Cost of Treating Patients Who Undergo Resection of Colorectal Liver Metastases: An Australian Pilot Study

ABSTRACT

Aim

The cost of treating cancer, especially metastatic disease, can place a financial burden on patients, caregivers, and the healthcare system. A previous systematic review assessing the cost of treating metastatic colorectal cancer in the era of personalized medicine showed that this was approximately 300,000 USD. Surprisingly, there is little data on the cost of treating patients with colorectal liver metastases (CRLM). The aim of this study was to conduct a pilot cost‐of‐illness study on patients who undergo liver resection for CRLM in Australia.

Methods

This was a retrospective cohort study examining the direct medical costs to the healthcare system and out‐of‐pocket costs to patients who underwent resection for CRLM in 1 year (2018). These costs were calculated in 2022 Australian dollars from the time of the diagnosis of the liver metastases to the time of the last follow‐up.

Results

A total of 25 patients underwent resection in 2018, but only 16 received all their treatment at the Northern Campus. The median total cost of treating each patient was $328,773.59 (range: $63,244.40–$525,873.10). The highest cost was for medical oncology treatment, with a median of $90,228.87 per patient (range: $2087.06–$241,514.80), followed by liver surgery and consultations, $57,131.76 (range: $12,455.10–$319,178.90), and colorectal surgery and consultations, $55,990.29 (range: $7234.19–$200,755.60). Patients considered to be cured of their disease had a lower median cost than patients who had either died from their disease or were undergoing treatment for recurrent disease ($152,476.50 vs. $351,487.50, p = 0.030).

Conclusion

This study demonstrates that the cost of treating patients with CRLM in Australia is moderate but is considerably lower than many other cancer types, especially for those who are cured, reflecting good value care.

The cost of treating cancer places a heavy financial burden on patients, caregivers, and the healthcare system. This was a pilot study on the cost of treating colorectal liver metastases in the Australian healthcare system.

1. Introduction

The cost of treating cancer places a heavy financial burden on healthcare expenditure. The Australian Institute of Health and Welfare found that the total health expenditure on cancer was AUD $10.1 billion in 2015–2016, which accounted for 8.6% of overall disease expenditure [1]. Furthermore, the incidence of cancer in Australia is rising yearly due to multiple factors, including an ageing population, and lifestyle and environmental factors [1, 2]. The development of new and emerging technologies and treatments (e.g., targeted biological therapy) is providing more management options but also increasing costs. Although, overall, cancer survival is increasing, the fact that more people now live beyond their diagnosis and require ongoing care also contributes to rising costs [3, 4]. Consequently, there is a need for high‐quality data to inform decision‐making by policymakers, health administrators, and health economists on the current costs of cancer; to undertake cost‐effectiveness studies of potential interventions; and to plan for the likely future cost increases.

Australian healthcare is a mixed public and private system [5]. The public sector is government‐funded and includes public hospitals, medicines funded by the Pharmaceutical Benefits Scheme (PBS) and medical services funded by the Medicare Benefits Scheme (MBS). The private system includes services provided by private hospitals, private clinics, and private organizations and these are funded through partial government assistance, private health insurance agencies, and contributions from patients and their caregivers.

In Australia, colorectal cancer (CRC) is the third most common cancer and the third highest in total expenditure, after breast and non‐melanoma skin cancer [6]. A systematic review of the international literature examining the cost of CRC by phase of care showed that the initial phase (first 12 months from diagnosis) ranged from US$7893 to $60,289; the continuing annual phase ranged from US$2323 to $15,744; and the end‐of‐life phase (last 12 months before death) ranged from US$15,916 to $99,687 [7]. Furthermore, another systematic review found that the average cost of treating metastatic CRC is high, ranging from US$12,346 to US$293,461 [8]. The reasons for the wide range of costs in all phases and in metastatic CRC are because of differences in the location of the tumor (colon or rectum), the stage of disease, and variations in the cost of treatment in different countries because of different healthcare systems, payment methods, and treatment methodologies, and the use of relatively expensive targeted biological therapies in metastatic CRC.

The liver is the most common site of metastatic disease in CRC, and unfortunately, many patients either present with or develop metastatic disease at some stage after the initial diagnosis [9]. Only a select group of patients with colorectal liver metastases (CRLM) undergo liver resection with curative intent, although a relatively large proportion of these patients can return to a normal life. Over the past decade, there has been an evolution of treatment with progress in treating metastatic CRC, leading to better survival outcomes [10]. Selected patients sometimes undergo curative‐intent lung or peritoneal disease resection in combination with systemic chemotherapy and targeted biological therapies [11]. This has further contributed to overall treatment costs, as well as to increased patient expectations [12]. The current alarming rise in CRC diagnoses in young patients suggests that these financial challenges will only increase in the future. To date, no studies have assessed the medical costs of CRLM to the healthcare system and the patient out‐of‐pocket (OOP) costs in Australia.

We hypothesize that (i) both healthcare costs and OOP costs will be considerable for patients who undergo resection of CRLM but will vary based on disease recurrence and treatment outcomes, and that (ii) patients who are cured of their disease will likely incur lower costs than patients who develop recurrence. This pilot study aimed to determine the median medical cost to the healthcare system and patient OOP costs in those who underwent a potentially curative liver resection for CRLM. Secondary aims included describing the distribution of costs and identifying factors associated with costs.

2. Materials and Methods

2.1. Study Design

A single‐unit retrospective cohort pilot study of patients who underwent potentially curative resection for CRLM in the year 2018 at two hospitals (Royal North Shore Hospital and North Shore Private Hospital) was undertaken. Patients with at least 1 year of follow‐up were retrospectively identified from a prospectively collected database. Only patients who had their entire treatment on the Northern Campus were included in the analysis. Patients who received any treatment outside the Northern Campus, including chemotherapy and radiotherapy, were excluded because of difficulties obtaining information from the different centers.

2.2. Calculating Costs

A detailed patient timeline from the diagnosis of their liver metastases to December 31, 2021 was constructed, listing all public and private hospital admissions, general practitioner (GP) consultation visits, private consultation visits, imaging, blood tests, and chemotherapy and radiation therapy doses. Any treatment after December 31, 2021 was excluded from the analysis. Cost data for public and private hospitals were obtained from the respective hospitals. GP consultation, imaging, and blood test costs were estimated using the MBS, funded by the government. Costs for consultations and operations were assumed to be as per the Australian Medical Association (AMA) fees list, which is a guideline for specialists. For the sake of simplicity, surgical and procedural costs at the private hospital included only the primary surgeon's fees, and excluded the assisting surgeon and anesthetist's fees. Medical oncology treatment was defined as chemotherapy, immunotherapy, consultations, and hospital admissions due to complications, hydration, and port flushes. Both colorectal surgery and liver surgery treatments included the surgery as well as surgical consultations. All costs were adjusted for inflation using an inflation calculator (https://www.rba.gov.au/calculator/) and were presented in Australian dollars as of the year 2022.

2.3. Demographics and Treatment

Information obtained from the database included demographic data (age and sex), pre‐existing medical conditions (Charlson Comorbidity Index [CCI] and American Society of Anesthesiologists [ASA]), radiological investigations (timing, primary sidedness), pre‐operative treatments (chemotherapy +/− biological agents, radiotherapy, trans arterial chemoembolization [TACE], selective internal radiation therapy [SIRT], liver ablation, or portal vein embolization), intra‐operative details (operative technique, extent of liver resection, resection margin, tumor differentiation), post‐operative outcomes (complications, length of stay, recurrence), and post‐operative treatment (chemotherapy +/− biological agents, liver ablation, locoregional therapy). Neoadjuvant chemotherapy and biological agents were defined as treatment provided 6 months before the liver resection. Synchronous disease was defined as patients who had a diagnosis of primary CRC and liver metastases within 6 months. Metachronous disease was defined as a diagnosis of liver metastases greater than 6 months after the diagnosis of their primary CRC. The extent of the liver resection was defined as either major or minor, where a major liver resection involved the removal of at least three contiguous Couinaud segments.

2.4. Analysis

Categorical variables were presented as counts and percentages. Continuous variables were expressed as a mean and standard deviation for normally distributed data. Continuous variables not normally distributed were expressed as a median and range. Cost data were presented as a median and range, as these data were skewed and analyzed using the Mann–Whitney U test to assess if there were differences in the demographic and treatment factors. A p value < 0.05 was statistically significant. All statistical analyses were performed using Stata BE for Windows version 17.1 (StataCorp, College Station, Texas, USA).

3. Results

A total of 25 patients had a liver resection for CRLM in the year 2018. Of these, 16 received all their treatment at the Northern Campus and nine patients were treated on the central coast campus and were therefore excluded from this study. Demographic characteristics, tumor biology, and the treatment received are shown in Table 1. Most patients were men (63%), and the median age at the time of the operation was 62 years. The median follow‐up from CRLM diagnosis to death or December 31, 2021 to determine costs was 4.5 years (range 1.6–5.6 years). Most patients were either ASA 1 or 2 (88%) and had an ECOG performance of 0 (76%). Most patients had their primary tumors removed first (56%), were diagnosed with a Dukes C primary tumor (63%), had a left‐sided tumor (63%), presented with synchronous disease (88%), received neoadjuvant chemotherapy (63%), had a minor liver resection (75%), had a post‐operative complication (55%), received post‐operative chemotherapy (81%), and developed recurrent disease (69%).

TABLE 1.

Demographic characteristics, tumor biology, and treatment received.

Variable
Gender, n (%)
Male	10 (63)
Female	6 (37)
Median age at operation (range)	62 (44–79)
American Society of Anesthesiologists, n (%)
1	7 (44)
2	7 (44)
3	2 (12)
Median Charlson Comorbidity Index (range)	9 (6–10)
Eastern Cooperative Oncology Group performance status, n (%)
0	12 (76)
1	2 (12)
2	2 (12)
Primary resection details
Order of resection
Primary first	9 (56)
Liver first	1 (6)
Combined	6 (38)
Dukes stage
Dukes A Dukes B	1 (6) 5 (31)
Dukes C	10 (63)
Primary site
Colon	13 (81)
Rectum	3 (19)
Sidedness
Left	10 (63)
Right	6 (37)
Timing of liver metastases
Metachronous	2 (12)
Synchronous	14 (88)
Treatment prior to liver resection
Neoadjuvant chemotherapy	10 (63)
SBRT or chemoradiotherapy	1 (6)
TACE	0 (0)
SIRT	0 (0)
RFA or MWA	1 (6)
Portal vein embolization	0 (0)
Operative details (n = 20)
Operative technique
Open	20 (100)
Liver resection
Major Minor	5 (25) 15 (75)
Median length of stay in days (range)	7 (4–83)
Complications	11 (55)
Tumor differentiation
Well	1 (5)
Moderate	6 (30)
Poor	2 (10)
Not stated	11 (55)
Treatment and outcomes following liver resection(s) (n = 16)
Chemotherapy	13 (81)
SBRT	2 (12)
Chemoembolization/TACE	2 (12)
SIRT	5 (31)
RFA or MWA	3 (19)
Recurrence	11 (69)

Abbreviations: MWA, microwave ablation; RFA, radiofrequency ablation; SBRT, stereotactic body radiotherapy; SIRT, selective internal radiation therapy; TACE, trans arterial chemoembolization.

Table 2 summarizes the total cost of each intervention and the median cost per patient of each intervention. The median overall cost of treating each patient was $328,773.59 (range $63,244.40–$525,873.10). The highest cost was for medical oncology treatment, with a median total cost of $90,228.87 per patient (range $2087.06–$241,514.80), followed by liver surgery treatment, with a median cost of $57,131.76 per patient (range $12,455.10–$319,178.90), and colorectal surgery treatment, with a median cost of $55,990.29 (range $7234.19–$200,755.60) per patient. Table A1 in the Supporting Information consists of patients diagnosed with synchronous disease (n = 14), where the median overall cost of treating each patient was $340,765.70 (range $104,244.80–$525,873.10). The highest cost was for medical oncology treatment, with a median total cost of $94,845.98 per patient ($17,736.68–$241,514.80), followed by colorectal surgery treatment, with a median cost of $62,259.18 per patient (range $24,290.13–$200,755.60) and liver surgery treatment with a median cost of $60,438.15 per patient (range $17,736.68–$319,178.90). The broad range in costs can be explained by some patients who had recurrence, requiring ongoing chemotherapy and more frequent consultations with their medical oncologist. Similarly, some patients who had a reversal of their ileostomy incurred a higher cost in their colorectal surgery treatment or had multiple liver resections for recurrence and consequently incurred higher costs in their liver surgery treatment.

TABLE 2.

Cost breakdown by the type of intervention for all patients.

Intervention	Total cost of cohort (n = 16)	Median cost per patient (range)
Blood tests	69,011.58	4565.14 (707.11–9089.45)
Colorectal	1,123,866.97	55,990.29 (7234.19–200,755.60)
General practitioner	10,439.70	625.60 (351.90–1133.90)
Gastroenterology	53,151.98	2774.79 (541.56–11,148.25)
Imaging	239,026.25	12,317.01 (3055.20–28,618.79)
Liver	1,272,415.05	57,131.76 (12,455.10–319,178.90)
Medical oncology	1,449,162.20	90,229.87 (2087.06–241,514.80)
Radiation oncology	43,313.30	6949.69 (240.00–16,858.14)
Interventional radiology	199,259.46	30,303.57 (6052.84–43,076.04)
Cardiology and/or vascular	47,992.64	6124.48 (25.2–26,559.50)
Neurology	1073.43	1073.43 (1073.43–1073.43)
Orthopedics	29,176.21	14,588.11 (13,427.66–15,748.55)
Urology	410.09	410.09 (410.09–410.09)
Dermatology	3728.97	900.29 (543.63–2285.05)
Palliative care	136,108.58	23,102.73 (17,246.24–44,391.55)
Endocrinology	1465.30	732.65 (381.49–1083.81)
Respiratory medicine	378.28	378.28 (378.28–378.28)
Allied health	27,748.77	9685.00 (3033.92–15,029.85)
Psych	3307.11	3307.11 (3307.11–3307.11)
Overall	4,711,035.87	328,773.59 (63,244.40–525,873.10)

Table 3 examines the relationship between patient and disease factors and the overall median cost of treatment per patient. There were no differences in cost between genders, age (above and below 65), ASA, CCI, sidedness, primary site, the extent of the liver resection, having a post‐operative complication following liver resection, and recurrent disease. Patients with multiple liver resections had a higher overall cost of treatment than those who only required one liver resection ($376,707.0 vs. $214,379.0, p = 0.042). Patients who were considered cured of the disease (defined as not receiving any treatment for the last 4 years up to December 2024) had a lower overall cost than deceased patients or those who were receiving treatment for recurrent disease ($152,476.5 vs. $351,487.5, p = 0.030).

TABLE 3.

Median cost and range examining the relationship between cost and patient/disease factors.

Variable	Median cost (range)	p value
Gender		0.264
Male	262,355.0 (63,244.4–525,873.1)
Female	354,844.8 (145,386.3–524,038.9)
Age		0.221
<60	200,728.3 (104,224.8–347,966.0)
≥60	355,008.9 (63,244.4–525,873.1)	0.142
<65	228,029.8 (63,244.4–515,364.2)
≥65	361,723.6 (139,742.3–525,873.1)
American Society of Anesthesiologists		0.758
1	228,029.8 (139,742.3–525,873.1)
2 or 3	333,565.4 (63,244.4–524,038.9)
Charlson Comorbidity Index		0.999
<9	228,029.8 (104,224.8–525,873.1)
≥9	333,565.4 (63,244.4–524,038.9)
Eastern Cooperative Oncology Group performance status		0.953
0	280,797.6 (63,244.4–525,873.1)
1 or 2	335,973.9 (145,386.3–391,690.4)
Sidedness		0.562
Left	276,005.8 (63,244.4–525,873.1)
Right	351,487.5 (139,742.3–524,038.9)
Primary site		0.439
Colon	347,966.0 (63,244.4–525,873.1)
Rectum	200,728.3 (150,467.6–323,981.8)
Liver resection		0.743
Major	347,966.0 (139,742.3–391,690.4)
Minor	228,029.8 (63,244.4–525,873.1)
Complications		0.267
Yes	228,029.8 (63,244.4–5240.38.9)
No	333,565.4 (200,728.3–525,873.1)
Multiple liver resections		0.042
Yes	376,707.0 (347,966.0–525,873.1)
No	214,379.0 (63,244.4–524,038.9)
Recurrence		0.377
Yes	347,966.0 (63,244.4–525,873.1)
No	200,728.3 (104,224.8–524,038.9)
Cured from the disease		0.030
Yes	152,476.5 (63,244.4–228,029.8)
No	351,487.5 (139,742.3–525,873.1)

4. Discussion

The healthcare costs for treating patients who undergo liver resection for CRLM are reasonable in Australia compared to other cancer types and compared to other health jurisdictions. Most patients had synchronous disease at diagnosis, and the highest contributors to the costs were medical oncology treatment, followed by liver surgery treatment and colorectal surgery treatment. Medical oncology treatment costs were the highest due to the prolonged treatment patients received when they had recurrent disease. Specific factors that led to the highest costs were having multiple liver resections and having recurrent disease requiring further treatment. The ongoing treatment included either surgery, radiotherapy, chemotherapy, and/or immunotherapy, as well as palliative care services. Not surprisingly, these patients had higher overall costs than those who were cured of their disease.

International studies assessing the cost of treating CRC in China, Vietnam, and Serbia found similar results where medical oncology costs were the highest, followed by surgery and radiotherapy [13, 14, 15]. Despite targeted therapy not being used routinely, these studies demonstrated that medical oncology costs, specifically the cost of chemotherapy, including managing the side effects of this treatment, superseded all other costs. Ananda et al. also conducted a study in Australia in 2016 examining the cost of treating CRC in the public healthcare system and found that with the addition of targeted therapy, the average cost of treating metastatic CRC was $81,000 per person [16]. In Australia, costs are at least partially covered by government funding, while in many other countries, some treatment options are unaffordable because they need to be self‐funded. As other treatments are developed, this will likely further increase costs, and this will likely result in inequity in healthcare in some countries. In the literature, costs vary considerably between different countries, possibly due to different treatment algorithms, the availability of different treatment options, or differences in funding sources and the local costs required to fund these treatments.

This study found that the cost of treating CRLM appears to be reasonable in Australia, especially in patients who were cured of their disease. Goldsbury et al. examined costs related to the treatment of multiple different cancer types in New South Wales, Australia [17]. CRC treatment costs were lower than the costs of treating hematological cancers such as leukemia and lymphoma due to the high cost of chemotherapy treatment. Similarly, Merollini et al. conducted a state‐wide assessment of the cost of different cancers in Queensland, Australia, finding that the cost of treating brain and pancreas cancer was double the cost of treating CRC and that irrespective of costs, these patients had the worst prognosis of all the cancers [18]. This study also found that hematological cancers such as myeloma and leukemia are more expensive than CRC due to the prolonged treatment of chemotherapy.

In this study, patients who survived treatment of their CRLM incurred overall lower costs than those who succumbed to the disease. The latter group required repeat or ongoing treatment for disease progression in the form of chemotherapy, immunotherapy, radiotherapy, and occasionally further surgery. The additional consultations, blood tests, imaging, and diagnostic procedures associated with ongoing treatment also contributed to the higher costs. The present study did not assess the health‐related quality‐of‐life (HRQoL), although we have previously examined this in long‐term survivors of CRLM, and most patients have good long‐term HRQoL and minimal ongoing symptoms after a curative liver resection [19]. Residual symptoms at long‐term follow‐up are mostly fatigue due to underlying comorbidities and age or due to ongoing neuropathy because of previous systemic chemotherapy. A limited number of studies have assessed long‐term HRQoL in patients who have undergone liver resection for CRLM. Wiering et al. and Rees et al. found that long‐term survivors had excellent HRQoL, while those with recurrence had poorer HRQoL [20, 21]. After curative intent resection of CRLM, patients can achieve a 50% 5‐year overall survival, which is relatively high compared to other gastrointestinal cancers [22]. This appears to justify the overall cost of treating advanced CRC in Australia.

4.1. Limitations and Directions for Future Research

The limitations of the present study should be considered, including the fact that the analysis of costs began from the diagnosis of liver metastases rather than from the diagnosis of the primary CRC. This means that the costs for some patients who already had treatment for their primary disease would not have been captured. The cancer pathway for each patient was mapped retrospectively to determine the treatment and interventions, and there may have been some missing consultations. Costing data were assumed to be based on rates set by the AMA guidelines for specialist fees. In contrast, imaging, blood tests, and GP consultations were assumed to be based on Medicare fees, possibly leading to an underestimation of some costs. This study also excluded the costs of the anesthetist and assisting surgeon because these data were unavailable. Furthermore, because the study included a cohort of patients in a relatively affluent part of Australia where many patients have private health insurance, the findings may not be representative of other health jurisdictions in Australia. Lastly, the sample size is relatively small and consists of a group of patients treated by a single liver surgeon. However, as noted previously, this is a pilot study and the first to assess the total cost of treating patients who have undergone liver resection for CRLM in Australia.

This study provides a foundation to understand the true medical costs of treating patients who have undergone liver resection for CRLM. While this is relevant for the mixed public‐private healthcare model studied, there are still implications for a more general application. These include helping policymakers allocate appropriate funds and resources, as well as identifying the cost drivers associated with treating CRLM, particularly in the era of personalized medicine, with newer and more expensive targeted therapies entering the market.

While this is a pilot study, a larger multicenter study is required to verify these results and identify other drivers of cost that could not be deciphered from this cohort. Furthermore, the findings are important for other reasons, including health service planning, equity and access, patient counseling, and value‐based care.

Important next steps that might arise from this pilot study include further research into this field, changes in clinical practice based on considering the financial impact of treatment decisions, and equity‐focused reforms, particularly for underprivileged patients.

5. Conclusion

This study demonstrates manageable costs for treating patients who undergo liver resection for CRLM in Australia. The highest cost was for medical oncology treatment. Patients who were cured of their disease incurred lower costs than those who required ongoing treatment for recurrence or who succumbed to their disease. Understanding these costs may assist policymakers in better allocating resources and provide patients and families with the opportunity to plan for their future. As Australia has a complex healthcare system and healthcare costs, especially for cancer, are rising, this study may stimulate further discussion about the costs of all cancer types, especially those with unfavorable outcomes.

Author Contributions

Nazim Bhimani: conceptualization, data curation, formal analysis, investigation, methodology, resources, software, writing – original draft. Rebecca Seton: conceptualization, supervision, writing – review and editing. David L Chan: conceptualization, supervision, writing – review and editing. Mbathio Dieng: conceptualization, supervision, writing – review and editing. Patrick J Kelly: conceptualization, supervision, writing – review and editing. Thomas J Hugh: conceptualization, supervision, writing – review and editing.

Funding

The authors have nothing to report.

Ethics Statement

Ethics approval was given by the Northern Sydney Local Health District Human Research Ethics Committee (2019/ETH12206).

Disclosure

This manuscript has not been a podium or poster meeting presentation.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table A1: Cost breakdown by the type of intervention for synchronous disease patients

Data Availability Statement

De‐identified data may be obtained by requesting the corresponding author.