Methicillin-resistant Staphylococcus aureus (MRSA) sepsis complicated by warm autoimmune haemolytic anaemia secondary to antimicrobial therapy

Abstract

A 61-year-old Caucasian woman presented to the emergency room complaining of left-sided chest pain and altered mentation for 3 days. Her medical history included liver cirrhosis and coronary artery disease. On admission, she was found to have methicillin-resistant Staphylococcus aureus (MRSA) bacteraemia. Due to a decline in mental status, a lumbar puncture was performed and cerebrospinal fluid cultures grew MRSA. She was treated initially with vancomycin. Ceftaroline was later added, due to the high burden of disease and difficulty in clearing her infection. After initiation of ceftaroline, bacteraemia cleared and mental status improved, however, she developed haemolytic anaemia. Ceftaroline was stopped and vancomycin continued. Staphylococcal meningitis is a rare occurrence, estimated at a rate of only 1%–10% of all bacterial meningitis cases. Ceftaroline seems to be a suitable option for disseminated MRSA infection, including MRSA meningitis, when the clinical response to vancomycin is inadequate. Further studies are warranted in order to establish adequate dosing while avoiding adverse effects.

Background

This case reflects how important it is to find good alternative treatments for methicillin-resistant Staphylococcus aureus (MRSA) disseminated infection, when not responding to traditional therapy.

Case presentation

A 61-year-old Caucasian woman presented to the emergency room with complaints of left-sided chest pain and left hip pain of 3-day duration. Her medical history was significant for liver cirrhosis secondary to hepatitis C infection and alcoholism, coronary artery disease and prior tobacco use. Vital signs at presentation showed a temperature of 36.2°C, blood pressure of 103/50 mm Hg, a heart rate of 82 beats per minute and a respiratory rate of 28 breaths per minute. She appeared dry, in respiratory distress, anxious and was writhing in pain. She was confused but able to move all her extremities, had intact reflexes, no nuchal rigidity, no cranial or peripheral neuronal deficits. Laboratory investigations revealed a white blood cell count of 21.5×10³/μL, haemoglobin 10.3 g/dL, platelet count 83 000/mm³ and a lactic acid of 5.50 mmol/L. She was initially admitted to the medical intensive care unit for septic shock management, requiring fluid resuscitation, introduction of broad-spectrum antimicrobials vancomycin and piperacillin–tazobactam, also vasopressors. An initial chest X-ray (figure 1) demonstrated a large left pleural effusion, and thoracentesis was performed. The purulent fluid was drained and a chest tube was placed. Pleural fluid cultures, as well as blood cultures, confirmed MRSA infection. Antimicrobial therapy was thus de-escalated to vancomycin alone. The patient’s mental status continued to decline and CT of her head was performed with normal results. MRI would have been the imaging of choice, however, CT scan of the head was performed instead, in order to obtain a rapid result and ensure that performing a lumbar puncture would be safe. A lumbar puncture was then performed, yielding turbid cerebrospinal fluid (CSF) (figure 2) that grew MRSA. Given disseminated MRSA infection with central nervous system involvement and persistent bacteraemia, ceftaroline was added for double coverage against high burden MRSA infection, on hospital day 3. A transthoracic echocardiogram was performed, with no visualisation of any valvular vegetations. Due to complaints of back pain, the concern for epidural abscess arose, and CT imaging of her lumbar spine was done (figure 3A,B), revealing mixed lucent and sclerotic endplate changes at the level of L2 to L3, suggestive of discitis. Her renal function declined leading to marked uraemia. The patient’s mentation continued to worsen with episodes of waxing and waning agitation. Renal replacement therapy was instituted via haemodialysis. Bacteraemia finally cleared 6 days after initiation of ceftaroline, however the patient was noted to have an acute drop in her haemoglobin from 8.7 to 4.9 g/dL. A direct Coombs test confirmed the acute development of warm autoimmune haemolytic anaemia and ceftaroline was stopped. Ceftaroline was cancelled due to proven autoimmune haemolytic anaemia. The patient had baseline anaemia, but not of the severity observed with ceftaroline. Multiple blood products including packed red blood cells, platelets and fresh frozen plasma were transfused. Her haemolysis was controlled with the addition of high-dose prednisone. The patient’s mentation improved significantly, returning to baseline. However, on hospital day 15, the patient opted out of dialysis, preferring comfort measures only, but with the continuation of antimicrobial therapy. Unfortunately, there was no repeat lumbar puncture prior to discharge as she refused any other invasive procedures. She was deemed competent to make her own medical decisions. A family meeting was held, and her family members were also supportive of her decision. She was discharged home with hospice care and her antimicrobial regimen was switched to linezolid monotherapy, orally, in order to complete 8 weeks of therapy.

Figure 1.

Chest X-ray.

Figure 2.

Lumbar puncture tubes.

Figure 3.

(A,B) CT of the lumbar spine.

Investigations

• Chest X-ray.

• CT of the spine.

• CSF analysis.

• Echocardiogram.

Treatment

• Vancomycin.

• Ceftaroline.

• Vasopressors.

• Renal replacement therapy.

• Supportive therapy.

Outcome and follow-up

Patient’s mentation improved back to her baseline and her bacteraemia cleared with our treatment, however she eventually opted out of dialysis and wanted hospice. Unfortunately, our patient passed away about 2 weeks after leaving the hospital, secondary to renal complications. There was no autopsy performed.

Discussion

Staphylococcal meningitis is a rare complication of MRSA sepsis. It represents only 1%–10% of all bacterial meningitis cases.^{1 2} Currently, two processes are thought to contribute to the development of the infection: surgical interventions (most commonly neurosurgical, including CSF devices and head trauma) and spontaneous.¹ In recent years, there has been an increasing number of reports describing MRSA meningitis, especially since the early 2000s.^{1 2} Literature reports that meningitis tends to affect more males, as well as those with comorbidities including coronary arterial disease, cerebrovascular disease, kidney and liver disease.^{1 3} Co-infections, especially endocarditis and pneumonia, are risk factors that appear to predispose to meningitis, as seen in this case.^{1 4} Clinical features most commonly reported by prevalence include fever, altered mental status and headache, but in some cases, patients have presented with seizures and septic shock.^1–3

In our patient, CSF studies (figure 2) revealed xanthochromia. Further results of the fluid analysis are included in table 1. In this setting, xanthochromia can be interpreted as secondary to the high concentration of proteins in the CSF, that is, >150 mg/dL.⁵ Gram stain revealed many neutrophils, no squamous epithelial cells and Gram-positive cocci in clusters suggestive of staphylococcus. Cultures confirmed MRSA the following day. These CSF findings met all classic characteristics of bacterial meningitis, including polymorphonuclear pleocytosis, hypoglycorrhachia and elevated CSF protein levels. For reference, a CSF glucose level <34.2 mg/dL, a CSF-blood glucose ratio <0.23, a CSF protein level >220 mg/dL, >2000 cells/μL CSF leukocytes or >1180 cells/μL CSF polymorphonuclear leukocytes were individual predictors of bacterial infection with 99% certainty or better.⁶

Table 1.

CSF analysis results

	Tube 1	Tube 4
Red blood cells	12 200 cells/μL	10 750 cells/μL
White blood cells	15 000 cells/μL	9200 cells/μL
Neutrophils	92%	81%
Lymphocytes	4%	6%
Glucose	2 mg/dL	Not reported
Protein	213 mg/dL	Not reported

Reference normal ranges: neutrophils range 0%–5%; lymphocytes range: 28%–96%; CSF glucose range: 60–80 mg/dL; and protein range: 14–45 mg/dL.

Tube 2 and 3 values not reported.

CSF, cerebrospinal fluid.

In our patient, CSF vancomycin minimum inhibitory concentration (MIC) was <0.5, blood vancomycin MIC was 1, pleural fluid vancomycin MIC was 1 and ceftaroline MIC was 0.5. As we can see, vancomycin was within the range of effectiveness. The rest of antimicrobial susceptibilities for MRSA in CSF are reflected in table 2.

Table 2.

MRSA antimicrobial susceptibility testing

	Methicillin/Oxacillin-resistant Staphylococcus aureus
Ceftaroline	Susc
Clindamycin	Susc
Erythromycin	R
Linezolid	Susc
Oxacillin	R
Tetracycline	Susc
Trimethoprim/Sulfamethoxazole	Susc
Vancomycin	Susc

MRSA, methicillin-resistant Staphylococcus aureus.

Vancomycin is the antibiotic most often used for the treatment of MRSA meningitis even though it is well recognised that it has poor penetration into the CSF.^{2 7 8} Penetration into the CSF is greatly dependent on inflammation. CSF to serum vancomycin ratio in non-inflamed meninges is 0–0.18, about 2% of blood concentration.^{3 9} Under the condition of meningeal inflammation, it has been reported that vancomycin levels in the CSF are found in higher concentration, CSF to serum vancomycin ratio of 0.36–0.48, approximately 5%.^{2 3 9} This demonstrates that there is an increase of at least twofolds of vancomycin in the CSF when the meninges are inflamed. Other than intravenous vancomycin, intrathecal/intraventricular vancomycin administration is an alternative. Nonetheless, the potential for irritation exists because of the high concentration of vancomycin.³ In our patient, the plan was to also start intrathecal vancomycin, however her mental status improved and her blood cultures cleared prior to this intervention, therefore treatment was continued with intravenous antimicrobials only. It is known that rifampin has good penetration through the blood–brain barrier.² Rifampin is most often used when the hardware that cannot be removed is found at the site of infection.⁸ Rifampin was not used in our patient as she did not have any hardware in place.

In this particular case, ceftaroline was used because our patient’s status continued to decline, with worsening mentation, despite being treated for several days with vancomycin. Vancomycin was continued and intravenous ceftaroline 600 mg every 8 hours was added. The frequency of ceftaroline had to be changed 3 days later, to every 12 hours, due to worsening renal function. The patient received a total of 6 days of ceftaroline prior to discontinuation secondary to haematological side effects. Ceftaroline binds to the penicillin-binding protein with great affinity resulting in anti-MRSA activity.⁹ As described above, the penetration of vancomycin into the CSF is poor and is greatly dependent on meningeal inflammation. In rabbit meningitis models, penetration of ceftaroline is around 9%–19%.⁹ It is reported that ceftaroline is typically well-tolerated, however about 3% of patients do experience side effects. Most have mild symptoms but there is a smaller percentage that have serious adverse effects, leading to discontinuation of ceftaroline.^{9 10} The most common adverse effects are haematological toxicity and severe dermatological reactions.¹⁰ Our patient developed warm autoimmune haemolytic anaemia that was confirmed by the Coombs test. Even though this is a rare adverse effect, it has been described with penicillin and cephalosporins.⁹ The precise mechanism by which this haematological reaction takes place is not entirely understood but it is clear that it is dose-dependent.¹⁰ Ceftaroline was used off-label in our case, for treatment of MRSA bacteraemia complicated by MRSA meningitis; however, this was done due to the high burden of disease and extreme difficulty in clearing her infection with vancomycin alone. After addition of ceftaroline, the patient’s bacteraemia cleared and her mental status rapidly improved.

Learning points.

• Metastatic infections in Staphylococcus aureus sepsis are very common. Methicillin-resistant Staphylococcus aureu (MRSA) meningitis is rare but potentially deadly if not treated promptly.

• Ceftaroline seems to be a suitable option for disseminated MRSA infection with multiple organ involvement, including MRSA meningitis, when the clinical response to vancomycin is inadequate however this remains an off-label indication at this time.

• Caution should be taken when using ceftaroline for off-label indications or for longer than 2 weeks.

• Treatment of MRSA, especially meningitis, continues to be a challenge. There are limited studies on what the right approach is.

• Further studies are warranted in order to establish adequate dosing while avoiding potentially deadly adverse effects of ceftaroline in the treatment of MRSA meningitis.