Risk Stratification Of Cancer Patients Research Articles

1863P Circulating biomarkers for risk stratification of cancer patients with chemotherapy associated febrile neutropenia

1863P Circulating biomarkers for risk stratification of cancer patients with chemotherapy associated febrile neutropenia

Three-dimensional telomere profiling predicts risk of progression in smoldering multiple myeloma.

Smoldering multiple myeloma (SMM) is a precursor stage that precedes multiple myeloma (MM). SMM is heterogenous with nearly 40% of patients progressing to MM in the first 5 years. The high rate of progression of SMM patients highlights the need for early intervention, which underscores the importance of identifying SMM patients with the highest risk of progression. Several risk stratification models showed utility in identifying high-risk SMM patients; however, these systems showed limited sensitivity. To date, identifying high-risk SMM patients remains an important clinical need. In this study, we present the 3-dimensional telomere profiling as a structural biomarker capable of stratifying SMM patients as a function of genomic instability. Quantifying telomere dysfunction using the TeloView technology showed utility in risk stratification of cancer patients, particularly hematological malignancies. In this study, we analyzed 168 SMM patients. We report an AUC in ROC analysis of 0.8 using a subset of the patients as a training dataset. We then conducted a blind validation on a different cohort and demonstrated a positive predictive value of 85% and negative predictive value of 73%, with sensitivity and specificity of 83% and 76%, respectively. We examined the correlation between the TeloView prediction and the 20-2-20 scoring system, and cytogenetic abnormalities. We report a correlation of 53% with the 20-2-20 scores and over 60% correlation with cytogenetic abnormalities. The result of this study presents the telomere profiling as an effective biomarker able to stratify SMM patients to their respective risk groups with high sensitivity and specificity.

Evaluation of metabolic parameters of microsatellites stable and instable colorectal cancer patients via PET/CT

Aims: Microsatellite instability has been determined as an important indicator in selecting chemotherapy drugs in colorectal cancer. Within the scope of this research, we aimed to elucidate the pathology reports and determine whether the metabolic parameters detected by PET/CT differ in MSI-positive and negative patients. Methods: A total of 35 patients were analyzed retrospectively. The patient population consisted of patients who applied to the Nuclear Medicine Department with a diagnosis of colon or rectum cancer, underwent PET/CT imaging for staging purposes, and were operated on. Results: A total of 35 colon or rectum cancer patients were included in this retrospective analysis. When microsatellite instability was analyzed among the patients, it was found that female patients comprised 4 microsatellite instability-positive and 16 microsatellite instability-negative individuals. On the other hand, 5 of the males were microsatellite instability positive, and 10 were microsatellite instability negative. The mean SUVmax value was 16.4±8.2, SUVmean was 8.1±1.9, TLG was 392.4±520.8, and MTV was 26.5±25.4 in the microsatellite instability-positive individuals. On the other hand, the mean SUVmax value was 22.7±9.7, SUVmean was 5.2±2.2, TLG was 316.4±325.7, and MTV was 21.7±21.7 in the microsatellite instability-negative individuals. Conclusion: With the advancement of image analysis technology, MTV, and TLG, volumetric indexes derived from 18F-FDG PET have been proposed for risk stratification of cancer patients. Regarding the outcomes of this research, the semiquantitative and metabolic parameters obtained by PET/CT are not different in colorectal cancer cases with instable and stable microsatellites.

Abstract LB104: Tumor fraction as a predictive factor of outcome in patients referred for oncology early phase clinical trials: Analyses of two precision medicine studies

Abstract Modern phase I trials have shown considerable advancement, demonstrating encouraging response rates and subsequently raising the demand for participant enrollment. Despite these advances, there remains a significant challenge: approximately 15% of patients enrolled in these trials pass away within the first 90 days, and nearly half of them do not experience any therapeutic benefit. These statistics underscore the critical need for more effective strategies in selecting suitable patients for trial participation. One promising avenue is the use of circulating tumor DNA (ctDNA) measurements to estimate tumor fraction (TF), a method that has shown potential in prognosticating outcomes for patients with solid tumors. Nevertheless, the effectiveness of TF as an indicator for predicting the clinical benefits of systemic treatments in patients referred for early phase studies is yet to be conclusively determined Methods: We analyzed adult patients with advanced solid tumors included from December 2020 to December 2021 in two precision medicine studies: BIP (NCT02534649) and STING (NCT04932525). Comprehensive genomic profiling was conducted using the FDA approved 324-gene FoundationOne Liquid Cdx assay. TF was categorized as ≥10% or <10%. Multivariate analysis incorporated established prognostic scores (Royal Marsden Hospital: RMH, Garrido-Laguna et al. 2012; and GRIm, Bigot et al., 2017) and their individual components (RMH: albumin, lactate dehydrogenase, number of metastatic sites; GRIm: neutrophil/lymphocyte ratio, lactate dehydrogenase, albumin). Results: A total of 1305 patients referred for early phase trials were evaluated: 871 in BIP and 434 in STING.The treatment regimens patients received post-TF assessment included: cytotoxic agents (BIP: 278 patients, 28.8%; STING: 106 patients, 24.4%), immunotherapy (BIP: 152 patients, 15.8%; STING: 76 patients, 17.5%), tyrosine kinase inhibitors (BIP: 126 patients, 13.1%; STING: 35 patients, 8.1%), hormonal therapy (BIP: 131 patients, 13.6%; STING: 9 patients, 2.1%), and various other therapies or combinations. High TF was significantly correlated with poorer objective response/stable disease rate, shorter progression-free survival (PFS), and lower overall survival (OS) in both studies. Multivariate analysis confirmed TF as an independent predictor of objective response, PFS, and OS, beyond the RMH and GRIm scores and their components. Conclusions: This is the first prospective analysis confirming that TF is a significant predictive marker of treatment efficacy in patients referred for early phase studies. Our findings suggest that TF could be a valuable tool in decision-making for phase I trial participation. Additionally, this research lays the groundwork for future exploration of TF in risk stratification of cancer patients in various clinical contexts. Citation Format: Laila Belcaid, Sophie Cousin, Arnaud Bayle, Melissa Alame, Etienne Rouleau, Laura Blouin, Ludovic Lacroix, Damien Vasseur, Maxime Brunet, Sofiane Taleb, Isabelle Soubeyran, Veronique Debien, Eurydice Angeli, Christophe Massard, Antoine Italiano. Tumor fraction as a predictive factor of outcome in patients referred for oncology early phase clinical trials: Analyses of two precision medicine studies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB104.

The 3D- Telomere Profiling Assay Identifies High Risk Smoldering Multiple Myeloma Patients with High Precision

Smoldering multiple myeloma (SMM) is an asymptomatic precursor stage of multiple myeloma (MM). According to the international Myeloma Working Group data (Blood Cancer J. 2020 Oct 16;10(10):102) the probability of progression to MM is 22% at 2 years, and 42% and 64% at 5 and 10 years, respectively. Thus, identification of patients with high risk of early progression to MM is crucial for optimal treatment management. Genomic instability is a sensitive indicator of disease progression in cancer. Telomere dysfunction is an early event in genomic instability. The 3-dimensional (3D) spatial profiling of telomeres using TeloView technology allows for quantification of telomere dysfunction and was shown to be instrumental in risk stratification of cancer patients generally, but particularly in selected hematological malignancies. Importantly, the utility of the 3D-telomere profiling was recently demonstrated in 2 clinical studies validating the utility of 3D telomere profiling as a structural biomarker to identify high risk SMM patients. The two studies achieved accuracy of >80% and specificity and sensitivity of over 80% & 76% respectively in a training dataset study followed by a blind validation (Kumar S.et al 2023. J Clin Oncol). To present the 3D telomere profiling assay as a reliable prognostic tool in the clinic, capable to precisely identify high risk SMM patients, it is important to demonstrate the repeatability and accuracy of the assay. In this study, we conducted a retrospective analytical validation including samples of 20 SMM patients with known progression outcome. The samples of each patient were blindly processed in triplets using the 3D-telomere assay, followed by the TeloView analysis. To eliminate any validation-bias the validation was conducted with uncontrolled variables. Samples were processed blindly by different operators on different dates using different instrumentation. We interrogated the repeatability and precision of the results of each patient analyzed on 2 levels. First, we calculated coefficient of variation (CV) of telomere predictors of each patient across the three runs; and secondly, we interrogated the repeatability of patient outcome using the scoring model developed and validated in (Kumar et al. 2023, J Clin Oncol) across the three runs of each patient. We set the acceptable concordance to 80% for CVs (CV <20%) and the acceptable repeatability of patient-outcome as predicted by the scoring model to 80% of the patients included in the study. We report CVs across all the telomere predictors, across the 3 runs of all patients ranging between 2.3 - 17.8. Furthermore, patient outcome as high/low risk of progression was consistent in 17 out of the 20 patients in all 3 runs (85%). The results of this study demonstrate the 3D-telomere profiling as an accurate and prognostic structural biomarker. The assay shows high level of precision, repeatability, and reliability, and presents a viable solution to accurately identify high risk SMM patients. This will potentially allow the treating physicians to make confident treatment decisions for high risk SMM patients based on the results of the 3D Telomere profiling assay.

Three-dimensional telomere profiling to predict risk of progression in smoldering multiple myeloma.

8056 Background: Multiple myeloma (MM) is preceded by monoclonal gammopathy of undetermined significance (MGUS). A transitional stage of smoldering multiple myeloma (SMM) can be identified between MGUS and MM. While MGUS carries a steady risk of progression of 1% per year, SMM is more heterogenous with nearly 40% of patients progressing in the first 5 years, 15% in the next 5 years, reaching the same low risk as MGUS after 10 years. SMM with its high risk of progression in the initial years after diagnosis presents a viable opportunity for early intervention. For implementing early intervention, the ability to identify SMM patients at the highest risk of progression is critical. This has led to the development of several risk stratification systems. Using these systems high risk SMM patients studied in phase 3 trials demonstrated delayed progression to MM and improved overall survival with early initiation of therapy. However, these approaches showed limited specificity exposing patients at lower risk of progression to therapy. To date, identifying high risk SMM patients and confirming disease stability in low risk SMM patients remain an important clinical need. Genomic instability has been shown to be a sensitive indicator of disease progression in cancer. Telomere dysfunction is an early event in genomic instability. The 3-dimensional spatial profiling of telomeres using TeloView technology allows for quantification of telomere dysfunction, and was shown to be instrumental in risk stratification of cancer patients generally, but particularly in selected hematological malignancies. Importantly, in a previous SMM proof-of-concept study telomeric parameters measured by TeloView technology was found to be significantly different between SMM patients who progressed to active MM within 2 years and those who remained stable for over 5 years. Methods: We analyzed a total of 162 SMM patients using TeloView technology. 88 patients were employed as training dataset in Receiver Operating Curve (ROC) modeling to develop a scoring model that stratifies individual SMM patients based on risk of progression to full stage MM. An additional cohort of 74 SMM patients was used for blind validation of the developed scoring model. Results: We report area-under-the-curve (AUC) in the ROC analysis of 0.8 (accuracy 80%) achieved by the scoring model developed using the training dataset. Furthermore, the independent blind validation achieved positive predictive value of 83% and negative predictive value of 71%, with sensitivity and specificity of 80% and 76% respectively. Conclusions: The result of this study supports presenting TeloView as an accurate prognostic biomarker which appears able to stratify SMM patients into their respective risk groups with high sensitivity and specificity. This will potentially allow for evidence-based treatment decisions for high risk SMM patients and confident monitoring of stable patients.

Relationship Between Frailty and Hospital Outcomes Among Cancer Patients in The United States

<h3>Research Objectives</h3> To understand the effects of frailty on hospital outcomes such as in-hospital mortality, length of stay, and healthcare cost among cancer patients using a nationally representative database. <h3>Design</h3> This was a retrospective observational analysis. <h3>Setting</h3> Nationwide Inpatient Sample (NIS) data collected during 2005-2014. <h3>Participants</h3> Cancer patients ≥45 years of age. <h3>Interventions</h3> Not applicable. <h3>Main Outcome Measures</h3> In-hospital mortality, hospital length of stay, and hospitalization cost. Propensity score match analysis was done to examine whether frailty was associated with length of stay and in-hospital mortality. <h3>Results</h3> There were 10,463,083 cancer hospitalizations during 2005-2014, of which 1,022,777 (9.8%) were frail. Patients having length of stay ≥8 days were significantly higher among frail group, compared to non-frail group (53.3% versus 25.3%, P < 0.001). Similarly, unadjusted mortality (12.0% versus 5.3%, P < 0.001) and hospitalization costs ($29,726 versus $18,595, P< 0.001) were significantly higher for frail patients. Nearly $28 billion was expended on hospitalization of frail cancer patients during the study period. In propensity score match analysis, the odds of in-hospital mortality (OR, 1.54; 95% CI, 1.50-1.58) and length of stay (OR, 2.23; 95% CI, 2.18-2.27) were significantly greater for frail patients. <h3>Conclusions</h3> Frailty was associated with adverse hospital outcome such as increased length of stay, mortality, and hospitalization cost among all cancer types. Our findings could be valuable for frailty-based risk stratification of cancer patients. Concerted efforts by the physiatrists, oncologists, and surgeons towards identifying frailty and incorporating it in risk estimation measures could help in optimizing management strategies for cancers. <h3>Author(s) Disclosures</h3> The authors report no real or perceived vested interests that relate to this article that could be construed as a conflict of interest.

Cancer Cure and Consequences on Survivorship Care: Position Paper from the Italian Alliance Against Cancer (ACC) Survivorship Care Working Group

A multidisciplinary panel of experts and cancer patients developed a position paper to highlight recent evidence on "cancer cure" (ie, the possibility of achieving the same life expectancy as the general population) and discuss the consequences of this concept on follow-up and rehabilitation strategies. The aim is to inform clinicians, patients, and health-care policy makers about strategies of survivorship care for cured cancer patients and consequences impacting patient lives, spurring public health authorities and research organizations to implement resources to the purpose. Two identifiable, measurable, and reproducible indicators of cancer cure are presented. Cure fraction (CF) is >60% for breast and prostate cancer patients, >50% for colorectal cancer patients, and >70% for patients with melanoma, Hodgkin lymphoma, and cancers of corpus uteri, testis (>90%), and thyroid. CF was >65% for patients diagnosed at ages 15-44 years and 30% for those aged 65-74 years. Time-to-cure was consistently <1 year for thyroid and testicular cancer patients and <10 years for patients with colorectal and cervical cancers, melanoma, and Hodgkin lymphoma. The working group agrees that the evidence allows risk stratification of cancer patients and implementation of personalized care models for timely diagnosis, as well as treatment of possible cancer relapses or related long-term complications, and preventive measures aimed at maintaining health status of cured patients. These aspects should be integrated to produce an appropriate follow-up program and survivorship care plan(s), avoiding stigma and supporting return to work, to a reproductive life, and full rehabilitation. The "right to be forgotten" law, adopted to date only in a few European countries, may contribute to these efforts for cured patients.

Improving on-treatment risk stratification of cancer patients with refined response classification and integration of circulating tumor DNA kinetics

BackgroundSignificant intertumoral heterogeneity exists as antitumor treatment is introduced. Heterogeneous therapeutic responses are conventionally evaluated by imaging examinations based on Response Evaluation Criteria in Solid Tumors (RECIST); nevertheless, there are increasing recognitions that they do not fully capture patient clinical benefits. Currently, there is a paucity of data regarding the clinical implication of biological responses assessed by liquid biopsy of on-treatment circulating tumor DNA (ctDNA). Here, we investigated whether biological response evaluated by ctDNA kinetics added critical information to the RECIST, and whether integrating on-treatment biological response information refined risk stratification of cancer patients.MethodsIn this population-based cohort study, we included 821 patients with Epstein-Barr virus (EBV)-associated nasopharynx of head and neck cancer (NPC) receiving sequential neoadjuvant chemotherapy (NAC) and chemoradiotherapy (CRT), who had pretreatment and on-treatment cfEBV DNA and magnetic resonance imaging (MRI) surveillance. Biological responses evaluated by cfEBV DNA were profiled and compared with conventional MRI-based RECIST evaluation. The inverse probability weighting (IPW)-adjusted survival analysis was performed for major survival endpoints. The Cox proportional hazard regression [CpH]-based model was developed to predict the on-treatment ctDNA-based individualized survival.ResultsOf 821 patients, 71.4% achieved complete biological response (cBR) upon NAC completion. RECIST-based response evaluations had 25.3% discordance with ctDNA-based evaluations. IPW-adjusted survival analysis revealed that cfEBV DNApost-NAC was a preferential prognosticator for all endpoints, especially for distant metastasis. In contrast, radiological response was more preferentially associated with locoregional recurrence. Intriguingly, cfEBV DNApost-NAC further stratified RECIST-responsive and non-responsive patients; RECIST-based non-responsive patients with cBR still derived substantial clinical benefits. Moreover, detectable cfEBV DNApost-NAC had 83.6% prediction sensitivity for detectable post-treatment ctDNA, which conferred early determination of treatment benefits. Finally, we established individualized risk prediction models and demonstrated that introducing on-treatment ctDNA significantly refined risk stratification.ConclusionsOur study helps advance the implementation of ctDNA-based testing in therapeutic response evaluation for a refined risk stratification. The dynamic and refined risk profiling would tailor future liquid biopsy-based risk-adapted personalized therapy.

Comparing single-target and multitarget approaches for postoperative circulating tumour DNA detection in stage II-III colorectal cancer patients.

Circulating tumour DNA (ctDNA) detection for postoperative risk stratification in cancer patients has great clinical potential. However, low ctDNA abundances complicates detection. Multitarget (MT) detection strategies have been developed to increase sensitivity. Yet, empirical evidence supporting performance gains of MT vs. single‐target (ST) strategies in a postoperative setting is limited. We compared ctDNA detection in 379 paired plasma samples from 112 stage II–III colorectal cancer patients by ST digital PCR and MT sequencing of 16 patient‐specific variants. The strategies exhibited good concordance (90%, Cohen's Kappa 0.79), with highly correlated ctDNA quantifications (Pearson r = 0.985). A difference was observed in ctDNA detection preoperatively (ST 72/92, MT 88/92). However, no difference was observed immediately after surgery in recurrence (ST 11/22, MT 10/22) or nonrecurrence (both 2/34) patients. In serial samples, detection was similar within recurrence (ST 13/16, MT 14/16) and nonrecurrence (ST 3/49, MT 1/49) patients. Both approaches yielded similar lead times to standard‐of‐care radiology (ST 4.0 months, MT 4.1 months). Our findings do not support significant performance gains of the MT strategy over the ST strategy for postoperative ctDNA detection.

Risk Stratification in Cancer Patients with Acute Upper Gastrointestinal Bleeding: Comparison of Glasgow-Blatchford, Rockall and AIMS65, and Development of a New Scoring System.

Background/AimsFew studies have measured the accuracy of prognostic scores for upper gastrointestinal bleeding (UGIB) among cancer patients. Thereby, we compared the prognostic scores for predicting major outcomes in cancer patients with UGIB. Secondarily, we developed a new model to detect patients who might require hemostatic care. MethodsA prospective research was performed in a tertiary hospital by enrolling cancer patients admitted with UGIB. Clinical and endoscopic findings were obtained through a prospective database. Multiple logistic regression analysis was performed to gauge the power of each score. ResultsFrom April 2015 to May 2016, 243 patients met the inclusion criteria. The AIMS65 (area under the curve [AUC] 0.85) best predicted intensive care unit admission, while the Glasgow-Blatchford score best predicted blood transfusion (AUC 0.82) and the low-risk group (AUC 0.92). All scores failed to predict hemostatic therapy and rebleeding. The new score was superior (AUC 0.74) in predicting hemostatic therapy. The AIMS65 (AUC 0.84) best predicted in-hospital mortality. ConclusionsThe scoring systems for prognostication were validated in the group of cancer patients with UGIB. A new score was developed to predict hemostatic therapy. Following this result, future prospective research should be performed to validate the new score.

Longitudinal Analysis of Circulating Tumor Cells in Colorectal Cancer Patients by a Cytological and Molecular Approach: Feasibility and Clinical Application.

IntroductionLiquid biopsies allowing for individualized risk stratification of cancer patients have become of high significance in individualized cancer diagnostics and treatment. The detection of circulating tumor cells (CTC) has proven to be highly relevant in risk prediction, e.g., in colorectal cancer (CRC) patients. In this study, we investigate the clinical relevance of longitudinal CTC detection over a course of follow-up after surgical resection of the tumor and correlate these findings with clinico-pathological characteristics.MethodsIn total, 49 patients with histologically proven colorectal carcinoma were recruited for this prospective study. Blood samples were analyzed for CTC presence by two methods: first by marker-dependent immunofluorescence staining combined with automated microscopy with the NYONE® cell imager and additionally, indirectly, by semi-quantitative Cytokeratin-20 (CK20) RT-qPCR. CTC quantification data were compared and correlated with the clinico-pathological parameters.ResultsDetection of CTC over a post-operative time course was feasible with both applied methods. In patients who were pre-operatively negative for CTCs with the NYONE® method or below the cut-off for relative CK20 mRNA expression after analysis by PCR, a statistically significant rise in the immediate post-operative CTC detection could be demonstrated. Further, in the cohort analyzed by PCR, we detected a lower CTC load in patients who were adjuvantly treated with chemotherapy compared to patients in the follow-up subgroup. This finding was contrary to the same patient subset analyzed with the NYONE® for CTC detection.ConclusionOur study investigates the occurrence of CTC in CRC patients after surgical resection of the primary tumor and during postoperative follow-up. The resection of the tumor has an impact on the CTC quantity and the longitudinal CTC analysis supports the significance of CTC as a prognostic biomarker. Future investigations with an even more extended follow-up period and larger patient cohorts will have to validate our results and may help to define an optimal longitudinal sampling scheme for liquid biopsies in the post-operative monitoring of cancer patients to enable tailored therapy concepts for precision medicine.

Abstract PO-090: Application of deep learning methods in predicting survival and identifying high-risk populations from large optum real world breast cancer data

Abstract Real-World Evidence (RWE) contains abundant observational data collected during routine clinical practice. With such large-scale and high-dimensional data, we can leverage the power of neural networks to increase the capacity of classical survival models in predicting the risk of breast cancer patients, and identifying potential prognosis-associated factors retrospectively. The target cohorts were extracted based on diagnosis code, biomarker, and mastectomy procedure code, we further identified metastatic breast cancer cases according to secondary diagnosis code and physician notes. Our features consist of demographic and vital data, cancer-related measurement, and medication history, which was divided into several lines of therapy. Considering the dynamic nature of cancer therapy, time-dependent survival models are more suitable to model the hazard given the current features. Therefore, we extended the proportional-hazard-based DeepSurv and logistic-hazard-based Nnet-survival to accommodate time-dependent features, and further incorporated recurrent neural networks to make better use of sequential medication information. We trained the model on the training set (70%), tuned the hyperparameters via Bayesian Optimization on the validation set (10%), and evaluated the model with c-index and integrated brier score on the test set (20%). Compared with the classical linear cox model and pooled logistic regression, the proposed model outperformed these baselines by a significant margin, indicating the promise of using deep learning and the time-dependent prognosis model for survival prediction and risk stratification in cancer patients. In addition, the output from the neural network was interpreted with SHAP values and Integrated Gradient (IG), which shed some light on the effect of adjuvant hormone therapy, especially AI and SERM, on recurrence and overall survival. Further, our model can be improved with more collected data and a better approach to handling error and missingness in the data. Citation Format: Han Wang, Xiaodong Luo, Meiqi Wu, LiMing Shen. Application of deep learning methods in predicting survival and identifying high-risk populations from large optum real world breast cancer data [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-090.

Prognostic and therapeutic significance of circulating tumor cells in patients with lung cancer.

Lung cancer is the second most common cancer and the main cause of cancer-related mortality worldwide. In spite of various efforts that have been made to facilitate the early diagnosis of lung cancer, most patients are diagnosed when the disease is already in stage IV, which is generally associated with the occurrence of distant metastases and a poor survival. Moreover, a large proportion of these patients will relapse after treatment, heralding the need for the stratification of lung cancer patients in addition to identifying those who are at a higher risk of relapse and, thus, require alternative and/or additional therapies. Recently, circulating tumor cells (CTCs) have been considered as valuable markers for the early diagnosis, prognosis and risk stratification of cancer patients, and they have been found to be able to predict the survival of patients with various types of cancer, including lung cancer. Additionally, the characterization of CTCs has recently provided fascinating insights into the heterogeneity of tumors, which may be instrumental for the development of novel targeted therapies. Here we review our current understanding of the significance of CTCs in lung cancer metastasis. We also discuss prominent studies reporting the utility of enumeration and characterization of CTCs in lung cancer patients as prognostic and pharmacodynamic biomarkers for those who are at a higher risk of metastasis and drug resistance.

Total metabolic tumor volume by 18F-FDG PET/CT for the prediction of outcome in patients with non-small cell lung cancer.

Metabolic tumor volume (MTV) and total lesion glycolysis (TLG) are imaging parameters derived from 18F-FDG PET/CT that have been proposed for risk stratification of cancer patients. The aim of our study was to test whether these whole-body volumetric imaging parameters may predict outcome in patients with non-small cell lung cancer (NSCLC). Sixty-five patients (45 men, 20 women; mean age ± SD, 65 ± 12years), with histologically proven NSCLC who had undergone 18F-FDG PET/CT scan before any therapy, were included in the study. Imaging parameters including SUVmax, SUVmean, total MTV (MTVTOT) and whole-body TLG (TLGWB) were determined. Univariate and multivariate analyses of clinical and imaging variables were performed using Cox proportional hazards regression. Survival analysis was performed using Kaplan-Meier method and log-rank tests. A total of 298 lesions were analyzed including 65 primary tumors, 114 metastatic lymph nodes and 119 distant metastases. MTVTOT and TLGWB could be determined in 276 lesions. Mean value of MTVTOT was 81.83ml ± 14.63ml (SE) whereas mean value of TLGWB was 459.88g ± 77.02g (SE). Univariate analysis showed that, among the variables tested, primary tumor diameter (p = 0.0470), MTV of primary tumor (p = 0.0299), stage (p < 0.0001), treatment (p < 0.0001), MTVTOT (p = 0.0003) and TLGWB (p = 0.0002) predicted progression-free survival in NSCLC patients, while age (p = 0.0550), MTV of primary tumor (p = 0.0375), stage (p < 0.0001), treatment (p < 0.0001), MTVTOT (p = 0.0001) and TLGWB (p = 0.0008) predicted overall survival. At multivariate analysis age, TLGWB and stage were retained in the model for prediction of progression-free survival (p < 0.0001), while age, MTVTOT and stage were retained in the model for prediction of overall survival (p < 0.0001). Survival analysis showed that patients with TLGWB ≤ 54.7g had a significantly prolonged progression-free survival as compared to patients with TLGWB > 54.7g (p < 0.0001). Moreover, overall survival was significantly better in patients showing a MTVTOT ≤ 9.5ml as compared to those having MTVTOT > 9.5ml (p < 0.0001). Similar results were obtained in a subgroup of 43 patients with advanced disease (stages III and IV). Whole-body PET-based volumetric imaging parameters are able to predict outcome in NSCLC patients.

Live-cell phenotypic-biomarker microfluidic assay for the risk stratification of cancer patients via machine learning.

The risk stratification of prostate cancer and breast cancer tumours from patients relies on histopathology, selective genomic testing, or on other methods employing fixed formalin tissue samples. However, static biomarker measurements from bulk fixed-tissue samples provide limited accuracy and actionability. Here, we report the development of a live-primary-cell phenotypic-biomarker assay with single-cell resolution, and its validation with prostate cancer and breast cancer tissue samples for the prediction of post-surgical adverse pathology. The assay includes a collagen-I/fibronectin extracellular-matrix formulation, dynamic live-cell biomarkers, a microfluidic device, machine-vision analysis and machine-learning algorithms, and generates predictive scores of adverse pathology at the time of surgery. Predictive scores for the risk stratification of 59 prostate cancer patients and 47 breast cancer patients, with values for area under the curve in receiver-operating-characteristic curves surpassing 80%, support the validation of the assay and its potential clinical applicability for the risk stratification of cancer patients.

Validation of the EPIPHANY index for predicting risk of serious complications in cancer patients with incidental pulmonary embolism.

The EPIPHANY index was developed to classify cancer associated pulmonary embolism (PE) into different risk categories using decision tree modeling. In this study, we tried to externally validate this index in a distinct group of patients solely composed of incidental PE (IPE). A retrospective study of patients diagnosed with IPE in two Emergency Departments in the USA and South Korea from 2013 to 2014 was performed. The primary outcome was the occurrence of a serious medical complication within 15days of presentation to ED. Thirty-day complication was the secondary outcome. Cumulative hazard curves for each prognostic category were drawn to show the change in hazards over time. A total of 258 patients with IPE were included (193 from MD Anderson Cancer Center and 65 from Asan Medical Center). Serious complication within 15days occurred in 23 (8.9%) patients. The risk of overall 15-day and 30-day serious complications increased with each category (low, intermediate, and high risk: 3.4, 8.9, and 23.8%, P = 0.033; 6.9, 9.5, and 33.3%, P = 0.011). Cumulative hazard curves for each prognostic category were drawn and the survival functions factored by prognostic categories were significantly different over 15days (P = 0.015) and 30days (P = 0.001). Our study suggests the EPIPHANY index could be a useful adjunct tool in risk stratification of cancer patients with IPE.

Management of febrile neutropenia in malignancy using the MASCC score and other factors: Feasibility and safety in routine clinical practice.

The current standards for empirical broad-spectrum intravenous antibiotic (AB) treatment, combined with hospitalization, are cautious and safe, but lead to over-treatment of a substantial group of patients. We need to validate parameters to identify these low-risk febrile-neutropenia (FN) patients, who could then be safely treated in an outpatient setting with minimal/no AB treatment. A retrospective analysis for validation of a risk-assessment model in FN patients was done on a patient population from January 2007 to December 2008. Inclusion criteria were a histological diagnosis of malignancy, FN secondary to chemotherapy, absolute-neutrophil-count of ≤ 500/μl, axillary temperature of ≥ 38°C, and age ≥ 14 years. Other clinical and laboratory parameters were explored for risk stratification during the FN episodes. Receiver-operating characteristic curves were used to find the threshold value, an Chi-square analysis was done to find the association between the outcome and the parameters. A total of 178 FN episodes were documented; 22 in solid tumors and 156 in hematolymphoid malignancies. Culture positivity was documented in 59 episodes; peripheral blood was the most common source, with Escherichia coli being the most common organism identified. Risk stratification was done using the Multinational Association of Supportive Care in Cancer (MASCC) risk-index score. The association between the MASCC score and risk stratification could not be established (P = not significant) at a score of ≤ 21; however, it was found to be significant at a score of ≤ 18. The total number of complications was 23 (sepsis 22, mortality 23). Other factors found to be significantly associated with a high risk of complications in the univariate analysis were, mucositis (P = 0.03), maximum temperature ≥ 103°F (P = 0.01), tachycardia (P < 0.001), tachypnea (P = <0.001), age (P = 0.006), high dose of steroid (P < 0.001), total duration of fever (≥ 2.5 days (for which sensitivity (S) and specificity (Sp) were 87 and 81%, respectively), serum-creatinine (≥ 0.45 mg%, S = 100%, Sp = 97%), serum-bilirubin (≥ 0.5 mg/dl, S = 100%. Sp = 90%), requirement of second-line antibiotics (P = 0.02), intensive-care (P ≤ 0.001), ventilatory support (P < 0.001), and requirement of packed cell (PC) transfusion (P = 0.02). In the multivariate analysis, mucositis (P = 0.02), HD steroid use (P = 0.026), and PC requirement (0.026) were identified as independent variables. The MASCC risk-index score was found to be meaningful at a score of ≤ 18. Other clinical and laboratory parameters were found to have a strong association with risk stratification in cancer patients during FN episodes.

PL-01 Predictive potential of haemostatic biomarkers for venous thromboembolism in cancer patients

Venous thromboembolism (VTE) is a common problem in cancer patients. However, the rates of VTE vary widely between different types of malignancies. Furthermore, patient- and treatment-related risk factors contribute to the risk of VTE. The prediction of the individual risk of VTE and the identification of patients with cancer that might benefit from thromboprophylaxis is a major clinical challenge, because the pathogenesis of cancer-associated VTE is multifactorial. Therefore, recent studies have focused on identification of predictive biomarkers for VTE. As there is a close interrelation between cancer and the haemostatic system, which leads to activation of haemostasis and fibrinolysis, biomarkers of the haemostatic system seem to be promising in predicting risk of developing VTE in cancer patients. Some candidate biomarkers or laboratory tests of the haemostatic system including platelet count, soluble P-selectin, tissue factor (TF) and TF-bearing microparticles, coagulation factor VIII, D-dimer, prothrombin fragment 1+2 and the thrombin generation assay have been reported to predict cancer-associated VTE. In addition, it has been shown that risk-scoring models, incorporating clinical parameters and biomarkers, allow risk stratification of cancer patients into groups at high- and at low risk of VTE. The accuracy of a previously established risk scoring model can be improved when it is expanded and biomarkers of the haemostatic system are added. The benefit of a targeted thromboprophylaxis for primary prevention of VTE in cancer patients based on risk assessment by measuring biomarkers or applying risk scoring models has to be shown in interventional clinical trials.

Validation of Circulating MMP-7 Level as an Independent Prognostic Marker of Poor Survival in Urinary Bladder Cancer

Molecular marker analyses aiming a more accurate disease characterization and risk stratification of cancer patients provided several promising marker candidates in the last few years. However, recent reviews underlined the paramount importance of validation, since many of the initially promising results could not be confirmed in independent patient cohorts. If serum or plasma is a more appropriate sample to test for prognostic markers is a matter of debate. We recently found serum MMP-7 levels to correlate with poor patients' prognosis in urinary bladder cancer. In this study, we examined associations of the MMP-7 plasma levels with clinical follow-up data in an independent cohort of bladder cancer patients to validate our former results and to assess if plasma is also suitable for MMP-7 analysis. Plasma levels of 97 patients and 22 controls were analyzed, using enzyme-linked immunosorbent assay. Associations between MMP-7 plasma concentrations and clinical data were assessed applying both univariate and multivariate analysis. Plasma MMP-7 levels were significantly higher in patients than in controls. Similarly to our former findings in sera, high MMP-7 plasma levels proved to be significant and independent predictors of both overall and disease-specific survival. In addition, we observed a metastasis-specific difference in MMP-7 levels between serum and plasma. In summary, we confirmed the prognostic relevance of circulating MMP-7 levels in an independent cohort of patients and concluded that circulating MMP-7 levels may help to identify bladder cancer patients at high-risk of disease progression who could benefit from an adjuvant chemotherapy or from an extended lymph node dissection.