Molecular testing for detection of

Mycobacterium tuberculosis

San Francisco Department of Public Health Laboratory

Aims

• Molecular Testing for MTB: laboratory considerations and challenges

• PCR-based testing at the SF Public Health Laboratory: GeneXpert (NOT FDA APPROVED)

Challenges for NAAT use

2 FDAv approved NAA tests

• Roche PCR and Gen-Probe MTD test– the only FDA-approved options

– Roche test is leaving / has left

– MTD is expensive and time consuming

– MTD may be the most sensitive method

Homebrew / RUO tests

• All are PCR -based

• Homebrew: rather inexpensive

• Performance can be excellent (See Halse & Musser JCM 2010, NYS lab)

But:

• They can require much more initial set-up / quality control

NAAT methods: challenges

• Culture still rules: higher sensitivity

-higher specimen volume is tested by culture than by NAAT (MTD & GeneXpert may be exceptions)

-TB is a bit tougher than other organisms

-sputum often doesn’t contain many MTB organisms(compared to viral specimens (herpes, flu etc..) for example)

Another challenge: logistics (# of specimens tested)

• Not necessarily a problem for all labs

• for medium, small labs: often very few specimens per day:

• Hence:

– you either run the entire assay on one or two specimens each day,

or

– you batch specimens, run one day a week, and lengthen your turn around times (and tick a lot of MD’s off)

PCR for TB

The San Francisco Public Health Lab Experience

2007 to present

Qiagen Real-Time(uses an IS6110 target)

• Automated extraction:Roche MagNAPure LC

• LightCycler real time PCR

• Internal inhibition control

• About 5 hours avg. from specimen to printed result

The Qiagen Real-Time: Light Cycler (uses an IS6110 target)

Sensitivity (culture as gold standard):

• Smear many/numerous: 100%• Smear Few/Rare: 75%• Smear negative: 50%

Specificity : 97.6%

Data based on analysis of 108 prospective patient specimens (sputum concentrates) and 50 frozen specimen

Qiagen RUO: issues

– Sensitivity was not as good as MTD

-we were getting 0 to 5 specimens per week, but running the test only on Wednesdays

-was costing a lot of micro time for that one day whether it was 1 or 5 specimens.

2010

Switched to Cepheid, GeneXpertMTB/RIF

-- NOT FDA APPROVED! --

The Device, Gene Xpert (Cepheid)

• Single use cartridges

• Extraction and amplification: in the cartridge

• Fully Automated

Clinical Specimen

Gene Xpert,results

Using the Cepheid Gene Xpert:

Treat with NALC-NaOH and make concentrate

Nested PCR: rpoB gene

• Take product of PCR 1, use as target in reaction 2

• Increase specificity by having two sets of primers needed for amplification

• Increase sensitivity by amplifying target prior to second PCR

1.

2.

Target DNA sequence: rpoB gene

• The target of rifampin: RNA polymerase subunit B

• PCR amplifies a small region relevant for rifampin resistance; uses 5 probes to assess for mutations

probes

Cepheid MTB: positive result

• five probes

-assay has anInternal PCRControl (for inhibitionassessment)

Test gives semi-quantitative results: “high”, “medium”, “low”, “very low” and “negative”

Cepheid MTB: negative result

Cepheid validation study

Validation: Sensitivity:sputum concentrates that became culture

positive:

--13/13 smear numerous, culture positives: 100%

--30/32 smear few / rare: 94%(missed one “few” and one “rare”)

--29/40 smear negative, culture positives: 72.5%

(Sm Neg/Cult Pos sensitivity of MTD test: ~72% (according to package insert)

Specificity

• 30 negative sputum concentrates– 0/30 positive

Of those:

- 10 smear+ / culture positive MOTTs tested:

0 were reactive

--100% specificity

Summary of sensitivities described in the literature:

  smear positive smear-negative

     

Moure et al (2011), JCM ND 75.30%

     

Boehme et al (2010) NEJM 98.20% 72.50%

     

Marlowe et al (2011) JCM 98% 72%

     

Helb et al (2010) JCM 98.40% 71.70%

     

Armand et al (2011) JCM 100% 48%

Rif Resistance

• In search of specimens

• 3 MGIT samples of resistant isolates tested: all three were called correctly

• 3/85 (sputum) specimens were called Rif resistant– but phenotype testing showed otherwise

• (96.6% specific)

How we are using itHow it is going…

• Sputum only (concentrated)

• Test is run: 3 days per week– Results follow: depending on when we get specimens:

either same day or next day

• All requests must go through public health TB control dept.

• No RIF susceptibility results reported right now: not validated

Prospective results of GeneXpertSince April, 2010:

126 matched GeneXpert, culture,smear specimens

(since 4/2010)

17 confirmed MTB + :

15 positive by GeneXpert

2 negative by GeneXpert

11 positive for MOTT, neg. for MTB

11 negative byGeneXpert

109 negative byculture

109 negative byGeneXpert

3 called Rif resistantBy GeneXpert

2 Rif resistantby culture

(1 false call)

The two “misses” by GeneXpert wereeach smear-negative.

All 15 “hits” were smear-positive.

Prospective results:

• 100% sensitivity for smear-positive

• 100% specificity

• 0/2 on smear negative specimens so far

• 67% PPV for rif resistance detection

• 100% NPV for rif resistance detection

Acknowledgements

• Anna Babst, Senior Microbiologist• Jonathan Carlson, Microbiologist• Sally Liska, DrPH, Lab Director

• Masae Kawamura, MD TB Control• Houmpheng Banouvong TB Control• Luke Davis, MD, SFGH• Adithya Cattamanchi, MD, SFGH