ABSTRACT Background The current management on rectal cancer based on TNM staging has some limitations. In early rectal cancer T stage can be not sufficient to predict the nodal status and, in locally advanced rectal cancer after neoadjuvant therapy the persistence of a complete response to therapy cannot be accurately predicted by the simple tumor regression grade. For both cases the current guidelines recommend the complete rectal resection with a total mesorectal excision. The implications for patients' quality of life are evident even in case of sphincter sparing surgery. Moreover, in both cases the cancer sample available for the analysis can be small or inexistent.

Hypothesis The main hypothesis underlying our research is that the aggressiveness of rectal cancer is determined by the complex interactions between the malignant cells and their immune microenvironment. The second hypothesis is that relevant trace of this cross talk between tumor cells and immune microenvironment can be detected in the normal mucosa surrounding the cancer according to the concept of field cancerization.

Aims The aim of this project is to analyze the healthy rectal mucosa surrounding the cancer to identify traces of immunosurveillance mechanisms and of field cancerization and to use them to obtain a composite prognostic test to predict nodal metastasis in early rectal cancer and recurrence after complete response at neoadjuvant therapy in case of locally advanced rectal cancer.

Experimental Design This prognostic test will be constructed on the combinatory analysis of the transcriptome, immune and epithelial cells cross-talk, immune checkpoints and miRNA expression in normal rectal mucosa surrounding cancer. The aim is to predict the presence of nodal metastasis in patients with early rectal cancer. In step A, we will retrospectively analyze archival tissue samples in order to identify the most performing biomarkers; in step B, we will validate the prognostic performance of the markers identified in phase I through a prospective analysis of rectal mucosa specimen.

Expected Results The anticipated outcome of this project is to generate one or different combination of markers to optimize rectal management and to predict rectal cancer patients outcome more accurately than traditional TNM staging or tumore regression grade. We expect to obtain a prognostic test from normal tissue that accurately predicts rectal cancer behavior even in case when the tumor samples are scarce (early rectal cancer) or absent (complete response to therapy) to avoid unnecessary total rectal excision.

Impact On Cancer An immunoscore specific for rectal cancer may predict tumor progression and clinical outcome more accurately and may contribute to better design a personalized therapeutic algorithm. Moreover, nowadays patients with early rectal cancer without nodal involvement and patients with potential complete response to neoadjuvant therapy still undergo total rectal excision which is a risky procedure that impairs quality of life. The use of this new prognostic test may make possible to adopt a minimally invasive approach or even simple observation if nodal involvement or residual disease are reasonably excluded.

Background The introduction of new technologies for diagnosis and screening programs led to an increasing rate of early detection of colorectal cancer [9]. This, combined to the evolution of techniques of local excision, led to the development of new strategies to reduce treatment related morbidity, especially for early rectal cancer. In fact, transanal local excision has recently received attention as an alternative to radical surgery for early rectal cancer [10]. Nevertheless, the definition of early rectal cancer and its staging and treatment algorithm are still under debate [9]. In, fact, to minimize the recurrence rate after local excision of rectal cancer, the false-negative rate of nodal staging should be minimized [11]. MRI can be helpful to better identify candidates suitable for local excision of early stage rectal cancer. However, strict MRI criteria for oncologic safety might result in considerable false-positive cases and limit the application of local excision [11]. Therefore, different markers are warranted to overcome this critical point in early rectal cancer management. A recent Swiss study on 126 rectal biopsies analyzed with microarray aimed to investigate whether immunoprobing might help in predicting lymph node involvement in this subgroup. Their results suggested that in early rectal cancers absence of CD8+ T-cell infiltration helps in predicting patients' nodal involvement [12]. Last but not least, a recent systematic review provided data suggesting that selected patients with T1/T2N0M0 rectal cancer may undergo local excision without compromising the oncological outcome otherwise conferred by total mesorectal excision [13]. High risk patients (i.e. those with low CD8 infiltrate) might benefit of chemoradiation. In fact, while guideline-concordant adoption of local excision for treatment of low-risk stage I rectal cancer is increasing, its use is also increasing for higher-risk rectal cancers that do not meet guideline criteria.

Treatment with local alone is associated with poorer long-term OS and additional studies are warranted to understand risk factor of poor outcome after local excision [14].

Hypothesis The clinical question is how to predict the presence of nodal metastasis in patients with early rectal cancer. These patients may potentially benefit of a minimally invasive treatment, a trans anal resection, provided that the nodal status can be considered negative for metastasis.

Study design Tissue samples will be obtained from cancer tissue (if possible) and from normal rectal mucosa adjacent to the cancer at surgery. The very large number of patients need to answer the two questions imply a multicentric design. Complete medical record and follow-up will be collected from each center. The analysis will be centralized mostly in Azienda Ospedaliera di Padova. The study will be articulated in two parts each of them aiming to answer to one of the above described questions. Each part of the study will be articulated in a retrospective and exploratory step (A) and in a prospective validation step (B).

Inclusion Criteria:

1. T1 and T2 rectal carcinoma operated on will be included in the study group. 2. A minimal number of 10 retrieved lymph nodes will be required for the inclusion in case of anterior resection or Miles abdominal perineal resection 3. Full availability of clinical records and at least 1 year of follow up

-

STEP A Taking into account a 15% of early rectal cancer patients with nodal metastasis and a maximum of 5 prognostic factors (including markers and clinical factors) in the model, we aim to enroll 334 patients in this part of the study.

STEP B The sample size was calculated to demonstrate a non-inferiority of AUC with respect to an area of 0.80, assuming an AUC of 0.90 and a non-inferiority margin of 0.05. With a type I error of 0.05 and a type II error of 0.20, we aim to analyze 66 patients with lymph node metastasis (N+) and 66 patients without lymph node metastasis (N0). Given an expected proportion of 15% of N+ patients among early rectal cancer, we estimate to enroll a total number of 66/0.15=440 patients with early rectal cancer. However, we will perform flow cytometry, transcriptomic analysis, mRNA microarray and mutational profiling only in 132 patients (66 N+ and 66 N0) in order to test the non-inferiority of AUC.