International Region of Birth and Long‐Term Outcomes for Patients Undergoing Colorectal Cancer Resection in an Australian Population

ABSTRACT

Background

Overseas birth is associated with a survival advantage in colorectal cancer in Australia. We wished to analyse this survival based on global region of birth, controlling for cancer‐specific and other causes of death.

Methods

A database of resected colorectal cancers from 2010 to 2016 (n = 1596) was grouped according to patient global region of birth. Chi‐squared testing was used to compare factors including patient demographics and AJCC stage. Kaplan–Meier and competing‐risk analysis were used to compare 5‐year survival outcomes between patients born in different regions, and regression analysis was used to control for age.

Results

751 (47.1%) patients were born overseas. 385 (24.1%) originated from Europe, 124 (7.8%) from East and SE Asia, 92 (5.8%) from the Middle East, and the remainder from other global regions. Immigrants from East or SE Asia were more likely to present with node‐positive disease (p = 0.048) than those born in Australia. Immigrants from East and SE Asia and the Middle East had significantly better all‐cause 5‐year survival than patients born in Australia (73.4% and 80.4% vs. 60.4%, p < 0.0001). Immigrants from the Middle East retained their cancer‐specific survival advantage after competing risk analysis (HR 0.76, p = 0.027). Immigrants from Europe displayed no significant difference in all‐cause or disease‐specific survival compared to individuals born in Australia.

Conclusion

Patients born in the Middle East appear to have a colorectal cancer‐specific survival advantage compared to those born in Australia, unrelated to stage at presentation. This has significant implications for prognosis and future research directions.

Australian‐born patients demonstrated the poorest 5‐year overall survival (52.3%) compared to immigrants born in East and South‐East Asia (66.9%), Middle East (73.9%), Europe (58.7%) and New Zealand and Pacific Islands (67.2%) for all colorectal cancer stages (log‐rank p < 0.001).

1. Introduction

Ethnicity and global region of birth have been extensively examined regarding their impact on long‐term outcomes after resection for colorectal cancer. In both the United States of America (USA) and United Kingdom (UK), African‐American and Afro‐Caribbean heritage have been associated with more advanced presentation stage and poorer 5‐year survival [1, 2, 3]. However, the confounding effect of relative socio‐economic deprivation makes it difficult to determine the impact of patient race alone [1, 2, 3]. Immigrants to the USA have also been shown to have poorer participation in colorectal cancer screening and subsequent worse survival [4, 5]. The proposed reasons for this include poorer health literacy and access to colonoscopy, which have significant implications for strategic preventative health‐care planning.

Australia and New Zealand have the highest per capita lifetime risk of developing colorectal cancer worldwide (7.38% [7.27–7.50]) [6]. In 2020, 29.7% of Australia's population was born overseas and it is this group that displays a significantly improved 5‐year survival for colorectal cancer compared to Australian‐born residents [7, 8]. The reasons for this are complex; a large population registry study from New South Wales (NSW) identified that patients born in Lebanon, the Philippines, and the UK were more likely to present with distant metastases than Australian‐born patients, and yet higher rates of 5‐year survival were demonstrated for all regions of overseas birth except New Zealand [9]. Our research group previously demonstrated that Australian place of birth was an independent risk factor (HR 1. 50, p < 0.001) for death in a cohort of patients undergoing surgery for colorectal cancer in Western Sydney [10]. This study aims to further investigate the effect of global region of birth on stage at presentation and survival for colorectal cancers undergoing resection in our Western Sydney population.

2. Methods

A database of consecutive colorectal cancer resections (the CONCRETE project) from January 2010 to October 2016, processed by the Institute for Clinical Pathology and Medical Research (ICPMR) of NSW Health Pathology was analysed for data extraction. The database was constructed from synoptic histopathology records and oncology medical correspondence from two tertiary cancer care centres, using a standardised extraction form. Surgery was carried out in six hospitals in Western Sydney, three public and three private. The study was designed following the STROBE checklist for cohort, case–control, and cross‐sectional studies [11]. Approval was sought and obtained for all centres by our local ethics committee (HREC LNR/16/WMEAD/398). Cause of death and survival from the date of surgery up to the 5th of December 2019 was calculated from data obtained from the NSW Registry of Births, Deaths, and Marriages. All‐cause and colorectal cancer‐specific mortality was calculated, along with median follow‐up. Exclusion criteria included trans‐anal or submucosal only resections and operations for recurrent disease.

Information relating to patient country of birth was obtained from hospital registration and then categorised into the following global regions: Australia and Australian territories, New Zealand and the Pacific Islands, East and South‐East Asia, the Middle East and Europe. Patients who did not originate from these territories (e.g., those from the Indian sub‐continent, Africa or the Americas) were felt to represent too small and heterogenous a group to examine; thus, they were not analysed (n = 88). For patients born in Australia Information on Aboriginal or Torres Strait Islander identification was also recorded. Socioeconomic deprivation was assessed using postcode information derived from the Australian Bureau of Statistics (ABS) Index of Relative Socio‐Economic Disadvantage 2016 [12].

Descriptive statistical analysis was performed for each of the five global regions of birth regarding demographics, histopathology, and American Joint Committee on Cancer (AJCC) stage. Student's t‐test and Chi‐squared testing were used to assess for differences between those born in Australia (the largest single group) and other regions of birth. Aboriginal and Torres Strait Islander patients were also compared using univariate analysis to other non‐indigenous Australian patients. Cox‐proportional hazard regression was used to determine independent risk‐factors for disease‐specific mortality, and multivariate regression analysis was performed to analyse and correct for median age differences across the patient groups. Kaplan–Meier and log‐rank testing were used to assess for differences in colorectal cancer‐specific and all‐cause mortality between patient groups born in Australia and the other global regions of birth being analysed. The five overseas‐born groups of interest then had a competing risk analysis of their 5‐year survival performed, using deaths from other causes as a competing risk against colorectal cancer‐specific mortality. This analysis was performed using the competing risk survival analysis package version 3.1 by Therneau [13]. Age‐specific and competing risk analysis was then performed to correct for non‐colorectal cancer‐related mortality. A p‐value of 0.05 was defined as the level of statistical significance. All statistical analysis was performed using SPSS version 27, and R Version 4.0.0.

3. Results

1507 eligible patients who underwent colorectal cancer resection were identified in our database originating from 88 countries (Table 1). Patients born in Australia (n = 845, 56.1%) were significantly older than all other groups except those born in Europe. East and South‐East Asian patients (n = 124, 8.2%) were more likely to present with AJCC stage III or IV disease as well as with perineural invasion. Patients of Middle Eastern birth (n = 92, 6.1%) were more likely to reside in the most socioeconomically deprived postcodes.

TABLE 1.

Demographic and histopathological features of patients by region of birth.

Region of birth	Australia n = 845 (56.1%)	Australian Aboriginal or Torres Strait Islander n = 19 (2.2% of Australian‐born group)	New Zealand & the Pacific Islands n = 61 (4.0%)	East & South‐East Asia n = 124 (8.2%)	The Middle East n = 92 (6.1%)	Europe n = 385 (25.5%)
Median age (years)	67.7	61.4 p = 0.005 a	60.6 p < 0.001 a	64.3 p = 0.007 a	64.3 p = 0.021 a	72.5 p < 0.001 a
Gender (male)	421 (49.9%)	8 (42.1%)	40 (65.6%)	68 (54.8%)	47 (51.1%)	227 (59%)
			p = 0.018 b	p = 0.303 b	p = 0.826 b	p = 0.003 b
Colonic	556 (66.4%)	15 (78.9%)	41 (67.2%)	84 (77.8%)	55 (59.8%)	159 (67.4%)
Rectal	281 (33.6%)	4 (21.1%)	20 (32.8%)	40 (32.3%)	37 (40.2%)	124 (32.3%)
		p = 0.092 c	p = 0.279 b	p = 0.199 b	p = 0.415 b	p = 0.503 b
AJCC stage
I and II	413 (48.9%)	4 (21.1%)	29 (47.5%)	46 (37.1%)	43 (46.8%)	197 (51.3%)
III	286 (33.8%)	6 (31.6%)	22 (36.1%)	54 (43.3%)	33 (35.9%)	127 (33.1%)
IV	146 (17.3%)	9 (47.4%)	10 (16.4%)	24 (19.2%)	16 (17.4%)	60 (15.6%)
		p = 0.004 c	p = 0.872 b	p = 0.048 b	p = 0.445 b	p = 0.743 b
Perineural invasion	172 (20.6%)	10 (52.6%)	13 (22%)	36 (29.5%)	28 (31.1%)	84 (22.5%)
		p = 0.002 c	p = 0.935 b	p = 0.08 b	p = 0.069 b	p = 0.62 b
Poorly differentiated tumour	178 (21.3%)	7 (36.8%)	16 (27.6%)	23 (19%)	20 (22%)	73 (19.5%)
		p = 0.093 c	p = 0.263 b	p = 0.56 b	p = 0.884 b	p = 0.463 b
Lymphatic invasion	243 (29.2%)	14 (73.7%)	13 (21.7%)	37 (30.3%)	27 (30%)	97 (26.1%)
		p < 0.0001 c	p < 0.0001 b	p = 0.793	p = 0.87	p = 0.27
Vascular invasion	178 (21.4%)	9 (47.4%)	16 (27.1%)	32 (26.2%)	21 (23.3%)	79 (21.2%)
		p = 0.005 c	p = 0.303 b	p = 0.229 b	p = 0.671 b	p = 0.933 b
Residence in bottom SEIFA quintile for deprivation	160 (19%)	8 (42.1%)	15 (24.6%)	24 (19.7%)	47 (52.2%)	88 (23%)
		p = 0.009 c	p = 0.291 b	p = 0.87 b	p < 0.0001 b	p = 0.108 b

Abbreviations: AJCC, American Joint Committee on Cancer; SEIFA, Socioeconomic Index for Areas.

^a T‐test of mean age of group versus mean age of patients born in Australia (control).

^b Chi‐squared testing of overseas‐born versus all Australian‐born patients.

^c Chi‐squared of Aboriginal or Torres Strait Islander versus all other Australian‐born patients.

2.2% (n = 19) of all Australian‐born patients identified as Aboriginal or Torres Strait Islander, which is consistent with the overall Australian percentage [14]. Regarding patients who identified as Aboriginal or Torres Strait Islander, almost half had evidence of metastatic disease at the time of surgery, and over half demonstrated lymphatic and perineural invasion on histopathology. 42.1% of this population resided within the lowest quintile for socioeconomic status for postcode compared with 19% of the entire Australian‐born cohort overall (p = 0.009).

On Cox regression analysis, Australian birthplace was demonstrated to be a significant predictor of both all‐cause and cancer‐specific mortality (p < 0.0001, OR 1.496, 95% CI 1.18–1.79), along with other well‐recognised predictors of mortality including AJCC stage and poor tumour differentiation (Table 2). On multivariate regression analysis correcting for age, patients from East and South‐East Asia, New Zealand, and the Pacific Islands maintained a significant survival advantage per year of age. Patients from the Middle East demonstrated no such survival advantage, while patients born in Europe had a non‐significant trend towards poorer survival per year of age (Table 3).

TABLE 2.

Cox‐regression analysis of independent predictors of cancer‐specific mortality.

	p	Hazard ratio	95% CI
Increasing age in years	< 0.001	1.011	1.004–1.02
Born in Australia	< 0.001	1.496	1.18–1.79
Residence in the most socioeconomically deprived quintile	0.169	1.107	0.95–1.55
Increasing tumour size in mm	0.001	1.007	1.002–1.01
Poor tumour differentiation	0.024	1.312	1.02–1.63
Perineural invasion	< 0.001	1.479	1.19–1.81
Lymphatic invasion	< 0.001	1.527	0.65–1.08
Vascular invasion	< 0.001	1.703	1.39–2.22
< 12 nodes examined	< 0.001	1.677	0.97–0.99
AJCC Stage	< 0.001	2.163	1.84–2.42
Macroscopic tumour rupture	0.033	1.298	1.01–1.6
Tumour infiltrating lymphocytes	0.251	0.861	0.65–1.08
Tumour site	0.582	0.93	0.72–1.2

TABLE 3.

Multivariate regression model results for all‐cause mortality on patients from different birth regions compared to patients born in Australia correcting for age.

Region	95% CI	Hazard ratio	p
East & South‐East Asia	0.052–0.139	0.085	0.014
Middle East	0.310–0.678	0.458	0.434
Europe	3.789–12.280	6.821	0.055
New Zealand & the Pacific Islands	0.071–0.214	0.214	0.036

Stage for stage, patients born overseas maintained a significant 5‐year survival advantage over those born in Australia with stage II, III, and IV disease (Figure 1a–c). Kaplan–Meier and log‐rank testing demonstrated that patients born in East and South‐East Asia and the Middle East had a significantly better 5‐year survival than those born in Australia (Figure 2). Using a competing‐risk model, comparing deaths from colorectal cancer with all other causes of death, patients born in Australia maintained a cancer‐specific survival disadvantage (HR 1.26, 95% CI 1.04–1.53, p = 0.021) (Figure 3a). On sub‐group analysis of different birth regions, patients born in the Middle East retained a survival advantage, with a hazard ratio of 0.76 (95% CI 0.59–0.97, p = 0.027) (Figure 3b). Patients who originated from East and South‐East Asia displayed a significant survival advantage regarding non‐colorectal cancer related causes of death compared to those born in Australia (HR 1.6, 95% CI 0.44–0.89, p = 0.014) (Figure 3c). Individuals born in New Zealand and the Pacific Islands retained a non‐significant advantage on competing‐risk analysis of non‐cancer specific survival, whilst those born in Europe appeared to have no significant survival advantage for either (Supporting Information S1 and S2).

FIGURE 1.

(a) Log‐rank testing and Kaplan–Meier analysis, 5‐year survival of Australian‐born (64.8%) and overseas‐born (77.2%) patients with stage II colorectal cancers (p = 0.01, 95% CI 4.23–4.46). (b) Log‐rank testing and Kaplan–Meier analysis, 5‐year survival of Australian‐born (50.3%) and overseas‐born (65.6%) patients with stage III colorectal cancer (p < 0.001, 95% CI 3.74–4.02). (c) Log‐rank testing and Kaplan–Meier analysis, 5‐year survival of Australian‐born (11.6%) and overseas‐born (13.7%) patients with stage IV colorectal cancer (p = 0.05, 95% CI 1.8–2.21).

FIGURE 2.

Log‐rank testing and Kaplan–Meier analysis, 5‐year survival of Australian‐born (52.3%), East and South‐East Asian‐born (66.9%), Middle Eastern‐born (73.9%), European‐born (58.7%), and New Zealand and Pacific Islander‐born (67.2%) patients with all cancer stages (p < 0.001, 95% CI 3.71–3.88).

FIGURE 3.

(a) Competing risk analysis of colorectal cancer‐specific death and all other‐cause death in Australian born versus all overseas‐born patients. Colorectal cancer‐specific death in Australian‐born patients HR 1.26, p = 0.021, 95% CI 1.04–1.53. All other cause death in Australian‐born patients HR = 1.27, p = 0.016, 95% CI 1.05–1.54. (b) Competing risk analysis of colorectal cancer‐specific death and all other‐cause death in Australian‐born versus Middle Eastern‐born patients. Colorectal cancer‐specific death in Australian‐born patients HR 1.32, p = 0.027, 95% CI 1.1–1.31. All other‐cause death in Australian‐born patients HR 1.8, p = 0.012, 95% CI 1.09–3. (c) Competing risk analysis of colorectal cancer‐specific death and all other‐cause death in Australian‐born versus East and South‐East Asian‐born patients. All other‐cause death in Australian‐born patients HR 1.31, p = 0.014, 95% CI 1.24–1.41. Colorectal cancer‐specific death p = 0.21.

4. Discussion

Our findings demonstrate a complex relationship between country of birth, age, socioeconomic status, and colorectal cancer survival in the Western Sydney population. In this group, patients born overseas have a colorectal cancer‐specific survival advantage over their Australian‐born counterparts. On sub‐group analysis, those originating from the Middle East maintain this cancer‐specific survival. This seems to have occurred despite patients from the Middle East demonstrating no significant differences in their stage at presentation or other adverse histological factors, and no difference in their overall survival corrected for age. One striking feature of this group is their socioeconomic profile, with over half of these patients residing in the most disadvantaged quintile by ABS scoring in the studied population. This may contribute to the anomaly we described in our previous work, which failed to find any difference in survival between the most and least disadvantaged patients in our cohort, despite a significantly increased risk of late presentation in the latter group [10].

Compared to Australian‐born patients, those originating from East or South‐East Asia had a significantly decreased risk of death from non‐colorectal cancer‐related causes but displayed no significant advantage in terms of cancer‐related death. This may be due to their significantly higher risk of presenting with node‐positive disease despite a significantly reduced risk of death per year of age added. In contrast to this, patients born in Europe had no significant difference in cancer‐specific and all‐cause mortality, despite being significantly older at diagnosis.

Rather than demonstrating a survival advantage for those born overseas, our findings illustrate a significant survival disadvantage in those patients in our group who were born in Australia, particularly with stage III disease. In our cohort, the 5‐year all‐cause survival for Australian‐born patients with stage III disease is only 50.3%, which contrasts sharply with the national relative 5‐year survival of 71.3% for stage III colorectal cancers provided by the Australian Institute of Health and Well‐Being [15]. Although the total number of patients who identified as Aboriginal or Torres Strait Islanders in our cohort was too small to perform any detailed analysis on, the much higher proportion of those presenting with metastatic disease (47.4% vs. 17.3%) is in keeping with findings from elsewhere around Australia [16]. Lack of access to services, embarrassment about symptoms, and fear or distrust of the healthcare system were all identified as factors contributing to delayed diagnosis in Indigenous Australians [17].

The reasons for the disparity between Australian born residents and those born overseas are unclear but are possibly due to a combination of factors. The ‘healthy immigrant paradox’ has been described in population‐level health care for some time and relates to health advantages enjoyed by immigrants to a particular country when compared to the native‐born population. This effect is attributed to pre‐immigration healthy habits and life‐style choices maintained by emigres, health screening processes implemented by destination countries, and the ‘salmon effect’ where immigrants sometimes return to their place of origin to die [18]. This has been demonstrated in immigrant groups in Australia and is associated with a modest improvement in disease‐free survival in a variety of cancers in countries with a similar public health care system [19, 20].

Another proposed contributing factor to the disparity between overseas and Australian‐born patients could be variation in their enteric microbiome. This has been shown to have significant global variety [21] and the predominance of certain species such as Fusobacterium has been associated with poorer oncological outcomes [22]. Ethnic or racial variation in the tumour microenvironment could also play a role, and there is an increasing body of evidence to suggest that racial differences exist in the lymphocytic immune response to colorectal cancer [23, 24, 25]. The degree of this response has been demonstrated to have a strong correlation with improved survival [26]. Unlike other health‐care systems such as those in the USA and the UK, NSW Health does not routinely collect information on patients' racial heritage other than country of birth, with the exception of Aboriginal and Torres Strait Islanders. This understandably makes it difficult to draw any definite conclusions on racial differences in our population, particularly when considering regions of global birth which may have considerable racial and ethnic heterogenicity, such as New Zealand and the Pacific Islands or the Middle East.

European‐born patients were our largest subgroup of patients born overseas and the only group significantly older than the Australian‐born population. One assumption we have made is that due to the historical immigration patterns to Australia, older native‐born Australians in our group are perhaps more likely to be of Caucasian Northern European extraction, which may partially explain their similar 5‐year survival to European‐born patients. Whilst diets are becoming increasingly globalised, Western diets are an established risk factor for colorectal cancer which both groups have in common [27]. Additionally, barriers to seeking health care such as language may not be as significant a concern for European‐born patients as in other subgroups. In one study performed in NSW, patients from the United Kingdom presented with later stage colorectal cancers without adversely affecting survival [9].

Socioeconomic deprivation was previously shown not to affect survival despite being associated with node positive, circumferential, and perforated colorectal cancers in our Western Sydney population [10]. A meta‐analysis of Australian studies of socioeconomic status and colorectal cancer demonstrated a highly variable effect on stage IV disease as well as survival [28]. Many included studies found no relationship, whilst large population series were more likely to demonstrate reduced survival associated with socioeconomic deprivation. Country of birth and socioeconomic status are certainly related; 53% of the lowest socioeconomic group was born overseas, and the protective effect of Middle Eastern birth found in the present study may explain similar survival in this group. The presence of patient subgroups that present with higher stage cancers but experience similar survival suggests there may be reduced participation in screening alongside appropriate access to care once a diagnosis has been made. Uptake of faecal occult blood testing has been shown to be lower in migrants to Australia as well as patients with low socioeconomic status [29, 30]. Further work will need to be performed to determine the current level of participation in screening programs for our population.

Aside from a lack of information on patient race, our study also suffers from other deficits common to retrospective analyses. We were unable to easily access information on whether our patients received post‐operative adjuvant therapy or not, and as our database was comprised of resected colorectal cancer specimens, we also did not have any information on patients who did not undergo surgery due to incurable disease or other health issues. It has recently been shown that more information on potential barriers to cancer care for ethnic minorities is needed in the Australian setting [31].

Despite the limitations of our retrospective study, we are confident that this paper highlights a significant disparity in long‐term outcomes after resection for colorectal cancer between patients with differing regions of birth despite being treated in a country with fairly universal health care. This has significant implications for both prognostication and identifying populations at risk of early death. More work is needed to determine the underlying causes for this anomaly, and Australia, with its racially and ethnically diverse population, may be able to provide some of these answers.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1.

Pasch J. A., Fok K. Y., Goltsman D., Premaratne C., and MacDermid E., “International Region of Birth and Long‐Term Outcomes for Patients Undergoing Colorectal Cancer Resection in an Australian Population,” ANZ Journal of Surgery 95, no. 7‐8 (2025): 1517–1524, 10.1111/ans.70160.