Although most patients with ovarian cancer initially respond to treatment, the majority will relapse and be offered chemotherapy. Patients who relapse within 6 months of completing the first line of chemotherapy are classified as “platinum resistant”; they have a median survival of 9 to 12 months and a 10% to 15% likelihood of responding to chemotherapy.1–5 Patients whose tumors progress while on treatment are classified as “platinum refractory” and have a poor prognosis (objective response rates to chemotherapy of <10%; median survival of 3–5 months).6
Randomized controlled trials of chemotherapy conducted in the platinum resistant/refractory population have focused on progression-free survival and overall survival as primary outcomes.7 However, some have also included health-related quality of life (HRQoL) as a secondary outcome using either the EORTC QLQ-C30 or the FACT-O/FOSI.8–11 Although well-validated and widely used measures of HRQoL, these instruments provide limited information about the symptom benefit of palliative chemotherapy, for reasons explored in the companion article. Participants in clinical trials are highly selected and may not represent the broader population of patients treated in clinical practice.12 Thus, the symptom burden of patients with platinum resistant/refractory recurrent ovarian cancer (PRROC) is not well documented, and the extent to which symptoms improve with palliative chemotherapy is unknown.
The 3rd Ovarian Cancer Consensus meeting concluded that objective response rates and progression-free survival alone were inadequate end points of treatment in patients with PRROC and recommended the development of an instrument to measure symptom benefit.13 This led to the establishment of the Symptom Benefit Working Group of the Gynecologic Cancer Intergroup (GCIG).
The GCIG Symptom Benefit Study (clinical trial register, ACTRN12607000603415) was designed in 2 stages. Stage 1 had 2 complementary aims: (1) to provide a comprehensive description of issues pertinent to the symptom burden of women with PRROC and (2) to determine the optimal instrument to measure subjective symptom benefit in clinical trials of palliative chemotherapy for recurrent ovarian cancer. The first aim is addressed in the current article; the second is addressed in the companion article (King et al).14 Specific objectives for the first aim were to document (1) the symptoms rated most severe by patients before chemotherapy; (2) the reasons for treatment, that is, whether for symptom control or due to elevated CA125 and/or radiological evidence of recurrence; (3) physician prediction of treatment duration; (4) the reasons for ceasing treatment; (5) patient expected versus perceived benefit from treatment; and (6) clinical benefit and objective response. We also aimed to compare patients’ with physicians’ reports of symptoms at baseline and toxicity during treatment.
Participants from 18 centers in Australia and Canada, aged 18 years or older, had been diagnosed with epithelial ovarian, primary peritoneal, or fallopian tube cancers and had recurrent cancer based on CA125, radiological, or clinical criteria. Most patients had PRROC but were also eligible providing they were receiving at least their third line of chemotherapy. All patients were required an ECOG PS 0 to 3 and a life expectancy of more than 3 months. They had to commence treatment within 2 weeks of registration and be able to complete all the questionnaires independently.
All patients were required to sign an informed consent. The protocol was approved by ethics committees at all the institutions involved in the study.
Physician Assessment of Patients
Physicians assessed participants before each of the first 4 cycles of chemotherapy, including ratings with version 3 of the NCI Common Terminology Criteria for Adverse Events (CTCAE),15 which includes symptoms as well as toxicities attributed to previous therapy. Objective tumor response was assessed after cycles 2 and 4 according to standard practice at the treating institution. Clinical benefit included CA125 response/RECIST response and/or symptomatic improvement as assessed by treating physician.
Patients self-completed a HRQoL booklet at baseline (within 2 weeks before the first cycle), before each cycle of chemotherapy (on day 1 of each cycle, for up to 4 cycles) and 4 weeks after the fourth cycle of treatment or at the end of treatment, whichever came first. The HRQoL booklet included 7 validated questionnaires: the Symptom Representation Questionnaire (SRQ),16 FACT-O,17 QLQ-C30,18 QLQ-OV28,19 the Patient Disease and Treatment Assessment Form (Pt DATA Form),20 the Hospital Anxiety and Depression Scale (HADS),21 and the Herth Hope Index,22 plus study-specific questions about the expected and perceived benefits of chemotherapy. This article focuses on symptoms, using data from the SRQ, QLQ-C30, QLQ-OV28, HADS, the Pt DATA Form, and patients’ perceptions of the benefits of chemotherapy. The SRQ, QLQ-C30, and QLQ-OV28 are described in detail in the companion article; the others are described later.
The Hospital and Anxiety Depression Scale
The HADS has 14 items, 7 in each of 2 subscales independently measuring anxiety and depression, with a recall period of 1 week.21 Scores for each subscale range from 0 to 21, with higher scores indicating greater levels of anxiety or depression. Scores greater than 10 indicate clinical levels of anxiety and depression.23
The Patient DATA Form
The Pt DATA Form has 47 items with a recall period of 1 week.20 Seventeen symptoms and 24 additional treatment-related concerns are rated on a numeric scale from 0 (“no trouble at all”) to 10 (“worst I can imagine”), anchored with the words “mild” at 2, “moderate” at 5, and “severe” at 8. Six aspects of well-being are rated on a scale from 0 (“worst possible”) to 10 (“best possible”).
Expected and Perceived Symptom Benefits
At baseline, patients were asked “How much do you expect your symptoms to improve with chemotherapy?” using a numeric rating scale from 0 (“none at all”) to 10 (“completely”). After completing 2 cycles of chemotherapy, before cycle 3 and at end of treatment, patients were asked “How much have your symptoms improved with chemotherapy?” using the same scale. If patients indicated an improvement in symptoms, they completed an additional 5-point scale asking whether their symptom improvement was enough to affect their overall quality of life (ranging from “not at all” to “very much so”).
The analyses are descriptive and exploratory rather than hypothesis testing. Most of the study objectives were addressed by expressing data as frequencies (percentages) for categorical variables and as means (SD) or medians (interquartile range) for normally and nonnormally distributed continuous variables, respectively. Spearman correlation coefficients (r) were calculated as a simple exploratory statistic to study the magnitude of agreement between on-study patient-reported symptoms and physician-reported adverse events. For key symptoms measured by both the Pt DATA Form and the CTCAE criteria, the worst Pt DATA Form score across all on-study assessments was identified, and the corresponding CTCAE grade formed the paired observation for each patient (r ranges from −1 to 1; values < .30 indicate a weak relationship, values from .30 to .50 indicate a moderate relationship, and values > .50 indicate a strong relationship).24 A threshold of 10 points was used to categorize patients into 3 groups (improved, stable, and deteriorated) based on change in score from baseline in the EORTC Global Health Status/QoL score and abdominal symptom score.25 The paired t test was used to compare expected versus actual treatment duration. All statistical analyses were conducted using SAS software, version 9.2 (SAS Institute). Two-sided Ps < 0.05 were considered statistically significant.
Demographics, Reasons for Treatment, and Questionnaire Compliance
One hundred twenty-three of the one hundred twenty-six participants had at least 1 cycle of chemotherapy. Their mean age was 62 years (range, 30–89 years). Most patients had an ECOG performance status of 0 or 1 (Table 1). Most patients had PRROC. Most patients had received at least 2 lines of chemotherapy before this study, with 38% having received 3 or more lines. For 75% of patients, symptom control was documented as the main reason for treatment. Most patients also had elevated CA125 (89%) and radiological evidence of recurrence (84%), which were also documented as reasons for treatment.
Compliance with questionnaire completion was excellent; with most of the participants still on study completing HRQoL assessments for the first 3 cycles (99% at baseline, 97% at cycle 1, 91% at cycle 2, and 93% at cycle 3), with a drop-off to 85% at cycle 4.
Baseline Symptoms Reported by Patients and Physicians
Based on the SRQ, only 1 patient reported being symptom free at baseline. Fifty-one percent of the patients had 7 or more symptoms that they rated 3 or higher on the SRQ’s 11-point scale (0, “did not have the symptom,” to 10, “as bad as I can imagine”), and 57% had 3 or more symptoms that they rated 5 or higher. The symptoms nominated most often by the patients as in their “top 3 symptoms” on the SRQ are shown in Table 2. The most commonly nominated symptoms were pain (44%), fatigue (37%), and abdominal bloating (32%). Almost half of the patients (45%) reported a constellation of gastrointestinal symptoms including nausea and vomiting, bowel dysfunction, and anorexia. Emotional distress and insomnia were also identified as important symptoms by patients. The HADS scores indicated that 26% of patients had clinical anxiety and 11% had clinical depression. Patients also reported a number of symptoms at baseline that most likely were related to previous surgery and chemotherapy, such as hot flashes and sensory neuropathy.
The symptoms and adverse events documented at baseline by the patient’s managing physician using the NCI CTCAE v3.0 are also summarized in Table 2. Forty-one patients were rated by physicians as having no adverse effects associated with previous therapy at baseline and no symptoms. However, all these patients reported symptoms with many patients scoring symptoms by 5 or higher on the SRQ (Table 4).
Patient Expectations Versus Perceptions of Symptom Benefit
Most patients expected that their symptoms would improve with treatment, with 91 (73%) scoring 6 or higher for expected improvement (where 0, “not at all,” and 10, “completely”). Twenty-four percent of the patients expected that their symptoms would completely or almost completely resolve. Of the 73 participants assessed before cycle 3, only one reported complete resolution of her symptoms, but 59% reported that their symptoms had improved by 5 or higher (where 0, “not at all,” and 10, “completely”).
Expected Versus Actual Duration of Treatment and Reasons for Ceasing
The median predicted duration of chemotherapy was 18 weeks, whereas the median actual duration was 12 weeks (P ≤ 0.001). Only 41% of the patients completed the expected number of cycles. Most patients stopped treatment early, for reasons including progression (33%), death (7%), adverse effects (6%), and physician or patient preference (6%). Patients received a median number of 3 cycles of chemotherapy, with 36% receiving 2 or more cycles and 17% receiving 5 or 6 cycles.
Clinical Benefit and Response Rates
Treating physicians recorded patients’ clinical benefit at the end of cycles 2 and 4 of chemotherapy. Criteria used to designate clinical benefit included clinical response, CA125 response, and RECIST response. The CA125 was the most commonly used criterion; it was used to assess response in 75% and 79% of the patients after cycles 2 and 4, respectively. Twenty-four percent of the participants had radiological investigations to assess response after cycle 2; and 32%, after cycle 4. Physicians reported that 40% of patients had a clinical benefit after their first 2 cycles of chemotherapy and that they planned on continuing the same chemotherapy in 96% of patients with clinical benefit. Four of the 47 patients (8.5%) who had radiological investigations had a RECIST response including 1 complete response and 3 partial responses.
Patient-Reported Symptom Benefit and Impact on HRQoL
Of the patients who were symptomatic at baseline (>10 on the QLQ-OV28 abdominal/gastrointestinal symptom scale) (ie, at least 10 points on a 0–100 scale), 36%, 44%, 48%, and 48% reported a clinically important improvement after cycles 1, 2, 3, and 4, respectively. The increasing proportion benefiting from treatment reflects that patients with stable disease or clinical benefit continued to receive chemotherapy, whereas patients who were not responding to chemotherapy came off treatment; the number receiving chemotherapy declined from 108 (before cycle 2) to 60 (before cycle 4). Similar trends were seen for change in QLQ Global Health Status/QoL, although the proportion improving at each cycle was smaller (data not shown).
For patients ceasing treatment after 1, 2, or 3 cycles, the number of moderate symptoms (SRQ > 5) seemed to increase before stopping treatment, reflecting the clinical decision to stop chemotherapy in patients who were not obtaining symptomatic benefit. In contrast, in patients who completed all 4 cycles and did not progress, the number of “moderate” symptoms decreased over time, reflecting a cumulative symptomatic benefit across the course of chemotherapy for a subgroup who responded to treatment.
Toxicity of Treatment
The adverse events of all grades reported by physicians are shown in Table 3. Thirty-seven percent of the patients had unplanned admissions to hospital either due to side effects of treatment or due to progression of disease. There were differences between patient- and physician-reported adverse effects of treatment. The correlation between the worst on study patient response to key symptoms on the Pt DATA Form and the corresponding physician-graded adverse event at the same visit was studied. A strong association between patient and physician ratings for vomiting was demonstrated (r = 0.58, P < 0.001). Moderate associations were demonstrated for diarrhea (r = 0.49, P < 0.001); neuropathy (sensory) (r = 0.40, P < 0.001) (Fig. 1); appetite (r = −0.39); nausea, constipation, and fatigue (r = 0.3 to <0.4, P ≤ 0.001). Although these correlations are statistically moderate to strong, they are not sufficiently robust to equate physician and patient reporting and highlight the need for rigorous patient-reported assessment.
Table 4 shows the baseline symptoms nominated most often in the top 3 symptoms in SRQ in the 41 patients reported by physicians to have no symptoms at baseline. There was discordance between patients and physicians in the reporting of symptoms, with all patients reporting a range of symptoms, many of which would be considered clinically significant using a cut point of 5 or greater on the SRQ.
The aim of treatment in patients with PRROC is palliation, but most clinical trials have not addressed symptom burden or impact of treatment on symptom benefit. Contrary to both physician and patient expectations, many patients progressed rapidly and did not seem to benefit from chemotherapy. One third of patients received less than 2 cycles of chemotherapy due to rapid progression or early death. It is worth noting that 89% of the patients were documented as having an ECOG performance status of 0 to 1 at study entry, which suggests that performance status was lower than reported. Both physicians and patients were overly optimistic about the benefits of chemotherapy, consistent with other studies of palliative chemotherapy.26 We have described the real-world experience of patients receiving palliative chemotherapy. These patients may be quite different to the selected population recruited into clinical trials. Our findings are a sober reminder of the need to focus on patient-reported outcomes in clinical trials as well as in the clinic.
We have described the complexity of the symptom array experienced by patients. It is important to identify cancer-related symptoms at baseline and their relative importance to the patient to judge the benefits of palliative chemotherapy. Careful assessment of all symptoms is also important as it should be possible to treat many of these symptoms (such as pain and depression).
There was an apparent discordance between patient reporting of symptoms at baseline and physician reporting of symptoms and adverse events, in keeping with other studies,27–29 although it should be noted that different forms were filled out by patients and physicians in the study. There were significant differences in the reporting of pain, abdominal bloating, dyspnea, insomnia, anorexia, and emotional problems. For example, mood disturbance was documented in only 4% of the patients by physicians in the baseline forms, whereas the HADS indicated that 26% of the patients had clinical anxiety and 11% had clinical depression. The HADS is a very well-regarded patient reported outcome (PRO) with proven validity.30 Physicians were not specifically asked whether the patient had a particular symptom but were asked to document what they thought the patient’s symptoms and adverse events were. This may explain some of the discrepancy in reporting mood disturbance, but it is still disconcerting. Insomnia is also an important issue to many patients, and this was not well documented by physicians, with only 4% of the patients reported to have insomnia.
The findings of this study illustrate the importance of asking patients to rate and grade their symptoms and side effects. It was notable that 41 patients were documented by physicians as not having any significant baseline symptoms, although the stated aim of treatment was palliation in most. Furthermore, all these patients reported symptoms, and this discordance is consistent with reports in other tumor types.31 It is noteworthy that, although response rates were low, physicians judged that 40% of the patients had a clinical benefit, which is similar to patient-reported symptom benefit. The results of this study illustrate the importance of developing a patient-reported, symptom-focused instrument to assess symptom benefit in patients with recurrent ovarian cancer who are receiving palliative chemotherapy. The instrument should also include side effects so that toxicity of treatment (as experienced by patients) can be documented. By providing a comprehensive description of the costs and benefits of palliative chemotherapy from the patients’ perspective, such a tool would have value as an outcome measure in clinical trials and as an aid to clinical decision.
Stage 1 of the GCIG Symptom Benefit Study has provided an insight into the complex array of symptoms experienced by patients with predominantly PRROC and the adverse effects associated with palliative chemotherapy, and it is a reality check. It illustrates the poor prognosis of this group of patients and the difficulty in predicting which patients are likely to benefit from palliative chemotherapy. The study also highlights differences and discordance between the reporting of symptoms and side effects by patients and their physicians. The development and validation of an instrument to measure symptom benefit with chemotherapy in clinical trials are essential, and this is the aim of stage 2 of the GCIG Symptom Benefit Study, which is open in 12 countries.
We would like to thank all the patients, physicians, and nurses who participated in the study as well as the support from the GCIG including Dr. Brigitte Miller and Monica Bacon.
1. Herzog TJ, Pothuri B. Ovarian cancer: a focus on management of recurrent disease. Nat Clin Pract Oncol. 2006; 3: 604–611.
2. Rustin GJ, Nelstrop AE, Tuxen MK, et al. Defining progression of ovarian carcinoma during follow-up according to CA 125: a North Thames Ovary Group Study. Ann Oncol. 1996; 7: 361–364.
3. Blackledge G, Lawton F, Redman C, et al. Response of patients in phase II studies of chemotherapy in ovarian cancer: implications for patient treatment and the design of phase II trials. Br J Cancer. 1989; 59: 650–653.
4. Markman M, Rothman R, Hakes T, et al. Second-line platinum therapy in patients with ovarian cancer previously treated with cisplatin. J Clin Oncol. 1991; 9: 389–393.
5. Gordon AN, Fleagle JT, Guthrie D, et al. Recurrent epithelial ovarian carcinoma: a randomized phase III study of pegylated liposomal doxorubicin versus topotecan. J Clin Oncol. 2001; 19: 3312–3322.
6. Fu S, Hennessy BT, Ng CS, et al. Perifosine plus docetaxel in patients with platinum and taxane resistant or refractory high-grade epithelial ovarian cancer. Gynecol Oncol. 2012; 126: 47–53.
7. Spriggs DR. Drug development for chronic cancers: time to think differently? J Clin Oncol. 2012; 30: 3779–3780.
8. Gordon AN, Asmar L, Messing MJ, et al. Phase II study of sequential doublets: topotecan and carboplatin, followed by paclitaxel and carboplatin, in patients with newly diagnosed advanced ovarian cancer. Gynecol Oncol. 2004; 94: 533–539.
9. Lortholary A, Largillier R, Weber B, et al. Weekly paclitaxel as a single agent or in combination with carboplatin or weekly topotecan in patients with resistant ovarian cancer: the CARTAXHY randomized phase II trial from Groupe d’Investigateurs Nationaux pour l’Etude des Cancers Ovariens (GINECO). Ann Oncol. 2012; 23: 346–352.
10. Mutch DG, Orlando M, Goss T, et al. Randomized phase III trial of gemcitabine compared with pegylated liposomal doxorubicin in patients with platinum-resistant ovarian cancer. J Clin Oncol. 2007; 25: 2811–2818.
11. Colombo N, Kutarska E, Dimopoulos M, et al. Randomized, open-label, phase III study comparing patupilone (EPO906) with pegylated liposomal doxorubicin in platinum-refractory or -resistant patients with recurrent epithelial ovarian, primary fallopian tube, or primary peritoneal cancer. J Clin Oncol. 2012; 30: 3841–3847.
12. Rothwell PM. External validity of randomised controlled trials: “to whom do the results of this trial apply?”. Lancet. 2005; 365: 82–93.
13. du Bois A, Quinn M, Thigpen T, 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–viii12.
14. King MT, Stockler MR, Butow P, et al. Development of the Measure of Ovarian Symptoms and Treatment Concerns: Aiming for Optimal Measurement of Patient-Reported Symptom Benefit With Chemotherapy for Symptomatic Ovarian Cancer. Int J Gynecol Cancer. 2014; 24: 865–873
15. National Cancer Institute. Common Toxicity Criteria for Adverse Events v3.0 (CTCAE). Bethesda, MD: National Cancer Institute; 2006 .
16. Donovan HS, Ward S, Sherwood P, et al. Evaluation of the Symptom Representation Questionnaire (SRQ) for assessing cancer-related symptoms. J Pain Symptom Manage. 2008; 35: 242–257.
17. Basen-Engquist K, Bodurka-Bevers D, Fitzgerald MA, et al. Reliability and validity of the functional assessment of cancer therapy–ovarian. J Clin Oncol. 2001; 19: 1809–1817.
18. Aaronson NK, Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst. 1993; 85: 365–376.
19. Greimel E, Bottomley A, Cull A, et al. An international field study of the reliability and validity of a disease-specific questionnaire module (the QLQ-OV28) in assessing the quality of life of patients with ovarian cancer. Eur J Cancer. 2003; 39: 1402–1408.
20. Stockler MR, O’Connell R, Nowak AK, et al. Effect of sertraline on symptoms and survival in patients with advanced cancer, but without major depression: a placebo-controlled double-blind randomised trial. Lancet Oncol. 2007; 8: 603–612.
21. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983; 67: 361–370.
22. Herth K. Abbreviated instrument to measure hope: development and psychometric evaluation. J Adv Nurs. 1992; 17: 1251–1259.
23. Bjelland I, Dahl AA, Haug TT, et al. The validity of the Hospital Anxiety and Depression Scale. An updated literature review. J Psychosom Res. 2002; 52: 69–77.
24. Cohen J. A power primer. Psychol Bull. 1992; 112: 155–159.
25. Osoba D, Bezjak A, Brundage M, et al. Analysis and interpretation of health-related quality-of-life data from clinical trials: basic approach of The National Cancer Institute of Canada Clinical Trials Group. Eur J Cancer. 2005; 41: 280–287.
26. Doyle C, Crump M, Pintilie M, et al. Does palliative chemotherapy palliate? Evaluation of expectations, outcomes, and costs in women receiving chemotherapy for advanced ovarian cancer. J Clin Oncol. 2001; 19: 1266–1274.
27. Fromme EK, Eilers KM, Mori M, et al. How accurate is clinician reporting of chemotherapy adverse effects? A comparison with patient-reported symptoms from the Quality-of-Life Questionnaire C30. J Clin Oncol. 2004; 22: 3485–3490.
28. Basch E, Jia X, Heller G, et al. Adverse symptom event reporting by patients vs clinicians: relationships with clinical outcomes. J Natl Cancer Inst. 2009; 101: 1624–1632.
29. Petersen MA, Larsen H, Pedersen L, et al. Assessing health-related quality of life in palliative care: comparing patient and physician assessments. Eur J Cancer. 2006; 42: 1159–1166.
30. Luckett T, Butow PN, King MT, et al. A review and recommendations for optimal outcome measures of anxiety, depression and general distress in studies evaluating psychosocial interventions for English-speaking adults with heterogeneous cancer diagnoses. Support Care Cancer. 2010; 18: 1241–1262.
31. Newell S, Sanson-Fisher RW, Girgis A, et al. How well do medical oncologists’ perceptions reflect their patients’ reported physical and psychosocial problems? Data from a survey of five oncologists. Cancer. 1998; 83: 1640–1651.