Blood Test for Viral DNA Outperforms Scans in Spotting Returning Throat Cancer

A simple blood test that detects fragments of tumor-derived human papillomavirus DNA circulating in the bloodstream has shown strikingly better accuracy than conventional post-treatment imaging at catching the return of HPV-associated throat cancer, according to a new retrospective study from the University of North Carolina at Chapel Hill. The research, published in BMC Cancer, evaluated the NavDx assay, which measures circulating tumor HPV DNA (ctHPVDNA), in patients whose oropharyngeal squamous cell carcinoma had been treated with definitive surgery. In a cohort of sixteen patients, the blood-based surveillance approach identified every biopsy-proven recurrence with no false positives, while standard imaging missed half of the locoregional recurrences and raised false alarms in three patients who were ultimately found to be disease-free.

Human papillomavirus-positive oropharyngeal squamous cell carcinoma, a cancer of the tonsils and base of tongue that has become the dominant form of head and neck cancer in many Western countries, generally carries a favorable prognosis compared with its HPV-negative counterpart. Even so, between 10 and 25 percent of patients experience progressive disease within three years of treatment, making reliable surveillance a central concern for clinicians. The standard follow-up regimen typically involves contrast-enhanced CT of the neck and chest or PET/CT scanning, interpreted using the Neck Imaging Reporting and Data System, known as NI-RADS, a structured risk classification designed to standardize how radiologists flag suspicious findings. Yet imaging after surgery is complicated by anatomical distortion, scar tissue, and inflammation, all of which can mimic or obscure tumor recurrence.

The UNC team, led by Makayla R. Matthews of the School of Medicine together with colleagues in otolaryngology, radiology, and radiation oncology, reviewed the records of patients treated with definitive surgical resection between 2020 and 2023. All patients underwent standard post-treatment imaging, and ctHPVDNA samples were also collected and independently assessed. The cohort was characteristic of the modern HPV-positive throat cancer epidemic: 81 percent of the patients were male, the median age was 58, and a majority had never smoked. Most tumors arose in the tonsil, followed by the base of tongue, and nearly all patients presented with lymph node involvement, ranging from N1 to N3 disease under the seventh edition of the American Joint Committee on Cancer staging system.

During follow-up, four of the sixteen patients, or 25 percent of the cohort, developed biopsy-proven recurrence, with half of those events occurring at locoregional sites and half at distant locations. The median time to detectable recurrence was 14 months, though the range was wide, spanning from 3 to 42 months after treatment. This recurrence burden aligns with the upper end of what has been reported for HPV-positive oropharyngeal cancer populations and underscores why surveillance strategies matter: catching a returning tumor early, while it is still small and localized, can open the door to potentially curative salvage therapy.

The performance gap between the two surveillance modalities was the study’s most consequential finding. When radiologists applied NI-RADS criteria to post-treatment imaging, they flagged four suspected locoregional recurrences. Only one of these proved to be a true positive on biopsy; the other three were false positives. That translates to a sensitivity of 50 percent, a specificity of 75 percent, a positive predictive value of just 25 percent, and a negative predictive value of 90 percent for imaging-based detection. In practical terms, imaging missed as many recurrences as it caught, and three-quarters of the positive imaging findings did not correspond to actual cancer.

The ctHPVDNA assay told a different story. The blood test correctly identified two locoregional recurrence events and two distant recurrence events, all four of which were confirmed by biopsy, and it produced no false positives. That yields 100 percent sensitivity, 100 percent specificity, 100 percent positive predictive value, and 100 percent negative predictive value within this small cohort. The assay works on a principle analogous to liquid biopsies developed for other cancers: tumors driven by HPV shed fragments of viral DNA that carry tumor-tissue-modified signatures into the circulation, where they can be quantified using digital PCR. Because the viral signal originates from malignant cells rather than normal tissue, a positive result carries strong biological specificity for active disease.

The technical logic of the test helps explain its precision. HPV-positive tumor cells release short fragments of viral DNA, modified by the tumor tissue, into the plasma. Digital PCR platforms such as the one behind NavDx amplify and count these target sequences, producing a quantitative measure of circulating tumor burden. Unlike imaging, which depends on anatomical changes large enough to be visually resolved, the molecular assay can in principle register subclinical disease before structural changes appear on scans. The study’s finding that ctHPVDNA detected distant recurrences as well as locoregional ones is particularly notable, since distant metastases in the lungs or elsewhere are often the hardest events for routine neck imaging to catch early.

The clinical implications extend beyond diagnostic statistics. False-positive imaging findings in this study triggered additional workup and, in some cases, raised the specter of unwarranted salvage operations, procedures that carry real morbidity, heighten patient anxiety, and add substantial financial toxicity. The authors argue that a surveillance strategy anchored in ctHPVDNA could reduce the need for repeat imaging and prevent unnecessary interventions, reserving invasive procedures for patients whose molecular signal confirms genuine disease. Conversely, the high negative predictive value suggests that a clear blood test could offer patients and clinicians greater confidence that no recurrence is present, potentially streamlining follow-up schedules that currently rely on serial scanning.

Important caveats temper these results. The study was a single-institution retrospective review with only sixteen patients and four recurrence events, a level of evidence the authors themselves classify as level 4. Perfect sensitivity and specificity in such a small sample will almost certainly narrow in larger, prospective validations, and the assay’s performance in patients treated primarily with radiation rather than surgery remains to be established in this dataset. Questions also persist about optimal testing frequency, the quantitative thresholds that best distinguish clinically meaningful disease from transient signal, and how ctHPVDNA results should be integrated with imaging rather than simply replacing it, since anatomical localization remains essential for planning any salvage treatment.

Nevertheless, the findings add to a growing body of evidence that molecular surveillance of HPV-associated cancers can outperform anatomical imaging, echoing the way prostate-specific antigen transformed monitoring of prostate cancer decades ago. For a cancer population that is young, often curatively treated, and destined to live for decades under surveillance, the prospect of a blood test that reliably rules recurrence in or out carries enormous weight, both medically and psychologically. Larger prospective studies will determine whether ctHPVDNA surveillance can move from promising retrospective data to standard practice, but this early signal from surgical patients suggests that the era of watching for returning throat cancer with blood rather than scanners may be drawing closer.

Subject of Research: Circulating tumor HPV DNA surveillance for recurrence detection in surgically treated HPV-positive oropharyngeal squamous cell carcinoma

Article Title: NavDx (ctHPVDNA) recurrence detection in surgically treated human papillomavirus-associated oropharyngeal squamous cell carcinoma

Article References: Matthews, M. R., Van Dorsten, A. L., Heetderks, K. A., Yarbrough, W. G., Oldan, J. D., Huang, B. Y., Shen, C., & Hackman, T. G. (2026). NavDx (ctHPVDNA) recurrence detection in surgically treated human papillomavirus-associated oropharyngeal squamous cell carcinoma. BMC Cancer. https://doi.org/10.1186/s12885-026-17011-0

Image Credits: AI Generated

DOI: 10.1186/s12885-026-17011-0

Keywords: HPV-positive oropharyngeal cancer, circulating tumor HPV DNA, liquid biopsy, NavDx, cancer recurrence surveillance, NI-RADS imaging, head and neck cancer, digital PCR, salvage surgery, tumor tissue modified viral DNA, post-treatment monitoring, BMC Cancer