NEW YORK – Half of all gynecologist visits in the US are motivated by an inflammation of the vagina called vaginitis. Gold-standard tests to determine the cause of the inflammation are time consuming and lack accuracy, and clinicians frequently resort to treating patients empirically, making adjustments after treatment fails.
Two molecular diagnostic tests launched earlier this year by Hologic aim to detect and distinguish three of the primary causes of vaginitis — bacterial vaginosis (BV), yeast infection caused by Candida strains, otherwise known as Candida vaginitis (CV), and infection with the protozoan parasite Trichomonas vaginalis (TV) — have recently shown high accuracy in a published study. The tests have also expanded Hologic's business further into women's health diagnostics, and increased competition for the only other cleared molecular test for BV, Becton Dickinson's BD Max Vaginal Panel.
The Hologic Aptima BV and Aptima CV/TV assays showed improved sensitivity and specificity compared to standard tests in a study published last month in the Journal of Clinical Microbiology, as well as higher accuracy than clinician's diagnoses and in-clinic assessments, leading study authors to conclude that the tests were more predictive of infection than traditional diagnostic methods.
To get treatment right the first time using molecular diagnostics, physicians need to be able to test for different possible targets, and the BV and CV/TV assays now bring the overall menu on Hologic's Panther system to 16 tests that customers can perform to detect 20 pathogens.
"We know in real-world practice women are going to present with co-infections, and with one swab and our collection vial you can run up to seven different tests," Kevin Thornal, division president of diagnostic solutions at Hologic, said in an interview.
Jane Schwebke, an infectious disease specialist at the University of Alabama Birmingham and senior author on the JCM study, said that the three conditions covered by these two Hologic tests account for a very high proportion of vaginitis cases among women of reproductive age.
The consequences of the inaccurate treatment inherent in gold-standard methods like culture and microscopy vary, Schwebke said.
In the short term, treating vaginitis empirically can lead to incorrect treatment for some women's infections. In these cases, "She is not going to feel any better, and she is going to waste her time and money on inappropriate treatment," Schwebke said.
Long-term consequences of vaginitis depend on the infectious agent, but Schwebke noted that both trichomonasis and BV have been linked to increased risk of acquisition and transmission of other sexually transmitted infections, including HIV, as well as to preterm birth. Also, the consensus among women's health experts is that more than half of all women treated for bacterial vaginosis have a recurrence within a year.
"We believe that is because they are initially getting the wrong treatment," Thornal said, adding that the Hologic tests "will allow physicians to prescribe the right treatment the first time," when they use the BV and CV/TV assays.
Bacterial vaginosis — a condition where bacteria other than the Lactobacillus strain that dominates the vaginal microbiome tend to overgrow — can be inferred from a number of tests.
Schwebke explained that one instrument, called the Amsel criteria, is a bedside method. Clinicians check the pH of a vaginal sample and look at it under the microscope for squamous epithelial cells that are coated with bacteria, called clue cells. "You also do what is called a whiff test, where you take some of the vaginal fluid and mix it with potassium hydroxide and smell it for a fishy odor," she said. There is a fourth criteria, which is the presence of a homogenous vaginal discharge, and any three of these criteria would result in a positive Amsel criteria test, she added.
Another instrument, called a Nugent score, involves performing a Gram stain of vaginal fluid to look at three different morphotypes of the bacteria. The proportion of non-Lactobacilli bacteria is estimated, and the sample is scored on a scale of one to 10.
Both of these assessments require expertise, and clinician as well as technician time. Clinicians are not always trained to do microscopy, or may be reluctant to get the required CLIA waiver to perform the testing, she said. "Right now, most clinicians just kind of guess what's going on," she said, and they offer empiric therapy.
By contrast, a fully-automated molecular test for BV has a much simpler workflow and requires less expertise, so "the accuracy is going to be much higher," Schwebke said.
As described in JCM, the trial involved more than 1,400 vaginitis patients at 21 trial sites. Six vaginal swabs, including one self-collected by the patient, were used to determine the accuracy of the two molecular tests, with results for each target of the Hologic tests compared to gold standard diagnostic testing and clinical evaluations.
The Aptima BV assay specifically reports positive or negative results for BV based on a mathematical algorithm analysis of ribosomal RNA of Lactobacillus species, Gardnerella vaginalis, and Atopobium vaginae. The Aptima CV/TV detects and differentiates RNA for three targets — a Candida species group that includes C. albicans, C. dubliniensis, C. parapsilosis, C. tropicalis, as well as C. glabrata alone, and Trichomonas vaginalis.
The reference testing for the Aptima BV test involved Nugent scores and modified Amsel criteria performed in the lab and in the clinic. For the Aptima CV/TV test, the Candida targets were compared to yeast culture and bidirectional sequencing for a subset of specimens, while the T. vaginalis target was compared to the Xpert TV molecular assay from Cepheid and an InPouch TV culture system from Biomed Diagnostics.
The BV test has a sensitivity of 95 percent and a specificity of 90 percent compared to the reference methods. For the CV/TV test, sensitivity and specificity were 92 percent and 86 percent, respectively, for Candida species group, 85 percent and 99 percent for C. glabrata, and 97 percent and 99 percent for T. vaginalis. Sensitivity and specificity were also similar in the clinician-collected and patient-collected samples.
The molecular tests detected more co-infections than the reference methods — 25 percent of patients versus 20 percent.
The Aptima tests were also compared to clinician diagnoses, which factor in bedside testing, other patient symptoms, and patient history. Overall clinician’s diagnosis and in-clinic assessments had lower sensitivity and sensitivity than the molecular testing, sometimes quite dramatically lower. The potassium hydroxide whiff test had a sensitivity of 27 percent to detect Candida infections, for example, and the clinician's diagnosis for trichomoniasis and yeast infections had sensitivities of 38 percent and 55 percent, respectively.
Thornal said that the publication of the JCM study now allows Hologic to get very specific in its marketing and point back to the clinical data that was behind that labeling. "It's easier to have the conversation of how much more of an improvement it is over the current method of testing … for physicians to understand the benefit of switching to this test," he said.
A rising tide lifts all boats?
The JCM study describes the results of the clinical trials for the two tests, which were assessed and submitted to the US Food and Drug Administration simultaneously and cleared together in May of this year.
Bundling the tests in this way was a strategic decision at Hologic, Thornal said. Indeed, using the firm's automated system, the two tests combined can function almost like a panel for the major causes of vaginitis.
The only other cleared molecular diagnostic test that can assess bacterial vaginosis is the BD Max Vaginal Panel, which also detects CV and TV in the same test. The BD panel was cleared by the FDA in late 2016, and Becton Dickinson recently said sales of the vaginal panel have supported strong overall growth of the BD Max platform and testing menu.
The BD panel includes the same targets as the Hologic panel, but also adds two additional targets for BV — called Bacterial vaginosis associated bacteria–2 (BVAB-2), and Megasphaera-1 — and one additional Candida strain, C. krusei, according to a white paper from BD.
Current Centers for Disease Control and Prevention guidelines, released in 2015, recommend testing for T. vaginalis in all women seeking treatment for vaginal discharge, specifically using nucleic acid amplification tests if possible.
But the CDC guidelines for yeast infections do not incorporate molecular testing, as there were no FDA-cleared tests when the guide was written. The guide states that culture for yeast is the gold standard for diagnosis, but the C. glabrata strain in particular is difficult to culture and identify with microscopy, and also particularly resistant to antifungal treatment. The guide for BV describes research use of PCR-based testing, but notes that at the time the guide was written, clinical utility evaluations were under way.
Theoretically, more FDA-cleared molecular testing should spur more clinical utility studies, which could in turn impact guidelines and payor decisions on reimbursement.
For labs and patients, having cleared tests from more vendors may be a case of "the more the merrier," Barbara Van Der Pol, an STI expert also at the University of Alabama at Birmingham. Van Der Pol was lead author on the clinical trial of the BD Vaginal Panel and led the clinical trials of CT/NG/Trich test Rheonix recently submitted to FDA, as well as the trial of the SpeeDx Mycoplasma genitalium test.
Having more FDA-cleared commercial tests available for vaginitis might also help make molecular testing the norm, she said in an email. And, overall a panel approach can help clinicians to assess all possible causes of the patient's symptoms, "which is really needed given the frequency of co-infections," she said.
Last year Van Der Pol co-authored a study utilizing the BD Vaginal panel and the BD triplex test for chlamydia, gonorrhea, and T. vaginalis, together, revealing that approximately one quarter of the women who had bacterial vaginosis alone or who were co-infected with Candida were also carrying a sexually transmitted infection.
The study noted that whether a woman is initially tested for vaginitis or an STI can often depend on whether she first presents at a family health clinic or at an STI clinic. However, an apparent increased risk for some STIs in women who are positive for BV or Candida provides a strong impetus for comprehensive testing, the authors wrote. "Integration of molecular testing for vaginitis and STI would establish consistent, objective, and sensitive testing methods, regardless of clinic type, to accurately identify and treat patients for these conditions."
Van Der Pol also said that "insurance companies are still refusing to reimburse for these tests," adding that it gives the impression "women’s health is not so very important after all."
Thornal said that Hologic is in the process of trying to get the reimbursement ducks in a row. "We're looking to get an additional [CPT] code or a combo code," he said.
Generally speaking, although molecular tests offer an improvement, Schwebke noted that one potential downside is the 24- to 48-hour turnaround time. "What am I going to do with that woman who is sitting on the exam table wanting treatment right now? I haven't quite figured that out," she said.
Schwebke herself still uses microscopy in the clinic. She also serves as medical director of the Jefferson County Department of Health STD Clinic, which she said uses the Hologic Aptima CT/GC/Trich test for screening of both men and women, but she couldn't say whether that clinic will add the vaginitis tests, which she referred to as a panel, in the future.
Still, "We have had a lot of experience with the Aptima CT/NG/Trich assay and have been very pleased with it" she said, adding "I don't have any doubt that this vaginitis panel is going to live up to what we published," she said.