Download Impact of New Diagnostic Technologies in the Clinical Microbiology

1/9/2014
Impact of New Diagnostic Technologies
in the Clinical Microbiology Laboratory
Randall J. Olsen, M.D., Ph.D.
Director, Molecular Diagnostics Laboratory
Co-Director, Microbiology Laboratory
Department of Pathology and Genomic Medicine
Houston Methodist Hospital
Houston Methodist Research Institute
Texas Society of Pathologists
93rd Annual Meeting
Houston, TX
January 17, 2014
Traditional microbiology
• Culture conditions and
growth characteristics
• Staining pattern and
microscopic morphology
• Biochemical phenotype
– Inexact
– Time consuming
– Limited to cultivable
organisms
A laboratory technologist reading culture plates.
Strategies for innovation in
the new microbiology lab
 Matrix Assisted Laser Desorption Ionization-Time of
Flight (MALDI-TOF) mass spectrometry
 Whole genome sequencing of microorganisms
 Rapid identification of pathogenic organisms
 Detection from patient specimens or primary cultures
 Rapid inferred antimicrobial susceptibility and virulence
Implementation of new technologies may
rapidly inform treatment decisions to
improve patient care and decrease costs
1
1/9/2014
New technologies:
MALDI-TOF Mass Spectrometry
• Rapid identification of bacteria, yeast & fungi
– Minimum of 24-96 h reduced turnaround time
compared to conventional culture-based methods.
– Even more dramatic results for certain organisms
(i.e. Legionella and Mycobacterium species).
• Test directly from positive blood cultures
– Bypasses time-consuming subculture steps.
Wimmer JL, J Clin Microbiol. 2012.
Perez, K. Arch Pathol Lab Med. 2012.
Gram-negative bacteremia causes
significant morbidity and mortality
• Account for 24% of nosocomial bloodstream
infections in intensive care units.
• Survival is directly associated with timely and
appropriate initiation of antibiotic therapy.
– For every hour that appropriate antibiotics are delayed,
survival decreases by 7.6%
• Treatment is increasingly becoming complicated by
the emergence of antimicrobial resistance.
Gaynes R. Clin Infect Dis. 2005.
Kang CI. Antimicrob Agents Chemother. 2005.
Kumar A. Crit Care Med. 2006.
MALDI-TOF in our
clinical laboratory
Identify Gram negative
bacteria directly from positive
blood culture media
• Transfer positive blood culture media to a serum
separator tube.
• Centrifuge to isolate the bacteria.
• Transfer an aliquot of the bacteria from the pellet to
the spotting plate.
• Immediately perform organism identification using
the MALDI-TOF.
• Concurrently perform antimicrobial susceptibility
testing using the pellet.
2
1/9/2014
Organism identification and antibiotic
susceptibility testing directly from
positive blood culture media.
• 110 blood cultures that recovered a Gramnegative organism were tested.
• Organism identification using the MALDI-TOF.
– 98% concordance with the reference method
– Reduced turnaround time by >24 hours
Wimmer, J. J Clin Micro. 2012.
Strategies to improve
patient care
Hypothesis:
Combining new rapid MALDI-TOF
technology with real-time clinical
interpretation will significantly
enhance patient care.
Innovation plan:
MALDI-TOF & antimicrobial stewardship
• Rapid identification (MALDI-TOF) and antibiotic susceptibility
testing directly from blood cultures that recover a Gramnegative organism.
• Microbiology laboratory results are immediately reported to the
infectious diseases pharmacist who is on-call 24/7.
• Pharmacist interprets clinical applicability of the MALDI-TOF
data and contacts the treating physician to recommend optimal
antibiotic therapy.
3
1/9/2014
MALDI-TOF & antimicrobial stewardship
Study Design
Patient Eligibility and Inclusion:
Bacteremia due to a
Gram-negative organism
Perez, K. Arch Pathol Lab Med. 2012.
MALDI-TOF & antimicrobial stewardship
significantly reduces turnaround times
Perez, K. Arch Pathol Lab Med. 2012.
MALDI-TOF & antimicrobial stewardship
significantly reduces length of stay
14
12
p = 0.01
Days
10
8
p = 0.05
Pre- Intervention
6
4
Intervention
2
0
Hospital LOS
ICU LOS
Perez, K. Arch Pathol Lab Med. 2012.
4
1/9/2014
MALDI-TOF & antimicrobial stewardship
significantly reduces costs
• The MALDI-TOF
antimicrobial stewardship
program saved our hospital
1.97 million dollars during
the study period.
Pre- Intervention
Intervention
$50,000
$40,000
p = 0.009
$30,000
• The program is estimated to
save our hospital
approximately $18 million
dollars annually.
$20,000
$10,000
$0
Total hospital cost per patient
Perez, K. Arch Pathol Lab Med. 2012.
MALDI-TOF & antimicrobial stewardship
• First investigation to evaluate the impact of
MALDI-TOF paired with antimicrobial
stewardship on clinical and economic outcomes.
– Earlier initiation of targeted therapy informed by rapid
identification and susceptibility testing significantly
improved patient care and decreased LOS and expenditures.
• Collaboration is key!
– An innovative microbiology laboratory paired with
an active pharmacy department was crucial to the
success of our program.
New technologies:
Whole genome sequencing
•
First bacteria genome
H. influenzae strain Rd
Published in 1995
> 1 year
> 1 million dollars
•
Today,
1 or a few genomes can be generated within 1 day
768 genomes can be generated in 3 weeks for ~$100
•
The $10 microorganism genome will soon be a reality
Flow cell for the Illumina MiSeq benchtop sequencer.
5
1/9/2014
Genome size and complexity
• Virus: few kb -> ~200 kb
• Bacterium: ~1 Mb -> 6 Mb; haploid
• Saccharomyces: 12.1 Mb
• C. elegans: 97 Mb
• Fruit fly: 180 Mb
• Human: 3,000 Mb; ~20,000 genes
Whole genome sequencing
in the clinical laboratory
“A day in the life of the microbiology laboratory”
 130 samples collected from 116 patient cultures on
a single day
 Isolated colonies and mixed samples
 Aerobic bacteria, anaerobic bacteria, acid-fast
bacilli and fungi
Long, et al, J. Clin Microbiol, 2013
Workflow for whole genome
sequencing of microorganisms
Extract genomic DNA
(Automated, Qiagen)
4 hours
Prepare sequencing libraries
(Automated, NexteraXT kit)
4 hours
Perform sequencing reaction
(Illumina MiSeq)
Analyze data
(Custom bioinformatics process)
6-39 hours
(overnight)
1-6 hours
6
1/9/2014
Results: A day in the life of
the microbiology laboratory
Whole genome sequencing identified:
– most “unknown” organisms
(88.5% concordance with reference method)
– Mycobacterium species in two samples
10 days before the conventional method
– multiple pathogens present in several
mixed samples
Challenges: A day in the life of
the microbiology laboratory
Whole genome sequencing could not identify
• 10 organisms due to their absence from the
reference database
• The lack of a comprehensive database of
human pathogens was particularly
problematic for medically important fungi
Whole genome sequencing of
microorganisms in the clinical lab
First run: June 26, 2013
• ~400 genome sequences analyzed in the
validation study
• Bacteria, fungi and influenza A virus
• ~100 genomes generated as an ordered test
7
1/9/2014
First organism to be identified by
our clinical laboratory
Mycobacterium abscessus subsp. massiliense
• Single gene sequencing assays can not
distinguish members of the M. abscessus group
• This organism has higher treatment response
rates to combination therapy since it usually
lacks inducible resistance to clarithromycin
Whole genome sequencing of
microorganisms in the clinical lab
Routine cases
•
Organisms that can not be identified using MALDI-TOF
•
•
•
•
Mycobacterium species
Fungus
Uncommon pathogens not represented in the database
Organisms requiring further classification for optimal
patient care or public health efforts
•
Serotyping Salmonella enterica and Neisseria meningitidis
Whole genome sequencing of
microorganisms in the clinical lab
Special cases
•
Organisms suspected of comprising an outbreak
•
Investigated a cluster of invasive group A Streptococcus
infections in Wyoming (Brown, et al., Emerg Infect Dis, in press)
•
Investigated a perceived outbreak of multiple-drug
resistant Klebsiella pneumoniae infections in our hospital
8
1/9/2014
different
location/
different “outbreak
patients”
service
same
location/
different
service
Perceived outbreak of multipledrug resistant K. pneumoniae
Patient
Specimen
Disposition
kp1
Abd fluid
expired
kp2
LRT
rehab facility
kp3
LRT
expired
kp4
Urine
inpatient
not avail
LRT
expired
kp5
LRT
inpatient
kp6
LRT
inpatient
We performed whole genome sequencing to investigate the possible
clonal relatedness of strains recovered from these patients.
Whole genome sequencing
identified three distinct lineages
Phylogenetic tree based on single nucleotide
polymorphisms (SNPs) in the sequenced strains
relative to the reference strain.
Strain key:
same location/ different service
“outbreak patients”
different location/ different service
Epidemiologically linked
strains kp3 & kp4 are
distantly related!
Whole genome sequencing of
microorganisms in the clinical lab
Special cases
•
Organisms with unusual antimicrobial resistance
•
Organisms causing unusually severe or unexpected
clinical presentations
•
Investigated a fatal anthrax-like infection (Wright, et al., Arch Pathol
Lab Med, 2011)
9
1/9/2014
Previously healthy male with severe
pneumonia and death within 72 h
BAL on presentation
Blood culture
with extremely high numbers
of rod-shaped bacteria
pure culture of
Bacillus cereus group organisms
Gram’s stain
Pap stain
India ink stain
Wright, et al., Arch Pathol Lab Med, 2011
The Bacillus cereus group troika
Bacillus anthracis
Bacillus cereus
Bacillus thuringiensis
•
•
•
Very closely related
Difficult to unambiguously classify using
conventional microbiology techniques
Mobile genetic elements (plasmids) are crucial
for virulence
We performed whole genome sequencing to investigate the
molecular basis of this severe, rapidly fatal infection.
Whole genome sequencing
identified the causative agent
 Identified etiologic agent of the fatal Anthrax-like infection
B. cereus with anthrax toxin genes, not B. anthracis
 Ruled out the likelihood of bioterrorism
No foreign DNA or evidence of other genetic manipulation
 Guided infection control response
No cases of secondary transmission
 Gained new insight to molecular pathogenesis
Launched a new line of investigation
10
1/9/2014
Economics of whole genome
sequencing of microorganisms
In a typical week, our Microbiology Laboratory recovers
8 Mycobacterium species, 1 fungus and 1 unusual bacterium
Number of Type of organism
strains
Probe
cost
4
Unknown Mycobacteria
600
4
Unknown Mycobacteria
600
1
1
Send out cost for
rDNA sequencing
Total cost for
conventional ID
2400
325
3700
Unknown Fungus
150
150
Unknown Bacterium
325
Total
325
$6575
Cost of whole genome sequencing for 10 organisms: $1255
(includes all reagents, labor and controls)
Take home message:
 Although some challenges remain, new technologies
such as MALDI-TOF and whole genome sequencing
can be readily integrated into clinical laboratories
 Enhance patient care
 Reduce turnaround times
 Reduce costs
Acknowledgments
• Department of Pathology and
Genomic Medicine
–
–
–
–
–
–
–
–
–
Steven Beres, PhD
Patricia Cernoch, MT(ASCP), SM, SV
James Davis, PhD
Heather Hendrickson, MT(ASCP)
Manuel Hinojosa, MHA, MT(ASCP)
Geof Land, PhD
S. Wesley Long, MD, PhD
Dustin Meuth, MBA
James Musser, MD, PhD
• Center for Biostatistics
– Leif E. Peterson, PhD
• Department of Pharmacy
–
–
–
–
–
William Musick, PharmD
Judy Ikwuagwu, PharmD
Michael Liebl, PharmD
Dan Metzen, PharmD, MBA
Katherine Perez, PharmD
• Houston Methodist Operations
– Clare Rose, MBA, FACHE
11