Real World Data on Adjuvant Chemotherapy in Stage II Colorectal Cancer: Risk Factors and Benefits

Adjuvant chemotherapy (ACh) benefits in stage II colorectal cancer (CRC) is still unclear. Risk factors (RF) as perineural (PNI) or lymphovascular (LVI) infiltrate in biopsy, T4 or less than 12 lymph nodes (LN) resected during surgery has associated with recurrence. We retrospectively evaluated Ach benefits in stage II CRC patients. This retrospective analysis collected information of CRC patients who were treated at University of Chile Clinical Hospital between 2011-2018 and had histological diagnosis of T3-4N0 colon adenocarcinoma. Clinical, surgical, histological variables, and RF were recorded. Disease‐free survival (DFS) was compared in patients who underwent ACh with patients who were observed. Of 91 patients, 51% received Ach and 49% observation. Median age was 60± 13 years. ACh were FOLFOX, CAPOX, capecitabine or FU/LV in 70%, 17%, 11%, and 2% of patients, respectively. DFS was 45 and 36 months in ACh and observation groups (p=0,277). RF were present in 62%, and of these 73% received ACh and 27% observation. There were no association of T4 stage, PNI and/or LVI with improve DFS and response to ACh. Recurrence was presented in 12,6% of patients and 75% of this presented one or more RF. In this retrospective cohort, ACh did not improve DFS compared to observation only. In our study, 25% of recurrences are in patients without RF, therefore it is necessary to improve strategies to predict recurrences and response to ACh. Further studies should evaluate other elements to optimize the decision to administer ACh in these patients, such as microsatellite instability and circulating DNA. Adjuvant RF: Risk PNI: Perineural Infiltrate, LVI: Lymphovascular Infiltrate, LN: Lymph Nodes, MSI: Microsatellite Instability, FOLFOX: Leucovorin, Fluorouracil and Oxaliplatin, CAPOX: Capecitabine and Oxaliplatin, FU/LV: Fluorouracil and Leucovorin, SD: Standard Deviation


Introduction
Colorectal cancer (CRC) have a growing incidence in developed countries [1,2], and is one of the main causes of mortality due to cancer worldwide [3]. Major improvements in disease-free survival (DFS) and overall survival (OS) in patients with non-metastatic colon cancer has been as a result of improvement of surgery quality [4][5][6] and continuous clinical trials investigation on adjuvant chemotherapy (ACh) [7][8][9].
Despite well-stablished role of ACh as standard treatment in patients with stage III, the current evidence does not support routine ACh in stage II patients [10]. It's have been shown a 70-80% of curative rates with only surgery in patients with stage II [11], in consequence, selection of patients who are candidate to ACh need to be rigorous. Due to benefit in stage II CRC is still unclear, it must be considered potential toxicity of chemotherapy agents, for instance, the potentially debilitating neuropathy related to oxaliplatin as well as the arguably benefit of oxaliplatin in patients >70 years of age [12,13].
It's have been identified risk factors (RF) based on pathologic and clinical characteristics, according to National Comprehensive Cancer Network (NCCN) guidelines, present of perineural (PNI) or lymphovascular (LVI) infiltrate in biopsy, T4 or less than 12 lymph nodes (LN) resected during surgery are associated with recurrence of disease [10]. Emerging advances for personalized therapy have found different molecular profiles in stage II, molecular markers as microsatellite stability (MSI) could help to decide which patients will benefit of ACh [14,15].
In our study, we retrospectively evaluated patients with stage II colon cancer, in real world data analysis with a specific emphasis on the practice of ACh selected by clinical and pathologic features.

Clinical and Pathological Data
In all patients age, sex, clinical comorbidities, surgical procedure, biopsies results, and blood samples were obtained by clinical records. Pathological examination and stage classification were according to TNM Classification of Malignant Tumours, eighth edition [16]. Also, RF were recorded, and were defined in concordance to

European Society for Medical Oncology (ESMO) Clinical Practice
Guidelines [17]. High risk patients were considered if they have at least one of these characteristics: lymph nodes sampling less than 12, LVI and/or PNI and pT4 stage.

Adjuvant Chemotherapy
The ACh protocols were determined by our institution Oncology Committee directed by local and international guidelines. High risk patients, according to previously mentioned criteria, were candidate to receive ACh. Chemotherapy agents include FOLFOX (leucovorin, fluorouracil and oxaliplatin), CAPOX (capecitabine and oxaliplatin), capecitabine or FU/LV (fluorouracil and leucovorin).
Duration and dose were following standard doses.

Statistical Analysis
Disease-free survival (DFS) was calculated starting the first day after surgery to recurrence of tumour determined by clinical and imaging records. Due to the features of observational and real-world data study, minimum sample size was not calculated. DFS were presented in Kaplan-Meier curves and statistical differences was assessed using the log-rank test. Univariate Cox's proportional hazards model (HR, 95% CI) was used in RF analysis. Statistical analysis and graphics were performed by GraphPad Prism 5.0 software. A significative result was considerate if p-value is less than 0.05.\   ACh group a total of 93% participants concluded a 6 months treatment and only 7% terminated a 3 months scheme.     [21], that reported a 13.3% recurrence rate in a retrospective follow-up.

Copy@ Hyungeun Song
The principal guidelines have suggested classifying stage II according to histopathological clinical characteristics [22]. High risk profile is considered if it has any of the following characteristics: present of perineural (PNI) or lymphovascular (LVI) infiltrate in biopsy, T4 or less than 12 lymph nodes (LN) resected during surgery [10]. In fact, a worse tumour-specific survival and DFS has been reported with the presence of more than one risk factor in a patient who did not receive chemotherapy [23]. In addition, a retrospective study of stage II patients shown decrease in general mortality in patients with high risk profile who underwent to ACh compared with only surgery group. However, this same effect was not seen in DFS in same groups [24]. In our study, did not find a benefit in DFS when comparing observation and ACh groups, in high risk profile patients. Recently, it has been proposed that the addition of oxaliplatin may not be superior to fluoropyrimidine-based regimens for high-risk stage II colon cancer [25]. In our cohort, 70% of patients received FOLFOX, this amount of patients treated with oxaliplatin should be considered in interpretation of non-differences outcomes between observation and ACh in high-risk groups.
Several study limitations should be considered, due to their characterises of retrospective and unicenter study could not be obtained a larger sample and evaluated other variables that might be a source of bias as socioeconomical status, performance status and comorbidities. The presence of microsatellite instability was not considered as a prognostic factor, since it is a measurement of recent practice in our center. Despite limitations a strength of this study is the use of real-world data methodology, that include a flexible inclusion criterion, that do not exclude older age or by comorbidities. These groups of patients have been underrepresented in randomised controlled trials [26].

Conclusion
In this retrospective and real-world data study is evidenced that usual clinical practice to select patient candidates to ACh in stage II colon cancer through high risk factor profile could be insufficient. It did not demonstrate a benefit in patients with a positive risk factor as PNI, LVI or T4. Even, a 25% of recurrences were in patients without RF. In this perspective, emerging strategies to predict a recurrences and response to ACh in stage II colon cancer are needed beyond the classical approaches of clinical and pathological risk factors.