Editor's Note: Liz Highleyman is a Treatment, Education and Advocacy Assistant at the San Francisco AIDS Foundation. Leslie Hanna is Associate Editor of BETA.
Pelvic inflammatory disease (PID) is an infection of the female upper reproductive tract. The term actually refers to a broad range of conditions. Various parts of the upper reproductive tract may become infected, including the fallopian tubes, the uterus and the ovaries. Advanced cases may affect the abdominal lining (peritoneum).
PID is a serious, potentially life-threatening condition in any woman. HIV-infected women have some unique considerations and risks. The 1993 expanded Centers for Disease Control and Prevention (CDC) case definition of AIDS includes PID as a Category B illness, or one that is attributable to or complicated by HIV. Although women with HIV respond as well to treatment for PID as women without HIV, HIV itself creates a biological environment, especially in the vagina and lower reproductive tract, that facilitates the development of PID. This consideration makes prevention of and screening for PID especially important for HIV positive women.
Causes of PID
Except for cases related to abortion, childbirth or pelvic surgical procedures, PID occurs when infection-causing organisms (pathogens) spread upward from the vagina through the cervix to the upper reproductive organs. Anything that facilitates this migration of pathogens may facilitate the development of PID. HIV-related physiological changes such as altered mucosal integrity may encourage PID. HIV-related disruption of the normal vaginal ecology may contribute to an imbalance that favors the proliferation of some bacteria over others. Resident bacteria (those that normally live in the vagina, e.g., Gardnerella vaginalis) may prevail against a weakened immune system and expand beyond their normal "territory."
Untreated sexually transmitted diseases (STD) cause most cases of PID. When STDs go undetected or untreated, the pathogens may proliferate and spread to infect new areas. Untreated or undertreated gonorrhea and chlamydia cause an estimated 90% of cases of PID. As many as 40% of untreated cases of chlamydia ascend to the upper reproductive tract. Chlamydia often goes undetected in women; 75% of women show no symptoms of chlamydia infection. Because women with gonorrhea and chlamydia are so often asymptomatic, Risa Denenberg, FNP, recommends STD screening every 6 months for women with AIDS and for HIV positive women who are symptomatic or sexually active. New non-culture-based screening tests for chlamydia have been developed that use DNA amplification technology and can be used on a urine sample. In addition, recently reported findings by D. Dean and colleagues, of the University of California at San Francisco, suggest that certain types of Chlamydia trachomatis, the organism that causes chlamydia, are more aggressive than others.
Advanced PID usually involves more than one pathogen. Some cases of PID involve infection with bacteria that do not commonly affect the reproductive tract, such as Mycobacterium tuberculosis. M. tuberculosis occasionally infects organs in the upper reproductive tract, after initially infecting the lungs. Cytomegalovirus (CMV), a common opportunistic pathogen in people with HIV/AIDS, also has been identified in reproductive organs. A. Moorman and colleagues, of the Multicenter HIV and PID Study Group, found higher rates of CMV in the cervical area of HIV positive women compared to HIV negative women. However, little is yet known about the clinical significance of the presence of CMV in the endocervix (upper, inner part of the cervix).
Certain anatomical or physiological states increase the risk of an acute episode of PID. Risk for PID is high during or immediately following the menstrual period, when the cervix is slightly dilated (more open), making it easier for pathogens to spread from the vagina to the upper reproductive tract. Seventy percent (70%) of women with gonoccocal PID develop symptoms during the first week after menstruation. Risk of PID is higher at points in the menstrual cycle when the cervical mucous is thinnest and most penetrable (e.g., during ovulation).
Oral contraceptives affect risk of PID in complicated, incompletely understood ways. Their use is associated with some degree of protection against symptomatic PID, possibly because they thicken cervical mucous, making it more difficult for pathogens to spread into the uterus. Conversely, oral contraceptive use appears to increase the frequency of chlamydia infections by exposing a type of vaginal tissue, called columnar epithelium, that is particularly vulnerable to infection.
PID may also develop following an incomplete or unsterile induced abortion or following childbirth. This type of PID may be accompanied by a severe infection of the bloodstream that is directly related to delivery or to an obstetric procedure (puerperal sepsis). Premature rupture of membranes or prolonged labor are particularly associated with PID, usually due to infection with an organism called Bacteroides. Any procedure in which the cervix is opened (e.g., using dilating rings) or an instrument is inserted into the uterus (e.g., a curette) may lead to pelvic infection. Women who use douches also may be at higher risk of PID. Douching removes protective mucous and normal bacterial flora, and may unintentionally flush vaginal bacteria into the upper reproductive tract.
Use of contraceptive intrauterine devices (IUD) is associated with a 3-fold increased risk of developing PID. The string of the IUD, which is left hanging through the cervix into the vagina, provides a conduit for pathogens to reach the upper reproductive organs. Therefore, IUDs increase the likelihood that an STD will progress to PID. If a woman using an IUD develops symptoms of PID, the device should be removed immediately. Because HIV positive women are especially susceptible to infection, this form of birth control is not recommended for them.
Barrier methods of contraception (e.g., condom, diaphragm) can help protect women against the organisms that cause PID. Test-tube studies show that spermicides kill Neisseria gonorrhoeae and C. trachomatis, the organisms that cause gonorrhea and chlamydia, respectively. Therefore, use of contraceptive foams and jellies made with these spermicides may help protect women against PID.
Consequences of PID
Untreated PID has serious consequences. The fallopian tubes and ovaries may become inflamed, swollen and covered with pockets of pus, becoming what are called tubo-ovarian abscesses; these abscesses occur in 10% of cases of PID. Pelvic pain may become chronic. Untreated PID related to gonococcal infection can lead to Fitz-Hugh-Curtis syndrome, or perihepatitis, an inflammation of the covering of the liver. Other potential outcomes are intestinal blockage, sepsis (the presence of bacteria or their toxins in the blood) and death.
Adverse reproductive consequences include infertility and ectopic or tubal pregnancy (in which the embryo begins to develop outside the uterus, e.g., in a fallopian tube). Infertility is due either to inflammatory changes in the uterine lining, which make it resistant to implantation of an embryo, or to scar tissue that blocks the reproductive organs. Significant pelvic scarring occurs in at least 20% of women who have had PID; some degree of scarring is likely to occur in nearly all cases. A woman who has had PID is 6-10 times more likely to have an ectopic pregnancy, a life-threatening condition that is the leading cause of maternal death among African-American women. Ten percent (10%) of women who have had PID will have an ectopic pregnancy when they become pregnant for the first time following an episode of PID. As many as 1 in 3 women have repeat episodes of PID after initial treatment. The rate of infertility is near 10% after 1 properly treated PID episode, near 20% after 2 episodes and near 50% after 3 episodes.
Symptoms of PID
Although the symptoms of PID may vary, they generally include moderate to severe pain, abdominal tenderness, inflammation of the cervix (cervicitis), pain during intercourse and fever. Vaginal discharge may be unusual, possibly pus-containing (purulent); non-menstrual bleeding may also occur. Urination may be painful and/or frequent. A palpable pelvic mass may be present.
Early symptoms are often absent or mild, and PID may progress to an advanced stage before it is diagnosed. PID due to chlamydia is less acute and less symptomatic than PID due to gonorrhea or other pathogens, but causes a more protracted illness than does gonorrheal infection. Chlamydial PID presents a greater potential for missed diagnosis and permanent damage.
PID can be a challenge to diagnose, since obtaining culture specimens from the upper reproductive tract is invasive and difficult. Organisms responsible for upper tract infection may not appear in vaginal or cervical samples. Therefore, PID is usually diagnosed on the basis of clinical findings. Since clinical diagnosis may be imprecise, many cases of PID go undetected. Because the consequences of untreated PID are so serious, the CDC recommends that healthcare providers "maintain a low threshold of diagnosis."
The CDC recommends empiric treatment for PID if all 3 of the following symptoms are present in a woman with no established cause of pelvic inflammation: lower abdominal tenderness, adnexal tenderness (tenderness of the pelvic reproductive organs) -- especially bilateral tenderness (on both sides of the body) -- and cervical motion/manipulation tenderness. The following additional criteria may be used to make a more specific diagnosis: oral temperature greater than 38.3 C (100.9 F) and abnormal cervical or vaginal discharge. Laboratory blood tests such as elevated blood erythrocyte sedimentation rate (ESR), elevated levels of C-reactive protein and increased white blood cell counts, as well as laboratory detection of N. gonorrhoeae or C. trachomatis, also may support a diagnosis of PID.
Additional diagnostic methods that are sometimes used include laparoscopy, biopsy and culdocentesis. Laparoscopy is a procedure in which a small, lighted lens-tip instrument is inserted through a narrow abdominal incision to allow visual examination of internal organs and tissue sampling. In an endometrial biopsy, a sample of tissue from the uterine lining (endometrium) is extracted and examined under a microscope. Culdocentesis is the removal (aspiration, via suction) of uterine fluid from the cul-de-sac, a recess between the uterus and the rectum, through a puncture of the vaginal wall. Ultrasonography (imaging using sound waves) may also be used to help diagnose PID, and can be especially useful for detecting fluid-filled tubo-ovarian abscesses and ectopic pregnancies. Because these procedures are potentially inconclusive and, except for ultrasonography, invasive, they should not be routinely substituted for clinical diagnosis.
PID in HIV Positive Women
There are no data to establish conclusive differences in clinical manifestations and course of PID in HIV positive women compared to HIV negative women. Large, controlled, unbiased studies have been difficult to conduct. However, anecdotal reports and some studies suggest that PID may be more common and more severe in women with HIV, and especially in women with AIDS. Abner Korn, MD, of San Francisco General Hospital, and colleagues note that "decreased mucosal and systemic immune function due to HIV could lead to increased ascent of pathogenic organisms from lower to upper genital tracts and lowered response to antimicrobial therapy once upper tract disease is established."
Some recent small studies have gathered data on immune responses to PID in HIV positive and HIV negative women. A central finding is that HIV-infected women with PID may not have elevated white blood cell counts in response to infection the way HIV negative women do. B. Hoegsberg and coworkers, C. Barbosa and coworkers, and Korn and coworkers have all reported relatively lower white blood cell counts (leukopenia) in HIV positive women with PID. Korn believes that this is partly related to the relatively lower incidence in HIV-infected women with PID of gonoccocal infections, which are associated with high white cell counts. The significance of this finding is that laboratory markers such as elevated white blood cell count and ESR cannot be considered reliable diagnostic indicators of PID in HIV positive women.
Hoegsberg's group reported a trend toward more PID-related surgical intervention in HIV positive women than in HIV negative women in Brooklyn (p=0.058; n=15 HIV positive, 95 HIV negative), as well as a trend toward more recurring infections. Korn and colleagues also found a higher incidence of surgical intervention in HIV positive women in San Francisco (p<0.05; n=23 HIV positive, 108 HIV negative). It appeared that the incidence of surgery increased as immune function decreased. Among asymptomatic HIV positive women with PID, there was no increase in surgical intervention, while 3 of 5 women with AIDS required surgery. Korn suggests that this higher incidence of surgery may be due to inadequate response to antibiotic therapy among HIV positive women. Still, Korn's study did not find a greater length of hospitalization or a greater incidence of tubo-ovarian abscesses in HIV positive women with PID compared to HIV negative controls. Barbosa and coworkers found that the incidences of tubo-ovarian abscesses and surgical intervention were the same in HIV positive women (n=13) and HIV negative women (n=138) with PID.
In a large, multicenter study sponsored by the CDC, Kathleen Irwin and colleagues, of the Multicenter HIV and PID Study Group, examined the clinical presentation and course of PID in HIV positive and HIV negative women (n=43 HIV positive, 162 HIV negative). HIV positive women had slightly more severe initial symptoms and were slightly more likely to have had PID in the past. Irwin found that HIV positive and HIV negative women had similar amounts of pain and cervical discharge, but that HIV positive women had significantly lower white blood cells counts.
In a microbiological substudy of tissue samples from the same women, Moorman and coworkers identified N. gonorrhoeae, C. trachomatis and various other types of bacteria in uterine lining (endometrium) and endocervical tissue at equal rates in HIV positive and HIV negative women. In these studies, HIV positive women were significant-ly more likely to have recently taken antibiotic drugs. Among women who had not recently used antibiotics, infection of the endometrium (endometritis) was indeed more common in women with HIV (65%) than in HIV negative women (30%), a statistically significant difference. Irwin s group found that 4-10 days after antibiotic therapy was completed, the HIV positive and HIV negative women were similarly improved.
Irwin's group found that initial symptoms and course of PID in HIV-infected women did not differ based on CD4 count. Irwin and colleagues attribute to bias the greater differences between HIV positive and HIV negative women found in previous studies; for example, the lower pain scores reported by HIV positive women with PID noted by Korn may be related to their use of opiate drugs.
Some investigators have reported lower rates of gonorrhea and chlamydia in HIV positive women with PID. Korn and coworkers postulate that this may be because PID in HIV-infected women is more often due to the ascent of resident vaginal organisms (e.g. Bacteroides, Staphylococcus, Streptococcus, Escherichia coli), which the compromised immune system can no longer keep in check. However, Irwin and R. Rice have countered that the lower rates of gonorrhea and chlamydia in HIV positive women may be due to their more frequent use of antibiotic drugs.
Treatment of PID
PID is typically treated with a combination of antibiotics. It is important to select an initial treatment regimen that provides broad activity against the range of organisms likely to cause infection, including various resident bacteria. It is also crucial to treat any initial infection such as gonorrhea or chlamydia. Since specimens are difficult to obtain, cultures take time and most cases of PID are polymicrobial (more than one pathogen is present), it is recommended that broad-spectrum combination antibiotic therapy not be delayed while awaiting laboratory results. Various studies, including the Irwin study discussed previously, have found that HIV status does not appear to influence cure rates of bacterial PID, and that HIV infection does not appear to negatively affect response to therapy.
PID treatment may be administered on an inpatient (in the hospital) or outpatient basis. Some studies have shown that PID may be more difficult to treat in women with HIV, and treatment may be more likely to fail. Thus, the CDC currently recommends hospitalization and intravenous (IV) antibiotic therapy for women with HIV infection and PID. Inpatient treatment is recommended for any woman who does not respond to outpatient therapy, in whom pelvic abscesses are suspected or who is too ill to take oral medication. Inpatient IV treatment is also recommended for pregnant and adolescent women. Two regimens for IV use are 1) cefoxitin and doxycycline, and 2) clindamycin and gentamycin. Cefoxitin and cefotetan are effective against gonococci and many other bacteria; doxycycline is active against chlamydia, mycoplasma and various resident bacteria. IV antibiotic treatment should continue for at least 4 days, and should continue for 2 days after fever subsides. Hospitalized patients should begin to improve in 2-3 days. By the seventh day of therapy, 80% of effectively treated women will have a normal pelvic exam. After IV therapy, according to an article in Scientific American Medicine, "a 14-day antibiotic course can be completed with the administration of doxycycline."
Because PID infection can be persistent, the CDC recommends re-examination 7-10 days following the completion of therapy to ensure that infectious organisms have been eradicated. Some physicians also recommend further follow-up at 4-6 weeks. Also, effective treatment for a woman with PID includes treating sexual partners for gonorrhea and chlamydia, even if partners are asymptomatic.
Severe, difficult to treat cases of PID often involve disturbances of the bacteria that normally reside in the vaginal area. For these cases, as well as cases that involve tubo-ovarian abscesses, IV clindamycin and gentamicin may be effective. Doses of gentamicin can be adjusted as needed relative to kidney function.
If a tubo-ovarian abscess does not respond to antibiotic therapy, surgical drainage may be required; immediate surgical intervention is needed if an abscess threatens to rupture. The most severe cases of PID may necessitate surgical removal of the uterus (hysterectomy), fallopian tubes and/or ovaries.
Non-steroidal anti-inflammatory drugs (NSAID) are often effective for the management of chronic pain due to PID; acupuncture may be a useful adjunct therapy. For more severe pain, narcotic pain medication such as codeine may be necessary.
Outpatient treatment is really only recommended for ambulatory (able to walk and move about normally, i.e., with relatively low-level pain and symptoms) HIV negative women with mild PID. Two recommended outpatient regimens are 1) an injected cephalosporin such as cefoxitin or ceftriaxone plus oral doxycycline or tetracycline (and possibly metronidazole), or 2) oral ofloxacin plus clindamycin or metronidazole. Ofloxacin is effective against both N. gonorrhea and C. trachomatis, and IV ofloxacin has been studied as monotherapy. However, at this time, the CDC does not recommend single-agent treatment for PID. Treatment should last at least 10-14 days. A follow-up examination should be done within 72 hours of the initiation of therapy to check for improvement. It is important that the full course of antibiotic therapy be taken, even if symptoms clear.
The conflicting findings on the clinical symptoms and course of PID in HIV positive women point to a need for continued research in this area, including how HIV-related immunosuppression influences the acquisition and course of STDs. Research is also needed on better methods to protect women from infection with the pathogens that can lead to PID.
Early treatment of PID is successful in HIV positive as well as in HIV negative women. It is therefore important that regular STD screening and thorough evaluation of pelvic pain and other symptoms be a part of the healthcare regimen of all women with HIV.
AIDS Weekly. Chlamydia causes infertility and increases risk of HIV transmission. Report of a CDC national interactive satellite symposium. 1995.
Barbosa C and others. Clinical course of pelvic inflammatry disease in HIV-infected women. Infectious Disease Society for Obstetrics and Gynecology. Stowe, VT. August 1993.
Boston Women's Health Book Collective. The New Our Bodies, Ourselves. Touchstone, New York. 1984, 1992.
Centers for Disease Control and Prevention. 1993 Sexually Transmitted Diseases Treatment Guidelines. Morbidity and Mortality Weekly Report 42:RR-14. September 24, 1993.
Dean D and others. Major outer membrane protein polymorphism of Chlamydia trachomatis is associated with severity of genital tract infections in San Francisco. ICAAC. San Francisco, CA. September 17-20, 1995. Abstract K89.
Denenberg R. Gynecological Care Manual for HIV Positive Women. Essential Medical Information Systems, Inc., Durant, OK. 1993.
Hoegsberg B and others. Sexually transmitted diseases and human immunodeficiency virus infection among women with pelvic inflammatory disease. American Journal of Obstetrics and Gynecology 163(4): 1135-1139. 1990.
Irwin K and others. The clinical presentation and course of pelvic inflammatory disease in HIV+ and HIV- women: preliminary results of a multicenter study. IX International Conference on AIDS. Berlin. June 1993. Abstract WS-B-07-1.
Irwin K and others. The influence of HIV on initial presentation and course of pelvic inflammatory disease: Final results of a multicenter study. ICAAC. San Francisco, CA. September 17-20, 1995. Abstract K95.
Irwin K and R Rice. Pelvic inflammatory disease in human immunodeficiency-infected women (letter). Obstetrics and Gynecology 83(3): 480-481. 1994.
Karchmer AW. Sexually transmitted disease. Scientific American Medicine 7: XXII:11-13. June 1994.
Kirkman R and Chantler E. Contraception and the prevention of sexually transmitted diseases. British Medical Bulletin 49(1):171-181. 1993.
Korn A and D Landers. Gynecologic disease in women infected with human immunodeficiency virus type 1. Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology 9:361-370. 1995.
Korn A and others. Pelvic inflammatory disease in human immunodeficiency virus-infected women. American Journal of Obstetrics and Gynecology 82:765-8. 1993.
Mesiwala S and V Parks. Guidelines for women with HIV/AIDS. Project Inform Fact Sheet. October 1994.
Moorman A and others. The influence of HIV infection on the microbiology and histopathology of acute pelvic inflammatory disease (PID): Final results of a multicenter study. ICAAC. San Francisco, CA. September 17-20, 1995. Abstract K94.
Safrin S and others. Seroprevalence and epidemiologic correlates of human immunodeficiency virus infection in women with acute pelvic inflammatory disease. Obstetrics and Gynecology 75:666-670. 1990.