Bacteria

Viruses

A number of viruses have been isolated from the upper genital tract in women with PID (Table 45.1) but their role in pathogenesis is unclear.

Clinical features

The clinical diagnosis of PID is based on the presence of lower abdominal pain, usually bilateral, combined with either adnexal tenderness or cervical excitation on vaginal examination (Fig. 45.1). A comprehensive medical history and examination including an accurate menstrual and sexual history may help to reach a diagnosis. A pelvic examination is essential and a speculum examination is useful for identifying lower genital tract inflammation and excluding foreign bodies in the vagina such as retained tampons. The poor specificity and associated low positive predictive value of this approach (65–90%) is justified because a delay in antibiotic therapy of even a few days may increase the risk of impaired fertility [8]. The risks of giving antibiotics to a woman who turns out not to have PID are low, although important differential diagnoses first need to be excluded.

Other clinical features can support a diagnosis of PID but are not essential before starting empirical therapy:

• intermenstrual or post‐coital bleeding, resulting from endometritis and cervicitis;
  
• deep dyspareunia;
  
• abnormal vaginal discharge, indicating lower genital tract infection;
  
• fever is non‐specific and usually only present in moderate to severe PID;
  
• nausea/vomiting may occur in severe PID but is more commonly associated with appendicitis.

PID caused by gonorrhoea presents more acutely and is more severe compared with chlamydial PID [9]. It is worth remembering that for every woman presenting with clinical features of PID there are two others who are asymptomatic.

Fitz‐Hugh–Curtis syndrome

Inflammation and infection of the liver capsule (perihepatitis) affects 10–20% of women with gonococcal or chlamydial PID and occasionally dominates the clinical presentation. Patients complain of right upper abdominal pain and have tenderness at the liver edge, occasionally accompanied by a hepatic friction rub.

Differential diagnosis

The main differential diagnoses are given in Table 45.2. The features that classically lead towards a diagnosis of PID are the typical ‘G string’ distribution of the pain and bilateral tenderness on pelvic examination. In bowel‐related disorders the pain tends to be higher in the abdomen and more central or to the left. Other conditions tend to give unilateral pain, at least at their onset. The main diagnoses to exclude are ectopic pregnancy and causes of an acute abdomen, which may require surgical intervention, such as appendicitis and ovarian ‘accident’ (e.g. torsion or persistent bleeding from a ruptured cyst). If the diagnosis is not clear, then empirical treatment with antibiotics should be commenced, but the patient kept under close observation to ensure that an alternative diagnosis has not been missed.

Microbiological tests

All women presenting with possible PID should be offered an NAAT to check for the presence of Chlamydia trachomatis and Neisseria gonorrhoeae on a vulvo‐vaginal swab (Fig. 45.2). The alternative enzyme‐linked immunosorbent assays lack sensitivity. NAATs for N. gonorrhoeae have greater sensitivity than culture but confirmation of a positive NAAT (using a second NAAT which has a different primer target) is required because of the risk of false‐positive results. Testing for Mycoplasma genitalium is advisable and should be performed when available since it may alter the choice of therapy.

The detection of gonorrhoea, chlamydia or mycoplasma in the lower genital tract greatly increases the likelihood of PID as the cause of lower abdominal pain, but many women with PID also have a negative infection screen from the lower genital tract.

A lack of polymorphs on a Gram‐stained smear of cervical discharge makes PID unlikely but their presence is non‐specific, i.e. the absence of polymorphs has good negative predictive value but their presence has poor positive predictive value for PID [10].

Screening for other STIs should be offered to women who test positive for gonorrhoea or chlamydia, and to those who are at higher risk of infection (e.g. two or more partners within the past year, lack of condom use or previous history of an STI). An appropriate screen would include:

• NAAT for Trichomonas vaginalis from a vulvo‐vaginal sample;
  
• endocervical swab for N. gonorrhoeae culture, which should be placed in transport medium (either Stuart or Amies) and arrive at the laboratory preferably within 6 hours but certainly within 24 hours, otherwise viability is rapidly lost;
  
• HIV antibody test; and
  
• syphilis serology.

If laparoscopy or laparotomy is performed, then specimens from the fallopian tube should also be sent requesting bacterial culture, including gonorrhoea. Chlamydia NAATs are not licensed for use for fallopian tube samples and therefore require cautious interpretation.

Radiology investigations

Transvaginal ultrasound of the pelvis may be useful where there is diagnostic difficulty. However, there are no features that are pathognomonic of acute PID. Free fluid in the pouch of Douglas is a common normal finding and is therefore not helpful. The value of ultrasonography generally lies in helping to exclude other pathology such as ectopic pregnancy, ovarian cysts or appendicitis, although it can also identify dilated fallopian tubes or a tubal abscess [11]. However, this investigation may not be readily available in an emergency setting.

MRI can assist in making the diagnosis where there is difficulty, but it is also not widely available and has not entered routine management. CT scanning in acute PID may show obscuring of the pelvic fascial planes, thickening of the uterosacral ligaments and accumulation of fluid in the tubes and endometrial canal. In the upper abdomen it can provide evidence of perihepatitis. Enhancement of the hepatic and splenic capsules on abdominal CT scan has been suggested as characteristic of Fitz‐Hugh–Curtis syndrome but it is of little value as a routine investigation.

Surgical investigation

For many years the definitive diagnostic procedure for PID was considered to be laparoscopy and it probably remains more sensitive than any other investigation currently available. In many cases there will be clear evidence of dilated hyperaemic tubes with an inflammatory fibrinous exudate covering the tubes and the fundus of the uterus. In mild cases, however, intraluminal inflammation of the tubes may be missed and significant inter‐ and intra‐observer variation in interpreting the appearance of salpingitis at laparoscopy has been reported [12]. It does enable swabs to be taken from the fimbrial ends of the tubes, which may be more accurate than endocervical swabs, but the principal benefit of laparoscopy is to exclude other diagnoses. As an invasive procedure it should be reserved for those cases where there is an element of doubt as to the diagnosis of acute PID or in cases where the patient fails to respond to antibiotics within 48–72 hours.

There is no evidence to support the routine use of hysteroscopy or endometrial biopsy in the diagnosis of acute PID. More invasive endoscopic techniques, such as fallaposcopy, may be potentially dangerous and have no place in management.

Histology and pathology

The spread of infection from the cervix to the endometrium leads to an acute, predominantly polymorph‐mediated endometritis [13]. Transcervical suction biopsy of the endometrium allows assessment of the endometrial inflammation, which correlates well with salpingitis. Unfortunately, the usefulness of this approach to diagnose PID is limited by the risk of introducing infection during the procedure, the time delay in fixing and staining the sample, and the uncertain significance of isolated endometritis.

The inflammatory response seen in the fallopian tubes depends on the underlying pathogen. Gonorrhoea infects the non‐ciliated epithelial cells but production of tumour necrosis factor and gamma interferon soon lead to collateral damage to the surrounding tissue and invasion of the submucosa. The tissue damage associated with Chlamydia is mediated primarily by the immune response to the infection that occurs as a result of a delayed‐type hypersensitivity reaction to a chlamydial heat shock protein. This is characterized by a low‐grade lymphocytic response compared with the acute neutrophil response of gonococcal salpingitis.

Recurrent infection with Chlamydia causes further immune stimulation, possibly mediated by a cross‐reaction between chlamydial and human heat shock protein 60 [14]. This exaggerated immune response following re‐exposure to Chlamydia may explain the exponential increase in the risk of tubal damage that occurs with repeated infection.

Severe inflammation is associated with tubal occlusion and the production of a tubo‐ovarian abscess or hydrosalpinx. Healing following acute inflammation may produce chronic fibrosis with associated damage to the ciliated epithelium, tubal blockage and/or pelvic adhesions. Histologically, this chronic damage produces lymphoid follicles and a mononuclear cell infiltrate.