Christer Borgfeldt

Machine learning-enhanced gas sensor technology identifies ovarian and endometrial cancer of all stages through plasma volatile organic compound patterns

Ovarian cancer presents with non-specific symptoms that make early diagnosis challenging and the prognosis poor. Ovarian and endometrial cancers exhibit similar genomic mutations and biomarker profiles. Endogenous volatile organic compounds (VOCs) are products of metabolic activity. In cancer, metabolites increase due to tumour necrosis, leading to cancer-specific VOC patterns. The aim of this study was to evaluate VOC analyses in plasma as diagnostic tests for early diagnosis of ovarian and endometrial cancer. Preoperative plasma from 133 women with ovarian cancer (stages I-IV or borderline ovarian tumors) and 41 women with endometrial cancer (stages I-IV) was compared to 115 healthy controls with highly sensitive gas sensors. Data analyses were performed using feature extraction from 32 gas sensors per sample. 85 features were extracted from each signal (including statistical, time-domain, and frequency-domain features), and training and test datasets were formed. The features underwent an iterative redundancy removal process for stepwise optimization of models. Model robustness was assessed using a train/test split scheme with unique datasets, leading to a model optimized for diagnostic performance. By implementing sequential binary classification boosting-based models, it was possible to determine not only the presence or not of cancer, but also to distinguish between ovarian- and endometrial cancer, and the stage of the cancer. The VOC analysis, powered by a 5-fold cross-validated ensemble classifier, achieved exceptional diagnostic performance. It correctly identified all 133 patients with ovarian cancer and borderline ovarian tumors, all 41 cases of endometrial cancer, and all 115 healthy controls. For staging, the model accurately classified 172 out of 174 (98.9%) cancer cases as stage I vs. II-IV. On validation data, the classifier yielded an accuracy of 96.63% (95% CI: 96.56%-96.70%), sensitivity of 96.42% (95% CI: 96.29%-96.54%), and specificity of 96.88% (95% CI: 96.76%-97.01%). These metrics were robustly replicated on the independent test set, with an accuracy of 97.13% (95% CI: 96.80%-97.45%), sensitivity of 96.92% (95% CI: 96.49%-97.35%), and specificity of 97.37% (95% CI: 96.97%-97.77%). Complementing this, four additional classifiers (each with accuracy exceeding 90%) were developed and integrated into a cascaded algorithm to enable multi-class discrimination (ovarian cancer and endometrial cancer vs. healthy controls), cancer typing (ovarian vs. endometrial), and staging (stage I vs. later stages). The analysis of VOCs correctly identified 133 out of 133 patients with ovarian cancer and borderline ovarian tumour. All 41 cases of endometrial cancer were correctly identified, as were all the 115 healthy controls. In 172 out of 174 (98.9%) cancer cases the model correctly classified stage I vs. II-IV. VOC analysis emitted to gas-phase from plasma demonstrates high sensitivity and specificity for diagnosing ovarian cancer, including borderline ovarian tumors and endometrial cancers, compared to healthy controls. VOC analyses accurately differentiated between early and advanced stages of both cancer types. Future external validation needs to be performed. The Strategic Innovation Programs Swelife and MedTech4Health, a joint venture of Vinnova, Formas and the Energy Agency (grant No. 2022-03464 and grant No. 2023-03874) and within the Convergence Accelerator Program (Track L - Real-World Chemical Sensing Applications), funded by the Swedish Research Council, Vetenskapsrådet (grant No. 2023-07219), and Sweden's Innovation Agency, Vinnova (grant No. 2023-04186), in collaboration with the US National Science Foundation (NSF). The computations, data handling, and machine learning model training and testing were conducted in the MATLAB environment and enabled by resources provided by the National Academic Infrastructure for Supercomputing in Sweden (NAISS), partially funded by the Swedish Research Council through grant agreement no. 2022-06725. This work received funding from the European Union's Horizon Europe Research and Innovation Programme under Grant Agreement No 101214318 (DISARM). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the Health and Digital Executive Agency (HaDEA). Neither the European Union nor HaDEA can be held responsible for them.

Survival in endometrial cancer in relation to minimally invasive surgery or open surgery – a Swedish Gynecologic Cancer Group (SweGCG) study

Abstract Background The aim of this study was to analyze overall survival in endometrial cancer patients’ FIGO stages I-III in relation to surgical approach; minimally invasive (MIS) or open surgery (laparotomy). Methods A population-based retrospective study of 7275 endometrial cancer patients included in the Swedish Quality Registry for Gynecologic Cancer diagnosed from 2010 to 2018. Cox proportional hazard models were used in univariable and multivariable survival analyses. Results In univariable analysis open surgery was associated with worse overall survival compared with MIS hazard ratio, HR, 1.39 (95% CI 1.18–1.63) while in the multivariable analysis, surgical approach (MIS vs open surgery) was not associated with overall survival after adjustment for known risk factors (HR 1.12, 95% CI 0.95–1.32). Higher FIGO stage, non-endometrioid histology, non-diploid tumors, lymphovascular space invasion and increasing age were independent risk factors for overall survival. Conclusion The minimal invasive or open surgical approach did not show any impact on survival for patients with endometrial cancer stages I-III when known prognostic risk factors were included in the multivariable analyses.

Risks of non-ovarian cancers in women with borderline ovarian tumor: a national cohort study in Sweden

Abstract Background Associations between different cancer types are known. The affirmation of the risk for non-ovarian cancer after ovarian borderline tumors (BOT) is, however, sparse. Aim To analyze the risk of subsequent or simultaneous cancers in women with BOTs compared with the general female Swedish population. Methods An open cohort study (1995–2018) was conducted where a diagnosis of BOTs as well as subsequent or simultaneous cancer diagnoses were obtained from the Swedish Cancer Register and matched to the Total Population Register. Each woman with BOT was followed until non-ovarian cancer, death or emigration and could only be included once for the outcome. Standardized incidence ratios (SIRs) with 95% confidence intervals (CIs) for specific non-ovarian cancers were analyzed. Results The 4998 women with serous and mucinous BOTs were diagnosed during 1995–2018 with a mean age of 55.7 years (SD 16.0) at diagnosis. Compared with the general female population, women with BOTs had increased risks for non-ovarian cancer in colon (SIR = 2.5; 95% CI 2.0–3.1), rectum (SIR = 1.7; 95% CI 1.1–2.5), small intestine (SIR = 5.0; 95% CI 2.3–9.5), cervix (SIR = 2.5; 95% CI 1.4–4.2), endometrium (SIR = 2.4; 95% CI 1.9–3.1), pancreas (SIR = 2.3; 95% CI 1.4–3.5), upper aerodigestive tract (SIR = 2.2; 95% CI 1.2–3.8), lung (SIR = 1.8; 95% CI 1.4–2.3), kidney (SIR = 2.3; 95% CI 1.4–3.7) and bladder (SIR = 1.8; 95% CI 1.1–2.8). Among women with serous BOTs, the risk of thyroid gland cancer (SIR = 3.1; 95% CI 1.2–6.4) was also increased. Lung and pancreas cancer showed increased risks more than 1 year after a diagnosis of BOT. Conclusions This Swedish population-based study demonstrated an increased risk of multiple malignancies including lung and pancreatic cancers beyond the first year of diagnosis in patients with borderline ovarian tumors (BOTs), suggesting a potential shared etiology.

Vulvar cancer incidence and net survival in Sweden 1960 to 2019: A population‐based national study

AbstractIntroductionVulvar cancer is a rare gynecological cancer affecting mostly older women. The aim of this population‐based study was to investigate the incidence and net survival of vulvar cancer in Swedish women from 1960 to 2019.Material and methodsData were retrieved from the mandatory Swedish Cancer Registry consisting of all women diagnosed with vulvar cancer between 1960 and 2019. Only women with a morphologically verified diagnosis of vulvar cancer were included. The individuals were then further matched with the Swedish Death Registry up until May 31, 2020.ResultsIn total, 8499 women were included with the following morphologies: squamous cell carcinoma 7250 (85.8%), malignant melanoma 539 (6.4%), adenocarcinoma 401 (4.8%) and other: 259 (3.1%). More than 50% of vulvar cancer cases occurred in women aged between 65 and 84 years of age. The 5‐year age‐standardized net survival increased from 53.0% (95% confidence interval [CI] 48.9–57.5) in 1960 to 72.1% (95% CI 68.8–75.5) in 2019. The proportion of adenocarcinoma among all cases increased from 2.0% to 8.7% between the 1960s and 2010s and an increase in age‐standardized 5‐year net survival was found for adenocarcinoma.ConclusionsThe age‐standardized incidence of vulvar cancer cases in Sweden was stable between 1960 and 2019. During the study period, an increase in adenocarcinoma and a decrease in malignant melanoma cases was found. Five‐year net survival increased by 20 percent units during the study period. For squamous cell carcinoma, an increased age‐specific 5‐year net survival was observed for all age groups, apart for women aged ≥85.

Cervical neoplasia in relation to socioeconomic and demographic factors – a nationwide cohort study (2002–2018)

AbstractIntroductionCervical cancer is a major cause of mortality and morbidity. We aimed to estimate the association between sociodemographic factors and cervical neoplasia.Material and methodsIn this Swedish nationwide open cohort study, 4 120 557 women aged ≥15 years at baseline were included between January 1, 2002 and December 31, 2018. The two outcomes were cervical cancer and carcinoma in situ identified in the Swedish Cancer Register. Sociodemographic factors (age, education level, family income level, region of residency, country of origin) were the main predictors. Incidence rates per 10 000 person‐years were calculated. Cox regression was used to estimate hazard ratios. Sensitivity analyses were conducted, including parity, urogenital infections, alcohol‐ and drug‐use disorders, and chronic obstructive pulmonary disease (used as a proxy for tobacco abuse).ResultsIn 38.9 million person‐years of follow‐up, 5781 (incidence rate: 1.5, 95% confidence interval [CI] 1.4–1.5) and 62 249 (incidence rate 16.9, 95% CI 15.9–16.1) women were diagnosed with cervical cancer and carcinoma in situ, respectively. Women from Eastern Europe had a hazard ratio of 1.18 (95% CI 1.05–1.33) for cervical cancer compared with Swedish‐born women, while women from non‐Western regions were inversely associated with cervical cancer and carcinoma in situ. Women with a low education level had a hazard ratio of 1.37 (95% CI 1.29–1.45) for cervical cancer compared with women with a high education level.ConclusionsWomen from the Middle East and Africa living in Sweden seem to suffer less from cervical neoplasia, whereas women with low education and women from Eastern Europe seem to suffer more from cervical cancer.

Time Trends for Incidence and Net Survival of Cervical Cancer in Sweden 1960–2014—A Nationwide Population-Based Study

Abstract Background: The aim was to investigate time trends for incidence and long-term net survival in the morphologic subtypes and stages of cervical cancer in Sweden during the period 1960 to 2014. Methods: Women with invasive cervical cancer were identified through the Swedish Cancer Registry. Incidence and net survival were calculated according to morphology, age at diagnosis, and FIGO stage at diagnosis. Results: In total, 29,579 cases of invasive cervical cancer between 1960 and 2014 were included. The age-standardized incidence for squamous cell carcinoma (SCC) decreased until 2000; thereafter, the incidence rate stagnated, and a small increase was found in 2014. The incidence of adenocarcinoma continuously increased. The age-standardized 5-year net survival increased. However, decreasing net survival with increasing age was found. A higher stage at diagnosis showed a worse net survival. SCC and adenocarcinoma did not statistically differ as regards net survival in the last years of the study. Conclusions: Age-standardized 5-year net survival improved between 1960 and 2014. A positive trend for short- and long-term net survival was seen for women ages 18 to 64 years but long-term net survival for women ≥75 years decreased. In this study, age and FIGO stage at diagnosis were found to be important prognostic factors in determining net survival. The morphologies, SCC, and adenocarcinoma did not statistically differ as regards net survival in the last years of the study. Impact: This study demonstrates longitudinal data on cervical cancer in Sweden for over 50 years with sub analyses on morphology, age, and stage at diagnosis.