A combination of molecular and clinical parameters provides a new strategy for high-grade serous ovarian cancer patient management

High-grade serous carcinoma (HGSC) remains the most common and deadly subtype of ovarian cancer, due to its diagnosis at advanced stages in over 80% of cases [1]. The standard of care of advanced stage HGSC is a combination of cytoreductive surgery and platinum-based chemotherapy (i.e. carboplatin and paclitaxel). Cytoreductive surgery can be performed at the time of cancer diagnosis or can be delayed after 3 or 4 cycles of neoadjuvant chemotherapy [2]. Recently, the use of a targeted management approach (i.e. anti-angiogenic agents and PARP inhibitors) has been introduced for specific subgroups of patients such as those with BRCA mutations [3, 4]. Although the results on BRCA mutation status can be available within 2–3 weeks from diagnosis in some referral oncological centres, in most clinical settings worldwide it is not available until first-line treatment has already commenced. Although HGSC is considered to be a platinum-sensitive disease, ~ 20–30% of patients fail to respond or experience disease recurrence within 6 months of completing chemotherapy [5]. In clinical practice, these patients are considered to be resistant to platinum, and second-line treatment is usually based on non-platinum agents (e.g. gemcitabine, pegylated liposomal doxorubicin, topotecan or weekly paclitaxel) and control of symptoms [2, 6]. Efforts to establish biological stratification profiles in HGSC are mainly being focused on genomic and transcriptomic markers [7‐10], while the evaluation of protein markers to assess treatment response and/or survival in HGSC tissue samples remain scarce [11‐13]. Beyond their clinical relevance as biological endpoints and drug targets, proteins are markers widely used in clinical tests for disease diagnosis and prognosis [13, 14]. In addition, the use of proteomic approaches (i.e. mass spectrometry) enable the systematic interrogation of proteomes from complex clinical samples which can later be translated into immunoassays for clinical use [15, 16]. Despite the numerous studies being carried out to identify novel protein biomarkers in ovarian cancer, there are currently no validated markers used in the clinic to predict response to first-line treatment and guide the management of patients with newly diagnosed HGSC. In this work, we present a new strategy that combines molecular and routine clinical parameters to anticipate the response to first-line treatment in HGSC patients (HGSC-1LTR). Prediction of those patients less likely to respond to first-line chemotherapy allows alternative treatment strategies to be considered at the time of cancer diagnosis (i.e. non-platinum drugs) thus improving patients’ response and survival. It also optimizes therapeutic decision making and allows an individualized management strategy by avoiding the use of futile treatments, thus improving patients’ quality of life. This strategy is therefore of special relevance for patients with advanced HGSC for whom a predictive method is lacking. Thus, in this work we aim to identify a protein biomarker signature able to predict response to first-line treatment in patients with newly diagnosed HGSC.

In this work we have established a new strategy that combines molecular and clinical parameters to predict the response to first-line treatment in HGSC patients (HGSC-1LTR) (Fig. 2E). There are currently no protein biomarkers available at the time of HGSC diagnosis able to predict patients individual response to first-line chemotherapy with carboplatin and paclitaxel. In addition, studies evaluating predictive protein markers in ovarian cancer tissues using proteomic approaches are scarce as highlighted in a recent review by our group [13]. The identification of these predictors of chemotherapy response, as the one presented here, allows the prioritization of platinum-based agents if the disease is sensitive, and the use of alternative treatments if resistant in order to improve patient management. This ability to predict a patient's response is particularly relevant in chemoresistant HGSC because alternative non-platinum based chemotherapy regimens could be considered or clinically evaluated in this subgroup of patients [33]. Additionally, it offers several clinical advantages and aid informed clinical decisions. Clinicians could address patients' individual needs and improve their quality of life by avoiding the use of futile treatments. It could also alleviate the economic burden of the healthcare system associated with the use of ineffective treatments. Although understanding why platinum-resistance occurs is essential for improving survival, new strategies able to discriminate between chemoresistant and chemosensitive patients at the time of cancer diagnosis are paramount for better HGSC patient management. Our findings are therefore an important step in advancing towards a stratified risk-management of HGSC patients through the identification of biological predictors of treatment response.

One of the main challenges in this type of projects is the selection of samples to be used for the discovery of protein candidates relevant to our study. In our study we chose tissues for two main reasons. Firstly, archival FFPE tumour samples represent a valuable resource for studying cancer biomarkers because they are widely available in hospitals and are associated to important clinical information (e.g. histology, response to treatment and outcomes). Secondly, all patients included in the study, and for whom the predictive biomarker is intended to be used (i.e. patients with newly diagnosed advanced stage HGSC), undergo a biopsy prior to starting chemotherapy treatment. Therefore, patients’ tissues are available without the need of additional invasive procedures. Moreover, tumour tissues are where potential protein biomarkers are more likely to be enriched. Although FFPE tissues have traditionally been associated to a high variation in protein quality due to formalin-induced chemical modifications and differences in storage times [34], recent studies have confirmed that proteomes are preserved to a comparable extent to those obtained from fresh frozen tissues and are not influenced by their storage [35, 36]. In the study by Coscia et al. [11] authors evaluated 25 advanced stage HGSC FFPE samples by discovery proteomics and identified CT45 as a platinum sensitivity mediator in ovarian cancer. We were able to observe an increased protein abundance by targeted proteomics in chemosensitive samples, in line with the results observed by Coscia et al. However, this protein was excluded from subsequent predictive analysis because its targeted peptides were not consistently detected.

Clinical parameters routinely recorded by healthcare professionals can be relevant to improve the classificatory ability of protein biomarkers. An example of this is the Risk of Ovarian Malignancy Algorithm (ROMA) which integrates patients menopausal status to serum CA125 and HE4 levels, to distinguishing between a benign and malignant pelvic mass [37]. We assessed the classificatory ability of our protein signature in combination with relevant clinical data such as the patients age, menopausal status, serum CA125 levels and the decision to treat with either primary cytoreductive surgery or neoadjuvant chemotherapy. Indeed, the addition of the patients data that is currently considered in clinical practice, improved the classification power of the protein biomarker combination. Although mutations in BRCA and other homologous recombination genes are known to predict response to chemotherapy, current clinical guidelines recommend referral for genetic testing at the time of ovarian cancer diagnosis. Hence, since the genetic test is not widely available in all clinical settings until chemotherapy has commenced, a predictor of first-line treatment response, at present, cannot include the genetic information. Although we have not integrated the BRCA mutational status in our HGSC-1LTR strategy, it would be a basic parameter to add in future validation phases together with the use of targeted therapies in the maintenance setting given the rapid advances in the field. In addition, our study was limited by the availability of HGSC samples and the quality of the FFPE tissues for protein extraction and MS analysis. This limitation is associated to differences in patients’ age between the chemoresistant and chemosenstivie groups and TFIp between the discovery and the verification cohorts. Given that we included patients undergoing either primary cytoreductive surgery or interval surgery, it would also be interesting to evaluate the HGSC-1LTR strategy in future studies comparing the molecular characteristics of tissues before and after neoadjuvant chemotherapy. Finally, for the widespread implementation of our signature in routine clinical practice, we would require validation in a higher number of patients and the availability of mass spectrometry technology in the clinics.

We have established a new strategy (HGSC-1LTR) that combines tissue levels of proteins TKT, LAMC1 and FUCO, together with patients’ age, menopausal status, serum CA125 levels, and treatment approach to predict the response to first-line treatment in HGSC patients. These data are obtained from ready-available biopsies in hospitals, and ready-available clinical data at the time of HGSC diagnosis, thus no additional interventions are required beyond current clinical practice. Because predictive tools are currently lacking for patients with advanced HGSC, this new strategy is clinically relevant for the prediction of chemoresistant patients. Identification of chemoresistance at the time of diagnosis can facilitate the study of alternative treatments aimed at improving the outcome for these patients. In addition, those patients classified as chemosensitive could undergo standard care with platinum-based agents. Therefore, the HGSC-1LTR strategy can allow optimization of therapeutic decision making and individualize HGSC patients’ care.