Prediction of relapse-free survival according to adjuvant chemotherapy and regulator of chromosome condensation 2 (RCC2) expression in colorectal cancer

Background There is a need for improved selection of patients for adjuvant chemotherapy after resection of non-metastatic colorectal cancer (CRC). Regulator of chromosome condensation 2 (RCC2) is a potential prognostic biomarker. We report on the establishment of a robust protocol for RCC2 expression analysis and prognostic tumour biomarker evaluation in patients who did and did not receive adjuvant chemotherapy. Materials and methods RCC2 was analysed in 2916 primary CRCs from the QUASAR2 randomised trial and two single-hospital Norwegian series. A new protocol using fluorescent antibody staining and digital image analysis was optimised. Biomarker value for 5-year relapse-free survival was analysed in relation to tumour stage, adjuvant chemotherapy and the molecular markers microsatellite instability, KRAS/BRAFV600E/TP53 mutations and CDX2 expression. Results Low RCC2 expression was scored in 41% of 2696 evaluable samples. Among patients with stage I–III CRC who had not received adjuvant chemotherapy, low RCC2 expression was an independent marker of inferior 5-year relapse-free survival in multivariable Cox models including clinicopathological factors and molecular markers (HR 1.45, 95% CI 1.09 to 1.94, p=0.012, N=521). RCC2 was not prognostic in patients who had received adjuvant chemotherapy, neither in QUASAR2 nor the pooled Norwegian series. The interaction between RCC2 and adjuvant chemotherapy for prediction of patient outcome was significant in stage III, and strongest among patients with microsatellite stable tumours (pinteraction=0.028). Conclusions Low expression of RCC2 is a biomarker for poor prognosis in patients with stage I–III CRC and seems to be a predictive biomarker for effect of adjuvant chemotherapy.


: Associations between RCC2 and five-year RFS according to cellular localization.
A) The continuous nuclear and cytoplasmic RCC2 scores are correlated. B) Both nuclear and cytoplasmic expression of RCC2 were dichotomized at the 41 st percentile of scores in the Norwegian series 1, and analyzed by Kaplan-Meier survival analysis. Although both measures showed an association between low expression of RCC2 and poor patient outcome, cytoplasmic expression was a stronger predictor.
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) State the marker examined, the study objectives, and any pre-specified hypotheses. 2

Patients 2
Describe the characteristics (e.g., disease stage or co-morbidities) of the study patients, including their source and inclusion and exclusion criteria. Table 1 3

2-4 &
Describe treatments received and how chosen (e.g., randomized or rule-based). 2 Specimen characteristics 4 Describe type of biological material used (including control samples) and methods of preservation and storage.

5
Specify the assay method used and provide (or reference) a detailed protocol, including specific reagents or kits used, quality control procedures, reproducibility assessments, quantitation methods, and scoring and reporting protocols. Specify whether and how assays were performed blinded to the study endpoint.

Study design 6
State the method of case selection, including whether prospective or retrospective and whether stratification or matching (e.g., by stage of disease or age) was used. Specify the time period from which cases were taken, the end of the follow-up period, and the median follow-up time.

7
Precisely define all clinical endpoints examined. 4 8 List all candidate variables initially examined or considered for inclusion in models. 4-5 & Table 1 9 Give rationale for sample size; if the study was designed to detect a specified effect size, give the target power and effect size. 2 Statistical analysis methods 10 Specify all statistical methods, including details of any variable selection procedures and other model-building issues, how model assumptions were verified, and how missing data were handled.

4-5 11
Clarify how marker values were handled in the analyses; if relevant, describe methods used for cutpoint determination. 4-5

Data 12
Describe the flow of patients through the study, including the number of patients included in each stage of the analysis (a diagram may be helpful) and reasons for dropout. Specifically, both overall and for each subgroup extensively examined report the numbers of patients and the number of events.
S. Fig. 1 13 Report distributions of basic demographic characteristics (at least age and sex), standard (disease-specific) prognostic variables, and tumor marker, including numbers of missing values. Table 1 Analysis and presentation 14 Show the relation of the marker to standard prognostic variables. Tables  S4,S5 15 Present univariable analyses showing the relation between the marker and outcome, with the estimated effect (e.g., hazard ratio and survival probability). Preferably provide similar analyses for all other variables being analyzed. For the effect of a tumor marker on a time-to-event outcome, a Kaplan-Meier plot is recommended. For key multivariable analyses, report estimated effects (e.g., hazard ratio) with confidence intervals for the marker and, at least for the final model, all other variables in the model. Table 2 17 Among reported results, provide estimated effects with confidence intervals from an analysis in which the marker and standard prognostic variables are included, regardless of their statistical significance. Table 2 18 If done, report results of further investigations, such as checking assumptions, sensitivity analyses, and internal validation. 5

19
Interpret the results in the context of the pre-specified hypotheses and other relevant studies; include a discussion of limitations of the study.   Table 2, multivariable analysis, Figure 2D - MSI scoring by immunohistochemistry. Two 4-plex stains were performed in the Norwegian series 2 against the four mismatch repair (MMR) proteins, MLH1, MSH6, MSH2 and PMS2. Multiplexed fluorescent IHC was performed on 4 μm thick tissue sections by following the Opal protocol (PerkinElmer/Akoya) with a 4-plex kit (NEL810001KT, PerkinElmer/Akoya). The Opal protocol was followed except that slide deparaffinization, antigen retrieval and antibody stripping were all performed using EnVision™ FLEX Target Retrieval Solution 3in-1, high/low pH, (Agilent/DAKO, codes K8005 and K8004) in the PT-link module (Agilent/DAKO). Staining against MLH1 and MSH6 was performed as follows: initial slide deparaffinization/antigen retrieval in High pH buffer followed by detection of anti-MLH1 (undiluted, Clone ES05, Agilent/DAKO) using Opal 570, antibody stripping in High pH buffer followed by detection of anti-MSH6 (undiluted, Clone EP49, Agilent/DAKO) using Opal 520, antibody stripping in High pH buffer followed by detection of a cocktail of antibodies (epithelialcocktail) consisting of anti-E-cadherin (1:20.000, Clone 36, BD-Biosciences), anti-pan Cytokeratin (1:4000, C-11, Abcam) and anti-pan Cytokeratin Type I/II (1:2000, AE1/AE3, Thermo Fisher) using Opal 690. After a final antibody stripping procedure in Low pH buffer, nuclei were stained with DAPI and sections were mounted in Prolong TM Diamond Antifade Mountant (Thermo Fisher). The same protocol was followed for MSH2 (undiluted, Clone FE11, Agilent/DAKO; detected with Opal 570) and PMS2 (undiluted, Clone EP51, Agilent/DAKO; detected with Opal 520).
The two stains against the MMR proteins were imaged with the Vectra 3 imaging platform (PerkinElmer/Akoya), visually examined and scored either positive or negative for each of the four proteins. If a sample was negative in the tumor compartment for at least one of the four proteins it was deemed microsatellite instable (MSI). If staining against all proteins was present it was deemed microsatellite stable (MSS). This analysis was done blinded to the MSI status that was available for 339 patients and which were previously assessed by PCR. Of the 798 samples in the Norwegian series 2, 734 were evaluable for MSI status by immunohistochemistry (IHC). Within the 315 samples for which we had overlapping data between methods we found a concordance of 96.5% (data not shown). For patients who were scored into differing categories between methods, the results from PCR were used.
Standard immunohistochemistry in the QUASAR2 cohort. Chromogenic IHC, visualized with the chromogen DAB, was performed using a rabbit polyclonal antibody against RCC2 (Novus Biologicals,, and samples were scored visually according to the Allred method, as previously described. 1 Samples with a cytoplasmic RCC2 Allred score <5 were classified as having low expression of RCC2. Immunocytochemistry. An RCC2 knockout cell-line was purchased from Horizon Discovery along with its parental HAP1 cell-line. These were grown according to the vendor's protocol prior to formalin-fixation and paraffin-embedding. Immunocytochemistry on 4 μm thick sections of the FFPE cellular pellets was performed by indirect detection with DAB (3,3'-diaminobenzidine) on an Autostainer Link 48 System (Agilent/DAKO) with a PT-link module. Deparaffinization and antigen retrieval were carried out for 20 minutes at 97°C using EnVision™ FLEX Target Retrieval Solution 3-in-1, high pH, (Agilent/Dako, K8004), in 65°C preheat mode. Subsequent staining was performed with the EnVision FLEX kit (Agilent/DAKO, code K800021-2) together with anti-RCC2 (1:100, monoclonal rabbit, Clone D14F3, Cell Signaling) and FLEX+ linker for rabbit primary antibody (Agilent/DAKO, code K8009). Mayer's hematoxylin (diluted 1:10, Agilent/DAKO, code S3309) was used as counterstain. Finally, the slides were dehydrated in a graded ethanol series and clarified in xylene before they were mounted onto glass slides using Richard-Allan Scientific Cytoseal mounting media (Thermo Fisher Scientific).

RCC2 staining by multiplexed fluorescent IHC.
Multiplexed fluorescent IHC was performed on 4 μm thick tissue sections by following the Opal protocol (PerkinElmer/Akoya) with a 4-plex kit (NEL810001KT, PerkinElmer/Akoya). The Opal protocol was followed except that slide deparaffinization, antigen retrieval and antibody stripping were all performed using EnVision™ FLEX Target Retrieval Solution 3-in-1, high/low pH, (Agilent/DAKO, codes K8005 and K8004) in the PT-link module (Agilent/DAKO). 4-plex staining against RCC2 and Cortactin (data on Cortactin not used in the current study) was performed as follows: initial slide deparaffinization/antigen retrieval in High pH buffer followed by detection of anti-RCC2 (1:100, monoclonal rabbit, Clone D14F3, Cell Signaling) using Opal 570, antibody stripping in High pH buffer followed by detection of anti-Cortactin (1:500, monoclonal mouse, Clone 771716, R&D systems) using Opal 520, antibody stripping in High pH buffer followed by detection of a cocktail of antibodies (epithelial-cocktail) consisting of anti-E-cadherin (1:10000, Clone 36, BD-Biosciences), anti-pan Cytokeratin (1:2000, C-11, Abcam) and anti-pan Cytokeratin Type I/II (1:1000, AE1/AE3, Thermo Fisher) using Opal 690. After a final antibody stripping procedure in Low pH buffer, nuclei were stained with DAPI and sections were mounted in Prolong TM Diamond BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) Antifade Mountant (Thermo Fisher). Determination of optimal primary antibody concentrations and testing for complete removal of antibodies between sequential rounds of staining was performed on separate test-TMAs prior to staining of the patient series (data not shown). A negative control slide was included during the staining procedures, in which primary antibody was omitted. No signal above noise was detected in the negative control (data not shown).
Digital image analysis, Vectra/Inform system (Oslo University Hospital, OUH). The samples were multispectrally imaged using the Vectra 3 Imaging platform (PerkinElmer/Akoya) at 20x magnification. The resulting multispectral images were then analyzed in Inform software (version 2.3.0, PerkinElmer). Spectra derived from images of samples stained with each fluorophore individually were used to spectrally un-mix the 4plex images, and tissue auto-fluorescence was removed by using a spectrum derived from unstained tissue. Signal from the epithelial-cocktail and DAPI channels were used to train an algorithm to classify the tissue into epithelial (tumor) and stromal regions. Individual nuclei were segmented using the DAPI signal and cytoplasmic area was constructed by setting the inner distance to nuclei to 1 pixel and the outer distance to 10 pixels. Data obtained with this platform were used to perform all downstream analyses.
Scoring of digitally analyzed RCC2. All digitally analyzed scores were calculated as the mean signal intensity within the cytoplasmic area of the tumor tissue compartment for each patient sample. Dichotomization of the continuous scores was performed within each cohort by setting a threshold based on the proportions of patients in the strong and weak categories of cytoplasmic RCC2 staining originally reported. 1 This cutoff was at the 41 st percentile. Continuous RCC2 scores were standardized within each cohort (Norwegian series 1 & 2) by meancentering and scaling by the standard deviation. The range of continuous RCC2 scores in the NS1 after scaling was from -1.23 to 8.11, and in the NS2 from -1.02 to 7.15.
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)