Identification Of The Predictive Factors For Patients With Metastatic Colorectal Cancer Response To Dendritic Cell Therapeutic Vaccine

Colorectal cancer (CRC) is one of the most common solid tumors in the world and becomes the one of major cause of cancer-related mortality death. Over the past years, in spite of higher response rate have been achieved with the improvement in pharmaceutical strategies, relatively resistant to systemic chemotherapy is also lead to poor prognosis. Almost 50% of the patients died for the disease progression. Developing a novel strategy for the treatment of metastatic colorectal cancer is necessary. Dendritic cell therapeutic vaccine has been proved to be effective in treatment of various advanced cancer such as prostate cancer and melenoma, In 2004, a random phase II clinical trial of sequential administration of OXA+5-Fu and DC vaccine in treatment of metastatic colorectal cancer were approved by SFDA of China. From 2005 to 2008,150 patients with metastatic colorectal cancers enrolled in the clinical trial. The patients treated with OXA+5-Fu and DC vaccine achieved 45.07%(CR+PR) clinical response significantly higher than patients treated with chemotherapy alone 25.35%. However, the underlying mechanisms and the immunological characteristics of patients sensitive to DC vaccine remain to be unidentified.Microsatellite instability (MSI) cancers have unique characteristics compared with microsatellite stability (MSS) cancers, such as proximal anatomic location and severe inflammatory cell infiltration. Studies have also indicated that patients with MSI CRC achieve better prognosis compared with patients with MSS CRC. In this study, Tumor samples obtained at baseline from 101 patients were suitable for MSI and immunohistochemistry analysis. Of them 60 patients received sequential administration of OXA+5-Fu/LV and DC vaccine (Group A),41 received OXA+5-Fu/LV (Group B) only. Genomic DNA of tumor and normal tissue was extracted with Fixed Tissue Purification module Promega Kit. Analysis of MSI status in genomic DNA samples was using the fluorescence-labeled primers of panel of five consensus markers recommended for the detection of MSI in colorectal cancer. This panel of markers includes three dinucleotide markers (D2S123, D5S346, and D17S250) and two mononucleotide markers (BAT25 and BAT26). PCR products were denatured and loaded onto the ABI 3730 sequencer. The data were processed by the Gene Scan software. Samples were classified as MSI-high if instability was detected at two or more consensus markers, MSI-low if instability was confined to only one consensus marker, and microsatellite stable if none of the consensus markers revealed instability.101 patients were classified into three kinds of type status of tumor:MSI-H, MSI-L and MSS. Of them,10 MSI-H,13 MSI-L and 37MSS in group A; 4 MSI-H,15 MSI-L,22 MSS in group B. Patients with MSI-H in Group A achieved 80%(8/10) objective clinical response, while 4 patients achieved 50%(2/4) objective clinical response in Group B.37 patients with MSS in Group A achieved 43.24%(16/37)objective clinical response.Although the presence of immune infiltrates of variable content in human solid tumors from different origins and different patients has long been established, the prognostic value of these components is still controversial. In this study, the primary tumor lesion was resected by surgery and Immunohistochemistry (IHC) was performed using CD8, CD45RO,IL-17, IL-10 and Hsp70 antibodies.4-μm sections of Paraffin-embedded tissue were used for immunostaining. Staining for Hsp70(1: 75 dilution, clone N27F3-4, Stressgen Bioreagents), CD45RO(1:200 dilution, clone UCHL1, Gene Tech), CD8 (1: 50 dilution, clone c8/144B, Daka) was performed as follows. Slides were dewaxed by immersion in xylene twice, each for 5 min, and industrial methylated spirits twice, each for 3 min, and were washed in phosphate buffered saline (PBS) three times for 3 min, Antigen retrieval in microwave with buffer solution PH 6,Cttrate buffer,20min, washed three times in PBS for 3 min. Endogenous peroxidase was blocked by incubating sections in 3 per cent (v/v) hydrogen peroxide in methanol for 20 min. Sections were washed three times in PBS for 3 min. Sections were incubated with 20% goat blood serum at room temperature, do not wash, and incubated with primary antibody at 37℃,2hours, washed three times in PBS for 3 min, then slides were incubated with secondary antibody(HRP-R/M, EnVision, Dako) at 37℃,30min.washed three times in PBS for 3 min. All antibodies were visualized using 3,3-diaminobenzidine as chromagen (DAB). Negative controls omitted the primary antibody. The location of tumor-infiltrating cells was classified as the invasive margin of cancer, center of tumor and cancer cell nests. Slides were analyzed using an image analysis software(Qwin, Leica DM2000,x200,0.18mm2), five random fields were chosen, the average density of positive cells was recorded. The cut-off values for CD8, CD45RO cell density were 50,10 cells/0.18mm2 in the center of the tumor; 50,25 cells/0.18mm2 in the invasive margin, and 2,2 cells/0.18mm2 in the tumor nests, respectively. The cut-off values for Hsp70 was 25% positive tumor cells/total tumor cells.High density CD45RO+ infiltrating cells in the center of tumor were detected in specimens from 63 patients. Of them,40 patients in Group A achieved 47.50%(19/40) objective clinical response, significantly higher than that of 23 patients who achieved 21.74% (5/23) objective clinical response in Group B (P=0.044). On the contrary, among 38 patients with low density of CD45RO+ infiltrating cells in the center of tumor,20 in Group A achieved 50.00%(10/20) objective response,18 patients in Group B achieved 44.44%(8/18) objective response.High density of CD45RO+ infiltrating cells in the tumor nests were detected in specimens from 61 patients. Of them,38 patients in Group A achieved 52.63%(20/38) objective clinical response, significantly higher than that of 23 patients who achieved 21.74%(5/23) objective clinical response in Group B (P=0.031). On the contrary, among 40 patients with low density of CD45RO+ infiltrating cells in the tumor nests,22 in Group A achieved 40.91%(9/20) objective response,18 patients in Group B achieved 44.44% (8/18) objective response.Low density of CD45RO+ cells infiltrating the invasive margin of tumor were detected in specimens from 53 patients. Of them,31 patients in Group A achieved 54.8%(17/31)objective clinical response, significantly higher than that of 22 patients who achieved 27.3%(6/22) objective clinical response in Group B (P=0.047). On the contrary, among 48 patients with high density of CD45RO+ infiltrating cells the invasive margin of tumor (IM),29 in Group A achieved 41.4%(12/29) objective response,19 patients in Group B achieved 36.8%(7/19) objective response.High density of Hsp70+ tumor cells were detected in specimens from 62 patients. Of them,37 patients in Group A achieved 62.16%(23/37) objective clinical response, significantly higher than that of 25 patients who achieved 36%(9/25) objective clinical response in Group B. On the contrary, among 39 patients with low Hsp70 expression,23 in Group A achieved 26.09%(6/23) objective response,16 patients in Group B achieved 25% (2/25) objective response.Comparison for different density of CD8+ cells infiltrating the tumor nests, center of tumor, invasive margin, and its clinical objective response rate, did not show statistical significance. Less expression of IL-17 (24/101) and IL-10 (8/101) were detected in all patients. The relationship of these marker expression and clinical response was still under analysis.The results suggested that high density of Hsp70+ tumor cells, high density CD45RO+ infiltrating cells in the tumor nests and center of tumor should be predictive factor for patients response to DC vaccine, The predictive value of CD8 and MSI-H should be further confirmed.