Second and subsequent lines of chemotherapy for metastatic breast cancer: what did we learn in the last two decades?

F. Cardoso, A. Di Leo,+, C. Lohrisch, C. Bernard, F. Ferreira and M. J. Piccart

Department of Medical Oncology, Jules Bordet Institute, Boulevard de Waterloo 125, 1000 Brussels, Belgium

Received 22 August 2001; revised 22 October 2001; accepted 15 November 2001.


    Abstract
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
‘It matters not how long we live, but how.’

Festus, Philip James Bailey

Despite almost 30 years of clinical cancer research, the true impact of second and subsequent lines of chemotherapy on the outcome of metastatic breast cancer patients, especially on the duration of survival, is still unknown. In the virtually incurable metastatic setting, issues like quality of life and patients’ preferences gain particular relevance. At the turn of the century, in-depth rethinking of the design of clinical trials run in this challenging disease setting appears to be warranted.

Key words: breast cancer, metastatic, quality of life, second line


    Introduction
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
Breast cancer is the most common cancer and the most common cause of cancer death in women. Over the last few decades there has been a stage migration, with a greater proportion of women being diagnosed with early stages of the disease, for which curative approaches are still possible. Nevertheless, 20–85% of these patients, depending on the initial stage, tumour biology and treatment strategy used, will develop distant metastases within 5 years of their initial diagnosis [1, 2]. An additional 6–10% present with metastatic disease at diagnosis [3, 4]. Once metastases are detected, the median survival is in the range of 18–24 months [38].

Both chemotherapy and hormonal therapy have been used to treat metastatic breast cancer (MBC) with different levels of success. Most patients experience objective responses associated with palliation of symptoms [1, 5, 7], but complete responses (CR) are uncommon and short-lived. Initial responses may last between 8 and 14 months [2, 4, 7]; however, progression of disease is inevitable, and durable responses to subsequent therapies are progressively fewer [1, 2, 5, 9, 10].

The true impact of chemotherapy on survival and quality of life (QOL) of MBC patients is still debated and under evaluation. The breast cancer mortality rate has remained stable for several decades, although a slight decrease is becoming evident in the last decade, primarily due to improvements in cure rate in early disease stages [1114]. This suggests that therapy for advanced breast cancer has not dramatically improved over the years [6], with the provision that nowadays most patients who relapse have received some form of adjuvant therapy and, presumably, have more resistant tumours. While a few randomised trials have shown improved survival for certain first-line regimens, chemotherapy beyond first line is associated with responses in fewer patients, and has no discernible or consistent effects on median survival. This fact has been the stimulus for developing newer and more effective drugs and therapeutic strategies. However, the value of second and subsequent lines of chemotherapy may be better appreciated with other outcome measures, such as the proportion of patients surviving at 1 and 2 years, and QOL improvements. These could help to define more clearly the value of such therapies in MBC patients. This paper reviews the available evidence regarding the impact of second and subsequent lines of chemotherapy on survival and QOL in patients with MBC.


    Impact of chemotherapy for MBC on survival
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
Chemotherapy was introduced as an anticancer therapy in the 1950s and combination chemotherapy started to be routinely used in the 1970s. Surprisingly, more than half a century later, its true impact on MBC patients’ survival is still a matter of debate.

There is a general consensus that the role of cytotoxic therapy in prolonging survival in MBC patients is modest, and that the efficacy of second and subsequent lines of chemotherapy, also called ‘salvage’ therapy regimens, is uniformly poor [5, 7, 15, 16]. Response rates are in the range of 20% and are short-lived, and median overall survival (OS) is usually <10 months (between 6 and 12 months) [5, 1719]. In an interesting retrospective study by Porkka et al. [18], out of 366 cycles (115 patients) of salvage chemotherapy, only one CR and 18 partial responses (PR) were seen, with the majority of them occurring with first salvage therapy and none being obtained beyond third-line salvage chemotherapy.

It seems clear that, for the majority of patients, the benefit of more than three lines of chemotherapy for metastatic disease is minimal if any. Notwithstanding, many patients still receive several lines of chemotherapy, sometimes given until death occurs. It must be remembered, however, that MBC is a heterogeneous disease, with substantial variations in growth rate and responsiveness to therapy, and its clinical outcome and prognosis in the individual patient do not always conform to the published data, which is based on population averages [6, 11].

Another difficult fact to reconcile is the observation that higher response rates (RR) and longer time to progression (TTP) do not always translate into detectable survival advantages. This might be related to the pattern of breast cancer growth. The typical Gompertzian curve shows that when the rate of cancer cell killing reaches a certain point, the rate of re-growth of the residual cancer cells starts to rise. This can, at least partially, explain the discrepancy between anticancer effects in the laboratory and those in the clinic [9]. Another plausible explanation for this lack of correlation may be that the individual trials conducted so far have usually been small and have lacked the statistical power necessary to detect modest survival gains [20].

A variety of drugs and regimens have been evaluated for treatment of MBC after failure of first-line chemotherapy. So far, none has unequivocally proved its superiority and thus no single standard salvage therapy exists, with the possible exception of taxanes, particularly taxotere, in anthracycline-resistant disease. In this setting, level 2 evidence exists to support the use of taxotere. This evidence is based on three randomised phase III trials, with a total of 851 patients, where single agent taxotere was superior to mitomycin + vinblastine [7], methotrexate + 5-fluorouracil (5-FU) (phase III trial) [15] and vinorelbine + 5-FU [21]; in the first study a survival gain was seen with taxotere. The majority of published studies on second and subsequent lines of chemotherapy for MBC are small, non-comparative trials that evaluate the feasibility of monotherapy or combinations of several cytotoxic agents. Few phase III randomised studies have compared two different chemotherapy regimens, and the usual end-points, apart from safety profile, were RR and TTP, with very few studies reporting data on survival.

There are no reported prospective randomised trials comparing chemotherapy with no chemotherapy, and it would be ethically impossible to conduct one today. We must rely on indirect methods of addressing the chemotherapy versus no chemotherapy question, such as non-randomised series, historical controls [20] or a randomised comparison between two different chemotherapy regimens. One can extrapolate that if one regimen shows superiority, namely in terms of survival, over another, then it would probably show the same or even greater superiority over a no treatment arm.


    Historical controls and non-randomised studies
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
Some studies have been conducted comparing chemotherapy and no chemotherapy using historical series of patients (also called ‘hospital-based studies with selective series of patients’). Results are contradictory, with three studies unable to demonstrate an improvement in survival with the introduction of chemotherapy [2224], and two studies being positive for a survival difference favouring chemotherapy [10, 25].

Powles et al. [23] retrospectively compared 78 patients who received no chemotherapy with 80 patients who received one of three chemotherapy regimens: oral cyclophosphamide, CMFV (cyclophosphamide + methotrexate + 5-FU + vincristine) or VAP (vincristine + doxorubicin + prednisone). Despite higher response rates in the chemotherapy groups, chemotherapy was not associated with an improvement in survival from the time of diagnosis of metastatic disease. However, in subgroup analysis responders to chemotherapy had an increase in survival benefit by a factor of four.

These results were duplicated in 1986 in a similar study by Patel et al. [22], who performed a retrospective review of 483 patients with MBC treated in one hospital between 1942 and 1975. They found no trend towards improved survival time and the ‘median metastatic time’ (time from first recurrence to death) remained steady at 21 months. Nevertheless, for responders chemotherapy had a significant impact, and this subgroup may benefit extended survival from chemotherapy.

Todd et al. [24] retrospectively analysed MBC patients’ survival from 1920 to 1980. Median survival increased steadily from 21 months in the 1920s to 41 months in the 1970s. However, 5-year survival reached 25% in the 1960s and, despite the introduction of combination chemotherapy during the 1970s, it remained near 25% in that decade.

On the other hand, Ross et al. [10] performed a comparison of consecutive series of patients in the 1950s, 1960s and 1970s and concluded that, except for patients with a very poor performance status, all subgroups showed a significantly improved survival in the 1970s compared with earlier decades. The introduction of combination chemotherapy was considered to be the major factor for survival improvement. The median improvement in survival was 9–12 months, similar to the median duration of response seen with various combination chemotherapy regimens.

Brincker et al. [25] compared survival after recurrence of patients entered in several protocols of the Danish Breast Cancer Cooperative Group between 1977 and 1982 and historical controls. Assuming that historical patients and protocol patients had comparable disease burden distribution, the authors concluded that the survival after recurrence and TTP benefit from chemotherapy for MBC was ~7 months.

The results from these studies were later pooled in a statistical overview performed by A’Hern et al. [20] in 1988. Despite probable publication bias, they suggested that there is a positive correlation between improved response rates and improved median survival. Because the individual trials were small, with modest differences in response rates, only minimal benefits in median survival could be observed in each of them.

There was also a population-based retrospective study, analysing 196 patients with recurrent breast cancer, 180 of whom had distant metastases, over a period of three decades. The authors concluded that chemotherapy in recurrent breast cancer prolongs survival by 9.5 months. This study had two basic assumptions: that the natural history of MBC did not change over the three decades analysed, and that the influence of endocrine therapy for MBC had remained constant over that period of time. The latter assumption, at least, is debatable [26].

An additional interesting finding was that previous adjuvant chemotherapy seemed to be associated with decreased efficacy of subsequent chemotherapy given for MBC, although no definite conclusions could be drawn due to the low number of patients who received adjuvant chemotherapy. Of note, in 2000, Salvadori et al. [27] presented a study analysing the possible effects that adjuvant anthracycline-containing chemotherapy might have on the activity of the same type of regimens as first-line treatment, and found no major influence.

Phase II trials have evaluated the efficacy and tolerability of several cytotoxic agents and combinations, after failure of first-line treatment of MBC. Although taxanes have proven their valuable role in anthracycline-resistant patients [7, 15, 21, 2831], other agents and regimens have low response rates and none has shown superiority over the others (although few randomised trials have been conducted) [15, 32].


    A critical look at randomised phase III trials
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
The way to evaluate properly the relative efficacy of a given cytotoxic drug or regimen is through a prospective randomised phase III trial. There are very few such trials addressing second or subsequent lines of chemotherapy, and even fewer report data on survival. A summary of those studies is presented in Table 1. Three of these studies showed improved survival [7, 33, 34].


View this table:
[in this window]
[in a new window]
 
Table 1. Randomised phase III trials: second or subsequent lines of chemotherapy
 
In 1995, Jones et al. [33] reported a prospective multicentric randomised trial that compared vinorelbine (NVB) with melphalan (ALK) in 183 patients with anthracycline-refractory advanced breast cancer. A 2:1 (NVB:ALK) randomisation ratio was used. No crossover was allowed. Primary end-points were TTP, QOL and relief of cancer-related symptoms. Secondary end-points were survival, time to treatment failure (TTF) and RR. The majority of patients (91% for NVB and 92% for ALK) had received anthracyclines in the advanced or advanced + adjuvant setting. NVB was associated with significantly longer TTP (P < 0.001), TTF (P < 0.001), RR (P = 0.06) and median OS (P = 0.034). For NVB and ALK, mean dose intensities were 19.3 mg/m2/week and 4.8 mg/m2/week, respectively, resulting in mean proportions of intended dose of 65% and 81%. In the NVB arm, dose reductions and dose delays were observed in 66% and 75% of patients, respectively. The incidence of grade 3 or 4 neutropenia was similar in both arms (75% versus 69%), as was the percentage of patients hospitalised for fever, infection, sepsis or pneumonia associated with neutropenia (10% versus 8%). There were no toxic deaths. A formal QOL assessment was performed and no significant differences between the two treatment arms were found. The authors note that ‘it was not considered ethical to deny these patients palliative treatment; therefore, in the absence of a therapeutic reference standard, ALK was selected for comparison’. A survival advantage was found in favour of NBV [33]. However, the Food and Drug Administration (FDA) did not accept this study on the basis of some methodological problems.

In 1999, Nabholtz et al. [7] published a prospective randomised trial of docetaxel versus mitomycin + vinblastine (MV), in 392 patients with MBC progressing despite previous anthracycline-containing chemotherapy. All patients in both arms previously received this type of chemotherapy in either adjuvant (17% and 21%), advanced (49% and 50%) or both (34% and 29%) settings. Crossover was not part of the study design; however, at the time of progression, 47% of patients in the docetaxel arm received further chemotherapy (12% of which with MV) and 54% of patients in the MV arm also received further chemotherapy (24% of which with docetaxel). In an intent-to-treat analysis, RR (P < 0.0001), median TTP (P = 0.001) and OS (P = 0.0097) were significantly longer with docetaxel. When adjusted for the crossover treatment, the difference in OS remained statistically significant (P = 0.007) in favour of docetaxel. The median relative dose intensity (RDI) was 0.94 (range 0.01–1.05) for docetaxel, 0.99 (range 0.65–1.43) for mitomycin and 0.97 (range 0.65–1.24) for vinblastine. Both arms had similar proportions of treatment delay (9.9% versus 9.3%), but dose reductions were more common with docetaxel (19.7% versus 4%). Withdrawal rates (11.8% versus 6.9%), toxic deaths (2% versus 1.6%) and QOL analysis were similar in both arms. The incidence of febrile neutropenia (9% versus 0.5%), grade 3/4 infections (11% versus 1.1%) and grade 3/4 neutropenia (93% versus 62%) were significantly higher with docetaxel. Grade 3/4 thrombocytopenia was significantly more frequent with MV (12% versus 4%). For non-haematologic side effects, stomatitis (9% versus 0.5%), diarrhoea (7.5% versus 0%), skin toxicity (4% versus 0%), asthenia (16% versus 6.4%), neurotoxicity (5% versus 0.5%) and nail disorders (2.5% versus 0%) were significantly more common with docetaxel, while constipation (0.5% versus 3.2%) was more frequent with MV [7].

At the 23rd Annual San Antonio Breast Cancer Symposium, in December 2000, O’Shaughnessy [34] presented a randomised phase III trial comparing docetaxel–capecitabine combination therapy versus docetaxel monotherapy in 511 MBC patients, resistant or relapsing after anthracycline-based therapy. The majority of patients, in both arms (65% and 67%), received the study drugs as second- or third-line chemotherapy. The primary end-point was TTP and secondary end-points were RR, OS, QOL, safety profile and medical care use. In the combination arm, 106 (42%) patients received prior anthracycline-based chemotherapy [46 in (neo)adjuvant and 60 in MBC setting]; in the monotherapy arm, 108 (42%) patients received that type of previous chemotherapy [44 as (neo)adjuvant and 64 as metastatic]. Crossover was not part of the study design; however, 20% of patients from the combination arm received single-agent docetaxel as post-study chemotherapy and 17% of patients from the monotherapy arm received single-agent capecitabine after the study; none received the combination. The capecitabine–docetaxel arm had superior TTP (P = 0.0001), overall RR (P = 0.006) and median OS (P = 0.0126), with a manageable side effect profile. Nevertheless, there were four (1.6%) toxic deaths in the combination arm as opposed to one (0.4%) in the single agent arm. Patients receiving monotherapy had a higher incidence of neutropenic fever, alopecia, myalgia and arthralgia, while diarrhoea, stomatitis, nausea/vomiting and hand-foot syndrome were more frequent in the combination arm. There were no significant differences regarding QOL. The proportion of dose reductions (65 versus 36%), treatment withdrawals (25 versus 18%) and treatment related hospitalisations (38 versus 36%) were higher in the combination arm [34]. Full publication of this study will be of interest.

Some phase III trials have been carried out in which a survival advantage was not found. These are reported below.

Mitoxantrone and doxorubicin were compared in a randomised phase II/III trial, as second-line chemotherapy for 325 MBC patients [35]. The main aim of this study was to determine accurately the activity, toxicity and degree of cross-resistance of mitoxantrone, as compared with doxorubicin. Crossover at objective evidence of progression was part of the study design. Among responders, difference in RR after crossover was not statistically significant; of the 39 patients resistant to doxorubicin, none responded to mitoxantrone, while six of the 70 patients resistant to mitoxantrone responded to doxorubicin. In an intent-to-treat analysis, doxorubicin yielded higher RR (P = 0.07) and higher TTF (P = 0.36), but also significantly higher toxicity. No differences in median OS were seen. The mean dose per course of mitoxantrone decreased slightly over time, while with doxorubicin this decrease was more rapid and more intense. Haematological toxicity (neutropenia and thrombocytopenia) was similar in both arms. For non-haematological toxicity, nausea/vomiting (77.3 versus 64.6), stomatitis/mucositis (31.8 versus 10.1) and alopecia (79.9 versus 27.8) were statistically more frequent with doxorubicin. An analysis of the probability of a cardiac event related to the cumulative dose of either drug showed a much lower risk for mitoxantrone-treated patients (P = 0.0005) [35].

Joensuu et al. [19] reported a prospective randomised trial comparing sequential low toxicity single-agent therapy with sequential combination therapy, in 303 MBC patients, as first- or second-line therapy. Single-agent therapy consisted of epirubicin given weekly at 20 mg/m2 (until progression or cumulative dose of 1000 mg/m2) followed by mitomycin. Combination first-line therapy was CEF (cyclophosphamide, epirubicin, 5-fluorouracil) and, as second line, the association mitomycin ± vinblastine was given. Primary end-point was survival and secondary end-points were QOL, toxicity, RR and TTP. It was assumed that 2-year survival for the standard arm (the combination arm) would be 0.30 and the study was 80% powered to detect a 1.5 ratio of median survival times. No crossover was allowed. There were no differences in OS or survival from the beginning of second-line therapy. Median survival after second-line chemotherapy was 10 and 8 months for single- and multi-agent therapy, respectively. With regard to first-line therapy, RR was higher for the combination regimen (53% versus 44%, not significant), whereas for second line, RR was higher for single-agent mitomycin (14% versus 6%, not significant). The sequential single-agent strategy was associated with less toxicity and better QOL. When comparing CEF with single agent epirubicin, haematologic side effects, alopecia, gastrointestinal toxicity, infections, fever, stomatitis, conjunctivitis, lethargy and neurological symptoms were all more frequent with CEF; only thrombosis and psychiatric symptoms were more commonly seen with epirubicin; incidence of cardiotoxicity was similar in both arms. Chemotherapy-related toxicity was also higher with the combination MV than with mitomycin alone, namely leucopenia (P = 0.005), alopecia (P = 0.003), nausea/vomiting (P = 0.01) and anaemia (P = 0.07). The median cumulative dose of epirubicin was 471 mg/m2 for CEF and 444 mg/m2 for low-dose epirubicin (P = 0.30). Treatment delays (42% versus 33%) and dose reductions (10% versus 0%) were more common with CEF than with single-agent epirubicin, which could explain at least partially why combination chemotherapy did not achieve a better survival [19].

The Scandinavian Breast Group compared docetaxel monotherapy with sequential methotrexate-5-FU in 283 patients with advanced breast cancer, after anthracycline failure. Primary end-point was TTP, secondary end-points were RR, toxicity and QOL, and an additional objective was to evaluate the influence on OS, of the order of administration of two salvage chemotherapy regimens. Crossover was recommended on progression. In an intent-to-treat analysis, overall response rate (ORR) (P < 0.001) and median TTP (P < 0.001) were higher in the docetaxel arm. Median OS, including the crossover phase, was not statistically different in both arms (P = 0.86). The median relative given dose per course (99% in both arms) and the median RDI (docetaxel 95% versus MF 94%) were similar in both arms, as were the percentages of treatment withdrawals (94% versus 95%). Overall side effects (including leucopenia, febrile neutropenia, infection, oedema, peripheral neuropathy, asthenia, nail changes, skin toxicity, stomatitis, alopecia, allergy and diarrhoea) were more frequent in the docetaxel arm, but grade 3 and 4 toxicities were infrequent in both treatment arms [15]. Conjunctivitis was the only side effect that occurred significantly more in the MF arm. There were three toxic deaths in the docetaxel arm and one in the MF arm. An evaluation of QOL was later performed and concluded that differences between the two arms were minor, and should not influence the choice between the two chemotherapy regimens [36].

Docetaxel was compared with doxorubicin in patients who had previously received alkylating agent-containing chemotherapy. Of the 326 patients randomised, 174 had received previous chemotherapy for MBC. Primary end-point was TTP. In an intent-to-treat analysis docetaxel produced a significantly better ORR (P = 0.008), but median TTP (P = 0.45) and median OS (P = 0.39) were not statistically different. No crossover was planned, but 28% of the docetaxel-treated group received anthracycline-based chemotherapy, and 26% of doxorubicin-treated patients received taxane-based chemotherapy, as first treatment after study therapy. When adjusted for crossover treatment, the difference in OS between the two treatment arms remained non-significant. The median RDI was 0.97 for docetaxel and 0.95 for doxorubicin. The percentages of withdrawals (54% versus 66%, P = 0.027) and of delayed cycles (7% versus 15%) were lower in the docetaxel arm. Dose reductions occurred in a similar number of cycles (5% in both arms). There were five (3%) toxic deaths in the doxorubicin arm, four of which were due to cardiotoxicity, and two (1.2%) toxics deaths in the docetaxel arm. The incidence of grade 4 neutropenia was similar with both drugs, but the incidence of severe neutropenic complications was higher with doxorubicin. Cardiotoxicity, nausea/vomiting and stomatitis were more frequent with doxorubicin, whereas fluid retention, diarrhoea, skin toxicity, allergy, nail disorder and neurotoxicity occurred more commonly with docetaxel. Although there were limitations in QOL assessment due to missing data, no major differences were seen between the two arms [28].

Norris et al. [37] compared combination doxorubicin ± vinorelbine with single-agent doxorubicin in a group of 303 MBC patients. In each treatment arm, 25% of patients (a total of 75 patients) had received one prior line of chemotherapy for MBC, although it could not have included anthracyclines or vinca alkaloids (most received CMF-like regimens). Primary end-point was median survival. In an intent-to-treat analysis, RR (38% versus 30%), duration of response (7.2 versus 6.8 months), TTP (6.2 versus 6.1 months), OS (13.8 versus 14.4 months) and QOL were not statistically significantly different. Both haematological and non-haematological toxicity were worse in the combination arm, namely neutropenia (100% versus 93%), febrile neutropenia (15% versus 10%; P = 0.2), neurotoxicity (6% versus 1%; P = 0.03) and constipation (47% versus 20%). Treatment withdrawals (11% versus 4%) were higher with the combination. There was no statistical difference in the received dose-intensity of doxorubicin by arm (P = 0.13), but there was a statistically significant difference in the total mean cumulative dose in favour of arm 2 (328 versus 260 mg/m2; P = 0.0001). Although the study had inadequate power to draw definite conclusions from subgroup analysis, there were also no statistically significant differences in RR (30% versus 24%), TTP (4.3 versus 5.3 months) or OS (9.4 versus 11.3 months) for the two treatments in the group of patients receiving them as second-line chemotherapy. Thus, the study hypothesis, that the combination of doxorubicin + vinorelbine was superior to doxorubicin alone, was not confirmed. However, this study was powered to detect a 50% relative increase in median OS, so it could not detect a 25% relative increase, which is probably the magnitude of gain for doxorubicin combinations as suggested by other studies.

A critical review of these randomised studies raises some possible explanations for the discrepancy in results (positive/negative). All three positive studies were performed in a selected population of MBC patients (all anthracycline-resistant), while this was the case in only one of the five negative studies. Since MBC patients are a quite heterogeneous population, a promising chemotherapy regimen may have a better chance of achieving positive results if it is applied to a well defined subgroup. Another important factor is the population size: the higher the number of patients entered in a study, the higher the probability of detecting a small, but significant, difference between the study arms. The trials by Nabholtz et al. [7] and O’Shaughnessy [34] are the two largest studies, with approximately 400–500 patients. In clinical practice, MBC patients are treated sequentially with different chemotherapy regimens. For that reason, a trial with built-in or allowed crossover between arms more accurately reflects clinical practice. However, from the data presented in Table 1 we can conclude that even when crossover is allowed or recommended, it only occurs in <30% of patients. It is essential that a drug (or combination of drugs) not only proves its superiority to the existing ones, but also that it is capable of maintaining this superiority even if a high rate of crossover occurs. On the other hand, a high percentage of crossover may ‘dilute’ the positive study results, particularly regarding OS, and therefore demands a larger study population. Among all the articles reviewed, the positive ones had the lowest rates of crossover, and in one of them [33] this was not even allowed.


    Impact of chemotherapy for MBC on QOL
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
Since treatment for MBC is not curative, QOL is an important issue to consider when selecting a treatment regimen [9, 36, 38, 39]. The gains in TTP must be balanced with treatment toxicity and inconvenience so that they can be ‘translated’ into gains in symptom control and better QOL. Clearly this is a more relevant problem for chemotherapy than hormone therapy, the latter being usually well tolerated. In this palliative setting, improving QOL or even maintaining it at an acceptable level is a main goal, and since no therapy confers a clear survival benefit, a positive influence on QOL might be the decisive factor in the choice of a chemotherapy regimen [36, 39, 40].

According to ASCO (American Society of Clinical Oncology) recommendations published in 1995, although survival is the most important outcome, alone it is not enough. The quality of survival and associated costs must also be addressed, leading to the concept of ‘quality-adjusted survival’ [41]. Regarding metastatic cancer, treatment can be recommended even in the absence of survival improvement, if it improves the QOL [41]. Also stated is that there is no minimal QOL benefit above which treatment is justified, rather the benefits must be balanced against toxicity and cost [41].

QOL is a multidimensional concept, subjective and dynamic in nature. It includes the patient’s own assessment of physical, psychological and social well-being, and functional capacity [39, 4144]. Its value as a measure of treatment efficacy has only recently been widely accepted. An important factor is the heterogeneity of existent QOL measurement instruments, none being universally accepted [40, 41], which makes accurate comparison of data difficult. Furthermore, all are time- and resource-consuming for patients and care givers, and require trained experts to analyse them [40]. Understandably, most randomised trials published so far do not provide data on the impact of therapy on QOL or evaluate QOL through the analysis of surrogates, such as toxicity measures, performance status evolution, decrease in analgesic consumption and treatment’s antitumour activity [36, 40, 44]. Each of these surrogates measure only limited aspects of QOL, and accurate, easy to perform, multidimensional measures are clearly needed [44]. As physicians have exponentially increased their focus on patients’ QOL in recent years [44], and as QOL is now recognised as an end-point of secondary importance only to survival [39, 41], systematic assessment of QOL in clinical trials is expected to become more frequent.


    Response rate as a surrogate for QOL
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
Several, but not all, studies have confirmed the correlation of RR with symptom control and, consequently, better QOL [41]. Early studies by Baum et al. [45] and Coates et al. [38] showed that in patients with advanced breast cancer, QOL was improved by inducing a tumour response. Carlson et al. [40] showed that objective response to chemotherapy can be a strong predictor of QOL, regardless of its acute toxicity (Report of a European Breast Cancer Expert Panel, submitted). Other studies also reported an association of objective tumour response with symptom control and improved QOL [4648].

In general, but not universally, there seems to exist a relationship between the degree of response and symptom improvement; higher responses (CR/PR) appear to be associated with better control of at least certain types of symptoms [48]. Another important factor is the severity of symptoms before treatment, since tumour response is more likely to positively affect QOL in a symptomatic patient [41]. Thus, it seems appropriate to consider the achievement of an objective response as a good surrogate marker of QOL, despite acute and chronic treatment toxicity [8, 38, 40, 45, 46]. Nevertheless, this assumption is based on retrospective trials and needs to be confirmed in properly designed prospective studies.


    Prediction of response
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
The expected overall RR is, therefore, a determinant factor for the choice of salvage regimen for MBC. A comprehensive univariate and multivariate analysis of prognostic factors in MBC was performed in a group of 619 patients receiving an anthracycline-containing regimen (FAC; 5-fluorouracil, adriamycin, cyclophosphamide) as first-line chemotherapy [49]. A few, most relevant, pre-treatment characteristics were incorporated in a statistical model, and the probability of response to therapy was accurately predicted. According to our knowledge, no similar analysis has ever been published regarding other lines of chemotherapy for MBC. Nevertheless, it is known that some clinical and biological parameters are particularly helpful for prediction of response to chemotherapy in this setting, and should be evaluated carefully. Among the clinical and laboratory parameters, the most important are the patient’s performance status, percentage of weight loss and co-existing medical conditions, some haematological and biochemical abnormalities, the number of previous lines of therapy and the obtained response, and the number (tumour burden) and site of metastases [1, 49]. ‘Classical’ primary tumour characteristics that directly influence the aggressiveness of the disease, such as number of positive nodes, tumour size and grade and hormone receptor status, are also helpful [1]. Other potentially useful prognostic factors are the HER2 receptor, p53 and bcl2 status, and measurements of cellular proliferation (i.e. Ki67, MIB1). In addition, in recent years, translational researchers have focused their efforts on defining specific predictor factors of responsiveness to certain types of chemotherapy and hormone therapy. The most extensively studied has been the HER2/neu receptor; its overexpression has been associated with a higher probability of response to anthracycline-based chemotherapy, a relative resistance to CMF and, possibly, a deleterious effect of tamoxifen. However, owing to the retrospective nature of the majority of published studies and to difficulties with respect to definition and standardisation of the best detection method, these results still await confirmation and universal acceptance.


    QOL is an independent prognostic factor for survival
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
QOL and survival might be directly related. As a result of the pioneering work of Greer and colleagues, the prognostic relevance of psychosocial factors has been a matter of debate and research [50, 51]. QOL has been shown to be an independent and strong prognostic factor for survival in several types of metastatic cancer, including MBC [46, 5155]. This prognostic value is independent of performance status and even of tumour response [46, 51]. Pre-treatment QOL might even have a stronger prognostic value than performance status and could, in the future, replace it as an eligibility criterion in clinical trials [44]. This, however, needs further research and confirmation.

The reason for the association of QOL scores and survival is still not completely clear. According to Hürny and Coates, two possible mechanisms exist. In the so-called ‘causal’ mechanism, perceived QOL actively influences the disease and the survival. In this case the prediction of survival should also occur in the adjuvant setting. The other possible mechanism, called ‘trivial’, relies on the fact that seriously ill patients are usually quite aware of the severity of their underlying disease and may perceive disease progression even before it can be documented [39, 46, 51]. This accurate perception of severity would be reflected in reported QOL scores, so that patients with worse underlying disease, and consequently worse expected survival, would report worse QOL [46, 51]. Two studies [51, 56] looked at the prognostic value of QOL in the adjuvant setting compared with after relapse, and concluded that in the adjuvant setting the prognostic significance of QOL is minimal or obscured by chemotherapy effects, while after relapse it becomes a strong prognostic factor and predicts for subsequent OS. These findings support the ‘trivial’ explanation of the prognostic significance of QOL.


    Patients’ preferences
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
To give a patient the opportunity to participate actively in the choice of treatment also seems to be an important factor for their successful adaptation and QOL. Patients’ preferences for treatment are becoming increasingly important in clinical decision-making and as a subject of medical research [44, 57]. In the early studies of QOL in patients with advanced incurable cancer, it was assumed that these patients were not able to make assessments of their own QOL. However, recently there is an increasing amount of evidence that patients’ judgments of their QOL are quite accurate and consistent, and thus patients’ views must be incorporated in treatment planning [39].

In MBC, curative treatment is not currently available, and different treatment alternatives, including no therapy, have different impacts on patients’ QOL. For these reasons, treatment options in this setting must be discussed in terms of the trade-offs between quantity and QOL. These are clearly extremely emotionally stressful and personal decisions, and therefore should involve the patient [41, 57]. Furthermore, the definition of QOL includes several dimensions of human experience, and the relative importance of each has a high individual variability. As it has been described elsewhere, ‘a person’s QOL is what he or she determines it to be’ [39, 58, 59], and it is influenced by his/her hopes and expectations [60]. As such, it is almost intuitive that the best judge of one’s QOL is oneself. Treatment decisions, particularly in the palliative metastatic setting, should be based on individual patients’ values, preferences and life priorities, which can be quite different from those of the physician and even of family relatives. Notwithstanding, in clinical practice it is often the physician who takes the leading role in treatment decisions. Moreover, at least in some countries, the decision is often made in conjunction with the family but not seldom leaving the patient aside. Several myths contribute to this situation. The assumption that patients with advanced cancer are not able or might not wish to accurately evaluate their QOL and treatment options [39], the fear that ‘too much’ information might render the patient anxious or depressed, or that QOL inevitably worsens in the terminally ill patient [39]. Furthermore, several QOL measurements erroneously assume that all symptoms are equally important to all patients [61].

Results from several studies have contradicted these assumptions. Patients’ willingness to participate actively in treatment decisions has been proven [62]. Other studies have shown that the attitudes and choices made by doctors or nurses can be quite different from those of cancer patients [44, 6268]. Although for most patients, the greater the treatment-associated toxicity the less likely it is that they would accept the therapy, the median benefit required by patients to make a treatment worthwhile is, almost universally, much smaller than the level of benefit that the care-giver would choose on behalf of the patient [62]. Understandably, other kinds of conflict may arise due to the fact that health policy makers’ and payers’ views generally do not conform to those of cancer patients [41]. In a study by Slevin et al. [62], most patients were willing to accept an intense chemotherapy regimen for only a 1% chance of cure or a 3 month prolongation of survival, while physicians and nurses were much less willing to accept such risks. Another study, by McQuellon et al. [57], showed that for some patients (~15%) a gain in life expectancy as little as 1 month would be enough to choose a high-risk treatment. Younger patients were more willing to assume the risks of treatment for a small increase in life expectancy. Regarding the palliative effect of chemotherapy, 75% of the patients would choose treatment, even if it only reduced symptoms of pain, without increasing life expectancy. Caution must be taken when interpreting these results, since this study was performed in women with non-MBC and 22 of them (19%) voluntarily said that their answer might have been different if they were actually facing the proposed scenario. Nevertheless, it has been ascertained that when faced with a real choice rather than a theoretical one, patients become even more likely to accept treatment in return for a minor benefit [62].

The patient must be allowed to make their decision based on facts and realistic hopes concerning the illness and the expected efficacy of the several treatment options. Accordingly, physicians must be willing to take the time to accurately and extensively discuss all these issues with the patient, trying to ignore their own hopes and fears and focusing on the patients’ preferences.


    Discussion
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
The existence of some studies that show a survival advantage of a second-line chemotherapy regimen over another is an indirect suggestion that chemotherapy may indeed prolong survival over no therapy in MBC after failure of first-line treatment. However, these studies are few in number and in some, the control arm is either a substandard or ‘old fashioned’ chemotherapy regimen. The question of survival benefit of further chemotherapy in previously treated women with MBC clearly needs further research. Future trials in MBC must follow one of two major lines of research. (i) The definition of new regimens appealing to poor clinical scenarios, incorporating newer and more efficacious agents; these trials must include OS as a primary end-point, and need a large number of patients, appropriate control arms and some control of crossover to have enough power to detect small yet clinically relevant differences. (ii) Trials exploring more ‘gentle’ regimens for situations that are not life-threatening, where QOL should be the main focus. Since at present MBC is virtually incurable, and second and subsequent lines of chemotherapy have yet to be proven to have an impact on duration of survival, it is of particular importance to explore their effect on QOL in all its dimensions, including symptom control, toxicity, treatment inconveniences, and family and economic problems.

In this delicate process of treatment decision-making, patients should always be encouraged to participate, and their preferences and expectations should be taken into consideration. One should not forget that for some patients, minimal gains in survival are reason enough to undergo treatment, regardless of anticipated side effects.

In conclusion, evidence-based medicine provides some support for the use of second-line and, to a lesser degree and in selected cases, third-line chemotherapy for MBC. Beyond third-line treatment, however, there are no data that suggest a clear potential benefit for such a therapeutic approach (Figure 1). Nevertheless, it is of utmost importance to emphasise that all therapeutic decisions must be individualised, targeting the specific histological and biological features that make a tumour unique, and the physical and psychological characteristics that make a patient unique. Translational research, focusing on reliable predictive markers and discovery of potential specific targets for new anticancer agents, plays the key role in this process. Clinical experience, scientific knowledge and some indispensable common sense will hopefully enable us to reach the best decision for each individual patient. As stated by Osoba [44], there seems to be a narrow margin of benefit between an overly aggressive treatment and one that is not aggressive enough. It is this balance that needs to be sought on an individual basis.



View larger version (17K):
[in this window]
[in a new window]
 
Figure 1. Salvage chemotherapy for MBC ‘beyond first line’. OS, overall survival; QOL, quality of life; CT, chemotherapy; ORR, overall response rate. **ORR depending mainly on performance status, previous number of chemotherapy lines, number and location of metastatic sites, response to previous chemotherapy, and potentially some biological markers (see text for details).

 


    Footnotes
 
+ Correspondence to: Tel: +32-2-5413182; Fax: +32-2-5413090; E-mail: angelo.dileo@bordet.be Back


    References
 Top
 Abstract
 Introduction
 Impact of chemotherapy for...
 Historical controls and non...
 A critical look at...
 Impact of chemotherapy for...
 Response rate as a...
 Prediction of response
 QOL is an independent...
 Patients’ preferences
 Discussion
 References
 
1. Harris JR, Morrow M, Bonadonna G. Cancer of the breast. In De Vitta VT Jr, Hellman S, Rosenberg SA (eds): Cancer. Principles and Practice of Oncology, 4th edition. Philadelphia, PA: Lippincott Company 1993; 1264–1332.

2. Greenberg P, Hortobagyi G, Smith T et al. Long-term follow-up of patients with complete remission following combination chemotherapy for metastatic breast cancer. J Clin Oncol 1996; 14: 2197–2205.[Abstract]

3. Miller KD, Sledge GW Jr. The role of chemotherapy for metastatic breast cancer. Hematol Oncol Clin North Am 1999; 13: 415–434.[ISI][Medline]

4. Van Oosterom AT. Docetaxel (Taxotere): An effective agent in the management of second-line breast cancer. Semin Oncol 1995; 22 (Suppl 13): 22–28.

5. Muss HB, Case D, Richards F et al. Interrupted versus continuous chemotherapy in patients with metastatic breast cancer. N Engl J Med 1991; 325: 1342–1348.[Abstract]

6. Honig SF. Treatment of metastatic disease—hormonal therapy and chemotherapy. In Harris JR, Lippman ME, Morrow M, Hellman S (eds): Diseases of the Breast. Philadelphia, PA: Lippincott-Raven 1996; 669–734.

7. Nabholtz JM, Senn HJ, Bezwoda WR et al. Prospective randomized trial of docetaxel versus mitomycin plus vinblastine in patients with MBC progressing despite previous anthracycline-containing chemotherapy. J Clin Oncol 1999; 17: 1413–1424.[Abstract/Free Full Text]

8. Stockler M, Wilcken NR, Ghersi D et al. Systematic reviews of chemotherapy and endocrine therapy in metastatic breast cancer. Cancer Treat Rev 2000; 26: 151–168.[ISI][Medline]

9. Norton L. Metastatic breast cancer—length and quality of life. N Engl J Med 1991; 325: 1370.[ISI][Medline]

10. Ross MB, Buzdar AU, Smith TL et al. Improved survival of patients with metastatic breast cancer receiving combination chemotherapy. Comparison of consecutive series of patients in 1950’s, 1960’s, and 1970’s. Cancer 1985; 55: 341–346.[ISI][Medline]

11. Hortobagyi GN. Treatment of breast cancer. N Engl J Med 1998; 339: 974–984.[Free Full Text]

12. Chu KC, Tarone RE, Kessler LG et al. Recent trends in US breast cancer incidence, survival, and mortality rates. J Natl Cancer Inst 1996; 88: 1571–1579.[Abstract/Free Full Text]

13. Garne JP, Aspegren K, Balldin G et al. Increasing incidence of and declining mortality from breast carcinoma: trends in Malmo, Sweden, 1961–1992. Cancer 1997; 79: 69–74.

14. Beral V, Hermon C, Reeves G et al. Sudden fall in breast cancer death rates in England and Wales. Lancet 1995; 345: 1642–1643.[ISI][Medline]

15. Sjostrom J, Blomqvist C, Mouridsen H et al. Docetaxel compared with sequential methotrexate and 5-fluorouracil in patients with advanced breast cancer after anthracycline failure: a randomised phase III study with crossover on progression by the Scandinavian Breast Group. Eur J Cancer 1999; 35: 1194–1201.[ISI][Medline]

16. Fossati R, Confalonieri C, Torri V et al. Cytotoxic and hormonal treatment for metastatic breast cancer: a systematic review of published randomized trials involving 31,510 women. J Clin Oncol 1998; 16: 3439–3460.[Abstract]

17. Buckner JC, Ingle JN, Everson LK et al. Results of salvage hormonal therapy and salvage chemotherapy in women failing adjuvant chemotherapy after mastectomy for breast cancer. Breast Cancer Res Treat 1989; 13: 135–142.[ISI][Medline]

18. Porkka K, Blomqvist C, Rissanen P et al. Salvage therapies in women who fail to respond to first-line treatment with fluorouracil, epirubicin, and cyclophosphamide for advanced breast cancer. J Clin Oncol 1994; 12: 1639–1647.[Abstract]

19. Joensuu H, Holli K, Heikkinen M et al. Combination chemotherapy versus single-agent therapy as first- and second-line treatment in metastatic breast cancer: a prospective randomized trial. J Clin Oncol 1998; 16: 3720–3730.[Abstract]

20. A’Hern RP, Ebbs SR, Baum MB. Does chemotherapy improve survival in advanced breast cancer? A statistical review. Br J Cancer 1988; 57: 615–618.[ISI][Medline]

21. Bonneterre J, Roche H, Monnier A et al. Taxotere® (TXT) versus 5-fluorouracil + Navelbine® (FUN) as second-line chemotherapy (CT) in patients (pts) with metastatic breast cancer (MBC) (preliminary results). Proc Am Soc Clin Oncol 1997; 16: 162a (Abstr 564).

22. Patel JK, Nemoto T, Vezeridis M et al. Does more intense palliative treatment improves overall survival in metastatic breast cancer patients? Cancer 1986; 57: 567–570.[ISI][Medline]

23. Powles TJ, Smith IE, Ford HT et al. Failure of chemotherapy to prolong survival in a group of patients with metastatic breast cancer. Lancet 1980; 1: 580–582.[Medline]

24. Todd M, Shoag M, Cadman E. Survival of women with metastatic breast cancer at Yale from 1920 to 1980. J Clin Oncol 1983; 1: 406–408.[Abstract]

25. Brincker H. Distant recurrence in breast cancer. Survival expectations and first choice of chemotherapy regimen. Acta Oncol 1988; 27: 729–732.[ISI][Medline]

26. Cold S, Jensen NV, Brincker H et al. The influence of chemotherapy on survival after recurrence in breast cancer—a population-based study of patients treated in the 1950’s, 1960’s and 1970’s. Eur J Cancer 1993; 29A: 1146–1152.

27. Salvadori B, Gennari A, Landucci E et al. Effects of adjuvant anthracyclines on the activity of epirubicin containing first line chemotherapy in metastatic breast cancer. Breast Cancer Res Treat 2000; 64: 86a (Abstr 341).

28. Chan S, Friedrichs K, Noel D et al. Prospective randomized trial of docetaxel versus doxorubicin in patients with metastatic breast cancer. J Clin Oncol 1999; 17: 2341–2354.[Abstract/Free Full Text]

29. Dieras V, Marty M, Tubiana N et al. Phase II randomized study of paclitaxel versus mitomycin in advanced breast cancer. Semin Oncol 1995; 22 (Suppl 8): 33–39.

30. Seidman AD, Reichman BS, Crown JP et al. Paclitaxel as second and subsequent therapy for metastatic breast cancer: activity independent of prior anthracycline response. J Clin Oncol 1995; 13: 1152–1159.[Abstract]

31. Clemons M, Leahy M, Valle J et al. Review of recent trials of chemotherapy for advanced breast cancer: the Taxanes. Eur J Cancer 1997; 33: 2183–2193.[ISI][Medline]

32. Clemons M, Leahy M, Valle J et al. Review of recent trials of chemotherapy for advanced breast cancer: Studies excluding Taxanes. Eur J Cancer 1997; 33: 2171–2182.[ISI][Medline]

33. Jones S, Winer E, Vogel C et al. Randomized comparison of vinorelbine and melphalan in anthracycline-refractory advanced breast cancer. J Clin Oncol 1995; 13: 2567–2574.[Abstract]

34. O’Shaughnessy J, on behalf of the Xeloda Breast Cancer Study Group. Results of a large phase III trial of Xeloda/taxotere combination therapy vs taxotere monotherapy in metastatic breast cancer (MBC). Breast Cancer Res Treat 2000; 64: (Abstr 381).

35. Henderson IC, Allegra JC, Woodcock T et al. Randomized clinical trial comparing mitoxantrone with doxorubicin in previously treated patients with metastatic breast cancer. J Clin Oncol 1989; 7: 560–571.[Abstract]

36. Hakamies-Blomqvist L, Luoma M-L, Sjostrom J et al. Quality of life in patients with metastatic breast cancer receiving either docetaxel or sequential methotrexate and 5-fluorouacil. A multicentre randomised phase III trial by the Scandinavian Breast Group. Eur J Cancer 2000; 36: 1411–1417.[ISI][Medline]

37. Norris B, Pritchard KI, James K et al. Phase III comparative study of vinorelbine combined with doxorubicin alone in disseminated metastatic/recurrent breast cancer: National Cancer Institute of Canada Trials Group Study MA8. J Clin Oncol 2000; 18: 2385–2394.[Abstract/Free Full Text]

38. Coates A, Gebski V, Stat M et al. Improving the quality-of-life during chemotherapy for advanced breast cancer. A comparison of intermittent and continuous treatment strategies. N Engl J Med 1987; 317: 1490–1495.[Abstract]

39. Waldron D, O’Boyle CA, Kearney M et al. Quality-of-life measurement in advanced breast cancer: assessing the individual. J Clin Oncol 1999; 17: 3603–3611.[Abstract/Free Full Text]

40. Carlson RW. Quality of life issues in the treatment of metastatic breast cancer. Oncology 1998; 12 (Suppl 4): 27–31.[Medline]

41. ASCO. Outcomes of cancer treatment for technology assessment and cancer treatment guidelines. American Society of Clinical Oncology. J Clin Oncol 1996; 14: 671–679.[Abstract]

42. Ganz PA. Treatment options for breast cancer—beyond survival. N Engl J Med 1992; 326: 1147.[ISI][Medline]

43. Kiebert GM, Curran D, Aaronson NK et al. Quality of Life as an endpoint in EORTC clinical trials. European Organization for Research and Treatment of Cancer. Stat Med 1998; 17: 561–569.[ISI][Medline]

44. Osoba D. Lessons learned from measuring health-related quality of life in oncology. J Clin Oncol 1994; 12: 608–616.[Abstract]

45. Baum M, Priestman T, West R et al. A comparison of subjective responses in a trial comparing endocrine with cytotoxic treatment in advanced carcinoma of the breast. In Proceedings of The Second EORTC Breast Cancer Working Conference. Pergamon Press: Oxford 1980; 223–226.

46. Coates A, Gebski V, Signorini D et al. Prognostic value of quality-of-life scores during chemotherapy for advanced breast cancer. The Australian New Zealand Breast Cancer Trials Group. J Clin Oncol 1992; 10: 1833–1838.[Abstract]

47. Ramirez AJ, Towlson KE, Leaning MS et al. Do patients with advanced breast cancer benefit from chemotherapy? Br J Cancer 1998; 78: 1488–1494.[ISI][Medline]

48. Geels P, Eisenhaur E, Bezjak A et al. Palliative effect of chemotherapy: objective tumor response is associated with symptom improvement in patients with metastatic breast cancer. J Clin Oncol 2000; 18: 2395–2405.[Abstract/Free Full Text]

49. Hortobagyi GN, Smith TL, Legha SS et al. Multivariate analysis of prognostic factors in metastatic breast cancer. J Clin Oncol 1983; 1: 776–786.[Abstract]

50. Greer S, Morris T, Pettingale KW. Psychological response to breast cancer: Effect on outcome. Lancet 1979; 2: 785–787.[Medline]

51. Coates A, Hürny C, Peterson HF et al. Quality-of-life scores predict outcome in metastatic but not early breast cancer. J Clin Oncol 2000; 18: 3768–3774.[Abstract/Free Full Text]

52. Coates AS, Thomson D, McLeod GR et al. Prognostic value of quality of life scores in a trial of chemotherapy with or without interferon in patients with metastatic melanoma. Eur J Cancer 1993; 29A: 1731–1734.

53. Ganz PA, Lee JJ, Siau J. Quality of life assessment. An independent prognostic variable for survival in lung cancer. Cancer 1991; 67: 3131–3135.[ISI][Medline]

54. Faller H, Bulzebruck H, Drings P et al. Coping, distress, and survival among patients with lung cancer. Arch Gen Psychiatry 1999; 56: 756–762.[Abstract/Free Full Text]

55. Wilsoff F, Hjorth M. Health-related quality-of-life assessed before and during chemotherapy predicts for survival in multiple myeloma. Br J Haematol 1997; 97: 29–37.[ISI][Medline]

56. Hürny C, Bernhard J, Coates AS et al. Impact of adjuvant therapy on quality of life in women with node-positive operable breast cancer. Lancet 2000; 347: 1279–1284.

57. McQuellon RP, Muss HB, Hoffman SL et al. Patient preferences for treatment of metastatic breast cancer: a study of women with early-stage breast cancer. J Clin Oncol 1995; 13: 858–886.[Abstract]

58. O’Boyle CA. The schedule for the evaluation of individual quality of life. Int J Mental Health 1994; 23: 3–23.

59. Joyce CRB. Health status and quality of life: which matters to the patient? J Cardiovasc Pharmacol 1994; 23 (Suppl 3): 26–33.

60. Calman KC. Quality of life in cancer patients: a hypothesis. J Med Ethics 1984; 10: 124–127.[Abstract]

61. Stiggelbout AM, Haes JCJM. Patient preference for cancer therapy: an overview of measurement approaches. J Clin Oncol 2001; 19: 220–230.[Abstract/Free Full Text]

62. Slevin ML, Stubbs L, Plant HJ et al. Attitudes to chemotherapy: Comparing views of patients with cancer with those of doctors. Br Med J 1990; 300: 1458–1460.[ISI][Medline]

63. Slevin ML, Plant H, Lynch D et al. Who should measure quality of life, the doctor or the patient? Br J Cancer 1988; 57: 109–112.[ISI][Medline]

64. Faden RR, Becker C, Freeman J et al. Disclosure of information to patients in medical care. Med Care 1981; 19: 718–733.[ISI][Medline]

65. McNeil BJ, Weichelsbaum R, Pauker SG et al. Speech and survival: trade-off between quality and quantity of life in laryngeal cancer. N Engl J Med 1981; 305: 982–987.[Abstract]

66. Sptizer WO, Dobson AJ, Hall J et al. Measuring quality of life of cancer patients: A concise QL-index for use by physicians. J Chronic Dis 1981; 34: 585–597.[ISI][Medline]

67. Presant CA. Quality of life in cancer patients. Who measures what? Am J Clin Oncol 1984; 7: 571–573.[ISI][Medline]

68. Kahn SB, Houts PS, Harding SP. Quality of life and patients with cancer: A comparative study of patient versus physicians’ perceptions and its implications for cancer education. J Cancer Educ 1992; 7: 241–249.[Medline]