The impact of preoperative treatments on the immune environment of rectal cancer

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ The impact of preoperative treatments on the immune environment of rectal cancer © 2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. Published version Wirta, Erkki‐Ville; Elomaa, Hanna; Ahtiainen, Maarit; Hyöty, Marja; Seppälä, Toni T.; Kuopio, Teijo; Böhm, Jan; Mecklin, Jukka‐Pekka; Väyrynen, Juha P. Wirta, E., Elomaa, H., Ahtiainen, M., Hyöty, M., Seppälä, T. T., Kuopio, T., Böhm, J., Mecklin, J., & Väyrynen, J. P. (2024). The impact of preoperative treatments on the immune environment of rectal cancer. Apmis, Early View. https://doi.org/10.1111/apm.13467 2024 The impact of preoperative treatments on the immune environment of rectal cancer ERKKI-VILLE WIRTA, 1,2 HANNA ELOMAA, 3,4 MAARIT AHTIAINEN, 5 MARJA HY € OTY, 1,2 TONI T. SEPP € AL € A, 1,2,6,7 TEIJO KUOPIO, 5 JAN B € OHM, 5 JUKKA-PEKKA MECKLIN 4,8 and JUHA P. V € AYRYNEN 9 1 Department of Gastroenterology and Alimentary Tract Surgery; 2 Faculty of Medicine and Health Technology, Tampere University and Tays Cancer Center, Tampere University Hospital, Tampere; 3 Department of Biological and Environmental Science, University of Jyv€ askyl€ a; 4 Department of Education and Research, The Wellbeing Services of Central Finland; 5 Department of Pathology, Wellbeing Services County of Central Finland, Jyv€ askyl€ a; 6 Department of Gastrointestinal Surgery, Helsinki University Central Hospital, University of Helsinki; 7 Applied Tumor Genomics, Research Program Unit, University of Helsinki, Helsinki; 8 Faculty of Sport and Health Sciences, University of Jyv€ askyl€ a, Jyv€ askyl€ a; and 9 Translational Medicine Research Unit, Medical Research Center Oulu, Oulu University Hospital and University of Oulu, Oulu, Finland Wirta E-V, Elomaa H, Ahtiainen M, Hy€ oty M, Sepp€ al€ a TT, Kuopio T, B€ ohm J, Mecklin J-P, V€ ayrynen JP. The impact of preoperative treatments on the immune environment of rectal cancer. APMIS. 2024. To improve local disease control, the use of preoperative radiotherapy either alone or combined with chemotherapy has become standard practice in rectal cancer, but it is unclear how these treatments modify the antitumoral immune response. We aimed to evaluate tumor histopathologic features and the prognostic effect of host immune response in rectal cancer with variable treatment modalities. Ninety-five rectal cancers with short-course radiotherapy (SRT), 97 with long-course chemoradiotherapy (CRT), and 154 without preoperative treatments, were evaluated for histopathologic features including Crohn’s-like reaction (CLR). CD3+and CD8+immunohistochemistry and tumor cells were analyzed from tumor tissue microarray samples to calculate T-cell densities and G-cross function values to estimate cancer cell–T-cell co-localization (proximity score). We found that lymphocyte densities were diminished after SRT, but CLR was scarcer after CRT. Proximity score and CLR density were prognostic for survival in cancer without preoperative treatments and could be combined into an enhanced prognostic score (immune grade). In the irradiated tumors, CLR density remained prognostic while the impact of T-cell infiltration was insufficient alone. In multivariable analysis, the immune grade proved to be an independent prognostic factor for survival. In conclusion, the immune contexture of rectal cancer harbors prognostic significance even after preoperative radiotherapy. Key words: Tumor-infiltrating lymphocytes; Crohn’s-like reaction; short-course radiotherapy; chemoradiotherapy; tumor regression. Erkki-Ville Wirta, Department of Gastroenterology and Alimentary Tract Surgery, Tampere University Hospital, El€ am€ anaukio, Kuntokatu 2, 33520 Tampere, Finland. e-mail: erkki-ville.wirta@tuni.fi Rectal cancer, although often referred to as part of colorectal cancer (CRC), has many unique features compared to colon cancer. The rectum gives rise to about one third of all CRCs, one of the most significant malignancies worldwide [1]. The partly extraperitoneal anatomic location in the small pelvis is surgically challenging, considering the limited space with adjacent pelvic vessels, nerves, and urinary and sexual organs [2]. In this regard, rectal cancer surgery is not only associated with high morbidity but also with a high local recurrence rate. The surgery first approach for Stages II–III cancer originally had a very high local recurrence rate. The appropriate surgical technique with total mesorectal excision (TME) reduces this risk, and the results for the early-stage disease are excellent, but the local recurrence rate for higher risk Stage III tumors without preoperative therapy remains at 20–30% [3, 4]. To further improve the locally advanced disease control, the use of preoperative radiotherapy either Received 28 November 2023. Accepted 29 August 2024 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 1 APMIS Ó2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. DOI 10.1111/apm.13467 alone or combined with chemotherapy has become standard practice in rectal cancer. The currently recommended preoperative treatments for locally advanced cancer are either short-course radiotherapy (SRT) of 25 Gy for 1 week (5 95 Gy), followed by surgery usually within 10 days of the first radiation fraction, or long-course chemoradiotherapy (CRT) of 45–50 Gy in 25–28 fractions typically combined with a fluoropyrimidine-based radiosensitizer, followed by surgery within 8–10 weeks of the last radiation fraction. CRT is advised when the circumferential margin is threatened according to the radiological assessment or there is a need for tumor regression to ensure an R0 resection [5]. Neoadjuvant CRT considerably improves the local control rate of the advanced disease, while this does not seem to improve the systemic control for overall survival, with distant metastases still occurring in 25–35% of patients [6]. The metastatic pattern in rectal cancer differs from that of colon originated disease. The venous drainage of the distal rectum through the iliac veins bypasses the portal system, and the iliac lymph nodes –a possible route for lymphatic cancer spread –are not removed during standard TME. Therefore, despite the liver being the most common site for distant metastasis, rectal cancer more frequently metastasizes to the lungs, nervous system, and bone, whereas peritoneal spreading is more common in colon cancer [7, 8]. Furthermore, differences in the carcinogenesis of rectal tumors have been recognized. Physical activity, lower body weight, and aspirin show a protective effect in colon cancer but not in rectal cancer [2]. Familial adenomatous polyposis causes cancers predominantly in the distal colon and rectum. However, most sporadic and Lynch syndrome-associated tumors with microsatellite instability (MSI) are generally hypermutated and characterized by a strong host immune reaction as a response to the high expression of neoantigens, and they arise usually in the proximal colon. In addition, BRAF V600E mutations are rare in rectal cancer [2, 9]. The host immune response has a well-established role in constraining cancer. High quantities of tumorinfiltrating CD3+and CD8+lymphocytes are associated with an improved prognosis in several cancer types [10], and this observation has led to numerous grading systems in an attempt to predict cancer behavior. Most notably, the ImmunoscoreÒ, derived from the CD3+and CD8+lymphocyte densities from the tumor invasive margin and center, has proven to be a strong prognostic marker in colon cancer [11].In addition, lymphoid aggregates, referred to as Crohn’- s-like lymphoid reaction (CLR) in CRC, play an important part in orchestrating the antitumoral reaction. CLR develops first from CD4+T-cell and antigen-presenting dendritic cell clusters, later including B cells and follicular dendritic cells, which through maturation form organized tertiary lymphoid structures with active germinal centers. CLR is more common in MSI tumors and related to the high amount of tumor infiltrating lymphocytes, as tertiary lymphoid structures enhance and sustain the antitumoral immune reaction by providing a local site for the tumor antigen presenting for dendritic cells, which lead to the activation, proliferation, and differentiation of T and B cells [12, 13]. However, it is unclear how preoperative radiotherapy in rectal cancer modifies the antitumoral immune response and associated cancer demeanor. The aim of this study was to evaluate the histopathologic features of rectal cancer after preoperative SRT or CRT, as well as the prognostic impact of the host immune response indicated by CRL and CD3+and CD8+lymphocytes. MATERIALS AND METHODS Patients The study population consisted of 346 rectal cancer patients, a part of a large cohort of 1479 CRC patients with surgical resection at Central Finland Central Hospital during 2000–2015 with recently reported age-adjusted Charlson comorbidity index (CCI) and associated multimodal management with updated survival data [14]. Adequate tumor samples for immunohistochemical studies were available for 1343 patients, of which 983 patients without preoperative therapy were analyzed in our previous study [15]. Here, we focus on rectal cancer patients, of whom 95 had preoperative SRT and 97 had CRT. For comparison, we included 154 rectal cancer patients without preoperative treatments, referred to as the nRT group (no radiotherapy). Complete responses were excluded from the study. Radical surgery was distributed as R0 (clean specimen marginals, n =298), R1 (tumor growth to less than 1 mm from specimen marginals, n =16), and R2 (unresectable primary tumor or distant metastasis, n =32). Histological tumor parameters; CLR density (according to V€ ayrynen et al. [16]); lymphovacular invasion (LVI); tumor regression grade (TRG, according to R€ odel et al. [17]); mucinous, stromal, and necrotic component; stroma maturity (according to Ueno et al. [18]); differentiation (according to the WHO criteria); and budding (according to the International Tumor Budding Consensus Conference [19]) were evaluated by a study pathologist (JPV) from hematoxylin and eosin (H&E)- stained whole slide samples. Stroma maturity was defined as: (i) mature with fine and elongated collagen fibers stratified into multiple layers; (ii) intermediate with keloid-like collagen intermingled with mature fibers; and (iii) immature consisting of myxoid stroma without mature fibers [18]. Immunohistochemical analyses Formalin-fixed paraffin-embedded (FFPE) tumor samples were used to prepare tissue microarray (TMA) blocks with 2Ó2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. WIRTA et al. 16000463, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apm.13467 by University Of Jyväskylä Library, Wiley Online Library on [11/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License a TMA Master II tissue microarrayer (3D Histech Ltd., Budapest, Hungary) containing two 1 mm diameter cores from representative areas of both the tumor center and the invasive margin. The blocks were then cut to 3.5 lm-thick sections and immunohistochemistry for CD3+and CD8+ T cells were performed by a BOND-III automated IHC stainer (Leica Biosystems, Buffalo Grove, IL, USA) with monoclonal antibodies and protocols, as described by Elomaa et al. [15] Immunohistochemical screening for DNA mismatch repair (MMR) deficiency with MLH1, MSH2, MSH6, and PMS2 expressions and for BRAF V600E mutation status was performed according to Sepp€ al€ a et al. [20]. CD3 and CD8 immunohistochemistry was analyzed by supervised machine learning built with the open-source bioimage analysis software QuPath [21] utilizing previously validated algorithms [15, 22]. The T-cell density score (DS) was calculated from the densities of CD3 + cells in the tumor center, CD3 + cells in the invasive margin, CD8 + cells in the tumor center, and CD8 + cells in the invasive margin, which were converted to percentiles (0–100) according to the principles of Immunoscore [11]. DS was determined by calculating the mean of the four percentiles and categorizing it into three groups: low (0–25), intermediate (>25–70) and high (>70–100) [15]. In our previous study, we introduced the T-cell proximity score (PS) as a measurement of tumor cellT-cell co-localization, which was associated with longer cancer-specific survival independent of T-cell densities. PS was calculated based on the G-cross (G Tumor:immune cell ) function values at a 20-lm radius (evaluating the likelihood of any tumor cell in the sample having at least one immune cell of the specified type within 20 lm radius), converted to percentiles and categorized into three groups (0–25, >25–70, and >70–100) [15]. CLR density, indicating the number of CLR follicles divided by the length of the analyzed invasive front, has been identified as a significant prognostic marker in CRC [16]. We selected a cutoff value of 0.25 follicles/mm obtained from a receiver operating characteristic (ROC) curve drawn in relation to disease-specific mortality. Examples of CLR density and T-cell proximity score analysis are shown in Fig. 1. CLR density, describing the local guidance of the host immune response, and PS, describing direct T cell-to-tumor cell interaction, were combined to form a more comprehensive parameter of the tumor immune environment, here referred to as the immune grade (IG). PS0 (0–25%), PS1 (>25–70%), and PS2 (<70–100%) were increased one category higher if the CLR density exceeded 0.25 follicles/mm, thus forming a four-step scale of IG. Statistical analysis Categorical data were compared using Pearson’s chi-square test. The Spearman correlation coefficient was used to determine correlations between immune cell infiltration and different treatment strategies. The Kaplan–Meier method was used to calculate disease-specific survival (DSS) and overall survival (OS), and the log-rank test was used to compare differences. A p-value of <0.05 was considered statistically significant. Survival times were from the date of surgery to the time of death or to the end of the follow-up. Survival analysis included only cancers with R0 resection, and cases with immediate postoperative deaths (n =5) were excluded. Multivariable Cox proportional hazards regression models were used to analyze prognostic factors for DSS and OS. Statistical analyses were performed using IBM SPSS Statistics (version 27.0; SPSS Inc., Chicago, IL, USA). RESULTS Clinicopathological features The median age of the patients was 69 (interquartile range, IQR 61–77) with a slight overrepresentation of the male gender (63%). The median follow-up time after surgery was 6.9 years (IQR 3.1–10.5). Clinicopathological variables in the different treatment groups are shown in Table 1. There was no statistically significant difference between the treatment groups according to age or sex. However, a higher CCI predicted exclusion from preoperative treatments (p =0.009). When considering tumor-associated parameters, there were no significant differences between treatment groups in TNM stage distribution, surgical radicality, tumor size, tumor budding, LVI, mucinous component, stroma maturity, or occurrence of distant metastases during follow-up. Increased proportions of tumor necrosis and intratumoral stroma as well as poorer tumor differentiation was seen after SRT and CRT (p <0.001). T-cell density score (DS) was lower after radiotherapy as DS0 was seen in 53% of the tumors after SRT and 24% after CRT vs 17% in the nRT group (p <0.001). DS and T-cell proximity score (PS) were distributed similarly in the nRT group. However, a slight shift toward higher PS was seen in SRT between DS and PS, but after CRT, the highest PS became less frequent. For CLR density, high densities were seen in 66% of the nRT group, in 56% after SRT, and only in 22% after CRT (p <0.001). The immune grade (IG) was generally lower after SRT and CRT, as IG0 was seen in 7% and Grade 3 in 20% of the directly operated tumors compared to IG0 in 22% and IG3 in 4% after SRT and IG0 in 20% and IG3 in 2% after CRT (Table 1;p<0.001). The calculated median densities of CD3+and CD8+cells after different pretreatment modalities are shown in Fig. S1, and Table S1 presents corresponding correlation analyses. Overall, lower CD3+ lymphocyte counts were observed after both SRT and CRT compared to the nRT group. Although a markedly diminished amount of CD8+cells were seen after SRT, there was no difference between the nRT and CRT groups. Immune contexture and association with histopathologic features Histopathologic features according to T-cell proximity score are shown in Table S2. Higher PS was associated with lower tumor grade and local tumor infiltration, less tumor spread to the lymph nodes, Ó2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. 3 THE IMMUNE ENVIRONMENT OF RECTAL CANCER 16000463, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apm.13467 by University Of Jyväskylä Library, Wiley Online Library on [11/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 4Ó2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. WIRTA et al. 16000463, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apm.13467 by University Of Jyväskylä Library, Wiley Online Library on [11/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License less distant metastasis, and overall lower TNM stage (p =0.005, p <0.001, p =0.013, p =0.007, and p <0.001 respectively). Also, LVI and tumor budding were more common with a lower PS (p <0.001 for both). Intratumoral stroma was more abundant in tumors with a low PS (p =0.004), yet almost in all (84%) tumors with high PS, the intratumoral stroma was composed of mature collagen fibers, while in tumors with a low PS, the stroma more commonly (in 49%) consisted of myxoid stroma without mature collagen (p <0.001). High CLR density was associated with higher PS (p =0.001). Significant associations were not identified between the PS and tumor size, tumor regression grade, mucinous tumor type or the amount of intratumoral necrosis. Only four tumors were MMR deficient, none with a high PS. The BRAF V600E mutation was found in nine tumors, three with PS0 and six with PS1. The occurrence of distant metastasis during the follow-up was significantly more infrequent with the highest PS (p =0.008). Table 2shows similar results with the immune grade. A higher IG was associated with lower TNM stage, less LVI and tumor budding, less intratumoral, yet more mature stroma and lower tumor grades (p =0.001, p =0.014, p <0.001, p=0.002, and p =0.010, respectively). Local recurrences and distant metastasis were more common with a lower IG (p =0.039 and p <0.001, respectively). Univariable survival analysis In the whole study population, CRT was associated with worse long-term survival (10-year DSS was 62% for CRT, 85% for SRT, and 77% for nRT, p=0.006). For meaningful case numbers in the further univariable survival analysis, the SRT and CRT groups were combined as the preoperative RT group (pRT). Kaplan–Meier survival analyses with clinicopathological variables for the nRT and pRT groups are shown in Table 3. As expected, a higher CCI was prognostic for a worse 5-year OS in both (p =0.003 for the nRT group and p <0.001 for the pRT group). A high TNM stage was significantly associated with a worse 5-year DSS and OS in the nRT group (p =0.005 and p =0.008, respectively) and with a worse DSS in the pRT group (p =0.004). Higher tumor budding was related to a worse DSS and OS in the nRT group (p <0.001 and p =0.008, respectively), but statistically significant differences in the pRT group were not identified (even when observing SRT and CRT separately). LVI was clearly associated with a higher disease-specific mortality in both treatment groups (p <0.001 for the nRT group and p=0.007, for the pRT group). Additionally, the 5-year OS was worse if LVI was identified, but significantly only within the nRT group (p =0.044). Tumor grade did not have statistical significance in any of the group analyses (Table 3). When considering immunological parameters, the density score had no distinct impact on survival. The proximity score performed well in the nRT group (5-year DSS for PS0 50% vs PS2 97%, p<0.001, and 5-year OS for PS0 39% vs PS2 85%, p=0.001), but it did not adduce a clear survival benefit in the pRT group (Table 3). However, when evaluating treatment groups separately, clear trend for improved survival for high PS was observed in CRT group (10-year DSS for high PS 92% vs 57% for combined PS0-1, p =0.072). In SRT group high PS was rare (only six tumors) and no significant differences were identified. High CLR density was associated with improved 5-year survival in all analyses (DSS 70% vs 90%, p =0.006, for the nRT group; 78% vs 93%, p =0.010, for the pRT group; OS 59% vs 83%, p =0.005, for the nRT group, and 65% vs 86%, p =0.003, for the pRT group). Immune grade, that is, a combination of the PS and CRL density, improved the identification of the prognostic extremities especially in the nRT group (Table 3). Because of the similar outcomes (Fig. 2, Table 3), IG0 to IG1 and IG2 to IG3 were combined as low and high immune grades. In nRT group, the 5-year DSS was for IG low 68% and for IG high 92% (p =0.001) and the 5-year OS for IG low 59% and for IG high 83% (p =0.005). In pRT group, the 5-year DSS for IG low was 78% and for IG High 93% (p =0.003), and the 5-year OS for IG low was 68% and for IG High 81% (p =0.027). Multivariable survival analysis Multivariable analysis with Cox proportional hazard model is shown in Table 4. The selected variables with significance in both treatment groups Fig. 1. Crohn’s like reaction density and T-cell proximity score analysis. (A) Analysis of Crohn’s like reaction density from a hematoxylinand eosin-stained whole-slide image. (B) Examples of CD3 and CD8 immunohistochemistry images from the tumor center and invasive margin of a single tumor. (C) Corresponding phenotyping maps for T cells, tumor cells, and other cells. (D) G-cross function curves, representing the likelihood of any tumor cell being co-located with at least one T cell within radius r. (E) T-cell proximity score calculation chart. G-cross function values at r =20 lm are converted into percentiles and, according to the mean of the four percentile values, the tumor is given a T-cell proximity score. Ó2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. 5 THE IMMUNE ENVIRONMENT OF RECTAL CANCER 16000463, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apm.13467 by University Of Jyväskylä Library, Wiley Online Library on [11/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Table 1. Clinicopathological variables in the different treatment groups nRT SRT CRT All p N of total 154 (% of column) N of total 95 (% of column) N of total 97 (% of column) N of total 346 (% of column) Age <65 41 (27) 40 (42) 41 (42) 122 (35) 0.053 65–75 61 (40) 32 (34) 30 (34) 123 (36) >75 52 (34) 23 (24) 26 (24) 101 (29) Sex Male 92 (60) 61 (64) 66 (68) 219 (63) 0.404 Female 62 (40) 34 (36) 31 (32) 127 (37) CCI 0–2 49 (32) 41 (44) 46 (47) 136 (40) 0.009 3 41 (27) 28 (30) 31 (32) 100 (29) ≥4 63 (41) 25 (27) 20 (21) 108 (31) TNM stage I 53 (34) 23 (24) 20 (21) 96 (28) 0.121 II 40 (26) 34 (36) 29 (30) 103 (30) III 42 (27) 31 (33) 34 (35) 107 (31) IV 19 (12) 7 (7) 14 (14) 40 (12) Radicality of surgery R0 130 (84) 88 (93) 80 (83) 298 (86) 0.265 R1 7 (5) 3 (3) 6 (6) 16 (5) R2 17 (11) 4 (4) 11 (11) 32 (9) T-cell density score 0 (low) 26 (17) 50 (53) 23 (24) 99 (29) <0.001 1 (intermediate) 93 (60) 43 (45) 50 (52) 186 (54) 2 (high) 35 (23) 2 (2) 24 (25) 61 (18) T-cell proximity score 0 (low) 26 (17) 37 (39) 22 (23) 85 (24.5) <0.001 1 (intermediate) 92 (60) 52 (55) 59 (61) 203 (58.5) 2 (high) 36 (23) 6 (6) 16 (16) 58 (17) Immune grade 0 11 (7) 21 (22) 19 (20) 51 (15) <0.001 1 51 (33) 35 (37) 46 (47) 132 (38) 2 61 (40) 35 (37) 30 (31) 126 (36) 3 31 (20) 4 (4) 2 (2) 37 (11) CLR density Low 52 (34) 42 (44) 76 (78) 170 (49) <0.001 High 102 (66) 53 (56) 21 (22) 176 (51) Tumor budding 0–4/0.785 mm 2 101 (66) 46 (48) 64 (66) 211 (61) 0.057 5–9/0.785 mm 2 33 (21) 29 (31) 22 (23) 84 (24) ≥10/0.785 mm 2 20 (13) 20 (21) 11 (11) 51 (15) Lymphovascular invasion No 118 (77) 77 (81) 79 (81) 274 (79) 0.573 Yes 36 (23) 18 (19) 18 (19) 72 (21) Tumor grade 1 46 (30) 11 (12) 13 (13) 70 (20) <0.001 2 97 (63) 74 (78) 69 (71) 240 (69) 3 11 (7) 10 (11) 15 (16) 36 (10) Mucinous tumor <50% 149 (97) 90 (95) 88 (91) 327 (95) 0.123 ≥50% 5 (3) 5 (5) 9 (9) 19 (5) Stroma maturity 1 (mature) 96 (62) 53 (56) 60 (62) 209 (60) 0.700 2 (intermediate) 22 (14) 20 (21) 15 (15) 57 (17) 3 (immature) 36 (23) 22 (23) 22 (23) 80 (23) Tumor size 0–40 mm 94 (63) 43 (46) 54 (56) 191 (57) 0.037 >40 mm 55 (37) 50 (54) 42 (44) 147 (43) 6Ó2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. WIRTA et al. 16000463, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apm.13467 by University Of Jyväskylä Library, Wiley Online Library on [11/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License according to the Kaplan–Meier analysis included TNM stage, LVI, and IG together with the constant variables age-adjusted CCI and sex. Additionally, the given preoperative therapy and TRG describing the achieved local effect at the time of the surgery were considered as key factors in standardizing the multivariable model. Immune grade was an independent prognostic factor for survival [IG low hazard ratio (HR) 3.17, 95% confidence interval (CI) 1.76–5.71, p <0.001 for DSS and HR 1.96, 95% CI 1.37–2.82, p <0.001 for OS]. High TNM stage was prognostic for worse survival (Stage IV HR 14.67, 95% CI 5.00–43.08, p<0.001 for DSS and HR 5.63, 95% CI 2.52– 12.56, p <0.001 for OS). Also, LVI impaired both DSS (HR 2.74, 95% CI 1.58–4.78, p <0.001) and OS (HR 1.52, 95% CI 1.01–2.29, p =0.044). Compared to CRT, patients with SRT had improved DSS (HR 0.25, 95% CI 0.11–0.57, p =0.004) and both nRT and SRT groups had better OS (HR 0.51, 95% CI 0.32–0.83 for both, p =0.010). Those with higher CCI had worse OS (CCI ≥4 HR 3.23, 95% CI 2.14–4.89, p <0.001). DISCUSSION We aimed to better understand the changes induced by preoperative (chemo)radiotherapy to the immune contexture of rectal cancer. The prognostic effect of the tumor infiltrating lymphocytes is well documented in colon cancer. However, the rectum differs embryologically, anatomically, and functionally from the colon, and they cannot be compared unconditionally. In addition, the (chemo)radiotherapy often administered in rectal cancer might affect the quality of the tumor-constraining immune response and subsequent survival benefit. Our results indicate that analysis of tumor immune contexture after preoperative treatments can still assist in predicting disease outcome in rectal cancer, as the combination of PS and CLR density (immune grade), was an independent prognostic factor for DSS and OS. Radiotherapy is known to cause a wide spectrum of changes in the tumor microenvironment, some favorable and some harmful for the antitumoral struggle. Irradiation-generated oxidative stressors cause damage to cancer cell DNA, followed by cell death, for example, through apoptosis, mitotic catastrophe, or cellular stress-induced permanent cell cycle arrest [23]. The dying cells express damage-associated molecular patterns (DAMP), which trigger the antitumoral immune response and immunogenic cell death. DAMPs enhance dendritic cell (DC) function with subsequent release of proinflammatory cytokines that activate cytotoxic T lymphocytes [23, 24]. As DCs are central operators in triggering the antitumoral immune response, the prognostic effect of CLRs functioning as local platforms for DC antigen presentation is not surprising. Tumor-induced neovasculature is often more prone to irradiation damage because of the fast rate of vascular endothelial cell proliferation with an immature structure. The destruction of tumor vasculature causes hypoxia, which reduces the irradiation-induced production of reactive oxygen Table 1 (continued) nRT SRT CRT All p N of total 154 (% of column) N of total 95 (% of column) N of total 97 (% of column) N of total 346 (% of column) Tumor necrosis <5% 32 (21) 5 (5) 12 (12) 49 (14) <0.001 5–15% 106 (69) 59 (62) 66 (68) 231 (67) >15% 16 (10) 31 (33) 19 (20) 66 (19) Intratumoral stroma <50% 70 (46) 21 (22) 29 (30) 120 (35) <0.001 ≥50% 84 (55) 74 (78) 68 (70) 226 (65) Local recurrence No 116 (89) 85 (97) 71 (89) 272 (91) 0.108 Yes 14 (11) 3 (3) 9 (11) 26 (9) Distant metastasis No 96 (74) 70 (80) 51 (64) 217 (73) 0.067 Yes 34 (26) 18 (20) 29 (36) 81 (27) CCI, Charlson comorbidity index; CLR, Crohn’s like reaction; CRT, chemoradiotherapy; nRT, no radiotherapy; SRT, short-course radiotherapy. CCI was determined without including the current colorectal cancer. CCI is missing from two patients. Tumor size is unknown in eight tumors. Local recurrence and occurrence of distant metastasis during follow-up are evaluated from only R0 resected patients (n =298). Ó2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. 7 THE IMMUNE ENVIRONMENT OF RECTAL CANCER 16000463, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apm.13467 by University Of Jyväskylä Library, Wiley Online Library on [11/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Table 2. Histopathologic features according to the immune grade 1 2 3 4 All p N of total 51 (% of column) N of total 132 (% of column) N of total 126 (% of column) N of total 37 (% of column) N of total 346 (% of column) T 1 1 (2) 8 (6) 12 (10) 6 (16) 27 (8) 0.001 2 8 (16) 27 (20) 37 (29) 18 (49) 90 (26) 3 37 (72) 85 (64) 67 (53) 13 (35) 202 (58) 4 5 (10) 12 (9) 10 (8) 0 (0) 27 (8) N 0 26 (51) 69 (52) 91 (72) 26 (70) 212 (61) 0.013 1 17 (33) 39 (30) 21 (17) 9 (24) 86 (25) 2 8 (16) 24 (18) 14 (11) 2 (5) 48 (14) M 0 42 (82) 114 (86) 113 (90) 37 (100) 306 (88) 0.059 1 9 (18) 18 (14) 13 (10) 0 (0) 40 (12) TNM stage 1 8 (16) 26 (20) 44 (35) 18 (49) 96 (28) 0.001 2 16 (31) 38 (29) 41 (33) 8 (22) 103 (30) 3 18 (35) 50 (38) 28 (22) 11 (30) 107 (31) 4 9 (18) 18 (14) 13 (10) 0 (0) 40 (12) Radicality of surgery R0 44 (86) 107 (81) 110 (87) 37 (0) 298 (86) 0.165 R1 2 (4) 8 (6) 6 (5) 0 (0) 16 (5) R2 5 (19) 17 (13) 10 (8) 0 (0) 32 (9) MMR status MMR proficient 51 (100) 130 (99) 124 (98) 37 (100) 342 (99) 0.709 MMR deficient 0 (0) 2 (1) 2 (2) 0 (0) 4 (1) BRAF Wild type 50 (98) 127 (96) 123 (98) 37 (100) 337 (97) 0.609 Mutation 1 (2) 5 (4) 3 (2) 0 (0) 9 (3) TRG Fibrosis <25% 26 (65) 43 (55) 51 (79) 5 (83) 125 (66) 0.078 Fibrosis 25– 50% 9 (23) 25 (32) 8 (12) 0 (0) 42 (22) Fibrosis >50% 5 (13) 10 (13) 6 (9) 1 (17) 22 (12) LVI No 37 (73) 97 (73) 105 (83) 35 (95) 274 (79) 0.014 Yes 14 (27) 35 (27) 21 (17) 2 (5) 72 (21) Tumor budding 0–4/0.785 mm 2 23 (45) 68 (52) 86 (68) 34 (92) 211 (61) <0.001 5–9/0.785 mm 2 16 (31) 44 (33) 21 (17) 3 (8) 84 (24) ≥10/0.785 mm 2 12 (24) 20 (15) 19 (15) 0 (0) 51 (15) Tumor size ≤40 mm 27 (54) 57 (44) 48 (39) 15 (44) 147 (43) 0.332 >40 mm 23 (46) 73 (56) 76 (61) 19 (56) 191 (57) Mucinous tumor 0–49% 48 (94) 123 (93) 119 (94) 37 (100) 327 (94) 0.455 50–100% 3 (6) 9 (7) 7 (6) 0 (0) 19 (6) Tumor necrosis <5% 4 (8) 18 (14) 21 (17) 6 (16) 49 (14) 0.164 5–15% 30 (59) 92 (70) 84 (67) 25 (68) 231 (67) >15% 17 (33) 22 (17) 21 (17) 6 (16) 66 (19) Stroma maturity 0 (mature) 19 (37) 70 (53) 86 (68) 34 (92) 209 (60) <0.001 1 (intermediate) 5 (10) 32 (24) 18 (14) 2 (5) 57 (17) 2 (immature) 27 (53) 30 (23) 22 (18) 1 (3) 80 (23) Intratumoral stroma <50% 10 (20) 37 (28) 55 (44) 18 (49) 120 (35) 0.002 ≥50% 41 (80) 95 (72) 71 (56) 19 (51) 226 (65) 8Ó2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. WIRTA et al. 16000463, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apm.13467 by University Of Jyväskylä Library, Wiley Online Library on [11/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 40. Posselt R, Erlenbach-W€ unsch K, Haas M, Jeßberger J, B€ uttner-Herold M, Haderlein M, et al. Spatial distribution of FoxP3+and CD8+tumour infiltrating T cells reflects their functional activity. Oncotarget. 2016;7:60383–94. https://doi.org/10.18632/oncotarget. 11039 41. Chiang SF, Huang CY, Ke TW, Chen TW, Lan YC, You YS, et al. Upregulation of tumor PD-L1 by neoadjuvant chemoradiotherapy (neoCRT) confers improved survival in patients with lymph node metastasis of locally advanced rectal cancers. Cancer Immunol Immunother. 2019;68:283–96. https://doi.org/10. 1007/s00262-018-2275-0 42. Liu L, Liu L, Yao HH, Zhu ZQ, Ning ZL, Huang Q. Stromal myofibroblasts are associated with poor prognosis in solid cancers: a meta-analysis of published studies. PLoS One. 2016;11:e0159947. https://doi. org/10.1371/journal.pone.0159947 43. Tsujino T, Seshimo I, Yamamoto H, Ngan CY, Ezumi K, Takemasa I, et al. Stromal myofibroblasts predict disease recurrence for colorectal cancer. Clin Cancer Res. 2007;13:2082–90. https://doi.org/10. 1158/1078-0432.CCR-06-2191 44. Trotsyuk I, Sparschuh H, M€ uller AJ, Neumann K, Kruschewski M, Horst D, et al. Tumor budding outperforms ypT and ypN classification in predicting outcome of rectal cancer after neoadjuvant chemoradiotherapy. BMC Cancer. 2019;19:1033. https://doi. org/10.1186/s12885-019-6261-5 45. Du C, Xue W, Li J, Cai Y, Gu J. Morphology and prognostic value of tumor budding in rectal cancer after neoadjuvant radiotherapy. Hum Pathol. 2012;43:1061– 7. https://doi.org/10.1016/j.humpath.2011.07.026 46. Roth AD, Tejpar S, Delorenzi M, Yan P, Fiocca R, Klingbiel D, et al. Prognostic role of KRAS and BRAF in stage II and III resected colon cancer: results of the translational study on the PETACC-3, EORTC 40993, SAKK 60-00 trial. J Clin Oncol. 2010;28:466–74. https://doi.org/10.1200/JCO.2009.23.3452 47. Sclafani F, Wilson SH, Cunningham D, Gonzalez de Castro D, Kalaitzaki E, Begum R, et al. Analysis of KRAS, NRAS, BRAF, PIK3CA and TP53 mutations in a large prospective series of locally advanced rectal cancer patients. Int J Cancer. 2020;146:94–102. https://doi.org/10.1002/ijc.32507 SUPPORTING INFORMATION Additional supporting information may be found online in the Supporting Information section at the end of the article. Fig. S1. Median densities of CD3+and CD8+ immune cells in different treatment groups: no radiotherapy (nRT), short-course radiotherapy (SRT), and long-course chemoradiotherapy (CRT). Table S1. Correlation between lymphocyte counts and treatment groups. Table S2. Histopathologic features according to the Proximity score. Ó2024 The Author(s). APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology. 15 THE IMMUNE ENVIRONMENT OF RECTAL CANCER 16000463, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apm.13467 by University Of Jyväskylä Library, Wiley Online Library on [11/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License