More than half of the colorectal cancer resections in South Africa have been done without a preoperative (confirmatory) biopsy, raising concerns about late presentation and deviations from standard diagnostic pathways, and highlighting the need for improved detection and access.
South African experts T MacQuene, M Lopez, N Sangana, N Parker, E Willson, H Garnie, G Lydon, J van Heerden, M C Gomperts, M Muchengeti, W C Chen, T Forgan, A Boutall and KM Chu unpack the findings in the SA Medical Journal.
They write:
Uninsured black and Asian patients, and other residents of certain provinces, face disproportionately higher risks of unplanned surgery, reflecting entrenched structural inequities in access to cancer diagnosis, care and treatment.
Addressing these inequities requires an integrated approach involving early detection, equitable access to diagnostic services and expansion of surgical oncology capacity. At the same time, development of a national colorectal cancer (CRC) screening programme and improving referral pathways should be key policy priorities to improve survival and reduce preventable mortality from the disease.
South Africa has one of the highest incidences of CRC in sub-Saharan Africa. Surgery is the mainstay of treatment and should be planned with appropriate staging and preoperative biopsy – a large proportion of unplanned colorectal cancer resections could indicate gaps in national oncology care.
Our national retrospective observational analysis aimed to estimate the proportion of unplanned CRC resections and to identify associated risk factors.
Operations without preoperative biopsy were used to estimate the proportion of unplanned resections; risk factors for these were modelled using multivariable logistic regression, and time to resection was reported in days from biopsy to resection.
Overall, 10 466 CRC resections were identified: 5 973 (57.1%) of them unplanned and 4 493 (42.9%) planned. On multivariable analysis, being uninsured (odds ratio (OR) 2.81, p<0.001), black (OR 1.83, p<0.001), or Asian (OR 1.26, p=0.021) were significantly associated with an increased risk of unplanned resection.
The median time to resection for the uninsured was much longer than for the insured for both colon cancer (26 v. 15 days, p<0.001) and rectal cancer (167.5 v. 95, p <0.001).
We concluded that most CRC resections were unplanned, highlighting the likelihood that many individuals present with surgical complications requiring emergency surgery
Common cause of cancer mortality
CRC is the second leading cause of cancer deaths globally, and the third most commonly diagnosed cancer. While rates in high-income countries (HICs) are four times higher than in low- and middle-income countries (LMICs), cases of CRC in LMICs are rising, mainly because of changes in demographic and lifestyle habits.
However, the apparent gap in incidence data between HICs and LMICs may partly reflect under-reporting in LMIC settings, where under-diagnosis, limited screening infrastructure and incomplete cancer registry coverage can lead to systematic under-estimation of the true disease burden.
Many HICs have implemented population-based screening with optimal care pathways, yet this has not occurred on a similar scale in LMICs, where CRC mortality rates are 17% higher.
Early detection is crucial, starting with screening. A sigmoidoscopy or colonoscopy confirms anatomical location and the diagnosis of CRC is made through endoscopic biopsy, allowing for better preoperative planning, especially with rectal malignancies, which often require neoadjuvant chemotherapy and radiation.
Globally, without appropriate screening or early detection, up to one-third of cases present with symptoms requiring acute or emergency (unplanned) surgical interventions. Even in HICs, up to one-third of CRC resections are emergent.
Operations done on individuals with CRC without preoperative biopsy are not standard of care, but may occur in extenuating circumstances, possibly in urgent settings such as secondary to large-bowel obstruction or perforation.
Although South Africa has one of the highest incidences of CRC in sub-Saharan Africa, there is no national screening programme. Our study aimed to estimate the proportion of unplanned CRC resections and assess time to surgical resection.
The secondary objective was to identify associated risk factors for unplanned resections.
Previous SA studies have been limited to single institutions or provinces. This is the first national retrospective observational study representing all CRC resections in both the private and public sectors: for our research we used pathology-confirmed CRC reports in the National Cancer Registry (NCR) from 2015 to 2020.
Variables recorded included age, sex, population group, province, year of diagnosis, public v. private sector laboratory, and a free text section. Cases from public labs were classified as medically uninsured, and those from the private ones as medically insured.
Age was categorised using a cut-off of 50 years to distinguish early- onset CRC from cases occurring at older age.
Findings
There were 10 466 CRC resections during the six-year study period, 5 973 (57.1%) unplanned and 4 493 (42.9%) planned. The median (interquartile range (IQR)) age was 64 (55 – 73), and 4 877 (46.6%) were female. For population group, 5 505 (53.6%) were white, 2 620 (25.5%) black, 1 559 (15.2%) mixed race and 586 (5.7%) Asian. Most were insured (n=8 267, 79.0%). Anatomical location was reported in 9 762 (93.3%) cases; 3 023 (28.9%) were rectal.
Surgical margins were reported in 9 616 (91.9%) cases. Positive surgical margins were found in 239 (2.3%): 166 (2.8%) for unplanned compared with 73 (1.6%) for planned resections (p<0.001).
Province was reported for public sector resections (n=2 549). The Western Cape performed most of these resections (n=1 207, 47.0%), followed by Gauteng (n=838, 32.9%) and the Eastern Cape (n=243, 9.5%). The proportion of unplanned operations by province ranged from 56.6% to 100%.
Associations with unplanned resection
On multivariable analysis, uninsured status (odds ratio (OR) 2.81, p<0.001), being black (OR 1.83, p<0.001), or being Asian (OR 1.26, p=0.021) were significantly associated with an increased risk of unplanned resection. Rectal cancer resections (OR 0.47, p<0.001) were significantly less likely to be unplanned.
Time to resection
Time to colon cancer resection was available for 818 (46.0%) of 1 777 planned cases. Median (IQR) time to colon cancer resection was 20 (9 – 43) days overall: 26 (12.5 – 49) days for uninsured and 15 (8 – 39) days for insured patients (p<0.001). Time to rectal cancer resection was available for 799 (45.05%) of 1 777 planned cases. Median (IQR) time to rectal cancer resection was 125 (26 – 217) days overall: 167.5 (57 – 237.5) days for the uninsured and 95 (19 – 201) days for the insured (p<0.001).
Discussion
This study, to our knowledge, is the first national analysis to report on CRC resection and biopsy patterns using two types of reports to estimate resections. More than half were performed without a preoperative biopsy, suggesting a potentially high proportion of emergency or non-standard-pathway operations.
There was significant variation by province, population group and insurance status, demonstrating inequity in access. Times to resection for colon and rectal cancer were significantly longer for the uninsured.
While we did not know the exact reasons for not having a preoperative biopsy, we surmised most of these cases might have been surgical emergencies, presenting with perforation, bleeding, or obstruction, but this could not be confirmed.
In a central hospital, 20% of CRC cases had malignant obstruction. More than half of CRC patients in SA present with a late-stage diagnosis, which is independently associated with higher postoperative mortality and morbidity, and lower long-term survival rates.
In addition, the higher frequency of positive margins in unplanned resections may reflect both later stage and limited preoperative planning, emphasising the importance of timely detection and structured surgical assessment to optimise R0 resection.
Overall, our findings highlight substantial challenges in the diagnosis and management of CRC in SA, particularly the high proportion of resections done without a preoperative biopsy.
These results suggest important gaps in access to diagnostic care within the existing health system.
To address this issue, population-based screening is needed. Previously cited barriers include limited infrastructure, insufficient endoscopic capacity, high costs of colonoscopy, shortages of trained personnel and pathologists, poor patient awareness, and cultural and socio-demographic obstacles.
In addition, delays to diagnosis and treatment for symptomatic individuals with CRC also occur.
We found population group and insurance status were independent risk factors for unplanned resection. An English study of a quarter of a million individuals with CRC showed that blacks, Hispanics and Chinese were at increased risk of emergency surgery.
Similarly, in our study, black and Asian patients showed an increased risk of unplanned resection. Marked inequities between those with and without health insurance were evident.
Even among the uninsured, there was geographical inequity in access to planned CRC operations.
There was a six-fold difference in absolute CRC resection numbers and a two-fold difference in the proportion of planned resections between provinces, suggesting that varying levels of urbanisation, health budget allocation and health-seeking behaviour could affect access to surgical oncology.
Patients with rectal cancer were more likely to have planned resections than individuals with colon cancer. There could be several reasons. Local symptoms of bleeding from rectal cancer might prompt earlier presentation; and histological confirmation for rectal cancers is easier than that for colon cancer: the latter requires colonoscopy, usually performed only by gastroenterologists and surgeons.
In contrast, 10% of rectal cancers can be diagnosed with digital rectal examination or rigid proctoscopy, which primary care doctors can perform. Also, primary care physicians can do rigid or flexible sigmoidoscopy to biopsy rectosigmoid cancers.
The median time to colon cancer resection was <30 days, even for the uninsured. Some treatment guidelines suggest colon cancer care should be initiated within 28 days of diagnosis, although a survival benefit has not been shown.
Rectal cancer treatment often begins with varying neoadjuvant chemoradiation regimens, so the longer median time to rectal cancer resection (125 days) was difficult to interpret.
We hypothesise that the insured had a shorter time to resection, possibly influenced by limitations in neoadjuvant options rather than reflecting a true delay to resection.
Study limitations
This study has several limitations: first was potential misclassification of non-biopsied individuals having emergency resection. It is possible, although less likely, that they had planned resection of non-histologically confirmed CRC, especially if oncological resection were surgically possible.
Preoperative biopsies may have been under-reported and/or not linked to the correct individual, thus over-estimating unplanned resections.
Second, reasons for the lack of preoperative biopsy were not known. For individuals without a preoperative biopsy, we assumed the resection was emergent; however, it is possible there were other reasons.
Geographical comparisons by province were limited to public sector data, so national comparisons are incomplete. Additionally, a potential bias is that the time-to-care results are mostly representative of the private sector.
Finally, the planned resection rate for rectal cancer may have been over-estimated by some emergent cases getting an initial diverting stoma without resection (and therefore not categorised as unplanned). This would have allowed delayed staging, biopsy, appropriate neoadjuvant therapy and planned resection, effectively moving them from the unplanned
to the planned cohort.
Despite these limitations, the large sample size enabled robust statistical analysis and has provided valuable insight into factors associated with unplanned CRC resection, as defined in this study by the absence of a preoperative biopsy.
These findings should therefore be interpreted within the context of this operational definition and do not imply a causal relationship between preoperative biopsy status and surgical planning.
The findings also highlight the urgent need for a national oncology care strategy in SA.
Integrating CRC screening with existing non-communicable disease platforms as well as with universal health coverage through the ongoing National Health Insurance planning offers a pragmatic step forward.
T MacQuene,1* BMedSci Hons, MSc ; M Lopez,2* MD ; N Sangana,2 MD ; N Parker,1 MB ChB ; E Willson,3 MB ChB ; H Garnie,3 MB ChB ; G Lydon,4 MD ; J van Heerden,5 BS ; M C Gomperts,6 BSc ; M Muchengeti,7,8,9 PhD ; W C Chen,7,10,11,12 PhD ; T Forgan,3 MMed (Surg) ; A Boutall,13 MB ChB, FCSSA ; K M Chu,1 MD
1 Centre for Global Surgery, Department of Surgical Sciences, Faculty of Medicine and Health Sciences, Stellenbosch University;
2 Department of Surgery, University of Texas Southwestern, Dallas, USA;
3 Division of Surgery, Faculty of Medicine and Health Sciences, Stellenbosch University;
4 Department of Surgery, Medical University of South Carolina, Charleston, USA;
5 Department of Surgery, Drexel University College of Medicine, Philadelphia, USA;
6 Department of Surgery, Indiana University School of Medicine, Indianapolis, USA;
7 National Cancer Registry, National Health Laboratory Service, Pretoria;
8 Division of Epidemiology and Biostatistics, School of Public Health, Faculty of Health Sciences, University of the Witwatersrand;
9 South African DSI-NRF Centre of Excellence in Epidemiological Modelling and Analysis (SACEMA), Stellenbosch University;
10 Strengthening Oncology Services Research Unit, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg;
11 Sydney Brenner Institute for Molecular Bioscience, Faculty of Health Sciences, University of the Witwatersrand;
12 Network for Oncology Research in Africa (NORA), Global Health Working Group, Martin-Luther-University, Halle-Wittenberg, Germany;
13 Colorectal Unit, Groote Schuur Hospital, University of Cape Town;
*Co-first authors
See more from MedicalBrief archives:
Long-term study focuses on colorectal cancer in young South Africans
Vital early signs of colon cancer in young adults – US study
Lack of basic care hastens colorectal cancer deaths in Africa – global analysis