Colorectal cancer (CRC) is one of the most commonly diagnosed tumors worldwide with about 0.78 million cancer-related deaths every year (1, 2). Approximately 60% of CRC patients were diagnosed with stage II and III disease at diagnosis (3–5). Despite remarkable improvement in new therapeutic techniques and strategies in the past two decades, postoperative recurrence occurs in approximately 30% of these patients (3–5). Furthermore, the benefit of adjutant chemotherapy is only noted in 5% of stage II and 15-20% of stage III patients (6–10). A significant proportion of CRC patients will experience adverse effects of chemotherapeutic drugs (3–10). These results highlight the unmet need to develop novel powerful prognostic biomarkers that allow for better recurrence stratification and optimal treatment strategies.

Accumulating evidence demonstrate that the immune microenvironment significantly affects tumor progression and metastasis (11).The immune microenvironment may provide abundant resources for identifying novel recurrence biomarkers (11). Macrophage is a main cellular subtype in the immune microenvironment and is commonly known as tumor-associated macrophages (TAMs) (12). Different macrophage subtypes have distinct functions in tumor progression. Generally, TAMs are separated into two distinct polarized states, including classically (M1), or alternatively (M2) activated macrophage (13–16). Classically activated (M1) TAMs, which express increased levels of iNOS, CD86, and CD169, play resistant roles in tumor progression and metastasis. In contrast, alternatively activated (M2) TAMs express increased levels of CD163, CD206, and CD204 (17–19). M2 TAMs in tumor tissues play a vital role in the suppression of tumor-associated immune responses and the enhancement of tumor invasion and metastasis (15–19). It should be noted that TAMs may be highly plastic cells and consist of a spectrum of activated states. And with M1-type and M2-type macrophages may only represent the extremes on each opposing end (18–20). We hypothesized that low CD86+ and high CD163+ TAM levels were clearly correlated with aggressive tumor phenotypes. Although the biology mechanism behind CD86/CD163 ratio is well known, the use of this ratio hasn’t been assessed for predicting tumor recurrence in CRC.

In this study, we evaluated the infiltration of macrophages marked with CD86 and CD163 in tumor tissues by the immunohistochemistry technique and focused on CD86/CD163 ratio, according to the STROBE guidelines (21). We aimed to provide enough evidence for the recurrence-risk stratification of stage II-III CRC.

By the immunohistochemistry staining, we constructed and validated a novel prognostic biomarker based on the stromal CD86/CD163 ratio to improve the prognostic stratification for stage II-III CRC. Our study demonstrated that CD86/CD163 ratio could effectively classify stage II-III patients with CRC into subgroups with high and low risks of postoperative recurrence. According to CD86/CD163 ratio and tumor stage, stage II-III patients were stratified into four recurrence-risk subgroups (high-ratio stage II, high-ratio stage III, low-ratio stage II, and low-ratio stage III). Of particular importance, this is the first study using standard immunohistochemical procedures that demonstrates the clinical utility of CD86/CD163 ratio as a postoperative prognostic tool in stage II-III CRC.

Adjutant chemotherapy after radical resection is recommended as the standard treatment strategy for stage III or high-risk stage II CRC (6–10). Previous evidence demonstrate that adjutant chemotherapy generally has limited benefits to stage II patients with CRC with an improved survival of 2-5% at 5 years after radical resection (10). In contrast, adjutant chemotherapy has shown robust efficacy in stage III CRC with an improved survival of 15-20% at 5 years (3–9). Six months of adjutant chemotherapy with oxaliplatin-based regimens has become the standard treatment for these patients. A significant proportion of CRC patients will experience cumulative neurotoxicity associated with oxaliplatin exposure. Grothey et al. (28) conducted a prospective pooled analysis and demonstrated that 3 months of adjutant treatment with either the CAPOX (oxaliplatin and capecitabine) or FOLFOX (oxaliplatin, fluorouracil, and leucovorin) regimen appeared to be sufficient in a lower-risk group (T1-3N1M0), especially when the CAPOX regimen was chosen. Six months of adjutant treatment with the CAPOX or FOLFOX regimen may be needed in a higher-risk group (T4N1M0/T1-4N2M0). Notably, our study demonstrated that tumor recurrence was heterogeneous even within the high-ratio and low-ratio subgroups of stage II-III CRC. Therefore, stage II-III CRC should be further stratified by CD86/CD163 ratio and stage into four recurrence-risk subgroups. This stratification may contribute to tailoring chemotherapy regimens and avoiding undertreatment or overtreatment in specific patients. We developed a prognostic nomogram based on CD86/CD163 ratio and tumor stage that allows for individualized estimation of the 48-month RFS probabilities among stage II-III CRC. Taken together with tumor stage, the stromal CD86/CD163 ratio may serve as a clinically useful tool to improve the prediction of tumor recurrence and guide more appropriate therapies for different risk subgroups.

The cross-network effect between the immune microenvironment and tumor cells may play critical roles in tumor occurrence and metastasis (11). The immune microenvironment may provide abundant resources for identifying novel recurrence biomarkers that allow for better stratification (11). Higher infiltration of CD3+ T cells, CD45RO+ memory T cells, and CD8+ cytotoxic T cells may be associated with decreased recurrence and mortality of CRC patients (29). Moreover, the quantitative analysis of two different lymphocyte subtypes (CD3/CD8, CD8/CD45RO, and CD3/CD45RO) revealed increased robustness and prognostic values for stages I-III CRC (30–32). These immune-based classifications are currently being introduced into the clinical settings for allowing the personal treatment.

Specifically, TAMs are one of the most represented lymphocyte subtypes in the immune microenvironment of solid tumors. TAMs are often separated into pro-inflammatory macrophages (M1-type) and anti-inflammatory macrophages (M2-type). Activated macrophages may be highly plastic immune cells and consist of a spectrum of activated states. And with M1-type and M2-type macrophages may only represent the extremes on two opposing ends of pro-inflammatory and anti-inflammatory states (18–20). M2 macrophages in tumor tissues was proved to promote angiogenesis and metastasis by secreting vascular endothelial growth factor (18, 33). Two previous studies involved only M2 TAMs and the total number of TAMs (M1+M2) instead of evaluating M1 and M2 TAMs separately (23, 34). In CRC patients, a high M2/(M1+M2) ratio may be significantly associated with increased recurrence and mortality of CRC patients (23, 34). This previous study implied that the infiltration of CD68+ TAMs wasn’t associated with recurrence and mortality of CRC, which may be attributed to the functional counterbalance modulated by the M1 and M2 TAMs (35). Although many studies have considered CD86 and CD163 as the cell surface markers of M1 and M2 TAMs, respectively, only a few studies have emphasized the clinical significance of CD86+ TAMs and CD163+ TAMs in solid tumors. High levels of CD86+ TAMs and low levels of CD163+ TAMs were closely related with a favorable prognosis in CRC patients (36, 37). And the polarization of M1 and M2 TAMs may be two ends of the macrophage spectrum. The combined analysis of CD86+ TAMs/CD163+ TAMs seems to be more appropriate for determining recurrence and mortality. In our study, high CD86/CD163 ratio subgroup hinted the polarization of M1 TAMs and served as a favorable prognostic factor for stage II-III CRC. On the other hand, low CD86/CD163 ratio subgroup exhibited the polarization of M2 TAMs and markedly correlated with decreased recurrence and mortality of CRC patients. These study results further emphasized the opposite functions of M1 and M2 TAMs. Until now, it was the first study about the ratio of M1/M2 (CD86+TAM/CD163+TAM) in stage II-III CRC.

But we should acknowledge some potential limitations for our study. First, our study design is retrospective in nature. So, this prognostic biomarker requires further validation in the prospective and multicenter cohorts before its clinical application. Secondly, our study is a real-world study. In such studies, some patients received adjuvant chemotherapy with the variant regimens, which may interfere with the study conclusions. In fact, the prognostic biomarkers must face such situations in the clinical practice. Thirdly, the detection method of TAMs needed the participation of experienced pathologists. The selection bias in the manual method couldn’t be completely avoided. However, the best way to solve this problem is to develop the automated method based on artificial intelligence.

In summary, our study demonstrated that CD86/CD163 ratio could effectively stratify stage II-III CRC into two different subgroups with a low and high risk of tumor recurrence and mortality. In addition, detection of CD86 and CD163 expression using the immunohistochemistry method is inexpensive and rapid and can be easily worked out in the pathology department of almost every hospital. Moreover, most patients could afford the cost associated with CD86 and CD163 detection for individual treatment. With further prospective studies of multiple-center cohorts, the CD86/CD163 ratio will aid in the personal treatment for stage II-III CRC.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by Yantai Yuhuangding Hospital and Guangzhou Red Cross Hospital. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

GX, YM, and CJ conceived of and designed the study. GX, LJ, GQ, and ZL performed the experiments. GX, LJ, and ZL performed the data analysis. GX and JC reviewed this manuscript. All authors contributed to the article and approved the submitted version.

This study protocol was supported by the High-level Talents Project of Liuzhou People Hospital (No. lrygcc202106), the Basic Research Ability Improvement Project of Young and Middle-aged Teachers from Guangxi colleges (No. 2021KY0111), and Self-funded project of Health Committee of Guangxi Zhuang Autonomous Region (Z20190901).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.