New findings on bacterial vaginosis

Nov 3 2005

Despite being one of the most common infections among women, scientists and doctors know little about the causes of bacterial vaginosis (BV), a usually benign disease that is also linked to serious health problems including pelvic inflammatory disease, an increase in the viral load of HIV from infected women and a two-fold increase in risk for pre-term labor and delivery.

Now researchers at the Fred Hutchinson Cancer Research Center have shed new light on BV by using genetic-sequencing technology to detect several new bacterial species - enough to almost double the number of known strains associated with the infection.

The findings - as well as a related editorial - will be published in the Nov. 3 edition of the New England Journal of Medicine.

These newly described uncultivated bacterial strains could be a reason for BV's high relapse rate - up to 30 percent - and could lead to the development of new diagnostic tests and better treatment approaches that target some of these novel bacteria. The knowledge also eventually will help doctors determine if particular uncultivated bacterial species are responsible for the serious complications associated with BV, according to David Fredricks, M.D., assistant member of the Infectious Diseases Program at the Hutchinson Center's Clinical Research Division.

BV is usually harmless and easily treated with antibiotics. Symptoms include malodorous discharge, itching and burning, however, some women with BV have no symptoms. Despite its prevalence (10 percent to 20 percent in Caucasian women and 30 percent to 50 percent in African-American women in the United States) the cause of BV is not well understood by scientists and doctors, Fredricks said.

"With most bacterial infections, you identify what the bacterium is and you treat that bacterium," said Fredricks, who is also an assistant professor in the Division of Allergy and Infectious Disease at the University of Washington School of Medicine. "The problem with BV is we don't know what we're treating. We know some of the strains associated with BV, however many strains that are cultured in the lab are not sensitive to the usual antibiotic treatments, yet patients may respond to therapy. We need to find out which bacteria cause BV and why some women either respond to antibiotic treatment or fail to be cured."

Fredricks and colleagues postulated that one reason for BV's persistent mystery is that scientists haven't identified the total community of bacteria that causes it. Indeed, the study found that most of the bacterial species detected in vaginal fluid are missed with conventional cultivation methods. The researchers catalogued a total of 35 bacterial species associated with BV, doubling the known species. Nineteen appear to be novel species, the study reported. The study identified three new strains, members of the Clostridium phylum, which researchers named BV-associative bacterium # 1, 2 and 3. They had specificities of 95 percent to 98 percent, meaning their detection in vaginal fluid reliably predicted the presence of BV. On the other hand, not all women with BV harbored these novel Clostridium-like bacteria.

"Numerous bacterial genera identified in this study have not, to our knowledge, been previously detected in the vaginal milieu with the use of cultivation methods," researchers said.

Rather than grow bacteria in a Petri dish -the standard identification method -Fredericks harvested DNA from bacteria, cloned segments of the DNA and used polymerase chain reaction (PCR) assays to analyze and sequence the segments. PCR analysis has a high degree of sensitivity and specificity. What made this possible is that every bacterium has the 16S ribosomal RNA gene, the gene that codes for ribosomes, which are the protein synthesizing machinery of cells.

"The 16S ribosomal RNA gene is unique for each different species of bacterium," said Fredricks. "By knowing the sequence of the gene you can identify the bacterium. We get a sort of molecular bar code." With these read outs, researchers went online to the GenBank database operated by the National Institutes of Health and looked for matches among the bacterial strains catalogued there.

Now that a more complete catalogue of bacterial strains has been assembled, Fredricks said work can begin looking for specific associations between individual species and the more adverse health problems linked to BV.

Fredricks said the larger lesson from PCR analysis is just how complex the human microbial ecosystem is. Other syndromes being investigated using PCR include ulcerative colitis, Crohn's disease, pneumonia and meningitis.