Despite being highly preventable, more than 300,000 women across the world died of cervical cancer in 2018. The majority of cervical cancer deaths occur in women from low- and middle-income countries with limited access to health care services. Conversely, cervical cancer death rates were reduced by nearly 50 percent in American women due to the widespread adoption of Papanicolaou (Pap) tests in clinical practice. Although wildly successful, the Pap test suffers some limitations and newer screening methods have been developed to address its limitations.
Classification of Cervical Lesions
|Cervical intraepithelial neoplasia (CIN) is a precancerous lesion of the uterine cervix. Based on the degree of cellular anomaly and the extent of tissue affected, it is classified as follows:
|CIN1: Has very few abnormal cells and involves about one third the thickness of the epithelium. Such a lesion is least likely to progress into cancer. These may also be called low-grade squamous intraepithelial lesions (LSIL) or mild dysplasia.
|CIN2: Lesions affecting one third to two thirds of the thickness of the epithelium.
|CIN3: Lesions affecting more than two thirds to nearly the entire thickness of the epithelium.
|CIN2 and CIN3 lesions are high-grade squamous intraepithelial lesions (HSIL) and may be termed moderate or severe dysplasia. Neoplasia that invades the epithelial basement membrane is classified as cancerous.
Most laboratories in the US report abnormalities according to the Bethesda System. These are ranked from lowest to highest severity as follows:
Pap test limitations
The Pap smear is a cytologic screening test performed by scraping cells from the cervix, fixing them onto glass slides, and examining them under a microscope. It identifies cellular changes or precancerous cells in the cervix before they transform into invasive tumors. Screening and early detection have proved crucial in improving therapeutic success and reducing mortality.
The reliability of the Pap test depends on the skill with which the cervical smear is obtained, the accuracy with which it is interpreted, and the frequency of testing.1 Thus, patient compliance and the availability of trained personnel are important considerations. Often, there may be a paucity of cytotechnologists and pathologists in resource-limited settings with a high cervical cancer burden. In addition, cultural barriers and gender norms may limit the uptake of the Pap test in parts of Asia and Africa.2 Lastly, the Pap smear has a low sensitivity ranging from 30 percent to 87 percent.3 Thus, abnormal results must be confirmed by visually examining the cervix using colposcopy or by biopsy.
Advances in cytological screening
The Pap smear is a low-cost, non-invasive method for cervical cancer screening. However, manual slide preparation may compromise specimen quality due to obscuration by blood, mucus, and inflammation, and uneven cellular distribution. Liquid-based cytology was developed to improve specimen quality and diagnostic reliability of Pap smears.
In liquid-based cytology, cellular material collected from the cervix is rinsed and fixed in a vial of preservative fluid before placing it on a slide as a thin layer. This preserves the cellular structure, clears debris, and prevents clumping—addressing issues that preclude the accurate interpretation of morphological features on conventional Pap smears. The technique also allows samples to be transported over long distances for analysis. This may be beneficial in areas that lack access to trained cytologists. However, the higher cost of consumables (vials, slides, test kits) for liquid-based cytology may be a disadvantage in regions with limited resources.
Whether liquid-based preparations offer significant diagnostic benefits over conventional Pap smears is debated. Several studies report that the sensitivity and rate of detection of cellular anomalies is comparable or only slightly higher for liquid-based preparations compared to conventional smears.4–6 Nonetheless, most commercial screening systems use liquid-based cytology for sample preparation.
Automated Pap smear
The Pap test requires skilled professionals to examine slides under the microscope for several hours each day to identify abnormal cells among thousands of microscopic objects. Relying entirely on human input to interpret Pap smears is prone to errors. Failure to flag abnormal cells introduces false-negative errors in reporting. This results in cervical cancer going undetected in its early non-invasive stage and developing into advanced disease. In countries with an established screening program, repetition, loss of concentration, and task fatigue may introduce false-negative rates of 5 to 10 percent.1 On the other hand, a false positive result may lead to women undergoing unnecessary procedures that place a financial and psychological burden on them.
Automation may help overcome the labor-intensive and subjective interpretation of cellular morphology. The first FDA-approved commercial analyzer used liquid-based cytology for sample preparation and flagged morphological anomalies for human inspection.7 SurePath™, ThinPrep™, and FocalPoint™ GS Imaging are commercial analyzers commonly used in clinical practice.
Over the past few years, advances in AI and big data have been used to improve the accuracy and efficiency of cervical cancer screening. A recent study combined digital microscopy and AI for screening in rural Kenya.8 Conventional Pap smears were collected at a rural clinic, digitized using a portable scanner, and used to train an algorithm to detect low-grade and high-grade lesions. The algorithm had high accuracy and sensitivity for detecting atypical cells, but also had a high rate of false positives for low-grade atypical cells that are susceptible to subjective interpretation on visual inspection. If used as a screening tool in resource limited settings, it can exclude the majority of normal samples, allowing pathologists to focus on verifying potentially abnormal slides. Similar AI-assisted cytology systems have also been deployed in low-resource settings in China. In an analysis of 70,000 women, AI-assisted cytology was more sensitive in detecting CIN2 lesions than manual examination by 5.8 percentage points, a statistically significant difference.9
Non-cytological screening for cervical cancer
The human papillomavirus (HPV) is a group of more than 200 sexually transmitted viruses. Although most HPV infections are naturally cleared by the immune system, persistent infection with high-risk oncogenic HPVs such as HPV 16 and HPV 18 are implicated in the majority of cervical cancer cases worldwide.
"Advances like liquid-based cytology and AI-based automation, as well as HPV testing, can help overcome these limitations and enhance point-of-care testing in areas with limited access to skilled health care professionals."
HPV co-testing along with a Pap smear was used to optimize the management of patients with ASC-US. Of the patients with ASC-US who undergo colposcopy, only about 20 percent are found to have a high-grade lesion. HPV co-testing identifies women at high risk of cervical cancer and avoids unnecessary testing in the low-risk cohort. In 2003, the FDA approved the HPV test as a primary screening test for cervical cancer. Randomized control trials have demonstrated that HPV testing has a significantly higher sensitivity of 95 percent for detecting high-grade CIN, compared to 55 percent with the Pap test.10 HPV testing has a high negative predictive value, low training requirements, high reproducibility, and high throughput. A negative HPV test requires screening only every five years as opposed to every three years for a Pap smear, reducing the need for follow-up tests.
HPV testing is performed using PCR amplification to detect DNA or mRNA of high-risk viral genotypes. Currently, there are 254 commercial HPV tests available to detect 425 viral variants.11 However, tests need to be validated and standardized for comparison to be used as clinically relevant diagnostics. Recently, a fully automated, high-throughput diagnostic system capable of detecting 14 high-risk HPV types in a single analysis received FDA approval.12 This system combines robotics and sample management software to automate labor-intensive and error-prone manual processes in clinical laboratory workflows.
The high cost of instrumentation, reagents, and infrastructure make primary HPV testing using commercially available devices difficult to implement in low-income settings. A battery-operated, portable PCR analyzer was recently used to detect four high-risk HPV genotypes at a clinic in India. In nearly 600 cervical samples, the sensitivity and specificity of the microanalyzer was found to be comparable to a commercially available validated HPV test.13
Other emerging diagnostic methods
Epigenetic silencing of tumor suppressor genes via DNA methylation is implicated in the progression of cervical cancer. DNA methylation is significantly higher in high-grade versus low-grade lesions. Furthermore, a meta-analysis of 43 studies found that DNA methylation had higher specificity than cytology and some types of HPV tests.14 Thus, DNA methylation tests are an objective approach to cervical cancer screening and risk stratification.
Moreover, certain host factors are modified by viral infection. For instance, p16INK4a–Ki-67 is upregulated by viral oncogenes. Such biomarkers may be identified using staining techniques and are found to have comparable results to liquid-based cytology and HPV testing.15
Moving beyond the conventional Pap smear
Although the Pap smear is the current gold standard for cervical cancer screening, it suffers some drawbacks. A labor-intensive process requiring skilled professionals, the conventional Pap smear analysis is prone to errors due to variation in specimen quality and subjective interpretation. Advances like liquid-based cytology and AI-based automation, as well as HPV testing, can help overcome these limitations and enhance point-of-care testing in areas with limited access to skilled health care professionals.
1. Coleman D. Clin Risk. 2001;7(6):235-240 doi:10.1258/1356262011928635
2. Nikumbh DB, Nikumbh RD, Kanthikar SN. South Asian J Cancer. 2016;5(2):79. doi:10.4103/2278-330X.181646
3. Nanda K, McCrory DC, Myers ER, et al. Ann Intern Med. 2000;132(10):810-819. doi:10.7326/0003-4819-132-10-200005160-00009
4. Siebers AG, Klinkhamer PJ, Grefte JM, et al. JAMA. 2009;302(16):1757-1764. doi:10.1001/JAMA.2009.1569
5. Nishio H, Iwata T, Nomura H, et al. Jpn J Clin Oncol. 2018;48(6):522-528. doi:10.1093/JJCO/HYY050
6. Akamatsu S, Kodama S, Himeji Y, Ikuta N, Shimagaki N. Acta Cytol. 2012;56(4):370-374. doi:10.1159/000337641
7. Bengtsson E, Malm P. Comput Math Methods Med. 2014;2014. doi:10.1155/2014/842037
8. Holmström O, Linder N, Kaingu H, et al. JAMA Netw Open. 2021;4(3):e211740-e211740. doi:10.1001/JAMANETWORKOPEN.2021.1740
9. Bao H, Sun X, Zhang Y, et al. Cancer Med. 2020;9(18):6896. doi:10.1002/CAM4.3296
10. Mayrand M-H, Duarte-Franco E, Rodrigues I, et al. N Engl J Med. 2009;357(16):1579-1588. doi:10.1056/NEJMOA071430
11. Poljak M, Oštrbenk Valen?ak A, Gimpelj Domjani? G, Xu L, Arbyn M. Clin Microbiol Infect. 2020;26(9):1144-1150. doi:10.1016/J.CMI.2020.03.033
12. https://news.bd.com/2021-08-25-BD-Launches-Fully-Automated-High-Throughput-Molecular-Diagnostic-Platform-for-U-S-Laboratories. Accessed October 1, 2021.
13. Hariprasad R, Tulsyan S, Babu R, et al. JCO Glob Oncol. 2020;(6):1147-1154. doi:10.1200/GO.20.00024
14. Kelly H, Benavente Y, Pavon MA, De Sanjose S, Mayaud P, Lorincz AT. Br J Cancer. 2019;121(11):954-965. doi:10.1038/s41416-019-0593-4
15. Song F, Du H, Xiao A, et al. Cancer Manag Res. 2020;12:9067-9075. doi:10.2147/CMAR.S273079
CIN2: Lesions affecting one third to two thirds of the thickness of the epithelium.