2025/03/01 by Jenny F. Seligmann · 1 voice
Medicine · #Colorectal Cancer Surgical Treatments #Colorectal Cancer Treatments and Studies #Colorectal and Anal Carcinomas
paper · pdf · doi:10.1093/bjs/znaf047
Modern management of colorectal cancer is complex and includes surgical resection, local ablative procedures, radiotherapy, molecular subtyping, and targeted systemic treatment. The level of evidence demonstrating heterogeneity between colorectal tumours is high, the metastatic potential of tumours varies widely, and once metastatic disease is established outcomes remain poor. Conversely many of our cancer treatments incur risk of significant long-term morbidity, and even mortality. Optimal patient selection is therefore highly desirable to direct treatment to those most likely to benefit, and conversely to avoid in those who either do not require it or are unlikely to respond. The TNM classification describes the anatomic extent of cancer. The first guidance was produced in 1958 by the UICC providing a common international language to describe the staging of solid tumours. Today, TNM staging remains the globally accepted method for uniformly describing the extent of cancer spread and is considered standard communication of risk. Tumours are classified both pretreatment (clinical staging) and after surgical resection (pathological staging). Each component is subcharacterized by extent of tumour spread (T 1–4), number and location of nodes (N 0–2) and presence or absence of metastatic disease (M0–1). This information leads to subclassifications into stage (usually AJCC or Duke’s), informing the likelihood of survival1. Finally, TNM provides standardization for clinical trials, ensuring that interventions are tested in a homogeneous patient population. Most of our evidence-based treatments are available to patients who meet the same TNM criteria as per the trial. Over time, the TNM classification for colorectal cancer has evolved, with the latest iteration (TNM 8) now including presence of tumour deposits1. However, given the heterogeneity in outcomes demonstrated, complexity of therapeutic options available, and an increased understanding of tumour biology, does TNM remain fit for purpose in modern cancer care? Despite prognostic significance, a major limitation of TNM is reliably identifying the high-risk patient, both for clinical staging (preoperatively) and pathological staging. This is pertinent for all stages of colorectal cancer, but is particularly topical in the locally advanced setting, where neoadjuvant treatment can be considered to reduce risk of both local recurrence and systemic relapse2,3. For preoperative clinical staging of localised disease, neither CT nor MRI show optimal concordance with pathological staging4. This has implications for patient selection for neoadjuvant treatment, leading to concerns regarding ‘over’ and ‘under’ treatment. Furthermore, biology and imaging appear to interplay together in microsatellite instability-high (MSI-H) tumours and are more likely to show lack of concordance5. However, the major trials of neoadjuvant treatment have utilized radiological TNM for patient selection with positive results. Additionally, we understand that there is interobserver variation for elements of pathological staging, including the degree of tumour penetration, tumour budding, lymphocytic infiltration, vascular invasion, and in particular in newer elements of this assessment such as tumour deposit count6. TNM does not assess tumour biology, which provides both prognostic information (likelihood of treatment recurrence in localized disease and overall survival in metastatic disease), or the likelihood of treatment response to systemic therapy. This is critical information for clinicians and patients in the era of multiple therapeutic modalities available for colorectal cancer. Information regarding tumour biology can be obtained either from the tumour itself (mutations, overexpression, fusions) or from circulating free DNA in the blood. Potential clinical scenarios that would benefit from increased confidence in patient selection include selecting patients for adjuvant chemotherapy, identification of suitable patients for organ preservation or limited surgery in locally advanced rectal cancer, and identification of metastatic patients who would benefit from extensive surgery such as cytoreductive surgery for peritoneal metastases or liver transplantation for liver-limited metastatic disease. The most pertinent example of this is deficient mismatch repair (dMMR) or MSI status, which occurs in approximately 15–20% of locally advanced colon cancers, 3–5% of locally advanced rectal cancers, and 5% of metastatic colorectal cancers7. Testing was initially recommended to screen for Lynch Syndrome, but this is now also utilized to identify patients that may respond to immunotherapy. This has had transformational implications in dMMR metastatic colorectal cancer, with over 50% of patients alive 5 years following diagnosis when treated with immunotherapy8. Similarly in locally advanced colon and rectal cancer impressive tumour responses following neoadjuvant immunotherapy are increasingly raising the potential of non-operative management9,10, particularly crucial in rectal cancer with the potential to avoid radiotherapy and surgery and the associated long-term morbidity of these treatments. Other molecular markers of relevance in colorectal cancer include mutations in RAS, BRAF, and HER2. In metastatic colorectal cancer these markers predict benefit for targeted therapy and results are ideally acquired prior to treatment initiation11. However, they also provide prognostic information in the locally advanced setting and targeted treatments are increasingly investigated in the neodadjuvant and adjuvant trial setting in high-risk populations12. Unlike when screening for Lynch Syndrome, testing for MSI or dMMR for guiding neoadjuvant treatment requires a quick turnaround time. Increasingly this is performed by next-generation sequencing panel testing in central genomic hubs where information regarding tumour mutations is also provided. Testing on all initial colorectal cancer biopsies has advantages on turnaround times both for neoadjuvant/adjuvant decision-making, and in the metastatic setting by obliviating the need to retest tumour blocks on metastatic relapse. The higher costs of increased testing would likely be offset by increased efficiency. Another area of ongoing interest is the use of ctDNA to guide treatment decisions at various opportunities in the continuum of care. Examples include use for molecular testing at diagnosis instead of tissue, detection of minimal residual disease following surgery, and monitoring response to therapy13. There is intense interest in detecting minimal residual disease14. It is highly prognostic (more so than any element of TNM) and may highlight patients who would benefit most from adjuvant chemotherapy, or adjuvant molecularly targeted therapy15. Encouraging results for stage II disease has shown that a ctDNA-guided decision algorithm was non-inferior to standard of care when chemotherapy was withdrawn in ctDNA-negative patients16. However, this is not currently routine practice due to concerns regarding the sensitivity and specificity of assays (particularly false negative results) and requires further phase III trial evidence prior to routine withdrawal of adjuvant chemotherapy. Implementation would require a highly sensitive assay that had quick turnaround times and was cost-effective. Although sparing patients who are already cured by surgery needless chemotherapy is an important goal, outcomes should not be compromised. At present, TNM remains central to decision-making in the management of colorectal cancer; however, it has limitations which should be understood. Contemporary gold-standard management should include some understanding of tumour biology—at the very least, MSI/MMR status prior to commencement of treatment. Routine reflex next-generation sequencing panel testing and perhaps ctDNA will provide further precision, but implementation with acceptable turnaround times, standardized assays, and cost-effectiveness will be critical. In the era of modern multidisciplinary cancer care, multimodality assessment of risk beyond TNM will better allow us to tailor our treatments and provide the best advice for patients. The authors have no funding to declare. The authors declare no conflict of interest. No new data were generated or analysed in support of this research.