Histological 'progression' from low (LSIL) to high (HSIL) squamous intraepithelial lesion is an uncommon event and an indication for quality assurance review

Abstract

An accurate assessment of 'progression' from a low (LSIL) to high (HSIL) grade squamous intraepithelial lesion (cervical intraepithelial neoplasia (CIN)2 or CIN3) of the cervix is critical to ascertaining HSIL outcome risk, the value of predictive biomarkers, and the need for excisional therapy. We obtained biopsy outcome data on a series of initially diagnosed LSIL to assess this risk. Consecutive biopsy diagnoses of LSIL were obtained from the archives, and the frequency of HSIL biopsy outcomes were ascertained by record and histological review. Then, a 'numerical severity score' was recorded for each diagnosis: LSIL (1-2), CIN2 (3-4) and CIN3 (5-6) with lower and higher values corresponding to the degree (low vs high) of histological severity within each category, respectively. Of 264 initial LSILs, 29 (11%) were reported with an HSIL outcome. However, histological review of 21 of these HSILs confirmed only 8 (38%) HSIL diagnoses by review with the numerical severity score: three cases scored as 5, three cases scored as 4, and two cases scored as 3; the remaining 13 cases were assigned a numerical severity score of 1 or 2. P16 immunostains of corresponding previous and subsequent biopsies were discordant in 4 of 12 cases (33%). In a blind review of a randomly selected series of HSILs from the same practice, HSIL was significantly more likely to be confirmed on re-review (10 of 13 (77%), P=0.024). These findings show that confirmed HSIL outcomes (on review) following an LSIL biopsy are infrequent ( approximately 3%). A diagnosis of HSIL following an LSIL should always be reviewed, as this diagnostic pairing may more likely be associated with a diagnostic error.