Adjuvant fluorouracil, epirubicin and cyclophosphamide in early breast cancer: Is it cost-effective?

Abstract

Adjuvant chemotherapy (ACT) in breast cancer exposes patients to morbidity, but improves survival. The FEC (fluorouracil, epirubicin, cyclophosphamide) regimen has taken over the prior role of CMF (cyclophosphamide, methotrexate, fluorouracil). In this model, efficacy, tolerability and quality of life (QoL) data from the literature were incorporated with Norwegian practice and cost data in a cost-effectiveness approach. The FEC efficacy was calculated 3–7% superior CMF. There was no difference in quality of life. An 80–100% dose intensity range was employed, one Euro (€ 1) was calculated NOK 8.78 and a 3% discount rate was used. The total cost of FEC employing the friction cost method on production loss, including amount spent on drugs, administration and travelling ranged between € 3 278–3 850 (human capital approach € 12 143–12 715). Money spent on drugs alone constituted 15–48%, depending on method chosen. A cost-effectiveness analysis revealed a cost per life year (LY) saved replacing FEC by CMF of € 3 575–15 125. Adjuvant FEC is cost effective in Norway.

Breast cancer is today the most common cancer among women in North America and throughout Europe. About one million women worldwide get the disease every year and it causes about 370 000 deaths [1]. A few years ago, cyclophosfamide, methotrexate and fluorouracil (CMF) was the standard adjuvant regimen employed for breast cancer. This regimen was launched back in the 1970s by Bonadonna and co-workers [2]. During the last decade, a new regimen [3] containing cyclophosfamide, epirubicin and fluorouracil (FEC or CEF) has become more and more popular in this setting and has in most countries taken over the role of CMF.

Whereas new regimens may offer improved survival figures, they frequently affect patients’ quality of life and health care budgets. New cytostatics may introduce new side effects that may diminish quality of life. In a situation with limited heath care resources, economic evaluations have to be implemented. Resources spent in one area of the battle against cancer may be allocated to other areas to achieve maximum value for money.

There are several instruments available in health economics. Among these are the cost-effectiveness, cost-minimization, cost-of-illness and cost-utility analysis. In this study, we implemented data from the literature on benefits in terms of quality of life and survival employing FEC and CMF together with national Norwegian cost data. We aimed to clarify the cost-effectiveness of the implementation of FEC in the place of CMF in adjuvant chemotherapy (ACT) in early breast cancer.

Material and methods

In a model-based cost-effective analysis (CEA), we have to implement data on life-years gained and direct and indirect costs. In this survey the following basis data were included:

Survival data

The PubMed database was searched for data on efficacy and costs related to FEC therapy. We assumed that there were no difference in quality of life (QoL) between FEC and CMF therapy and searched for QoL-data to “justify” that a cost-utility-analysis was not included in the project. Data on CMF was implemented from a prior survey [4]. The following inclusion criteria were employed: English written articles published in the time period from January 1998 to June 2004 and including the words “breast cancer” and adjuvant”. A total of 3 994 articles were disclosed. They were further searched for the words “FEC”, “CEF”, “CMF”, “cost”, “life years gained” or “quality of life”. Articles dealing with metastatic, locally advanced or male breast cancer and studies just mentioning “CEF” or “FEC” without reporting any clinical study as well as meta-analysis were excluded. Following inclusion and exclusion criteria, a total of 64 articles were disclosed and searched manually for quality of life and survival data related to the FEC/CEF/CMF regimens. Many articles were overview reports based on the same studies. They were excluded and the final survival data were based on the studies summarised in Table I. They all reported overall survival (OS) figures except for two articles [26], [28] reporting disease-free-survival (DFS). DFS may be a surrogate for OS and was implemented to support the life years gained (LYG) by exchanging the CMF with FEC regimen. We did not calculate any weighted mean survival benefit in terms of LYG, but implemented the range in median survival reported as shown in Table I.

Table I.  An overview of data from the English written literature on the efficacy of adjuvant FEC and CMF in early breast cancer.

					Efficacy FEC		Efficacy CMF
Ref. no	Study	TNM	No	Mean age (yrs)	5-yrs	10-yrs	5-yrs	10-yrs
[13] 2003	Randomised	N+, M0	621	42.6	65%	53%
[14] 2001	Randomised	T1-2,N0,M0	248	52	81%
[15] 2001	Randomised	N+, M0	565	50.6	66% ^1
[27] 1998	Randomised	N+, M0	360		77%		70%
					63%^2		53% ^2
[16] 2001	Overview	T1-2,N + /−,M0	710	<50	77%		70%
[12] 2002	Meta-analysis	Tx, N + /−,M0	18 000	<50	72%		69%
[29] 1999	Randomised	T2/3, N-,M0	390	45	85%	68%
[17] 2001	Phase III	Tx, N+, M0	777	<50	68%		63%
				>50	61%
[28] 2001	Overview	Tx, N + /−,M0			71.5%
					84.8%^2
[26] 2003	Overview	T, N+, M0	2005		85–81% ^3

¹FEC 100 (Epirubicin 100 mg/m²), ² Disease free survival (DFS), ³ 3-years DFS at different dose intensity.

The mean age of Norwegian women undergoing adjuvant chemotherapy (ACT) has been reported 50 years [4]. The life expectancy of a Norwegian woman aged 50 years is 28.57 years [5]. The corresponding figure of a 60-year old woman is 19.45 years. We know that relapse of breast cancer is usually diagnosed within the first ten years and relapsing women will respond to chemo- and/or hormonal therapy regimens before most of them die of metastatic breast cancer. Based on this knowledge and the prior report [4], we implemented the cost-effectiveness analysis women cured by FEC and not by CMF, living for a mean time of 20 years.

It could be argued that the FEC regimen also may gain life years over the CMF regimen simply by delaying time to relapse. However, we did not reveal any data supporting this assumption.

Costs

The cost of CMF differs from FEC only in the cost of chemotherapy as methotrexate is exchanged with epirubicin. There is no difference in travelling, number of courses or other drugs delivered, anti-emetics, interval, outpatient clinic and indirect costs. In this CEA the drug cost should be the only one included. Whereas it may look unnecessarily complicated to implement other costs (travelling, outpatient therapy, hospital administration and anti-emetics), we decided to implement them to cover these relevant aspects for decision-makers in health care.

Direct costs

Direct costs include drug costs (cytostatics, antiemetics), travelling, outpatient clinic costs and drug administration costs. One Euro (€ 1) was calculated 8.78 Norwegian krone (NOK).

Cytostatics are delivered at a dose per square meter (m²). We implemented the mean body surface area (BSA) among 5 000 women treated for breast cancer of 1.73 m² in our model [6]. The standard adjuvant FEC regimen employed in Norway and at the University Hospital of North Norway (UNN) is shown in Table II. The pricelist of cytostatics within Norway is available on the Internet (www.felleskatalogen.no). We employed the costs as in September 2004. Norwegian health authorities negotiate with the pharmaceutical companies and get a discount. This is called the LIS-agreement. The LIS cost as in September 2004 at the pharmacy at the UNN was employed. In this survey both alternatives for drug costs (feleskatalogen and LIS) were implemented.

Table II.  The FEC regimen as employed in Norway.

Drug	Dose (mg/m^2)	Solution	Frequency	No. of cycles
5-Fluorouracil	600	100 ml saline	Every 3rd week	6
Epirubicin	60	100 ml dextrose	Every 3rd week	6
Cyclophosphamide	600	100 ml saline	Every 3rd week	6

According to local tradition at the UNN, ondansetron 8 mg and 16 mg dexametasone is given i.v. at the start of chemotherapy and the ondansetrone dose is repeated as a single oral dose in the evening. Drugs administered as i.v. infusion have to be diluted in saline or dextrose solutions. The cost of these together with the hospital pharmacy's administration costs € 22.6 (NOK 198) as in August 2004 was included in the analysis. Outpatient clinic cost € 27.9 (NOK 245) was calculated according to a national price list [7]. The individual patient and the National Insurance Administration (NIA) cover the amount.

According to the rights of Norwegian patients, the hospitals have to cover patient transportation costs, except for a minor share taken care of by the patients themselves. The Regional health authority of Northern Norway paid in 2002 a total of NOK 506 million in patient transportation costs (€ 113.6 or NOK 997/patient/transport). However, many of these patients are transported by air (airplane/helicopter) or ambulance due to emergency settings. To indicate the cost of travelling for adjuvant chemotherapy in an outpatient setting, we registered the distance of travelling of 21 consecutive women travelling to the UNN for outpatient FEC therapy. The cost per kilometer (km) was calculated according to the national tariff of NOK 3/km. The mean distance (two ways) was 209 km (Be aware that the population of Northern Norway is scattered within a significant geographic area).

Indirect costs

The side effect of chemotherapy is significant [4], [8], [9]. Therefore most women stay out of the workforce during adjuvant FEC therapy. They are reported ill by their medical doctors and receive a pension from the NIA. To clarify the percentage of patients staying out of workforce, we questioned 8 senior medical oncologists dealing with breast cancer patients at the UNN. They reported 88% (range 75–95%) of women staying out of work force for a median time of 6 months. According to Statistics Norway [5], 83% of Norwegian women aged between 50–54 years are in the working force. The mean age of Norwegian women undergoing ACT is 50 years [4], [10]. The mean income in this group was NOK 236 800/year [5]. This includes wages, pension, capital income and income of self-employed tradesmen. The indirect costs were calculated employing the friction cost method [11]. In this analysis, the cost of simply replacing one worker by another is reflected. The cost was calculated € 985 [((NOK 236 800*0.83)/8.78*2*10%)*0.88]. Employing another main method for calculation of production loss, the human capital approach, the corresponding figure was € 9 850.

Method of analysis

In this study we employed a cost-effectiveness-analysis (CEA) indicating cost/life years gained changing policy from the CMF to the FEC regimen in adjuvant treatment of early breast cancer.

Results

Life years gained

In a CEA, only real differences in survival matters. As disease-free-survival (DFS) in fact is of no interest in a CEA, we implemented such data only in the overview and not in the final analysis. Several studies have shown a benefit of adjuvant FEC over CMF. One such meta-analysis by the Early Breast Cancer Trialists’ Collaborative Group (EBCTG) included more than 6 000 women [12]. In this study antracycline containing regimens (FEC, FAC) improved 5 years survival from 69 to 72% (difference is 3%). Fumoleau and coworkers [13] indicated 6 courses with FEC superior to 3 courses. Ten years survival was 64.3% and 56.6%, respectively. A randomised study by Paradiso et al. [14] included 248 women with early breast cancer (T1-2, N0, M0) where 50% had a histological grade 3 tumor into a randomised study with locoregional therapy +/ − FEC. Five years disease free survival (DFS) were (81 and 69%) in favour of FEC.

A French study [15] included 565 women with node positive disease (N+, more than 3 affected nodes, histological grade 2 or 3, estrogen receptor negative). Patients were randomised to receive FEC50 (50 mg/m² of epirubicin) or FEC100. 5 years DFS were 54.8 and 66.3 in the FEC50 and FEC100 groups, respectively. Another study by Levine and coworkers [3] including 710 patients revealed CEF superior to CMF. The 5-year survival was improved by 7% in the CEF arm. Trudeau et al. [16] run an overview of the literature and performed a randomised study comparing CMF and FEC. They concluded the 5-year survival data of FEC 7% better than those of CMF (77% vs 70%).

Whereas many studies strongly indicate FEC superior CMF, the study of Piccart et al. [17] did not show any difference between CMF and two EC regimens among women with node positive breast cancer.

Several studies employing different anthracycline containing regimens report a benefit of such regimens without specifying the difference to CMF. Other publications revealed were overview articles including data from the other articles already presented in this analysis.

Based on the literature, we concluded the FEC-regimen improving 5 years survival in early breast cancer by 3–7% when compared to CMF. This estimate is simply the range of survival benefits derived from the studies above comparing CMF and FEC. No weighted mean or calculated mean value was used. An overview of the articles is shown in Table I.

Quality of life

Several studies have focused QoL connected to adjuvant chemotherapy. Bergh et al. [18] performed an overview of the literature. They concluded that based on randomised adjuvant polychemotherapy studies, all regimens diminish QoL initially during therapy, but improves QoL later on compared with untreated patients. Brandberg et al. [19] reported employing the questionnaire developed by the European Organisation for Research and Treatment of Cancer (EORTC QLQ C-30) [20] a QoL before, after 16 weeks and after 54 weeks of FEC therapy of 71.5, 50.1 and 73.2, respectively. Del Mastro et al. [21] documented a higher level of distress during therapy when FEC was administered every two weeks compared to standard FEC.

The MA5 study [16] did also employ the EORTC QLQ C-30 instrument and revealed no significant difference in quality of life between CMF and FEC. Based on this important study, we concluded no difference in QoL between CMF and FEC and consequently this justified not using any cost-utility analysis.

Treatment costs using FEC

The total cost of chemotherapy employing FEC in early breast cancer was calculated ranging between € 3 278 and 3 850 (NOK 28 781–33 803) depending on pricelist employed and dose intensity (friction cost method). Employing the human capital approach, the corresponding figures were € 12 143–12 715 (NOK 106 616–111 638). Details are shown in Table III. Drug cost was the major factor influencing cost and constituted 39–48% of the total amount when the friction cost method was employed. Epirubicin alone acccounted for three-quarters of the drug costs. Between 11–13% of the cost was due to hospitalisation and the indirect costs constituted 26–30%. In the human capital approach setting, the production loss became the most important figure (77–81%).

Table III.  The calculation of cost in Euro (€) employing the FEC regimen. A dose intensity (DI) of 100%, 90% and 80% is implemented (€ 1 = 8.78 NOK).

	Cost (LIS)			Cost (Felleskatalogen)
Dose intensity	DI 100%	DI 90%	DI 80%	DI 100%	DI 90%	DI 80%
Fluorouracil	63	57	50	68	62	55
Epirubicin	1 226	1 103	981	1 438	1 294	1 123
Cyclo-phosphamide	122	110	97	126	113	101
Ondansetrone	82			116
Dexametasone	25			51
Saline	14			18
Dextrose	9			13
Infusion sets	33			33
Pharmacy	406
Outpatient cost	167
Transportation	428
Indirect cost	985^1
	9 850^2
Total cost^1	3 560	3 419	3 278	3 850	3 686	3 495
Total cost ^2	12 425	12 284	12 143	12 715	12 551	12 360

¹–Friction cost method.

²–Human capital approach.

A prior overview of the literature [4] revealed CMF saving 2.45 life years per patient treated. Although this study was performed back in 2000, we still believe the conclusion is valid. However, as most breast cancer groups today have left the CMF regimen and implemented the FEC regimen, international data from the very last years are almost absent. Calculating three dose intensity levels (80, 90 and 100%), a 5% improvement by FEC and employing the LIS agreement, the cost per life year saved ranged between € 7 150 and € 9 075 [(Cost FEC-cost CMF)/(2.45 LY*1.05-2.45 LY)]. Employing the recommended 3% discount rate by Murray and Lopez [22] and calculating the 5% benefit (0.12 years/patient treated) equally distributed over 20 years, the corresponding figures were € 9 404 and € 11 935. A sensitivity analysis varying discount rate (1, 3 and 5%) and survival gain (3, 5 and 10%) is shown in Table IV. The range in cost per life year saved was between € 3 575 and € 15 125.

Table IV.  Sensitivity analysis. All figures are in Euro (€).

	Dose intensity
Alternatives	100%	90%	80%
	9 075	8 114	7 150
Survival benefit
3%	15 125	13 523	11 916
5%	9 075	8 114	7 150
10%	4 538	4 057	3 575
Discount rate
1%	9 967	8 912	7 853
3%	11 935	10 672	9 404
5%	14 146	12 648	11 146

Discussion

Changing standard adjuvant therapy based on quality of life, disease free survival or overall survival data may introduce significant economic burdens to health care budgets as minor improvements may introduce significant health care costs. In this survey we have demonstrated that according to the literature, there are no benefits in terms of quality of life by altering standard therapy from CMF to FEC. The improved 5-year survival benefit is between 3–7%. By simply replacing the old drug methotrexate by epirubicin, the drug costs were significantly altered, but the therapy is still cost-effective with a raised cost per life year saved between € 3 575 and € 15 125.

The CMF was standard ACT for several decades before the anthracyline containing regimens took over the place of CMF. Especially the cardiotoxicity and the risk of secondary leukemias [4] of anthracyclines may have played a role in the delayed shift from CMF to FEC. However, there is today no doubt that the introduction of anthracyclines in ACT in breast cancer has improved survival [12]. An overview published in Lancet [23] in 2001 included 18 000 women and concluded the regimen especially beneficial to women below 50 years of age in terms of risk for relapse and death of disease.

In this survey, epirubicin was a major contributor to cost. Epirubicin is among the ”younger” drugs and there are reasons to believe the drug will become cheaper in the future. Other anthracyclines as doxorubicin may be employed to save money, but there is a clear risk of a raised frequency of heart failure employing doxorubicin. Thus money saved initially may easily be lost on early cardiac deaths or diminished quality of life.

In this model analysis, we calculated a cost of the FEC regimen (100% dose intensity) of € 1 411 (12 367 NOK) employing the LIS agreement. Looking at the refund from the NIA of NOK 2474/course, there is a balance between hospital drug cost and the NIA refund (6*NOK 2 474 = NOK 14 844). This may indicate that the figures employed reflect true costs.

We have calculated the cost of therapy when patients employ the nearest hospital in terms of geography. In Norway, patients are free to select hospital for their therapy. Raised transportation cost due to “free hospital selection” was mainly covered by the NIA. In the time period 1990–1998 the cost of transportation covered by the NIA increased by 57%. In the near future it will be of interest to see if the “free hospital selection” right will influence on the total cost.

In this study, the survival benefit was 3–7% employing the FEC instead of CMF and there were no difference in QoL. This benefit seems reasonable for a shift in standard therapy. Jansen and coworkers [24] clarified women's “benefit limit” to undergo ACT. At 1% improved survival, 50% of women would undergo ACT. At 5% improved survival, 72% of the women requested ACT.

In this survey, the cost-effectiveness analysis revealed a cost per life year saved ranging between € 3 575 and € 15 125. Several studies have pointed on reasonable cut-off limits for the introduction of new therapies or strategies in cancer care. One such Scandinavian study [25] argues for a limit somewhere between SKR 100 000 and 250 000 (about € 11 100–27 800). Our figures are below this limit and the cost-effectiveness of the FEC regimen is strongly supported.

Whereas today's ACT improves survival in breast cancer, there is still a room for improvements. Taxanes may be of interest in this setting as there are reports indicating an even better survival implementing these regimens [18], [26]. Monoclonal antibodies as trastuzumab (Herceptin®) may be another alternative coming up in this setting in the near future. However, these drugs will significantly raise the costs of ACT and there is without any doubt a need for health economic evaluations when decisions on the implementation of these drugs into daily practice are to be taken.

Conclusion

Adjuvant FEC chemotherapy in breast cancer does not improve quality of life, but improves 5-years survival rates by 3–7%. The total cost of FEC was estimated between € 3 278 and 12 715, depending on method chosen. The additional cost per life year by simply changing from adjuvant CMF to FEC was calculated ranging between € 3 575 and € 15 125. This figure is below reasonable cut-offs and strongly indicates adjuvant FEC cost-effective.