HPV E6/E7 mRNA fell after LEEP, but recurrence prediction did not improve
In women treated for cervical intraepithelial neoplasia, the biologic logic behind HPV E6/E7 mRNA testing is straightforward: if oncogenic viral activity persists, the risk of residual or recurrent disease should remain elevated. This retrospective cohort from the First People’s Hospital of Nanning tested that assumption after LEEP, following 91 patients with high-grade cervical lesions for 24 months. The central finding was blunt. Preoperative and postoperative HPV E6/E7 mRNA status did not predict recurrence after excision, even though postoperative positivity collapsed dramatically.
Three patients, or 3.30%, developed recurrence during follow-up. That small number alone shaped much of the statistical uncertainty, but the pattern was still disappointing for the biomarker. Recurrence occurred at nearly identical rates in women who were preoperatively E6/E7 mRNA positive and those who were negative, and postoperative status fared no better as a discriminator. In practical terms, the assay changed after surgery without tracking the clinical endpoint it was supposed to anticipate.
The biomarker shifted, yet the endpoint did not follow
Before surgery, 68.1% of patients were positive for E6/E7 mRNA. After LEEP, positivity fell to 1.1%. A separate paired analysis found that 66.7% of patients converted from positive to negative after excision, a highly significant change. On paper, that looks encouraging. It suggests the procedure removed tissue with active viral transcription in many patients and sharply reduced the measurable oncogenic signal.
The problem is that biomarker movement did not translate into useful risk stratification. Among patients with preoperative negativity, recurrence was 3.45%; among those with preoperative positivity, recurrence was 3.23%. Postoperative negativity was followed by recurrence in 5.26% of patients, while no recurrences were observed in the postoperative positive group, a counterintuitive pattern that likely reflects the very small number of events rather than a protective effect of persistent positivity. None of these comparisons reached statistical significance.
A one-step biomarker can look elegant in a laboratory assay and still fail at the bedside.
Why a biologically plausible marker can still miss recurrence
The appeal of E6/E7 mRNA lies in its specificity. Unlike tests that capture transient HPV exposure, E6/E7 transcripts reflect oncogene expression from high-risk viral types, which in theory should correspond more closely to biologically active lesions. In cervical carcinogenesis, E6 and E7 disrupt normal cell-cycle control and support persistence of dysplastic epithelium. That makes the assay attractive for follow-up after excision, where clinicians want to know whether oncogenic activity remains in the cervical field.
Yet post-LEEP recurrence is not a single-variable event. Residual microscopic disease, sampling variability, host immune response, the quality of excision, and the timing of testing all influence what the assay sees. A negative mRNA result may simply indicate that the sampled cervix no longer contains enough transcriptionally active virus to register, while recurrence may still arise later from unsampled residual lesions or new infection. A positive result may persist without immediately translating into histologic recurrence. The biology is real; the timing and tissue sampling are the weak links.
Sensitivity was the larger problem than specificity
The study’s diagnostic performance numbers explain why the test failed to gain clinical traction here. Preoperative E6/E7 mRNA testing showed a sensitivity of 66.67% and specificity of 31.82%. Postoperative testing showed a sensitivity of 0% and specificity of 97.30%. Those figures tell two very different stories. Before surgery, the assay missed too many future recurrences and produced too many false positives to be dependable as a gatekeeper. After surgery, it became highly specific but completely insensitive for recurrence detection in this cohort.
That tradeoff matters because follow-up after LEEP depends on identifying the small subset of patients who truly need closer surveillance or additional intervention. A test with poor sensitivity cannot safely rule out recurrence risk. A test with high specificity but zero sensitivity after surgery offers little help when the goal is to catch disease early enough to change management.
Margin-negative LEEP did not make E6/E7 mRNA more useful
All included patients had histologically negative margins after LEEP, which removed one of the most obvious confounders in post-treatment surveillance. Even in that cleaner setting, E6/E7 mRNA failed to separate recurrence from non-recurrence. Baseline age and body mass index also did not differ significantly between groups, and the recurrence rate remained low overall. The study therefore focused attention on the biomarker itself rather than on gross surgical failure or obvious demographic imbalance.
A three-patient recurrence signal is not enough to support confident biomarker adoption, especially when the confidence intervals were so wide they stretched from near-zero to infinity in the odds-ratio analysis. The postoperative OR estimates were statistically unstable, and the absence of significance should be interpreted with caution. Still, the direction of the results did not suggest hidden clinical utility waiting beneath a small sample size. The assay simply did not perform as a recurrence predictor in this cohort.
Follow-up care still needs a better molecular signal
The bedside implication is narrow but important: clinicians should not rely on HPV E6/E7 mRNA status alone to guide surveillance after LEEP for CIN. The assay can fall rapidly after excision and may still fail to identify the patients who later recur. ThinPrep cytology, HPV screening, colposcopy, and biopsy remain central to follow-up because they do not depend on a single molecular readout to capture disease return.
This study also exposes a familiar problem in gynecologic oncology follow-up: a biomarker can be mechanistically credible, analytically clean, and still clinically blunt. The search for a better predictor of residual or recurrent CIN after excision remains open, especially one that combines molecular specificity with a meaningful link to future histologic disease. Here, E6/E7 mRNA moved in the right direction biologically, but it did not solve the surveillance problem it was designed to address.