Interpretation
Current recommended primary cervical screening methods are hrHPV tests, cervical cytology or co-testing. However, high sensitivity of hrHPV tests may cause patients’ psychological burden, over-referral to colposcopy examination and over-treatment for HPV infection. Cervical cytology assessment has a lower sensitivity and requires qualified cytopathologists. In many rural areas in China, cytopathologists are lacked [24]. Besides, women usually need to wait for several days or longer to be informed of their results. Since COVID-19 broke out in 2020, routine cervical screening has become a challenge for all women and gynecologists, as both HPV test and cervical cytology examination could not avoid specimen contact and increase the risk of COVID-19 exposure. Until today, in COVID-19 post-pandemic, the same concern still exists. Therefore, a simple, non-invasive and immediate screening method is warrant.
TS detection requires no cytopathologists because of the easiness of operation and objective results. In previous studies, TS has demonstrated a promising diagnostic efficacy. A mate-analysis reported the pooled sensitivity, specificity and AUC of TS was 76%, 69% and 0.7859, respectively [20]. Compared with it, TS in our study had a relatively lower sensitivity (53.1% for CIN1+, 65.7% for CIN2+ and 67.3% for CIN3+) but a higher specificity detecting CIN1+ (77.1%) and similar specificity detecting CIN2+ (66.7%) and CIN3+ (62.7%). A recent study demonstrated similar diagnostic value of TS applied in patients with abnormal Pap smear results (the sensitivity and specificity were 65% and 55%, respectively) [19].
Among 17 ASC-H patients enrolled, 2 were TS negative. One was pathologically confirmed CIN1, the other one was CIN3. The misdiagnose of CIN3 by TS was TZ type I and TS-examined at outpatient department. Colposcopically-directed three biopsies (1, 6, 11 o’clock) were obtained. Only 1 o’clock was pathologically confirmed CIN3 while the other two reported no lesion. We inferred that the tip of the device did not cover the lesion since the cervix was not located in the middle and leaning to the right, leading to the 1 o’clock area not exposed satisfactorily.
For the missed cases of CIN2+ by TS in our study, 24 of 36 cases were with cervical TZ type II and III (Figure 2 A3-D3). No cancer was missed by TS. We deduced the relatively low sensitivity of TS was because of the undetectable cervical canal and endocervix. Therefore, we defined TZ type II and III as incomplete cervical TZ type because of the unseen SCJ to compare with TZ type I. We unexpectedly found that TS had a better diagnostic performance in incomplete cervical TZ type. Higher sensitivities and significantly higher specificities and NPVs were observed in this group. This might be attributed to the systematic bias, as the TS operator would put the tip of device as inwardly into the cervical canal as possible when the SCJ unexposed naturally, aiming not to miss any lesion. Currently, women with incomplete cervical TZ type are recommended to undergo ECC which may cause discomfort for patients as well as operating difficulty for colposcopists. Our result indicated that TS could reduce misdiagnosis of CIN and predict no CIN occurred effectively, thus might decrease the frequency of ECC.
A recent study about TS applied in hrHPV infected women reported TS combined with HPV 16/18 had the highest specificity (83.6%) comparing with TS alone or HPV 16/18 alone [18]. Our study reported similar specificities of TS combined with hrHPV of significancy (82.8% for CIN1+, 73.3% for CIN2+) and significantly higher PPV of CIN1+ comparing with hrHPV alone. This demonstrated that combination of TS and hrHPV test could reduce misdiagnosis of CIN1+ and predict CIN1+ effectively comparing with hrHPV alone.
Post-menopausal women usually present with cervical TZ type III. To analyze the influence of menopause on TS detection, we compare it between non-menopausal and post-menopausal women with cervical TZ type III. As a result, sensitivities of non-menopausal women detecting CIN1+ and CIN3+ were significantly higher comparing with post-menopausal women. This suggested that TS was more effective in non-menopausal women.
In current COVID-19 post-pandemic context in China, we are eager to find an effective screening triage method to lower the rate of colposcopy. In our study, 89.2% and 80.3% women with negative TS result in ASCUS group and LSIL group, respectively, were pathologically confirmed ≤CIN1. To decrease COVID-19 exposure, gynecologists could suggest these patients to follow up with 6 months instead of further colposcopy or biopsy.
At last, we analyze the correlation of TS between IHC staining p16 and Ki-67, which are important auxiliary indicators in diagnosing CIN2+ [25, 26]. We found the correlation between TS and p16, TS and Ki-67 in overall participants and ASCUS group, but all the correlation strength were not strong. Even so, the correlations proved TS was of high quality in diagnosing CIN.