Medical Device Link.

MOLECULAR DIAGNOSTICS

Figure 1. Women in Barshi, India waiting for cervical cancer screening. Easy-to-use, rapid, objective tests are needed to give healthcare workers additional tools for screening in low-resource settings.

Cervical cancer, the second most common type of cancer afflicting women worldwide, generally develops from a persistent infection caused by a high-risk type of human papillomavirus (HPV). HPVs are ubiquitous viruses that infect epithelial cells. They are spread through skin-to-skin contact, frequently as a result of sexual activity.

There are more than 100 types of HPV, and their manifestations in infected persons vary from asymptomatic to nongenital and genital warts. About 40 of these HPV varieties infect human genitalia. More than 50% of sexually active people will be infected with at least one type of HPV in their lifetime. The genital HPV viruses are divided into two categories: low-risk, such as HPV 6 and HPV 11, which are associated with genital warts, and high-risk, including HPV 16 and HPV 18, which are associated with cervical cancer. Not all of the viruses will cause health concerns, and only a portion of chronic high-risk infections have a probability of progressing to precancerous or cancerous lesions.

Typically, cervical cancer is a slow-progressing disease that originates as cellular dysplasia and advances to become invasive cancer. When the disease is detected early as cellular dysplasia, treatment typically prevents its progression to cervical cancer, a devastating disease. Invasive cervical cancer produces great pain and discomfort and causes the death of women during their productive years, when their families and communities rely on them. Annually, 490,000 new cases of cervical cancer are diagnosed worldwide, and approximately 270,000 victims die from the disease.¹ The World Health Organization estimates that 3.2 million disability-adjusted life years are lost each year to cervical cancer.²

Women in the developing world disproportionately carry 80% of the global burden of cervical cancer. Success in reducing the incidence of cervical cancer in the developed world came largely from access to healthcare programs involving annual screening through high-quality cytology based on the Pap smear, and from adequate follow-up with women who test positive by cytology. The incidence of cervical cancer in developing countries has remained high, however, owing to women’s limited access to routine healthcare and the lack of good-quality, high-coverage Pap smear programs followed by effective treatment (see Figure 1).

This article reviews current screening technology and considers what is needed for determining risk of developing cervical cancer in women in low-resource settings. One testing platform that shows potential in this regard is described below.

Current Methods

Techniques now used to identify women at risk of developing cervical cancer include cytology, DNA testing, and visual inspection of the cervix. All of these methods require a speculum examination by a healthcare worker for collection of an adequate cervical sample.

The cytological approach is based on an examination of cervical epithelial cells for atypical physical abnormalities that are indicative of cancer or a precancerous condition. Samples are typically collected by scraping cells from the surface of the cervix with a spatula or specialized brush. Traditional Pap smears are made directly on the glass slide, while slide preparation for liquid-based cytology can be automated in a laboratory. Cytology requires a trained technician to stain the slide and a skilled and experienced professional to examine the cells using a microscope. Interpretation of cytology slides is highly subjective, even among the most highly trained healthcare workers. Also, the sensitivity of this technique, even in high-end laboratories, ranges between 60 and 70%. With regard to throughput, turnaround times in cytology can range from half a day to several weeks, depending on the laboratory methods used and the availability of trained cytologists to interpret the results.

Nucleic acid tests have recently been approved by regulatory agencies for the identification of women infected with high-risk types of HPV. Additionally, in 2005, the World Health Organization found HPV DNA tests acceptable as a primary screening modality.³ The digene HPV Hybrid Capture II (HC2) test developed by Qiagen Gaithersburg Inc. (Gaithersburg, MD) is the most widely accepted and used HPV nucleic acid test. The assay uses a cervical sample collected the way a cytology specimen is. Once the sample is processed to denature the DNA, it is incubated with RNA probes specifically designed to hybridize with any DNA of high-risk HPV present in the sample. An antibody that recognizes the DNA-RNA hybrid is used, and signal is detected by chemiluminescence. The assay requires a well-equipped laboratory, skilled technicians, and specialized equipment. HC2 performance is good: a study found it to have a sensitivity of 97.1% and specificity of 85.6%.⁴

However, the mere presence of HPV does not correlate well with progression to cancer, because many infections are self-limiting. The assay is run on a 96-well plate, so throughput is most efficient when samples are batched. Depending on patient volume, time to results can vary from half a day to several days.

Visual inspection techniques are simple methods in which differences in cervical epithelium are identified in vivo. Visual inspection with acetic acid utilizes a vinegar solution that is applied to the cervix. After about two minutes, the cervix is observed with the naked eye. A white patch is suggestive of abnormal cells. Visual inspection with Lugol’s iodine (also known as Lugol’s solution) is a similar process, but this method employs an iodine solution, and the presence of a bright yellow patch is the suggestion that cells are abnormal.

Visual techniques are inexpensive, and they can be conducted by trained healthcare workers. Thus, they are attractive for use in low-resource settings. The disadvantages of visual inspection are several. Turnover among practitioners generates a need for frequent training. Because determination of test results is subjective, the sensitivity and specificity of visual inspection methods for identifying women at risk of developing cervical cancer are highly variable, ranging from 30 to 80% depending on the setting in which the tests are implemented.

Throughput is slow, because visual inspection is performed one patient at a time during the speculum examination of each. This speed disadvantage is offset, however, by the immediate availability of results and the ready identification of women who need additional testing or treatment.

The Unmet Need

Cytology, DNA testing for high-risk HPV, and visual inspection can be effective tools when implemented in the proper environments with appropriate controls. However, these methods do not adequately fill the needs for screening tests that accurately predict individual women’s risk of developing cervical cancer.

While cytology is good for detecting abnormal cellular morphology, HPV DNA tests for identifying the presence of infection, and visual inspection for spotting problematic tissue characteristics in vivo, none of these methods can accurately identify a particular precancerous stage—whether cellular intraepithelial neoplasia grade 1 (CIN1), 2 (CIN2), or 3 (CIN3)—or tell whether a lesion will progress to cervical cancer or regress. Healthcare providers need novel tests that utilize markers that are more indicative of likely progression to cervical cancer. But the field of biomarker research is challenged by an ethical constraint: researchers cannot follow women with high-grade CIN2 or CIN3 precancerous lesions for progression to cancer as it is unethical for them not to provide treatment for this group.

Clearly, the field needs diagnostics appropriate for use where the prevalence of cervical cancer is greatest—in low-resource settings. The challenge of reducing cervical cancer in developing countries is too complex to be overcome simply by introducing current detection methods into high-incidence areas. The infrastructure—facilities, equipment, and personnel—and training required for conducting a proper examination, collecting a high-quality sample, and adequately performing tests of the types currently in use present significant barriers in many parts of the world. These factors hamper effective implementation of new screening technologies in low-resource settings.

To bridge the gaps between the capabilities of current methods and the need to combat cervical cancer globally requires new screening tools, more suitable for use in low-resource settings, that can more accurately identify women at risk of developing cervical cancer and thus reduce the number of women referred for diagnosis and treatment. Such tests would also bring health equity to cervical cancer prevention. New tests with better positive and negative predictive value would allow cervical cancer screening to be conducted every 5-10 years, or two or three times in a woman’s life, therefore making it possible for the limited resources in developing countries to be used more efficiently.

Recent Developments

The market under discussion clearly has adequate room for multiple new tools taking a variety of approaches to cervical cancer diagnostic testing. Researchers generally have a choice between two options when developing new detection tools: they can adapt existing technologies to better meet the need in low-resource areas, or they can develop novel approaches to solving the problem.

One of these approaches is not better than the other. Basing methods development on existing technologies offers the benefit of a foundation already laid, which often means reduced development time and costs. However, the developer may be limited to working with the bounded capabilities of the preexisting technology. On the other hand, while brand-new technologies can be uniquely fashioned to meet the requirements of specific needs, their development, validation, and adoption can have long timelines and can be costly.

Another approach to DNA testing for high-risk HPV involves polymerase chain reaction (PCR) assays for the detection, quantification, and typing of HPV DNA. The region selected for amplification is typically the gene that encodes for the HPV structural proteins L1 and L2. By careful selection of the targeted gene sequence, HPV DNA PCR assays can identify and differentiate the type or types of infection.

PCR assays have high analytical sensitivity, but they are sensitive also to contamination. In addition, the clinical utility of such tests for HPV has not been validated. The instrumentation and skills necessary to perform a PCR assay correctly are highly complex and technical, and sample preparation to remove contaminants and purify the DNA is a complex process as well. For these reasons, PCR assays are not appropriate for use in healthcare settings with limited resources. They are better suited for well-equipped laboratories where samples can be batched, resulting in high throughput. As HPV DNA PCR tests are indicative only of infection, they cannot identify women with neoplasia.

In a different area of development, the investigation of alternative markers predictive of cellular dysplasia or cervical cancer has begun to yield new approaches to testing. Cellular transformation resulting in a cancerous cell requires two nonstructural regulatory HPV-encoded oncoproteins, E6 and E7. These proteins affect cellular proteins such as p53 tumor suppressor and p16 kinase and therefore disrupt the cell cycle. It has been necessary for researchers to study these biomarkers thoroughly in order to determine their predictive validity for progression to cervical cancer and thus their utility for a diagnostic test.

One approach, building on the high sensitivity of nucleic acid testing platforms and the promise of alternative HPV markers that are more predictive of cancer, focuses on viral messenger RNA (mRNA) transcripts for E6 and E7. Because elevated levels of E6 and E7 proteins are necessary for cellular transformation, a logical thing to do is to look for increased levels of mRNA, which would differentiate an infection leading to dysplasia or cancer from a merely transient HPV infection. Two mRNA tests that do this are the Pretect HPV-Proofer by NorChip A/S (Oslo, Norway) and the Aptima HPV assay developed by Gen-Probe Inc. (San Diego). Studies have shown that the presence of elevated levels of E6 and E7 mRNAs may be indicative of high-risk HPV infections progressing to cervical cancer.⁵

Detection of an overexpression of the cellular protein p16 is the approach adopted by mtm Laboratories AG (Heidelberg, Germany), whose CINtec test detects overexpression of the variant p16INK4a in biopsied tissue and cytology slides via staining and the Cervatec test detects the protein from cell scrapings in a 96-well immunoassay. Data suggest that women with elevated levels of p16INK4a have an increased risk of progressing to cellular dysplasia or cervical cancer.⁶

Serology testing for exposure to high-risk HPV is inadequate for determining whether the woman is at risk of developing cervical cancer. It indicates only whether an individual has been exposed to HPV, not whether the infection is active or persistent.

Figure 2. Field evaluation of a prototypical cervical cancer screening test in a county hospital in Shanxi, China.

PATH (Seattle), with its Screening Technologies to Advance Rapid Testing for cervical cancer (START) project, provides an example of diagnostic tools developed specifically for low-resource settings (see Figure 2). START project activities included addressing user needs and policymakers’ preferences, conducting product research and development, performing sample collection and clinical evaluation, and evaluating the economics of the technologies in application. They were carried out in order to inform the R&D of the new tests. Users’ and policymakers’ feedback pointed to a need for tests that would:

• Provide results on the same day as sample collection in order to minimize lost follow-up opportunities.
• Require no special storage conditions.
• Tolerate temperatures as high as 35°C.
• Utilize both cervical and vaginal samples.
• Require no expensive instrumentation.
• Be able to run without electricity or else on a rechargeable battery.
• Maintain high clinical sensitivity and specificity for detecting high-grade disease (CIN2 or greater).
• Be practically priced.

Investigators found that the screening coverage—the number of women screened—was more important than assay sensitivity and specificity or the cost of the screening tests in reducing the impact of cervical cancer on a community. However, to facilitate adoption and sustainable employment of screening tests, the costs do have to be low.

New Assay Platforms

As part of the START project, PATH partnered with Qiagen to develop and assess the feasibility of a new platform for high-risk-HPV DNA testing that would be suitable for low-resource settings by virtue of being affordable and acceptable to women, healthcare workers, and policymakers.

Figure 3. The Qiagen careHPV system with all reagents and instrumentation required to perform the test. The system is designed to have a small footprint and use battery power so testing can be conducted in settings with limited infrastructure.

The careHPV test is based on the technology of Qiagen’s HC2 but is implemented in a novel format specifically designed for low-resource settings (see Figure 3). The assay is designed to work with provider-collected cervical samples or self-collected vaginal samples. Each sample is placed in an assay-specific solubilizing buffer and manually mixed. Up to 90 samples are batched and added to the plate and denatured. Complementary RNA probes specific for 14 types of high-risk HPV are added. If DNA from high-risk HPV is present, the probes will hybridize with molecules in the sample. The DNA-RNA hybrids are bound by specific enzyme-conjugated antibodies. The signal is detected by chemiluminescence. Incubation steps are conducted in a specially designed heater-shaker instrument, and plates are read on a simplified luminometer.

This test requires no running water. The complete assay footprint is small, and its instrumentation can be powered by a rechargeable battery. Disposables necessary for running the test have been minimized through innovative reagent delivery strategies. The test has been designed to be stored without refrigeration; it features dry reagent technologies and proprietary reagent preservation methods. Time to results is approximately two hours, which enables healthcare providers to identify the women who need follow-up attention on the same day they are screened.

The platform for the new test has been evaluated for performance in low-resource settings, with encouraging results. One evaluation, conducted in a county hospital in Shanxi, China, and involving 2530 women recruited from the general population, found that clinical sensitivity and specificity for cervical specimens were 90% and 84.2%, respectively, and for vaginal specimens were 81.4% and 82.4%, respectively.⁴ Similar programs to demonstrate the utility of careHPV in various settings in India, Uganda, and Nicaragua are slated to begin in 2009.

Qiagen will commercialize the new test, offering it to public-sector health entities in low-resource areas at cost-effective pricing that will allow sustainable mass campaigns for cervical cancer screening to be conducted.

In addition to an HPV DNA–based assay, the START project assessed the feasibility of a biomarker-based assay. PATH partnered with Arbor Vita Corp. (Sunnyvale, CA) to develop a strip test for detecting elevated levels of the E6 oncoprotein from high-risk types of HPV.

The platform is based on an immunochromatography strip that is suitable for testing small numbers of women—up to 10 at a time—at or near the point of care. The test utilizes a PDZ protein for specific capture of high-risk-HPV E6 oncoprotein and HPV type-specific anti-E6 antibodies for detection. PDZs are a conserved class of protein domains that engage in protein-protein interactions and are involved in many biological functions, including cell-to-cell contact, intercellular signaling, and cell polarity. (Their name is derived from the first three proteins in which the domain was described: postsynaptic density 95, Drosophila large disk, and zona occludens.)

Figure 4. (click to enlarge) The Arbor Vita E6 strip test utilizes a nitrocellulose strip with an absorbent pad with an immobilized stripe of PDZ (a). The sample is mixed with an anti-E6 gold conjugate (b). The strip is added to the sample which migrates up the strip (c). If E6 from a high-risk HPV type is present, the E6-gold complex will bind with the PDZ (d) producing a visibly red signal (e).

The anti-E6 monoclonal antibodies (mAb) are linked to gold particles; therefore, formation of the PDZ–E6 oncoprotein–anti-E6 mAb complex on the strip becomes visible as a red line (see Figure 4). The capturing PDZ protein derives from a cellular PDZ domain containing protein that recognizes E6 from high-risk types of HPV but not low-risk types.

Experimental evidence from mouse models supports a role for the E6-PDZ interaction in the pathogenesis of cervical cancer.⁷ Initial data on a small cohort of CIN1-positive women followed for two years for disease progression indicate that the presence of E6 in a sample results in a high risk for developing cervical dysplasia.⁸ Also, they suggest that the presence of E6 in lesions of grade CIN2 or greater positively correlates with their severity.⁹ These data indicate the presence of E6 as a risk factor for cervical cancer.

Additional work is being performed to increase the sample set and determine the quantity of E6 oncoprotein that correlates to disease state. Ongoing research and development is focused on giving the strip sufficient analytical sensitivity for clinical relevance. A strip test that has no special storage requirements, can be run with minimal laboratory infrastructure, and can determine whether a woman has, or predict whether she will progress to, cervical cancer would be a useful tool for screening and triaging women in low-resource settings.

Vaccines and Prevention

The recent development and approval of HPV vaccines have given hope for eventual eradication of cervical cancer. Vaccines that are now available provide direct protection against the two most prevalent types of high-risk HPV that cause cervical cancer. Vaccination against HPV 16 and HPV 18 should significantly reduce the burden of this disease. Approximately 70% of cases are caused by these two types. And as new vaccines that provide additional high-risk-HPV coverage become available, the burden of disease is expected to decrease even further.

Because vaccines have maximum efficacy in girls and young women who have not reached sexual debut, vaccinating older women is not an effective strategy for immediately reducing the incidence of cervical cancer. It could take 30 years before the full impact of HPV vaccination strategies becomes apparent. Incomplete coverage or slow adoption of the vaccine will only prolong the time frame for the realization of benefits. Substantial hurdles to full adoption and coverage are the vaccine’s required three-dose regimen and its cost. Current negotiated U.S. government contracts for the vaccine place the total cost of the three-dose regimen at $288 (the price will be substantially less in low-resource settings that could benefit most, however). For these reasons, it is important to continue cervical cancer screening campaigns in both developed countries and low-resource regions.

Conclusion

Cervical cancer is one of the few cancers for which the cause is known. Also well understood is the way to detect and treat precancerous lesions. Current technologies for identifying women at risk have benefited only those who have access to well-resourced health services, however. New technologies are needed to differentiate a simple HPV infection from a precancerous lesion, especially in populations where effective screening strategies or HPV vaccination have reduced disease prevalence. Additionally needed are new tools appropriate for use in low-resource settings that can be adopted by the healthcare systems in those areas.

Given the proper tools, practitioners could foreseeably cut down the disproportionately high burden of cervical cancer borne by women in the developing world. Cervical cancer screening services are a cost-effective means of reducing this burden. Vaccination against high-risk HPV is an important element of the assault on the disease, but access to screening, diagnosis, and treatment is necessary to make the largest possible impact.

Work to develop tools for use where such services are needed most has begun. Diagnostic tools that do not rely on the support of a standard laboratory are in great demand. Platform and device developers should be encouraged to work closely with disease specialists, healthcare workers, and policymakers to develop, validate, and implement creative and simple solutions that are appropriate for the administration areas most in need.

Roger B. Peck is a research scientist in the diagnostics development team of the Technology Solutions Global Program at PATH (Seattle). He can be reached at rpeck@path.org.	Jose A. Jeronimo, MD, is director of START in the Reproductive Health Global Program at PATH. He can be reached at jjeronimo@path.org.	Bernhard H. Weigl, PhD, is deputy director of START and group leader of the diagnostics development team of the Technology Solutions Global Program at PATH. He can be reached at bweigl@path.org.	John W. Sellors, MD, is past medical director of the Reproductive Health Global Program at PATH and a professor in the Department of Family Medicine at McMaster University (Hamilton, ON, Canada). He can be reached at ve3fos@yahoo.com.