RELATION OF QUALITY OF LIFE AND RESPONSE TO …98 following second-line chemotherapy among women...
Transcript of RELATION OF QUALITY OF LIFE AND RESPONSE TO …98 following second-line chemotherapy among women...
1
Title: Quality of life and treatment response among women with platinum-resistant versus 1
platinum-sensitive ovarian cancer treated for progression: A prospective analysis. 2
3
Authors: Vanessa L Beesley1, Adele C Green2, David K Wyld3, Peter O’Rourke4, Leesa F 4
Wockner4, Anna deFazio5, Phyllis N Butow6, Melanie A Price6, Keith R Horwood7 5
Alexandra M Clavarino8, Australian Ovarian Cancer Study Group1, Australian Ovarian 6
Cancer Study–Quality of Life Study Investigators7, Penelope M Webb1 7
8
Affiliations: 9
1Gynaecological Cancers Group, QIMR Berghofer Medical Research Institute, Herston, 10
QLD, Australia 11
2Cancer and Population Studies Group, QIMR Berghofer Medical Research Institute, 12
Herston, QLD, Australia 13
3Department of Medical Oncology, Royal Brisbane and Women’s Hospital, Herston, QLD, 14
Australia 15
4Statistics Unit, QIMR Berghofer Medical Research Institute, Herston, QLD, Australia 16
5 Westmead Institute of Cancer Research, University of Sydney at Westmead Millennium 17
Institute and Westmead Hospital, Sydney, NSW, Australia 18
6Centre for Medical Psychology and Evidence-Based Decision-Making, School of 19
Psychology, The University of Sydney, Sydney, NSW, Australia 20
7Greenslopes Specialist Centre, Greenslopes Private Hospital, Greenslopes, QLD, Australia 21
8School of Pharmacy, University of Queensland, St. Lucia, QLD, Australia 22
23
24
2
Correspondence to: 25
Dr Vanessa L Beesley 26
QIMR Berghofer Medical Research Institute, 27
Gynaecological Cancers Group, 28
Locked Bag 2000, Royal Brisbane Hospital, 29
Herston, QLD 4029, Australia. 30
E-mail: [email protected] 31
Tel: +61 7 3362 0270 32
Fax: +61 7 3845 3503 33
34
35
36 37
38
3
Abstract: 39
Objective. Most women with ovarian cancer relapse and undergo further chemotherapy 40
however evidence regarding the benefits of this for women with platinum-resistant disease is 41
limited. Our objective was to determine whether there was a quality of life improvement or 42
treatment response among women treated for platinum-resistant recurrent ovarian cancer. 43
Methods. We combined data from 2 studies where women treated with chemotherapy for 44
recurrent ovarian cancer (n=172) completed a quality of life questionnaire every 3 months. 45
Cancers were classified as platinum-resistant if they progressed within 6 months of 46
completing first-line chemotherapy. Mixed effects models were used to analyse change in 47
quality of life during the first 6 months after second-line chemotherapy. 48
Results. One-quarter of women (n=44) were classified as having platinum-resistant disease. 49
Overall, their quality of life did not significantly increase or decrease, following 50
commencement of second-line chemotherapy (least square mean scores = 107, 105, 103 at 51
chemotherapy start, 3 and 6 months later, respectively), although 26% of these women 52
reported a meaningful increase and 31% reported a meaningful decline. One-third of the 53
platinum-resistant group responded (11% complete and 21% partial response) to second-line 54
chemotherapy, and this figure increased to 54% among the subset (36%) re-treated with 55
platinum-based agents with or without other agents. Preliminary analyses suggest quality of 56
life may be higher at chemotherapy initiation in women whose disease responded (median 57
score 121 vs 110). 58
Conclusions. Overall, quality of life appears to be maintained in women with platinum-59
resistant ovarian cancer who receive further chemotherapy and some women respond to re-60
treatment. 61
Keywords: ovarian cancer, oncology, progression, recurrence, chemotherapy, platinum-62
resistance, quality of life 63
4
64
Introduction 65
66
Ovarian cancer is the fifth most common cause of cancer death in women [1]. Two-thirds of 67
women diagnosed with ovarian cancer present with advanced disease and, although most 68
initially respond to treatment, the vast majority eventually experience disease progression [2]. 69
With few exceptions, recurrent ovarian cancer is not curable, yet end-points of most clinical 70
trials in this group include response to treatment, progression-free survival and overall 71
survival. In such a palliative setting, quality of life may be a more meaningful end-point. 72
73
After first-line chemotherapy for ovarian cancer, time to progression can be used to predict 74
how groups of patients might respond to further chemotherapy at relapse [3]. A cancer with a 75
long progression-free interval after completion of chemotherapy (>6 months) is associated 76
with higher response rates to re-treatment, and is classified as ‘platinum-sensitive’. 77
Management of recurrent disease in this group generally involves further platinum-based 78
chemotherapy with response rates ranging from 30-60% and a median progression-free 79
survival of 6-12 months [4, 5]. 80
81
Women whose disease progresses within 6 months of completing first-line chemotherapy are 82
defined as having ‘platinum-resistant’ disease. They have a 10-25% likelihood of responding 83
to second-line chemotherapy, a median progression-free survival of 3.5-4 months and a 84
median overall survival of 9-10 months [4, 6]. ‘Platinum-refractory’ disease refers to disease 85
that progresses during chemotherapy and the objective response rate to second-line 86
chemotherapy in these patients is <10% [4]. While a subset of women in these groups derive 87
a survival benefit from second-line chemotherapy, some women with platinum-88
5
resistant/refractory ovarian cancer continue to receive aggressive and expensive 89
chemotherapy until days before their death [7]. The goal of end of life care is to maximize 90
quality of life yet to date only 1 small study (n=9 with complete data) with a highly selected 91
patient group of women with good performance status, has considered quality of life 92
specifically among women with platinum-resistant recurrent ovarian cancer [8]. Other studies 93
which have combined women with platinum-sensitive and resistant disease, found quality of 94
life improvements after chemotherapy [9], however these pooled results are likely to be 95
driven by the high proportion of women with platinum-sensitive disease. We have used 2 96
population-based longitudinal datasets to determine if there is a quality of life improvement 97
following second-line chemotherapy among women with platinum-resistant ovarian cancer 98
and have compared this group to women with platinum-sensitive disease. We also consider 99
treatment response among these subgroups. 100
101
Methods 102
103
Study designs 104
105
We combined data from 2 longitudinal studies of women with ovarian cancer (Figure 1). 106
Ethics approval was obtained from the Human Research Ethics Committees of the QIMR 107
Berghofer Medical Research Institute, The University of Sydney and all participating 108
hospitals. Participants from both studies provided informed consent. 109
110
The first study, PROSPECT [10], recruited 74 women with newly-diagnosed and 48 with 111
recurrent disease from 7 hospitals in 2 Australian states between April 2003 and January 112
2005. Eligible patients were women aged 18 to 79 years referred for chemotherapy for 113
6
primary epithelial ovarian cancer and who were able to complete the study documents in 114
English. Participants were mailed questionnaires every 3 months for 2 years or until they 115
withdrew or died. At study entry participants were between 1 month and 20 years post-116
diagnosis (median 15 months). 117
118
The second study included 798 women who had participated in a previous population-based 119
case-control study [11-13]. The Australian Ovarian Cancer Study (AOCS) recruited women 120
aged 18–79 years with invasive epithelial ovarian cancer diagnosed between January 2002 121
and June 2006 through gynaecological oncology units and state-based cancer registries in all 122
Australian states and territories. Women who participated in AOCS and were still alive 123
between May 2005 and March 2007 were invited to take part in a quality of life substudy 124
(AOCS-QoL) [14]. Women who consented were mailed questionnaires at 3-monthly intervals 125
for up to 2 years, beginning 3 months to 5 years after diagnosis (median 26 months). 126
127
Women from the PROSPECT and AOCS-QoL studies were excluded from this analysis if 128
they did not experience a progression of their disease during data collection (n=434), or did 129
not complete a quality of life questionnaire within 2 months of starting chemotherapy for 130
their first disease progression (n=304, including 2 with platinum-resistant and 7 with 131
platinum-sensitive disease who did not have further chemotherapy) or if their disease was not 132
classifiable as platinum-resistant or –sensitive (n=7 of whom 2 were platinum-refractory). We 133
also removed duplicate data for 3 women who were in both studies leaving 172 women for 134
analysis. 135
136
7
Data collection instruments and classification 137
138
Platinum status and treatment response: We reviewed medical records for date of diagnosis, 139
start and completion dates of each chemotherapy course and cancer antigen (CA) -125 blood 140
levels at each follow-up with the treating doctor. Platinum status was classified based on 141
response to first-line treatment (CA-125 and CT scans when available) [4, 15, 16]. Women 142
who had at least a partial response to first-line platinum-based chemotherapy and no disease 143
progression within the first 6 months after completion of chemotherapy, were classified as 144
platinum-sensitive and those who had no response and/or disease progression within 6 145
months of completing chemotherapy were classified as platinum-resistant. Response to 146
chemotherapy after second-line treatment was determined based on CA-125 criteria alone. 147
That is, compared to a pretreatment CA-125, a complete response was defined as 148
normalization of CA-125 maintained for 28 days, a partial response was defined as at least a 149
50% reduction maintained for 28 days and no response was defined as no sustained reduction 150
in CA-125. 151
152
Quality of Life: This was measured using the 39-item Functional Assessment of Cancer 153
Therapy—Ovarian (FACT-O). The FACT-O is a multi-dimensional, ovarian cancer-specific 154
instrument, assessing 4 general sub-scales (physical, social, emotional and functional well-155
being) plus an ovarian cancer and treatment specific subscale. Overall quality of life is 156
derived from these 5 subscales. This instrument has undergone extensive reliability and 157
validity testing showing that internal consistency and test-retest reliability are adequate [17]. 158
159
Covariates: At baseline we collected self-reported demographic information including: date 160
of birth, marital status and education level. We also collected clinical data from medical 161
8
records including: International Federation of Gynecology and Obstetrics (FIGO) disease 162
stage at diagnosis and agents/regimen of each chemotherapy course. 163
164
Survival: Dates of disease progression and death were obtained from medical records. 165
Clinical follow-up of progression and survival status continued until September 2010 for 166
PROSPECT and July 2010 for AOCS. Survival time was calculated as the date of death, or if 167
alive, end of clinical follow-up period, minus the date of first progression. 168
169
Statistical methods 170
171
We analyzed data collected during the first 6 months after second-line chemotherapy. After 172
this there were fewer than 20 participants remaining per group. For the 14% of woman who 173
were missing data at the 3 or 6 month data-points yet remained active in the study, we 174
imputed the value as the average score of the time-points before and after the missing time-175
point. We documented reasons for dropout and calculated, at each time-point, the proportion 176
of women who dropped out due to death or incapacity (i.e. not missing completely at random) 177
by platinum status. To determine the likely effect of missingness [18] we graphed quality of 178
life scores by patterns of dropout, stratified by platinum status. To determine predictors of 179
missingness [18] we cross-tabulated the characteristics of those who dropped out versus those 180
who did not. 181
182
We used descriptive statistics, stratified by platinum status, to calculate the proportion of 183
women re-treated with different chemotherapy regimens and who had a response to 184
chemotherapy. Cox regression survival analysis was used to obtain hazard ratios and 95% 185
confidence intervals for post-progression survival by platinum status after adjustment for age. 186
9
187
A mixed effects repeated measures model with a random intercept and random slope was 188
used to estimate the least squares mean scores of quality of life and wellbeing subscales by 189
platinum status at baseline (0-2 months after the start of second-line chemotherapy) and 3 and 190
6 months later. We fitted the model using terms for platinum status, the interaction of 191
platinum status with time, and covariates including age at progression and whether a woman 192
was on chemotherapy at the time she completed the quality of life questionnaire. An 193
unstructured covariance matrix for the random effects was used as it had the best fit for this 194
dataset. Other key demographic and clinical covariates including marital status, education 195
level, second-line chemotherapy with a platinum-based drug (yes, no) and second-line 196
chemotherapy with liposomal doxorubicin (yes, no) were not included in the final models as 197
they were not significantly associated with quality of life in fully adjusted analyses. Although 198
not significant, age was retained as a standard covariate. Disease stage at diagnosis was not 199
included as almost all participants (96%) were classified as late stage. 200
201
We repeated our analyses excluding women (a) who dropped out due to death or incapacity 202
(n=31) and (b) with baseline data only (n=44), to assess the robustness of the results when 203
different data distributions were included. Furthermore, as all models have assumptions 204
about missing data we also conducted sensitivity analyses by fitting a repeated measures 205
ANOVA model with last observation carried forward and a weighted generalized estimating 206
equations model to see if the estimates were different from the mixed effects model. The 207
results across all models were similar and thus the estimates were deemed robust (data not 208
shown). 209
10
210
Clinically important changes and group differences (±) in outcomes were defined a priori as 211
follows: 8 for the overall FACT-O scale, 2 for the FACT-O wellbeing subscales and 2.5 for 212
cancer-specific concerns. These minimal differences were part way between the conservative 213
half standard deviation guideline [19] and what others have reported as clinically meaningful 214
differences in the gynecological cancer setting [20]. P-value were considered at the 215
conventional p<0.05 level. 216
217
Results 218
219
Participants 220
221
On average, women in our analysis were 60 years of age at diagnosis (SD = 10), most (71%) 222
were married or living with their partner, approximately three-quarters (74%) were classified 223
as platinum-sensitive and almost all had late stage disease at diagnosis (96%) (Table 1). 224
225
Characterizing missingness 226
227
Forty percent of women dropped out of the study due to death, incapacity, unknown reason or 228
because they had completed their commitment to the parent study (Figure 1). The proportion 229
of dropouts due to death or incapacity was lower in the platinum-sensitive group (33%) than 230
in the platinum-resistant group (75%). Irrespective of platinum status, women who dropped 231
out for any reason had lower mean quality of life at baseline than those with complete data 232
(Figure 2). Those who dropped out at the 6-month follow-up had similar absolute changes in 233
quality of life prior to that time-point to the women with complete data (Figure 2). 234
11
235
There were no systematic differences in the characteristics of the 37 women who dropped out 236
due to study completion or for unknown reasons and the 104 women with complete data 237
(Table 1). However the 31 women who dropped out due to death or incapacity were less 238
likely to be platinum-sensitive, less likely to have a partner, had lower baseline quality of life 239
and had shorter survival time compared to those with complete data (Table 1). 240
241
Second-line chemotherapy agents, responses and survival by platinum-status 242
243
Most women (87%, n=111) with platinum-sensitive disease were re-treated with a platinum-244
based regimen (Figure 3). Approximately one-third (36%, n=16) of women with platinum-245
resistant disease were re-treated with platinum with or without other chemotherapy agents, 246
while just under half (43%, n=19) were re-treated with liposomal doxorubicin (Figure 3). 247
248
Most women (84%, n=86/102 with adequate CA-125 data for evaluation) with platinum-249
sensitive disease had a CA-125 response to second line chemotherapy (complete response 250
65%, n=66; partial response 19%, n=20) compared to only 32% (n=12/38) of women with 251
platinum-resistant disease (complete response 11%, n=4; partial response 21%, n=8). Of the 252
women re-treated with platinum-based agents, 88% (n=79/90) of those who were platinum-253
sensitive and 54% (n=7/13) of those who were platinum-resistant, had a complete or partial 254
response. Many of these women also had other agents combined with platinum that they had 255
not received before. Five of the 7 women with platinum-resistant disease who had a response 256
were re-treated with platinum plus paclitaxel, docetaxel, gemcitabine or liposomal 257
doxorubicin. 258
259
12
Overall a total of 119 of the 172 women died during the clinical follow-up period; comprising 260
83 of the 128 women with platinum-sensitive disease and 36 of the 44 women with platinum-261
resistant disease. Women who had platinum-resistant disease had significantly worse post-262
progression survival than women who had platinum-sensitive disease (multivariate HR 2.34, 263
95% CI 1.57-3.49, p<0.001; median post-progression survival 13 versus 30 months). Among 264
women with platinum-resistant disease, median post-progression survival was 21 months for 265
the 16 women re-treated with platinum (and 31 months among the 7 women who responded 266
to re-treatment) versus 10 months for the 28 women not re-treated with platinum. 267
268
Quality of life changes during and after second-line chemotherapy 269
270
Among women with follow-up data and platinum-sensitive disease, 51% (n=49) had an 271
increase, 40% (n=39) had no meaningful change and 9% (n=9) experienced a decrease in 272
quality of life over the 6 months after starting second-line chemotherapy. In contrast, among 273
women with follow-up and platinum-resistant disease, 26% (n=8) had an increase, 42% 274
(n=13) had no meaningful change and 31% (n=10) reported a decrease in quality of life. 275
276
Group differences in quality of life at chemotherapy initiation 277
278
An unadjusted comparison among women with platinum-resistant disease and follow-up data 279
indicated that quality of life at chemotherapy initiation was no different in the 10 women who 280
experienced a decline in quality of life compared to the 21 who had preserved or improved 281
quality of life (median score 114 vs 112). However, among women with platinum-resistant 282
disease median quality of life at chemotherapy initiation was higher (median score 121 vs 283
110) in women whose disease responded to treatment (n=12) compared to women whose 284
13
disease progressed (n=26). 285
286
Platinum status as a predictor of quality of life during and after second-line chemotherapy 287
288
Mixed effects models revealed a clinically meaningful and statistically significant group 289
improvement in quality of life over the 6 months after starting second-line chemotherapy 290
among women with platinum-sensitive disease, whereas there was no significant or 291
precipitate decline among those with platinum-resistant disease (figure 4a). These results 292
were replicated in sensitivity analyses (figure 4b & 4c). The changes in quality of life over 293
time were statistically significantly different between the 2 groups in all models (p<0.05). 294
Current receipt of chemotherapy was the only other factor that was significantly associated 295
with (decreased) quality of life (p=0.008). Marital status, education level, platinum-based and 296
liposomal doxorubicin second-line chemotherapy agents were not significantly associated 297
with quality of life. 298
299
Within the quality of life subscales, a clinically significant but not statistically significant 300
reduction in physical wellbeing was reported over time in women with platinum-resistant 301
cancer, whereas a clinically and statistically significant increase in physical and functional 302
wellbeing and an improvement in ovarian cancer-specific symptoms was reported in women 303
with platinum-sensitive disease (Figure 5). 304
305
Discussion 306
307
This is the first population-based study to evaluate quality of life separately in women 308
classified as having platinum-sensitive and platinum-resistant disease. We have shown that, 309
14
overall, among women classified as having platinum-resistant disease, retreatment with 310
chemotherapy when they experience a recurrence does not significantly improve or diminish 311
their quality of life. However, one-quarter and one-third of this group respectively report 312
meaningful increases and decreases in quality of life. Like others [5, 6] we confirmed that for 313
women with platinum-resistant ovarian cancer, complete tumor response is rare and survival 314
is significantly poorer than for women with platinum-sensitive disease. Importantly though, a 315
subset of women with apparently platinum-resistant tumors do respond to re-treatment. 316
Notably, as reported here and elsewhere [21], the response in women with platinum-resistant 317
disease appears to be better for those re-treated with platinum-based agents ± other agents 318
than the response in those re-treated with liposomal doxorubicin or taxane alone. Larger 319
studies adjusting for characteristics of women selected to receive platinum are needed to 320
confirm this result and to determine if there is also a quality of life improvement or 321
maintenance. 322
323
In women with platinum-resistant disease maintenance of quality of life may be a useful 324
outcome, as there is a possibility that without chemotherapy quality of life would have 325
rapidly declined due to disease progression. This conclusion however is to be taken with 326
caution. We were unable to compare the quality of life trajectory of women not given second-327
line chemotherapy following disease progression since almost all women went on to receive 328
second-line chemotherapy. We also show that aside from platinum-status the only other 329
factor that was associated with improved quality of life within our overall sample was 330
cessation of chemotherapy. Furthermore, our analyses may underestimate decline in quality 331
of life among women with platinum-resistant disease as computation of missingness in the 332
mixed effects model is based on the observed data distribution and those women who 333
dropped out of the study due to incapacity were likely to have poorer quality of life. 334
15
335
Our results highlight the need for detailed investigation into which women with platinum-336
resistant ovarian cancer will and will not benefit from chemotherapy in the later stages of 337
their disease course. Our preliminary subgroup analysis gives an indication that quality of life 338
at the start of chemotherapy may be correlated with having a response to treatment. If this 339
were to be confirmed with a larger sample it is possible that quality of life screening at the 340
start of treatment may be a useful clinical practice in this particular group of patients. This is 341
important preliminary information since the decision to re-treat women with platinum-342
resistant disease may come at considerable cost, both in terms of increased patient morbidity, 343
and to the health system in terms of expenditure which is estimated at USD $20,744 on 344
average per woman after first-line chemotherapy [22]. 345
346
In conclusion, this study raises important ethical and economic questions about the re-347
treatment of women with platinum-resistant ovarian cancer with chemotherapy. Disease 348
response rates are particularly low in those not treated with platinum and approximately one 349
third of women in the platinum-resistant group when treated with second-line chemotherapy 350
have significant declines in quality of life. Further research in a larger series of patients with 351
platinum-resistant disease should be conducted to identify the groups most likely to benefit 352
from re-treatment with chemotherapy. Given that the majority of women with ovarian cancer 353
will have a progression and that approximately one-quarter will be classified as having 354
platinum-resistant disease, evaluation of both the costs and benefits of treating this disease is 355
essential for efficient use of limited health care resources and optimal patient-centered care 356
[23]. 357
358
16
Acknowledgments 359
Data for this analysis came from (1) the PROSPECT study which was funded by Cancer 360
Council Queensland, (2) The Australian Ovarian Cancer Study which was supported by the 361
US Army Medical Research and Materiel Command under DAMD17-01-1-0729, The Cancer 362
Councils of New South Wales, Queensland, South Australia, Tasmania and Victoria, The 363
Cancer Foundation of Western Australia and the National Health and Medical Research 364
Council (NHMRC) of Australia (199600 and 400413) and (3) The AOCS-QoL study which 365
was funded by the Cancer Councils of New South Wales and Queensland (RG 36/05). 366
367
Vanessa Beesley was funded by an NHMRC early career fellowship, Phyllis Butow and 368
Penelope Webb by NHMRC senior research fellowships, Melanie Price by a post-doctoral 369
fellowship from the School of Psychology, The University of Sydney and Alexandra 370
Clavarino by NHMRC Capacity Building Grant. 371
372
We thank Lynley Aldridge, Diana Grivas, Naomi McGowan, Ann Ward, Colleen Loos, 373
Rebekah Cicero and Alessandra Francesconi for their help with the data tracking, 374
classification and management for the PROSPECT and AOCS-QoL studies. We also 375
acknowledge all participating institutions and investigators represented within the AOCS 376
group listed at www.aocstudy.org. 377
378
Conflicts of interests 379
A. deFazio holds a patent unrelated to this manuscript and has received a Speaker's 380
Honorarium from Roche. All remaining authors have declared no conflicts of interest. 381
382
383
17
References 384
[1] Australian Institute of Health and Welfare, National Breast and Ovarian Cancer 385
Centre. Ovarian Cancer in Australia: an overview, 2010. In. Canberra: Australian Institute of 386
Health and Welfare & National Breast and Ovarian Cancer Centre; 2010. 387
[2] Lockwood-Rayermann S. Survivorship issues in ovarian cancer: a review. Oncol Nurs 388
Forum; 2006;33: 553-562. 389
[3] Eisenhauer EA, Vermorken JB, van Glabbeke M. Predictors of response to 390
subsequent chemotherapy in platinum pretreated ovarian cancer: a multivariate analysis of 391
704 patients. Ann Oncol; 1997;8: 963-968. 392
[4] Friedlander M, Butow P, Stockler M, Gainford C, Martyn J, Oza A, Donovan HS, 393
Miller B, King M. Symptom control in patients with recurrent ovarian cancer: measuring the 394
benefit of palliative chemotherapy in women with platinum refractory/resistant ovarian 395
cancer. Int J Gynecol Cancer; 2009;19 Suppl 2: S44-S48. 396
[5] Colombo N, Gore M. Treatment of recurrent ovarian cancer relapsing 6-12 months 397
post platinum-based chemotherapy. Crit Rev Oncol Hematol; 2007;64: 129-138. 398
[6] Blackledge G, Lawton F, Redman C, Kelly K. Response of patients in phase II studies 399
of chemotherapy in ovarian cancer: implications for patient treatment and the design of phase 400
II trials. Br J Cancer; 1989;59: 650-653. 401
[7] Emanuel EJ, Young-Xu Y, Levinsky NG, Gazelle G, Saynina O, Ash AS. 402
Chemotherapy use among Medicare beneficiaries at the end of life. Ann Intern Med; 403
2003;138: 639-643. 404
[8] Goff BA, Thompson T, Greer BE, Jacobs A, Storer B. Treatment of recurrent 405
platinum resistant ovarian or peritoneal cancer with gemcitabine and doxorubicin: A phase 406
I/II trial of the Puget Sound Oncology Consortium (PSOC 1602). Am J Obstet Gynecol; 407
2003;188: 1556-1562. 408
18
[9] Gordon AN, Fleagle JT, Guthrie D, Parkin DE, Gore ME, Lacave AJ. Recurrent 409
epithelial ovarian carcinoma: a randomized phase III study of pegylated liposomal 410
doxorubicin versus topotecan. J Clin Oncol; 2001;19: 3312-3322. 411
[10] Beesley V, Clavarino A, Webb P, Wyld D, Francesconi A, Horwood K, Doecke J, 412
Loos C, Green A. Ranked importance of outcomes of first-line and repeated chemotherapy 413
among ovarian cancer patients. Support Care Cancer; 2010;18: 943-949. 414
[11] Merritt MA, Green AlC, Nagle CM, Webb PM. Talcum powder, chronic pelvic 415
inflammation and NSAIDs in relation to risk of epithelial ovarian cancer. Int J Cancer; 416
2008;122: 170-176. 417
[12] Tothill RW, Tinker AV, George J, Brown R, Fox SB, Lade S, Johnson DS, Trivett 418
MK, Etemadmoghadam D, Locandro B, Traficante N, Fereday S, Hung JA, Chiew Y-E, 419
Haviv I, Gertig D, DeFazio A, Bowtell DDL. Novel molecular subtypes of serous and 420
endometrioid ovarian cancer linked to clinical outcome. Clin Cancer Res; 2008;14: 5198-421
5208. 422
[13] Jordan SJ, Green AlC, Whiteman DC, Moore SP, Bain CJ, Gertig DM, Webb PM. 423
Serous ovarian, fallopian tube and primary peritoneal cancers: a comparative epidemiological 424
analysis. Int J Cancer; 2008;122: 1598-1603. 425
[14] Beesley V, Price M, Butow P, Green A, Olsen C, Webb P, Australian Ovarian Cancer 426
Study, AOCS-QoL Study Investigators. Physical activity in women with ovarian cancer and 427
its association with decreased distress and improved quality of life. Psychooncology; 428
2011;20: 1161-9. 429
[15] Pectasides D, Psyrri A, Pectasides M, Economopoulos T. Optimal therapy for 430
platinum-resistant recurrent ovarian cancer: doxorubicin, gemcitabine or topotecan? Expert 431
Opin Pharmacother; 2006;7: 975-987. 432
19
[16] Rocconi RP, Case AS, Straughn JM, Jr., Estes JM, Partridge EE. Role of 433
chemotherapy for patients with recurrent platinum-resistant advanced epithelial ovarian 434
cancer: A cost-effectiveness analysis. Cancer; 2006;107: 536-543. 435
[17] Basen-Engquist K, Bodurka-Bevers D, Fitzgerald M, Webster K, Cella D, Hu S, 436
Gershenson D. Reliability and validity of the Functional Assessment of Cancer Therapy-437
Ovarian. J Clin Oncol; 2001;19: 1809-1817. 438
[18] Ibrahim JG, Molenberghs G. Missing data methods in longitudinal studies: a review. 439
Test (Madrid, Spain); 2009;18: 1-43. 440
[19] Sloan JA, Cella D, Hays RD. Clinical significance of patient-reported questionnaire 441
data: another step toward consensus. J Clin Epidemiol; 2005;58: 1217-1219. 442
[20] Courneya KS, Karvinen KH, Campbell KL, Pearcey RG, Dundas G, Capstick V, 443
Tonkin KS. Associations among exercise, body weight, and quality of life in a population-444
based sample of endometrial cancer survivors. Gynecol Oncol; 2005;97: 422-430. 445
[21] Alsop K, Fereday S, Meldrum C, deFazio A, Emmanuel C, George J, Dobrovic A, 446
Birrer M, Webb P, Stewart C, Friedlander M, Fox S, Bowtell D, Mitchell G. BRCA mutation 447
frequency and patterns of treatment response in BRCA mutation-positive women with 448
ovarian cancer: a report from the Australian Ovarian Cancer Study Group. J Clin Oncol; 449
2012;30: 2654-2663. 450
[22] Gordon L, Scuffham P, Beesley V, Green A, DeFazio A, Wyld D, Clavarino A, 451
Australian Ovarian Cancer Study, Webb P. Outcomes and related medical costs for women 452
with ovarian cancer in Australia: a patient-level analysis over 2½ years. Int J Gynecol 453
Cancer; 2010;20: 757-65. 454
[23] Shih Y-CT, Halpern MT. Economic evaluations of medical care interventions for 455
cancer patients: how, why, and what does it mean? CA Cancer J Clin; 2008;58: 231-244. 456
457
20
Tables and Figures 458
Table 1: Demographic and clinical characteristics of all participants and by missingness type 459
Figure 1. Flow of participant recruitment and data contribution to this analysis 460
Figure 2. Mean quality of life score by dropout time stratified by platinum status 461
Figure 3. Chemotherapy regimen administered after first progression, by platinum status 462
Figure 4a. Least squares mean quality of life scores following second-line chemotherapy 463
in women with ovarian cancer treated for platinum-sensitive (n=128) versus -resistant 464
(n=44) disease. Main model: including all available data. 465
Figure 4b. Least squares mean quality of life scores following second-line chemotherapy in 466
women with ovarian cancer treated for platinum-sensitive (n=112) versus -resistant (n=29) 467
disease. Sensitivity model: not including women who dropped out due to death or incapacity. 468
Figure 4c. Least squares mean quality of life scores following second-line chemotherapy in 469
women with ovarian cancer treated for platinum-sensitive (n=97) versus -resistant (n=31) 470
disease. Sensitivity model: not including women with baseline only data. 471
Figure 5. Wellbeing subscale breakdown of overall quality of life presented in figure 4a. 472 473
21
Table 1: Demographic and clinical characteristics of all participants and by missingness 474 type 475 476 All
participants (n=172)
Participants with complete data (n=104)
Dropout due to death or incapacity
(n=31)
Dropout due to completed
study or unknown
reason (n=37) Platinum sensitive, % 74 77 52* 86
Age at progression, mean (SD)
60 (10) 60 (9) 62 (10) 58 (12)
Had partner, % 71 74 53* 76
Stage III or IV at diagnosis, %
96 96 97 95
QoL at second-line chemotherapy initiation, mean (SD)
105 (22) 108 (22) 96 (22)* 105 (21)
Survival from second-line chemotherapy initiation (months), median
26 30 9 * 29
* Different from all participants 477 478
22
479 480
481
482 483 484 485 486 487 488 489 490 491 492 493
494
495
496
497
498
499
500
501
Figure 1. Flow of participant recruitment and data contribution to this analysis 502 a Completed follow-up for parent study 503
PROSPECT study April 2003 and January 2005 N = 122
PROSPECT study April 2003 and January 2005
N = 122
AOCS-QOL study May 2005 and March 2007
N = 798
Completed questionnaire within 2 months of starting chemotherapy for first progression
N = 52
Completed questionnaire within 2 months of starting chemotherapy for first progression
N = 130
Platinum status was classifiable i.e. data used in this analysis N = 172 Platinum-sensitive (PS) N = 128, 74%
Platinum-resistant (PR) N = 44, 26%
3 month follow-up data = 128 (74%) • PS withdrawals: death = 7, incapacity = 4, completed follow-upa = 7, unknown = 13. • PR withdrawals: death = 4, incapacity = 5, unknown = 4.
6 month follow-up data = 104 (60%) • PS withdrawals: death = 2, incapacity = 3, completed follow-upa = 4, unknown = 8. • PR withdrawals: death = 4, incapacity = 2, unknown = 1.
23
Platinum-sensitive Platinum-resistant 504
505 Figure 2. Mean quality of life score by dropout time stratified by platinum status 506
95
100
105
110
115
120
2nd-linechemotherapy
start
3 months 6 months
Qua
lity
of li
fe sc
ale
(FAC
T-0)
rang
ing
from
0-1
48
Complete data n=80
6 month dropout n=17
3 month dropout n=31
95
100
105
110
115
120
2nd-linechemotherapy
start
3 months 6 months
Complete data n=24
6 month dropout n=7
3 month dropout n=13
24
507
508
509
510
511
512
513
514
515
516
517
518
519
Figure 3. Chemotherapy regimen administered after first progression, by platinum 520
status 521
45%
26%
16%
10%
1% 2%
7%
18%
11%
43%
11% 9%
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
Plat
in a
ndta
xane
age
nts
Plat
in &
oth
erag
ent(
s)
Plat
in o
nly
Lipo
som
aldo
xoru
bici
n
Taxa
ne o
nly
Oth
er
% re
ceiv
ed re
gim
en
Platinum-sensitive Platinum-resistant
25
Figure 4a. Least squares mean quality of life scores following second-line chemotherapy in women with ovarian cancer treated for platinum-sensitive (n=128) versus -resistant (n=44) disease. Main model: including all available data.
Figure 4b. Least squares mean quality of life scores following second-line chemotherapy in women with ovarian cancer treated for platinum-sensitive (n=112) versus -resistant (n=29) disease. Sensitivity model: not including women who dropped out due to death or incapacity. Figure 4c. Least squares mean quality of life scores following second-line chemotherapy in women with ovarian cancer treated for platinum-sensitive (n=97) versus -resistant (n=31) disease. Sensitivity model: not including women with baseline only data. Note: All models adjusted for age at progression and whether currently on chemotherapy. Marital status, education, second-line chemotherapy with platinum or liposomal doxorubicin were tested and not included as covariates due to their lack of association. Note: A difference of 8 on the FACT-O scale is considered clinically significant
105
111
115
107
105 103
90
100
110
120
2nd-line chemotherapystart
3 months 6 monthsQua
lity
of li
fe ra
ngin
g fr
om 0
-148
107
113 116
112
111
108
90
100
110
120
2nd-line chemotherapystart
3 months 6 monthsQua
lity
of li
fe ra
ngin
g fr
om 0
-148
106
112 115
109
107
104
90
100
110
120
2nd-line chemotherapystart
3 months 6 monthsQua
lity
of li
fe ra
ngin
g fr
om 0
-148
platinum-sensitive platinum-resistant
Change in platinum-sensitive group p<0.001 Change in platinum-resistant group p=0.229
Change in platinum-sensitive group p<0.001 Change in platinum-resistant group p= 0.260
Change in platinum-sensitive group p<0.001 Change in platinum-resistant group p= 0.167
26
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 Figure 5. Wellbeing subscale breakdown of overall quality of life presented in figure 4a. 37 38 Note: A difference of 2 on the wellbeing scales and 2.5 on the ovarian cancer-specific concerns scale is considered to be a 39 clinically significant least squares mean difference. 40 41 Statistical support (p<0.05) for changes were noted in physical, functional, emotional wellbeing and ovarian cancer-specific 42 concerns among women who had platinum-sensitive disease. 43 44
18.7
20.1
22.4
20.5
19.1 18.5
15
17
19
21
23
25
2nd-linechemotherapy
start
3 months 6 months
Phys
ical
wel
lbei
ng 0
-28
22.1 22.5 22.3
22.1 22.4 22.2
15
17
19
21
23
25
2nd-linechemotherapy
start
3 months 6 monthsSoci
al/f
amily
wel
lbei
ng 0
-28
17.2 18.2 18.3
16.6 17.2 17.7 15
17
19
21
23
25
2nd-linechemotherapy
start
3 months 6 months
Emot
iona
l wel
lbei
ng 0
-24
17.6
19.1 19.6 18.8
18.2 17.4 15
17
19
21
23
25
2nd-linechemotherapy
start
3 months 6 months
Func
tiona
l wel
lbei
ng 0
-28
29.5 31.1 32.1
28.7 28.1 27.5
22
25
28
31
34
37
2nd-linechemotherapy
start
3 months 6 months
Ova
rian
canc
er-s
peci
fic c
once
rns
0-40
platinum-sensitive platinum-resistant