Uncommon cause of pelvic inflammatory disease leading to toxic shock syndrome

Abstract

A 44-year-old Caucasian female with a history of endometriosis is admitted to the intensive care unit due to severe left lower quadrant abdominal pain, nausea and vomiting. With patients’ positive chandelier sign on pelvic examination, leucocytosis, elevated erythrocyte sedimentation rate and elevated C-reactive protein indicated that she had pelvic inflammatory disease (PID). PCR tests were negative for Neisseria gonorrhoeae and Chlamydia trachomatis; however, her blood and urine cultures grew Group A streptococci (GAS) with a negative rapid Streptococcus throat swab and no known exposure to Streptococcus. On further review, patient met criteria for GAS toxic shock syndrome based on diagnostic guidelines. The patient was promptly treated with intravenous antibiotics and supportive care, and she acutely recovered. This case demonstrates a rare cause of PID and an atypical aetiology of severe sepsis. It illuminates the importance of considering PID as a source of infection for undifferentiated bacteraemia.

Background

Pelvic inflammatory disease (PID) is a commonly diagnosed condition that is conventionally thought of as a sexually transmitted infection (STI).^{1 2} Most clinicians typically manage PID with a combination of medications targeting the most common causes of STI associated PID, Neisseria gonorrhoeae and Chlamydia trachomatis (GC), in addition to management of syndromic symptoms.³ However, we present a case of PID that is caused by group A beta-haemolytic streptococcus (GAS) that led to our patient developing streptococcal associated toxic shock syndrome (TSS). While polymicrobial PID is somewhat common,⁴ there are only a handful of reported cases of PID caused by GAS.⁵ This case demonstrates the importance of considering other causes of PID in an effort to prevent serious complications associated with potential pathogens that are more invasive than GC.

Case presentation

A 44-year-old gravida 2, para 2 premenopausal female was transferred to our intensive care unit (ICU) after sepsis was identified in an emergency department at her local hospital. The patient presented with a 4-day history of progressively worsening, severe lower abdominal pain. She had associated general malaise, nausea, vomiting and diarrhoea; however, she denied experiencing any fever or chills in the days preceding her presentation. The patient had a medical history significant for endometriosis treated with medroxyprogesterone. The patient had a hysteroscopy with dilation and curettage on year and 3 months prior to presentation. The patient is a never smoker and is a married homemaker. Her family history was significant for paternal type 2 diabetes mellitus and multiple relatives with autoimmune diseases.

On physical examination, the patient was noted to be afebrile, tachycardic, tachypnic and slightly hypotensive. She was in mild distress at the time of presentation but did not appear toxic. The patient had left lower abdominal tenderness to palpation with voluntary guarding but no other peritoneal signs. A pelvic examination revealed obvious purulent cervicitis with purulent discharge from cervix. A bimanual examination elicited significant cervical motion tenderness and positive chandelier sign as well as uterine and adnexal tenderness.

Investigations

On admission, our patient underwent a battery of tests to identify the source of her abdominal pain. A complete blood count initially revealed that the patient’s white cell count (WCC) was 2.3×10³/L, but on hospital day 3, the patient’s WCC increased to 13.9×10³/L. Her WCC then continued to increase for 2 days. During this time, efforts were made to identify the patient’s source of infection. A CT scan of the abdomen and pelvis was performed that showed extensive inflammatory stranding in the pelvis near the bladder and uterus that is seen in figure 1. This was also seen on later MRI imaging seen in figure 2. Based on this finding and the patient’s physical examination, it was thought that she had PID. The patient was then placed on empiric antibacterial therapy. The underlying aetiology was unclear and at that time PCR was performed to look for N. gonorrhoeae and C. trachomatis, which were both negative. HIV antibodies were tested and negative as well. A urine pregnancy test was negative.

Figure 1.

CT scan of the abdomen and pelvis that showed extensive inflammatory stranding in the pelvis near the bladder and uterus.

Figure 2.

MRI of pelvis with and without contrast that further shows inflammation and stranding near the bladder and uterus that is further seen on CT scan in figure 1.

At that time, the prior facility was contacted and patient was found to have group A beta-haemolytic Streptococcus growing in two sets of blood cultures ×2. Urine cultures collected at the prior facility were also found to be positive with group A beta-haemolytic Streptococcus. Although, repeat urine cultures at our facility exhibited no GAS growth. Throat culture was obtained and ended up growing Group F Streptococcus. A complete metabolic panel (CMP) revealed mild acute kidney injury with creatinine 1.2. The CMP also revealed elevated aspartate aminotransferase and alanine aminotransferase levels at 197 and 73, respectively. An MRI of the pelvis was performed that showed oedema and enhancement of the left iliacus muscle that was suggestive of myositis seen in figure 3. An erythrocyte sedimentation rate was found to be elevated at 120 mm/hour, and a C-reactive protein was found to be elevated at 37.8 mg/dL. A subsequent autoimmune workup was performed including antinuclear antibodies, anti-Ro, anti-La, anti-ribonucleoprotein, anti-Smith and anti-Jo 1 that were all negative.

Figure 3.

MRI showing oedema and enhancement of the left iliacus muscle that was suggestive of myositis.

Differential diagnosis

• Sepsis.

• PID.

• Salpingitis.

• Endometritis.

• TSS.

Treatment

The patient was treated with intravenous antimicrobials in addition to general supportive care in the ICU. When the patient initially presented she was treated with intravenous cefoxitin and oral doxycycline for empiric treatment of typical PID. In spite of this treatment, her condition deteriorated. Once cultures were known and with her worsening condition, she was then treated with ampicillin–sulbactam, clindamycin and gentamicin. Infectious disease along with gynaecology helped with her care during her stay in the ICU and provided recommendations for appropriate antibiotic therapy.

When the patient began developing signs and symptoms consistent with severe sepsis, intravenous fluid resuscitation was provided with lactated ringers. Additionally, the patient’s haemodynamic status was monitored closely in the ICU. Our patient did become slightly hypotensive with a mean arterial pressure of 71. However, due to prompt treatment, the patient’s blood pressure began increasing, and she did not require vasopressor support to maintain adequate organ perfusion. Patient was discharged with oral antibiotic regimen and improved clinically.

Outcome and follow-up

This patient remained in the ICU for 4 days followed by 6 days of inpatient intravenous antibiotic therapy. While the patient’s white cell count and inflammatory markers decreased, she continued to have significant pain and discomfort, which led to her prolonged hospital course. At the time of discharge, the patient remained afebrile and her cluster of symptoms continued to improve. The patient followed-up with infectious diseases, her primary care physician and her gynaecologist. She completed the course of oral antibiotics and improved with treatment.

Discussion

PID consists of inflammation of the upper genital tract involving any or all of the following structures the uterus, fallopian tubes, ovaries, peritoneal cavity and all encompassing organ structures. PID often goes undiagnosed due to the absence of symptoms in subclinical disease. The paramount importance of treating PID arises from its long-term sequalae especially infertility, ectopic pregnancy and chronic pelvic pain. Acute PID can present with nausea, vomiting and diarrhoea, which is the constellation of symptoms that initially led our patient to seek treatment.³

Differential diagnoses for PID include appendicitis, endometriosis, ectopic pregnancy, ovarian cyst and tubo-ovarian abscess. Imaging can be used to help differentiate among these potential diagnoses. In our patient’s case, there was noted extensive inflammatory stranding in the pelvis near the bladder and uterus on CT scan. This along with her physical examination and clinical picture further supported the diagnosis of PID.

First-line antimicrobial regimen for treatment of PID, as suggested by the Centers for Disease Control and Prevention (CDC), consists of doxycycline with or without metronidazole plus either ceftriaxone intramuscular in a single dose, cefoxitin or another third-generation cephalosporin. Inpatient regimen consists of intravenous cefotetan or cefoxitin plus doxycycline or clindamycin plus gentamicin. Our patient received cefoxitin and doxycycline initially. Even with proper empiric therapy, our patient clinically deteriorated. This further pointed to an uncommon aetiology. Due to this, she was then changed to ampicillin–sulbactam, clindamycin and gentamicin based on patient’s culture results. The standard of care for PID includes at least 2 weeks of oral therapy once patient clinically improves.³ In our case, the patient received 10 days of intravenous antibiotics for PID and TSS followed by 21 days of oral antibiotics for GAS-associated myositis in her iliacus muscle given her complicated clinical course.

The most common route of bacterial inoculation is ascending infection starting in the lower genital tract such as the vaginal canal. The risk factors for invasive GAS disease include age greater than 65, chronic cardiac and pulmonary disease, diabetes mellitus, HIV, cancer, corticosteroid use, alcohol abuse, injection drug use and varicella infection; none of which our patient possessed.⁶ Gynaecological literature suggests that a precursor for Streptcoccus pyogenes associated PID is the mechanical disruption of the mucosal epithelial barrier, which can be seen during postpartum state, after intrauterine device insertion, or after endometrial biopsy.⁷ Our patient did have a dilation and curettage, although it was thought to be too far out from patients presentation. It is thought that oral pathogens including S. pyogenes, Haemophilus influenzae, Streptococcus pneumoniae, anaerobes and Staphylococcus aureus can colonise the vagina for a short period of time secondary to self-contamination from the pharynx, a close contact source or gastrointestinal carriage.

In cases where pathogens are found in places that they usually do not infect, this can cause large amounts of inflammation when infection does occur.^{2 8} Prior to patient’s illness, both husband and daughter recently had an upper respiratory tract infection, although this was not confirmed GAS. This could be a possible source of infection given her close contact. In our patients’ case, throat and vaginal cultures did not match and her rapid streptococcus antigen test was positive for Group F Streptococcus. The patient did not have a recently recorded vaginal culture and could have previously been colonised with vaginal GAS prior to infection.

Here, we present a case of PID caused by S. pyogenes. PID is a commonly diagnosed syndrome. However, less than 15% of acute PID is caused by enteric or respiratory pathogens such as Group A streptococci.⁸ There are only a handful of documented cases of PID caused by S. pyogenes in postpubertal women that are not pregnant or puerperal peppering the literature.⁸

The significance in our case was the development of Streptococcal TSS is a systemic illness that occurs in the setting of PID with group A Streptococcus (GAS).⁹ The most common sources of GAS are superficial skin infections, respiratory infections and necrotising fasciitis.⁶ A rare cause of TSS is GAS vaginitis among postpubertal women. Typically, TSS from GAS, there is higher incidence of invasive infections in pregnant and postpartum women with a 20-fold increase in incidence.¹⁰ However, due to the risk of GAS vaginitis progressing to streptococcal TSS any woman diagnosed with GAS vaginitis needs to be treated promptly.⁸

Cases have been reported of postmenopausal women with streptococcal TSS due to GAS vaginitis. This is likely thought to be secondary to vaginal atrophy in the postmenopausal period which leads to disruption in mucosal defences and changes in vaginal pH that increase susceptibility to infection of the lower genital tract. However, our patient was noted to be premenopausal.⁸

Our patient met the CDC criteria for streptococcal caused TSS which includes hypotension defined as a systolic blood pressure less than or equal to 90 mm Hg and multiorgan failure characterised by two or more organs involved. This is defined by renal impairment, coagulopathy, liver involvement, acute respiratory distress syndrome, a generalised erythematous macular rash that may desquamate and soft tissue necrosis that includes myositis. Also included in the criteria is the isolation of group A Streptococcus.¹¹ The standard treatment consists of clindamycin plus penicillin. Length of treatment varies on a case-to-case basis. However, patients with bacteraemia should be treated for at least 14 days.¹² Adjunctive therapy with intravenous immunoglobulin, antitumour necrosis factor and or hyperbaric oxygen can be used if no improvement is seen with antibiotic therapy.¹³ In our patient’s case, she was improving with ampicillin–sulbactam and she did not require adjunctive treatment. In rare occurrences, PID can be polymicrobial since a respiratory pathogen was involved and anaerobes could not be ruled out, our patient was continued on ampiclillin–sulbactam and changed to amoxicillin–clavulanate on discharge.¹⁴

Learning points.

• Group A Streptococcus is a uncommon cause of pelvic inflammatory disease (PID).

• Streptococcal toxic shock syndrome can be a complication of PID.

• If patient worsens with empiric therapy for PID, then uncommon causes need to be considered.