RESUMO
BACKGROUND: AICyte has previously demonstrated a potential role in cervical cytology screening for reducing the workload by using a 50% negative cutoff value. The aim of the current study is to evaluate this hypothesis. METHODS: The authors used the Ruiqian WSI-2400 (with the registered trademark AICyte) to evaluate a collection of 163,848 original cervical cytology cases from 2018 to 2023 that were collected from four different hospital systems in China. A breakdown of cases included 46,060 from Shenzhen, 67,472 from Zhengzhou, 25,667 from Shijiazhuang, and 24,649 from Jinan. These collected cases were evaluated using the AICyte system, and the data collected were statistically compared with the original interpretative results. RESULTS: In 98.80% of all artificial intelligence cases that were designated as not needing further review, the corresponding original diagnosis was also determined to be negative. For any cases that were designated atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion or higher, the sensitivity and negative predictive value were 90.77% and 98.80%, respectively. The sensitivity and negative predictive value were greater in cases designated as low-grade squamous intraepithelial lesion or higher at 98.92% and 99.94%, respectively. Of the 49 low-grade squamous intraepithelial lesion or higher that were designed by AICyte as not needing further review, the cytohistologic correlation revealed eight cases of cervical intraepithelial neoplasia 1 and 18 negative cases; and the remaining cases were without histologic follow-up. In practice, AICyte used at a 50% negative cutoff value could reduce the anticipated workload if a protocol were implemented to label cases that qualified within the negative cutoff value as not needing further review, thereby finalizing the case as negative for intraepithelial lesions and malignancy. CONCLUSIONS: For pathologic practices that do not have cytotechnologists or in which the workflow is sought to be optimized, the artificial intelligence system AICyte alone to be an independent screening tool by using a 50% negative cutoff value, which is a potential assistive method for cervical cancer screening.
RESUMO
OBJECTIVES: To examine the associated risk of cervical intraepithelial neoplasm grade 3+ (CIN3+) lesions in patients with AGC and extensive human papillomavirus (HPV) genotyping. METHODS: Cases with atypical glandular cell (AGC) interpretation on a Papanicolaou (Pap) test were identified along with associated extensive HPV genotyping and histologic follow-up results. RESULTS: Within this cohort of 469,694 Pap tests, 0.4% were diagnosed as AGCs. In total, 1267 cases had concurrent high-risk HPV (hrHPV) genotyping, and 40.3% were hrHPV positive. The percentage of AGC cases with cervical CIN3+ on histologic follow-up was 52.2% when hrHPV was positive, whereas it was 4.9% with a negative hrHPV result. The top 5 hrHPV genotypes associated with cervical CIN3+ in this cohort were HPV16, HPV18, HPV58, HPV52, and HPV33. Indeed, 92.8% of the hrHPV-associated CIN3+ lesions identified in this cohort were positive for at least one of these HPV genotypes. The sensitivity of detecting cervical CIN3+ lesions was 85.6% with the top 5 hrHPV genotypes (HPV16/18/58/52/33) and only increased to 89.0% when the additional 12 genotypes were included. CONCLUSIONS: In patients with an AGC Pap, the risk of having a cervical CIN3+ lesion is greatly increased by positivity for hrHPV types 16, 18, 58, 52, and/or 33. Incorporating comprehensive HPV genotyping into AGC cytology allows for refined risk stratification and more tailored management strategies.
RESUMO
BACKGROUND: A cytologic diagnosis of atypical squamous cells, cannot exclude high-grade squamous lesion (ASC-H) poses a disproportionately high risk of cervical cancer development. The objective of this study was to analyze type-specific risks by mapping human papillomavirus (HPV) genotypes in ASC-H cytology. METHODS: In total, 1,048,581 Papanicolaou tests that had ASC-H cytology were retrieved. Concurrent HPV genotyping using proprietary multiplex real-time (MRT) and polymerase chain reaction (PCR) HPV tests and histologic follow-up findings were analyzed. RESULTS: Among 1678 patients who had ASC-H findings (0.16%), 1414 (84.3%) underwent concurrent HPV genotyping (MRT, 857; HPV PCR test, 557). The overall high-risk HPV (hrHPV)-positive rate was 84.4%. Of the 857 MRT cases, 63.9% were infected with a single hrHPV, and 24.4% had multiple genotypes. The most prevalent HPV types were HPV16/52/58/33/31. Lesions that were identified as cervical intraepithelial neoplasia 2 or worse (CIN2+) were detected in 498 of 906 cases (55.0%), including 81 cervical carcinomas (8.9%). The risk of CIN2+ for the composite group of HPV16/52/58/33/31-positive cases was 62.7%, representing 90.7% (264 of 291) of total CIN2+ lesions in ASC-H/hrHPV-positive cases by MRT. CIN2+ lesions were detected in 108 of 142 (76.1%) HPV16-positive and/or HPV18-positive women by the PCR the HPV test. Among 128 hrHPV-negative ASC-H cases by both methods, CIN2+ lesions were identified in 21 of 128 (16.4%), including five cervical carcinomas (3.9%). The sensitivity, specificity, positive predictive value, and negative predictive value for patients in the composite group with HPV16/52/58/33/31 were 88.0%, 40.8%, 62.7%, and 75.0%, respectively. CONCLUSIONS: Papanicolaou tests classified as ASC-H are associated with a high CIN2+ rate and warrant colposcopy, regardless of HPV status. The extent to which the risk-stratification provided by comprehensive HPV genotyping can inform the management of ASC-H cytology remains to be explored.
RESUMO
The role of artificial intelligence (AI) in pathology offers many exciting new possibilities for improving patient care. This study contributes to this development by identifying the viability of the AICyte assistive system for cervical screening, and investigating the utility of the system in assisting with workflow and diagnostic capability. In this study, a novel scanner was developed using a Ruiqian WSI-2400, trademarked AICyte assistive system, to create an AI-generated gallery of the most diagnostically relevant images, objects of interest (OOI), and provide categorical assessment, according to Bethesda category, for cervical ThinPrep Pap slides. For validation purposes, 2 pathologists reviewed OOIs from 32,451 cases of ThinPrep Paps independently, and their interpretations were correlated with the original ThinPrep interpretations (OTPI). The analysis was focused on the comparison of reporting rates, correlation between cytological results and histologic follow-up findings, and the assessment of independent AICyte screening utility. Pathologists using the AICyte system had a mean reading time of 55.14 seconds for the first 3000 cases trending down to 12.90 seconds in the last 6000 cases. Overall average reading time was 22.23 seconds per case compared with a manual reading time approximation of 180 seconds. Usage of AICyte compared with OTPI had similar sensitivity (97.89% vs 97.89%) and a statistically significant increase in specificity (16.19% vs 6.77%) for the detection of cervical intraepithelial neoplsia 2 and above lesions. When AICyte was run alone at a 50% negative cutoff value, it was able to read slides with a sensitivity of 99.30% and a specificity of 9.87%. When AICyte was run independently at this cutoff value, no sole case of high-grade squamous intraepithelial lesions/squamous cell carcinoma squamous lesion was missed. AICyte can provide a potential tool to help pathologists in both diagnostic capability and efficiency, which remained reliable compared with the baseline standard. Also unique for AICyte is the development of a negative cutoff value for which AICyte can categorize cases as "not needed for review" to triage cases and lower pathologist workload. This is the largest case number study that pathologists reviewed OOI with an AI-assistive system. The study demonstrates that AI-assistive system can be broadly applied for cervical cancer screening.