Cost-minimisation analysis of rectal cancer neoadjuvant chemoradiotherapy based on fluoropyrimidines (capecitabine versus 5-fluorouracil)

Abstract

Objectives

The current standard treatment for patients with rectal cancer stage II–III is neoadjuvant chemoradiotherapy followed by surgery. Neoadjuvant chemoradiotherapy can be performed with 5-fluorouracil (5-FU) or capecitabine (CPC) considered to be equivalent therapies. Medication cost is higher for CPC than for 5-FU, however, the administration of continuous 5-FU intravenous infusion is related to other costs such as those associated with outpatient facilities or central venous catheter insertion.

Methods

This retrospective study analysed the direct sanitary costs associated with the treatments and their complications from a hospital perspective. Costs in patients treated with 5-FU or CPC were measured between January 2010 and July 2018 at Mataró Hospital. The aim of this study was to perform a cost-minimisation analysis between the two treatments. We aimed to assess the cost associated with the complications related to each drug and the economic impact of applying the most efficient option.

Results

Ninety-eight patients were analysed: 32 were treated with CPC and 66 with 5-FU. Treatment cost was significantly higher for 5-FU than for CPC (2560.86±99.17 and 563.10±9.52 respectively, P=0.0001). No significant differences were found in the costs associated with treatment complications between groups (148.21±934.91 and 41.41±102.50 euros respectively, P=0.322).

Conclusions

Considering the clinical equivalence shown in the available trials and previous reviews, the most efficient treatment is neoadjuvant chemoradiotherapy with CPC. Complications associated with the treatments did not significantly modify these results. Other studies gave similar results both in the neoadjuvant and adjuvant context, reaffirmed in this study.

Introduction

Rectal cancer constitutes nearly one-third of colorectal cancer cases and a high number of cancer-related deaths around the world.¹ The incidence of rectal cancer in the European Union is 125,000 cases per year with a mortality rate of 4–10/100,000 inhabitants per year and a mean age of diagnosis of 70.² Establishing local and systemic disease control is the main objective of solid tumour oncology in the curative setting.³ Rectal cancer is characterised by its difficult surgical approach which complicates the possibility of achieving negative surgical margins after resection. After rectal resection for early rectal cancer (T1–2, N0), oncology outcome is positive, whereas for locally advanced rectal cancer (stage II–III) the loco-regional recurrence rate is around 15% to 65%. Even with total mesorectal excision, stage III local regional recurrence is around 20% to 30%, neoadjuvant chemoradiotherapy treatment being needed for locally advanced patients.⁴ Currently, optimal treatment in patients with stage II–III rectal cancer is neoadjuvant chemoradiotherapy followed by surgery, and, if necessary, adjuvant chemotherapy.² While combined treatment with chemoradiotherapy plus surgery reduces loco-regional recurrences, poor overall survival has been reported in large randomised clinical trials which report a slightly positive effect on 5-year survival rates.⁵ Chemoradiotherapy before surgery can reduce tumour mass, block tumour invasion, increase tumour resection and anus retention rates, and reduce iatrogenic dissemination during operation, and local recurrence rates by up to 6%.⁴ This strategy achieved pathologic complete response (pCR, defined as no residual cancer found on histological examination of the specimen) in 15% to 27% of patients. The 5-year survival rate in patients who achieved pCR after chemoradiotherapy was higher than in those who did not (83% vs 66%, P<0.0001) as was the 5-year distal metastases-free survival rate (89% vs 75%, P<0.0001).⁶ The optimal neoadjuvant approach has been evaluated in several clinical trials. The NSABP R-03 (National Surgical Adjuvant Breast and Bowel Project R-03) showed an increase in 5-year disease-free survival and overall survival for patients who underwent neoadjuvant chemoradiotherapy 8 weeks prior to surgery compared with those ones who underwent adjuvant therapy at least 4 weeks after surgery (64.7% vs 53.4%, P=0.011% and 74.5% vs 65.6%, P=0.065; respectively).⁷ The long-term results of the trial performed by the CAO/AROAIO-94 (Working Group of Surgical Oncology/Working Group of Radiation Oncology/Working Group of Medical Oncology of the German Cancer Society) showed an improvement in the 10-year cumulative incidence of local relapse in patients who underwent neoadjuvant chemoradiotherapy compared with adjuvant treatment (7.1 vs 10.1% P=0.048).⁸ In addition, continuous infusion of 5-fluorouracil (5-FU) proved superior to treatment with 5-FU bolus and has become the chemotherapeutic treatment of choice in the neoadjuvant context.⁹ Later, the National Surgical Adjuvant Breast and Bowel Project R-04 study showed non-inferiority of Capecitabine (CPC, prodrug of 5-FU) compared with continuous infusion with 5-FU.¹⁰ These results were reaffirmed by the meta-analysis of five retrospective studies, two prospective studies, and two randomised studies performed by Zou et al in 2017.¹¹ Adverse effects for both regimens were gastrointestinal-related (eg, diarrhoea and mucositis), haematologic (leukocytopenia and neutropenia) and hand-foot syndrome episodes. With the available evidence, neoadjuvant chemotherapeutic treatment schemes with oral CPC or 5-FU continuous infusion during 5 weeks in concomitance with radiotherapy are considered the standard treatment. Both presented equivalent results regarding efficacy.

Improvements in prevention, diagnosis, and treatment of cancer have been accompanied by major increases in healthcare costs. Currently, the economic impact of these new interventions together with the limited benefits they have provided are being evaluated with increasing interest by healthcare financers and institutions.¹² In this context, pharmacoeconomic models have been proposed as a better strategy to that of others such as restricting access to treatments.¹³

In our hospital, the procedure based on continuous 5-FU along with radiotherapy has been used for many years. The procedure based on CPC has been frequently used during recent years due to the current available evidence. Oral administration of CPC might improve patients’ quality of life although there is no literature to support this hypothesis. However, oral treatment reduces costs associated with outpatient facilities, insertion, and complications related to implantable venous-access port, and sterile pharmacy preparations.

This study aims to measure all factors that could condition the efficiency of both treatments. The main objective of this study was to perform a cost-minimisation analysis comparing the neoadjuvant chemotherapy with CPC or 5-FU. We aimed to calculate the savings that could be generated by applying the most efficient therapy for each patient. As secondary objectives, we aimed to: perform a cost evaluation associated to the adverse effects of each alternative using a retrospective cohort of patients in our centre and discuss how this could affect the results of the primary objective; and assess annual budget impact of using the most efficient option for the total of treated patients in our centre over 1 year.

Methods

An observational, longitudinal, and retrospective cost-minimisation study was performed comparing chemoradiotherapy with CPC and 5-FU in neoadjuvant therapy for stage II–III rectal cancer. Current clinical and economic data was used. The time frame analysed went from the beginning of the chemoradiotherapy treatment to the surgical intervention day.

Patients

All patients who had been treated with neoadjuvant chemoradiotherapy based on CPC or 5-FU for stage II and III rectal cancer between January 2010 and July 2018 at Mataró Hospital (Barcelona, Spain) were included. Patients were identified using the statistical module of Oncofarm software.

Demographical, anthropometric, and clinical data

Data on age, sex, and body surface area of patients was collected. The presence of the following comorbidities was obtained: heart failure or ischaemic cardiopathology; arterial hypertension; dyslipidaemia; diabetes mellitus; chronic renal failure; previous neoplasm; cirrhosis; AIDS; asthma; and COPD. Data on the stage (II or III) of rectal cancer was obtained.

Data on treatment and complications

Regarding treatment, the following data was obtained for each patient: received total doses; number of days of treatment (for CPC) or number of cycles (for 5-FU); dose reduction during treatment; and early discontinuation of treatment for any reason. Received doses were obtained through Oncofarm software. The following data was obtained to analyse the impact of treatment complications: number of visits to the emergency department due to treatment-related complications; number of urgent visits to outpatient facilities due to treatment-related complications; number of admission days due to treatment-related toxicities; number of additional oncology medical visits due to treatment-related toxicities; and number of additional medical visits to other specialists due to treatment-related toxicities. Data on complications was obtained from the patient medical records.

Study economic methodology

The study was performed following the minimisation-cost analysis methodology. Direct healthcare costs from a hospital perspective were analysed. The following sanitary resources were measured: drugs; visits to oncology; pharmacy visits; CPC pharmacy dispensing; 5-FU preparation in biological safety cabinet; sessions in outpatient facility; totally implantable venous-access port insertion and maintenance for patients treated with 5-FU; the device for continuous infusion of 5-FU; visits to the emergency department; urgent visit to outpatient facility; additional visit to oncology; additional visit to another specialist; and hospitalisation days. Costs were measured in euros and updated to the value of 2018.

Calculation of medication, treatment, and complications costs

This study calculated medication, treatment, and complications costs separately. Medication cost was calculated based on the received total doses of drugs and their cost. For the treatment cost we added all the costs related to preparation, dispensing, and administration of the drug. Thus, costs of activities related to totally implantable venous-access port insertion and maintenance, medical visits, pharmacy visits and pharmacy dispensing, preparation of 5-FU in a biological safety cabinet, nursing care and sessions in outpatient facility were added to the medication costs. In the case of complications, cost of emergency visits, urgent visits to outpatient facility, days of hospitalisation, additional visits to oncology, and additional visits to another specialist were measured.

Costs data

Drugs and devices for continuous infusion costs were obtained from costs corresponding to 2018. Visits to all facilities, hospitalisation costs, and Port-a-cath insertion and maintenance were obtained from the published public health fees of the Catalan Health Service “Orden SLT/30/2013, de 20 de febrero, del Servicio Catalán de la Salud”.¹⁴ Pharmacy visits and pharmacy dispensing costs were calculated based on billing data of our hospital. The cost of the 5-FU preparation in biological safety cabinet was obtained accounting for the fact that preparation lasts 5 min. It was calculated 5 min of pharmacy technical staff (1.55 euros considering the company cost) and the cabinet cost (0.19 euros considering the cabinet cost of 22 329 euros with a 15% annual amortisation divided into 365 days per year and considering a use of 5 out of the 240 min that the biological safety cabinet was used per day).

Statistical analysis

Quantitative variables are reported as the mean (SD, ±) and dichotomous variables as the number of events (percentage, %). The χ² test was used to compare categorical variables and the t-student test to compare the continuous variables. Those variables that followed a non-normally distribution were compared using the Mann-Whitney U test. Differences were considered significant for P<0.05 values. The statistical analysis of the data was performed using the SPSS-PC software.

Results

Epidemiological, anthropometric, and clinical data of patients

A total of 98 patients were analysed. They were treated with neoadjuvant chemoradiotherapy for rectal cancer stage II–III at the Mataró Hospital from January 2010 to July 2018. Sixty-six patients were treated with 5-FU and 32 with CPC. No patients were removed from the analysis due to data loss or any other reason. Table 1 broadly describes the epidemiological, anthropometric, and clinical features of both groups. A higher percentage of patients treated with CPC had an initial diagnosis of stage III rectal cancer. Twenty-nine (90.60%) patients had an initial diagnosis of stage III disease in the case of CPC and 43 (65.20%) in the case of 5-FU. No differences in comorbidities were found at the beginning of the treatment between both groups.

Table 1.

Patient characteristics

Variable	5-fluorouracil	Capecitabine	P value
No. of patients	66	32
Age, years
Mean (SD)	61.47 (10.47)	63.09 (10.07)	0.468
Sex, no. of patients (%)
Male	47 (71.20)	24 (75.00)	0.694
Female	19 (28.80)	8 (25.00)
Body surface area
Mean, square metres (SD)	1.81(0.16)	1.81(0.17)	0.881
Clinical stage, no. of patients(%)
II	23 (34.80)	3 (9.40)	0.007
III	43 (65.20)	29 (90.60)
Comorbidities, no. of patients (%)
Ischaemic cardiopathology /heart failure	3 (4.50)	1 (3.10)	1
Arterial hypertension	30(45.45)	14 (43.75)	0.874
Dyslipidaemia	26 (39.40)	13 (40.60)	0.907
Diabetes mellitus	10 (15.20)	3 (9.40)	0.537
Chronic renal failure	2 (3.00)	2 (6.30)	0.595
Previous neoplasm	1 (1.50)	0 (0)	1
Cirrhosis	0 (0)	0 (0)
AIDS	0 (0)	0 (0)
Asthma or COPD	4 (6.06)	4 (12.50)	0.432

AIDS, acquired immune deficiency syndrome; COPD, chronic obstructive pulmonary disease; no, number.

Treatment

In the 5-FU group, the mean dose was 15 130.08 mg whereas in the CPC group the received mean dose was 107,914.06 mg. Patients treated with 5-FU underwent a mean of 4.95 1-week treatment cycles (complete treatment is five cycles). Patients treated with CPC underwent a mean of 37.16 days of treatment depending on the radiotherapy duration and early interruptions due to intolerance of treatment. In the 66 patients treated with 5-FU, one dose adjustment and one early interruption during treatment were described. In the 32 patients treated with CPC, two dose adjustments and four early interruptions during treatment were described.

Medication cost

The cost of one vial of 5000 mg of 5-FU was 6.69 euros, the mg cost was 0.0013 euros, given that all vials were reused in the biological safety cabinet. CPC has two different presentations, 150 mg with a cost of 0.08 euros per tablet and 500 mg with a cost of 0.21 euros per tablet. In the group of patients treated with CPC, the usage ratio of 500 mg and 150 mg tablets was 4:1, giving a calculated cost of 0.000466 euros per mg of CPC. The mean medication cost was significantly lower for those patients treated with 5-FU than for those treated with CPC (19.67±3.57 and 50.29±9.51 euros respectively, P=0.0001). These results are described in tables 2 and 3.

Table 2.

Medication and treatment costs, 5-fluorouracil

5-fluorouracil, variable	Resource consumed (mean per patient)	Monetary value (euros)	Mean cost per patient (euros)
Medication resource consumption and cost (SD)	15 130.08 mg	0.0013/mg	19.67 (3.57)
Visits to oncology	4	118	472
Totally implantable venous-access port insertion	1	506.85	506.85
Totally implantable venous-access port maintenance	2	183	366
Sessions in outpatient facility	5.93	183	1086.90
Preparation on safety cabinet	4.95	1.74	8.62
Device for continuous infusion	4.95	20.35	100.82
Treatment cost (SD)			2560.86 (99.17)

mg, milligrams.

Table 3.

Medication and treatment costs, capecitabine

Capecitabine, variable	Resource consumed (mean per patient)	Monetary value (euros)	Mean cost per patient (euros)
Medication resource consumption and cost (SD)	107 914.06 mg	0.000466/mg	50.29 (9.51)
Visits to oncology	4	118	472
Pharmacy visits	1	32.93	32.93
Pharmacy dispensing	2	3.94	7.88
Treatment cost (SD)			563.10 (9.52)

mg, milligrams.

Treatment cost

Unitary resources consumed for each of these categories along with their costs are described in tables 2 and 3. The mean treatment cost was significantly higher for those patients treated with 5-FU than for those treated with CPC (2560.86± 99.17 and 563.10±9.52 respectively, P=0.0001) even including the medication costs.

Treatment-related complications cost

There were no significant differences in the number of urgency visits, urgent visits to outpatient facility, days of hospitalisation, additional visits to oncology, and additional visits to another specialist between both groups (table 4).

Table 4.

Treatment-related complications of 5-fluorouracile and capecitabine

Variable	Capecitabine (no. of events)	5-fluorouracil (no. of events)	P-value	Monetary value, euros
Visits to the emergency department	6	2	0.170	151
Urgent visits to outpatient facility	1	4	0.583	183
Additional visit to oncology	1	2	0.980	118
Additional visit to another specialist	1	2	0.610	118
Hospitalisation (days)	0	18	0.322	536 (first 5 days) 425 (sixth day and later)

Unitary resources consumed for each of these categories (amount of visits and days of hospitalisation) along with their costs are described in table 4. The mean cost associated with complications related to 5-FU was 148.2±934.91 euros and to CPC was 41.41±102.50 euros. Although there was a tendency towards higher costs related to treatment-associated complications with 5-FU, no statistically significant differences were found between both costs (P=0.322). Cost results for both groups in this study are summarised in table 5.

Table 5.

Comparative costs of 5-fluorouracile and capecitabine

Variable	5-fluorouracil	Capecitabine	P-value
Medication cost Mean (SD), euros	19.67 (3.57)	50.29 (9.51)	0.0001
Treatment cost+ medication cost Mean (SD), euros	2560.86 (99.17)	563.10 (9.52)	0.0001
Cost of associated complications Mean (SD), euros	148.21 (934.91)	41.41 (102.50)	0.322

Budgetary impact on the implementation of the treatment associated with lower costs

The mean cost of treatment with CPC was 1997.76 euros less than the cost with 5-FU per patient without taking into account the associated costs of complications. This could involve a cost reduction of up to 19 500 euros yearly for an average hospital that treats approximately 10 patients per year and much higher for bigger hospitals.

Discussion

The cost of medication was more than double in patients treated with CPC in this study. Despite this, when the treatment total cost was considered, the cost was significantly higher for the group treated with 5-FU compared with the group treated with CPC (2560.86± 99.17 vs 563.10±9.52 euros; P=0.0001). Complications cost did not show significant differences between both groups although a tendency towards an increase in these costs for patients treated with 5-FU was shown. Although a higher number of dose reductions and early interruptions were described in patients treated with CPC, they always occurred in the last days of treatment and did not condition complication results.

Epidemiologic, anthropometric, and comorbidity data was comparable for both groups although a higher number of patients affected by a stage III disease in the group treated with CPC was found. Despite this, a posterior statistical analysis confirmed that differences in staging did not confuse results in costs.

A literature search on costs related to both these neoadjuvant therapies for rectal cancer allowed us to compare our results to those of other researchers. A study carried out in Australia and published in 2015 showed a mean saving of 1042 Australian dollars per patient when treated with CPC. This study showed an increase in both treatment costs and those associated with secondary complications for those patients treated with 5-FU.¹⁵ A recent cost-utility study carried out in Thailand in the adjuvant context also showed higher efficiency for CPC treatment compared with 5-FU plus leucovorin. The authors also described a higher cost for CPC when medication cost was considered.¹⁶ In the case of advanced or metastatic colorectal cancer, important resource-use savings,¹⁷cost savings,^18–20and societal cost savings²⁰ for patients treated with CPC compared with 5-FU has been reported in several settings and from different perspectives (provider, third payer, and society).^17–20

The estimation of costs associated with a particular treatment is an important issue for health policy decision-making. Economic evaluations in healthcare aims to compare the costs and the consequences on health of various alternatives (treatments, diagnostic tools, programmes, and so on). This enables counting both the clinical benefits and costs/resources of several alternatives to select the most efficient. Cost minimisation analyses are necessary in order to increase efficiency in our clinical practice, especially oncology where medication is very expensive. When two treatment options are equivalent from a clinical perspective, these studies should be taken into account when selecting the optimal treatment.

Clinical equivalence of both treatments has been demonstrated in previous clinical trials and meta-analysis.¹¹ The results of this study showed that the most efficient treatment was neoadjuvant chemoradiotherapy with CPC. Cost reduction involved in the treatment with CPC could reach up to 19 590.9 euros annually for a medium-size hospital treating approximately 10 patients per year.

Treatment-associated complications or other factors related to the study’s internal validity did not modify this study’s results. Other studies have shown similar results in other settings both in neoadjuvant and adjuvant contexts. This study’s results confirm these results in our setting. At a time when new clinically effective therapies are continually appearing, it is very important to take into account pharmacoeconomic studies in order to prioritise the most efficient interventions. The evaluation and posterior inclusion of the most efficient therapies should consider before other strategies such as restrictions to new treatments or co-payments of treatments.

What this paper adds.

What is already known on this subject

• The current standard treatment for patients with rectal cancer stage II–III is neoadjuvant chemoradiotherapy based on 5-fluorouracil or capecitabine in concomitance with radiotherapy followed by surgery.

• 5-fluorouracil and capecitabine are considered equivalent therapies according to non-inferiority studies and meta-analyses.

What this study adds

• The results of this study show that the most efficient treatment is neoadjuvant chemoradiotherapy with capecitabine taking into account direct healthcare costs from the hospital perspective.

• Complications cost before surgery did not show significant differences between patients treated with 5-fluorouracil or capecitabine.

Data availability statement

All data relevant to the study are included in the article. Because of the sensitive nature of the data collected for this study, requests to access the dataset from qualified researchers trained in human subject confidentiality protocols may be sent to the corresponding author.

Ethics statements

Patient consent for publication

Not required.