Involved Margins, High-Risk HPV Post-Treatment Predict Treatment Failure in Cervical Cancer

Results from a systemic review and meta-analysis of data collected from nearly 45,000 women treated for cervical precancer found that high-risk HPV posttreatment was a more accurate predictor of treatment failure than incomplete rescission.

The failure rate of excisional treatment, defined as persistent or recurrent cervical intraepithelial neoplasia of grade 2 or worse (CIN2+), is estimated to be 4% to 18%. Most of these cases occur within 2 years after primary treatment.

In 24 studies with at least 18 months of follow-up, 6.6% (95% CI, 4.9-8.4) of women had residual or recurrent CIN2+. Overall, the risk for residual or recurrent CIN2+ post-treatment was 17.1% (95% CI, 12.7-22.1) for women with positive margins. The pretest—post-test probability plots show that positive resection margins are associated with an average risk of post-treatment CIN2+ not reaching 20% and that negative resection margins are associated with post-treatment CIN2+ risk exceeding 2%.

However, investigators found that the presence of high-risk HPV post-treatment was associated with treatment failure rate of 28.4%. The treatment failure rate for patients who were negative for high-risk HPV was just 0.8%.

“This meta-analysis confirms that the risk of residual or recurrent CIN2+ is significantly increased with positive excision margins compared with negative excision margins,” Marc Arbyn, MD, Unit of Cancer Epidemiology, Belgian Cancer Centre, Scientific Institute of Public Health, and co-investigators wrote. “However, high-risk HPV post-treatment predicts treatment failure more accurately than margin status. Combined results of the margin and post-treatment HPV status could be used to stratify risk and diversify management.”

Investigators reviewed nearly 100 studies collected in the Pubmed MEDLINE, Embase, and CENTRAL databases from January 1, 1975 to February 1, 2016. Data from 44,446 women were included in this analysis.

The overall proportion of incomplete excisions was 23.1% (95% CI, 20.4-25.9) and was highly variable (range, 2.8%-59.5%; I² = 97.7%; P heterogeneity <.0001). Large loop excision of the transformation zone was associated with the highest proportion of incomplete excisions (25.9%; 95% CI, 22.3-29.6), followed by cold-knife conization (20.2%; 95% CI, 14.3-26.7), and laser conization (17.8%; 95% CI, 12.9-23.2).

The risk for CIN2+ in women with clear margins was 3.7% (95% CI, 2.5-5.1) and investigators found no significant differences by treatment procedure. Relative risk for CIN2+ was 4.8 (95% CI, 3.2-7.2; P <.001) for women with involved versus clear margins.

Failure rates were heterogeneous (range, 1.4%-18.3%; I2 >90%; P <.0001) and varied by treatment procedure, from roughly 2% for cold-knife conization and laser conization to almost 7% for large loop excision of the transformation zone. Treated CIN3+ lesions were not more prone to therapeutic failure than treated CIN2+ lesions (P = .94).

In pooled data from studies in which women were treated for histologically confirmed CIN2+, the sensitivity of the margin status to predict residual or recurrent CIN2+ was 55.8% (95% CI, 45.8-65.5) and specificity was 84.4% (95% CI, 79.5-88.4).

High-risk HPV testing, performed in 18 of the 25 studies, showed a pooled sensitivity of 91.0% (95% CI, 82.3-95.5) and a specificity of 83.8% (95% CI, 77.7-88.7). Investigators concluded that margin status was 38% less sensitive (sensitivity ratio, 0.62; 95% CI, 0.53-0.72) but equally as specific (specificity ratio, 1.01; 95% CI, 0.97-1.06) as post-treatment high-risk HPV testing to predict residual or recurrent CIN2+.

Researchers in 5 studies evaluated accuracy data for the combination of the margin status and post- treatment HPV testing. The sensitivity of the 2 combined tests was 99.1% (95% CI, 94.7-100) and specificity was 57.6% (95% CI, 47.4-67.5). The combination was no more sensitive (ratio, 1.04; 95% CI, 0.97-1.11) and significantly less specific (ratio, 0.75; 95% CI, 0.67-0.84) than HPV testing alone.

Arbyn M, Redman CWE, Verdoodt F, et al. Incomplete excision of cervical precancer as a predictor of treatment failure: a systematic review and meta-analysis. Lancet Oncol [published online November 7, 2017]. doi: 10.1016/S1470-2045(17)30700-3.