Abstract Background: The ability of cancer cells to metastasise influences the mortality rate of patients with cancer. Extracellular matrix (ECM) from stromal organization is associated with tumorigenesis and metastasis in breast cancer. It is proposed that morphological features of collagen, based on second harmonic generation (SHG) microscopy, within the tumour environment can be a quantitative image-based biomarker for to predict the survival rate of triple-negative breast cancer (TNBC) patients. To this, we have developed quantitative fibrosis indices by combining multiple collagen features in breast tumour to reveal the diverse survivals of TNBC patients. Method: The patients (n = 216) in this study were diagnosed with TNBC in Singapore from 2003 to 2015. Disease-free survival (DFS) and overall survival (OS) were defined as time from the point of diagnosis to recurrence or to death/the date of the last follow-up, respectively. The constructed tissue microarrays (TMA) of breast tumours were scanned by the SHG microscope (Genesis, HistoIndex Pte. Ltd., Singapore). 33 collagen parameters were quantified from each sample. These collagen parameters were used to build disease free survival (DFS) index, overall survival (OS) index for prediction of early recurrence (DFS< 1 year) and early death (OS< 4 years), respectively. Kaplan-Meier survival analysis was further performed to assess long-term survival of TNBC patients with high and low risk as stratified by the indices, tumour grade and tumour size. The indices were validated using leave-one-out method. Results: Both DFS-index and OS-index were created using 10 collagen parameters chosen by sequential selection methods. Due to insufficient follow-up time, 12 patients and 81 patients were excluded from the early recurrence analysis and early death analysis, respectively. The DFS-index could differentiate low-risk patients with DFS≥1 year (n=179) and high-risk patients with DFS< 1 year(n=25) (training p< 0.001; validation p=0.157) with a cut-off value DFS-index=0.880. The OS-index could differentiate the low-risk patients with OS≥4 years (n=101) and high-risk patients with OS< 4 years (n=34) (training p< 0.001; validation p=0.011) with a cut-off value OS-index=0.703. The log-rank test showed DFS-index (training p = 0.001; validation p=0.025) and OS-index (training p < 0.001; validation p=0.011) could be used for the prediction of disease-free survival and overall survival. Kaplan-Meier survival analysis revealed tumour size>20mm (DFS, p=0.605; OS, p=0.136) and tumour grade=3 (DFS, p=0.328; OS, p=0.768) had poor predictive value in this study. Conclusion: Quantitative assessment of fibrosis in breast cancer correlates with long-term survival of TNBC patients. This study used DFS-index and OS-index combined complex morphological collagen features and obtained better prediction results than tumour size and tumour grade. Table. The difference between low and high risk groups differentiated by developed indices, tumour size and tumour grade. Low-risk group has a longer DFS and OS months based on DFS-index and OS-index. Citation Format: Yayun Ren, Dean Tai, Ying Zhao, Puay Hoon Tan. Improved quantitative fibrosis indices reveal diverse survivals of triple negative breast cancer patients [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P6-04-11.