Cancer Of Unknown Primary Patients Research Articles

Exploring the molecular landscape of cancer of unknown primary: A comparative analysis with other metastatic cancers.

Cancer of unknown primary (CUP) tumors are biologically very heterogeneous, which complicates stratification of patients for treatment. Consequently, these patients face limited treatment options and a poor prognosis. With this study, we aim to expand on the current knowledge of CUP biology by analyzing two cohorts: a well-characterized cohort of 44 CUP patients, and 213 metastatic patients with known primary. These cohorts were treated at the same institution and characterized by identical molecular assessments. Through comparative analysis of genomic and transcriptomic data, we found that CUP tumors were characterized by high expression of immune-related genes and pathways compared to other metastatic tumors. Moreover, CUP tumors uniformly demonstrated high levels of tumor-infiltrating leukocytes and circulating Tcells, indicating a strong immune response. Finally, the genetic landscape of CUP tumors resembled that of other metastatic cancers and demonstrated mutations in established cancer genes. In conclusion, CUP tumors possess a distinct immunophenotype that distinguishes them from other metastatic cancers. These results may suggest an immune response in CUP that facilitates metastatic tumor growth while limiting growth of the primary tumor.

Seeing the primary tumor because of all the trees: Cancer type prediction on low-dimensional data.

The Cancer of Unknown Primary (CUP) syndrome is characterized by identifiable metastases while the primary tumor remains hidden. In recent years, various data-driven approaches have been suggested to predict the location of the primary tumor (LOP) in CUP patients promising improved diagnosis and outcome. These LOP prediction approaches use high-dimensional input data like images or genetic data. However, leveraging such data is challenging, resource-intensive and therefore a potential translational barrier. Instead of using high-dimensional data, we analyzed the LOP prediction performance of low-dimensional data from routine medical care. With our findings, we show that such low-dimensional routine clinical information suffices as input data for tree-based LOP prediction models. The best model reached a mean Accuracy of 94% and a mean Matthews correlation coefficient (MCC) score of 0.92 in 10-fold nested cross-validation (NCV) when distinguishing four types of cancer. When considering eight types of cancer, this model achieved a mean Accuracy of 85% and a mean MCC score of 0.81. This is comparable to the performance achieved by approaches using high-dimensional input data. Additionally, the distribution pattern of metastases appears to be important information in predicting the LOP.

PD-L1 expression associates with favorable survival of patients with cancer of unknown primary (CUP) not treated with checkpoint inhibitors

PurposeCancer of unknown primary (CUP) is a heterogeneous entity with limited overall survival (OS) in most patients. Prognostic biomarkers are needed, particularly for treatment stratification. We investigated the impact of programmed death-ligand 1 (PD-L1) expression as prognostic marker in immunotherapy-naïve CUP patients. MethodsClinical data from patients with confirmed CUP diagnosis according to ESMO guidelines, treated at the West German Cancer Center, Essen from 2015 to 2021, were analyzed. Patients treated with checkpoint inhibitors were excluded. PD-L1 expression was assessed in tumor tissues following established guidelines. ResultsOf a cohort of 132 patients, 62 patients, including 30 patients with prognostically unfavorable CUP, met inclusion criteria and were evaluable for PD-L1 expression. Comparing PD-L1 Tumor Proportional Score (TPS) and Combined Positive Score (CPS) revealed almost complete concordance (96.2 %). Patients with PD-L1-positive CUP (TPS ≥1 %; n = 36; 58 %) had superior overall survival (median not reached) as compared to patients with PD-L1-negative CUP (median 14 months, 95 %CI 10.0–18.0; HR 0.3, p < 0.001). The benefit of PD-L1 positivity (n = 12; 40 %) was maintained in the unfavorable CUP subgroup (median 20 months, 95 %CI 3.0–37.0 versus 10 months, 95 %CI 3.1–16.9; HR 0.4, p = 0.032). In PD-L1-positive CUP there was a positive correlation between the level of PD-L1 expression and survival (TPS ≥50 %: median survival not reached, HR 0.1; TPS 1 to 49 %: OS: 77 months, HR 0.4). ConclusionPD-L1 expression associates with favorable survival in checkpoint inhibitor-naïve CUP patients. This implies a role of cancer immunity in CUP biology and may nominate PD-L1 for patient stratification in further studies and clinical care.

Synergic activity of FGFR2 and MEK inhibitors in the treatment of FGFR2-amplified cancers of unknown primary

Patients with cancer of unknown primary (CUP) carry the double burden of an aggressive disease and reduced access to therapies. Experimental models are pivotal for CUP biology investigation and drug testing. We derived two CUP cell lines (CUP#55 and #96), and corresponding patient-derived xenografts (PDXs), from ascites tumor cells. CUP cell lines and PDXs underwent histological, immune-phenotypical, molecular, and genomic characterization confirming the features of the original tumor. The tissue-of-origin prediction was obtained from the tumor microRNA expression profile and confirmed by single-cell transcriptomics. Genomic testing and FISH analysis identified FGFR2 gene amplification in both models, in the form of homogenously staining region (HSR) in CUP#55 and double minutes in CUP#96. FGFR2 was recognized as the main oncogenic driver and therapeutic target. FGFR2-targeting drug BGJ-398 (infigratinib) in combination with the MEK inhibitor trametinib proved to be synergic and exceptionally active, both in vitro and in vivo. The effects of the combined treatment by single-cell gene expression analysis revealed a remarkable plasticity of tumor cells and the greater sensitivity of cells with epithelial phenotype. This study brings personalized therapy closer to CUP patients and provides the rationale for FGFR2 and MEK targeting in metastatic tumors with FGFR2 pathway activation.

The evolution of molecular management of carcinoma of unknown primary.

There is significant need to improve diagnostic and therapeutic options for patients with cancer of unknown primary (CUP). In this review, we discuss the evolving landscape of molecular profiling in CUP. Molecular profiling is becoming accepted into the diagnostic work-up of CUP patients with tumour mutation profiling now described in international CUP guidelines. Although tissue-of-origin (ToO) molecular tests utilising gene-expression and DNA methylation have existed some time, their clinical benefit remains unclear. Novel technologies utilising whole genome sequencing and machine learning algorithms are showing promise in determining ToO, however further research is required prior to clinical application. A recent international clinical trial found patients treated with molecularly-guided therapy based on comprehensive-panel DNA sequencing had improved progression-free survival compared to chemotherapy alone, confirming utility of performing genomic profiling early in the patient journey. Small phase 2 trials have demonstrated that some CUP patients are responsive to immunotherapy, but the best way to select patients for treatment is not clear. Management of CUP requires a multifaceted approach incorporating clinical, histopathological, radiological and molecular sequencing results to assist with identifying the likely ToO and clinically actionable genomic alternations. Rapidly identifying a subset of CUP patients who are likely to benefit from site specific therapy, targeted therapy and/or immunotherapy will improve patient outcomes.

A multi-centre, single arm, phase 2 trial of pembrolizumab in treatment-naïve patients with carcinoma of unknown primary site.

2659 Background: Cancer of unknown primary (CUP) represents a heterogeneous group of cancers that present with metastatic epithelial disease with no identifiable primary lesion at the time of diagnosis. Patients with poor prognostic features have limited effective systemic therapy options. Researchers hypothesize that these primary lesions are occult, in part, due to anti-cancer immune surveillance. We hypothesize, pembrolizumab will have activity in this setting. Methods: This single arm phase 2 study evaluated the use of pembrolizumab in treatment naïve patients with poor prognosis CUP. Patients ≥18 years of age with ECOG PS 0-1 and histologically proven, measurable metastatic carcinoma with no primary site identified were eligible. Pembrolizumab was administered at 200 mg intravenously every 3-weeks for up to 24 months. The primary endpoint was safety and objective response rate (ORR) per RECIST 1.1. Secondary objectives included progression free survival (PFS) and overall survival (OS). The study used a Simon’s two stage design based on a one-sided test with a type 1 error rate of 0.1, null hypothesis for ORR ≤ 20% yielding a power of 0.8 with an alternate hypothesis of ≥40%. The null hypothesis will be rejected if 8 or more responses are observed in 25 patients. Mismatch repair (MMR) and microsatellite instability (MSI) testing was not mandatory for patients. Results: There were 27 evaluable and 6 unevaluable (no response assessment, included in adverse event analysis) patients. The median age was 64 (range 21-80) and the majority were male (57.6%) and ECOG 1 (84.8%). Four patients experienced grade 3 treatment related adverse events (pain, anorexia, fatigue, confusion, and anemia). There were no treatment related deaths or new signals of toxicity. Of the evaluable patients, the ORR was 18.5% (4 partial responses and 1 complete response), disease control rate was 51.9% (9 stable disease) and thus the null hypothesis could not be rejected. For evaluable patients, the median PFS and OS were 2.56 (95% CI 2.10 – 7.46) and 9.79 (95% CI 3.84 – 16.03) months respectively. The one patient who developed a CR had MMR deficiency and has completed 2 years of therapy. Of the 4 patients with a partial response, one had an MMR proficient tumour, while the remaining 3 had unknown MMR status. Conclusions: Although not meeting the primary endpoint, pembrolizumab has some activity in CUP patients. Patients with CUP should be screened for MMR or MSI. Future studies should evaluate combination therapies with the addition of pembrolizumab. Clinical trial information: NCT03391973 .

Sequential PET/CT and pathological biomarker crosstalk predict response to PD-1 blockers alone or combined with sunitinib in propensity score-matched cohorts of cancer of unknown primary treatment.

The efficacy of immune checkpoint inhibitors (ICIs), including toripalimab and pembrolizumab, has not been confirmed in the treatment of cancer of unknown primary (CUP), which has a very poor prognosis. Combined with anti-angiogenic therapies, ICIs are hypothesized to be effective in prolonging overall survival. The study aims to give evidence on the treatment effects of sunitinib combined with ICIs, find pathological biomarkers associated with changes in volumetric 18F FDG PET/CT parameters, and investigate inner associations among these markers associated with response on PET/CT. The study recruited patients receiving combined treatment (ICIs + sunitinib), compared the effects of combined treatment with those of separate treatment and age-matched negative controls, and analyzed propensity score-matched (PSM) pairs. Markers associated with survival were identified, and their inner associations were tested using structural equation modeling. A total of 292 patients were enrolled in the final analysis, with 53 patients receiving combined treatment. Survival analysis demonstrated significantly prolonged survival in either combined or separate treatment, with the combined arm showing better response when PSM-paired using pre-treatment whole-body PET/CT parameters. The angiogenic markers KDR and VEGF mediate the PD-1 blockade impact on volumetric value changes in positive and negative manners. The anti-angiogenic agent sunitinib may potentiate PD-1 blockade by diminishing angiogenesis or its downstream effects. The combined separate treatment increased the survival of CUP patients, and the responses could be evaluated using volumetric PET/CT parameters.

Management and prognosis of patients with cancer of unknown primary: 20 years of experience.

Cancer of unknown primary (CUP) is a difficult clinical entity to manage. The aim of the study was to investigate the sociodemographic and pathological characteristics, treatment options, and factors affecting overall survival (OS) in CUP patients whose primary tumor was not detected during follow-up. A total of 243 CUP patients whose primary tumors could not be detected during follow-up were included in the study. Their demographic characteristics, survival outcomes, and prognostic factors were investigated. Of the 243 patients included in this study, 61.7% were male and 38.3% were female, and the median age was 61 (range: 19-90) years. The most common histological type was adenocarcinoma (79%). The median follow-up time of the patients was 30.3 months (95% CI: 11.4-49.3), the median OS time was 9.1 months (95% CI: 7.2-11.0), and 72.4% of the patients received at least 1 line of chemotherapy (CT). The difference in survival between the patients who did and did not receive CT was statistically significant (median OS: 10.1 vs. 4.2 months, p = 0.003). According to the multivariate analysis, the presence of cholestasis (HR: 0.48, 95% CI: 0.29-0.79, p = 0.004), lung metastasis (HR: 0.69, 95% CI: 0.51-0.95, p = 0.001), second-line chemotherapy (HR: 1.69, 95% CI: 1.14-2.49, p < 0.001), and Eastern Cooperative Oncology Group (ECOG) performance status (HR: 0.20, 95% CI: 0.10-0.40, p < 0.001) were independent prognostic factors influencing OS. CUP patients who receive multiple lines of chemotherapy tend to have longer survival. This is the first study to report cholestasis as a prognostic factor in CUP patients. In addition, the presence of lung metastases, not receiving second-line chemotherapy, and ECOG performance status (≥2) were found to be independent poor prognostic factors.

Cancer of unknown primary—state of the art

SummaryThe Cancer of Unknown Primary (CUP) syndrome does not represent a single clinical entity but comprises a diverse group of malignant tumors. CUP is defined as a histologically confirmed metastatic malignant disease for which the primary tumor cannot be identified after completing initial diagnostics. Although its incidence has decreased, managing CUP patients remains a significant clinical challenge, particularly for those in the poor prognosis group, where the overall survival is merely 3 months. In this context, accurate diagnostics and prompt treatment initiation are crucial. While platinum-based chemotherapy still plays a central role in CUP therapy, the future holds promise with the advent of next-generation targeted therapies and immunotherapies guided by the results of advanced sequencing tools. These advancements offer hope for improving the bleak outcomes faced by these patients.

Factors associated with unknown primary status in head and neck squamous cell carcinoma

ObjectivesHead and neck cancer of unknown primary (CUP) poses significant therapeutic challenges. We compare CUP and oropharyngeal primary (OP) cases to identify factors associated with tumor detection. MethodsThe 2004–2019 National Cancer Database was queried to identify CUP and OP cases based on clinical and pathologic TNM staging. Clinical and demographic characteristics were compared by primary detection and HPV status with descriptive statistics. Multivariable logistic regression models to characterize OP detection were constructed. Among HPV-positive and negative patients, respectively, OP and CUP patients were matched by clinical nodal disease. Cox proportional-hazards models were constructed using matched cohorts to characterize survival. Results81,053 CUP and OP cases were identified; 64.3 % were HPV-positive. OP detection increased over time in HPV-positive and negative disease. HPV-positive status had higher odds of OP detection (odds ratio (OR) = 1.77, p < 0.001), while females (OR = 0.95, p = 0.008), and black (OR = 0.82, p < 0.001) and Asian (OR = 0.7, p < 0.001) patients had lower odds compared to males and whites, respectively. In HPV-positive and negative disease, OP patients had higher 2 and 5-year survival rates than CUP (p < 0.001). Primary detection status conferred lower death risk in HPV-positive (hazard ratio (HR) = 0.85, p < 0.001) and negative disease (HR = 0.87, p < 0.001) when controlling for age, sex, race, comorbidities, insurance, treatment facility, and income. ConclusionIn the largest cohort of CUP to date, we report a survival benefit in primary tumor detection regardless of HPV status. Groups with higher persistent CUP rates, including non-white, female, HPV-negative, and low income patients, may benefit from increased diagnostic workup to improve detection and treatment.

Patient characteristics associated with definitive diagnosis of metastatic pancreatic cancer in those initially diagnosed with cancer of unknown primary.

Cancer of unknown primary (CUP) and pancreatic cancer (PC) are malignancies associated with poor prognosis. CUP is the fourth most common cause of cancer mortality in the US, and median survival time is 3-4 months. PC is the third most common cause of cancer mortality in the US, and median survival time for patients with stage 3 or 4 PC is 2-3 months. The present study aimed to understand the patient characteristics of those initially misdiagnosed with CUP who ultimately received a diagnosis of PC. The present study used 2010-2015 Surveillance, Epidemiology, and End Results-Medicare data, a US population-based cancer registry linked to Medicare health insurance claims. Odds ratios (ORs) and 95% confidence intervals were calculated using two binary logistic regression models to compare the characteristics of patients who received definitive diagnosis between the CUP-PC group (those with an initial diagnosis of CUP who eventually received a stage 3 or 4 PC diagnosis) and the PC group (those diagnosed with stage 3 or 4 PC only). Approximately 26% of patients who received a definitive diagnosis of metastatic PC started with an initial diagnosis of CUP (n=17,565). The odds of definitive PC diagnosis in patients with CUP were lower for those with a comorbidity score of 0 [OR, 0.85 (95% CI: 0.79, 0.91)] and epithelial/unspecified histology [OR, 0.76 (95% CI: 0.71, 0.82)]. The odds of definitive PC diagnosis in patients with CUP were higher for patients of other race [OR, 1.27 (95% CI: 1.13, 1.43)] compared with white patients. Definitive diagnosis of PC in patients with CUP was lower in patients who were older with fewer or no comorbidities and unspecified histology. The complexity of CUP diagnosis and patient performance status may influence delays in diagnosis to a known primary site.

Early Improved Functional Outcomes in Head and Neck Cancer Patients with Primary Tumor Detection.

We characterize functional outcomes in head and neck cancer of unknown primary (CUP) based on primary site identification. In this retrospective study, CUP cases were categorized as known primaries (KP) if a tumor was localized after diagnostic workup or persisting unknown primaries (UP). Age, sex, HPV status, diagnostic methods, and treatments regimens were collected. Pretreatment and short-term posttreatment (3-6 months after completion of treatment) weights, PHQ-9, Eating Assessment Tool (EAT-10), and Voice Handicap Index (VHI-10) scores were compared between UP and KP. Among 67 CUP patients, 35 (52.2%) had identified primaries (91.4% oropharyngeal and 8.6% nasopharyngeal). KP patients were younger (58 vs. 64, p = 0.04) and more likely to be HPV-positive (88.6% vs. 50%, p = 0.002). Overall detection rates were 16.7% for PET/CT, 34.7% for direct laryngoscopy, and 46.6% for transoral robotic oropharyngectomy. Diagnostic workup was not significantly different between groups. Patients with KP received smaller intermediate radiation dose volumes (436.5 vs. 278.9 cc, p = 0.03) and lower doses to the cricopharyngeal muscle (41.6 vs. 24.6 Gy, p = 0.03).Pretreatment weights, PHQ-9, EAT-10, and VHI-10 scores did not differ between groups. However, posttreatment, UP had greater relative weight loss (-14.1% vs. -7.6%, p = 0.032), higher EAT-10 scores (12.5 vs. 3, p = 0.004), and higher PHQ-9 scores (6 vs. 1.4, p = 0.017). Specifically, UP reported more stressful swallowing, difficulty swallowing solids and pills, and swallowing affecting public eating. KP patients experienced less weight loss, depression, and reduced swallowing dysfunction, highlighting an early functional benefit of primary tumor identification likely driven by reduced radiation treatment volumes. 4 Laryngoscope, 134:701-707, 2024.

SUVmax Above 20 in 18F-FDG PET/CT at Initial Diagnostic Workup Associates with Favorable Survival in Patients with Cancer of Unknown Primary.

Cancer of unknown primary (CUP) is a heterogeneous entity with a limited prognosis. Novel prognostic markers are needed for patient stratification in prospective clinical trials exploring innovative therapies. Methods: In CUP patients treated at the West German Cancer Center Essen, the prognostic value of 18F-FDG PET/CT at the initial diagnostic workup was analyzed by comparing overall survival (OS) in patients who underwent 18F-FDG PET/CT with those who did not. Results: Of 154 patients with a CUP diagnosis, 76 underwent 18F-FDG PET/CT at the initial diagnostic workup. The median overall survival (OS) of the full analysis set was 20.0 mo. Within the PET/CT subgroup, an SUVmax above 20 was associated with significantly superior OS (median OS, not reached vs. 32.0 mo; hazard ratio, 0.261; 95% CI, 0.095-0.713; P = 0.009). Conclusion: Our retrospective work shows that an SUVmax above 20 on 18F-FDG PET/CT at the initial diagnostic workup is a favorable prognostic factor in patients with CUP. This finding deserves further prospective studies for validation.

Cancer of unknown primary (CUP) through the lens of precision oncology: a single institution perspective

PurposeFor patients with cancer of unknown primary (CUP), treatment options are limited. Precision oncology, the interplay of comprehensive genomic profiling (CGP) and targeted therapies, aims to offer additional treatment options to patients with advanced and hard-to-treat cancers. We aimed to highlight the use of a molecular tumor board (MTB) in the therapeutic management of CUP patients.MethodsIn this single-center observational study, CUP patients, presented to the MTB of the Comprehensive Cancer Center Munich LMU, a tertiary care center, were analyzed retrospectively. Descriptive statistics were applied to describe relevant findings.ResultsBetween June 2016 and February 2022, 61 patients with unfavorable CUP were presented to the MTB, detected clinically relevant variants in 74% (45/61) of patients, of which 64% (29/45) led to therapeutic recommendation. In four out of 29 patients (14%), the treatment recommendations were implemented, unfortunately without resulting in clinical benefit. Reasons for not following the therapeutic recommendation were mainly caused by the physicians’ choice of another therapy (9/25, 36%), especially in the context of worsening of general condition, lost to follow-up (7/25, 28%) and death (6/25, 24%).ConclusionCGP and subsequent presentation to a molecular tumor board led to a high rate of therapeutic recommendations in patients with CUP. Recommendations were only implemented at a low rate; however, late GCP diagnostic and, respectively, MTB referral were found more frequent for the patients with implemented treatment. This contrast underscores the need for early implementation of CGP into the management of CUP patients.

Abstract 5429: Using machine learning to predict tissue of origin from somatic mutation features

Abstract Background: Predicting tissue of origin (ToO) using clinical and molecular data improves diagnostic accuracy up to 95% in patients with Cancer of Unknown Primary (CUP). It is hypothesized that better treatment stratification of CUP patients using omics and machine learning (ML) classifiers may improve prognosis. Methods: We used publicly available whole exome somatic mutation data from 4733 primary solid tissue samples, across 11 tumor types from the TCGA database, and employed a ML classifier to predict their ToO. We used 5 sets of modeling features: 1) Non-silent somatic mutation burden of 230 cancer-related genes 2) Frequency of SNP substitution type 3) Trinucleotide mutation frequency 4) Copy number variation of the 230 cancer-related genes 5) Presence of hotspot mutations. We trained a Support Vector Machine on a training subset (80% of samples) and tuned the hyperparameters maximizing a 5-fold cross-validation F1-score. We then tested the model performance on a validation subset (20% of samples) and on a limited (n=6) dataset of metastatic samples present in the TCGA database. Results: On the primary tumor validation set, we achieved an average AUC of 0.98(std: 0.02) and top 1, top 2 and top 3 accuracies of 80%(std: 0.11), 90%(std: 0.08) and 95%(std: 0.04) respectively, across 11 tumor types. The classification accuracy plateaus after ~300 samples, suggesting further data collection may benefit low performing tumor types. The 2 worst performers: esophageal and stomach cancers were mostly misclassified with colorectal cancers, reflecting their relative similarity. On metastatic samples (n=6) the model achieved a 67% accuracy, this is work in progress. Conclusion: Our study confirms the potential for a DNA-based machine learning approach to improve prognosis in CUP patients by aiding diagnosis of ToO. To this end, we plan to take this study further by applying this approach to large, independent datasets derived from metastatic samples and liquid biopsies from CUP patient cohorts. Table 1. Top_1_acc Top_2_acc Top_3_acc Precision Recall F1_score Training_size Breast 0.88 0.97 0.99 0.88 0.84 0.86 756 Colorectal 0.85 0.97 0.98 0.83 0.83 0.83 460 Oesophagus 0.5 0.75 0.89 0.51 0.5 0.51 144 Liver 0.83 0.92 0.96 0.82 0.89 0.85 288 Lung 0.86 0.91 0.94 0.98 0.81 0.88 396 Ovary 0.87 0.94 0.97 0.77 0.9 0.83 312 Pancreas 0.73 0.79 0.85 0.62 0.7 0.66 132 Prostate 0.88 0.94 0.98 0.75 0.88 0.81 380 Sarcoma 0.78 0.84 0.96 0.8 0.82 0.81 180 Stomach 0.73 0.91 0.95 0.68 0.58 0.62 340 Endometrial 0.9 0.97 0.99 0.86 0.87 0.87 400 Mean 80.09% 90.09% 95.09% 0.77 0.78 0.78 344 Top_1_acc_n Top_2_acc_n Top_3_acc_n n samples Metastatic breast 2 2 2 2 Metastatic prostate 0 0 0 1 Metastatic pancreas 0 0 0 1 Metastatic sarcoma 1 1 1 1 Metastatic oesophagus 1 1 1 1 Mean accuracy 67% 67% 67% Citation Format: Andrea Giorni, Prabu Sivasubramiam, Aidan Kubeyev, Jordan Laurie, Luiz Silva, Matthew Foster, Uzma Asghar, Matthew Griffiths. Using machine learning to predict tissue of origin from somatic mutation features. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5429.

Abstract 5405: An ML-based tool for predicting tissue of origin for cancer of unknown primary (CUP) based on genomic and transcriptomic data

Abstract The ability to locate a tumor primary site for patients with a cancer of unknown primary (CUP) is a major obstacle in providing personalized therapeutic options and access to clinical trials. Despite the recent use of molecular-based tools to identify the tumor tissue-of-origin (TOO), overall survival for CUP patients remains low. Here, we present an AI-based tool that predicts the TOO by using genomic and transcriptomic data to classify CUP tumors into hierarchically-organized molecular subgroups. The TOO predictor was composed of DNA, RNA, and Consensus classifiers that were hierarchically organized with respect to molecular diagnosis with upper level clusters based on common molecular features reflecting similar cell of origin, and lower levels containing specific diagnoses for further classification. The ML-based DNA classifier was trained on a dataset of publicly available genomic data generated from 8,000 samples, and independently validated using more than 5,500 samples. The ML-based RNA classifier was trained on a dataset of publicly available transcriptomic data created from more than 10,100 samples with tumor- and normal-specific features for each cancer type, and independently validated using 20,000 samples. The Consensus classifier, combining outputs from both DNA and RNA algorithms, was trained on a dataset of genomic and transcriptomic data from 1,000 samples, and validated on an independent dataset of 2,000 samples. Each classifier contained features selected based on data analysis according to the weighted F1-score, and the best hyperparameters for the final model. The 3-classifier algorithm predicts TOO for 33 cancer types and subtypes belonging to solid neoplasms, independently of sample source, sequencing methods, and cohort. Validation of the Consensus classifier showed a higher accuracy (95% f1-score) compared to the DNA and RNA classifiers (79% and 93% f1-scores, respectively), along with a high sensitivity (95%), specificity (99%), and precision (96%), as it takes into account both genomic events and expression patterns. The TOO predictor was prospectively validated on approximately 298 clinical samples with a known diagnosis using all classifiers. The diagnosis was identified in 90% of clinical cases (295/297) by the Consensus classifier with 90% sensitivity and 94% precision. The call rate for the DNA and RNA classifiers was above 95%. Of note, sensitivity of the top 4 predicted diagnoses was &amp;gt; 90% for all 3 classifiers, and the calculated rule-out accuracy of the Consensus classifier was 97%. In conclusion, an ML-based algorithm was developed that utilizes genomic and transcriptomic data to accurately predict the TOO for CUP tumors. Utilizing the Consensus classifier after DNA and RNA classifiers helps to identify the TOO of the tumor with high specificity, which can guide precision oncology therapeutic options. Citation Format: Zoia Antysheva, Daria Kiriy, Anton Sivkov, Alexander Sarachackov, Alexandra Boyko, Naira Samarina, Nara Shin, Jessica H. Brown, Ivan Kozlov, Viktor Svekolkin, Alexander Bagaev, Nathan Fowler, Nikita Kotlov. An ML-based tool for predicting tissue of origin for cancer of unknown primary (CUP) based on genomic and transcriptomic data. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5405.

Breast-like cancer of unknow primary : Implications for radiological diagnostics

Cancer of unknown primary (CUP) is defined by the presence of metastatic disease with an undetectable primary tumor at the time of presentation despite standard-of-care imaging. Although the prognosis of most CUP patients is poor, certain subgroups with more favorable prognosis have been defined. Women with isolated axillary lymph node metastases and confirmed histologic adenocarcinoma or poorly differentiated subtype, no other distant metastases, and no evidence of aprimary cancer including primary breast carcinoma evaluated by clinical examination, computed tomography of thorax and abdomen, mammography, breast ultrasound, and breast magnetic resonance imaging (MRI) represent apotentially curable subgroup of patients with CUP. Breast MRI is the most important radiological modality in the diagnostic workup of breast-like CUP to exclude aprimary cancer in the breast. Breast-like CUP patients are treated according to guidelines for patients with node-positive breast cancer. Standard-of-care adjuvant systemic therapy should be given. Axillary lymph node dissection (ALND) is indicated. If no primary cancer in the breast is detected, surgery of the ipsilateral breast should not be performed. Radiotherapy of the ipsilateral breast and supra-/infraclavicular lymph nodes should be discussed.

Significance of Distinct Liquid Biopsy Compartments in Evaluating Somatic Mutations for Targeted Therapy Selection in Cancer of Unknown Primary.

Cancer of unknown primary (CUP) accounts for 2-5% of all cancer diagnoses, wherein standard investigations fail to reveal the original tumor site. Basket trials allocate targeted therapeutics based on actionable somatic mutations, independent of tumor entity. These trials, however, mostly rely on variants identified in tissue biopsies. Since liquid biopsies (LB) represent the overall tumor genomic landscape, they may provide an ideal diagnostic source in CUP patients. To identify the most informative liquid biopsy compartment, we compared the utility of genomic variant analysis for therapy stratification in two LB compartments (circulating cell-free (cf) and extracellular vesicle (ev) DNA). CfDNA and evDNA from 23 CUP patients were analyzed using a targeted gene panel covering 151 genes. Identified genetic variants were interpreted regarding diagnostic and therapeutic relevance using the MetaKB knowledgebase. LB revealed a total of 22 somatic mutations in evDNA and/or cfDNA in 11/23 patients. Out of the 22 identified somatic variants, 14 are classified as Tier I druggable somatic variants. Comparison of variants identified in evDNA and cfDNA revealed an overlap of 58% of somatic variants in both LB compartments, whereas over 40% of variants were only found in one or the other compartment. We observed substantial overlap between somatic variants identified in evDNA and cfDNA of CUP patients. Nonetheless, interrogation of both LB compartments can potentially increase the rate of druggable alterations, stressing the significance of liquid biopsies for possible primary-independent basket and umbrella trial inclusion.

A qualitative study of patients with Cancer of Unknown Primary: Perceptions of communication, understanding of diagnosis and genomic testing, and information needs.

Patients with Cancer of Unknown Primary (CUP) commonly report poor understanding of their illness and high levels of psychological distress. Despite the potential benefits to CUP patients, there is a paucity of research exploring the reasons behind poor understanding of a CUP diagnosis. The aim of this study was to understand patients' experiences of communication with doctors, their understanding of diagnosis and the role of genomic testing, as well as their information needs. Semi-structured interviews explored CUP patients' perceptions of communication with their doctors, understanding of their illness, and their needs regarding medical information. Qualitative inductive thematic analysis of transcribed audio-recordings was employed. Nineteen patients were recruited from within a prospective cohort study involving routine genomic testing of CUP patients. CUP patients had varied perceptions of communication with doctors as well as different levels of need, readiness, and capacity for information. Some patients felt well understood and supported by their doctors while others did not. Many patients reported feeling overwhelmed and shocked when receiving their cancer diagnosis and emphasized the importance of family support in receiving and understanding medical information. While patients understood the implications of genomic testing for treatment and diagnosis, few had a detailed understanding of genomic testing. Patients' experience of communication and understanding of CUP could be potentially improved by clinicians' assessment of the communication style preferred by each patient and their family and the development of online resources to meet their evolving information needs.

MutFusVAE: Mutational Fusion Variational Autoencoder for Predicting Primary Sites of Cancer

The metastatic propensity of malignant primary tumors is a recurring theme when it comes to the cause of mortality in cancer. Establishing the primary site of a metastatic cancer is a significant but challenging task. There are ∼3% of metastatic cancer cases diagnosed as cancer of unknown primary (CUP), and the conventional diagnostic process fails to detect the primary site for 80% of CUP patients. Benefiting from the explosion of the information available from large-scale tumor DNA sequencing projects, it became favorable to predict the cancer primary sites from genomic perspective. The existing methods on the task intensively studied the mutational and oncogenic features with assists of machine learning (ML) and deep learning (DL) techniques, yet lack of development of a model architecture tailored for the mutational features. To address the gap, in this research, we aimed to develop a DL methodology specialized for the mutational data. A mutational fusion variational autoencoder (MutFusVAE) deep architecture is proposed to actualize the idea1. We downloaded mutational profiles meeting our criteria of 2,603 tumor samples, which get split into 2,082 training samples and 521 (20%) held-out testing samples, from the International Cancer Genome Consortium (ICGC) database. The proposed methods achieved 94% overall classification accuracy for differentiating among seven primary sites on the held-out testing set. The results show the discriminative power brought by a specialized design of deep models for the mutational data and gain insights to facilitate DL-based genomic diagnostics for cancer from a modeling view.