ThinPrep Integrated Imager
Device Facts
| Record ID | P950039S036 |
|---|---|
| Device Name | ThinPrep Integrated Imager |
| Applicant | Hologic, Inc. |
| Product Code | MNM · Pathology |
| Decision Date | Apr 18, 2018 |
| Decision | APPR |
| Device Class | Class 3 |
| Attributes | Real-World Evidence |
Real-World Evidence
| Submission | Device | Sponsor | RWD Sources | RWE Use Summary | Key Tags |
|---|---|---|---|---|---|
| P950039S036 · Apr 18, 2018 | ThinPrep Integrated Imager | Hologic, Inc. | Retrospective clinical study cohort of 1,260 de-identified patient cervical specimens (ThinPrep Pap Test slides) | The study used a retrospective cohort of previously collected and adjudicated clinical slides to compare the sensitivity and specificity of the ThinPrep Integrated Imager (I2) against the approved ThinPrep Imaging System (TIS). | Retrospective cohort; Cervical cytology; Diagnostic performance; Adjudicated truth |
Clinical Evidence
| Study Design | Population | Comparator | Key Endpoints |
|---|---|---|---|
| Multi-Center Evaluation of the ThinPrep Integrated Imager; Multi-center, two-armed retrospective clinical study; Follow-up/Duration: Not applicable for this study | 1,260 patient cervical specimens (2,520 slides total); Sample Size: 1,260 patient cases; Number of Sites: 3 | ThinPrep Imaging System (TIS) | Sensitivity, specificity, and likelihood ratios at various diagnostic thresholds (ASCUS+, LSIL+, ASC-H+, HSIL+) |
Indications for Use
The Hologic ThinPrep® Integrated Imager is a device that uses computer imaging technology to assist in primary cervical cancer screening of ThinPrep® Pap Test slides for the presence of atypical cells, cervical neoplasia, including its precursor lesions (Low Grade Squamous Intraepithelial Lesions, High Grade Squamous Intraepithelial Lesions), and carcinoma as well as all other cytologic criteria as defined by the Bethesda System: Terminology for Reporting Results of Cervical Cytology¹.
Device Story
ThinPrep Integrated Imager (I2) is a desktop system combining imaging and review functions for cervical cytology. It processes ThinPrep Pap Test slides prepared via ThinPrep 2000 or 5000 systems. During imaging, the device captures high-magnification frames at >500 x-y locations across the cell spot; it predicts z-locations (focal planes) based on focus quality and fiducial points. Software identifies objects of interest based on integrated optical density; 22 fields of view (FOVs) are recorded. During review, a cytotechnologist (CT) evaluates these 22 FOVs (Autolocate). If abnormalities are detected, the CT performs a full slide review (Autoscan). The system operates in Sequential (imaging then immediate review) or Batched (imaging then later review) modalities. The output assists the CT in identifying abnormal cells, potentially reducing screening time while maintaining diagnostic accuracy compared to manual review or previous imaging systems. It is intended for use in clinical laboratories by trained cytotechnologists.
Clinical Evidence
Multi-center clinical study (3 sites, 1,260 patient cases, 2,520 slides) compared I2-assisted review to TIS-assisted review. Primary endpoint: sensitivity and specificity at ASCUS+ threshold using pathologist adjudication as the reference standard. Results: I2 showed higher sensitivity than TIS across all thresholds (e.g., ASCUS+ sensitivity 89.8% vs 86.0%; LSIL+ 83.7% vs 77.8%; HSIL+ 67.5% vs 59.6%) with a slight decrease in specificity. Bench testing and clinical workload assessments confirmed CT screening rates and accuracy are consistent with CLIA guidelines.
Technological Characteristics
Desktop computer-based system; includes microscope with imaging camera, slide ID reader, automated stage, and touchscreen interface. Uses integrated optical density analysis for object identification. Connectivity: database-driven for FOV coordinate storage. Software: proprietary system application. Sterilization: N/A (IVD device). Complies with IEC 61010-1, IEC 61326-1, and IEC 61326-2-6 for electrical safety and EMC.
Indications for Use
Indicated for primary cervical cancer screening of ThinPrep Pap Test slides to detect atypical cells, cervical neoplasia (LSIL, HSIL), and carcinoma in adult women.
Predicate Devices
- ThinPrep Imaging System (P020002)
- ThinPrep 2000 System (P950039)
- ThinPrep 5000 Processor (P950039/S026)
Reference Devices
- BD FocalPoint™ GS Imaging System (P950009/S008)
Submission Summary (Full Text)
{0}
# SUMMARY OF SAFETY AND EFFECTIVENESS DATA (SSED)
## I. GENERAL INFORMATION
| Device Generic Name: | Reader, Cervical Cytology Slide, Automated |
| --- | --- |
| Device Trade Name: | ThinPrep Integrated Imager |
| Device Procode: | MKQ, MNM |
| Applicant's Name and Address: | Hologic, Inc. 250 Campus Drive Marlborough, MA 01752 |
| Date(s) of Panel Recommendation: | None |
| Premarket Approval Application (PMA) Number: | P950039/S036 |
| Date of FDA Notice of Approval: | April 18, 2018 |
The ThinPrep Imaging System (TIS) was originally approved on June 6, 2003, under original PMA P020002 (which was later combined with P950039), for assisting in primary cervical cancer screening of ThinPrep Pap Test slides for the presence of atypical cells, cervical neoplasia, including its precursor lesions (Low Grade Squamous Intraepithelial Lesions, High Grade Squamous Intraepithelial Lesions), and carcinoma as well as all other cytological criteria as defined by The 2001 Bestheda System: Terminology for Reporting Results of Cervical Cytology. The SSED to support the indication is available on the CDRH website and is incorporated by reference here. The current supplement was submitted for the ThinPrep Integrated Imager, which is intended in primary cervical cancer screening of ThinPrep Pap Test slides for the presence of abnormal cells.
## II. INDICATIONS FOR USE
The Hologic ThinPrep® Integrated Imager is a device that uses computer imaging technology to assist in primary cervical cancer screening of ThinPrep® Pap Test slides for the presence of atypical cells, cervical neoplasia, including its precursor lesions (Low Grade Squamous Intraepithelial Lesions, High Grade Squamous Intraepithelial Lesions), and carcinoma as well as all other cytologic criteria as defined by the Bethesda System: Terminology for Reporting Results of Cervical Cytology¹.
## III. CONTRAINDICATIONS
There are no known contraindications.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 1
{1}
# IV. WARNINGS AND PRECAUTIONS
The warnings and precautions can be found in the Thinprep Integrated Imager labeling.
# V. DEVICE DESCRIPTION
The ThinPrep Integrated Imager (I2) is a computer-based, desktop device which combines the functionality of a ThinPrep Imaging station and a review scope into a single automated imaging and review system. The I2 is intended for use with ThinPrep Pap Test slides prepared with the ThinPrep 2000 System (approved under P950039) or the ThinPrep 5000 Processor (approved under P950039/S026) and subsequently stained with ThinPrep stain.
As illustrated in Figure 1, the three major subsystems of the I2 are the Microscope, Controller and Computer. The Microscope has an imaging camera, slide ID reader, automated stage, hand controls and adjustable touch screen user interface. The Controller controls the electromechanical components of the Microscope. The Computer hosts the system application software and system database.
Figure 1: Three major subsystems of I2
The operation and design of the I2 are grouped into two major functions: imaging and review. First, during imaging, the I2 takes high magnification frames at more than 500 x-y locations covering the entire cell spot and this process takes approximately 90 seconds. The z-locations (focal plane) of the first several frames are calculated based on the x-y locations of these frames, the x-y locations of the fiducial points, and the z-locations of the fiducial points on the slide. After that, for each frame, the I2 predicts the z-location of the current frame based on the focus quality changes and z-locations of existing frames. This allows the cell spot to be scanned with a single image taken at each x-y location. The system software analyzes images of the cell spot and identifies objects of interest based on integrated optical density. The coordinates of 22 of fields of view (FOVs) are recorded with the slide ID, and stored in the system database.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 2
{2}
Second, during review, the I2 retrieves the locations of the objects of interest from the system database and sequentially positions these locations for evaluation and interpretation by the cytotechnologist (CT). The CT uses the review control or touchscreen to step through each of the fields of view (Autolocate). Additionally, the system provides a method for automated marking of objects for further review. If the CT identifies any of these fields as containing abnormal objects, that field may be marked electronically. The I2 will require the CT to conduct a review of the entire cell spot for any slide with electronically marked fields (Autoscan).
The ThinPrep Imager is designed to work in two modalities: Sequential Modality and Bached Modality. In Sequential Modality, after being loading on the Microscope, the slide is imaged and then reviewed immediately by the CT. As an alternative workflow, in Batched Modality, slides can be imaged in succession, with the coordinates stored in the computer database, for review by the CT or pathologist at a later time.
## VI. ALTERNATIVE PRACTICES AND PROCEDURES
The procedures for primary cervical cancer screening of ThinPrep Pap Test slides include review of the entire slides using manual microscopy (full manual review) or automated imaging systems. The following automated imaging systems are currently approved:
- BD FocalPoint™ GS Imaging System (approved under P950009/S008).
## VII. MARKETING HISTORY
The ThinPrep Integrated Imager (I2) has not been distributed for IVD use in the United States. The I2 was CE Marked on September 21, 2009. Table 1 lists the countries and other areas in which the I2 has been marketed.
**Table 1: Countries and other areas in which the I2 has been marketed**
| Australia | Hong Kong | Macau | Spain |
| --- | --- | --- | --- |
| Austria | India | Netherlands | Sweden |
| Bahrain | Indonesia | New Zealand | Switzerland |
| Belgium | Israel | Romania | Thailand |
| France | Italy | Slovakia | Turkey |
| Germany | Japan | Saudi Arabia | United Arab Emirates |
| Greece | Korea, South | South Africa | United Kingdom |
## VIII. PROBABLE ADVERSE EFFECTS OF THE DEVICE ON HEALTH
The ThinPrep Integrated Imager (I2) is not intended to direct therapy or treatment of a patient. However, potential risks of the device when used to assist in the screening of slides for cervical cytology are associated with the I2's ability to correctly determine the 22 fields most likely to contain abnormal cells. If none of the 22 fields contains abnormal cells when disease is present, this may result in a false negative diagnosis. If the CT interprets a slide as
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 3
{3}
containing abnormalities when no disease is present, this may result in a false positive diagnosis.
A false positive diagnosis may result in unnecessary colposcopy exam for the patient or the patient may be referred for a biopsy, both are generally considered low risk procedures. A false negative interpretation may result in delayed diagnosis or treatment for the patient.
### IX. SUMMARY OF NONCLINICAL STUDIES
Non-clinical studies were designed to evaluate and validate the performance of the ThinPrep Integrated Imager. Studies included hardware testing (Optical System Alignment, camera performance) and system characteristics. Subsystem equivalency tests - such as Illumination Brightness and Uniformity, Focus and Slide Error Threshold testing - were performed to demonstrate similarities of the ThinPrep Integrated Imager to the previously approved ThinPrep Imaging System.
#### A. Laboratory Studies
##### 1. Within-instrument Reproducibility
The objective of this study was to characterize the within-instrument reproducibility metrics for the ThinPrep Integrated Imager (I2) with regard to the repeatability of the selection of equivalently ranked diagnostic content for each slide's selected 22 FOVs, abnormal FOV count, and autoscan referral rates. The same reproducibility metrics were also assessed for the approved ThinPrep Imaging System (TIS) for comparative purposes.
Slides that were previously prepared from cervical specimens and had an adjudicated cytology diagnosis were used. The study set included 260 slides with 130 slides prepared on ThinPrep 2000 System and 130 slides prepared on ThinPrep 5000 processor. The study also included specimens with varying diagnoses, as indicated in the study protocol.
Each slide was imaged a total of six times: three times on a single TIS system and three times on a single I2 system. Three CTs reviewed the 22 FOVs for each run of each slide. The highest ranked diagnosis from review of 22 FOVs and the numbers of abnormal FOVs were recorded for each of three runs for both TIS review and I2 review. The washout period for any given slide was 14 days in order to minimize memory bias. Full slide reviews were not carried out for this evaluation.
In Table 2, the within-instrument results are summarized for each diagnostic category of slides (according to adjudicated truth results). For each category, the following metrics are reported:
- % Abnormal – the proportion of slides for which any abnormal FOVs were observed. (For NILM or UNSAT slides, the % Normal column is used to record the proportion that are not abnormal).
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 4
{4}
- % Category+ – the proportion of slides for which at least one FOV was observed with content of the slide’s true category or higher.
- % N/A – the proportion of slides in that category that are excluded from analysis (slide not able to be imaged by imager or missing data).
- FOV % zero – the proportion of slides for which zero abnormal FOV were observed.
- FOV Median – the median number of abnormal FOV observed (out of 22 total).
Table 2: Within-instrument study results
| Dx | Imager | % Abnormal | % Category+ | % Normal | % N/A | FOV | |
| --- | --- | --- | --- | --- | --- | --- | --- |
| | | | | | | % Zero | Median |
| NILM | TIS | | | 69.6% | 11.0% | 70.4% | 0 |
| | I2 | | | 78.1% | 4.3% | 78.4% | 0 |
| ASCUS | TIS | 75.9% | 75.9% | | 13.3% | 25.0% | 6 |
| | I2 | 71.9% | 71.9% | | 5.0% | 28.1% | 7 |
| LSIL | TIS | 97.3% | 93.2% | | 3.3% | 2.8% | 14 |
| | I2 | 96.0% | 94.0% | | 0.7% | 4.0% | 15 |
| ASC-H | TIS | 93.3% | 86.7% | | 0.0% | 6.7% | 11.5 |
| | I2 | 100% | 83.3% | | 0.0% | 0.0% | 14 |
| AGUS | TIS | 63.0% | 51.9% | | 6.7% | 35.7% | 2 |
| | I2 | 55.6% | 48.1% | | 10.0% | 44.4% | 2 |
| HSIL | TIS | 98.0% | 77.3% | | 0.0% | 2.0% | 20 |
| | I2 | 97.3% | 71.3% | | 0.7% | 2.7% | 20 |
| CANCER | TIS | 100% | 46.7% | | 0.0% | 0.0% | 22 |
| | I2 | 100% | 53.3% | | 0.0% | 0.0% | 22 |
| UNSAT | TIS | | | 72.2% | 40.0% | 72.2% | 0 |
| | I2 | | | 85.7% | 36.7% | 94.7% | 0 |
2. Between-instrument Reproducibility
Between-instrument reproducibility results were derived from the clinical study (described in Section X below). Briefly, in the clinical study, three CT/pathologist pairs reviewed 2,514 slides once each; first on different ThinPrep Imaging Systems (TIS) and then on different ThinPrep Integrated Imagers (I2).
In Table 3, the between-instrument reproducibility results are summarized for each diagnostic category of slides (according to adjudicated truth results). For each category, the following metrics are reported:
- % Abnormal – the proportion of slides for which any abnormal diagnosis was recorded. (For NILM or UNSAT slides, the % Normal column is used to record the proportion that are not abnormal).
- % Category+ - the proportion of slides for which the site diagnosis was equal to or higher than the slide’s adjudicated category.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 5
{5}
**Table 3: Between-instrument study results**
| Dx | Imager | % Abnormal | % Category+ | % Normal |
| --- | --- | --- | --- | --- |
| NILM | TIS | -- | -- | 90.0% |
| | I2 | -- | -- | 88.1% |
| ASCUS | TIS | 64.4% | 64.4% | -- |
| | I2 | 71.7% | 71.7% | -- |
| LSIL | TIS | 95.0% | 75.0% | -- |
| | I2 | 96.9% | 80.6% | -- |
| ASC-H | TIS | 87.7% | 62.6% | -- |
| | I2 | 92.8% | 63.6% | -- |
| AGUS | TIS | 53.8% | 37.6% | -- |
| | I2 | 67.5% | 57.3% | -- |
| HSIL | TIS | 97.7% | 54.7% | -- |
| | I2 | 99.3% | 64.7% | -- |
| CANCER | TIS | 100% | 63.2% | -- |
| | I2 | 100% | 63.2% | -- |
| UNSAT | TIS | -- | -- | 95.2% |
| | I2 | -- | -- | 93.2% |
# 3. FOV Reproducibility Study
A total of 100 slides from the clinical study specimens were included in this study with the abnormality spectrum shown in Table 4 below. The slides were imaged twice on one ThinPrep Imaging System (TIS), and then imaged twice on one ThinPrep Integrated Imager (I2). FOV locations were compared separately between the two TIS runs and the two I2 runs, and the matching FOV locations were counted. Any FOVs with coordinates within 400 µm were considered an FOV match. Two slides were excluded (one ASCUS slide and one LSIL slide), because of error encountered during both of the TIS runs. As shown in Table 5, the mean FOV reproducibilities between two I2 runs are lower than those between two TIS runs in all categories except LSIL.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 6
{6}
**Table 4: Mean FOV reproducibility breakdown for diagnostic categories**
| Category | # of slides | Mean FOV Reproducibility (% [95% CI]) | | |
| --- | --- | --- | --- | --- |
| | | TIS | I2 | Difference |
| **UNSAT** | 10 | 58.2% [49.5%, 65.0%] | 45.5% [39.1%, 52.7%] | -12.7% [-20.9%, -5.5%] |
| **NILM** | 50 | 58.5% [55.3%, 61.5%] | 52.1% [48.6%, 55.2%] | -6.4% [-10.2%, -2.9%] |
| **ASCUS** | 10* | 65.7% [59.6%, 71.4%] | 56.4% [49.5%, 62.3%] | -10.1% [-17.3%, -3.1%] |
| **LSIL** | 10* | 59.6% [53.4%, 66.9%] | 69.5% [61.8%, 75.9%] | 9.1% [1.5%, 15.2%] |
| **ASC-H** | 5 | 71.8% [64.5%, 79.1%] | 70.9% [64.5%, 80.9%] | -0.9% [-8.2%, 5.5%] |
| **HSIL** | 10 | 67.3% [60.5%, 72.7%] | 63.6% [58.2%, 68.6%] | -3.6% [-7.7%, 0.0%] |
| **CANCER** | 5 | 60.0% [49.1%, 70.9%] | 47.3% [36.4%, 58.2%] | -12.7% [-17.3%, -6.4%] |
| **All Abnormal** | 40 | 64.7% [61.2%, 68.4%] | 62.2% [58.1%, 66.3%] | -3.0% [-7.0%, 0.7%] |
| **Overall** | 100 | 60.9% [58.6%, 63.1%] | 55.5% [52.9%, 58.1%] | -5.7% [-8.1%, -3.1%] |
\* One ASCUS slide and one LSIL slide were excluded.
**Table 5: Mean and median distances between matched pair FOVs**
| Category | Mean Distance Between Matched FOVs (µm) | | Median Distance Between Matched FOVs (µm) | |
| --- | --- | --- | --- | --- |
| | TIS | I2 | TIS | I2 |
| **UNSAT** | 17 | 41 | 1 | 12 |
| **NILM** | 28 | 39 | 1 | 4 |
| **ASCUS** | 23 | 29 | 1 | 5 |
| **LSIL** | 29 | 33 | 1 | 4 |
| **ASC-H** | 13 | 29 | 1 | 4 |
| **HSIL** | 22 | 23 | 1 | 4 |
| **CANCER** | 22 | 27 | 1 | 3 |
| **All Abnormal** | 22 | 28 | 1 | 4 |
| **Overall** | 25 | 35 | 1 | 4 |
Positive and negative percent agreements (PPA and NPA, respectively) at the ASCUS+ diagnostic threshold, along with the proportion of UNSAT cases correctly identified, are shown in Table 6 below. The results from the I2 review demonstrate higher sensitivity than those from the TIS review.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 7
{7}
Table 6: PPV, NPV and proportion of UNSAT cases
| Parameter | TIS | I2 |
| --- | --- | --- |
| PPA (ASCUS+) | 83% | 89% |
| NPA ( | 85% | 84% |
| UNSAT agreement | 63% | 63% |
### B. Animal Studies
None.
### C. Additional Studies
Testing for Electrical Safety and EMC (Electromagnetic Compatibility) demonstrated compliance to following safety and electrical requirements:
- IEC 61010-1, Safety requirements for electrical equipment for measurement, control, and laboratory use - Part 1: General requirements
- IEC 61326-1, Electrical equipment for measurement, control and laboratory use – EMC requirements – Part 1: General requirements
- IEC 61326-2-6, Electrical equipment for measurement, control and laboratory use – EMC requirements – Part 2-6 Particular requirements-IVD medical equipment
Shipping and Package testing demonstrated compliance to functional requirements in ISO/IEC 17025:2005(E).
### X. SUMMARY OF PRIMARY CLINICAL STUDY
The applicant performed a clinical study to establish a reasonable assurance of safety and effectiveness of the ThinPrep Integrated Imager for assisting in primary cervical cancer screening of ThinPrep Pap Test slides in the US. Data from this clinical study and the supplemental study for cytotechnologist workload assessment were the basis for the PMA approval decision. A summary of the clinical study is presented below.
### A. Study Design
A multi-center, two-armed clinical study titled, “Multi-Center Evaluation of the ThinPrep Integrated Imager,” was performed at three sites within the US. The objective of the study was to show that routine screening of ThinPrep Pap Test slides prepared on the ThinPrep 2000 System and the ThinPrep 5000 processor using the ThinPrep Integrated Imager (I2) is similar to the review of ThinPrep slides using the ThinPrep Imaging System (TIS) for all categories used for cytologic diagnosis (specimen adequacy and descriptive diagnosis), as defined by the Bethesda System criteria¹.
The study included 1,260 patient cases that covered all cytologic diagnosis categories. Two slides were prepared from each case; one slide was prepared on the ThinPrep 2000 System and the other slide was prepared on the ThinPrep 5000 processor. Each slide was screened first using the Sequential Modality of the I2 and then, after two-week lag,
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 8
{8}
screened with the TIS. The slides were randomized prior to slide review in each study arm. All slides were reviewed independently three times by three CT/pathologist pairs at each site for both study arms. The unadjudicated site diagnoses obtained from TIS review and I2 review were compared to each other, as well as to the adjudicated diagnoses (which were used as truth).
# 1. Clinical Inclusion and Exclusion Criteria
The slides included in the clinical study were produced, reviewed manually and adjudicated during the execution of a previous study (used to support P950039/S026). The slides were de-identified and not used for clinical use.
The ThinPrep Pap Test slides from three sites included the following:
- NILM: 1260 slides from 630 cases
- ASC-US: 300 slides from 150 cases
- LSIL: 300 slides from 150 cases
- ASC-H: 300 slides from 150 cases
- AGUS: 30 slides from 15 cases
- HSIL: 300 slides from 150 cases
- Cancers: 30 slides from 15 cases
A slide was withdrawn and not available for interpretation if it was broken or determined to be unreadable for the purposes of this study.
# 2. Follow-up Schedule
There is no follow-up schedule in this study.
# 3. Clinical Endpoints
# Primary Endpoint
The primary objective of this study was to estimate the sensitivity, specificity, and likelihood ratios when diagnosing specimens imaged and reviewed on the ThinPrep Integrated Imager (I2) as compared with the ThinPrep Imaging System (TIS). The reference standard for this study was Pathologist adjudication consensus. The unadjudicated site diagnoses obtained from TIS review and I2 review are compared to adjudicated diagnoses (truth). The results were to be deemed acceptable if the sensitivity and specificity, at the ASCUS+ cutoff threshold, with I2 review were non-inferior to those metrics with TIS review.
# B. Accountability of PMA Cohort
From 1,260 patient cervical specimens, a total of 2,520 slides (840 slides or 420 patient cervical specimens from each site) were enrolled in this study. Six slides were excluded from review and analysis as they were broken and unreadable. Each slide was reviewed three times, once per site. The data set contains 7,542 analyzed records [(2,520 slides– 6 broken / excluded) × 3 reviews / slide = 7,542].
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 9
{9}
### C. Study Population Demographics and Baseline Parameters
Basic demographic information was collected for each patient specimen enrolled at each site to aid the CT in making a diagnosis for the resulting slides. A summary of the demographic information is presented in Table 7 for all sites.
**Table 7: Site demographics**
| Site Number | Median Age (years) | # Hysterectomy (% of enrolled) | # Postmenopausal (% of enrolled) |
| --- | --- | --- | --- |
| 1 | 36 | 11 (2.6%) | 30 (7.1%) |
| 2 | 33 | 15 (3.6%) | 25 (6.0%) |
| 3 | 37 | 25 (6.0%) | 51 (12.1%) |
| Overall | 35 | 51 (4.0%) | 106 (8.4%) |
### D. Safety and Effectiveness Results
### D. Safety and Effectiveness Results
No adverse events occurred nor were anticipated to occur in this study. The ThinPrep Integrated Imager is an in vitro diagnostic device and as such the potential for adverse events is relative to an “incorrect” diagnosis leading to either unnecessary care or delayed follow up care. The diagnoses made for the slides in this study were prepared from de-identified patient specimens and thus not able to be used in patient care.
#### 2. Effectiveness Results
##### Unadjudicated Descriptive Diagnosis Data
The major diagnostic categories of the Bethesda System$^{1}$ were used to examine the agreement between the ThinPrep Imaging System-assisted review (TIS review) findings and the ThinPrep Integrated Imager-assisted review (I2 review) findings for the unadjudicated data.
Table 8 below shows the unadjudicated descriptive diagnosis from all three sites combined for each arm.
**Table 8: Unadjudicated 8x8 diagnostic contingency table (I2 vs. TIS)**
| | TIS | | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | UNSAT | NILM | ASCUS | LSIL | ASC-H | AGUS | HSIL | CANCER |
| I2 | UNSAT | 185 | 33 | 6 | 0 | 4 | 2 | 0 | 0 |
| | NILM | 31 | 3949 | 200 | 20 | 20 | 16 | 7 | 1 |
| | ASCUS | 9 | 296 | 372 | 85 | 58 | 1 | 12 | 1 |
| | LSIL | 1 | 54 | 152 | 605 | 31 | 0 | 27 | 0 |
| | ASC-H | 3 | 53 | 81 | 34 | 171 | 6 | 73 | 2 |
| | AGUS | 4 | 27 | 10 | 1 | 3 | 27 | 4 | 9 |
| | HSIL | 1 | 23 | 31 | 73 | 106 | 11 | 472 | 7 |
| | CANCER | 2 | 2 | 4 | 4 | 0 | 7 | 22 | 91 |
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 10
{10}
### Adjudicated Descriptive Diagnosis Data
Table 9 through Table 15 show the performance of ThinPrep Imaging System (TIS) review and ThinPrep Integrated Imager (I2) review compared to adjudicated diagnosis made by the adjudication panel (truth) for the major descriptive diagnosis classifications of the Bethesda System$^{1}$.
**Table 9: “True Negative” (NILM) 8x8 contingency table, all sites combined**
| | | TIS | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | UNSAT | NILM | ASCUS | LSIL | ASC-H | AGUS | HSIL | CANCER |
| I2 | UNSAT | 75 | 29 | 2 | 0 | 1 | 1 | 0 | 0 |
| | NILM | 25 | 3735 | 147 | 5 | 13 | 7 | 3 | 0 |
| | ASCUS | 5 | 187 | 123 | 11 | 16 | 1 | 1 | 0 |
| | LSIL | 0 | 21 | 22 | 14 | 2 | 0 | 2 | 0 |
| | ASC-H | 1 | 29 | 20 | 1 | 23 | 1 | 4 | 0 |
| | AGUS | 1 | 15 | 3 | 0 | 0 | 5 | 0 | 0 |
| | HSIL | 0 | 8 | 4 | 0 | 10 | 0 | 10 | 0 |
| | CANCER | 0 | 0 | 2 | 0 | 0 | 1 | 0 | 4 |
**Table 10: “True ASCUS” 8x8 contingency table, all sites combined**
| | | TIS | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | UNSAT | NILM | ASCUS | LSIL | ASC-H | AGUS | HSIL | CANCER |
| I2 | UNSAT | 2 | 0 | 1 | 0 | 2 | 0 | 0 | 0 |
| | NILM | 1 | 143 | 36 | 7 | 4 | 5 | 2 | 1 |
| | ASCUS | 0 | 76 | 113 | 23 | 15 | 0 | 3 | 0 |
| | LSIL | 1 | 11 | 33 | 45 | 5 | 0 | 2 | 0 |
| | ASC-H | 0 | 16 | 18 | 5 | 37 | 1 | 19 | 0 |
| | AGUS | 1 | 0 | 0 | 0 | 1 | 2 | 0 | 0 |
| | HSIL | 0 | 5 | 6 | 5 | 19 | 0 | 53 | 0 |
| | CANCER | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
**Table 11: “True LSIL” 8x8 contingency table, all sites combined**
| | | TIS | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | UNSAT | NILM | ASCUS | LSIL | ASC-H | AGUS | HSIL | CANCER |
| I2 | UNSAT | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| | NILM | 0 | 13 | 11 | 8 | 0 | 0 | 1 | 0 |
| | ASCUS | 0 | 18 | 107 | 49 | 4 | 0 | 1 | 0 |
| | LSIL | 0 | 19 | 86 | 516 | 10 | 0 | 17 | 0 |
| | ASC-H | 0 | 3 | 12 | 13 | 16 | 1 | 16 | 9 |
| | AGUS | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| | HSIL | 0 | 1 | 3 | 40 | 11 | 2 | 107 | 0 |
| | CANCER | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 1 |
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 11
{11}
**Table 12: “True ASC-H” 8x8 contingency table, all sites combined**
| | | TIS | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | UNSAT | NILM | ASCUS | LSIL | ASC-H | AGUS | HSIL | CANCER |
| I2 | UNSAT | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 |
| | NILM | 0 | 5 | 4 | 0 | 2 | 1 | 1 | 0 |
| | ASCUS | 0 | 9 | 16 | 1 | 13 | 0 | 4 | 0 |
| | LSIL | 0 | 1 | 3 | 2 | 7 | 0 | 1 | 0 |
| | ASC-H | 0 | 4 | 14 | 1 | 31 | 1 | 9 | 0 |
| | AGUS | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
| | HSIL | 0 | 4 | 4 | 2 | 17 | 0 | 31 | 1 |
| | CANCER | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 2 |
**Table 13: “True AGUS” 8x8 contingency table, all sites combined**
| | | TIS | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | UNSAT | NILM | ASCUS | LSIL | ASC-H | AGUS | HSIL | CANCER |
| I2 | UNSAT | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| | NILM | 1 | 30 | 2 | 0 | 1 | 3 | 0 | 0 |
| | ASCUS | 0 | 2 | 0 | 0 | 1 | 0 | 1 | 0 |
| | LSIL | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| | ASC-H | 0 | 1 | 0 | 0 | 4 | 1 | 2 | 0 |
| | AGUS | 2 | 10 | 3 | 0 | 1 | 12 | 1 | 1 |
| | HSIL | 1 | 2 | 2 | 0 | 4 | 3 | 9 | 0 |
| | CANCER | 2 | 2 | 1 | 0 | 0 | 1 | 1 | 9 |
**Table 14: “True HSIL” 8x8 contingency table, all sites combined**
| | | TIS | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | UNSAT | NILM | ASCUS | LSIL | ASC-H | AGUS | HSIL | CANCER |
| I2 | UNSAT | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| | NILM | 0 | 4 | 0 | 0 | 0 | 0 | 0 | 0 |
| | ASCUS | 0 | 3 | 12 | 1 | 7 | 0 | 2 | 1 |
| | LSIL | 0 | 2 | 7 | 28 | 7 | 0 | 5 | 0 |
| | ASC-H | 0 | 0 | 16 | 13 | 58 | 1 | 23 | 2 |
| | AGUS | 0 | 1 | 3 | 0 | 1 | 1 | 3 | 0 |
| | HSIL | 0 | 3 | 12 | 26 | 44 | 6 | 243 | 5 |
| | CANCER | 0 | 0 | 0 | 1 | 0 | 1 | 16 | 12 |
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 12
{12}
**Table 15: “True Cancer” 8x8 contingency table, all sites combined**
| | TIS | | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | UNSAT | NILM | ASCUS | LSIL | ASC-H | AGUS | HSIL | CANCER |
| I2 | UNSAT | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| | NILM | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| | ASCUS | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 |
| | LSIL | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 |
| | ASC-H | 0 | 0 | 1 | 1 | 2 | 0 | 0 | 0 |
| | AGUS | 0 | 0 | 0 | 1 | 0 | 6 | 0 | 8 |
| | HSIL | 0 | 0 | 0 | 0 | 1 | 0 | 19 | 1 |
| | CANCER | 0 | 0 | 0 | 0 | 0 | 4 | 5 | 63 |
Tables 17 through 19 summarize the descriptive diagnosis sensitivity and specificity estimates with 95% confidence intervals for all slides, and for slides prepared on ThinPrep 2000 System and ThinPrep 5000 Processor separately. Abbreviations for diagnostic thresholds are shown in Table 16:
**Table 16: Abbreviations for diagnostic thresholds**
| Threshold | Negative | Positive |
| --- | --- | --- |
| ASCUS+ | NILM | ASCUS, LSIL, ASC-H, AGUS, HSIL, Cancer |
| LSIL+ | NILM, ASCUS | LSIL, ASC-H, AGUS, HSIL, Cancer |
| ASC-H+ | NILM, ASCUS, LSIL | ASC-H, AGUS, HSIL, Cancer |
| HSIL+ | NILM, ASCUS, LSIL, ASC-H, AGUS | HSIL, Cancer |
Table 17 summarizes the descriptive diagnosis sensitivity and specificity estimates with 95% confidence intervals for all three sites combined. In all abnormal categories, the sensitivity for the I2 was higher than the TIS across all thresholds listed in Table 17. For specificity, there was a slight decrease for the I2 as compared to the TIS.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 13
{13}
**Table 17: TIS vs. I2, descriptive diagnosis summary (all slides)**
| Threshold | Sensitivity | | | Specificity | | |
| --- | --- | --- | --- | --- | --- | --- |
| | TIS (95% CI) | I2 (95% CI) | Difference (95% CI) | TIS (95% CI) | I2 (95% CI) | Difference (95% CI) |
| **ASCUS+** | 86.0% (84.7%, 87.3%) | 89.8% (88.6%, 90.9%) | 3.8% (2.6%, 5.0%) | 89.8% (88.9%, 90.6%) | 87.9% (86.9%, 88.8%) | -1.9% (-2.8%, -1.0%) |
| **LSIL+** | 77.8% (76.0%, 79.6%) | 83.7% (82.0%, 85.2%) | 5.8% (4.1%, 7.5%) | 92.5% (91.7%, 93.2%) | 90.6% (89.8%, 91.4%) | -1.9% (-2.6%, -1.2%) |
| **ASC-H+** | 73.3% (70.4%, 75.9%) | 80.7% (78.1%, 83.0%) | 7.4% (4.7%, 10.1%) | 92.7% (92.0%, 93.3%) | 91.1% (90.4%, 91.8%) | -1.6% (-2.1%, -1.0%) |
| **HSIL+** | 59.6% (55.9%, 63.3%) | 67.5% (63.9%, 70.9%) | 7.9% (4.5%, 11.2%) | 95.1% (94.6%, 95.6%) | 94.0% (93.4%, 94.6%) | -1.1% (-1.6%, -0.6%) |
| **UNSAT** | 78.9% (71.6%, 84.7%) | 77.6% (70.2%, 83.5%) | -1.4% (-7.3%, 4.5%) | 98.4% (98.1%, 98.6%) | 98.4% (98.1%, 98.7%) | 0.1% (-0.2%, 0.3%) |
In addition, the data presented in Table 18 and Table 19 are stratified by the type of processor used (ThinPrep 2000 System or ThinPrep 5000 Processor). In all abnormal cases, the sensitivity for the I2 was higher than the TIS across all thresholds. There was a slight decrease in specificity for the I2 as compared to the TIS.
**Table 18: TIS vs. I2, descriptive diagnosis summary (ThinPrep 2000 System-processed slides only)**
| Threshold | Sensitivity | | | Specificity | | |
| --- | --- | --- | --- | --- | --- | --- |
| | TIS [# of reads] (95% CI) | I2 [# of reads] (95% CI) | Difference [# of reads] (95% CI) | TIS [# of reads] (95% CI) | I2 [# of reads] (95% CI) | Difference [# of reads] (95% CI) |
| **ASCUS+** | 85.7% [1209/1411] (83.8%, 87.4%) | 90.0% [1270/1411] (88.3%, 1.5%) | 4.3% [61/1411] (2.6%, 6.1%) | 90.3% [2006/2222] (89.0%, 91.4%) | 88.9% [1975/2222] (87.5%, 90.1%) | -1.4% [-31/2222] (-2.7%, -0.1%) |
| **LSIL+** | 77.6% [820/1057] (75.0%, 80.0%) | 84.3% [891/1057] (82.0%, 86.4%) | 6.7% [71/1057] (4.3%, 9.1%) | 92.7% [2388/2576] (91.6%, 93.6%) | 91.3% [2353/2576] (90.2%, 92.4%) | -1.4% [-35/2576] (-2.3%, -0.4%) |
| **ASC-H+** | 73.1% [370/506] (69.1%, 76.8%) | 81.8% [414/506] (78.2%, 84.9%) | 8.7% [44/506] (4.9% to 12.5%) | 92.8% [2903/3127] (91.9% to 93.7%) | 91.1% [2849/3127] (90.1% to 92.1%) | -1.7% [-54/3127] (-2.5%, -1.0%) |
| **HSIL+** | 59.0% [214/363] (53.8%, 63.9%) | 70.2% [255/363] (65.4%, 74.7%) | 11.3% [41/363] (6.4%, 16.1%) | 95.4% [3118/3270] (94.6%, 96.0%) | 94.2% [3081/3270] (93.4%, 95.0%) | -1.1% [-37/3270] (-1.8%, -0.5%) |
| **UNSAT** | 83.3% [65/78] (73.5%, 90.0%) | 82.1% [64/78] (72.1%, 89.0%) | -1.3% [1/78] (-8.9%, 6.2%) | 98.6% [3647/3699] (98.2%, 98.9%) | 98.6% [3649/3699] (98.2%, 99.0%) | 0.1% [2/3699] (-0.3%, 0.4%) |
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 14
{14}
**Table 19: TIS vs. I2, descriptive diagnosis summary (ThinPrep 5000 Sytem-processed slides only)**
| Threshold | Sensitivity | | | Specificity | | |
| --- | --- | --- | --- | --- | --- | --- |
| | TIS [# of reads] (95% CI) | I2 [# of reads] (95% CI) | Difference [# of reads] (95% CI) | TIS [# of reads] (95% CI) | I2 [# of reads] (95% CI) | Difference [# of reads] (95% CI) |
| **ASCUS+** | 86.4% [1190/1377] (84.5%, 88.1%) | 89.6% [1234/1377] (87.9%, 91.1%) | 3.2% [44/1377] (1.6%, 4.8%) | 89.3% [1989/2228] (87.9%, 90.5%) | 86.8% [1935/2228] (85.4%, 88.2%) | -2.4% [-54/2228] (-3.8%, -1.1%) |
| **LSIL+** | 78.1% [796/1019] (75.5%, 80.5%) | 83.0% [846/1019] (80.6%, 85.2%) | 4.9% [50/1019] (2.5%, 7.3%) | 92.2% [2385/2586] (91.1%, 93.2%) | 89.9% [2324/2586] (88.6%, 91.0%) | -2.4% [-61/2586] (-3.4%, -1.4%) |
| **ASC-H+** | 73.4% [354/482] (69.3%, 77.2%) | 79.5% [383/482] (75.6%, 82.8%) | 6.0% [29/482] (2.2%, 9.8%) | 92.5% [2888/3123] (91.5%, 93.3%) | 91.1% [2845/3123] (90.0%, 92.0%) | -1.4% [-43/3123] (-2.2%, -0.6%) |
| **HSIL+** | 60.4% [194/321] (55.0%, 65.6%) | 64.5% [207/321] (59.1%, 69.5%) | 4.0% [13/321] (-0.6%, 8.6%) | 94.9% [3116/3284] (94.1%, 95.6%) | 93.8% [3082/3284] (93.0%, 94.6%) | -1.0% [-34/3284] (-1.7%, -0.3%) |
| **UNSAT** | 73.9% [51/69] (62.5%, 82.8%) | 72.5% [50/69] (61.0%, 81.6%) | -1.4% [1/69] (-11.3%, 8.4%) | 98.2% [3628/3696] (97.7%, 98.5%) | 98.2% [3630/3696] (97.7%, 98.6%) | 0.1% [2/3696] (-0.3%, 0.4%) |
# Unadjudicated Marginal Frequencies
Table 20 shows the descriptive diagnosis marginal frequencies for benign cellular changes for all sites combined. The results demonstrate that the unadjudicated marginal frequencies for the TIS vs. I2 reviews are comparable.
**Table 20: Unadjudicated marginal frequencies – all sites combined**
| | TIS Review | | I2 Review | |
| --- | --- | --- | --- | --- |
| Number of reads | 7542 | | 7542 | |
| Descriptive Diagnosis | N | % | N | % |
| **Benign Cellular Changes** | 402 | 5.3% | 420 | 5.6% |
| **Organisms** | | | | |
| Trichomonas vaginalis | 20 | 0.3% | 28 | 0.4% |
| Fungal organisms consistent with Candida spp. | 122 | 1.6% | 128 | 1.7% |
| Shift in Flora s/o bacterial vaginosis | 183 | 2.4% | 208 | 2.8% |
| Bacteria consistent with Actinomyces spp. | 2 | 0.0% | 3 | 0.0% |
| Cellular changes consistent with Herpes virus | 2 | 0.0% | 1 | 0.0% |
| Other infection | 0 | 0.0% | 0 | 0.0% |
| **Other Non-Neoplastic Findings** | | | | |
| Reactive cellular changes assoc. w/ inflammation | 34 | 0.5% | 16 | 0.2% |
| Atrophy | 33 | 0.4% | 26 | 0.3% |
| Reactive cellular changes assoc. w/ radiation | 0 | 0.0% | 0 | 0.0% |
| Reactive cellular changes assoc. w/ IUD | 0 | 0.0% | 1 | 0.0% |
| Glandular cells status post hysterectomy | 0 | 0.0% | 0 | 0.0% |
| Endometrial cells in a woman ≥ 45 yrs of age | 6 | 0.1% | 9 | 0.1% |
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 15
{15}
# 3. Subgroup Analyses
No further subgroup analyses were performed.
# 4. Pediatric Extrapolation
In this premarket application, existing clinical data was not leveraged to support approval of a pediatric patient population.
# **E. Financial Disclosure**
The Financial Disclosure by Clinical Investigators regulation (21 CFR 54) requires applicants who submit a marketing application to include certain information concerning the compensation to, and financial interests and arrangement of, any clinical investigator conducting clinical studies covered by the regulation. The pivotal clinical study included 3 investigators. None of the clinical investigators had disclosable financial interests/arrangements as defined in sections 54.2(a), (b), (c), and (f). The information provided does not raise any questions about the reliability of the data.
# **XI. SUMMARY OF SUPPLEMENTAL CLINICAL INFORMATION**
# **A. Cytotechnologist Workload Assessment Studies**
To determine the workload limit for the ThinPrep Integrated Imager (I2) in Sequential Modality, the CT screening rates were recorded during the clinical study when the I2 was used in Sequential Modality. Additionally, bench testing was conducted for both the Sequential Modality and Batched Modality.
# 1. CT Workload Assessment in the Clinical Study
To assess the CT workload in the clinical setting, results from the clinical study (described in Section X) were evaluated. Nine CTs recorded the number of hours worked each day and the number of slides screened for both the ThinPrep Imaging System (TIS) and ThinPrep Integrated Imager (I2) reviews. Only the Sequential Modality of I2 review was evaluated. The experience levels of the cytologists ranged from 4 to 30 years. During the study, the CT screening times for both TIS review and I2 review included automated screening of the 22 FOVs (FOV only), full microscopic review (FMR only) of the slide if the automated screening was not applicable, and automated screening of the 22 FOVs followed by full microscopic review (FOV + FMR) of the slide when abnormal cells were identified during automated screening.
The number of hours each CT screened slides per day varied due to logistical issues and scheduling. Using the I2, CTs scanned and reviewed 67% of the slides that CTs reviewed when using TIS. The data are summarized below.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 16
{16}
**Table 21: CT screening rates**
| | **TIS Average slides/hour** | **I2 Average slides/hour** |
| --- | --- | --- |
| **Site 1** | | |
| **CT 1** | 9.8 | 9.9 |
| **CT 2** | 10.4 | 9.7 |
| **CT 3** | 11.1 | 8.1 |
| **Site 2** | | |
| **CT 1** | 6.2 | 6.1 |
| **CT 2** | 9.0 | 6.4 |
| **CT 3** | 9.1 | 6.5 |
| **Site 3** | | |
| **CT 1** | 9.2 | 6.6 |
| **CT 2** | 9.9 | 6.8 |
| **CT 3** | 10.1 | 6.5 |
| **Combined Median** | **9.8** | **6.6** |
2. CT Workload Assessment in Bench Testing
Bench testing was also performed to characterize the screening volumes for CTs.
The ThinPrep Integrated Imager (I2) was used to collect data in two ways:
1) Each slide was imaged and then reviewed by a CT using the I2. This is referred to as Sequential Modality (i.e., imaging and slide review is performed consecutively by the CT).
2) All slides were imaged in advance as a batch using the I2 and then the CT reviewed slides as a batch. This is referred to as Batched Modality.
Three CTs participated in this study. The CTs independently reviewed slides over three days (screening slides for an 8 hour day) for each arm of the study. All slides were prepared from ThinPrep specimens of known cytology diagnoses, on a ThinPrep processor, and stained with ThinPrep Stain. Sets of 400 randomized slides per CT, each with approximately 10% abnormal diagnosis were provided. The CTs were blinded to the diagnoses. A minimum one week washout period occurred between study arms for each CT. The table below shows the total breakdown of the types of reviews performed.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 17
{17}
**Table 22: Total slides reviewed by review type**
| | Sequential Modality | | | | Batched Modality | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | CT #1 | CT #2 | CT #3 | Overall | CT #1 | CT #2 | CT #3 | Overall |
| **Total # slides reviewed** | 255 | 285 | 300 | 840 | 365 | 340 | 353 | 1058 |
| **# FOV only** | 212 | 179 | 239 | 630 | 308 | 226 | 265 | 799 |
| **# (FOV+FMR)** | 42 | 100 | 37 | 179 | 51 | 109 | 75 | 235 |
| **# FMR Only** | 1 | 6 | 4 | 11 | 6 | 5 | 13 | 24 |
| **% Autoscan* Referral** | 16% | 35% | 19% | 24% | 14% | 32% | 21% | 22% |
\* % autoscan = #(FOV+FMR) / Total # slides reviewed over 3 days
The study results are shown in Table 23. The median number of slides screened per day when the I2 was used for screening and reviewing of slides was 92 slides. CTs using the I2 on the Sequential Modality and Batched Modality reviewed 67% and 86%, respectively, of the slides that CTs could have reviewed when using ThinPrep Imaging System (TIS). The demonstrated rate for the Sequential Modality is consistent with the results from the clinical study.
**Table 23: Cytotechnologist daily slide review rates**
| Modality | CT | # Slides Reviewed | | | | Overall Daily Median |
| --- | --- | --- | --- | --- | --- | --- |
| | | Day 1 | Day 2 | Day 3 | Daily Median | |
| **Sequential Modality** | **1** | 87 | 80 | 88 | 87 | 92 (67%*) |
| | **2** | 90 | 100 | 95 | 95 | |
| | **3** | 92 | 108 | 100 | 100 | |
| **Batched Modality** | **1** | 119 | 123 | 123 | 123 | 119 (86%*) |
| | **2** | 124 | 106 | 110 | 110 | |
| | **3** | 119 | 120 | 114 | 119 | |
\* Percentage with regard to TIS being 100%
The agreement of the CT diagnoses was determined based on the adjudicated results, as shown in Table 24. The study results indicate that the accuracy from the bench testing is comparable to that observed from the clinical study.
**Table 24: PPA and NPA results based on adjudicated results\***
| CT | Sequential Modality | | Batched Modality | |
| --- | --- | --- | --- | --- |
| | PPA | NPA | PPA | NPA |
| **1** | 100% | 97% | 97% | 96% |
| **2** | 100% | 76% | 100% | 79% |
| **3** | 91% | 94% | 100% | 90% |
| **Overall** | 97% | 89% | 99% | 89% |
\* Positive results mean ASC-US+
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 18
{18}
# 3. CT Workload Limits for Sequential Modality and Batched Modality
Workload is defined by Clinical Laboratory Improvement Amendments (CLIA) of 1988 as a maximum limit of 100 slides in no less than an 8-hour workday$^{2}$. This refers to a full manual review of 100 slides.
Based on the results from the clinical study and bench testing, when ThinPrep Integrated Imager (I2) was used in the Sequential Modality, CTs could review 67% of the slides that CTs reviewed when using ThinPrep Imaging System (TIS). To help laboratories determine the workload limits when using the I2, Hologic recommends the values and method specified in Table 25, should be used for Sequential Modality.
**Table 25: Values for calculating workload, Sequential Modality**
| FMR = 1 slide |
| --- |
| FOV = 0.85 slide |
| FMR + FOV = 1.85 slides |
| Upper Limit = 100 slides |
| 0.85 * FOV + 1.85 * (FOV + FMR) + 1.0 * FMR = 100 |
The bench testing suggested that when I2 was used in Batched Modality, CTs could review 86% of the slides that CTs reviewed when using TIS. To help laboratories determine the workload limits when using the I2, Hologic recommends the values and method specified in Table 26 should be used for Batched Modality.
**Table 26: Values for calculating workload, Batched Modality**
| FMR = 1 slide |
| --- |
| FOV = 0.65 slide |
| FMR + FOV = 1.65 slides |
| Upper Limit = 100 slides |
| 0.65 * FOV + 1.65 * (FOV + FMR) + 1.0 * FMR = 100 |
# **XII. PANEL MEETING RECOMMENDATION AND FDA'S POST-PANEL ACTION**
In accordance with the provisions of section 515(c)(3) of the act as amended by the Safe Medical Devices Act of 1990, this PMA was not referred to the Hematology and Pathology Devices, an FDA advisory committee, for review and recommendation because the information in the PMA substantially duplicates information previously reviewed by this panel.
# **XIII. CONCLUSIONS DRAWN FROM PRECLINICAL AND CLINICAL STUDIES**
# **A. Effectiveness Conclusions**
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 19
{19}
A clinical study was conducted to demonstrate the performance of the ThinPrep Integrated Imager (I2) for assisting in primary cervical cancer screening of ThinPrep Pap Test slides. In the clinical study, the I2 was compared to the ThinPrep Imaging System (TIS) to determine the sensitivity and specificity when diagnosing specimens. Upon comparison of I2 review vs. TIS review, significantly higher sensitivity was observed in all abnormal categories, as defined by the Bethesda System¹, and there was a slight decrease in specificity, although not significant.
- For ASCUS+ slides, the increase of sensitivity was 3.8% (95% CI: 2.6% to 5.0%) and a decrease of specificity was -1.9% (95% CI: -2.8% to -1.0%).
- For LSIL+ slides, the increase in sensitivity was 5.8% (95% CI: 4.1% to 7.5%) and a decrease of specificity was -1.9% (95% CI: -2.6 to -1.2%).
- For HSIL+ the increase in sensitivity was 7.9% (95% CI: 4.5% to 11.2%) and a decrease in specificity of -1.1% (95% CI: -1.6% to -0.6%).
In the clinical study and bench testing, workload assessment was conducted. The screening volumes for the CTs was recorded when using the I2, and it was concluded that the imaging and review of slides is within the CLIA guidelines for the total number of slides that can be screened in one day, based on the values in Tables 25 and 26. The number of slides that a CT can scan and review in one day is less on the I2 than the TIS.
In the non-clinical studies, the within- and between-instrument reproducibilities and FOV reproducibility of the I2 was investigated. In all studies the performance of the I2 was comparable to that of the TIS, in identifying abnormalities in the 22 FOVs presented during the Autolocate mode of slide review.
Taken together, the data from the clinical study and non-clinical study demonstrate that the I2 is effective for its intended use to assist in primary cervical cancer screening of ThinPrep Pap Test slides for the presence of atypical cells, cervical neoplasia, including its precursor lesions, and carcinoma as well as all other cytological criteria as defined by the Bethesda System¹.
## **B. Safety Conclusions**
Potential for adverse events is relative to an incorrect diagnosis leading either to unnecessary care or delayed follow up care. (See effectiveness conclusions above for test sensitivity and specificity.) The diagnoses made for the slides in this study were prepared from de-identified patient specimens and thus not able to be used in patient care.
## **C. Benefit-Risk Determination**
The results of the clinical investigation demonstrated that ThinPrep Pap Test slides reviewed with the ThinPrep Integrated Imager result in equivalent diagnosis to that for a slide reviewed using the ThinPrep Imaging System (TIS) when used with the Bethesda System: Terminology for Reporting Results of Cervical Cytology¹.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 20
{20}
The risks are associated with the Integrated Imager’s ability to correctly determine the 22 fields most likely to contain abnormal cells. If none of the 22 fields contains abnormal cells when the disease is present, this may result in a false negative interpretation. If the pathologist interprets a slide as containing abnormalities when no disease is present, this may result in false positive interpretation. A false positive interpretation may result in an unnecessary colposcopy exam which may include one or more biopsies. Colposcopy with biopsy is a low risk procedure. A false negative test result may result in delayed diagnosis or treatment; this is the major risk of concern with this device. Based on the demonstrated performance characteristics of the device, a false negative test result is no more likely to occur with the new device as compared to the approved TIS device.
The benefits of cervical precancer screening are well-documented, and the performance characteristics of this device indicate that similar benefits are expected with its use. The risks of a false positive test result are insignificant. The risks of a false negative test result are of more concern, and in the worst case scenario, a woman can die of cervical cancer as an indirect, downstream result of a false negative test.
The worst case scenario of a false negative test result (noted earlier) is mitigated by multiple factors which include aspects of the standard of care for women in the context of cervical precancer screening. These include repeat testing according to published screening guidelines, and the natural history of cervical precancers which generally evolve into invasive cancers over years to a decade or more (if ever). Even when invasive (cancerous), the disease remains curable in all but the most advanced stages of disease. Accordingly, after a false negative test, there are likely to be multiple additional opportunities to identify the disease. Despite the lack of actual clinical outcome data, though a limitation of the assessment of the performance characteristics of the device, the available data are sufficient to provide reasonable assurance of safety and effectiveness of the device.
# 1. 1. Patient Perspectives
This submission did not include specific information on patient perspectives for this device.
# **D. Overall Conclusions**
The data in this application support the reasonable assurance of safety and effectiveness of this device when used in accordance with the indications for use and product labeling. The provided studies support use of the ThinPrep Integrated Imager to assist in primary cervical cancer screening of ThinPrep Pap Test slides.
# **XIV. CDRH DECISION**
CDRH issued an approval order on April 18, 2018.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 21
{21}
The applicant's manufacturing facilities have been inspected and found to be in compliance with the device Quality System (QS) regulation (21 CFR 820).
# **XV. APPROVAL SPECIFICATIONS**
Directions for use: See device labeling.
Hazards to Health from Use of the Device: See Indications, Contraindications, Warnings, Precautions, and Adverse Events in the device labeling.
Post-approval Requirements and Restrictions: See approval order.
# **XVI. REFERENCES**
1. Nayar R, Wilbur DC. (eds). The Bethesda System for Reporting Cervical Cytology: Definitions, Criteria, and Explanatory Notes. 3rd ed. Cham, Switzerland: Springer: 2015.
2. Clinical Laboratory Improvement Amendments of 1988; Final Rule. Federal Register. Feb. 28 1992; 57: 493.1483.
PMA P950039/S036: FDA Summary of Safety and Effectiveness Data
Page 22
