The Study Of Relationship Of Lymph Node Ratio And Prognosis Of Breast Cancer And Preliminary Study Of Small RNA Characterization Of Invasive Micropapillary Carcinomas Of The Breast

Part1Objective:To investigate if lymph node ratio (LNR) is better than positive lymph nodes (PLN) to evaluate recurrence hazard and overall survival time of breast carcinoma (BC) patients and the feasibility of LNR and HER-2expression status in stratifying the risk category of BC.Methods:We retrospectively analyzed1095medical records from Tianjin Medical University Cancer Institute and Hospital between January2002and December2005. All the patients were node-positive primary invasive breast caner and followed up at least five years. Univariate and multivariate analysis were performed to evaluate and compare the impact of PLN and LNR on replase free survival (RFS) and overall survival (OS). Log-rank test were used to compare survival differences between LNR groups, and then identified the best cut-off value. Comparing survival differences between the risk categories defined by PLN combined with HER-2expression status and those defined by the LNR in combination with HER-2expression status.Results:1. In univariate analyses, pT-stage, grade, ER/PR/HER-2status, PLN, LNR, total lymph nodes dissected, extranodal soft tissue invasion, adjuvant chemotherapy and endocrine therapy were associated significantly with replase free survival (RFS) and overall survival (OS)(P<0.05).2. In multivariate analyses, when PLN or LNR entered into the Cox hazard ratio model as covariate respectively, both PLN and LNR were the dependent prognostic factors of RFS (P<0.001) and OS (P<0.001); when PLN and LNR entered into the Cox hazard ratio model as covariate at the same time, we found that LNR still was the dependent prognostic factor of RFS (P<0.001) and OS (P=0.002) but PLN lost significance (RFS:P=0.792, OS:P=0.124).3. Log-rank test demonstrated significantly different RFS and OS among three LNR groups (<0.10,0.11-0.30and>0.30). The most significant difference was observed between LNR<0.30and LNR>0.30.4. Kaplan-Meier analysis on patients in the intermediate and high-risk PLN groups showed that the RFS were82.0%and57.1%(χ2=66.374, P<0.001), respectively and the OS were92.3%and74.2%(χ2=50.139, P<0.001), respectively. For the intermediate and high-risk LNR groups, the RFS rates were80.6%and53.5%(χ2=85.117, P<0.001), respectively and the OS rates were91.0%and71.8%(χ2=61.281, P<0.001), respectively.5. The high-risk PLN group was confirmed to be associated with increased hazard ratio (HR) of2.88(95%CI,2.20-3.76) for RFS and3.78(95%CI,2.54-5.62) for OS, in comparing with the intermediate risk group. The high-risk LNR group demonstrated the same trend for RFS (increased HR of2.93;95%CI,2.31-3.73) and OS (increased HR of3.57;95%CI,2.54-5.02).Concclusions:This study identified LNR as an independent prognostic predictor of node-positive BC patients and compared to the number of involved lymph nodes, LNR can predict recurrence hazard and OS more efficiently. The risk categories defined by LNR combined with HER-2status were compatible to those defined by the PLN in combination with HER-2status. Because of its better equilibrium and more accurate reflection of axillary lymph nodes status, we propose that LNR, in conjunction with HER2expression, be taken into consideration in defining risk categories of node-positive breast cancer patients to provide more powerful reference for clinicians to make adjuvant therapy decisions. Part2Objective:Invasive micropapillary carcinoma (IMPC) is a special type of invasive breast cancer. Due to its special growth pattern of tumor cell clusters arranged in solid micropapillary or tubular configuration, contain no the central fibrovascular cores and its high propensity for lymphovascular invasion and axillary lymph node metastasis, more and more clinicians and pathologists attach great importance to IMPC. Recently, emerging evidence suggested that some micro-RNA(miRNA) can function as oncogenes or tumor suppressors in tumor development and progression and this provided new ideas for elucidating the pathogenesis and treatment of tumor. The current study was designed to investigate possible role of miRNA in IMPC high lymphatic vessel invasion and high lymph node metastasis.Methods:Six IMPC (five formalin-fixed paraffin-embedded, FFPE, one fresh-frozen) and six IDC-NSTs (five FFPE, one fresh-frozen) specimens were selected and then analyzed by using next-generation sequencing. To validate the miRNA expression pattern profiled by deep sequencing in more patients, miRNA-specific realtime RT-qPCR was performed on total RNA isolated from IMPC (n=22) and IDC-NSTs (n=24) FFPE samples.Results:1. A total of197,504,842reads were obtained from the sequencing machine. After removing adaptors, filtering out low quality tags and cleaning up the contaminations,175,300,661(88.76%) clean reads were obtained. The length distribution analysis revealed that the RNA sequences were mainly within a range of21-23nt the size of most known small RNAs.2. A total of45miRNAs were identified to be differentially expressed between IMPC and IDC-NSTs. In an unsupervised cluster analysis, these differently expressed miRNAs generated a tree with clear distinction between IMPC and IDC-NSTs classes. Furthermore paired FFPE and fresh-frozen tissues in two patients showed a good correlation of miRNA expression, with Pearson’s correlation coefficients of0.9909(p<0.001) and0.9904(p<0.001), respectively.3. By using realtime RT-qPCR we further investigated miRNA expression and found let-7b, miR-30c, miR-148a, miR-181a, miR-181a*and miR-181b were significantly differently expressed between the two groups, which is consistent with the results of deep sequencing.4. We also elucidated several features of miRNA in these breast cancer tissues, including5’variability, miRNA editing, and3’untemplated addition. Drosha defines more precise5’ends than Dicer in both IMPC and IDC-NSTs tissues. More miRNA editing was observed in IDC-NSTs than IMPC (p<0.05). C-to-T and A-to-G alterations account for36.4%～44.0%and8.8%～11.4%of all miRNA editing, respectively. In terms of the spatial location of editing within the mature miRNAs, nucleotides5,7,8and13-16showed a relatively low frequency of editing.3’additions were presented in7.4%of all miRNAs and the majority of these additions were adenine addition (58.9%) and uridine addition (33.6%).Conclusions:1. Deep sequencing revealed that there exist differentially expressed miRNA between IMPC and IDC-NSTs, which could separate IMPC versus IDC-NSTs by unsupervised cluster analysis.2. Compared with IDC-NSTs, the low-expression of let-7b, miR-30c, miR-148a, miR-181a, miR-181a*and miR-181b may play an important role in high lymphatic vessel invasion and high lymph node metastasis of IMPC.3. Both IMPC and IDC-NSTs presented with5’variability, miRNA editing, and3’ untemplated addition, but they are not specific key factors that affect high propensity of IMPC for lymphovascular invasion and axillary lymph node metastasis.4. Small RNA characterization revealed by deep sequencing will help us to better understand the molecular mechanism underlying the invasive property of IMPC.