Circulating cell-free DNA (cfDNA) is fragmented DNA released by tissues; its size can range from 100 to 800 bp (1). cfDNA is of genomic, mitochondrial, or viral origin and is released through various cellular processes, including apoptosis, necrosis, and other cell death pathways, or vesicle release (2). The clinical role of cfDNA detection and analysis is evolving, and is expected to be used in several diseases, including prenatal testing, rheumatic diseases, organ transplantation, and cancer.

Because cancer tissues are characterized by rapid cancer cell proliferation and high rates of programmed cell death and necrosis due to impaired blood flow and hypoxic/acidic tumoral microenvironmental conditions, a significant proportion of the cfDNA in patients with cancer is of tumoral origin (named as circulating tumor DNA; ctDNA). Plasma cfDNA can be assessed using reverse transcription polymerase chain reaction (RT-PCR) or next-generation sequencing (NGS) techniques. Quantitative analysis of cfDNA and ctDNA, as well as analysis of methylation or gene mutations using NGS, provide important tools for cancer diagnosis and screening, and can also be used to develop prognostic and predictive biomarkers (3).

In healthy individuals, the cfDNA concentration in the blood ranges from 0-100 ng/ml (4,5). Concentration increases as the size of the primary tumor and the extent of lymph node involvement advances (6-8). Metastatic disease is also linked to higher cfDNA levels (9).

The eventual role of cfDNA quantification in predicting therapeutic efficacy has been suggested as early as 1977 (10). Surgical resection of the tumor and chemotherapy have been shown to reduce the cfDNA concentration (11,12). Several studies noted that the reduction of cfDNA in the plasma is associated with a better and sustained response to chemotherapy in gastric, pancreatic, and colorectal cancers (13,14). Conflicting results have been reported regarding the pre-chemotherapy prognostic value of cfDNA levels in NSCLC (15,16).

In the current report, we provide the results of a prospective study focusing on the role of plasma cfDNA concentration in the outcome of radiotherapy (RT) for squamous cell head neck cancer (SCHNC).

Ethical considerations. The study was approved by the local Ethics and Research Committee (ES1 23-01-2019 and ES2 22-02-2019) and performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Written informed consent to enter the trial was obtained from all patients before therapy. Patient consent included permission to anonymously publish their clinical and laboratory data for research and educational purposes.

Patients. Between 2020 and 2022, 47 patients with histologically diagnosed SCHNC treated with RT combined with chemotherapy were prospectively enrolled. Table I shows the patient and disease characteristics. The median follow-up of patients was 16 months (range=1-36 months). For patients alive at the last follow-up, this was 18 months (range=6-36 months).

Treatment details. Volumetric Modulated Arc Therapy (VMAT) was applied using a 6 MV ELEKTA Infinity™ Linear Accelerator (Elekta, Stockholm, Sweden) endowed with an Agility™ head (Elekta). Image-guided radiotherapy (IGRT), using daily cone-beam CT, was applied. All patients were treated with a simultaneous integrated boost (SIB) technique delivering 22 daily fractions within 30 days. For prophylactic neck irradiation, a daily dose of 2.15 Gy was applied, while the primary tumor received fractions of 2.6-2.7 Gy. Enlarged nodes received 2.3-2.5 Gy per fraction. The inversed plans were produced using Monaco TPS version 5.11.03 (Elekta CMS, Maryland Heights, MO, USA), and each planning target volume (PTV) received at least 95% of the prescribed dose to 98% of its volume.

Patients received concurrent chemo-RT (CRT) with cisplatin (35-40 mg/m² per week; 12 patients) without or with cetuximab (250 mg/m²/week; 28 patients).

Response evaluation and follow-up. Tumor response was assessed two months after RT completion with a computed tomography (CT-scan) or magnetic resonance imaging (MRI) as appropriate. CT evaluation was repeated at four months and every six months thereafter. We used the WHO criteria (17) for response assessment with in-house modifications as follows: Complete response (CR) was defined as a 95-100% reduction of 2D dimensions of all measurable lesions. Any residual scar measuring less than 5% of the initial tumor volume that does not progress for at least four months following response documentation was considered to indicate a complete response. Partial (PR) and minimal response (MR) refers to 50-95% and 25-49% reduction of tumor 2D dimensions, respectively. Tumor reductions between 0-24% that lasted at least two months were considered stable disease. All other cases were considered progressive disease (PgD), regardless of the initial response.

Plasma collection. Blood was collected in in vials containing ethylenediaminetetraacetic acid (EDTA) (BD Biosciences, Franklin Lakes, NJ, USA). Plasma and peripheral blood mononuclear cell (PBMCs) isolation was performed with density gradient centrifugation using a 1,077 g/ml synthetic epichlorohydrin sucrose polymer Lymphosep (Lymphocyte Separation Media-500 ml; density 1,077 g/ml; Cat no: LM-T1702/500; Biosera, Cholet, France), at a 1:1 ratio of the total volume. Centrifugation of blood samples was performed at 1,500 rpm (400 × g), for 30 min at 25˚C. The plasma layer was transferred in new sterile Eppendorf tubes, and then gradually frozen at –20˚C and then at –80˚C for long-term storage.

Plasma-cfDNA quantification. The MagMAX Cell-Free (Cat. no: Α36716; Thermo Fisher Scientific, Waltham, MA, USA) isolation kit, specifically designed for nucleic acid enrichment in liquid biopsies such as the plasma, was used to isolate total cell-free nucleic acids. The isolation was based on MagMAX magnetic bead technology to recover high-quality nucleic acids suitable for NGS. Isolation was performed using 4 ml of plasma from each patient, before and at the end of treatment. Quantification of cfDNA concentration was performed using the Qubit fluorometer and the Qubit 1X dsDNA HS (Cat. no: Q33231; Thermo Fisher Scientific), for specific and selective quantification of double-stranded DNA (dsDNA).

Statistical analysis. The GraphPad Prism 7.0 package (Boston, MA, USA) was used for statistical analysis and graph presentation. The Wilcoxon matched pairs non-parametric two-tailed test was used for testing relationships between paired continuous variables. The Mann–Whitney non-parametric test was used to compare non-paired continuous variables. Locoregional progression free survival (LRFS), distant metastasis-free survival (DMFS), and disease-specific overall survival (OS) curves were plotted using the method of Kaplan–Meier. A Cox-regression analysis model was used to assess the independent significance of variables in the treatment outcome. A p-value of less than 0.05 was considered statistically significant.

cfDNA levels. The plasma cfDNA levels before CRT ranged between 0.09-5.2 ng/μl [median 0.91, 95% confidence interval (95%CI)=1.06-1.75], while after CRT ranged between 0.18-19.40 (median 1.01, 95%CI=1.2-3.1) (Figure 1). The overall increase did not reach significance (p=0.09). The ratio between cfDNA levels after/before CRT (A/B-ratio) ranged from 0.08 to 9.2 (median 1.1; 95%CI=1.2-2.4). Twenty-six out of 47 patients (55.3%) had an A/B-ratio >1 (median 1.8, range=1.1-9.2, 95%CI=1.7-3.7), thus increased cfDNA after CRT (Figure 1A).

cfDNA levels and histopathological variables. Comparison of T4-stage vs. all other stages showed a significant increase in cfDNA levels after therapy (3.3 ng/μl vs. 1.3 ng/μl, respectively, p=0.007; Figure 1B). Analysis according to N-stage (N0 vs. N1 vs. N2 vs. N3) did not demonstrate any significant association (p>0.56) (data not shown). Regarding the histopathological grade, a marginally higher cfDNA A/B-ratio was noted after therapy in grade 3 tumors compared to grade 1,2 (median value 1.2 ng/μl vs. 0.68 ng/μl, respectively, p=0.08).

cfDNA and response to CRT. Patients with MR or PgD after therapy had significantly higher cfDNA levels before therapy compared to patients with PR or CR (mean value 2.7 ng/μl vs. 1.2 ng/μl, respectively, p=0.03; Figure 1C). No significant difference was noted for the cfDNA levels after therapy or for A/B-ratio (p=0.30 and 0.65, respectively).

cfDNA and survival. Kaplan–Meier survival curves showed no statistically significant impact of an A/B-ratio >1 on OS (p=0.12; Figure 2A). A significantly worse LRFS was recorded in patients with an A/B-ratio >1 (p=0.01; hazard ratio 3.5, 95%CI=1.2-9.7; Figure 2B). No association was noted for the DMFS (p=0.53; Figure 2C). A multivariate Cox-regression analysis model including the AB/ratio (<1 vs. >1), T-stage (1, 2, 3 vs. 4), N-stage (0, 1, 2 vs. 3), and histopathological grade (1, 2 vs. 3), showed that the A/B-ratio was the only independent prognostic variable predicting locoregional relapse (p=0.03, hazard ratio 3.9, 95%CI=1.2-13).

Taking the median cfDNA value as a cut-off point, survival analysis was performed for the levels of cfDNA before and after CRT. No association of cfDNA levels (either before or after therapy) with OS and LRFS was noted (Figure 3A-F). A non-significant trend for worse DMFS was observed in patients with higher than the median cfDNA levels before and after CRT.

cfDNA is readily detectable in the plasma of patients. Overall higher levels of cfDNA are present in cancer patients compared to healthy controls. In the current prospective trial, we found that the plasma cfDNA in patients with SCHNC ranged from 0.09 to 5.2 ng/μl. In our study, higher pretreatment levels were significantly associated with advanced T-stage, which is in accordance with a previous study by Muhanna et al. (8). This could be a result of increased tumor burden, frequently accompanying T4-stage tumors, which eventually is associated with increased release of fragmented DNA from a higher number of dying cancer cells, and, eventually, from necrotic tissue areas existing in large tumors.

In 2016, Mazurek et al. reported a study on 200 patients with HNC where plasma cfDNA was measured using a technique based on telomerase reverse transcriptase amplification (18). Patients with oropharyngeal cancer had significantly higher levels compared to other tumor types. Moreover, stage IV cases and lymph node involvement were linked with higher cfDNA plasma levels. The study, however, did not report an analysis of treatment outcomes. As the incidence of distant metastasis is overall low in SCHNCs, in the range of 10% (19), the association of high cfDNA levels with metastatic disease is interesting. Such a biomarker could be useful for identifying patients at a higher risk of metastasis. Our study found an association between cfDNA levels with DMFS and LRFS, though with no statistical significance. This discrepancy could be due to the short clinical follow-up, the small number of patients, and the low incidence of distant metastasis.

Monitoring the kinetics of cfDNA levels in patients with SCHNC undergoing CRT, we observed that levels increased at the end of therapy in 55.3% of patients. In a study by Verma et al., 24 patients with HNC were studied by monitoring the cfDNA levels before, during, and after CRT (20). The authors found decreasing cfDNA levels in responders three months after the end of therapy. However, increasing overall levels were noted in patients with persistent disease. In our study, patients with minimal response or progressive disease after CRT had more than 2-fold higher cfDNA levels before therapy than responders. We suggest that high cfDNA levels before treatment reflect radioresistant tumors with high proliferating activity and necrosis that lead to an intense release of DNA fragments in the peripheral blood. Kaplan-Meier analysis showed that the subgroup of patients with increasing cfDNA levels after CRT had a significantly worse LRFS. Although no similar analysis is available in the literature regarding patients with SCHNC, studies in patients with lung and hepatocellular cancer shoed better disease control rates in patients with low cfDNA after CRT (21,22). Whether this finding represents a high intrinsic radio-resistance or a rapid cancer proliferation activated during radiotherapy, both conditions leading to radiotherapy failure, are questions that require further investigation.

The role of the kinetics of cfDNA during RT in predicting locoregional recurrence is clinically relevant. As the assessment of response of SCHNC and other carcinomas after CRT is usually performed two months after the end of therapy, the use of cfDNA levels can be important to identify high-risk patients immediately after treatment. Such patients could be candidates for immediate consolidation chemotherapy or, even better, be recruited in immunotherapy protocols to enhance the eradication of the remnant disease by taking advantage of the postulated radio-vaccination effect of otherwise ineffective CRT (23,24).

The assessment of cfDNA levels in the plasma of patients with SCHNC before and after CRT provides useful information regarding the locoregional control and recurrence probability. A high after/before CRT cfDNA levels ratio indicates incomplete tumor eradication and poor LRFS. This information becomes available earlier than the routine CT-assessment, which is performed several weeks after treatment completion. Whether high levels of cfDNA before and after therapy are markers of high risk for the development of metastasis requires further investigation in larger cohorts of patients. cfDNA indexes may prove important in stratifying patients with SCHNC at a high risk of local and distant relapse into trials of consolidation chemo-immunotherapy.

There are no conflicts of interest to report in relation to this study.

Conceptualization, M.I.K., C.N.B.; methodology, I.K., E.X, N.K.; validation, S.P.F., I.M.K.; M.I.K., E.X.; investigation, M.I.K., I.M.K.; writing – original draft preparation, M.I.K., E.X.; writing – review and editing, I.M.K., S.P.F., N.K., I.K., C.N.B.; supervision, M.I.K., C.N.B.; All Authors have read and agreed to the published version of the manuscript.

This research has been co-financed by the European Regional Development Fund of the European Union and Greek National Funds through the Operational Program Competitiveness, Entrepreneurship, and Innovation, under the call RESEARCH–CREATE–INNOVATE (project code: T2EDK-03266, project acronym, and title: "BIOKAKETRA—Identification of genomic and transcriptomic prognostic bio-signatures in head and neck cancer").