Early detection of head and neck cancer relapse by expression profiling of blood platelets and/or ctDNA analysis in plasma and oral rinses

Head and neck squamous cell carcinomas (HNSCC) arise in the mucosal lining of
the upper aerodigestive tract and are caused by smoking and alcohol consumption
or human papillomavirus (HPV) infection. Most HNSCC patients present with
advanced stage of disease. Tumors are treated by radiotherapy,
chemoradiotherapy, the combination of cisplatin with concomitant radiotherapy,
or surgery with or without postoperative (chemo)radiotherapy. Locoregional
recurrences occur in 30-40% of advanced stage patients and are difficult to
manage as these are mostly detected late. When detected early, re-irradiation
and particularly salvage surgery are curative treatment options, but at present
only a minority of patients qualify for salvage surgery. Most reliable methods
for diagnosing recurrent disease are FDG-PET and examination under general
anesthesia with biopsy, but both are unsuited for screening in routine. Other
imaging modalities are too insensitive. Biomarkers for early detection in body
fluids as a screening assay for early diagnosis of relapse would be a major
improvement in the early detection and management of recurrent disease. Liquid
biopsy may guide FDG-PET imaging and/or examination under general anesthesia,
and lead to earlier detection of recurrent disease and consequently improved
clinical management survival of patients. In this study we focus on detection
of circulating tumor DNA (ctDNA) in plasma and oral rinses as well as
tumor-educated platelets (TEPs).

Tumor DNA circulates in the bloodstream and can be detected in plasma by DNA
sequencing. Both copy number alteration (CNA) profiling as well as ultradeep
target enrichment sequencing for mutations have been employed in preliminary
studies. We have combined low coverage whole genome sequencing for copy number
changes and HPV presence, with high coverage target-enrichment sequencing for
mutations in a single assay. We have analyzed ctDNA of 40 HNSCC patients and 20
controls, and ctDNA was found in 78% of patients irrespective of HPV status,
and not in controls.

A second potential screenings assay is the detection of tumor-educated
platelets (TEPs). Blood platelets circulate through the body including the
tumor and take up exosomes from the tumor, which educates them to TEPs. It has
been shown that the RNA profiles of these TEPs can be used to detect the
presence of a tumor. In a small pilot study, we have shown that this is also
the case for head and neck squamous cell carcinoma (HNSCC). In this pre-study,
analyses of 45 samples from HNSCC patients showed that HNSCC could be detected
with an accuracy of 91%, and in more recent analyses this improved to 98%.

Accuracy of early detection of recurrent/metastatic disease in HNSCC patients
treated for cure, and the median and range of time intervals between relapse
and presence of ctDNA in plasma and oral rinses as well as TEPs.

This is a prospective observational cohort study. Tumor material, plasma and
oral rinses will be collected at baseline and during follow-up of enrolled
patients. In this study we will assess the accuracy of ctDNA detection in oral
rinses and blood and TEP profiles for early detection of relapsed disease.

Under general anesthesia 1-3 extra biopsies next to the standard biopsy will be
taken of primary tumor and relapse, and blood and an oral rinse will be
collected. For surgically treated patients we will also collect the persurgical
rinse fluid in addition. Risk and discomfort will be minimal. At all scheduled
follow-up visits (on average 10x) additional blood samples and oral rinses will
be collected, which is of low risk and limited burden. Adverse events from
extra biopsies or blood withdrawal are not expected. The tests are still
experimental, and positive findings have formally no scientific meaning.
However, the mutations in the tumor are known and when mutations disappear and
reappear in the blood, we will report this to the treating physician and
discuss it in the multidisciplinary team that might decide on examination under
general anesthesia. This will not influence the main endpoint of the study:
early detection of relapse. We will record changes in routine clinical
management.