Download Cross-Canada Spread of Methicillin

819
Cross-Canada Spread of Methicillin-Resistant Staphylococcus aureus via
Transplant Organs
Lynn Johnston, Linda Chui, Nicholas Chang,
Sheila Macdonald, Margaret McKenzie,
William Kennedy, David Haldane, Robert Bethune,
Geoff Taylor, Martha Hanakowski, and Gregory Tyrrell
From the Queen Elizabeth II Health Sciences Centre, the IWK-Grace
Health Centre for Children, Women, and Families, and the Departments
of Medicine, Pathology and Laboratory Medicine, and Pediatrics,
Dalhousie University, Halifax, Nova Scotia; and Provincial Laboratory
of Public Health, University of Alberta Hospital Microbiology
Laboratory, Department of Laboratory Medicine and Pathology,
Division of Infection Control, and Department of Medicine, University
of Alberta, Edmonton, Alberta, Canada
We report our investigation of the transmission of methicillin-resistant Staphylococcus aureus
(MRSA) through transplantation. The kidneys, liver, and corneas were harvested from a child who
died in Nova Scotia. Several days postmortem it was learned that culture of a premortem endotracheal tube aspirate from the donor yielded MRSA. Both kidneys were transplanted into a child in
Nova Scotia and the liver into a child in Alberta. Both recipients subsequently became blood
culture–positive for MRSA. One corneal ring from the donor was MRSA-positive. All four MRSA
isolates were mecA-positive by polymerase chain reaction (PCR). The relatedness of the MRSA
isolates was examined by restriction fragment length polymorphism (RFLP) analysis, a 16S–23S
ribosomal PCR typing method, and comparison of antibiograms. Results were identical for all four
MRSA isolates. These findings indicate that MRSA from the donor was transferred to recipients
during implantation of harvested organs in Alberta and Nova Scotia, a cross-Canada spread.
Infectious complications represent a significant cause of
morbidity and deaths among transplant recipients. The potential sources of pathogens are several [1, 2]. Transplant recipients are at risk for acquiring organisms that may be responsible
for active or latent infection in the donor at the time of organ
retrieval. Organs and tissues may become microbially contaminated during harvesting, processing, preservation, and transportation. Transplant recipients are also at risk of reactivation
of their own latent infections.
Very frequently it is difficult, if not impossible, to determine
the origin of the infectious agent: the donor, exogenous
sources, or the latently infected recipient. However, it has been
suggested that donor-to-recipient transmission of bacterial infection most commonly is due to organ/tissue contamination
during processing and preservation [3]. Recently, molecular
techniques have enabled us to determine the relatedness of
bacterial isolates and whether the bacteria cultured as part of an
epidemiological investigation are derived from a common
source [4].
Staphylococcus aureus has been documented as a bacterial
contaminant of donor kidneys and corneas [3, 5, 6] and has
been linked to infection of kidney and cornea recipients [6, 7].
Received 1 December 1998; revised 17 May 1999.
Reprints or correspondence: Dr. Lynn Johnston, QEII Health Sciences Centre, Room 5014 ACC, 1278 Tower Road, Halifax, Nova Scotia, B3H 2Y9
([email protected]).
Clinical Infectious Diseases 1999;29:819 –23
© 1999 by the Infectious Diseases Society of America. All rights reserved.
1058 – 4838/99/2904 – 0016$03.00
Methicillin-resistant S. aureus (MRSA), although an increasingly frequent and serious nosocomial pathogen, has been
reported to be the cause of donor-to-recipient infection only
once [8].
Microbial contamination of donor organs and tissues appears
to occur considerably more commonly than subsequent recipient infection [1, 2]. When kidney recipients have developed
transplantation-transmitted bacterial infection, the outcome has
tended to be poor, with death or transplant nephrectomy occurring in the vast majority of cases [1].
This study describes cross-country donor-to-recipient transmission of MRSA, confirmed by molecular typing of the donor
and recipient isolates.
Background
On 10 February 1997 the Infection Control Department at
the Queen Elizabeth II Health Sciences Centre (QEII), in
Halifax, Nova Scotia, was notified by its microbiology laboratory of an MRSA-positive culture from a donor’s corneal ring.
On 3 February 1997 the Infection Control Department at the
University of Alberta Hospital (UAH; Edmonton, Alberta,
Canada) had been notified by its microbiology laboratory of an
MRSA-positive culture of blood that had been obtained immediately postoperatively (on 30 January) from a liver transplant
recipient. Investigations were undertaken to determine the
source of the MRSA and whether nosocomial transmission had
occurred. Organ and tissue retrieval from the donor had taken
place at the IWK-Grace Health Centre for Children, Women,
and Families (IWK), also in Halifax. The kidneys and corneas
were transplanted in Halifax and the liver was sent to the UAH.
820
Johnston
Materials and Methods
Epidemiological investigation. The charts and microbiological data of the organ donor and the four recipients of organs
and tissue (liver, kidneys, and 2 corneas) were reviewed. The
donor MRSA isolate and MRSA isolates from the kidney and
liver recipients’ blood and from the donor corneal ring were
compared at a phenotypic level (antibiograms) as well as at a
genotypic level to determine genetic relatedness. In addition, a
strain (Halifax A) that caused an MRSA outbreak at the QEII
the previous year, another Halifax MRSA strain (Halifax B),
two MRSA strains (Aberdeen A and B) from another area of
Nova Scotia, and two strains from the UAH (UAH A and B)
were subjected to the same assays to assess relatedness.
It had initially been questioned whether the donor had been
resuscitated at the hospital where Aberdeen strains A and B had
been isolated. This possibility was later excluded, and it was
determined that there were no epidemiological links between
the donor and that area of the province. To determine if
medical staff–to-patient spread of MRSA had occurred at the
UAH, six members of the surgical transplantation team were
screened for MRSA carriage, as were seven patients sharing
the same preoperative patient care unit. At the IWK, cultures of
specimens from patients sharing the same intensive care unit as
the donor and recipient were monitored for MRSA.
Bacterial strains. MRSA isolates were collected from the
kidney and liver transplant recipients’ blood cultures, a donor
corneal ring culture, and the donor’s endotracheal tube aspirate.
In addition, as stated in the previous section, two MRSA
isolates circulating in Halifax, two from another area of Nova
Scotia, and two from the UAH were collected for genotype
comparisons with the donor and recipient strains. The isolates
were demonstrated to be oxacillin-resistant with use of
Mueller-Hinton agar with 4% NaCl and oxacillin (6 mg/L), as
per National Committee for Clinical Laboratory Standards
(NCCLS) guidelines [9].
This was also confirmed by amplification of the mecA gene
with use of PCR as previously described [10]. The template for
PCR was prepared as follows. Colonies growing on a blood
agar plate were scraped off and resuspended in lysis buffer
(100 mM of NaCl, 10 mM of Tris-HCl [pH, 8.3], l mM of
EDTA [pH, 9.0], and 1% Triton X-100 [Sigma Chemical, St.
Louis, MO]). The suspension was boiled for 10 minutes and
then subjected to centrifugation at 16,000g for 10 minutes.
Approximately 2 mL of the supernatant was used as template in
a 50-mL PCR reaction [10]. The susceptibilities of the MRSA
isolates to cefazolin, oxacillin, ciprofloxacin, clindamycin,
erythromycin, trimethoprim-sulfamethoxazole, and vancomycin were determined according to NCCLS guidelines, with use
of disk diffusion [9].
Molecular typing. The MRSA isolates were typed at the
molecular level by intergenic spacer PCR (ITS-PCR) and restriction fragment length polymorphism (RFLP) analysis by
pulsed-field gel electrophoresis (PFGE). For ITS-PCR, the
CID 1999;29 (October)
MRSA target DNA was extracted by means of standard procedures. ITS-PCR was performed as previously reported [11,
12]. Amplification was carried out with an automated thermal
cycler (Perkin-Elmer Cetus, Norwalk, CT). All amplicons were
analyzed on a 1.5% agarose gel.
In preparation of samples for PFGE, bacteria were embedded in 1.5% low-melting-point (LMP) agarose [13]. The agarose plugs were lysed in a lysostaphin lysing solution (0.25 mol
of EDTA [pH, 9.0], 1% lauroylsarcosine, and lysostaphin [10
mg/mL]) for 30 minutes at 37°C. The plugs were then incubated in ESP solution (0.5 mol of EDTA [pH, 9.0], 0.5%
lauroylsarcosine, and proteinase K [0.5 mg/mL]) for 24 hours
at 50°C. After this, the plugs were rinsed twice with TE buffer
containing 1 mM of phenylmethylsulfonylfluoride, followed by
three washes in TE buffer. They were then preincubated in Sma
I restriction buffer (Gibco BRL, Burlington, Ontario, Canada)
for 35 minutes, after which time the buffer was removed and
replaced with 100 mL of fresh buffer containing 20 units of
Sma I (Gibco BRL). Digestion was carried out at 25°C for 1.5
hours. Plugs were then subjected to PFGE with use of 1%
agarose and the CHEF-DRIII system (BioRad Laboratories,
Mississauga, Ontario, Canada) with an initial ramping time of
5 seconds and a final ramping time of 25 seconds for 24 hours
at 10°C.
Results
Organ and tissue retrieval and transplantation took place
over a 6-day period between 29 January and 3 February 1997.
Donor. A 23-month-old girl was transferred from her local
hospital to the IWK on 26 January 1997 with severe ischemic
encephalopathy secondary to foreign body aspiration. She was
stabilized at her local hospital and transported by air within 12
hours. At the time of intubation she was noted to reflux gastric
contents and aspirate. Chest radiography revealed bilateral
pneumothoraces, pneumomediastinum, and progressive pulmonary infiltrates, felt due to pulmonary hemorrhage. She continued to do poorly neurologically and was pronounced braindead on 29 January. Liver, both kidneys, and both corneas were
then harvested for transplantation. Throughout her hospitalization she was afebrile and infection was not suspected clinically.
However, a blood culture performed 27 January later yielded
Klebsiella pneumoniae (reported 30 January), and an endotracheal tube aspirate culture done 29 January yielded both
K. pneumoniae and MRSA. She had previously been in good
health, with no prior hospitalizations. She was an only child
whose parents were not employed in the health care field.
Organ recipient number 1. A 14-year-old girl from New
Brunswick with chronic renal failure secondary to glomerulosclerosis received both kidneys. She had been dialysisdependent since 1995, and most recently her condition had
been managed with hemodialysis. She was admitted to the
IWK on 29 January and underwent transplantation several
hours later, on 30 January. Several days previously, she had
CID 1999;29 (October)
Spread of MRSA via Transplants
been bacteremic with methicillin-susceptible S. aureus, and on
admission she was receiving cephalexin. Postoperatively she
received cefotaxime (as it was now known that the donor had
been bacteremic with K. pneumoniae prior to death) and vancomycin for her previous S. aureus infection. Immediately
postoperatively, blood cultures were obtained.
There was no documentation of why these cultures were
performed. One of these three sets was positive for MRSA.
Many subsequent blood cultures were negative, and no focus of
infection other than the bloodstream was identified. She received cefotaxime for 8 days and vancomycin for 16 days.
Follow-up cultures of specimens from the axilla, groin, nares,
and perineum were all negative for MRSA.
Organ recipient number 2. A 6-month-old boy was admitted to the UAH in November 1996 for treatment of biliary
atresia. Nasal screening cultures of the patient were negative
for MRSA on admission and in December 1996. On 30 January
the patient underwent transplantation of the donor liver from
Halifax. A blood culture performed at the end of the procedure
was reported positive for MRSA 2 days postoperatively, and
vancomycin therapy was initiated. Subsequently, the patient
developed multiple abscesses involving pleural, pericardial,
and intraabdominal sites, from which MRSA was grown. He
required several drainage procedures and prolonged antimicrobial therapy.
Cornea recipients. A 25-year-old man from New Brunswick underwent cornea transplantation on 3 February at the
QEII and was discharged from the hospital the following day.
A culture from the donor’s corneal ring was positive for
MRSA. Surveillance cultures of specimens from this recipient
(nares, axilla, groin, and rectal) were all negative. The second
cornea was transplanted into a 39-year-old woman from Nova
Scotia. The culture from this donor ring was negative for
MRSA. Follow-up surveillance cultures were not done for this
recipient. Both cornea recipients received subconjunctival injections of cefazolin and tobramycin intraoperatively and tobramycin eyedrops postoperatively. Neither had infectious
complications related to the corneal transplant.
Contact tracing. Neither of the hospitals involved in the
solid organ transplants has endemic MRSA. The UAH handles,
on average, ,20 sporadic cases per year, and the IWK, ,10
cases per year. There was an outbreak of MRSA the previous
year at the QEII (where the cornea transplantations were performed), but ophthalmology patients were not involved, and
this facility also did not have endemic MRSA. There were no
MRSA carriers identified among patients or health care workers who were screened.
Bacterial strains. All donor and recipient S. aureus isolates
were mecA-positive, as determined by PCR directed toward the
mecA gene. The isolates had a similar antimicrobial susceptibility pattern, different from that of the QEII outbreak strain
of the previous year. All strains were resistant to cefazolin
and oxacillin. The donor and recipient strains were susceptible
821
Figure 1. Intergenic spacer (ITS)–PCR profiles of methicillinresistant Staphylococcus aureus (MRSA) strains. Following PCR
amplification, 5 mL of amplicon was applied to a 1.5% agarose gel for
electrophoretic separation. Lane 1 contains the donor MRSA strain;
lane 2, the donor corneal ring strain; lane 3, the liver recipient strain;
lane 4, the kidney recipient strain; lane 5, Aberdeen A strain; lane 6,
Aberdeen B strain; lane 7, Halifax A strain; lane 8, ATCC 33591
S. aureus; and lane 9, a 1-kb ladder. The donor and recipient MRSA
isolates and the Aberdeen A isolates are identical in their ITS-PCR
profiles, whereas the Aberdeen B and Halifax A MRSA isolates are
not. S. aureus ATCC 33591 was used as a positive control.
to ciprofloxacin, clindamycin, erythromycin, trimethoprimsulfamethoxazole, and vancomycin. The remaining strains
(Halifax A and B, Aberdeen A and B, and UAH A and B) were
susceptible only to trimethoprim-sulfamethoxazole and vancomycin. The susceptibility data alone suggest that the donor and
recipient MRSA strains were not related to any MRSA strains
circulating in the hospitals in question.
Molecular typing. ITS-PCR and RFLP analysis were used
to type the isolates at a molecular level. Both typing methods
identified the donor and recipient isolates as identical. ITSPCR did not reveal a difference between Aberdeen A and the
outbreak strains. Aberdeen B was clearly different (figure 1).
However, the RFLP data analysis indicated that the Aberdeen
A strain and the donor/recipient strains were possibly related if
the guidelines for interpreting RFLP patterns for PFGE, as
published by Tenover et al. [14], are used (figure 2).
There are five band differences between donor/recipient
strains and the Aberdeen A strain by PFGE. This would suggest ITS-PCR is not as discriminatory a typing method for
MRSA as PFGE. In addition, RFLP analysis clearly shows that
the two Halifax strains and the two UAH strains are not related
to the donor/recipient MRSA (figure 2).
822
Johnston
Figure 2. Restriction fragment length polymorphism (RFLP) analysis of MRSA strains by pulsed-field gel electrophoresis. Lane 1
contains molecular size markers as indicated; lane 2, the donor
MRSA strain; lane 3, the donor corneal ring strain; lane 4, the liver
recipient strain; lane 5, the kidney recipient strain; lane 6, Aberdeen
A strain; lane 7, Aberdeen B strain; lane 8, Halifax A strain; lane 9,
Halifax B strain; lane 10, UAH A strain; lane 11, UAH B strain; and
lane 12, molecular size markers as indicated. Only the donor and
recipient strains are identical in their RFLP profiles.
Discussion
This study illustrates several interesting features relating to
infectious complications arising directly from transplantation.
It conclusively documents donor-to-recipient transmission of a
bacterial pathogen through kidney and liver transplantation,
confirming that this represents a real although unquantified risk
to recipients. Such transmissions may occur more commonly
than has been recognized.
A recent report describes donor-to-recipient transmission of
MRSA through heart transplantation [8]. It may be that these
infections would not have been recognized as donor-related
had more commonly seen pathogens been isolated. Since
MRSA was not endemic to the three involved facilities (UAH,
QEII, and IWK) during the time of the transplantation, its
isolation prompted consideration that the transmissions were
from a contaminated donor source. Thus, the donor as source
should always be considered when bacterial infection occurs
early in the transplant period.
If donor-to-recipient transmission is more common than
previously appreciated, strategies for its recognition and
opportunities for prevention will have to be developed.
Preventive strategies are important both from a patient care
perspective and for infection control, especially when
antimicrobial-resistant pathogens are at issue. However, it is
not clear what the most cost-effective or feasible method of
screening would be.
Given the time constraints involved in organ transplantation,
delays to await culture results are not possible. Screening on
CID 1999;29 (October)
clinical grounds would have missed this donor’s infection, as
infection was not suspected premortem; nor was MRSA colonization suspected in the donor before organ retrieval. MRSA
was and is not known to be endemic in the area where the
donor lived. Furthermore, MRSA carriage in a healthy toddler
is considered quite unusual. Screening for MRSA colonization
would not, therefore, be routine in this situation.
As other antimicrobial-resistant bacteria are identified, they
too would have to be considered as organisms that could
possibly be spread through transplantation. The benefits of
routine culture to identify potential bacterial colonization of
donor organs have not been demonstrated. This clearly represents an area where more data are required. Whether routine
bacterial culture screening of all donors is justified in areas of
low endemicity as a preventive measure remains to be determined.
There was a delay in notifying the recipients’ institutions of
the donor’s MRSA colonization. It is not clear that this information would have altered the outcome of these transmissions.
However, in situations where recipients develop infectious
complications and culture results are negative, this might prove
to be very useful information upon which empirical antimicrobial therapy may be based.
Microbiology laboratory staff members, infection control
personnel, and transplant coordinators must work together to
develop strategies to notify recipient facilities when donors are
found to be infected or colonized with significant organisms.
This outbreak is reflective of the spectrum of potential outcomes following donor-to-recipient transmission of bacterial
pathogens. The cornea recipient had no clinical sequela despite
not receiving specific antimicrobial therapy and did not become
infected or colonized with MRSA. The kidney recipient fortuitously received “preemptive” therapy with vancomycin, beginning within hours of transplantation and before her positive
blood culture results were available. She likewise had no
morbidity related to receipt of a contaminated organ.
The liver recipient developed disseminated staphylococcal
infection complicating his transplantation. MRSA was not
identified until 2 days postoperatively, and thus vancomycin
administration was unavoidably delayed. The earlier and fortuitous administration of vancomycin in the kidney recipient
may have prevented the serious infection experienced by the
liver recipient.
The literature is scant but would suggest that the outcome of
S. aureus donor-to-recipient-transmitted infection is poor in
cases of solid organ transplantation. Doig et al. reported death
following transplant nephrectomy in one recipient and transplant nephrectomy in the second in which such infection occurred [7]. Coll et al. reported the death of a heart recipient due
to fulminant MRSA myocarditis; in this case the donor had
MRSA bacteremia [8]. Our data suggest that the course of
these infections may be modified by early antimicrobial therapy, as was demonstrated in the kidney recipient.
CID 1999;29 (October)
Spread of MRSA via Transplants
This report suggests greater specificity of RFLP analysis by
PFGE in comparison to ITS-PCR. ITS-PCR of the donor/
recipient MRSA strains resulted in an identical pattern to that
of the Aberdeen A MRSA strain. This apparent genetic relatedness was not associated with an epidemiological link between the donor and the area where these isolates were obtained. However, RFLP analysis demonstrated that the donor/
recipient MRSA strain pattern was not identical to that of the
Aberdeen A strain. Close examination of the PFGE gel (figure
2) shows that there are five band differences between the
strains.
Previous investigators have found ITS-PCR (referred to as
RS-PCR in their work) to be not as discriminatory as PFGE
[15]. However, ITS-PCR can aid in determining if an outbreak
of MRSA is occurring in a particular situation, as all MRSA
connected to that outbreak should have identical ITS-PCR
profiles.
In summary, this report highlights the potential for wide
geographic spread of multiresistant bacterial pathogens through
organ and tissue transplantation. As organs are increasingly
transported across borders, this potential will grow and possibly introduce MRSA into areas of nonendemicity.
References
1. Gottesdiener KM. Transplanted infections: donor-to-host transmission
with the allograft. Ann Intern Med 1989;110:1001–16.
2. Kakaiya R, Miller WV, Gudino MD. Tissue transplant–transmitted infections. Transfusion 1991;31:277– 84.
3. Spees EK, Light JA, Oakes DD, Reinmuth B. Experiences with cadaver
renal allograft contamination before transplantation. Br J Surg 1982;69:
482–5.
823
4. Tompkins LS. Current concepts: the use of molecular methods in infectious diseases. N Engl J Med 1992;327:1290 –9.
5. Bijnen AB, Weimer W, Dik P, Oberop H, Jeekel J. The hazard of
transplanting contaminated kidneys. Transplant Proc 1984;16:27– 8.
6. Escapini H Jr, Olson RJ, Kaufman HE. Donor cornea contamination with
McCarey-Kaufman medium preservation. Am J Ophthalmol 1979;88:
59 – 62.
7. Doig RL, Boyd PJR, Eykyn S. Staphylococcus aureus transmitted in
transplanted kidneys. Lancet 1975;2:243–5.
8. Cole P, Montserrat I, Ballester M, et al. Epidemiological evidence of
transmission of donor-related bacterial infection through a transplanted
heart. J Heart Lung Transplant 1997;16:464 –7.
9. National Committee for Clinical Laboratory Standards. Approved standard
M7-A3. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. Villanova, Pennsylvania: National Committee for Clinical Laboratory Standards, 1995.
10. Kobayashi N, Wu H, Kojima K, et al. Detection of mecA, femA, and femB
genes in clinical strains of staphylococci using polymerase chain reaction. Epidemiol Infect 1994;113:259 – 66.
11. Jensen MA, Webster JA, Strauss N. Rapid identification of bacteria on the
basis of polymerase chain reaction–amplified ribosomal DNA spacer
polymorphisms. Applied Environ Microbiol 1993;59:945–52.
12. Tyrrell GJ, Bethune RN, Willey B, Low DE. Species identification of
enterococci via intergenic ribosomal PCR. J Clin Microbiol 1997;35:
1054 – 60.
13. Chang N, Chui L. A standardized protocol for the rapid preparation of
bacterial DNA for pulsed-field electrophoresis. Diagn Microbiol Infect
Dis 1998;31:275–9.
14. Tenover FC, Arbeit RD, Goering RV, et al. Interpreting chromosomal
DNA restriction patterns produced by pulsed-field gel electrophoresis:
criteria for bacterial strain typing. J Clin Microbiol 1995;33:2233–9.
15. Kumari DNP, Keer V, Hawkey PM, et al. Comparison and application of
ribosome spacer DNA amplicon polymorphisms and pulsed-field gel
electrophoresis for differentiation of methicillin-resistant Staphylococcus aureus strains. J Clin Microbiol 1997;35:881–5.