Clinical Performance of the Consensus Immunoscore in Colon Cancer in the Asian Population from the Multicenter International SITC Study

Simple Summary

Research interest in Immuno-oncology and the role of the adaptative immune system in the progression and prognosis of colon cancer (CC) is growing. In this study, we evaluated the prognostic value of the consensus Immunoscore in 423 patients with AJCC/UICC-TNM stages I–III CC from Asian care centers. Immunoscore (IS) is a bench-to-digital pathology assay that quantifies CD3+ and cytotoxic CD8+ T-lymphocyte densities within the tumor and its invasive margin, stratifying patients into three categories: Low IS, Intermediate IS, and High IS. Multivariable Cox models stratified by center were used to assess the associations between Immunoscore and outcomes, adjusting for potential confounders, including gender, T-stage, N-stage, sidedness, and MSI. A comparison of the performance of risk prediction models was performed using the likelihood ratio test p-value. In uni/multivariable analyses, a High Immunoscore was significantly associated with prolonged survival of CC patients within the Asian population.

Abstract

BACKGROUND: In this study, we evaluated the prognostic value of Immunoscore in patients with stage I–III colon cancer (CC) in the Asian population. These patients were originally included in an international study led by the Society for Immunotherapy of Cancer (SITC) on 2681 patients with AJCC/UICC-TNM stages I–III CC. METHODS: CD3+ and cytotoxic CD8+ T-lymphocyte densities were quantified in the tumor and invasive margin by digital pathology. The association of Immunoscore with prognosis was evaluated for time to recurrence (TTR), disease-free survival (DFS), and overall survival (OS). RESULTS: Immunoscore stratified Asian patients (n = 423) into different risk categories and was not impacted by age. Recurrence-free rates at 3 years were 78.5%, 85.2%, and 98.3% for a Low, Intermediate, and High Immunoscore, respectively (HR[Low-vs-High] = 7.26 (95% CI 1.75−30.19); p = 0.0064). A High Immunoscore showed a significant association with prolonged TTR, OS, and DFS (p < 0.05). In Cox multivariable analysis stratified by center, Immunoscore association with TTR was independent (HR[Low-vs-Int+High] = 2.22 (95% CI 1.10–4.55) p = 0.0269) of the patient’s gender, T-stage, N-stage, sidedness, and MSI status. A significant association of a High Immunoscore with prolonged TTR was also found among MSS (HR[Low-vs-Int+High] = 4.58 (95% CI 2.27−9.23); p ≤ 0.0001), stage II (HR[Low-vs-Int+High] = 2.72 (95% CI 1.35−5.51); p = 0.0052), low-risk stage-II (HR[Low-vs-Int+High] = 2.62 (95% CI 1.21−5.68); p = 0.0146), and high-risk stage II patients (HR[Low-vs-Int+High] = 3.11 (95% CI 1.39−6.91); p = 0.0055). CONCLUSION: A High Immunoscore is significantly associated with the prolonged survival of CC patients within the Asian population.

1. Introduction

The AJCC/UICC-TNM classification system based on the anatomopathological evaluation of tumors provides useful yet limited prognostic data [1]. Recent methods established to classify cancer that focus on tumor cells have demonstrated limitations in their clinical efficiency to reliably estimate outcomes [1,2]. Nevertheless, extensive studies have shed light on the adequate prognostic accuracy of the in situ immune cell infiltrate in tumors [1,3,4,5,6,7,8,9,10,11,12]. Our previous works on colorectal cancer (CRC) have shown important correlations between tumor recurrence, overall survival, and the strength of the in situ adaptive immune response [3,8,12,13,14] at the center of the tumor (CT) and its invasive margin (IM). A systematic review of 200 relevant publications depicting the role of immune cell subpopulations in the prognosis of cancer patients in 20 different cancer types showed that, in 97% of the studies, cytotoxic CD8+ T cells were associated with a good prognosis [15]. We have also reported that within specific regions of primary tumors, tumor recurrence and overall survival rates of patients with CC were mostly dependent on the presence of cytotoxic and memory T cells. In our earlier clinical study on human CRC, we showed that cytotoxic and memory T cells could predict the clinical outcome in early-stage (I/II) CRC patients. Furthermore, we revealed that the state of the local immune reaction was correlated with the histopathology-based prognostic factors of CRC. In combined tumor regions, the analysis of CD8+ cytotoxic T-lymphocyte density proved to be a better indicator of tumor recurrence than the TNM staging score [16,17,18]. This indicates that the patient’s intratumoral native adaptive immune reaction is of utmost importance for survival, strongly hinting that the immune parameters are more relevant than tumor progression and invasion classifications. This immune response was defined as the “Immunoscore” [15,19,20,21].

An international consortium of 14 care centers enrolled patients with TNM stage I–III CC and showed that Immunoscore was the first worldwide standardized consensus assay to quantify pre-existing immunity. According to these results, the consensus Immunoscore is recognized as a pertinent and powerful tool to predict the prognosis of patients [22]. The consensus Immunoscore provides a reliable assessment method for predicting the recurrence risk in CC, as confirmed by a meta-analysis of the prognostic value of Immunoscore on more than 10,000 patients [23].

Recent publications have demonstrated the prognostic value of Immunoscore in stage III CC patients and its predictive value for response to chemotherapy, thus reinforcing Immunoscore’s clinical relevance [24,25]. In the latest (5th) edition of the WHO Digestive System Tumours classification, the immune response evaluated with the consensus Immunoscore was defined as an “essential and desirable diagnostic criteria for colorectal cancer”. Immunoscore was also introduced into the 2020 ESMO Clinical Practice Guidelines for CC to improve the prognosis and thus adjust the chemotherapy decision-making process in stage II and even in low-risk stage III patients. However, the clinical performance of the consensus Immunoscore in the Asian population remained to be established.

2. Materials and Methods

2.1. Patients

An international consortium composed of 14 pathology expert centers from 13 countries was initiated to evaluate the standardized Immunoscore assay in primary tumors from 2681 patients with stage I/II/III CC. The selected patients are a subset of the SITC study cohort based on an Asian population of 423 patients (Centers from Japan, China, and India). The results of this particular cohort (Asia) have not been reported before and were not shown in Pages et al. [22]. Clinical data from Asia and the complete international consortium datasets are presented in Table S1. The outcomes of interest were time to recurrence (TTR), defined as time from surgery to disease recurrence; overall survival (OS), defined as time from surgery to death due to any cause; and disease-free survival (DFS), defined as time from surgery to disease recurrence or death from any cause. Ethical, legal, and social implications were approved by the ethical review board of each center.

2.2. Immunohistochemistry

At every care center, a tumor block containing the CT and IM was selected for each patient by the center’s pathologist. Two FFPE slides of 4 microns were generated per block and processed for immunohistochemistry according to a protocol recommended by the reference center and as previously described [22]. An example image of CD3 and CD8 staining is provided in Supplementary Figure S2. Digital slides were obtained with a 20× magnification and a resolution of 0.45 µm/pixel.

2.3. Image Analysis

The stained CD3 and CD8 cell densities were determined in CT and IM regions using in-house Immunoscore software (INSERM, Paris, France). The means and distributions of staining intensities and cell densities were monitored, with an internal quality control for each slide.

2.4. Immunoscore Determination

For each slide, the Immunoscore was assessed: CD3 and CD8 densities in CT and IM regions were converted into percentiles, as previously described [22]. The mean of the four percentiles obtained (two markers, two regions) was calculated and translated into the Immunoscore scoring system. The Immunoscore categories were previously defined independently of clinical data [22]. These pre-defined categories were used herein, with three Immunoscore categories being defined as follows: mean percentiles of 0–25%, >25–70%, and >70–100% were Low (Lo), Intermediate (Int), and High (Hi) Immunoscore, respectively. Additionally, analyses were performed with two Immunoscore categories (Low (0–25%) and Int+Hi (25–100%) groups) and five Immunoscore categories (I0 (0–10%), I1 (>10–25%), I2 (>25–70%), I3 (>70–95%), and I4 (>95–100%) groups).

2.5. Monitoring of the Study

The biomarker reference center (Immunomonitoring platform, Hôpital Européen Georges Pompidou AP-HP, INSERM, Paris, France) optimized immunostaining protocols, provided the Immunoscore software user’s manual, and validated data from each cohort analyzed within each of the 14 participating centers [22]. Exclusion criteria include: missing counts at either tumor region, poor/low staining intensity (≤152 AU), damaged FFPE slides during staining, and several (>3) failed attempts at antigen retrieval. After quality control exclusion, analyses were performed on 423 Asian patients and compared to the 2681 patients included in the international consortium.

2.6. Statistics

Statistical analyses of demographics and disease characteristics were descriptively compared across Asia and the rest of the world and compared by t-test, Fisher’s exact test, and Chi-square test when applicable. The bivariable association between Immunoscore and time-to-event outcomes was evaluated by the log-rank test and by a participating-center-stratified Cox proportional hazards model. Multivariable Cox models stratified by center were used to assess the associations between Immunoscore and outcomes, adjusting for potential confounders (survival, R package). Model performance was assessed by Harrell’s C-statistics. The centers were used as the stratification factors, and the variables adjusted in the multivariable models were Immunoscore, gender, T-stage, N-stage, sidedness, and MSI. A comparison of the performance of risk prediction models was performed using the likelihood ratio test p-value. The relative importance of each parameter to survival risk was assessed using the chi-squared proportion (χ²) (rms, R package). An alternative measure of the survival time distribution was used, the restricted mean survival time (RMST), for two-sample comparisons (survRM2, R package) [26].

3. Results

3.1. Immune Densities and Immunoscore in Relation to the Age of the Patients

Biomarker data from 423 colon cancer patients from the Asian population (Japan, China, and India) from the AJCC/UICC-TNM stage I–III part of the consensus Immunoscore international validation study [22] were investigated. Clinical characteristics of patients from the Asian population (n = 423) were compared to the 2681 patients from the SITC international study (Table S1). Balanced clinical characteristics were observed, with no statistical differences between cohorts in gender, T-stage, N-stage, or UICC/AJCC-TNM stages (Table S1). However, the Asian population had slightly fewer dMMR (MSI-H) patients (6.1% vs. 11.3%), and Asian patients were more frequently younger (47.3% vs. 38.2% below 65 years old) and more frequently received chemotherapy (62.7% vs. 28%) (Table S1). Overall, Asian patients were 54.6% male, with a mean age of 64.7 ± 12.1 years. The mean number of lymph nodes (LN) examined was 16.3 ± 9.9. Across all patients analyzed, 65 relapses (15.4% of patients) and 62 deaths (14.7% of patients) were observed. The median follow-up times (95% CI) were 73.6 (69.8–76.9), 75.0 (71.1–78.4), and 78.3 (75.0–81.6) months for TTR, DFS, and OS, respectively. The 5-year relapse or survival rates were 83.0% (79.3–86.9), 81.0% (77.2–85.0), and 87.9% (84.7–91.1) for TTR, DFS and OS, respectively (Table S2).

Pre-defined consensus Immunoscore cut-points [22] were applied to the Asian cohort to convert CD3 and CD8 immune densities into percentiles and Immunoscore categories. The intra-tumoral densities quantified in the core of the tumor and in the invasive margin were not influenced by the age of the patients (Figure 1A). Similarly, the proportions of High-, Intermediate- and Low-Immunoscore patients were independent of the age interval (Figure 1B). Thus, Immunoscore did not significantly differ between young and elderly Asian patients.

3.2. Immunoscore and the Outcome of Asian Colon Cancer Patients

The prognostic value of two, three, and five categories of Immunoscore for TTR, DFS, and OS of 423 stage I–III CC patients was further evaluated in the Asian population using pre-defined cut-points (Figure 2 and

Table 1. Stage I–III bivariable analysis for clinical parameters for TTR.

		Time to Recurrence (TTR)
	Number of	Rate at		Unadjusted			RMST
	Patients (%)	3 yr % (95% CI)	5 yr % (95% CI)	HR (95% CI)	p-Value *	C-Index (95% CI)	Rel. Months (95% CI)	p-Value **
Age at surgery (5 groups)						0.55 (0.49−0.62)
<60	134 (31.7)	82.4 (75.7–89.8)	81.3 (74.3–88.9)	1.0 (reference)			0.0 (reference)
≥60–70	134 (31.7)	85.3 (79.2–91.8)	83.5 (77.2–90.4)	0.87 (0.47−1.59)	0.6500		2.2 (−5.1–9.6)	0.5517
≥70–85	149 (35.2)	86.4 (80.8–92.2)	85.6 (80–91.7)	0.72 (0.39−1.32)	0.2854		3.9 (−3.2–10.9)	0.2832
>85	6 (1.4)	66.7 (37.9–100)	50 (22.5–100)	2.85 (0.85−9.56)	0.0897		−15.8 (−43.3–11.7)	0.2609
Gender						0.5 (0.44−0.56)
Male	231 (54.6)	84.2 (79.3–89.4)	83.6 (78.6–88.9)	1.0 (reference)			0.0 (reference)
Female	192 (45.4)	84.9 (79.7–90.3)	82.4 (77–88.3)	1.03 (0.63−1.67)	0.9133		−0.4 (−13.1–12.3)	0.9520
T stage						0.62 (0.56−0.67)
T1	20 (4.7)	100 (100–100)	100 (100–100)	1.0 (reference)			0.0 (reference)
T2	53 (12.5)	93.7 (87–100)	93.7 (87–100)	Inf (0-Inf)	NA		−8.2 (−17.1–0.8)	0.0737
T3	289 (68.3)	84.9 (80.7–89.4)	82.9 (78.4–87.6)	Inf (0-Inf)	NA		−20.5 (−26–15)	<0.0001
T4	61 (14.4)	68.9 (57.6–82.4)	68.9 (57.6–82.4)	Inf (0-Inf)	NA		−38.7 (−54.1–23.4)	<0.0001
N stage						0.64 (0.58−0.71)
N0	318 (75.2)	89.8 (86.4–93.3)	88.3 (84.7–92.1)	1.0 (reference)			0.0 (reference)
N1	66 (15.6)	79.2 (69.3–90.4)	79.2 (69.3–90.4)	1.94 (1.01−3.74)	0.0477		−11.3 (−24.6–1.9)	0.0944
N2	39 (9.2)	38.5 (23.5–62.8)	34.2 (19.9–58.8)	6.91 (3.89−12.27)	<.0001		−59.5 (−80.5–38.6)	<0.0001
AJCC/UICC-TNM stage						0.67 (0.62−0.72)
I	67 (15.8)	98.3 (95.1–100)	98.3 (95.1–100)	1.0 (reference)			0.0 (reference)
II	251 (59.3)	87.6 (83.4–91.9)	85.7 (81.3–90.4)	9.34 (1.28−68.23)	0.0277		−16.9 (−24.4–9.5)	<0.0001
III	105 (24.8)	66.2 (56.8–77.1)	64.9 (55.4–76)	25.62 (3.49−187.95)	0.0014		−44.3 (−58.5–30.1)	<0.0001
Differentiation grade						0.62 (0.56−0.67)
Well	118 (28.2)	91.3 (86.4–96.6)	91.3 (86.4–96.6)	1.0 (reference)			0.0 (reference)
Moderate	261 (62.4)	83.3 (78.6–88.3)	81.4 (76.5–86.7)	2.35 (1.18−4.67)	0.0152		−14.9 (−25.1–4.7)	0.0043
Poor–undiff.	39 (9.3)	68.4 (53.9–86.8)	64.1 (49–84)	5.15 (2.19−12.14)	0.0002		−39.3 (−64.9–13.6)	0.0027
Proximal vs. Distal Primary (Tumor)						0.51 (0.45−0.57)
Proximal	184 (44.2)	84.2 (78.7–90.1)	83.5 (78–89.5)	1.0 (reference)			0.0 (reference)
Distal	232 (55.8)	84.7 (80–89.6)	82.6 (77.7–87.9)	1.07 (0.65−1.77)	0.7848		−1.6 (−14.5–11.3)	0.8042
VELIPI						0.53 (0.48−0.58)
NO	122 (28.8)	88.3 (82.1–95.1)	88.3 (82.1–95.1)	1.0 (reference)			0.0 (reference)
YES	301 (71.2)	83.3 (79–87.7)	81.4 (77–86.1)	1.44 (0.77−2.7)	0.2514		−9.1 (−24.4–6.2)	0.2433
Mucinous colloid type						0.51 (0.48−0.54)
NO	367 (95.3)	84.9 (81.2–88.7)	83.3 (79.5–87.3)	1.0 (reference)			0.0 (reference)
YES	18 (4.7)	75.1 (56.6–99.7)	75.1 (56.6–99.7)	1.54 (0.56−4.23)	0.4061		−11.9 (−43.7–19.9)	0.4632
MSI Status (Derived)						0.52 (0.49−0.56)
MSS	246 (90.4)	86.2 (82–90.6)	84.5 (80.1–89.2)	1.0 (reference)			0.0 (reference)
MSI–H	26 (9.6)	92.3 (82.6–100)	92.3 (82.6–100)	0.48 (0.12−2.01)	0.3168		9.6 (−5.7–24.8)	0.2178
Adjuvant chemotherapy						0.57 (0.5-0.63)
NO	146 (35.5)	89.9 (85–95.1)	89.9 (85–95.1)	1.0 (reference)			0.0 (reference)
YES	265 (64.5)	81.8 (76.9–86.9)	79.4 (74.2–84.8)	2 (1.12−3.57)	0.0198		−15.4 (−27.8–3.1)	0.0145
Immunoscore Lo vs. Int+Hi						0.58 (0.52−0.64)
Lo (0–25%)	158 (37.4)	78.5 (72.1–85.4)	77 (70.5–84.1)	1.9 (1.17−3.1)	0.0097		−19.3 (−34.1–4.5)	0.0106
Int+Hi (25–100%)	265 (62.6)	88.4 (84.3–92.6)	86.9 (82.7–91.4)	1.0 (reference)			0.0 (reference)
Immunoscore Lo vs. Int vs. Hi						0.6 (0.55−0.66)
Lo (0–25%)	158 (37.4)	78.5 (72.1–85.4)	77 (70.5–84.1)	7.26 (1.75−30.19)	0.0064		−33.5 (−47.2–19.9)	<0.0001
Int (25–70%)	197 (46.6)	85.2 (80.1–90.6)	84 (78.7–89.6)	4.77 (1.14−20.04)	0.0327		−21.1 (−32.9–9.3)	0.0005
Hi (70–100%)	68 (16.1)	98.3 (95–100)	96.3 (91.3–100)	1.0 (reference)			0.0 (reference)
Immunoscore						0.61 (0.55−0.67)
I0 (0–10%)	71 (16.8)	75.1 (65.5–86.1)	73.5 (63.6–84.8)	7.75 (1.8−33.4)	0.0060		−16.1 (−22.7–9.6)	<0.0001
I1 (10–25%)	87 (20.6)	81.3 (73.2–90.3)	80 (71.7–89.3)	6.05 (1.4−26.18)	0.0161		−12.4 (−17.9–6.9)	<0.0001
I2 (25–70%)	197 (46.6)	85.2 (80.1–90.6)	84 (78.7–89.6)	4.48 (1.07−18.82)	0.0404		−10 (−13.4–6.5)	<0.0001
I3 (70–95%)	62 (14.7)	98.1 (94.6–100)	96 (90.7–100)	1.0 (reference)			−1.8 (−4.5–0.9)	0.1969
I4 (95–100%)	6 (1.4)	100 (100–100)	100 (100–100)	Inf (0-Inf)	NA		0.0 (reference)
Immunoscore Lo vs. Int+Hi and High risk (T4 and VELIPI+) vs. Low risk (all others)						0.58 (0.52−0.65)
0–25% High Risk	11 (2.6)	63.6 (40.7–99.5)	63.6 (40.7–99.5)	3.48 (1.22−9.91)	0.0198		−27.1 (−59.4–5.3)	0.1007
0–25% Low Risk	147 (34.8)	79.7 (73.2–86.7)	78.1 (71.5–85.4)	1.79 (1.08−2.99)	0.0250		−9.7 (−18.5–0.9)	0.0302
25–100% High Risk	16 (3.8)	87.5 (72.7–100)	87.5 (72.7–100)	0.97 (0.23−4.06)	0.9643		−0.4 (−19.4–18.6)	0.9678
25–100% Low Risk	249 (58.9)	88.4 (84.2–92.8)	86.9 (82.4–91.5)	1.0 (reference)			0.0 (reference)

* Wald p-value. ** Restricted mean survival time (RMST) p-value. MSS: proficient mismatch repair (pMMR).

).

The distribution of Immunoscore was 62.6% High+Int and 37.4% Low in two categories; 16.1% High, 46.6% Intermediate, and 37.4% Low in three categories; and 1.4%, 14.7%, 46.6%, 20.6%, and 16.8% in Immunoscore I4, I3, I2, I1, and I0, respectively. The two categories of Immunoscore enabled the identification of patients with distinct clinical outcomes for TTR (Figure 2A and Table 1). High-Immunoscore patients (63%) had a significantly longer survival for TTR (Hazard Ratio of HR_Lo/Int+Hi = 1.9 (1.17−3.1), p = 0.0097) and a higher 5-year recurrence-free rate (Hi: 86.9% (82.7–91.4%); Lo: 77% (70.5–84.1%)). The three categories of Immunoscore also enabled the identification of patients with distinct clinical outcomes for TTR (Figure 2B, Table 1). High-Immunoscore patients had a significantly longer survival for TTR (HR_Lo/Hi = 7.26 (1.75−30.19), p = 0.0064, and Trend p = 0.0025) and a higher 5-year recurrence-free rate (Hi: 96.3% (91.3–100%), Int: 84% (78.7–89.6%), and Lo: 77% (70.5–84.1%)). Even more striking differences were observed for the Immunoscore in five categories (Figure 2C and Table 1). High-Immunoscore patients had a significantly longer survival for TTR (HR_(I0/I3) = 7.75 (1.8−33.4), p = 0.006, and Trend p = 0.0019) and a higher 5-year recurrence-free rate (I4: 100% (100–100%), I3: 96% (90.7–100%), I2: 84% (78.7–89.6%), I1: 80% (71.7–89.3%), and I0: 73.5% (63.6–84.8%)) (Table 1). Forest plots are illustrated in Figure S1. Similar results were found for OS and DFS (Table S2).

3.3. Immunoscore in Microsatellite-Stable (MSS) Tumors

The impact of Immunoscore was further investigated in relation to the MMR status of the patients. When stratified into two Immunoscore categories, MSS tumors were associated with an Int+Hi Immunoscore in 61.8% (152/246) of cases, while a Low Immunoscore was observed in 38.2% (94/246) of MSS tumors. For the Immunoscore in two categories, MSS patients with an Int+Hi Immunoscore had significantly longer TTR (Figure 3A and Table S3) (HR_Lo/Int+Hi = 4.58 (2.27−9.23), p < 0.0001) and a higher 5-year recurrence-free rate (Int+Hi: 92.8% (88.7–97%) and Lo: 71.2% (62.7–81%)).

The 5-year OS survival rates were: 98.0% for Int+Hi and 80.9% for Lo; HR_Lo/Int+Hi = 5.3 (2.25–12.49), p = 0.0001. Patients with highly infiltrated MSS tumors had a survival advantage in both TTR (5-year recurrence rate, Hi: 96.9%, Int: 91.7%, and Lo: 71.2%; HR_Lo+Hi = 10.93 (1.49–80.49), p = 0.0188) and OS (5-year survival rate, Hi: 96.9%, Int: 98.3%, and Lo: 80.9%; HR_Lo+Hi = 7.67 (1.03–57.09), p = 0.0466) compared to patients with weakly infiltrated tumors. Similar results were found for DFS (Table S3).

A similar profile was observed when three (Figure 3B) and five (Figure 3C) Immunoscore categories were applied. This analysis identified low-risk MSS patients (I4) with a significantly longer TTR compared to Immunoscore I0 MSS patients (Figure 3C). High-Immunoscore patients had a significantly longer survival for TTR (HR _I0/I3 = 12.85 (1.68−98.28), p = 0.0139, and Trend p = 0.0005) and a higher 5-year recurrence-free rate (I4: 100% (100–100%), I3: 96.8% (90.8–100%), I2: 91.7% (86.9–96.7%), I1: 74.5% (63.9–87%), and I0: 66.6% (53.3–83.2%)). Similar results were found for OS and DFS (Table S3).

3.4. Immunoscore and Time-to-Event Analysis among Patients with Stage II Colon Cancer

Stage II patients from the Asian population (n = 251) were analyzed. Low-risk patients with an Int+Hi Immunoscore presented significantly better outcomes for TTR compared to Low-Immunoscore patients (Figure 4A). The 5-year recurrence rate for patients with a High Immunoscore was 91.1% (86.4–96.1%) and only 78.3% (70.4–86.9%) for those with a Low Immunoscore. High-Immunoscore patients had a significantly longer survival for TTR (HR_Lo/Int+Hi = 2.72 (1.35−5.51), p = 0.0052). The three categories of Immunoscore also enabled the identification of patients with distinct clinical outcomes for TTR (Figure 4B). High-Immunoscore patients had a significantly longer survival for TTR (HR_Lo/Hi = 3.82 (0.9−16.24), p = 0.0697, and Trend p = 0.0089) and a higher 5-year recurrence-free rate (Hi: 93.5% (85.2–100%), Int: 90.5% (85–96.3%), and Lo: 78.3% (70.4–86.9%)) (Table S3). For the Immunoscore in five categories, High-Immunoscore patients had a significantly longer survival for TTR (HR_I0/I3 = 3.89 (0.86−17.57), Trend p = 0.0771) and a higher 5-year recurrence-free rate (I4: 100% (100–100%), I3: 93.2% (84.6–100%), I2: 90.5% (85–96.3%), I1: 80.4% (70.3–92.2%), and I0: 75.8% (64.3–89.4%)) (Figure 4C, Table S3).

Among all stage II patients (n = 251), patient risk groups were defined using histopathological parameters: low risk, high risk (the extent of the primary tumor T4 or VELIPI+), and very high risk (T4 primary tumors and VELIPI+). In all risk groups (low risk (n = 224), high risk (n = 185), and very high risk (n = 27)), a High Immunoscore was associated with prolonged survival (Figure 5A–C, Table S3).

In low-risk stage II patients, Immunoscore was also significantly associated with TTR (unadjusted HR_Lo/Int+Hi = 2.62 (1.21−5.68), p = 0.0146), and the 5-year recurrence rate was 91.6% (86.7–96.7%) for patients with a High Immunoscore and only 80.2% (72.2–89.1%) for those with a Low Immunoscore (Figure 5A). Thus, patients with a Low Immunoscore at low pathological risk were in fact at high risk of recurrence. In high-risk stage II patients (T4 or VELIPI+), Immunoscore was significantly associated with TTR (unadjusted HR_Lo/Int+Hi = 3.11 (1.39−6.91), p = 0.0055), and the 5-year recurrence rate was 91.5% (86.4–97%) for patients with a High Immunoscore and only 75.8% (66.6–86.2%) for those with a Low Immunoscore (Figure 5B). In very high-risk stage II patients, Immunoscore was also associated with TTR, with a 5-year recurrence rate of 87.5% (72.7–100%) for patients with a High Immunoscore and only 63.6% (40.7–99.5%) for those with a Low Immunoscore (Figure 5C). Strikingly, patients with high-risk or very high-risk stage II and a High Immunoscore had a good outcome, similar to the rest of the Stage II cohort with lower risk (Figure 5D,E; Table S3).

3.5. Performance of Immunoscore in Multivariable Analyses

Cox multivariable analyses adjusted for Immunoscore, age, gender, T-stage, N-stage, sidedness, and MSI and stratified by city center revealed the significant prognostic value of Immunoscore (Figure 6). In a multivariable model, age, gender, T-stage (T3 vs. T1–2), N-stage (N1 vs. N0), sidedness, and MSI were not significant. Among tumor-related parameters, only T-stage ((T4 vs. T1–2), HR = 14.35 (1.73–119.27), p = 0.0137) and N-stage ((N2 vs. N0), HR = 2.35 (1.13–4.89), p = 0.0223) were significant for TTR (Table S4). The Immunoscore in two categories (Lo/Int+Hi) was significant in Cox multivariable analyses, with HR = 2.22 (1.10–4.55), p = 0.0269. Cox multivariable analyses for OS and DFS showed similar results, with significant p-values for Immunoscore in OS (p = 0.0304) and DFS (p = 0.0516).

The power of Immunoscore in OS was also evaluated using the contribution to risk. Indeed, variables with the most important relative contribution to the risk of death (Chi² proportion) were: Immunoscore (25%), T-Stage (28%), age (18%), MSI (17%), N-Stage (12%), and gender (1%). In multivariable analysis, the only variables shown to be of significant predictive value were Immunoscore, T-stage, and N-stage. Moreover, the predictive power of Immunoscore for recurrence (likelihood ratio test, p < 0.0001) and death (likelihood ratio test, p < 0.0001) was further strengthened when adding it to a model combining all clinical variables.

4. Discussion

The major prognostic impact of the immune contexture has been demonstrated in several studies [27,28,29]. The powerful assessment of immune cells in the tumor using digital pathology led to the international validation of the Immunoscore assay in stage I/II/III CC [22], as well as in stage III patients [24,25,30], and in two randomized phase 3 clinical trials [24,25]. The prognostic impact of the tumor microenvironment and Immunoscore has been clearly established, from pre-cancer lesions [31] to primary tumors [3,7,8,12,14,22,27,32] to metastasis [29,33,34,35,36]. The study complied with the STARD reporting guidelines (Table S5). Beyond the results obtained for stages I/II/III [3,8,22], for localized cancers [14,22], and for metastatic diseases (stage IV) [29,33,34,35,36,37], the relevance of the consensus Immunoscore in the Asian population remained to be established. Based on immune parameters alone, we highlighted the ability of the consensus Immunoscore to accurately layer all patients and, on an anatomopathological basis, defined high- and low-risk patients with significant differences in clinical survival. Interestingly, one of the most used tools in clinical oncology (i.e., MSI status) was shown to be dependent on the Immunoscore, as presented in our Cox multivariable analyses. We also found that the local intra-tumoral immune environment was not affected by patient age in the Asian population, thus contrasting with previous reports suggesting that peripheral and intra-tumoral immunity were known to decline over time [30,38].

In addition to its strong prognostic value, Immunoscore also predicted the response to chemotherapy in an international cohort study [30] and in a randomized phase 3 clinical trial [24]. Many guidelines include chemotherapy as a potent treatment for all stage III CC. Indeed, following surgical resection, the risk of death decreases by 10% to 15% when patients are treated with 5-FU and by 20% when treated with the oxaliplatin–fluoropyrimidine combination [39,40,41]. However, in stage III CC, a mere 20% of patients can benefit from adjuvant chemotherapy (AC), leaving 80% of patients susceptible to unneeded toxicity. In fact, 50% of those patients could be cured by surgery alone, and even with AC treatment, 30% of patients experience events of recurrence that lead to death within 2-3 years [42].

Previously, it has been shown that chemotherapy’s anti-tumoral activity is tightly linked to the immune response within the tumor, as it can modulate the immune system both positively and negatively [43,44,45,46]. Accordingly, Immunoscore was developed to help segregate patients who could benefit from chemotherapy. In this study, we showed that patients with better pre-built immunity (i.e., Intermediate and High Immunoscores) do benefit the most from chemotherapy, whereas Low-Immunoscore patients fall short in response to chemotherapy. Similar findings were observed in all stage III and low-risk and high-risk stage III patients, suggesting that effective chemotherapy partly relies on the modulation of the immune system and the high density of pre-existing tumor-infiltrating T cells, a hallmark of immune surveillance. Interestingly, none of the few patients with the highest Immunoscores (I4) relapsed, even when they were not treated with chemotherapy [30], supporting the idea of sparing these patients from unnecessary chemotherapy.

A limitation of the study might be the heterogeneity of the patient population, having come from three large countries, namely, China, India, and Japan. However, this non-randomized approach aimed at enhancing the robustness of the consensus Immunoscore within the Asian population. In particular, the use of chemotherapy and its impact on survival cannot be analyzed in an overall population including stages I, II, and III, since these patients have different outcomes and do not receive chemotherapy to equal extents (stage I does not receive chemotherapy, stage II may be provided with chemotherapy depending on risk factors, and stage III should undergo chemotherapy based on international recommendations).

For this aim, subgroup analysis has to be performed. However, in our present study, the sample size did not allow us to appropriately evaluate the benefit of chemotherapy.

Within stage III (n = 105 patients), only 11 did not receive chemotherapy. This is related to different reasons, such as patient refusal or a critical health condition. Sub-dividing these 11 patients into Immunoscore categories would not lead to ultimate statistical conclusions. Indeed, there was no significant difference in survival between patients receiving or not receiving chemotherapy in this cohort.

Within stage II (n = 251 patients), 148 received chemotherapy. Since the use of chemotherapy is not recommended for all stage II patients, decisions were made based on risk factors. So far, no randomized studies have shown a significant benefit of chemotherapy within the subgroup of stage II patients. Thus, the limited number of patients analyzed herein would not provide statistical conclusions, and much larger stage II groups should be analyzed to reach significant conclusions.

Moving forward, it will be important to further validate the standardized Immunoscore assay in randomized clinical trials of stage II and/or III CC treated with adjuvant chemotherapy in the Asian population [47,48].

A meta-analysis of the prognostic value of Immunoscore conducted on more than 10,000 patients confirmed that Immunoscore provided a reliable estimate of the recurrence risk in colon cancer [23].

5. Conclusions

The present study further enhances the clinical utility of Immunoscore in Asian CC patients. Developed as an in vitro diagnostic test, the Immunoscore assay is available in FDA CLIA-certified laboratories and in China for clinical use (CE-IVD). Moreover, the 5th edition of the WHO Digestive System Tumours classification introduced, for the first time, the immune response as “essential and desirable diagnostic criteria for colorectal cancer” while citing the consensus Immunoscore as the “best clinical evidence in colon cancer”. In fact, Immunoscore was also introduced in the 2020 ESMO Clinical Practice Guidelines for colon cancer patient support, allowing physicians to refine the prognosis and thus adjust the chemotherapy decision-making process in stage II and even in low-risk stage III patients [49]. Recently, Immunoscore was introduced into the Pan-Asian-adapted ESMO Clinical Practice Guidelines for the diagnosis, treatment, and follow-up of patients with localized colon cancer. Immunoscore was considered for its full-range indication in colon cancers, stage II and stage III, with the inclusion of all risk groups [50]. Supported by the multicentric international SITC study, the results of Immunoscore in the Asian population and its recent inclusion in the above-cited guidelines argue for the benefit of implementing Immunoscore in routine clinical practice as well as its introduction in other international guidelines. This would allow patients and physicians to benefit from this powerful predictive tool in colorectal cancer support.

6. Patents

J.G., F.P. and B.M. have patents associated with the immune prognostic biomarkers. Immunoscore^® is a registered trademark owned by the National Institute of Health and Medical Research (INSERM) and licensed to Veracyte. Michael Roehrl is a member of the Scientific Advisory Boards of Azenta and Universal DX. All other authors declare no conflict of interest.