RESUMO
OBJECTIVE: Artificial intelligence (AI) could automatedly detect abnormalities in digital cytological images, however, the effect in cervical cancer screening is inconclusive. We aim to evaluate the performance of AI-assisted cytology for the detection of histologically cervical intraepithelial lesions (CIN) or cancer. METHODS: We trained a supervised deep learning algorithm based on 188,542 digital cytological images. Between Mar 13, 2017, and Oct 20, 2018, 2145 referral women from organized screening were enrolled in a multicenter, clinical-based, observational study. Cervical specimen was sampled to generate two liquid-based slides: one random slide was allocated to AI-assisted reading, and the other to manual reading conducted by skilled cytologists from senior hospital and cytology doctors from primary hospitals. HPV testing and colposcopy-directed biopsy was performed, and histological result was regarded as reference. We calculated the relative sensitivity and relative specificity of AI-assisted reading compared to manual reading for CIN2+. This trial was registered, number ChiCTR2000034131. RESULTS: In the referral population, AI-assisted reading detected 92.6% of CIN 2 and 96.1% of CIN 3+, significantly higher than or similar to manual reading. AI-assisted reading had equivalent sensitivity (relative sensitivity 1.01, 95%CI, 0.97-1.05) and higher specificity (relative specificity 1.26, 1.20-1.32) compared to skilled cytologists; whereas higher sensitivity (1.12, 1.05-1.20) and specificity (1.36, 1.25-1.48) compared to cytology doctors. In HPV-positive women, AI-assisted reading improved specificity for CIN1 or less at no expense of reduction of sensitivity compared to manual reading. CONCLUSIONS: AI-assisted cytology may contribute to the primary cytology screening or triage. Further studies are needed in general population.
Assuntos
Colo do Útero/patologia , Aprendizado Profundo , Processamento de Imagem Assistida por Computador/métodos , Infecções por Papillomavirus/diagnóstico , Displasia do Colo do Útero/diagnóstico , Neoplasias do Colo do Útero/diagnóstico , Adulto , Biópsia , Colposcopia , Conjuntos de Dados como Assunto , Detecção Precoce de Câncer/métodos , Feminino , Humanos , Pessoa de Meia-Idade , Invasividade Neoplásica , Papillomaviridae/isolamento & purificação , Infecções por Papillomavirus/patologia , Infecções por Papillomavirus/virologia , Valor Preditivo dos Testes , Índice de Gravidade de Doença , Triagem/métodos , Neoplasias do Colo do Útero/patologia , Esfregaço Vaginal , Displasia do Colo do Útero/patologia , Displasia do Colo do Útero/virologiaRESUMO
BACKGROUND: This study aimed to provide more information for cancer prevention strategies by determining the distribution of human papilloma virus (HPV) genotype prevalence in invasive cervical carcinoma (ICC) and precancerous lesion patients in the Yangtze River Delta area in China. METHODS: This multi-centre descriptive cross-sectional study involves four university hospitals in the Jiangzhehu area. Women with histologically confirmed cervical intraepithelial neoplasia (CIN) 1, CIN2, CIN3 or ICC who were diagnosed and treated in the four selected hospitals between February 2012 and April 2014 were eligible for recruitment. The average age of the patients was 40.93 ± 11.87 years old, among whom the youngest was 17 years old and the oldest was 76 years old.Those with immunodeficiency diseases or a previous history of cancer or CIN were excluded. HPV genotyping was performed by a central laboratory. The distribution and age and disease specificity of the HPV genotype prevalence were analysed. RESULTS: Of the 2181 collected samples, 251 were ICC and 1930 were CIN. The mean age of cervical cancer and CIN patients was 40.93 ± 11.8 years (range, 17-76 years). The five most commonly identified HPV types in each lesion class were as follows: CIN1: 52, 58, 16, 33, and CP; CIN2: 16, 58, 52, 33, and 31; CIN3: 16, 58, 33, 52, and 31; and ICC: 16, 58, 18, 52, and 33. CIN1 had an earlier age of onset (30-40 years) than CIN2, CIN3, and cervical cancer. The age of onset of cervical cancer exhibited two peaks at 40-44 and 50-54 years of age. In all infected patients, the frequency of HPV infection with a single type was 62.9%, and with multiple types, it was 38.1%. There was no difference in the frequencies of multiple types amongst the different cervical lesions. CONCLUSIONS: The most prevalent genotypes in the investigated area (52, 58, 16 and 18) justify the necessity of anti-HPV vaccination in teenagers and young girls under 24 years old in the Yangtze River Delta area in China. Infection with multiple high-risk HPV types versus single infection does not increase the risk for ≥ CIN2 in ICC development.
Assuntos
Alphapapillomavirus/classificação , Alphapapillomavirus/genética , Genótipo , Infecções por Papillomavirus/complicações , Infecções por Papillomavirus/virologia , Lesões Pré-Cancerosas , Neoplasias do Colo do Útero/etiologia , Neoplasias do Colo do Útero/patologia , Adolescente , Adulto , Idoso , China/epidemiologia , Feminino , Humanos , Pessoa de Meia-Idade , Estadiamento de Neoplasias , Neoplasias do Colo do Útero/epidemiologia , Adulto Jovem , Displasia do Colo do Útero/epidemiologia , Displasia do Colo do Útero/etiologia , Displasia do Colo do Útero/patologiaRESUMO
OBJECTIVE: To describe the natural history of cervical intraepithelial neoplasia (CIN) I and the biologic factors associated with the progression of CIN I and to analyze the predictive values of p16(INK4a) protein for the progression of CIN I. METHODS: From August 2010 to July 2013, 104 patients referred for abnormal cytology [≤ low-grade squamous intraepithelial lesion (LSIL); including negative for intraepithelial lesion or malignancy (NILM), atypical squamous cells of undetermined significance (ASCUS), LSIL] and high-risk (HR) HPV positive, and were diagnosed CIN I by colposcopy-assisted biopsy and followed at 1-year intervals in the First Affiliated Hospital of Nanjing Medical University. In order to analyze the relationship between the progression of CIN I with clinical biologic factors, including patient age, cervical cytology before colposcopy, loads of HR HPV, HPV L1 capsid protein, p16(INK4a) protein, χ(2) tests was used to compare the different frequencies of factors in groups of progressed and persisted/regressed CIN I, then five factors with progressed CIN I were processed into binary logistic regression analysis. RESULTS: (1) In the first year of follow-up, among 104 patients (including 15 cases NILM, 78 cases ASCUS, 11 cases LSIL), 52 cases of them were NILM and HR HPV negative, 30 cases were negative for intraepithelial lesion, 10 cases were CIN I, 5 cases were CIN II and 7 cases were CIN III. In total, 82 cases (78.8%, 82/104) cases had regressed, 10 cases (9.6%, 10/104) persisted, 12 cases (11.5%, 12/104) progressed [including 5 cases (4.8% , 5/104) progressed to CIN II, 7 cases (6.7% , 7/104) progressed to CIN III, none progressed to invasive cancer]. (2) All patients, 88 cases of them accepted immunohistochemical detection the expression of p16(INK4a) protein. The result shown that 30 cases (34%, 30/88) were positive and 58 cases (66%, 58/88) were negative. And 94 cases accepted immunocytochemical detection the expression of HPV L1 capsid protein, 49 cases (52% , 49/94) were positive and 45 cases (48% , 45/94) were negative. (3) Univariate analysis showed that age of the patient, loads of HR HPV, cervical cytology before colposcopy and the expression of HPV L1 capsid protein were not risk factors of the progression of CIN I (all P>0.05) except for the expression of p16(INK4a) protein (P<0.05). Multivariable analysis found that p16(INK4a) protein positive was associated with progression of CIN I (OR=5.1, 95%CI: 1.162-22.387, P=0.031). (4) Thirty-one cases were p16(INK4a) protein positive, 8 cases (27%, 8/30) of them progressed, while 4 cases (7%, 4/58) of 58 cases with p16(INK4a) protein negative progressed,in which there were significant difference (P<0.05). The sensitivity was 75%, the specificity was 71%, the positive predictive value was 27% and the negative predictive value was 93% for progression to CIN II-III of p16(INK4a) protein staining. CONCLUSIONS: The progression rate of CIN I with abnormal cytology (≤LSIL) and HR HPV positive was lower, and there was no progression to invasion at 1-year intervals. Immunostaining of p16(INK4a) protein as the risk factors of CIN I progression could have a role in prediction of CIN I and the management of high-risk CIN I.
Assuntos
Biomarcadores Tumorais/análise , Inibidor p16 de Quinase Dependente de Ciclina/metabolismo , Progressão da Doença , Displasia do Colo do Útero/patologia , Neoplasias do Colo do Útero/patologia , Biomarcadores , Biópsia , Proteínas do Capsídeo/metabolismo , Colposcopia , Citodiagnóstico , Feminino , Humanos , Gravidez , Sensibilidade e Especificidade , Neoplasias do Colo do Útero/metabolismo , Esfregaço Vaginal , Displasia do Colo do Útero/metabolismoRESUMO
INTRODUCTION: High-risk human papilloma virus (HPV)-associated cervical cancer is the fourth most common cancer in women worldwide. Current treatments of high-grade squamous intraepithelial lesion (HSIL) of the cervix are based on invasive surgical interventions, compromising cervical competence and functionality. APRICITY is a multicentre, prospective, double-blind, randomised controlled phase 3 study further evaluating the efficacy and safety of Cevira, an integrated drug-delivery and light-delivery device for hexaminolevulinate photodynamic therapy, which shows promise as a novel, non-invasive outpatient therapy for women with HSIL. METHODS AND ANALYSIS: Patients with biopsy-confirmed HSIL histology are invited to participate in the study planned to be conducted at 47 sites in China and 25 sites in Ukraine, Russia and the European Union. The aim is to include at least 384 patients, which will be randomised to either Cevira or placebo group (2:1). All patients will be assessed 3 months after first treatment and a second treatment will be administered in patients who are HPV positive or have at least low-grade squamous intraepithelial lesion. Primary endpoint is the proportion of the responders 6 months after first treatment. Secondary efficacy and safety endpoints will be assessed at 6 months, and data for secondary performance endpoints of the Cevira device will be collected at 3 months and 6 months, in case second treatment was administered. All patients in the Cevira group will be enrolled in an open, long-term extension study for 6 months to collect additional efficacy and safety data (study extension endpoints). ETHICS AND DISSEMINATION: The study was approved by the ethics committee of the Peking Union Medical College Hospital and Hannover Medical University, Germany. Findings will be disseminated through peer review publications and conference presentations. TRIAL REGISTRATION NUMBER: NCT04484415; clinicaltrials.gov.
Assuntos
Carcinoma in Situ , Infecções por Papillomavirus , Fotoquimioterapia , Lesões Intraepiteliais Escamosas , Neoplasias do Colo do Útero , Ácido Aminolevulínico/análogos & derivados , Ensaios Clínicos Fase III como Assunto , Feminino , Humanos , Estudos Multicêntricos como Assunto , Infecções por Papillomavirus/complicações , Fotoquimioterapia/métodos , Estudos Prospectivos , Ensaios Clínicos Controlados Aleatórios como Assunto , Neoplasias do Colo do Útero/patologiaRESUMO
Background: This retrospective multi-center study aimed to describe the epidemiological characteristics, clinical features, and management of patients with cervical cancer in pregnancy (CCIP) and evaluate maternal and infant outcomes. Methods: The data of patients with CCIP were retrospectively collected from those diagnosed and treated in 17 hospitals in 12 provinces in China between January 2009 and November 2017. The information retrieved included patients' age, clinical features of the tumor, medical management (during pregnancy or postpartum), obstetrical indicators (i.e., gestational age at diagnosis, delivery mode, and birth weight), and maternal and neonatal outcomes. Survival analyses were performed using Kaplan-Meier survival curves and log-rank tests that estimated the overall survival of patients. Results: One-hundred and five women diagnosed with CCIP (median age = 35 years) were identified from ~45,600 cervical cancer patients (0.23%) and 525,000 pregnant women (0.020%). The median gestational age at cancer diagnosis was 20.0 weeks. The clinical-stage of 93.3% of the patients with CCIP was IB1, 81.9% visited the clinic because of vaginal bleeding during pregnancy, and 72.4% had not been screened for cervical cancer in more than 5 years. To analyze cancer treatments during pregnancy, patients were grouped into two groups, termination of pregnancy (TOP, n = 67) and continuation of pregnancy (COP, n = 38). Analyses suggested that the TOP group was more likely to be diagnosed at an earlier gestational stage than the COP group (14.8 vs. 30.8 weeks, p < 0.001). The unadjusted hazard ratio for the COP group's overall survival was 1.063 times that of the TOP group (95% confidence interval = 0.24, 4.71). There were no significant differences between the TOP and COP groups in maternal survival (p = 0.964). Thirty-three of the infants of patients with CCIP were healthy at the end of the follow-up period, with a median age of 18 ± 2.8 months. Conclusions: Most patients with CCIP had not been screened for cervical cancer in over 5 years. The oncologic outcomes of the TOP and COP groups were similar. A platinum-based neoadjuvant chemotherapy regimen could be a favorable choice for the management of CCIP during the second and third trimesters of pregnancy.
RESUMO
OBJECTIVE: To compare the specificity and sensitivity of two genotyping approaches for human papillomavirus (HPV). METHOD: HPV DNA was amplified and detected in clinical specimens by polymerase chain reaction in a pair of universal primers MY09/11, and then genotyped with either sequencing method or liquid chip hybridization method (luminex method). RESULT: Sequencing method obtained precise genotyping results in single-type HPV infection, while luminex method obtained accurate genotyping results in multiple-type HPV infection. CONCLUSION: A combined method using both sequencing and luminex method is suitable for the genotyping of HPV-infected specimens.