A Patient with autoimmune hepatitis and transverse myelitis presented with persistent Staphylococcus aureus bacteremia, the discrepancies in assessing susceptibility; VISA versus Non-VISA

Vancomycin-Intermediate Staphylococcus aureus (VISA) is still uncommon among MRSA isolates. In our region, we rarely encounter a case of VISA and/or GISA bacteremia. Here, we report a man who suffered from autoimmune hepatitis on immunosuppressive therapy and thoracic transverse myelitis suspected to be due to polyomavirus infection; he developed persistent MRSA blood stream infection, PVL-positive and MLST clonal complex 88 which is reported most commonly from Africa. A strain with Vancomycin susceptibility of 4 – 6 μg/ml (VISA) was initially identified, retested again elsewhere and showed MIC of 2μg/ml and Teicoplanin susceptibility of 4μg/ml. Treatment failure occurred while attaining higher serum vancomycin levels than recommended and died.


Introduction
Soon after the initial cases of Vancomycin-Intermediate Staphylococcus aureus (VISA) were reported form Japan in1997, it relentlessly seems to be reported worldwide, predisposing for more vancomycin resistant strains i.e. VRSA, though it is still a rare sporadic isolate. However, VISA strains, a more frequent isolates compromise glycopeptides activity and complicate the treatment regimens of patients infected with these strains [1].In an international study, hVISA phenotype (defined as the presence of subpopulations typically at a rate of 1 organism per 10 5 -10 6 of MRSA with VISA) occurred in more than one-quarter of MRSA blood isolates from patients with infective endocarditis, though it varied in frequency by geographic region [2,3].In Taiwan, isolates from ten centers collected 1000 MRSA, the prevalence of VISA isolates were 0.2% and hVISA isolates 0.7% [4].
In our region, reports from the Arab Countries have been published; In Jordan a case of VISA (MIC = 4 mg/dl) was reported in the year 2006 by F. Bakri and coworkers in a patient with erythrodermic psoriasis who suffered three months persistent Staphylococcus aureus bacteremia [5]. MRSA strains with decreased susceptibility for vancomycin were reported from Qassim region in Saudi Arabia [6]. VISA, hVISA persistent blood stream infection in spite of high blood vancomycin levels were reported from Saudi Arabia as well [7]. Another interesting report came from Oman on the treatment of infective endocarditis due to VISA; a patient was treated with a combination of intravenous linezolid and fusidic acid without a need for surgical intervention and resulted in complete cure [8]. Moreover, a report on glycopeptide-intermediate Staphylococcus aureus (GISA) were reported form north Lebanon, five strains out of 113 (4.4%) were resistant to teicoplanin (GISA) by MIC dilution method and no strain was resistant to vancomycin. Although all MRSA isolates were reported as susceptible by Muller-Hinton agar diffusion method [9], we should consider other methods for testing MRSA for MIC when it is needed.
Here, we present a patient who was immunosuppressed and had had suffered from persistent Staphylococcus aureus blood stream infection in spite of being on unintentional high serum levels of vancomycin. The isolate show discrepant susceptibility testing between different laboratories, once was interpreted as VISA, at another institution was not.

Case Report
A 47 year old male patient from Yemen with type 2 diabetes mellitus on insulin, and autoimmune hepatitis since 1996 on Immunosuppressive therapy; dexamethasone 4 mg/day and azathioprine 50 mg twice daily. He was admitted on November 1, 2014 in Jordan hospital complaining of bilateral lower limb weakness of one month duration,it was sudden in onset and progressive, it was noticed when lifting heavy objects, and later falling down spontaneously on walking. Dorsal spine MRI imaging showed lesion in the anterior part of dorsal spine (Figure 1). Differential diagnosis included viral transverse myeli- tis, demyelination, and low grade neoplasm.
He was transferred from another hospital with decrease level of consciousness; he was admitted to the ICU, and on admission he was sleepy, generally week and could not move his legs. CSF analysis showed WBC = 44/mm 3 , RBC =100 /mm 3 , Protein=308 mg/dl, Glucose =148 mg/dl, he was started on ceftriaxone 2gm/12 hours and vancomycin 1gm/12 hours. Dorsal MRI showed a lesion of the anterior part of the spinal cord ,extending opposite the lower part of D5 to the lower part of the D6 vertebral bodies, it show no significant enhancement in the post contrast images (Figure 1). Dexamethasone was discontinued on November 3 rd and methylprednisolone 1gm/24hours was started. On November 6 th he started fever (38.6 0 C orally) when an infectious diseases consultation was sought. Laboratory tests for Clostridium difficile toxins, CMV IgM, toxoplasma IgM, urine legionella antigen, mycoplasma IgM, and Chlamydophila pneumoniae IgM, HIV antibodies, HCV Antibodies and HBsAg were all negative. Urine sediment for decoy cells was positive, and BK-Polyomavius by Taq-Man real time PCR (Anatolia GeneworkInc, Istanbul, Turkey) was positive. Immunosuppressive therapy was advised to be tapered and ceftriaxone and vancomycin were discontinued as aerobic blood cultures twice showed no growth with abating fever. Few days later he redeveloped fever, and blood culture grew Enterobacter cloacae, he was started on pipracillin/ tazobactam 4.5 gm/8hours, fever subsided for several days and the patient was relatively stable with his baseline neuro-deficit. A repeat dorsal spine MRI showed no significant changes from previous imaging and no evidence of new lesions.
About a week later, the patient developed high fever (38.4 0 C), and his blood cultures grew methicillin-resistant Staphylococcus aureus (MRSA), vancomycin 2gm IV loading followed by 1gm/12hours was started, 2D-echocardiogram was normal, and bone scan was reported as multiple abnormal areas of increased radiotracer uptake seen at both sterno-clavicular joints, both knee joints, and irregular tracer uptake is seen at D5-D6 and D12 vertebral bodies. Five days into vancomycin treatment; on November 19 th fever persisted, and empiric fluconazole 600mg IV loading then 400mg IV/24hours was added, vancomycin trough levels were monitored aiming to keep levels 15 -20 µg/ml, althoughin most instances levels were unintentionally elevated due to measured deterioration in renal function (Figure 3) (Figure 2). Linezolid 600 mg IV/12hours and imipenem/cilastatin 250 mg/6 hours were started; the later and fluconazole were discontinued as repeat blood cultures showed the same S. aureus isolate. The patient showed continuous clinical de- terioration, white lungs on chest radiogram, poor arterial blood gases, electrolytes imbalance, further renal function deterioration, intractable shock and met his demise.

Discussion
In our region, we rarely document patients with MRSA-induced sepsis with elevated or creeping MIC phenomena, and persistent bacteremia while on vancomycin treatment, due to resources limitations. A case was presented by my group (Rula Rashed MD, Fadi Shaqlous PharmD) earlier in2010, in a regional meeting; the isolate demonstrated creeping vancomycin MIC [10] while the patient was on treatment with therapeutic vancomycin levels, MIC creped in a month time from 0.125 µg/ ml, to 0.75 µg/ml, to 1.5 µg/ml as was verified by E-test. The patient continued to have pus discharge from a traumatized combat-related infected foot wound with repeated surgical debridement, till he was switched to linezolid with almost brisk response and discharged home (unpublished report).
Our current patient was immunocompromised, and had thoracic spinal cord transverse myelitis that was suspected to be due polyomavirus infection, which may have been predisposed by immunosuppressive therapy. Again he showed lack of response to vancomycin therapy, though the strain was within the susceptibility range, and vancomycin serum levels were much higher than what was desired (Figure 3). Failure of vancomycin therapy for VISA strains is well described, as the target attainment (AUC/MIC ≥ 400) cannot be reached [11]. Higher failure rates were also reported with "creped" vancomycin MIC, though still within susceptibility ranges of ≤ 2 µg/ml; the upper CLSI limit for vancomycin susceptibility [12,13]. Our S. aureus isolate as retested in (Antimicrobial Resistance and Healthcare-Associated Infections Reference Unit. Public Health England, 61 Colindale Avenue, London NW9 5EQ), it exhibited vancomycin susceptibility and did not meet CLSI MIC level for GISA [12,14].The isolate belongs to spa type that has been associated with PVL-MRSA and belonging to MLST clonal complex 88 reported most commonly from Africa. This mecA positive S. aureus encodes Pantone-Valentine leucocidin (PVL) genes. Infection with PVL-SA most commonly presents as primary cutaneous lesions [15,16]. However PVL-positive S. aureus was reported to be associated with serious life threatening disease, namely community-associated pneumonia (CAP) in four patients, all did not have the classical risk factors for infection with strains carrying PVL genes [17].The patient was started on Linezolid but the duration of therapy was not enough to appreciate a response, and he died. In our case, a lesson to be learned; laboratories should pay attention for microorganisms' antimicrobial susceptibility testing, taking into account the imported tests kits source, storage and quality; here the E-tests strips that showed discrepant MIC values for .S aureus in the same laboratory and tests strips brand, though labeled as not expired; E-test (bioMérieux Marcy l'Etoile. 376, chemin de l'Orme 69280 Marcy l'Etoile).
Another lesson learned from this case, that a persistent fever in -compromised patients is a risk factor for the development of more antibiotic resistant patterns of S. aureus infections and moreover in PVL-positive strains. This study also suggests the early implementation of antibiotic treatment based on suspicion of the frank pathogen, in order to cover strains with higher MICs, before the actual MIC is available [18].