HPV cfDNA Demonstrates 95% Sensitivity for Detecting HPV-Driven Oropharyngeal Cancer and Predicts Recurrence Months Before Imaging
Key Insights
A prospective study of 58 OPC patients found HPV cfDNA digital PCR had 95% sensitivity and 95% specificity for detecting HPV-driven oropharyngeal cancer (search) at diagnosis.
Post-treatment HPV cfDNA monitoring detected recurrence 3 to 8 months earlier than imaging, with a positive predictive value of 75% and negative predictive value of 89%.
HPV-negative OPC patients had significantly worse outcomes, with 35% mortality and 43% recurrence rates compared to 8.8% recurrence in HPV-driven cases.
A prospective observational study from the University Hospital of Zurich (search) has demonstrated that circulating HPV cell-free DNA (search) (cfDNA) detected via digital PCR can identify HPV-driven oropharyngeal cancer (search) (OPC) with high accuracy at diagnosis and may predict disease recurrence months before conventional imaging. The findings, published in Nature, add to growing evidence supporting liquid biopsy approaches for head and neck cancer surveillance.
The study enrolled 58 OPC patients and one cancer of unknown primary (CUP) patient between March 2020 and May 2023. Tumor HPV status was rigorously determined using a three-biomarker algorithm combining p16 immunohistochemistry, HPV DNA testing, and HPV multiplex serology. Of the cohort, 39 patients had HPV-driven tumors and 20 were HPV-negative.
Diagnostic Performance at Baseline
Using a multiplex digital PCR assay covering eight high-risk HPV types (HPV16 (search), 18, 31, 33, 35, 45, 52, 58), HPV cfDNA was detected in 36 of 38 evaluable HPV-OPC patients at diagnosis, yielding a sensitivity of 95% (95% CI 82–99%). Among 20 HPV-negative OPC patients, 19 tested negative, resulting in specificity of 95% (95% CI 75–100%).
The two false-negative cases both involved stage I disease with minor lymph node involvement (N1). The single false-positive result — an HPV-negative patient with 8.5 copies/ml of HPV16 (search) cfDNA — was negative for both p16 IHC and HPV16 E6 antibodies, suggesting the tumor may have harbored transcriptionally inactive HPV DNA.
HPV cfDNA concentrations at diagnosis ranged from 3 to 104,009 copies/ml (median 560 copies/ml) for HPV16 (search)-driven tumors, and from 7 to 226 copies/ml (median 26 copies/ml) for other HPV types. HPV-type specific sensitivity was 97% for HPV16 (33/34, 95% CI 85–100%).
Post-Treatment Surveillance and Recurrence Detection
Twenty-four HPV-OPC patients with positive baseline HPV cfDNA were followed for a median of 1.5 years, with 91 plasma samples collected across a median of three follow-up visits per patient. Four patients experienced recurrence or persistence during follow-up.
In two patients, HPV cfDNA became detectable again prior to clinical recurrence. One patient showed HPV16 (search) cfDNA rising to 82 copies/ml three months before recurrence was diagnosed, later increasing to 494 copies/ml. A second patient had detectable HPV cfDNA (4 copies/ml) approximately eight months before clinical diagnosis of pulmonary metastases.
On a per-test basis across 59 post-treatment samples, HPV cfDNA detection had a positive predictive value (PPV) of 75% (3/4, 95% CI 19–99%) for recurrence within 12 months, and a negative predictive value (NPV) of 89% (49/55, 95% CI 78–96%). Specificity was 98% (49/50, 95% CI 89–100%).
One patient with locoregional recurrence had negative HPV cfDNA at the last blood draw seven months prior, and one patient with persistent disease confirmed by imaging and biopsy also remained HPV cfDNA-negative after treatment.
HPV cfDNA Clearance Kinetics
Patients treated with surgery alone cleared HPV cfDNA more rapidly, with all five surgically treated patients testing negative at the first follow-up visit (mean 37 days post-diagnosis). Among 15 patients treated with radiotherapy or chemoradiation, the last positive sample occurred at a mean of 34 days after diagnosis, with the first negative result at a mean of 68 days. In some cases, HPV cfDNA levels initially increased after treatment initiation before eventual clearance.
Clinical Outcomes by HPV Status
The study confirmed stark prognostic differences between HPV-driven and HPV-negative OPC. All seven deaths during follow-up occurred in the HPV-negative group (35%, 95% CI 15–59%), with deaths at a median of 7 months after diagnosis. Recurrence was more common in HPV-negative patients (43%, 95% CI 18–71%) compared to HPV-driven cases (8.8%, 95% CI 2–24%).
HPV-negative patients had higher rates of alcohol consumption (55% vs. 13%) and smoking (90% vs. 49%), and more than 90% presented with advanced stage III or IV disease, compared to 26% of HPV-OPC patients.
Study Strengths and Limitations
The study benefited from close patient monitoring with higher-frequency follow-up than routine aftercare, and tumor HPV status was determined using three established biomarkers to mitigate single-biomarker limitations. The dPCR assay covered eight high-risk HPV types plus a beta-globin reference gene.
However, the authors acknowledge several limitations: the cohort was relatively small at 59 patients, and follow-up samples were unavailable for 14 of 39 HPV-OPC patients — a dropout partly attributable to the COVID-19 pandemic that may have introduced selection bias. With only four recurrence events, performance metrics could not be precisely determined. The researchers emphasize that larger studies with more participants and longer follow-up are needed to confirm these findings.
"Evaluating its clinical value in terms of potentially improved patient outcomes requires clinical trials that compare follow-up protocols incorporating HPV cfDNA testing with the current standard of care," the authors conclude.