Panther Hologic Service Manual
Hologic instruments are adaptable to changing workflow and diverse testing needs for all types of molecular labs. The Panther ® system: Unprecedented agility on a consolidated platform. The Panther system is a fully automated testing platform with true sample-to-result automation, adaptable workflow options and a consolidated testing menu. MODEL QDR®4500 QDR FOR WINDOWS® FAN BEAM X-RAY BONE DENSITOMETER. Hologic, Inc. Has made all reasonable effort to ensure that the information in this manual is accurate and complete. Hologic, Inc. Shall not, however, be liable for any technical or editorial errors or omissions contained herein, or for incidental, special or consequential.
- Panther Hologic Service Manual Pdf
- Service Manual Tv
- Panther Hologic Service Manual Online
- Hologic Panther Maintenance
- Service Manual Bmw
We would like to show you a description here but the site won’t allow us. View & download of more than 75 Panther PDF user manuals, service manuals, operating guides. Air compressor user manuals, operating guides & specifications. Service Technicians C-Arm Forum Manual For Hologic Insight 2 tri lam SJOH I need service manual for Hologic Insight 2, please. Run HIV-1, HCV and HBV in parallel with sexually transmitted disease/sexually transmitted infection (STD/ STI) testing assays on the Panther system: CT/NG, HPV and Trichomonas vaginalis. Run up to 4 assays in parallel. Deliver continuous results to physicians without waiting to fill a batch; Comprehensive information for complex coinfections.
Abstract
To evaluate the performance of the Hologic Gen-Probe (San Diego, CA) PANTHER system.
Free sites like google earth maps. You’ll hardly find a person who is still using paper maps while driving. Let’s find out what online services you may use on your desktop or portable gadget to find a direction or a bus route.SEE ALSO:SEE ALSO: Google MapsNo doubt, is the most famous web mapping service on the Net. Numerous online maps have almost squeezed their predecessors and have got their space on almost every smartphone.
The performance of PANTHER was compared with the Hologic Gen-Probe TIGRIS and/or Roche (Indianapolis, IN) COBAS AMPLICOR systems through testing of patient specimens and the spiked-urine matrix.
After discrepant resolution, PANTHER demonstrated a 99.3% (95% confidence interval [CI], 96.0%–99.9%) positive and 100% (98.5%–100.0%) negative agreement for Chlamydia trachomatis (CT) and 100% (96.6%–100.0%) positive and 100% (98.6%–100.0%) negative agreement for Neisseria gonorrhoeae (NG) for all male, female, unsexed, and NG-spiked female urine specimens combined. For other specimen types collectively, the PANTHER demonstrated 100% (95% CI, 90.6%–100.0%) positive and 100% (88.3%–100.0%) negative agreement for CT and 90.9% (62.8%–98.4%) positive and 100% (93.5%–100.0%) negative agreement for NG. Analytical sensitivity of the PANTHER in urine matrix was similar to the TIGRIS system.
The PANTHER system provides an excellent new addition to options for detecting CT and NG, is appropriate for testing urine samples, and will facilitate high-throughput testing in the clinical laboratory.
The prevalence of Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) infection continues to increase or remains high around the world.1–3 Therefore, the US Preventive Services Task Force,4 Centers for Disease Control and Prevention,5 and professional organizations6 have advocated an intensive screening effort to reduce disease burden and prevent the sequelae of untreated infection, such as pelvic inflammatory disease and associated infertility. For example, recent recommendations in the United States include screening all sexually active women younger than 25 years for CT.4
Rapid, sensitive, and accurate nucleic acid amplification testing methods are needed to respond to this public health mandate. Several CT/NG nucleic acid amplification testing platforms have been introduced over the past decade. However, some assays have shown suboptimal performance characteristics (ie, cross-reactivity with nongonococcal Neisseria species). This is of special concern in a lower-prevalence screening population because of the consequent low positive predictive value of these tests.7
Recently, the Hologic Gen-Probe (San Diego, CA) APTIMA Combo 2 (AC2; PANTHER system) was cleared by the US Food and Drug Administration (FDA) in the United States for the detection of CT and NG in vaginal and endocervical swabs, male urethral swabs, and ThinPrep PreservCyt (Hologic) cervical cytology specimens. To our knowledge, there has been only one previously published study examining the performance of the AC2 assay on the PANTHER system, limited to examination of the analytical specificity for NG.8 However, extensive literature supports the robust performance of the AC2 assay run in manual, semiautomated, and fully automated TIGRIS DTS systems (Hologic Gen-Probe).9,10 Therefore, here we sought to verify the clinical performance characteristics of the Hologic Gen-Probe AC2 (PANTHER system) assay (hereafter referred to as PANTHER) and to determine whether performance characteristics of the AC2 were affected by the platform on which it was run. Hence, specific comparison was made to performance on the TIGRIS DTS platform. Additional comparisons were also made for a subset of patients who were initially tested for CT/NG using the Roche (Indianapolis, IN) COBAS AMPLICOR system (COBAS) and for whom all NG+ results were confirmed by additional testing. Notably, the PANTHER was not cleared by the FDA for testing female and male urine specimens. Therefore, we also sought to validate off-label use of urine specimens on the PANTHER to expand the sample types available for outpatient screening.
Materials and Methods
Characteristics of Clinical Samples and Testing Methods
The described analysis was performed as part of normal quality assurance activities related to verification and validation of methods being considered for adoption in our clinical laboratory practice. The underlying intent of these studies was not to generate generalizable knowledge. As such, institutional review board approval was not required. Five neighboring institutions (Brigham and Women’s Hospital, Cambridge Hospital, Boston Medical Center, Massachusetts General Hospital, and Needham Harvard Vanguard) supplied us with deidentified coded samples that were previously tested and scored as positive or negative for CT and NG using the TIGRIS. Sample types included male urethral and female endocervical swabs, male and female urine samples, and a small number of ThinPrep PreservCyt samples. Urine samples from one institution were not differentiated as to female or male sex and therefore are referred to as unsexed urine specimens.
Notably, the PANTHER requires use of the same sample transport media as the TIGRIS. Therefore, all samples previously collected for the TIGRIS were run on the PANTHER without modification. Samples were preselected to enrich for NG+ and CT+ specimens to evaluate positive agreement with greater power. Both the PANTHER and TIGRIS use the AC2 assay, which is based on transcription-mediated amplification of species-specific targets in the 23S ribosomal RNA (rRNA) of CT and 16S rRNA of NG. Specific amplification of these targets leads to distinct luminescent signals, which are quantified as relative light units (RLUs) and used to assign samples to negative, equivocal, and positive categories. Samples were kept frozen at −20°C after initial testing and thawed in our laboratory immediately before testing in this study.
We also tested additional male and female urine samples that were previously analyzed in our laboratory using the COBAS. These samples were stored at −20°C in their original states (ie, without addition of transport media with nucleic acid stabilizers per the FDA-cleared collection protocol for the COBAS) and preselected for analysis based on prior results (ie, to enrich for CT+/NG+ specimens). In contrast to the AC2, the COBAS uses conventional end-point polymerase chain reaction (PCR) to detect the multicopy cryptic plasmid from CT and the single-copy M.NgoPII gene from NG.11 Because of known COBAS cross-reactivity issues, any COBAS NG+ sample was scored as a true positive only after confirmation using the Light Cycler (LC)–based (Roche), real-time PCR assays specific for the NG 16S rRNA and PAP genes.7 We previously determined that these confirmatory assays were more sensitive than the COBAS assay7 and therefore can reliably conclude that COBAS NG+ confirmatory test negative samples are true negatives.
Prior to retesting by the PANTHER, these samples were thawed to room temperature and vortexed for even resuspension. Then, 3 mL was transferred to an APTIMA urine transport tube. Subsequent to PANTHER testing, these samples were retested on the TIGRIS for comparative evaluation and to verify coding by the outside laboratories. Thus, results from in-house collected urines were compared across three assay platforms (COBAS, TIGRIS, and PANTHER).
Accuracy
Of the 464 samples evaluated, 395 were urine, 54 were swabs, and 15 were ThinPrep PreservCyt samples. Among these samples, results of nine urine and three swab samples ultimately were not included due to unresolvable discrepancies, as explained in the text. Our goal was to test at least 50 CT+, 50 NG+, and 100 CT−/NG− male and female urine samples per recommended guidelines in Cumitech 31a.12 However, because of the low prevalence of NG in our patient population, we were able to obtain only a small number of NG+ female urine samples. To compensate for this limited number, we performed spiking experiments in female urine matrix.
Discrepant Resolution
Any discrepant sample initially tested on the TIGRIS was tested again on both the PANTHER and TIGRIS. Any discrepant sample initially tested on the COBAS was tested again on the COBAS, PANTHER, and TIGRIS if sufficient specimen were available. In addition, NG discrepant samples were also retested using the LC confirmatory assays described above. Discrepancies were resolved based on agreement achieved on retesting, as discussed in the Results section, taking into account the potential for specimen degradation in unpreserved specimens. Samples with an unresolved discrepancy were excluded from subsequent analysis for both CT and NG analytes.
Spiking Studies
Because of the low prevalence of NG in female patients in the greater Boston vicinity, additional analysis of potential female urine matrix effects was done through the use of spiking studies of 47 COBAS CT− and one COBAS CT+ female urine specimens. In total, 2 mL of urine from each sample was transferred into an APTIMA urine transport tube and screened for CT and NG on the PANTHER. A second set of 2-mL aliquots was spiked with 100 μL of a fresh suspension (500 colony-forming units [CFU]/mL based on McFarland conversion and subsequently confirmed by CFU counts from suspension dilutions) made from an overnight (24-hour) NG (43069, American Type Culture Collection [ATCC], Manassas, VA) agar culture. Each of these NG-spiked female urine specimens was then mixed with 2 mL of APTIMA transport media for a final concentration of 12.5 CFU/mL of NG per sample and tested on the PANTHER.
Analytical Specificity
A series of ATCC quality control strains and clinical isolates, identified based on phenotypic characteristics or biochemical identification by the Vitek 2 instrument (Bio-Mérieux, Durham, NC), were grown on chocolate or sheep blood agar for 24 hours, resuspended in sterile saline to 0.5 McFarland, and diluted in pooled CT−/NG− urine to give an approximate concentration of 5 × 104 CFU/mL. This concentration, found in urinary tract infections or in significantly contaminated urine specimens, was deemed relevant for examining analytical specificity. Then, 3 mL of each spiked urine specimen was transferred into APTIMA urine transport tubes and evaluated on the PANTHER.
Relative Analytical Sensitivity in the Urine Matrix
Limiting dilution studies of the AmpliPROBE CT/GC control material (Bio-Rad, Hercules, CA) were used to compare the relative analytical sensitivity of the PANTHER and TIGRIS. The AmpliPROBE reagent contains a stabilized mixture of CT LGV type II strain 434 elementary bodies and whole-cell lysate of NG.13 However, the absolute concentration of CT and NG is not provided by the manufacturer. Therefore, relative but not absolute limit of detection (LOD) was determined. To appropriately simulate patient urine samples and ensure accurate dilution of control material across multiple replicates and multiple dilutions, we prepared test samples with two components consisting of urine and control material stabilized in urine transport buffer. Notably, the control RNA target is intrinsically unstable in urine. Therefore, control material was diluted in transport media first and then mixed with urine rather than the converse. This is similar to the procedure used by the manufacturer in limiting dilution studies described in the PANTHER14 and TIGRIS15 package inserts.
Specifically, to perform this analysis, a series of 10-fold dilutions of the AmpliPROBE control material were made in APTIMA urine transport media, ranging from 50- to 108-fold dilutions. Then, 2 mL of each dilution was transferred to an empty, labeled APTIMA urine transport tube. Four tubes were prepared from each dilution, and these were used for initial screening. Thirty different patient urine samples that had previously been confirmed negative for both CT and NG were pooled and used as the urine sample matrix. Of this pooled urine, 3 mL was then added to each sample tube containing specific dilutions of the CT/NG control material and mixed by gentle inversion.
After the initial quadruplicate screening, replicate screening was expanded to allow statistical comparison that was reasonably powered. Serial dilutions at, below, and above the lowest concentration (highest dilution) that exhibited reactivity for CT and NG were prepared, now in replicates of 20. These replicate dilutions (ranging from 7.5 × 104-fold to 2.5 ×106-fold) were then tested in parallel on the PANTHER and TIGRIS to determine the relative LOD.
Statistical Analysis
Positive, negative, and overall percent agreement and 95% confidence intervals (listed as a range in parentheses following percent agreement) were calculated as recommended in Clinical and Laboratory Standards Institute’s guidelines.16 Percent agreement was determined preferentially instead of clinical sensitivity and specificity because of the study design, which was a comparison study not referenced to patient infectious status and an analytical gold standard.16 LOD was determined by probit analysis as previously described17 using SPSS version 18.0 (SPSS, Chicago, IL). For the most accurate determination of the LOD, only data below the 100% positivity level were used for extrapolation of the LOD.17
Results
The PANTHER platform was evaluated for accuracy and precision. A major goal was to verify the off-label use of the PANTHER for detection of CT and NG in male and female urine specimens. Therefore, a more extensive analysis was performed for these samples types, including determination of relative analytical sensitivity in comparison with the TIGRIS.
Male Urine Samples
In total, 172 male urine specimens were evaluated for performance on the PANTHER. Data are summarized in Table 1. Forty-six urine samples were collected in the APTIMA transport medium and initially tested by the TIGRIS (four CT+, one NG+, one CT+/NG+, and 40 CT−/NG−). On retesting by the PANTHER, 45 yielded identical results. One discrepant sample, designated as CT+/NG+ by an outside institution, tested CT−/NG− on the PANTHER and on repeat testing on both the PANTHER and TIGRIS. This sample was presumed to have been misclassified by the outside laboratory and was therefore reassigned to the CT−/NG− category for data analysis.
Comparison of Test Results for Male Urine Specimensa
Comparison of Test Results for Male Urine Specimensa
In addition, 126 male urine samples initially tested by the COBAS (47 CT+, 47 NG+, one CT+/NG+, and 31 CT−/NG−) were retested on both the PANTHER and TIGRIS. Among COBAS NG+ samples, there were 13 discrepancies. All were initially NG+ by the COBAS prior to frozen storage in the absence of transport medium. However, on the PANTHER, the results were as follows: nine NG+, three NG± (equivocal), and one NG−. Interestingly, further diminution of positivity was observed on subsequent testing on the TIGRIS, with four NG± and nine NG−. To resolve these discrepancies, the samples were again retested on the PANTHER, TIGRIS, COBAS, and LC. On repeat PANTHER testing, all results were negative, whereas on repeat TIGRIS testing, all but two were negative. Interestingly, when retested on the COBAS, eight were NG+, four were NG±, and one was NG−, very similar to initial PANTHER results. Last, all samples were positive when retested by LC assays. However, four samples had a much delayed cycle threshold compared with LC when performed during the original clinical testing.
The simplest explanation for these findings is that urine samples were stored without preservative per the COBAS testing protocol. Notably, both the COBAS and LC detect DNA in contrast to the AC2, which detects RNA. RNA is intrinsically much less stable than DNA. Therefore, these discrepant results likely reflect the preferential degradation of RNA target in samples that do not have a guanidinium-based transport buffer to preserve RNA integrity. However, it is likely that DNA was also degraded to some extent since even the DNA-based COBAS assay showed some loss of positivity. By design, LC assays, as described previously by our laboratory,7 are more sensitive than COBAS assays since extracted nucleic acids are concentrated by ultrafiltration prior to PCR reaction setup (by a factor of 10-fold or more). This enhanced sensitivity, in combination with a DNA-based amplification assay, provides a reasonable explanation for the ability of LC to detect NG in all samples, even after partial degradation of target. Since nine specimens agreed on the PANTHER, initial testing on the COBAS, and follow-up testing by LC, they were scored as concordant NG+. The four remaining PANTHER discrepancies (including one CT+/NG+ sample) were excluded from subsequent analysis since they were not reliably detected by the PANTHER or the TIGRIS comparator.
Similar issues confounded the analysis of the COBAS CT+ male urine samples. Among COBAS CT+ samples, there were five potentially discrepant samples. Among these, the PANTHER identified two as CT+ and three as CT−. The two CT+ samples were equivocal on the TIGRIS and variably equivocal, positive, or negative on repeat PANTHER and TIGRIS testing. Of the three PANTHER negative samples, only one was positive by the TIGRIS, but this sample became negative on TIGRIS repeat and positive on PANTHER repeat. Taken together, two of the five discrepant samples (initially CT+ on COBAS) were resolved as concordant CT+ since the initial results on the PANTHER agreed with the original COBAS results. As noted above, the remaining discrepancies can best be explained by sample degradation in urine samples collected without preservative and variable positivity due to stochastic effects. We therefore excluded the three PANTHER negative, discrepant specimens from subsequent analysis.
License GrantNumber of Users: In consideration for the license fee paid, Flat-it type foundry grants to you only, the Licensee, the non-exclusive, non-transferable right to use and display the Flat-it type foundry Product purchased through this service on up to five (5) CPUs at your site connected to any number of printers at your own site. GeneralThis document is a legal agreement between you, the end user, and Flat-it type foundry. By using or installing this software, you agree to be bound by the terms of this Agreement.In this document the term gfont softwareh refers to any program or piece of code and/or computer information containing data that can be used to output letterforms or designs to any output or display device currently known or unknown. Standard License agreementLICENSING AGREEMENT 0. 
Among 31 CT−/NG− male urine samples originally tested by the COBAS, one sample was negative on the PANTHER but positive on the TIGRIS. On subsequent retesting, the sample was negative on the TIGRIS and borderline positive by the COBAS. This sample was excluded from analysis since CT was not consistently detected by any of the three methods.
Female Urine Samples
In total, 130 female urine samples were evaluated for performance of the PANTHER. Data are summarized in Table 2. Ninety-seven samples were collected in the APTIMA urine transport tubes and initially tested on the TIGRIS at outside institutions (17 CT+, eight NG+, two CT+/NG+, and 70 CT−/NG−). Thirty-three were initially tested by the COBAS (33 CT+).
Comparison of Female Urine Specimensa
Comparison of Female Urine Specimensa
Among CT+ samples, there were six discrepancies. Three of these samples, originally designated as CT+ on the TIGRIS by outside institutions, were confirmed negative on repeat testing by the TIGRIS and PANTHER. These discrepancies likely were due to miscoding of the results on the sample tubes. Therefore, they were reassigned to the CT−/NG− category for comparison. The remaining three were COBAS CT+ samples that tested CT+ on initial testing by the PANTHER. However, on testing by the TIGRIS, two of these were negative and one gave an invalid result. On retesting by the TIGRIS, the negatives remained negative and the invalid sample became equivocal. The TIGRIS negative samples were tested again on the PANTHER and COBAS. Both samples became negative on the PANTHER. On the COBAS, one became CT− and the second remained CT+. As explained above, these results are most likely explained by specimen degradation in unpreserved urine specimens. Taken together, results were resolved as concordant CT+ since initial PANTHER CT+ results matched the COBAS and later results could reasonably be explained by temporal target degradation.
Among NG+ samples, one was negative on the PANTHER and retesting on the TIGRIS. This sample was assumed to have been miscoded by the outside laboratory and was resolved as CT−/NG− for specimen comparison. Notably, the number of NG+ female urine samples (ie, nine) was much lower than desired because of the low prevalence of NG in women in the greater Boston area. Therefore, additional NG-spiked female urine samples were used for additional evaluation as discussed below.
NG-Spiked Female Urine Specimens
Spiking studies are a recommended method for evaluating rare analytes when diluted into the appropriate matrix.12 The analytical sensitivity claim for the PANTHER on approved sample types is 50 CFU/mL.14 Therefore, our goal was to determine whether NG could be detected in distinct female urine samples spiked with levels below this threshold. Hence, we spiked 48 COBAS CT– clinical female urine samples with 12.5 CFU/mL of NG ATCC 43069 at a concentration fourfold lower than detection claims for approved PANTHER specimen types and the APTIMA positive control. Prior to spiking, samples were first tested by the PANTHER for both CT and NG. Unexpectedly, one sample was CT+ on the PANTHER (RLU = 412, a low value, perhaps accounting for negativity by the COBAS), a positive result reconfirmed on the TIGRIS. After spiking, NG+ (RLU range, 339–1,248) was detected in 47 samples. The remaining sample was equivocal for NG (RLU = 71) by the PANTHER and borderline negative when retested on the TIGRIS (RLU = 54). Since we had purposefully used an NG+ concentration below the analytical sensitivity claim for the PANTHER, such a result for a single sample is not surprising. Specifically, among 48 spike samples, it is statistically plausible that some samples may have stochastically received smaller quanta of targets just beyond the detection threshold of the assay. This would account for equivocal/negative results by both the PANTHER and TIGRIS methods, whose RLU measurements were essentially the same.
Unsexed Urine Specimens
Forty-seven deidentified urine samples, previously tested by the TIGRIS, were provided to us without designation as to the sex of patients (39 CT+ and eight NG+). Among these samples, two designated CT+ at an outside institution tested CT− on the PANTHER. The first sample was CT− on the TIGRIS and PANTHER repeat and was reassigned as CT−/NG−. The second sample on retesting was positive on both the PANTHER (RLU = 115) and TIGRIS (RLU = 102), respectively, very close to the positivity cutoff of 100. The very low level of positivity likely explained the variability in target detection. Nevertheless, the second sample was tabulated as a PANTHER false-negative based on initial testing results. The combined data for female, male, unsexed, and spiked urine specimens are shown in Table 3.
Comparison of Combined Results for Urine Samples Including Spiked and Unsexed Samplesa
Comparison of Combined Results for Urine Samples Including Spiked and Unsexed Samplesa
Female and Male Swabs and ThinPrep PreservCyt
A more limited analysis was performed for approved sample types, including 52 endocervical swabs (33 CT+, eight NG+, one CT+/NG+, and ten CT−/NG−), two male swabs (one NG+ and one CT+/NG+), and 15 ThinPrep PreservCyt samples (five CT+ and ten CT−/NG−), all previously tested on the TIGRIS, with combined results shown in Table 4. Notably, the PANTHER failed to detect NG in one of nine NG+ endocervical swab samples, both initially (RLU = 8) and on repeat testing (RLU = 20). However, the sample was positive when retested on the TIGRIS (RLU = 1,351) and was therefore resolved as a PANTHER false-negative. Three female swabs yielded invalid results on the PANTHER and were excluded from analysis since there was insufficient sample for retesting. Only two male swabs were available for testing: one was CT−/NG+ and the other was CT+/NG+. No NG+ ThinPrep PreservCyt specimens were available. All male swab and ThinPrep PreservCyt results were confirmed on the PANTHER.
Comparison of Combined Test Results for Female Swabs, Male Swabs, and ThinPrep PreservCyt Specimensa
Comparison of Combined Test Results for Female Swabs, Male Swabs, and ThinPrep PreservCyt Specimensa
Analytical Specificity
To examine potential effects of urine matrix on analytical specificity, we seeded CT−/NG− urine with organisms frequently found as either urine contaminants or pathogens. As shown in Table 5, no cross-reactivity was observed when tested by the PANTHER, and all RLU were within the range found in CT−/NG− urine samples.
Analytical Specificity in the Urine Matrix
Analytical Specificity in the Urine Matrix
Analytical Sensitivity in Urine Matrix
The relative LODs for the PANTHER and TIGRIS were determined through testing dilutions of the AmpliPROBE CT/NG control material in urine matrix Table 6. For CT, the PANTHER LOD, 4.9 (95% CI, 4.3–5.2) log dilution, was nominally higher than the TIGRIS LOD, 5.0 (95% CI, 4.0–5.2) log dilution, as determined by probit analysis. For NG, the LODs for the PANTHER and TIGRIS were equivalent (5.9 log dilution). Overall, the analytical sensitivity appeared to be equivalent in urine matrix within the power of the analysis.
Relative Limit of Detection Assessed by Replicate Testing of AmpliProbe CT/NG Control Serially Diluted in Urine Matrix
Relative Limit of Detection Assessed by Replicate Testing of AmpliProbe CT/NG Control Serially Diluted in Urine Matrix
Discussion
We expected that the PANTHER and TIGRIS would have almost identical performance characteristics based on their use of the same AC2 reagents and amplification conditions. However, it is conceivable that the PANTHER automation might somehow compromise performance relative to other AC2-based assays. Therefore, we evaluated the performance of the PANTHER compared with the well-established TIGRIS and/or COBAS assays. Taking into account evidence for specimen degradation in unpreserved urine and likely misclassification of several specimens at outside institutions, we found that the PANTHER demonstrated nearly 100% positive and negative agreement with the TIGRIS and/or COBAS (Tables 1–3) for urine samples. Furthermore, the PANTHER and TIGRIS demonstrated an equivalent LOD within the power of analysis. A more limited evaluation was performed for approved specimen types in which agreement between the PANTHER and TIGRIS was 100% for CT−, CT+, and NG− and 90.9% for NG samples with larger 95% confidence intervals implicit in the small numbers of samples tested (see Table 4).
During our study, we noted that several urine samples stored in the absence of APTIMA transport medium (ie, without guanidinium-containing denaturant) showed loss of CT/NG detection, consistent with temporal degradation of the RNA target during storage. We previously observed similar lability in cervical swab samples on storage, even for the more stable COBAS DNA targets.7 These observations underscore the importance of adding samples to transport medium as soon as practicable to avoid specimen degradation and false-negative results.
Although we did not directly assess clinical sensitivity and specificity of the PANTHER, we infer that clinical sensitivity and specificity should be maintained relative to other AC2-based assays based on comparisons with the TIGRIS. Furthermore, the instrument automation and user interface were facile. No additional operator intervention is required after loading reagents and samples. Moreover, multiple assays can be run at once, even on the same specimen, providing intriguing possibilities for laboratory automation as additional assays are approved for the PANTHER. Taken together, the PANTHER provides an automated, high-throughput addition to methods for CT/NG detection with excellent performance characteristics on a wide range of sample types, including male and female urine samples.
Acknowledgments
We thank the clinical microbiology laboratories at Brigham and Women’s Hospital, Cambridge Hospital, Boston Medical Center, Massachusetts General Hospital, and Needham Harvard Vanguard for providing deidentified samples and Hologic Gen-Probe for providing reagents for this study.