Keywords
overall survival - toxicity - drug metabolism - geriatric patients - ctDNA - MRD
Introduction
Adjuvant chemotherapy (CT) has shown survival benefit in stage 3 and selected stage 2 colon cancers.[1] Initially, fluoropyrimidines were given and later addition of oxaliplatin was found to be useful (Mayo and Degramont).[2] In the MOSAIC trial, the standard arm of fluorouracil plus leucovorin (LV5FU2) was compared to the study arm where oxaliplatin was added (FOLFOX4) in the adjuvant setting for stage II and III colon cancer.[3] Among 2,246 patients, the 6-year overall survival (OS) was 78.5% in the FOLFOX4 arm and 765 in the standard arm of LV5FU2 (p = 0.046). The 5-year disease-free survival (DFS) was also better for the study arm (73.3% for FOLFOX4 and 67.4% for LV5FU2 group patients; p = 0.003). And the 6-year OS rates were 78.5% with FOLFOX4 and 76.0% with LV5FU2 (p = 0.046). In the FOLFOX4 arm incidence of grade 3 peripheral sensory neuropathy was 1.3% at 12 months after treatment and 0.7% at 4 years. This data cemented the value of FOLFOX4 in the adjuvant management of stage II or III colon cancer.
However, protocols with more drugs and a longer duration, leads to significant toxicities affecting quality of life (QoL), the classic example being protocols containing irinotecan that failed. More recent studies are designed to replace intravenous infusions with user-friendly oral-based regimens and shorten the duration of adjuvant CT. We present an updated review for the current management of adjuvant therapy of colon cancer.
Three versus Six Months Adjuvant Therapy
Six months of adjuvant CT has been the standard of care in colon cancer, duration of which would be either 8 cycles of 3 weekly CAPOX or 12 cycles of 2 weekly FOLFOX.[3]
[4] The toxicities associated with longer duration, especially cumulative neurotoxicity with oxaliplatin, is troublesome to most patients.[5]
[6] Reducing the duration of adjuvant CT reduces toxicities. It also helps economically (especially the health care budget) in countries where the funding of treatment is undertaken by the government. Other than doing dose modifications for dose-limiting toxicities the scenario remained unchanged—until the International Duration Evaluation of Adjuvant Chemotherapy (IDEA) study results came in 2017.[7]
[8] The pooled analysis of six randomized controlled trials done concurrently included data from SCOT (U.K., Denmark, Spain, Sweden, Australia, New Zealand), TOSCA (Italy), Alliance/SWOG80702 (U.S., Canada), IDEA (France), ACHIEVE (Japan), and HORG (Greece).[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16] This included 12,834 patients with stage 3 colon cancer. A total of 40% of the patients received CAPOX and the remaining 60% FOLFOX. Three months of CAPOX was noninferior to 6 months (hazard ratio [HR] = 0.95; 95% confidence interval [CI], 0.85–1.06) but 3 months of FOLFOX was inferior to 6 months (HR = 1.16; 95% CI, 1.06–1.26). Grade 2 to 4 toxicities (especially neurotoxicity and diarrhea) were lesser in the 3 months arm as expected. Patient compliance and completion of planned therapy was also more in the 3 months arm. The results allowed us to divide patients into low-risk colon cancer group (T1 to T3 and N1; 60% of patients) and high-risk colon cancer group (T4 and/or N2).
The TOSCA noninferiority phase 3 randomized clinical study evaluated 3 versus 6 months of FOLFOX (fluorouracil, leucovorin, and oxaliplatin) or CAPOX (capecitabine plus oxaliplatin) in the adjuvant setting for patients with high-risk stage II resected colorectal cancer (CRC). The 5-year relapse-free survival (RFS) data was in 1,254 evaluable patients. A total of 776 patients (61.9%) received FOLFOX and 478 (38.1%) received CAPOX. The 5-year RFS was 82.2% for the 3-month arm and 88.2% for the 6-month arm (p = 0.86 for noninferiority), as confirmed in the pooled analysis.[17] For FOLFOX, the data supported the longer duration of 6 months. This was associated with neurotoxicity being five times higher in the 6 months arm. The final recommendation was to use 3 months of CAPOX or 6 months of FOLFOX, when opting for oxaliplatin doublet adjuvant therapy.
The phase 3 SCOT randomized clinical trial had also compared 3 versus 6 months of CT in the adjuvant setting. This was also for high-risk stage II or stage III nonmetastatic CRC patients with potentially curative surgery. The key difference was that patients were also analyzed for early or late commencement of the adjuvant therapy (with respect to surgery—less than 6 weeks vs. more than 6 weeks). Of the 5,719 patients, 914 were in the early-start group and 4,805 were in the late-start group. If adjuvant CT was commenced more than 6 weeks after surgery, DFS (primary endpoint of the study) was worse (p = 0.01) ([Table 1]).[18]
Table 1
What should be the duration of adjuvant therapy—evidence from published trials[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
|
Trial
|
Common regimen
|
3 y DFS
3 vs. 6 mo
|
Additional remarks
|
|
SCOT
|
CAPOX
|
76.7% vs. 77.1%
|
Noninferior
CAPOX- noninferior
FOLFOX- not noninferior
Neuropathy more in 6 months
|
|
TOSCA
|
FOLFOX
|
81.1 vs. 83
|
6 months better
|
|
ACHIEVE
|
CAPOX
|
79.5 vs. 77.9%
|
3 months for low risk
6 months for high risk
|
|
IDEA
|
mFOLFOX6
|
72.1 vs. 75.7
|
6 months superior to 3 months
|
|
HORG
|
CAPOX
|
77.2 vs. 77.9
|
6 months better in high risk
|
Abbreviation: DFS, disease-free survival.
The sustained benefit of CAPOX poststoppage could be due to adherence to intended dose intensity and higher dose of oxaliplatin (130 vs. 85 mg/m2).
The IDEA investigators and the ESMO committee divided patients into “fighters” (only prepared to accept a 1–2% reduction in efficacy when using 3 months adjuvant therapy) or “fatalists” (prepared to accept a > 2% reduction in efficacy). Suggested treatment recommendations (in four scenarios) are shown in [Table 2]. In 2020, updated results of the IDEA data showed 5 year OS rate was 82.4% in the 3 months arm and 82.8% in the 6 months arm (absolute 5-year OS difference of –0.4% [95% CI –2.1 to 1.3%]). Hence, the recommendations remained unchanged.
Table 2
Management recommendations for fatalists versus fighters
|
High-risk fatalist
3 months CAPOX
|
N2 high-risk fighter
Majority 3 months CAPOX
Minority 6 months CAPOX
|
|
T4 high-risk fighter
6 months CAPOX/FOLFOX
|
Low-risk fighter
3 months CAPOX
|
CAPOX versus FOLFOX
CAPOX and FOLFOX are the two commonly practiced adjuvant CT regimes in colon cancer. Preference between the two varies on logistics and institutional practice majority of the times. Despite being only two common regimens there are no good randomized controlled trials comparing the two, especially using uniform dosing of 5FU, oxaliplatin, or capecitabine. The advantages of CAPOX regime are oral route, can be given on day care basis, no continuous infusions, no central lines or port insertions requirement, reduced risk of thrombophlebitis, or coronary vasospasm. There are differences in toxicity between the two regimens.[19]
[20] While patients on FOLFOX had increased mucositis and neutropenia, CAPOX was associated with increased diarrhea and hand–foot syndrome.[21] Aitini et al did a cost analysis study to do an economic comparison between CAPOX and FOLFOX.[22] They found that central line resulted in 70% cost in infusional regimes and one cycle of CAPOX was 53% cheaper than FOLFOX. Hence, in resource-limited countries CAPOX would be more economically feasible. Impact on QoL should also be taken into consideration based on unique features of individual patients and their preferences.[23]
A recently published systematic review looked at 543 published articles and analyzed the 29 ones that met their criteria.[24] These collectively included 7,028 patients. In general, the results of the eligible studies indicated that adjuvant therapy after resection of colorectal liver metastases led to improved RFS/DFS rates, but this benefit did not contribute to a statistically significant prolongation of OS.
Patient Selection
As we have seen, about a third of patients with resectable/early CRC relapse in spite of receiving current standard of care. On the other hand, several potentially cured cases might be subject to toxicities of unnecessary therapy. Hence, there is a need to find better means of identifying high-risk patients that require adjuvant CT.
Are Warburg Categories Useful?
The Netherlands Cohort Study involved 1,793 patients.[25] This highlighted the importance of using Warburg categories to identify high-risk patients. Warburg-low CRC did not benefit from adjuvant CT (HRCRC-specific 1.07; 95% CI 0.76–1.52, HRoverall 0.95; 95% CI 0.70–1.30). On the other hand, Warburg-moderate CRC (HRCRC-specific 0.64; 95% CI 0.47–0.86, HRoverall 0.61; 95% CI 0.47–0.80) and possibly Warburg-high CRC (HRCRC-specific 0.86; 95% CI 0.65–1.14, HRoverall 0.82; 95% CI 0.64–1.05) had survival benefit from adjuvant therapy.
What if the Patient Had Obstructive CRC and Has Already Undergone Stent Placement?
Matsuda et al reported on 129 such patients in a retrospective study.[26] The 3-year RFS between those that did not get adjuvant CT (56.4%) and those that did (78.5%) was in favor of giving adjuvant CT even in this subgroup (p = 0.003). The data was stronger for patients with high-risk features (high carcinoembryonic antigen, T4, and lymphovascular invasion).
What about Biomarkers in Identifying High-Risk Patients?
Williams et al reports on the role of CD3 and CD8 in tumor tissue of 868 patients receiving adjuvant CT in the QUASAR trial.[27] Using artificial intelligence (AI), they showed that risk of recurrence was double in tumors having all parameters of high risk (CD3-CT: rate ratio, 2.00, p = 0.0008; CD3-IM: 2.38, p < 0.00001; CD8-CT: 2.17, p = 0.0001; CD8-IM: 2.13, p = 0.0001). This was seen in both the training set as well as the validation group.
Similarly, Jiang et al published their data on pathomic signatures in a retrospective study that indicated benefit of adjuvant CT.[28] A total of 114 pathomic features were identified in hematoxylin and eosin-stained slides of 785 patients using Cox regression model. Patients with a low pathomic signature benefited more from CT (DFS p = 0.001; OS p < 0.001).
Other studies looking at analysis of resected tumor tissue focused mainly on mutational or transcriptomic signatures. No robust prognostic biomarkers could be identified as yet. However, the application of recent AI-driven tools and refined digital imaging is expected to make available new algorithms that could stratify patients into clinically meaningful distinct prognostic groups.[29]
What about Older Patients (Whose Life Expectancy is Limited)?
Okamoto et al published the results of adjuvant CT in 1,138 Japanese patients operated for high-risk stage II or stage III CRC.[30] Patients were analyzed according to age (cutoff 70 years) and adjuvant chemotherapeutic regimens. Older patients (N = 507; 45% ≥ 70 years old) were less likely to receive adjuvant CT (p < 0.001). Survival was longer in the younger age group patients (p = 0.006); especially if they received adjuvant CT (p = 0.005). The outcome was not different between the various adjuvant CT used.
Another Surveillance, Epidemiology, and End Results evaluation of 90,347 CRC geriatric patients (more than 70 years of age) showed that the use of appropriate surgery led to better OS (significant in the whole group as well as age-stratified analysis of four age groups [70–74, 75–79, 80–84, ≥ 85; all p < 0.001]).[31] Use of adjuvant therapy was also significant prognostic factor for OS in the elderly CRC patients (all p < 0.001). Survival advantage with adjuvant CT was maximum for stage III and IV colon cancer patients. Among rectal cancer patients, the OS benefit was seen with adjuvant chemoradiotherapy (stage II, III, and IV cases).
Impact of Drug Metabolism
Single-nucleotide polymorphisms (SNPs) in dihydropyrimidine dehydrogenase (DPYD) and other genes involved in the metabolism and bioavailability of 5FU influence blood levels, efficacy, and especially toxicity.[32]
[33]
[34]
[35]
[36] Systematic review of published literature (13,929 patients) showed that the incidence of DPYD variants (heterozygous or homozygous) was in 4.1% of patients.[32] It was associated with treatment-related deaths in 0.1% in patients without identified DPYD variants and 2.3% of those with known DPYD variants. Another cohort study included 161 patients. It showed that CES1 rs71647871-A phenotype led to severe hand–foot syndrome (p = 0.030; GG vs. A) and CDA rs1048977-CC (p = 0.030; T vs. CC) was associated with toxicity-related interruption of capecitabine monotherapy (p = 0.003).[33]
Pooled Indian data from 2,000 patients indicated that V732I, S534N, and rs3918290 variants were associated with 5FU/capecitabine toxicity, whereas C29R, I543V and M166V variants did not lead to extra toxicity.[34]
Pavitran reported on common variant in Asian population (c.496A > G). In his series, 47 /375 (12.5%) had DPYD mutation (32/47 mutation variant c.496A > G; 15/47 mutation IVS14 + 1 G > A). A total of 35 out 47 (74.5%) patients had grade II to III toxicity even after dose reduction during first CT cycle. This included acute myocardial infarct,[1] febrile neutropenia,[4] neutropenia,[25] hand–foot syndrome,[18] diarrhea,[15] and mucositis.[7]
[35]
Minimal Residual Disease (MDR) and Adjuvant CT
A systematic review of adjuvant CT randomized trails found 1,469 publications of which 18 were eligible for analysis.[37] This included 16,682 patients. Among those with stage II to IV CRC, better RFS was attributed to early divergences during active cytotoxic CT. Adjuvant CT did not seem to reduce the risk of late recurrences. Hence, there is a need to find better biomarkers to identify patients likely to benefit most from adjuvant CT.[29]
[38]
The CIRCULATE-Japan GALAXY observational trial involved medial follow-up of 23 months in 2,240 patients.[38] It showed that circulating tumor deoxyribonucleic acid (ctDNA) positivity correlated with inferior DFS (p < 0.0001) and OS (p < 0.0001). This was also true for patients who converted to ctDNA positivity while on follow-up (shorter OS; p < 0.0001). On the other hand, if ctDNA clearance persisted, the 24-month DFS was better (89.0% vs. 3.3%) and the 24-month OS was also better (100.0% vs. 82.3%).
Chidharla et al reported on a meta-analysis of postsurgical ctDNA among patients with stage I to IV (oligometastatic) CRC after they had undergone curative resection.[39] This included 23 studies and 3,568 patients. Pooled analysis for RFS in postop ctDNA positive cases was inferior (p < 0.00001).
Adjuvant Immunotherapy
It is well established that immunotherapy has an important role in metastatic CRC having mismatch repair deficient or microsatellite instability-high (dMMR/MSI-H) tumors. What about nonmetastatic CRC? Burley et al reported on dMMR/MSI-H stage III CRC treated with definitive surgery and adjuvant oxaliplatin-based CT that had early recurrence.[40] This was associated with persistently high plasma ctDNA levels. Change in therapy to pembrolizumab (immune checkpoint blockade) led to ctDNA clearance.
Conclusion
Adjuvant therapy in CRC is constantly evolving. Accumulation of data over the years has enabled us to better identify patients that would benefit most from such an approach. Whether to give treatment for short or long duration is also being refined. Individualizing patient management allows us to meet their preference with minimal toxicity and better QoL. Molecular monitoring allows a more precise decision making. Use of AI tools and algorithms may further improve outcomes in the future.
