Molecular Mechanisms of Ovarian Carcinoma Metastasis: Key Genes and Regulatory MicroRNAs

Getting to know ovarian cancer: Focusing on the effect of LncRNAs in this cancer and the effective signaling pathways

This article undertakes a comprehensive investigation of ovarian cancer, examining the complex nature of this challenging disease. The main focus is on understanding the role of long non-coding RNAs (lncRNAs) in the context of ovarian cancer (OC), and their regulatory functions in disease progression. Through extensive research, the article identifies specific lncRNAs that play significant roles in the intricate molecular processes of OC. Furthermore, the study examines the signaling pathways involved in the development of OC, providing a detailed comprehension of the underlying molecular mechanisms. By connecting lncRNA dynamics with signaling pathways, this exploration not only advances our understanding of ovarian cancer but also reveals potential targets for therapeutic interventions. The findings open up opportunities for targeted treatments, highlighting the importance of personalized approaches in addressing this complex disease and driving progress in ovarian cancer research and treatment strategies.

MicroRNA-31 induced by Fusobacterium nucleatum infection promotes colorectal cancer tumorigenesis

Persistent Fusobacterium nucleatum infection is associated with the development of human colorectal cancer (CRC) and promotes tumorigenicity, but the underlying mechanisms remain unclear. Here, we reported that F. nucleatum promoted the tumorigenicity of CRC, which was associated with F. nucleatum-induced microRNA-31 (miR-31) expression in CRC tissues and cells. F. nucleatum infection inhibited autophagic flux by miR-31 through inhibiting syntaxin-12 (STX12) and was associated with the increased intracellular survival of F. nucleatum. Overexpression of miR-31 in CRC cells promoted their tumorigenicity by targeting eukaryotic initiation factor 4F-binding protein 1/2 (eIF4EBP1/2), whereas miR-31 knockout mice were resistant to the formation of colorectal tumors. In conclusion, F. nucleatum, miR-31, and STX12 form a closed loop in the autophagy pathway, and continuous F. nucleatum-induced miR-31 expression promotes the tumorigenicity of CRC cells by targeting eIF4EBP1/2. These findings reveal miR-31 as a potential diagnostic biomarker and therapeutic target in CRC patients with F. nucleatum infection.

Long Non-Coding RNAs and microRNAs Groups in the Regulation of Expression Level of a Number of Tumor-Associated Genes in Ovarian Cancer

The search for interacting long non-coding RNAs (lncRNAs), microRNAs (miRNAs), and mRNAs of protein-coding genes through the mechanism of competing endogenous RNAs in tumors of ovarian cancer patients was carried out. The levels of expression of 24 lncRNAs, 20 miRNAs, and 28 mRNAs of protein-coding genes involved in oncogenesis were determined by real-time PCR on a set of representative samples. Correlations between lncRNAs/miRNA and miRNA/mRNA levels in ovarian cancer samples were analyzed. We identified 8 pairs of lncRNAs/miRNA and 17 pairs of miRNA/mRNA, the expression levels of which have a negative correlation. Five triplets of potentially interacting lncRNAs/miRNA/mRNA have been identified, among which the most significant triplet is the OIP5-AS1/miR-203a-3p/ZEB1. The data obtained determine new epigenetic profiles, as well as new potential biomarkers and targets for targeted therapy of ovarian cancer patients.

Integration of Transcriptome and Epigenome to Identify and Develop Prognostic Markers for Ovarian Cancer

DNA methylation is a widely researched epigenetic modification. It is associated with the occurrence and development of cancer and has helped evaluate patients' prognoses. However, most existing DNA methylation prognosis models have not simultaneously considered the changes of the downstream transcriptome. Methods. The RNA-Sequencing data and DNA methylation omics data of ovarian cancer patients were downloaded from The Cancer Genome Atlas (TCGA) database. The Consensus Cluster Plus algorithm was used to construct the methylated molecular subtypes of the ovary. Lasso regression was employed to build a multi-gene signature. An independent data set was applied to verify the prognostic value of the signature. The Gene Set Variation Analysis (GSVA) was used to carry out the enrichment analysis of the pathways linked to the gene signature. The IMvigor 210 cohort was used to explore the predictive efficacy of the gene signature for immunotherapy response. Results. We distinguished ovarian cancer samples into two subtypes with different prognosis, based on the omics data of DNA methylation. Differentially expressed genes and enrichment analysis among subtypes indicated that DNA methylation was related to fatty acid metabolism and the extracellular matrix (ECM)-receptor. Furthermore, we constructed an 8-gene signature, which proved to be efficient and stable in predicting prognostics in ovarian cancer patients with different data sets and distinctive pathological characteristics. Finally, the 8-gene signature could predict patients' responses to immunotherapy. The polymerase chain reaction experiment was further used to verify the expression of 8 genes. Conclusion. We analyzed the prognostic value of the related genes of methylation in ovarian cancer. The 8-gene signature predicted the prognosis and immunotherapy response of ovarian cancer patients well and is expected to be valuable in clinical application.

MicroRNA-31: a pivotal oncogenic factor in oral squamous cell carcinoma

Oral squamous cell carcinoma (OSCC) continuously constitutes a major challenge for treatment and prognosis due to approximately half of treated OSCC patients dying from locoregional recurrences and distant metastases. MicroRNA-31 (miR-31), an early mammalian miRNA identified, has been gaining importance in the field of OSCC research in recent years. This comprehensive review was conducted for the first time to summarize the current evidence on the association between miR-31 and OSCC. The vast majority of relevant studies (20/21, 95%) demonstrated that miR-31 was an oncogenic factor in the tumorigenesis and progression of OSCC. miR-31 expression is significantly upregulated in plasma, saliva, and tumor tissue of OSCC. miR-31 played an essential role in OSCC development by constituting a complex network with its targeted genes (e.g. RhoA, FIH, ACOX1, VEGF, SIRT3, LATS2, KANK1, and NUMB) and the signaling cascades (e.g. EGF-AKT signaling axis, ERK-MMP9 cascade, Hippo pathway, Wnt signaling, and MCT1/MCT4 regulatory cascade). This review highlights that miR-31 might function as a potential diagnostic, prognostic, and predictive biomarker for OSCC. Further studies are still warranted to better illuminate the clinicopathological features and the molecular mechanisms of miR-31-mediated OSCC development.

Aberrant Methylation of 20 miRNA Genes Specifically Involved in Various Steps of Ovarian Carcinoma Spread: From Primary Tumors to Peritoneal Macroscopic Metastases

Our work aimed to differentiate 20 aberrantly methylated miRNA genes that participate at different stages of development and metastasis of ovarian carcinoma (OvCa) using methylation-specific qPCR in a representative set of clinical samples: 102 primary tumors without and with metastases (to lymph nodes, peritoneum, or distant organs) and 30 peritoneal macroscopic metastases (PMM). Thirteen miRNA genes (MIR107, MIR124-2, MIR124-3, MIR125B-1, MIR127, MIR129-2, MIR130B, MIR132, MIR193A, MIR339, MIR34B/C, MIR9-1, and MIR9-3) were hypermethylated already at the early stages of OvCa, while hypermethylation of MIR1258, MIR137, MIR203A, and MIR375 was pronounced in metastatic tumors, and MIR148A showed high methylation levels specifically in PMM. We confirmed the significant relationship between methylation and expression levels for 11 out of 12 miRNAs analyzed by qRT-PCR. Moreover, expression levels of six miRNAs were significantly decreased in metastatic tumors in comparison with nonmetastatic ones, and downregulation of miR-203a-3p was the most significant. We revealed an inverse relationship between expression levels of miR-203a-3p and those of ZEB1 and ZEB2 genes, which are EMT drivers. We also identified three miRNA genes (MIR148A, MIR9-1, and MIR193A) that likely regulate EMT–MET reversion in the colonization of PMM. According to the Kaplan–Meier analysis, hypermethylation of several examined miRNA genes was associated with poorer overall survival of OvCa patients, and high methylation levels of MIR130B and MIR9-1 were related to the greatest relative risk of death.

CircLNPEP promotes the progression of ovarian cancer through regulating miR-876-3p/ WNT5A axis

CircRNA LNPEP has been shown to promote the development of hepatocellular carcinoma, but its function in ovarian cancer (OC) remains unclear. The Kaplan–Meier method was used to analyze the clinical significance of circLNPEP expression in OC patients. The stability of circLNPEP was detected by actinomycin D and RNase R treatment. The correlations between miR-876-3p and two genes (circLNPEP and WNT5A) were predicted by bioinformatics analysis and confirmed by dual-luciferase reporter assay. Expressions of circLNPEP, miR-876-3p, and WNT5A were determined by qRT-PCR and western blot. The effect of circLNPEP/miR-876-3p/WNT5A axis on viability, proliferation, migration, and invasion, and angiogenesis of cells was determined by cell function experiment and rescue experiment. Xenograft tumor mice were constructed for in vivo verification. Expressions of apoptosis, epithelial mesenchymal transition (EMT)-related genes, and CD34 were determined by qRT-PCR, western blot and immunohistochemistry. High level of circLNPEP was related to poor prognosis in OC. CircLNPEP was highly expressed in OC tissues and cell lines, mainly distributed in the cytoplasm, while miR-876-3p was the opposite. MiR-876-3p targeted and negatively correlated with circLNPEP and WNT5A. Sh-circLNPEP repressed cell viability, proliferation, migration, invasion, angiogenesis, and EMT but promoted apoptosis, which were related to its regulation of expression of EMT- and apoptosis-related genes, WNT5A, and CD34. The regulatory effect of sh-circLNPEP can be reversed by miR-876-3p inhibitor, and that of miR-876-3p inhibitor can be reversed by sh-WNT5A. CircLNPEP promoted cancer cell proliferation, EMT and angiogenesis, and inhibited apoptosis by regulating miR-876-3p/WNT5A axis.

Impact of Selected Signaling Proteins on SNAIL 1 and SNAIL 2 Expression in Ovarian Cancer Cell Lines in Relation to Cells' Cisplatin Resistance and EMT Markers Level

It has been increasingly recognized that SNAIL1 and SNAIL2, as major EMT-inducers, might also be involved in drug resistance of cancer cells. We sought to determine a relation between SNAIL1/2, E-cadherin and N-cadherin expression, as well as ovarian cancer cells’ resistance to cisplatin and EMT markers’ level. Thus, four ovarian cancer cell lines, were used: A2780, A2780cis, SK-OV-3 and OVCAR-3. We assessed the impact of ERK1/2, AKT and STAT3 proteins (chosen by the profiling activity of over 40 signaling proteins) on SNAIL1/2 expression, along with E-cadherin and N-cadherin levels. We showed that expression of SNAIL1 and N-cadherin are the highest in cisplatin-resistant A2780cis and SK-OV-3 cells, while high SNAIL2 and E-cadherin levels were observed in cisplatin-sensitive A2780 cells. The highest E-cadherin level was noticed in OVCAR-3 cells. SNAIL1/2 expression was dependent on ERK1/2 activity in cisplatin-resistant and potentially invasive SK-OV-3 and OVCAR-3 cells. STAT-3 regulates expression of SNAIL1/2 and leads to the so-called “cadherin switch” in cancer cells, independently of their chemoresistance. In conclusion, SNAIL1, but not SNAIL2, seems to be involved in ovarian cancer cells’ cisplatin resistance. STAT3 is a universal factor determining the expression of SNAIL1/2 in ovarian cancer cells regardless of their chemoresitance or invasive capabilities.

LncRNAs in Ovarian Cancer Progression, Metastasis, and Main Pathways: ceRNA and Alternative Mechanisms

Ovarian cancer (OvCa) develops asymptomatically until it reaches the advanced stages with metastasis, chemoresistance, and poor prognosis. Our review focuses on the analysis of regulatory long non-coding RNAs (lncRNAs) competing with protein-coding mRNAs for binding to miRNAs according to the model of competitive endogenous RNA (ceRNA) in OvCa. Analysis of publications showed that most lncRNAs acting as ceRNAs participate in OvCa progression: migration, invasion, epithelial-mesenchymal transition (EMT), and metastasis. More than 30 lncRNAs turned out to be predictors of survival and/or response to therapy in patients with OvCa. For a number of oncogenic (CCAT1, HOTAIR, NEAT1, and TUG1 among others) and some suppressive lncRNAs, several lncRNA/miRNA/mRNA axes were identified, which revealed various functions for each of them. Our review also considers examples of alternative mechanisms of actions for lncRNAs besides being ceRNAs, including binding directly to mRNA or protein, and some of them (DANCR, GAS5, MALAT1, and UCA1 among others) act by both mechanisms depending on the target protein. A systematic analysis based on the data from literature and Panther or KEGG (Kyoto Encyclopedia of Genes and Genomes) databases showed that a significant part of lncRNAs affects the key pathways involved in OvCa metastasis, EMT, and chemoresistance.

MiRNAs: A Powerful Tool in Deciphering Gynecological Malignancies

Accumulated evidence on the clinical roles of microRNAs (miRNAs) in cancer prevention and control has revealed the emergence of new genetic techniques that have improved the understanding of the mechanisms essential for pathology induction and progression. Comprehension of the modifications and individual differences of miRNAs and their interactions in the pathogenesis of gynecological malignancies, together with an understanding of the phenotypic variations have considerably improved the management of the diagnosis and personalized treatment for different forms of cancer. In recent years, miRNAs have emerged as signaling molecules in biological pathways involved in different categories of cancer and it has been demonstrated that these molecules could regulate cancer-relevant processes, our focus being on malignancies of the gynecologic tract. The aim of this paper is to summarize novel research findings in the literature regarding the parts that miRNAs play in cancer-relevant processes, specifically regarding gynecological malignancy, while emphasizing their pivotal role in the disruption of cancer-related signaling pathways.

The Role of microRNAs in Epithelial Ovarian Cancer Metastasis

Epithelial ovarian cancer (EOC) is the deadliest gynecological cancer, and the major cause of death is mainly attributed to metastasis. MicroRNAs (miRNAs) are a group of small non-coding RNAs that exert important regulatory functions in many biological processes through their effects on regulating gene expression. In most cases, miRNAs interact with the 3′ UTRs of target mRNAs to induce their degradation and suppress their translation. Aberrant expression of miRNAs has been detected in EOC tumors and/or the biological fluids of EOC patients. Such dysregulation occurs as the result of alterations in DNA copy numbers, epigenetic regulation, and miRNA biogenesis. Many studies have demonstrated that miRNAs can promote or suppress events related to EOC metastasis, such as cell migration, invasion, epithelial-to-mesenchymal transition, and interaction with the tumor microenvironment. In this review, we provide a brief overview of miRNA biogenesis and highlight some key events and regulations related to EOC metastasis. We summarize current knowledge on how miRNAs are dysregulated, focusing on those that have been reported to regulate metastasis. Furthermore, we discuss the role of miRNAs in promoting and inhibiting EOC metastasis. Finally, we point out some limitations of current findings and suggest future research directions in the field.

[Clinical significance of methylation of a group of miRNA genes in patients with ovarian cancer.]

It was found that the proportion of microRNA genes inactivated by methylation of regulatory CpG islands is several times higher than the genes encoding proteins, which increases their attractiveness as promising markers of cancer. The aim of this work is to evaluate the clinical significance of methylation of 13 tumor-associated microRNA genes (MIR-124a-2, MIR-124a-3, MIR-125-B1, MIR-127, MIR-129-2, MIR-132, MIR-137, MIR-203a, MIR-34b/c, MIR-375, MIR-9-1, MIR-9-3, MIR-339) in 26 patients with ovarian cancer. Methylation level was evaluated by the method of methylation-specific PCR in real time. The data obtained in primary tumors (26), histologically unchanged ovarian tissues (15) and peritoneal metastases (19) were compared using a number of statistical programs. For all 13 genes, an increase in the level of methylation was revealed during the transition from unchanged tissue to primary tumors and further from primary tumors to peritoneal metastases; moreover, in the genes MIR-203a, MIR-375 and MIR-339, the level of methylation in metastases increased most significantly (in 2 and more times). A correlation was observed for the first time, showing a consistency between the increase in methylation level in some miRNA pairs, for example, MIR-129-2/MIR-132 (r_s> 0,7; p<0,0001), both in primary tumors and in metastases. An analysis of microRNA gene methylation in clinical samples of ovarian cancer showed a correlation between the observed molecular changes both with the initial stages of tumor formation and with the progression and dissemination of ovarian cancer, with the presence of metastases in a large omentum and with the appearance of ascites. The revealed dependencies deepen the understanding of the mechanism of peritoneal metastasis and can be used to select new diagnostic and prognostic markers of ovarian cancer.

Discovery of potential serum and urine-based microRNA as minimally-invasive biomarkers for breast and gynecological cancer

BACKGROUND

Deregulated microRNAs (miRNAs) in breast and gynecological cancer might contribute to improve early detection of female malignancies.

OBJECTIVE

Specification of miRNA types in serum and urine as minimally-invasive biomarkers for breast (BC), endometrial (EC) and ovarian cancer (OC).

METHODS

In a discovery phase, serum and urine samples from 17 BC, five EC and five OC patients vs. ten healthy controls (CTRL) were analyzed with Agilent human miRNA microarray chip. Selected miRNA types were further investigated by RT-qPCR in serum (31 BC, 13 EC, 15 OC patients, 32 CTRL) and urine (25 BC, 10 EC, 10 OC patients, 30 CTRL) applying two-sample t-tests.

RESULTS

Several miRNA biomarker candidates exhibited diagnostic features due to distinctive expression levels (serum: 26; urine: 22). Among these, miR-518b, -4719 and -6757-3p were found specifically deregulated in BC serum. Four, non-entity-specific, novel biomarker candidates with unknown functional roles were identified in urine (miR-3973; -4426; -5089-5p and -6841). RT-qPCR identified miR-484/-23a (all p⩽ 0.001) in serum as potential diagnostic markers for EC and OC while miR-23a may also serve as an endogenous control in BC diagnosis.

CONCLUSIONS

Promising miRNAs as liquid biopsy-based tools in the detection of BC, EC and OC qualified for external validation in larger cohorts.

Circulating non-coding RNAs in recurrent and metastatic ovarian cancer

Ovarian cancer has a poor outcome because it is usually detected at advanced tumor stages, and the majority of the patients develop disease relapse as a result of chemotherapy resistance. This most lethal gynecological malignancy metastasizes within the peritoneal fluid or ascites to pelvic and distal organs. In ovarian cancer progression and metastasis, small non-coding RNAs (ncRNAs), including long noncoding RNAs and microRNAs have been recognized as important regulators. Their dysregulation modulates gene expression and cellular signal pathways and can be detected in liquid biopsies. In this review, we provide an overview on circulating plasma and serum ncRNAs participating in tumor cell migration and invasion, and contributing to recurrence and metastasis of ovarian cancer. We will also discuss the development of potential, novel therapies using ncRNAs as target molecules or tumor markers for ovarian cancer.

Role of miRNA-Regulated Cancer Stem Cells in the Pathogenesis of Human Malignancies

Recent biomedical discoveries have revolutionized the concept and understanding of carcinogenesis, a complex and multistep phenomenon which involves accretion of genetic, epigenetic, biochemical, and histological changes, with special reference to MicroRNAs (miRNAs) and cancer stem cells (CSCs). miRNAs are small noncoding molecules known to regulate expression of more than 60% of the human genes, and their aberrant expression has been associated with the pathogenesis of human cancers and the regulation of stemness features of CSCs. CSCs are the small population of cells present in human malignancies well-known for cancer resistance, relapse, tumorigenesis, and poor clinical outcome which compels the development of novel and effective therapeutic protocols for better clinical outcome. Interestingly, the role of miRNAs in maintaining and regulating the functioning of CSCs through targeting various oncogenic signaling pathways, such as Notch, wingless (WNT)/β-Catenin, janus kinases/ signal transducer and activator of transcription (JAK/STAT), phosphatidylinositol 3-kinase/ protein kinase B (PI3/AKT), and nuclear factor kappa-light-chain-enhancer of activated B (NF-kB), is critical and poses a huge challenge to cancer treatment. Based on recent findings, here, we have documented the regulatory action or the underlying mechanisms of how miRNAs affect the signaling pathways attributed to stemness features of CSCs, such as self-renewal, differentiation, epithelial to mesenchymal transition (EMT), metastasis, resistance and recurrence etc., associated with the pathogenesis of various types of human malignancies including colorectal cancer, lung cancer, breast cancer, head and neck cancer, prostate cancer, liver cancer, etc. We also shed light on the fact that the targeted attenuation of deregulated functioning of miRNA related to stemness in human carcinogenesis could be a viable approach for cancer treatment.

MicroRNA in Ovarian Cancer: Biology, Pathogenesis, and Therapeutic Opportunities

Ovarian cancer comprises one of the three major malignant tumor types in the female reproductive system. The mortality rate of this cancer is the highest among all gynecological tumors, with ovarian cancer metastasis constituting an important cause of death. Therefore, markers for disease prediction and prognosis are highly desirable for early diagnosis as well as for helping optimize and personalize treatment. Recently, microRNAs (miRNAs), which consist of short-sequence RNAs that do not encode a protein, have emerged as new biomarkers in the clinical diagnosis and treatment of ovarian cancer. By pairing with bases specific to the target messenger RNA (mRNA), miRNAs cause degradation of the target mRNA or inhibit its translation, thereby regulating various cellular processes including cell proliferation and adhesion. Increasing numbers of studies have shown that miRNA expression abnormality plays an important role in the development of ovarian cancer. In this review, we discuss the mechanisms of miRNA action, current research regarding their role in the suppression or promotion of ovarian cancer, and their use as markers for diagnosis of prognosis or as therapeutic targets for this disease. Finally, we present future perspectives regarding the clinical management of ovarian cancer and the role for miRNAs therein.

HLA-G 3' untranslated region variants +3187G/G, +3196G/G and +3035T define diametrical clinical status and disease outcome in epithelial ovarian cancer

Expression of the non-classical human leukocyte antigen-G (HLA-G) promotes cancer progression in various malignancies including epithelial ovarian cancer (EOC). As single nucleotide polymorphisms (SNPs) in the HLA-G 3' untranslated region (UTR) regulate HLA-G expression, we investigated HLA-G 3'UTR haplotypes arranged by SNPs in healthy controls (n = 75) and primary EOC patients (n = 79) and determined soluble HLA-G (sHLA-G) levels. Results were related to the clinical status and outcome. Although haplotype frequencies were similar in patients and controls, (i) sHLA-G levels were increased in EOC independent of the haplotype, (ii) homozygosity for UTR-1 or UTR-2 genotypes were significantly associated with metastases formation and presence of circulating tumor cells before therapy, whereas (iii) the UTR-5 and UTR-7 haplotypes were significantly associated with a beneficial clinical outcome regarding negative nodal status, early FIGO staging, and improved overall survival. Lastly, (iv) the ambivalent impact on clinical EOC aspects could be deduced to specific SNPs in the HLA-G 3'UTR: +3187G, +3196G and +3035T alleles. Our results give evidence that even if the genetic background of the HLA-G 3'UTR is identical between patients and controls, certain SNPs have the potential to contribute to diametrical clinical status/outcome in EOC.

Upregulation of microRNA-3129 suppresses epithelial ovarian cancer through CD44

The purpose of this work is to evaluate whether human microRNA-3129 (hsa-miR-3129) may functionally regulate cancer development, possibly through downstream target CD44 in human epithelial ovarian cancer (EOC). Direct targeting of hsa-miR-3129 on human CD44 transcript was evaluated using a dual-luciferase reporter assay. Gene expression of hsa-miR-3129 in immortal EOC cell lines was evaluated by qRT-PCR. Lentivirus-mediated hsa-miR-3129 upregulation or downregulation was conducted in SK-OV-3 and CAOV-3 cells, in which endogenous hsa-miR-3129 and CD44 expressions were then measured. In hsa-miR-3129 upregulated or downregulated EOC cells, functional assays were applied to evaluate EOC proliferation, bufalin chemoresistance in vitro, or xenotransplantation in vivo. Moreover, CD44 was ectopically overexposed in hsa-miR-3129 upregulated EOC cells to functionally evaluate the correlation between hsa-miR-3129 and CD44 in EOC. Dual-luciferase reporter assay confirmed hsa-miR-3129 directly binds CD44. QRT-PCR revealed that hsa-miR-3129 was substantially downregulated in EOC cell lines. In SK-OV-3 and CAOV-3 cells, lentivirus-induced hsa-miR-3129 upregulation downregulated CD44 whereas hsa-miR-3129 downregulation did not affect CD44 expression. Hsa-miR-3129 upregulation had significant anti-cancer effects by inhibiting EOC proliferation, increasing bufalin chemoresistance, and suppressing xenotransplantation. On the other hand, overexpressing CD44 reversed the anti-cancer functions by hsa-miR-3129 upregulation in EOC cells. In conclusion, Has-miR-3129 is a functional regulator, possibly through reverse targeting on CD44, in EOC.

Radiomics of high-grade serous ovarian cancer: association between quantitative CT features, residual tumour and disease progression within 12 months

OBJECTIVES

To determine if radiomic features, alone or combined with clinical data, are associated with residual tumour (RT) at surgery, and predict the risk of disease progression within 12 months (PD12) in ovarian cancer (OC) patients.

METHODS

This retrospective study enrolled 101 patients according to the following inclusion parameters: cytoreductive surgery performed at our institution (9 May 2007-23 February 2016), assessment of BRCA mutational status, preoperative CT available. Radiomic features of the ovarian masses were extracted from 3D structures drawn on CT images. A phantom experiment was performed to assess the reproducibility of radiomic features. The final radiomic features included in the analysis (n = 516) were grouped into clusters using a hierarchical clustering procedure. The association of each cluster's representative radiomic feature with RT and PD12 was assessed by chi-square test. Multivariate analysis was performed using logistic regression models. P values < 0.05 were considered significant.

RESULTS

Patients with values of F2-Shape/Compactness1 below the median, of F1- GrayLevelCooccurenceMatrix25/0-1InformationMeasureCorr2 below the median and of F1-GrayLevelCooccurenceMatrix25/-333-1InverseVariance above the median showed higher risk of RT (36%, 36% and 35%, respectively, as opposed to 18%, 18% and 18%). Patients with values of F4-GrayLevelRunLengthMatrix25/-333RunPercentage above the median, of F2 shape/Max3DDiameter below the median and F1-GrayLevelCooccurenceMatrix25/45-1InverseVariance above the median showed higher risk of PD12 (22%, 24% and 23%, respectively, as opposed to 6%, 5% and 6%). At multivariate analysis F2-Shape/Max3DDiameter remained significant (odds ratio (95% CI) = 11.86 (1.41-99.88)). To predict PD12, a clinical radiomics model performed better than a base clinical model.

CONCLUSION

This study demonstrated significant associations between radiomic features and prognostic factors such as RT and PD12.

KEY POINTS

• No residual tumour (RT) at surgery is the most important prognostic factor in OC. • Radiomic features related to mass size, randomness and homogeneity were associated with RT. • Progression of disease within 12 months (PD12) indicates worse prognosis in OC. • A model including clinical and radiomic features performed better than only-clinical model to predict PD12.