Artificial Intelligence in Oncology - Supporting scientific research
(Project manager Doctor N. van Trommel, CGOA/VUMC, initiated 3rd quarter 2017)
Project 1. Developing a new screening test for cervical cancer
Project 2. Developing screening test for ovarium carcinoma
Project 1. Developing a new screening test for cervical cancer
Half of Dutch women eligible for cervical cancer screening, do not attend the screening program. There are many reasons to not participate, varying from lack of time, and fear for the examination, psychological barriers or not seeing the importance. In this project, we examined ways to make this investigation easier and less invasive. Since collection of urine is easy and non-invasive, we focused on this medium. In this project we looked for biomarkers in urine both selective and sensitive enough to detect and/or exclude cervical cancer and its precursor lesions in urine. Parallel to this project, we collaborate with the TU Enschede to develop a lab-on-a-chip on which we intend to detect cervical cancer and its precursor lesions with our biomarkers and this technique.
In this project, we found:
Project 2. Developing screening test for ovarium carcinoma
In project 2, we showed that serum biomarker HE4, is useful in predicting whether an ovarian tumor is malignant or benign. About 50% of the referrals of women with an ovarian tumor turn out be incorrect after final pathology. This is a psychological burden for patients but also a financial burden for hospitals and society.
It is also a cost effective diagnostic tool. Circulating tumor DNA fragments can be useful to determine which tumors are ovarian cancer. Women with ovarian tumors are satisfied with their referral and treatment independent of the final diagnosis. Possible future developments include detection of hypermethylated DNA in urine or measurement of RNA in tumor educated platelets (TEPs). These different (molecular) biomarkers, can be added to ultrasound models and other patient variables to further optimize referral strategies and treatment.
Project 1. Developing a new screening test for cervical cancer
Cervical cancer is a rare disease of which precursor stages can be detected and treated. Women between the ages of thirty and sixty are invited to participate in the population screening. For this, the general practitioner will take a smear from the cervix. Unfortunately, four out of ten women do not respond to the call to participate in the population screening. Previous research has shown that most cases of cervical cancer occur in women who did not participate in the population screening. For this reason, this project is investigating whether cervical cancer and its precursor stages can be detected in urine by means of different disease markers (Human Papillomavirus, DNA methylation). In collaboration with the Technical University of Twente, we are working on the development of a urine chip "lab-on-a-chip". This could be a solution for women who would otherwise not participate in the population screening. In the future, this urine chip could also be used for population screening in developing countries where there is less good access to care.
In order to be able to research and develop this, urine, self-collected smear from the vagina and smear from the cervix are currently being collected from women with cervical cancer and precursor stages of cervical cancer. In addition, it is being investigated whether urine can also contribute to the detection of other gynecological cancers, such as endometrial cancer and ovarian cancer.
Project 2. Developing screening test for ovarium carcinoma
Ovarian cancer is a rare disease and in most cases it is detected at a late stage. In that case, the disease has a poor prognosis. Patients with early stage disease have a good prognosis. Patients with suspected ovarian cancer undergo staging surgery by a gynecologist-oncologist to determine the stage and thus the need for additional therapy. However, the current diagnostics consisting, which includes biomarkers, is not optimal, which means that it regularly occurs that patients with a benign ovarian tumor undergo surgery by a gynecologist-oncologist. This might lead to anxiety in the patients since they are referred because of an increased cancer risk and it also leads to increased health care costs as patients must again undergo several diagnostic tests. It also occurs that patients with ovarian cancer needs a second (staging) surgical procedure when they had their first surgical procedure in a peripheral center conducted by a general gynecologist. This is uncomfortable for the patients since they undergo two surgical procedures. Besides, it also leads to increased health care costs and it might lead to a delay in starting systemic therapy (such as chemotherapy) if indicated. Therefore, we are currently conducting a study in which we investigate the diagnostic value of other serum biomarkers in the differentiation between benign and malignant ovarian tumors. We look at, among other things, human epididymis protein 4 (HE4). The literature shows that HE4 is more specific than the most commonly used biomarker Cancer-Antigen 125 (CA125), which leads to fewer false-positive results. In addition, it has recently been shown that cancer can leave free DNA and RNA fragments in the blood. Various analyzes are used to determine whether these 'liquid biopsies' can be of value in the diagnosis of ovarian tumors. In addition to biomarkers in the blood, research is also being carried out into the value of structural change of DNA, so-called "DNA methylation", in urine and smears of patients with an ovarian tumor.