect recovery. The optimal timing of chemotherapy is yet to be determined. Authors of a 2009 review of seven studies could not reach a definite conclusion because of the disparate nature (methods, regimens and intervals) of the studies [22]. The results of more recent studies, including only women treated with platinum/taxane-based regimens, have also been conflicting (Table IV). Two studies, including the largest (n 3326) [23], like us, found no overall association between the interval between surgery and chemotherapy and survival. However, the latter study did report slightly worse survival for each additional week of delay of chemotherapy in women without residual disease after surgery. This contrasts with findings from another study (n 191) which showed that women with residual disease who commenced chemotherapy more than
four weeks after surgery had significantly poorer survival [10]. We found no evidence that residual disease modified the association although our analysis was insufficiently powered to detect small effects. The reason for these conflicting results is not immediately clear, however, our results along with those from similar studies provide some reassurance that up until 4–5 weeks post-surgery, the timing of chemotherapy commencement has a negligible effect on survival. A strength of our study is the population-based design, which means our results are generalisable to the broader group of women diagnosed with EOC, not just those from select centres or participants in clinical trials. Further strengths are the long duration of follow-up and the availability of detailed clinical information for most women. Complete chemotherapy dose information was not available for some women because appropriate medical records were not accessible but our comparison of characteristics
Chemotherapy and survival in ovarian cancer
7
Table IV. Summary of recent studies on the association of time interval (TI) between surgery and initiation of chemotherapy with survival in epithelial ovarian cancer. Years of diagnosis
Number of women
FIGO Stage
Hofstetter, 2013 [10] (Germany/Belgium/ Austria)
2005
191
III– IV (serous subtype)
carboplatin/ taxane
28
Mahner, 2013 [23] (Germany/France)
1995– 2002 (three RCTs)
3326
IIB–IV
platinum/ paclitaxel (/- study drug)
19
Lydiksen, 2014 [24] (Denmark)
2005– 2006
650
I–IV
carboplatin/ taxane
32
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Study, year (place)
Chemotherapy
across patients with and without complete chemotherapy data suggests there were no major differences. Performance status was often not recorded in clinical notes so the lack of these data, which could have influenced chemotherapy doses and timing, is another limitation of our study although we have adjusted our results for comorbidities and age. For consistent comparisons we have only included women who had primary cytoreductive surgery followed by intravenous carboplatin and paclitaxel commencing at three-weekly intervals; this was standard treatment when the women in our study were diagnosed. While this regimen is still a recommended first-line approach [25], other regimens (neoadjuvant, intraperitoneal, dose-dense) are now also recommended and increasingly used. Our findings may not be directly applicable to women receiving those regimens. Finally, our study was observational with the potential for bias inherent to all such studies. While randomisedcontrolled trials are the preferred method for assessing treatment effects, such a study to assess the effects of variation in RDI would not be feasible. In conclusion, our results suggest that, among women with advanced EOC, maintaining dose intensity during chemotherapy may help improve survival. Clinical trials to examine the use and costeffectiveness of growth factors or other options to maintain RDI are warranted. However, with respect to the time between primary surgery and commencing chemotherapy, our results provide reassurance that intervals of up to five weeks are unlikely to negatively impact upon survival. Acknowledgements We acknowledge the members of the Patterns of Care Study Group; and the research nurses and associated
Median TI (days)
Hazard ratio for TI
Adjustment factors
Age, FIGO stage, HR 1.73 (1.08– residual disease, 2.78) for 28 days extent of surgery, vs. 28 days Any and treatment residual disease centre HR 2.24 (1.08– 4.66) Overall no significant Age, FIGO stage, association No residual disease, residual disease grade and ascites HR 1.087 (1.005–1.176) per additional week TI Residual disease, HR 1.13 (0.92– FIGO stage 1.39) 32 days vs. 32 days
staff for assistance with data collection. We also acknowledge the Australian Ovarian Cancer Study Group for providing clinical information for women who participated in that study. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This work was supported by Cancer Australia (formerly National Breast and Ovarian Cancer Centre). Collection of clinical data from the Australian Ovarian Cancer Study was funded by the National Health and Medical Research Council (NHMRC) (400281, 400413); Cancer Councils of Victoria, Queensland, New South Wales, South Australia and Tasmania; and the Cancer Foundation of Western Australia. Susan Jordan and Penelope Webb are funded by fellowships from the NHMRC. This study was also supported by the RioTinto Ride to Conquer Cancer Grant through QIMR Berghofer Medical Research Institute. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. References [1] Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012;62:10–29. [2] Metzger-Filho O, Moulin C, D’Hondt V. First-line systemic treatment of ovarian cancer: A critical review of available evidence and expectations for future directions. Curr Opin Oncol 2010;22:513–20. [3] Jordan S, Steer C, DeFazio A, Quinn M, Obermair A, Friedlander M, et al. Patterns of chemotherapy treatment for women with invasive epithelial ovarian cancer – a populationbased study. Gynecol Oncol 2013;129:310–7. [4] Repetto L, Pace M, Mammoliti S, Bruzzone M, Chiara S, Oliva C, et al. The impact of received dose intensity on the outcome of advanced ovarian cancer. Eur J Cancer 1993; 29A:181–4.
Downloaded by [OARE Consortium] at 18:06 30 June 2015
8
S. Anuradha et al.
[5] Nagel CI, Backes FJ, Hade EM, Cohn DE, Eisenhauer EL, O’Malley DM, et al. Effect of chemotherapy delays and dose reductions on progression free and overall survival in the treatment of epithelial ovarian cancer. Gynecol Oncol 2012;124:221–4. [6] Hanna RK, Poniewierski MS, Laskey RA, Lopez MA, Shafer A, Van Le L, et al. Predictors of reduced relative dose intensity and its relationship to mortality in women receiving multiagent chemotherapy for epithelial ovarian cancer. Gynecol Oncol 2013;129:74–80. [7] Fauci JM, Whitworth JM, Schneider KE, Subramaniam A, Zhang B, Frederick PJ, et al. Prognostic significance of the relative dose intensity of chemotherapy in primary treatment of epithelial ovarian cancer. Gynecol Oncol 2011; 122:532–5. [8] Gadducci A, Sartori E, Landoni F, Zola P, Maggino T, Maggioni A, et al. Relationship between time interval from primary surgery to the start of taxane-plus platinum-based chemotherapy and clinical outcome of patients with advanced epithelial ovarian cancer: Results of a multicenter retrospective Italian study. J Clin Oncol 2005;23:751–8. [9] Flynn PM, Paul J, Cruickshank DJ, Scottish Gynaecological Cancer Trials G. Does the interval from primary surgery to chemotherapy influence progression-free survival in ovarian cancer? Gynecol Oncol 2002;86:354–7. [10] Hofstetter G, Concin N, Braicu I, Chekerov R, Sehouli J, Cadron I, et al. The time interval from surgery to start of chemotherapy significantly impacts prognosis in patients with advanced serous ovarian carcinoma – analysis of patient data in the prospective OVCAD study. Gynecol Oncol 2013;131:15–20. [11] Wright JD, Herzog TJ, Neugut AI, Burke WM, Lu YS, Lewin SN, et al. Effect of radical cytoreductive surgery on omission and delay of chemotherapy for advanced-stage ovarian cancer. Obstet Gynecol 2012;120:871–81. [12] Anuradha S, Webb PM, Blomfield P, Brand AH, Friedlander M, Leung Y, et al. Survival of Australian women with invasive epithelial ovarian cancer: A population-based study. Med J Aust 2014;201:283–8. [13] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373–83. [14] du Bois A, Quinn M, Thigpen T, Vermorken J, Avall-Lundqvist E, Bookman M, et al. 2004 consensus statements on the management of ovarian cancer: Final document of the 3rd International Gynecologic Cancer Intergroup
Ovarian Cancer Consensus Conference (GCIG OCCC 2004). Ann Oncol 2005;16(Suppl 8):viii7–12. [15] Loibl S, Skacel T, Nekljudova V, Luck HJ, Schwenkglenks M, Brodowicz T, et al. Evaluating the impact of relative total dose intensity (RTDI) on patients’ short and long-term outcome in taxane- and anthracycline-based chemotherapy of metastatic breast cancer – a pooled analysis. BMC Cancer 2011;11:131. [16] Merritt MA, Green AC, Nagle CM, Webb PM. Talcum powder, chronic pelvic inflammation and NSAIDs in relation to risk of epithelial ovarian cancer. Int J Cancer 2008; 122:170–6. [17] Hryniuk W, Bush H. The importance of dose intensity in chemotherapy of metastatic breast cancer. J Clin Oncol 1984;2:1281–8. [18] Lyman GH, Dale DC, Tomita D, Whittaker S, Crawford J. A retrospective evaluation of chemotherapy dose intensity and supportive care for early-stage breast cancer in a curative setting. Breast Cancer Res Treat 2013;139:863–72. [19] Lyman GH, Dale DC, Culakova E, Poniewierski MS, Wolff DA, Kuderer NM, et al. The impact of the granulocyte colony-stimulating factor on chemotherapy dose intensity and cancer survival: A systematic review and meta-analysis of randomized controlled trials. Ann Oncol 2013;24:2475–84. [20] Aziz MT, Good BL, Lowe DK. Serotonin-norepinephrine reuptake inhibitors for the management of chemotherapyinduced peripheral neuropathy. Ann Pharmacother 2014; 48:626–32. [21] Cividalli A, Creton G, Ceciarelli F, Strigari L, Tirindelli Danesi D, Benassi M. Influence of time interval between surgery and radiotherapy on tumor regrowth. J Exp Clin Canc Res 2005;24:109–16. [22] Larsen E, Blaakaer J. Epithelial ovarian cancer: Does the time interval between primary surgery and postoperative chemotherapy have any prognostic importance? Acta Obstet Gynecol Scand 2009;88:373–7. [23] Mahner S, Eulenburg C, Staehle A, Wegscheider K, Reuss A, Pujade-Lauraine E, et al. Prognostic impact of the time interval between surgery and chemotherapy in advanced ovarian cancer: Analysis of prospective randomised phase III trials. Eur J Cancer 2013;49:142–9. [24] Lydiksen L, Jensen-Fangel S, Blaakaer J. Is it possible to define an optimal time for chemotherapy after surgery for ovarian cancer? Gynecol Oncol 2014;133:454–9. [25] National Comprehensive Cancer Network Guidelines – Ovarian Cancer Version 3.2012. [cited 2014 Oct 10]. Available from: http://www.nccn.org/professionals/physician_gls/f_guidelines.asp.
Acta Oncologica Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/ionc20
Variations in adjuvant chemotherapy and survival in women with epithelial ovarian cancer – a populationbased study ab
b
b
c
Satyamurthy Anuradha , Peter J. Donovan , Penelope M. Webb , Alison H. Brand , Jeffrey d
e
f
g
h
Goh , Michael Friedlander , Martin K. Oehler , Michael Quinn , Christopher Steer & Susan J. Jordan a
b
The University of Queensland, School of Public Health, Brisbane, Queensland, Australia
b
The QIMR Berghofer Medical Research Institute, Gynaecological Cancers Group, Brisbane, Queensland, Australia c
Department of Gynaecological Oncology, Westmead Hospital, University of Sydney, Westmead, New South Wales, Australia d
Royal Brisbane and Women's Hospital, Brisbane, Queensland, Australia
e
Department of Medical Oncology, Prince of Wales Hospital, Sydney, New South Wales, Australia f
Department of Gynaecological Oncology, Royal Adelaide Hospital, Adelaide, South Australia, Australia g
Department of Obstetrics and Gynaecology University of Melbourne, Melbourne, Victoria, Australia h
Border Medical Oncology, Wodonga, Victoria, Australia Published online: 17 Jun 2015.
To cite this article: Satyamurthy Anuradha, Peter J. Donovan, Penelope M. Webb, Alison H. Brand, Jeffrey Goh, Michael Friedlander, Martin K. Oehler, Michael Quinn, Christopher Steer & Susan J. Jordan (2015): Variations in adjuvant chemotherapy and survival in women with epithelial ovarian cancer – a population-based study, Acta Oncologica To link to this article: http://dx.doi.org/10.3109/0284186X.2015.1054950
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
Downloaded by [OARE Consortium] at 18:06 30 June 2015
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Acta Oncologica, 2015; Early Online: 1–8
ORIGINAL ARTICLE
Variations in adjuvant chemotherapy and survival in women with epithelial ovarian cancer – a population-based study
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Satyamurthy Anuradha1,2, Peter J. Donovan2, Penelope M. Webb2, Alison H. Brand3, Jeffrey Goh4, Michael Friedlander5, Martin K. Oehler6, Michael Quinn7, Christopher Steer8 & Susan J. Jordan2 1The
University of Queensland, School of Public Health, Brisbane, Queensland, Australia, 2The QIMR Berghofer Medical Research Institute, Gynaecological Cancers Group, Brisbane, Queensland, Australia, 3Department of Gynaecological Oncology, Westmead Hospital, University of Sydney, Westmead, New South Wales, Australia, 4Royal Brisbane and Women’s Hospital, Brisbane, Queensland, Australia, 5Department of Medical Oncology, Prince of Wales Hospital, Sydney, New South Wales, Australia, 6Department of Gynaecological Oncology, Royal Adelaide Hospital, Adelaide, South Australia, Australia, 7Department of Obstetrics and Gynaecology University of Melbourne, Melbourne, Victoria, Australia and 8Border Medical Oncology, Wodonga, Victoria, Australia Abstract Background. To investigate whether variations in primary chemotherapy were associated with survival in a nationally complete cohort of Australian women with epithelial ovarian cancer (EOC). Material and methods. All 1192 women diagnosed with invasive EOC in Australia in 2005 were identified through state-based cancer registries. Medical record information including details of primary chemotherapy treatment was obtained and survival data updated in 2012. Those started on standard chemotherapy (carboplatin and paclitaxel given at three-weekly intervals) after primary cytoreductive surgery were included (n 351) and the relative dose intensity (RDI) was calculated. Time interval between surgery and start of chemotherapy was analysed in weeks. Hazard ratios [HR, 95% confidence interval (CI)] were calculated using multivariable Cox proportional hazards models. Results. Compared to women with RDI of 91–100%, those with RDI of 70% had significantly poor survival (HRadj 1.62, 95% CI 1.05–2.49). This association was stronger among women with advanced (FIGO stage III/IV) disease at diagnosis (HRadj 1.90, 95% CI 1.22–2.96). The interval between primary surgery and chemotherapy was not related to survival (HRadj 0.98, 95% CI 0.93–1.03 for every week of delay), at least up to a period of five weeks. Conclusion. Our results suggest that RDI of 70% or less was associated with poorer survival, particularly in women with advanced stage EOC. In contrast, the interval duration between primary surgery and chemotherapy was not related to survival, irrespective of disease stage or residual disease. These results provide some reassurance that, at least up until five weeks post-surgery, timing of chemotherapy commencement has a negligible effect on survival. Long-term survival for women with invasive epithelial ovarian cancer (EOC) is poor with five-year relative survival of 50% [1]. There is a concerted effort to improve outcomes through earlier diagnosis and better treatments but, until those longer term goals are realised, optimising use of existing evidencebased treatments may represent one achievable option for improving outcomes for women with ovarian cancer.
Standard treatment for women with EOC includes surgery to achieve maximal cytoreduction and combination platinum/taxane chemotherapy [2]. Variations in the timing of commencing chemotherapy and subsequent cycles as well as dose modifications are common because of individual patient/clinical circumstances and treatment-related side effects [3]. Such modifications occur much more commonly in the general population than in the more selected
Correspondence: S. Anuradha, 120 Fleming Road, Chapel Hill, Queensland 4069, Australia. Tel: 61 401 621977. E-mail: [email protected] (Received 14 March 2014; accepted 20 May 2015) ISSN 0284-186X print/ISSN 1651-226X online © 2015 Informa Healthcare DOI: 10.3109/0284186X.2015.1054950
Downloaded by [OARE Consortium] at 18:06 30 June 2015
2
S. Anuradha et al.
clinical trial populations where the benefits of these regimens were initially tested and, as a result, the effects of such variations on survival are not well established. Some studies have found no associations between survival and modifications in chemotherapy dosing and timing [4,5], while others reported that reduced chemotherapy relative dose intensity (RDI) was associated with poorer progression-free and/or overall survival in women with EOC [6,7]. In addition, the optimal interval between primary cytoreductive surgery and commencing chemotherapy has not been defined, with some studies [8,9] showing that this interval does not influence survival and others [10,11] reporting that delaying chemotherapy could compromise survival. The aim of our study was to clarify these important questions by investigating the association between common variations in chemotherapy timing and dosage and survival in an Australian populationbased national cohort of women diagnosed with EOC in 2005. Material and methods In 2009, Australian state-and territory-based cancer registries identified all women aged 18 years that were diagnosed with primary invasive EOC (including fallopian tube/primary peritoneal cancers) in Australia during 2005. Registration of all cancer diagnoses is mandated by Australian law. The cancer registries provided de-identified information on women’s age, grade and histological subtype of their cancer, and, if applicable, death details. The registries also provided research nurses with information to enable location of each woman’s medical record. From these, comprehensive clinical information including patient characteristics (age, comorbidities); cancer characteristics [International Federation of Gynaecology and Obstetrics (FIGO, 1988) stage and sub-stage, histological subtype, grade] and treatment (surgery, chemotherapy, other treatments) was abstracted [3,12]. We derived comorbidity scores and categorised women into four groups (score 0, 1, 2 or 3) based on the Charlson Comorbidity Index [13] which assigns points according to number and severity of diseases. The histological subtype of cancer was classified as serous, mucinous, endometrioid, clear cell, carcinosarcoma or other; and tumour grade as well differentiated (G1), moderately differentiated (G2) or poorly differentiated (G3). Where stage information was not recorded, we assumed a woman had advanced stage disease (stage III or IV) if she had macroscopic residual disease and/or poorly differentiated cancer with serous histological subtype (15 women in this analysis) as among those with stage information, most in these groups had advanced disease. Surgical details
collected included the date, type (primary/interval cytoreduction) and extent of surgery; amount of residual disease; and surgical complications (e.g. perioperative haemorrhage, pulmonary embolism). We categorised residual disease as present or absent. Chemotherapy information collected included: treatment dates; drugs and doses given; and reasons for delays, dose reductions or drug cessation. In 2005 standard treatment was considered to be combination carboplatin (5 or 6 AUC) and paclitaxel (175 mg/m2) given at three-weekly intervals [14]. For these analyses we included only women who had had primary cytoreductive surgery with adjuvant carboplatin and paclitaxel administered at threeweekly intervals. Women who died or had progressive disease before completing chemotherapy were excluded. We calculated RDI as the planned dose intensity divided by the actual dose intensity. Dose intensity for each drug (carboplatin and paclitaxel) was calculated by multiplying the initial dose of each drug by six and dividing this by 105 (five 3-week intervals over the six cycles; 5 21 days 105 days). We calculated the actual total dose intensity by summing up individual doses received and dividing by the total time taken to receive the doses. In cases where drugs were permanently discontinued for reason other than disease progression or death, remaining cycles were calculated with the standard duration and zero dose [15]. The RDI was averaged across the two drugs. We initially analysed RDI in increments of 10% using 91–100% RDI as the reference category. As our results suggested that survival declined with RDIs below the reference category, we also analysed the effect of a 5% decrement in RDI (from 100% to 70%). Analyses were repeated for RDI of the individual drugs and the combined total dose of the drugs using similar cut-points and reference categories. Women who received six cycles of combination chemotherapy were also categorised according to the cumulative number of days of delay over their course of chemotherapy (0–3 days; 4–21 days; 22–42 days; 42 days). The time interval between surgery and commencing chemotherapy was analysed in weeks both categorically (six categories: 7 days; 8–14 days; 15–21 days; 22–28 days; 29–35 days and 35 days) and continuously (per week). All analyses were repeated restricting to those with advanced (FIGO stage III/IV) disease at diagnosis (n 253, 72%) and, for analyses of the interval between surgery and chemotherapy, those who had residual disease. Survival data The outcome was all-cause mortality; over 95% of deaths were from ovarian cancer. Survival data were obtained from medical records and cancer registries
Chemotherapy and survival in ovarian cancer
in 2009 with updated data obtained from all but one cancer registry in 2012. Updated data for one-half of the women identified through this latter cancer registry were available up until October 2011 through linkage via another study [16] to the Australian National Death Index. Participants not known to have died were censored at the last date of data linkage. Survival time was calculated from date of diagnosis until death or censoring.
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Statistical analysis We calculated hazard ratios (HR) and 95% confidence intervals (CI) for the association between chemotherapy variations and survival using multivariable Cox proportional hazards models to adjust for potential confounders (age and comorbidity in RDI analyses; age, comorbidity, stage, grade, subtype, residual disease, extent of surgery, and surgical complications in analyses of interval between surgery and chemotherapy). Analyses were conducted using STATA v11.2 (Stata Corporation, College Station, TX, USA). The study was approved by the Human Research Ethics Committees at the QIMR Berghofer Medical Research Institute and all participating institutions. Results Overall, 1192 women were identified as being diagnosed with invasive EOC in Australia in 2005. Of these, 351 met inclusion criteria for the current analyses. Figure 1 shows reasons for exclusion. Overall, 334 women had either no cytoreductive surgery and/ or no chemotherapy. Mostly this was because of frailty or comorbidities, although 42 had low-grade/early
3
stage cancer and therefore chemotherapy was not indicated [3]. For 94 women who had primary cytoreductive surgery and adjuvant combination chemotherapy some dose information was missing and they were thus excluded from these analyses. However, there were no statistically significant differences in the characteristics of these women when compared to those included in the analyses (Table I). The average age of the included women was 57.7 years, the majority (72%) had advanced disease at diagnosis (FIGO stage III or IV) and most (67%) had serous cancers. Table II shows the associations between chemotherapy variations (RDI, total dose and delays) and survival. The median RDI was 92% (range 17– 116%). Compared to those whose RDI was 91– 100%, those whose RDI was 70% were just over 60% more likely to die over the observation period (HRadj 1.62, 95% CI 1.05–2.49). This association appeared somewhat stronger among women with advanced (FIGO stage III/IV) disease (HRadj 1.90, 95% CI 1.22–2.96). Figure 2 shows Kaplan-Meier survival curves by RDI of carboplatin and paclitaxel among women with advanced disease. No significant association was found among those with FIGO stage I or II disease (HRadj 0.51, 95% CI 0.06–4.16) although numbers in this group were small. Among those with RDI 70%, for 43% of women, the reason for the dose reduction or delay was recorded as ‘toxicity’ without specific details. Of those with information available, side effects included neutropenia/ anaemia (45%), gastrointestinal (14%), reaction to taxol (10%) and others (31%). We also considered the RDIs for each drug separately. The median RDI for carboplatin was 95% (range 17–133%) and 91% (range 10–109%) for
Figure 1. Flowchart showing inclusions and exclusions for women with ovarian cancer included in the study analysis (N 351).
4
S. Anuradha et al. Table I. Characteristics of women included in the analyses compared with those excluded due to missing chemotherapy dose data. Characteristic
Included N 351 n (%)
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Age (Mean in years) Comorbidity scoreb 0 1 2 3 Missing FIGO stage Early (I & II) Advanced (III & IV) Missing Histological subtype Serous Mucinous Endometroid Clear cell Carcinosarcoma Other epithelial Residual disease (cm) 0 0 Missing
57.7
No dose data N 94 n (%) 57.4 n (%)
p-Valuea 0.78
290 (83) 29 (8) 24 (7) 6 (2) 2
72 (79) 6 (7) 9 (10) 4 (4) 3
0.31
96 (28) 253 (72) 2
23 (25) 71 (75) 0
0.70
234 7 30 49 6 25
63 3 14 8 3 3
(67) (3) (15) (9) (3) (3)
0.17
34 (45) 41 (55) 19
0.97
(67) (2) (8) (14) (2) (7)
143 (45) 173 (55) 35
based on t-test for age and c2 statistic for other variables; bweighted score based on the Charlson Comorbidity Index. a p-Value
paclitaxel. The HRs were greater than one for those who received 70% RDI of either carboplatin or paclitaxel (HRadj 1.33, 95% CI 0.82–2.15 and HRadj 1.19, 95% CI 0.79–1.81, respectively) compared to those who received 91–100% but neither estimate reached statistical significance. Total dose received and treatment delays were also considered separately but the results (Table II) did not indicate that one element was more important than the other although these analyses were limited by small numbers. Time interval between surgery and chemotherapy The median interval between primary surgery and commencing chemotherapy was 17 days (range 2–178 days). Compared with women who commenced chemotherapy 8–14 days after surgery, the survival of those who commenced treatment either earlier or later (up to five weeks after surgery) than this was not significantly worse. There was no significant effect modification by disease stage or residual disease (Table III). Discussion This population-based study shows that among women with EOC who have primary cytoreductive surgery, adjuvant chemotherapy RDI of 70% was
associated with poorer survival, particularly in women with advanced disease at diagnosis. In contrast, we found that, up to five weeks, the interval duration between primary surgery and commencing chemotherapy, was not related to survival. While the importance of chemotherapy RDI for survival has been established for some other cancers including breast cancer and lymphoma [17,18], information on ovarian cancer is limited. Our finding of poorer survival among women with reduced RDI, particularly those with advanced cancer, is consistent with two previous studies which found that RDIs of less than 85% and 70%, respectively [6,7] were associated with worse overall or progression-free survival. An older study in women with advanced ovarian cancer (n 226) found no association between RDI and progression-free survival [4], however, the different drug regimens in that study preclude direct comparisons with our results. Another small study of women with advanced ovarian cancer (n 157) investigated associations between dose reductions and delays individually rather than calculating RDI and found these variations were not associated with overall survival [5]. Our analyses of these separate elements suggested that both may affect survival although estimates were not statistically significant. Both our analyses and those by Nagel et al. [5] were limited by small numbers. Similarly, our analyses of
Chemotherapy and survival in ovarian cancer
5
Table II. Hazard ratios for the association between survival and relative dose intensity (RDI), total chemotherapy dose, chemotherapy delays amongst women diagnosed with invasive epithelial ovarian cancer in Australia in 2005.
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Unadjusted HR (95% CI) Variable
N
Total RDI% 100b 91–100 81–90 71–80 70 Per 5% decrease in RDIc
13 175 84 28 51
Adjusted HR (95% CI)a N
All women N 351 1.65 (0.80–3.42) Ref 1.04 (0.71–1.51) 1.18 (0.64–2.16) 1.85 (1.23–2.78)
Unadjusted HR (95% CI)
1.50 (0.73–3.12) Ref 1.00 (0.68–1.45) 1.18 (0.64–2.16) 1.62 (1.05–2.49)
1.08 (1.01–1.16) 1.05 (0.98–1.13) % planned combined dose of carboplatin and paclitaxel 25 1.64 (0.97–2.78) 1.46 (0.86–2.48) 100d 91–100 230 Ref Ref 81–90 44 1.24 (0.79–1.95) 1.16 (0.74–1.83) 71–80 17 1.12 (0.55–2.30) 1.05 (0.51–2.16) 35 1.31 (0.79–2.16) 1.06 (0.61–1.83) 70 Per 5% decrease in % planned combined dose of carboplatin and paclitaxelc 1.04 (0.96–1.12) 1.00 (0.92–1.09) Duration of delay during the course of chemotherapy (days) 327e 0–3 175 Ref Ref 4–21 113 0.97 (0.69–1.38) 0.96 (0.68–1.36) 22–42 26 1.91 (1.18–3.08) 1.75 (1.08–2.85) 13 2.07 (1.04–4.12) 1.94 (0.97–3.86) 42 Per week of delay during the course of chemotherapy 1.14 (1.04–1.25) 1.12 (1.03–1.23)
13 120 62 22 36
24 162 34 12 21
244e 122 85 25 12
Adjusted HR (95% CI)a
Women with advanced disease N 253 1.14 (0.64–2.77) Ref 1.03 (0.70–1.54) 1.22 (0.64–2.19) 2.08 (1.35–3.19)
1.07 (0.51–2.23) Ref 1.00 (0.67–1.48) 1.16 (0.63–2.13) 1.90 (1.22–2.96)
1.09 (1.02–1.18)
1.08 (1.00–1.16)
1.19 (0.70–2.01) Ref 1.24 (0.78–1.96) 1.20 (0.56–2.58) 1.80 (1.06–3.06)
1.11 (0.65–1.88) Ref 1.21 (0.76–1.92) 1.10 (0.51–2.32) 1.57 (0.91–2.73)
1.09 (1.00–1.18)
1.06 (0.97–1.16)
Ref 0.91 (0.63–1.30) 1.26 (0.77–2.07) 1.50 (0.75–2.99)
Ref 0.89 (0.62– 1.28) 1.18 (0.71 –1.97) 1.44 (0.72 – 2.89)
1.07 (0.98–1.18)
1.06 (0.97–1.16)
aadjusted
for age, comorbidity; bThose in this category had RDI of 101–110 except one woman with RDI of 116%; cMeasures above 100% were excluded from this calculation; dThose in this category had a % planned dose of 101–110 except one woman with a dose of 116%; eOnly those who had at least 6 cycles of chemotherapy were included. All the bold values were significant at P 0.05
the RDIs of individual drugs were also limited by the small sample. We had expected that the RDI of carboplatin might have a stronger association with survival than that for paclitaxel but our results did not clearly show this. This issue should be further explored in a larger data set. Overall, our findings suggest that wherever possible, dose reductions and delays should be avoided and that measures to maintain RDI need to be
Figure 2. Kaplan-Meier survival curves for ovarian cancer by relative dose intensity (RDI %) of carboplatin and paclitaxel among women with advanced disease (N 253).
explored. A meta-analysis of randomised-controlled trials of granulocyte colony stimulating factor (G-CSF) use in adults receiving chemotherapy for solid tumours or lymphoma found that G-CSF support increase delivered dose intensity and reduced allcause mortality [19]. It is not clear how many ovarian cancer cases were included in this meta-analysis, and currently G-CSF is not approved in Australia for use in this patient group. Based on our findings, consideration should be given to conducting a trial specifically addressing the potential of growth factors to maintain RDI in women with EOC. Unfortunately, at present there are no effective interventions to prevent paclitaxel-related neurotoxicity which is a common cause for dose reductions, although some evidence suggests that serotonin-norepinephrine reuptake inhibitors (e.g. duloxetine) may help reduce associated neuropathic pain [20]. In contrast to our findings in relation to RDI, our results did not suggest that the duration of the interval between surgery and commencing chemotherapy was associated with survival. It has been suggested that surgery alters tumour biology resulting in accelerated repopulation of tumour cells [21]. Delay in commencing chemotherapy might there-
6
S. Anuradha et al. Table III. Hazard ratios for the association between time interval (TI) between surgery and initiation of chemotherapy and survival amongst women diagnosed with invasive epithelial ovarian cancer in Australia in 2005. Time interval (days)
Unadjusted HR (95% CI)
Adjusted HRa (95% CI)
1.21 (0.77–1.89) Ref 0.86 (0.54–1.37) 0.82 (0.50–1.35) 0.95 (0.51–1.79) 0.75 (0.48–1.19)
0.89 (0.55–1.43) Ref 0.95 (0.57–1.59) 0.80 (0.48–1.35) 0.93 (0.42–2.11) 1.02 (1.00–1.04)
0.98 (0.93–1.03)
1.00 (0.94–1.06)
173 36 49 30 26 11 21
0.84 (0.50–1.42) Ref 1.14 (0.68–1.92) 0.68 (0.36–1.27) 2.10 (1.03–4.29) 0.84 (0.45–1.57)
0.77 (0.44–1.35) Ref 1.06 (0.59–1.93) 0.71 (0.37–1.36) 1.30 (0.47–3.58) 0.81 (0.40–1.64)
143 11 37 24 24 12 35
1.39 (0.48–4.01) Ref 0.24 (0.05–1.10) 1.05 (0.43–2.63) 0.47 (0.10–2.13) 1.09 (0.47–2.51)
2.10 (0.68–6.51) Ref 0.49 (0.10–2.33) 0.61 (0.21–1.74) 0.62 (0.07–5.53) 2.36 (0.90–6.18)
253 75 44 38 37 15 44
0.96 (0.61–1.53) Ref 1.11 (0.70–1.78) 0.68 (0.39–1.17) 1.37 (0.71–2.64) 0.88 (0.55–1.41)
0.83 (0.50–1.35) Ref 1.02 (0.61–1.73) 0.67 (0.38–1.18) 1.11 (0.49–2.55) 1.03 (0.61–1.75)
N
All women N 351 7 8–14 15–21 22–28 29–35 35 Per week increase in TI
351 50 95 60 51 24 71
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Women with residual disease N 173 7 8–14 15–21 22–28 29–35 35 Women without residual disease N 143 7 8–14 15–21 22–28 29–35 35 Women with advanced disease N 253 7 8–14 15–21 22–28 29–35 35 a adjusted
for age, comorbidity, stage, grade, subtype, residual disease, extent of surgery, surgical complications.
fore negatively affect women’s survival. Alternatively, chemotherapy given too soon after surgery might adversely affect recovery. The optimal timing of chemotherapy is yet to be determined. Authors of a 2009 review of seven studies could not reach a definite conclusion because of the disparate nature (methods, regimens and intervals) of the studies [22]. The results of more recent studies, including only women treated with platinum/taxane-based regimens, have also been conflicting (Table IV). Two studies, including the largest (n 3326) [23], like us, found no overall association between the interval between surgery and chemotherapy and survival. However, the latter study did report slightly worse survival for each additional week of delay of chemotherapy in women without residual disease after surgery. This contrasts with findings from another study (n 191) which showed that women with residual disease who commenced chemotherapy more than
four weeks after surgery had significantly poorer survival [10]. We found no evidence that residual disease modified the association although our analysis was insufficiently powered to detect small effects. The reason for these conflicting results is not immediately clear, however, our results along with those from similar studies provide some reassurance that up until 4–5 weeks post-surgery, the timing of chemotherapy commencement has a negligible effect on survival. A strength of our study is the population-based design, which means our results are generalisable to the broader group of women diagnosed with EOC, not just those from select centres or participants in clinical trials. Further strengths are the long duration of follow-up and the availability of detailed clinical information for most women. Complete chemotherapy dose information was not available for some women because appropriate medical records were not accessible but our comparison of characteristics
Chemotherapy and survival in ovarian cancer
7
Table IV. Summary of recent studies on the association of time interval (TI) between surgery and initiation of chemotherapy with survival in epithelial ovarian cancer. Years of diagnosis
Number of women
FIGO Stage
Hofstetter, 2013 [10] (Germany/Belgium/ Austria)
2005
191
III– IV (serous subtype)
carboplatin/ taxane
28
Mahner, 2013 [23] (Germany/France)
1995– 2002 (three RCTs)
3326
IIB–IV
platinum/ paclitaxel (/- study drug)
19
Lydiksen, 2014 [24] (Denmark)
2005– 2006
650
I–IV
carboplatin/ taxane
32
Downloaded by [OARE Consortium] at 18:06 30 June 2015
Study, year (place)
Chemotherapy
across patients with and without complete chemotherapy data suggests there were no major differences. Performance status was often not recorded in clinical notes so the lack of these data, which could have influenced chemotherapy doses and timing, is another limitation of our study although we have adjusted our results for comorbidities and age. For consistent comparisons we have only included women who had primary cytoreductive surgery followed by intravenous carboplatin and paclitaxel commencing at three-weekly intervals; this was standard treatment when the women in our study were diagnosed. While this regimen is still a recommended first-line approach [25], other regimens (neoadjuvant, intraperitoneal, dose-dense) are now also recommended and increasingly used. Our findings may not be directly applicable to women receiving those regimens. Finally, our study was observational with the potential for bias inherent to all such studies. While randomisedcontrolled trials are the preferred method for assessing treatment effects, such a study to assess the effects of variation in RDI would not be feasible. In conclusion, our results suggest that, among women with advanced EOC, maintaining dose intensity during chemotherapy may help improve survival. Clinical trials to examine the use and costeffectiveness of growth factors or other options to maintain RDI are warranted. However, with respect to the time between primary surgery and commencing chemotherapy, our results provide reassurance that intervals of up to five weeks are unlikely to negatively impact upon survival. Acknowledgements We acknowledge the members of the Patterns of Care Study Group; and the research nurses and associated
Median TI (days)
Hazard ratio for TI
Adjustment factors
Age, FIGO stage, HR 1.73 (1.08– residual disease, 2.78) for 28 days extent of surgery, vs. 28 days Any and treatment residual disease centre HR 2.24 (1.08– 4.66) Overall no significant Age, FIGO stage, association No residual disease, residual disease grade and ascites HR 1.087 (1.005–1.176) per additional week TI Residual disease, HR 1.13 (0.92– FIGO stage 1.39) 32 days vs. 32 days
staff for assistance with data collection. We also acknowledge the Australian Ovarian Cancer Study Group for providing clinical information for women who participated in that study. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This work was supported by Cancer Australia (formerly National Breast and Ovarian Cancer Centre). Collection of clinical data from the Australian Ovarian Cancer Study was funded by the National Health and Medical Research Council (NHMRC) (400281, 400413); Cancer Councils of Victoria, Queensland, New South Wales, South Australia and Tasmania; and the Cancer Foundation of Western Australia. Susan Jordan and Penelope Webb are funded by fellowships from the NHMRC. This study was also supported by the RioTinto Ride to Conquer Cancer Grant through QIMR Berghofer Medical Research Institute. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. References [1] Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012;62:10–29. [2] Metzger-Filho O, Moulin C, D’Hondt V. First-line systemic treatment of ovarian cancer: A critical review of available evidence and expectations for future directions. Curr Opin Oncol 2010;22:513–20. [3] Jordan S, Steer C, DeFazio A, Quinn M, Obermair A, Friedlander M, et al. Patterns of chemotherapy treatment for women with invasive epithelial ovarian cancer – a populationbased study. Gynecol Oncol 2013;129:310–7. [4] Repetto L, Pace M, Mammoliti S, Bruzzone M, Chiara S, Oliva C, et al. The impact of received dose intensity on the outcome of advanced ovarian cancer. Eur J Cancer 1993; 29A:181–4.
Downloaded by [OARE Consortium] at 18:06 30 June 2015
8
S. Anuradha et al.
[5] Nagel CI, Backes FJ, Hade EM, Cohn DE, Eisenhauer EL, O’Malley DM, et al. Effect of chemotherapy delays and dose reductions on progression free and overall survival in the treatment of epithelial ovarian cancer. Gynecol Oncol 2012;124:221–4. [6] Hanna RK, Poniewierski MS, Laskey RA, Lopez MA, Shafer A, Van Le L, et al. Predictors of reduced relative dose intensity and its relationship to mortality in women receiving multiagent chemotherapy for epithelial ovarian cancer. Gynecol Oncol 2013;129:74–80. [7] Fauci JM, Whitworth JM, Schneider KE, Subramaniam A, Zhang B, Frederick PJ, et al. Prognostic significance of the relative dose intensity of chemotherapy in primary treatment of epithelial ovarian cancer. Gynecol Oncol 2011; 122:532–5. [8] Gadducci A, Sartori E, Landoni F, Zola P, Maggino T, Maggioni A, et al. Relationship between time interval from primary surgery to the start of taxane-plus platinum-based chemotherapy and clinical outcome of patients with advanced epithelial ovarian cancer: Results of a multicenter retrospective Italian study. J Clin Oncol 2005;23:751–8. [9] Flynn PM, Paul J, Cruickshank DJ, Scottish Gynaecological Cancer Trials G. Does the interval from primary surgery to chemotherapy influence progression-free survival in ovarian cancer? Gynecol Oncol 2002;86:354–7. [10] Hofstetter G, Concin N, Braicu I, Chekerov R, Sehouli J, Cadron I, et al. The time interval from surgery to start of chemotherapy significantly impacts prognosis in patients with advanced serous ovarian carcinoma – analysis of patient data in the prospective OVCAD study. Gynecol Oncol 2013;131:15–20. [11] Wright JD, Herzog TJ, Neugut AI, Burke WM, Lu YS, Lewin SN, et al. Effect of radical cytoreductive surgery on omission and delay of chemotherapy for advanced-stage ovarian cancer. Obstet Gynecol 2012;120:871–81. [12] Anuradha S, Webb PM, Blomfield P, Brand AH, Friedlander M, Leung Y, et al. Survival of Australian women with invasive epithelial ovarian cancer: A population-based study. Med J Aust 2014;201:283–8. [13] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373–83. [14] du Bois A, Quinn M, Thigpen T, Vermorken J, Avall-Lundqvist E, Bookman M, et al. 2004 consensus statements on the management of ovarian cancer: Final document of the 3rd International Gynecologic Cancer Intergroup
Ovarian Cancer Consensus Conference (GCIG OCCC 2004). Ann Oncol 2005;16(Suppl 8):viii7–12. [15] Loibl S, Skacel T, Nekljudova V, Luck HJ, Schwenkglenks M, Brodowicz T, et al. Evaluating the impact of relative total dose intensity (RTDI) on patients’ short and long-term outcome in taxane- and anthracycline-based chemotherapy of metastatic breast cancer – a pooled analysis. BMC Cancer 2011;11:131. [16] Merritt MA, Green AC, Nagle CM, Webb PM. Talcum powder, chronic pelvic inflammation and NSAIDs in relation to risk of epithelial ovarian cancer. Int J Cancer 2008; 122:170–6. [17] Hryniuk W, Bush H. The importance of dose intensity in chemotherapy of metastatic breast cancer. J Clin Oncol 1984;2:1281–8. [18] Lyman GH, Dale DC, Tomita D, Whittaker S, Crawford J. A retrospective evaluation of chemotherapy dose intensity and supportive care for early-stage breast cancer in a curative setting. Breast Cancer Res Treat 2013;139:863–72. [19] Lyman GH, Dale DC, Culakova E, Poniewierski MS, Wolff DA, Kuderer NM, et al. The impact of the granulocyte colony-stimulating factor on chemotherapy dose intensity and cancer survival: A systematic review and meta-analysis of randomized controlled trials. Ann Oncol 2013;24:2475–84. [20] Aziz MT, Good BL, Lowe DK. Serotonin-norepinephrine reuptake inhibitors for the management of chemotherapyinduced peripheral neuropathy. Ann Pharmacother 2014; 48:626–32. [21] Cividalli A, Creton G, Ceciarelli F, Strigari L, Tirindelli Danesi D, Benassi M. Influence of time interval between surgery and radiotherapy on tumor regrowth. J Exp Clin Canc Res 2005;24:109–16. [22] Larsen E, Blaakaer J. Epithelial ovarian cancer: Does the time interval between primary surgery and postoperative chemotherapy have any prognostic importance? Acta Obstet Gynecol Scand 2009;88:373–7. [23] Mahner S, Eulenburg C, Staehle A, Wegscheider K, Reuss A, Pujade-Lauraine E, et al. Prognostic impact of the time interval between surgery and chemotherapy in advanced ovarian cancer: Analysis of prospective randomised phase III trials. Eur J Cancer 2013;49:142–9. [24] Lydiksen L, Jensen-Fangel S, Blaakaer J. Is it possible to define an optimal time for chemotherapy after surgery for ovarian cancer? Gynecol Oncol 2014;133:454–9. [25] National Comprehensive Cancer Network Guidelines – Ovarian Cancer Version 3.2012. [cited 2014 Oct 10]. Available from: http://www.nccn.org/professionals/physician_gls/f_guidelines.asp.
