ThinPrep 5000 Processor Instructions for Use Rev 001

INTENDED USE The ThinPrep® 5000 processor is intended as a replacement for the conventional method of Pap smear preparation for use in screening for the presence of atypical cells, cervical cancer, or its precursor lesions (Low-grade Squamous Intraepithelial Lesions, High-grade Squamous Intraepithelial Lesions), as well as all other cytologic categories as defined by The Bethesda System for Reporting Cervical/Vaginal Cytologic Diagnoses1.

SUMMARY AND EXPLANATION OF THE SYSTEM The ThinPrep process begins with the patient’s gynecologic sample being collected by the clinician using a cervical sampling device which, rather than being smeared on a microscope slide, is immersed and rinsed in a vial filled with 20 mL of PreservCyt® Solution (PreservCyt). The ThinPrep sample vial is then capped, labeled, and sent to a laboratory equipped with a ThinPrep 5000 processor. At the laboratory, the PreservCyt sample vial is barcoded along with the test request form to establish a sample chain of custody and is placed into a ThinPrep 5000 processor. A glass slide bearing the same sample identification number as on the sample vial is loaded into the processor. A gentle dispersion step mixes the cell sample by currents in the fluid that are strong enough to separate debris and disperse mucus, but gentle enough to have no adverse effect on cell appearance. The cells are then captured on a gynecological ThinPrep Pap test filter that is specifically designed to collect cells. The ThinPrep 5000 processor constantly monitors the rate of flow through the ThinPrep Pap test filter during the collection process in order to prevent the cellular presentation from being too scant or too dense. A thin layer of cells is then transferred to a glass slide in a 20 mm-diameter circle, and the slide is automatically deposited into a fixative solution.

MAN-04008-001 Rev. 001

Page 2 of 27

The ThinPrep Sample Preparation Process

(1) Dispersion

(2) Cell Collection

(3) Cell Transfer

The ThinPrep Pap test filter rotates within the sample vial, creating currents in the fluid that are strong enough to separate debris and disperse mucus, but gentle enough to have no adverse effect on cell appearance.

A gentle vacuum is created within the ThinPrep Pap test filter, which collects cells on the exterior surface of the membrane. Cell collection is controlled by the ThinPrep 5000 processor’s software that monitors the rate of flow through the ThinPrep Pap test filter.

After the cells are collected on the membrane, the ThinPrep Pap test filter is inverted and gently pressed against the ThinPrep microscope slide. Natural attraction and slight positive air pressure cause the cells to adhere to the ThinPrep microscope slide resulting in an even distribution of cells in a defined circular area.

As with conventional Pap smears, slides prepared with the ThinPrep® 5000 processor are examined in the context of the patient’s clinical history and information provided by other diagnostic procedures such as colposcopy, biopsy, and human papillomavirus (HPV) testing, to determine patient management. The PreservCyt® Solution component of the ThinPrep 5000 system is an alternative collection and transport medium for gynecologic specimens tested with Hologic’s APTIMA COMBO 2® CT/NG Assay and the Digene Hybrid Capture System HPV DNA assay. Refer to the respective manufacturer’s package inserts for instructions for using PreservCyt Solution for collection, transport, storage, and preparation of specimens for use in those systems. The PreservCyt Solution component of the ThinPrep 5000 system is also an alternative collection and transport medium for gynecologic specimens tested with the Roche Diagnostics COBAS AMPLICORTM CT/NG assay. Refer to Hologic’s labeling (Document #MAN-02063-001) for instructions for using PreservCyt Solution for collection, transport, storage, and preparation of specimens and to the Roche Diagnostics COBAS AMPLICOR CT/NG package insert for instructions for use of that system.

MAN-04008-001 Rev. 001

Page 3 of 27

LIMITATIONS 

Gynecologic samples collected for preparation using the ThinPrep 5000 processor should be collected using a broom-type or endocervical brush/plastic spatula combination collection devices.



Preparation of microscope slides using the ThinPrep 5000 processor should be performed only by personnel who have been trained by Hologic or by organizations or individuals designated by Hologic.



Evaluation of microscope slides produced with the ThinPrep 5000 processor should be performed only by cytotechnologists and pathologists who have been trained to evaluate ThinPrep-prepared slides by Hologic or by organizations or individuals designated by Hologic.



Supplies used by the ThinPrep 5000 processor are those designed and supplied by Hologic specifically for the ThinPrep 5000 processor. These include PreservCyt Solution vials, ThinPrep Pap test filters, and ThinPrep microscope slides. These supplies are required for proper performance of the system and cannot be substituted. Product performance will be compromised if other supplies are used. After use, supplies should be disposed of in accordance with local, state, and federal regulations.



A ThinPrep Pap test filter must be used only once and cannot be reused.



The performance of HPV DNA and CT/NG testing on reprocessed sample vials has not been evaluated.

CONTRAINDICATIONS 

Chlamydia trachomatis and Neisseria gonorrhoeae testing using the Hologic APTIMA COMBO 2® CT/NG assay and the Roche Diagnostics COBAS AMPLICOR assay should not be performed on a sample that has already been processed using the ThinPrep 5000 processor.

WARNINGS 

For In Vitro Diagnostic Use



Danger. PreservCyt Solution contains methanol. Toxic if swallowed. Toxic if inhaled. Causes damage to organs. Flammable liquid and vapor. Keep away from heat, sparks, open flames and hot surfaces. Other solutions cannot be substituted for PreservCyt Solution. PreservCyt Solution should be stored and disposed of in accordance with all applicable regulations.

PRECAUTIONS 

This equipment generates, uses and can radiate radio frequency energy, and if not installed and used in accordance with the operator’s manual, may cause interference to radio communications. Operation of this equipment in a residential area is likely to cause harmful interference, in which case the user will be required to correct the interference at his/her own expense.



PreservCyt Solution with cytologic sample intended for ThinPrep Pap testing must be stored between 15oC (59oF) and 30oC (86oF) and tested within 6 weeks of collection.



PreservCyt Solution with cytologic sample intended for CT/NG testing using the Roche Diagnostics COBAS AMPLICOR CT/NG test must be stored between 4oC (39oF) and 25oC (77oF) and tested within 6 weeks of collection.

MAN-04008-001 Rev. 001

Page 4 of 27



PreservCyt Solution was challenged with a variety of microbial and viral organisms. The following table presents the starting concentrations of viable organisms, and the log reduction of viable organisms found after 15 minutes in the PreservCyt Solution. As with all laboratory procedures, universal precautions should be followed.

Initial Concentration

Log Reduction After 15 Minutes

Candida albicans

5.5 x 105 CFU/mL

>4.7

Aspergillus niger*

4.8 x 105 CFU/mL

>2.7

Escherichia coli

2.8 x 105 CFU/mL

>4.4

Staphylococcus aureus

2.3 x 105 CFU/mL

>4.4

2.5 x 10 CFU/mL

5

>4.4

9.4 x 10 CFU/mL

5

>4.9

6.0 x 106 PFU/mL

>5.5***

1.0 x 107.5 TCID50/mL

>7.0***

Organism

Pseudomonas aeruginosa Mycobacterium tuberculosis** Rabbitpox virus HIV-1

* After 1 hour >4.7 log reduction ** After 1 hour >5.7 log reduction *** Data is for 5 minutes

PERFORMANCE CHARACTERISTICS: REPORT OF CLINICAL STUDIES The ThinPrep 5000 processor is technologically similar to the ThinPrep 2000 system. The performance characteristics of the ThinPrep 5000 processor are predicated on those of the ThinPrep 2000 system. Both the clinical studies for the ThinPrep 2000 system and those comparing the ThinPrep 5000 processor to the ThinPrep 2000 are described in the following sections.

ThinPrep 2000 System Compared to Conventional Pap Smear A prospective multi-center clinical study was conducted to evaluate the performance of the ThinPrep 2000 system in direct comparison to the conventional Pap smear. The objective of the ThinPrep clinical study was to demonstrate that gynecologic specimens prepared using the ThinPrep 2000 system were at least as effective as conventional Pap smears for the detection of atypical cells and cervical cancer or its precursor lesions in a variety of patient populations. In addition, an assessment of specimen adequacy was performed. The initial clinical study protocol was a blinded, split sample, matched pair study, for which a conventional Pap smear was prepared first, and the remainder of the sample (the portion that normally would have been discarded) was immersed and rinsed into a vial of PreservCyt Solution. At the laboratory, the PreservCyt sample vial was placed into a ThinPrep 2000 processor and a slide was then prepared from the patient’s sample. ThinPrep and conventional

MAN-04008-001 Rev. 001

Page 5 of 27

Pap smear slides were examined and diagnosed independently. Reporting forms containing patient history as well as a checklist of all possible categories of The Bethesda System were used to record the results of the screening. A single independent pathologist reviewed all discrepant and positive slides from all sites in a blinded fashion to provide a further objective review of the results. Laboratory and Patient Characteristics Cytology laboratories at three screening centers (designated as S1, S2, and S3) and three hospital centers (designated as H1, H2, and H3) participated in the clinical study. The screening centers in the study serve patient populations (screening populations) with rates of abnormality (Low-grade Squamous Intraepithelial Lesion [LSIL] and more severe lesions) similar to the United States average of less than 5%.2 The hospital centers in the study serve a high risk referral patient population (hospital populations) characterized by high rates (>10%) of cervical abnormality. Data on race demographics was obtained for 70% of the patients that participated in the study. The study population consisted of the following race groups: Caucasian (41.2%), Asian (2.3%), Hispanic (9.7%), African American (15.2%), Native American (1.0%) and other groups (0.6%). Table 1 describes the laboratories and the patient populations. Table 1: Site Characteristics

Site

Laboratory Characteristics Laboratory Type of Patient Volume - Smears Cases Population per Year

Clinical Study Demographics Previous Convent. Patient PostAbnormal Prevalence Age Range Menopausal Pap Smear LSIL+

S1

Screening

300,000

1,386

18.0–84.0

10.6%

8.8%

2.3%

S2

Screening

100,000

1,668

18.0–60.6

0.3%

10.7%

2.9%

S3

Screening

96,000

1,093

18.0–48.8

0.0%

7.1%

3.8%

H1

Hospital

35,000

1,046

18.1–89.1

8.1%

40.4%

9.9%

H2

Hospital

40,000

1,049

18.1–84.4

2.1%

18.2%

12.9%

H3

Hospital

37,000

981

18.2–78.8

11.1%

38.2%

24.2%

Clinical Study Results The diagnostic categories of The Bethesda System were used as the basis of the comparison between conventional and ThinPrep® findings from the clinical study. The diagnostic classification data and statistical analyses for all clinical sites are presented in Tables 2 through 11. Cases with incorrect paperwork, patient’s age less than 18 years, cytologically unsatisfactory slides, or patients with a hysterectomy were excluded from this analysis. Few cases of cervical cancer (0.02%3) were represented in the clinical study, as is typical in the United States patient population.

MAN-04008-001 Rev. 001

Page 6 of 27

Table 2: Diagnostic Classification Table, All Categories Conventional NEG ASCUS AGUS LSIL HSIL SQ CA GL CA TOTAL 5224 295 3 60 11 0 0 5593 ThinPrep NEG 125 2 45 7 0 0 497 ASCUS 318 AGUS 13 2 3 0 1 0 1 20 LSIL 114 84 0 227 44 0 0 469 HSIL 11 15 0 35 104 2 0 167 SQ CA 0 0 0 0 0 1 0 1 GL CA 0 0 0 0 0 0 0 0 TOTAL 5680 521 8 367 167 3 1 6747 Abbreviations for Diagnoses: NEG = Normal or negative, ASCUS = Atypical Squamous Cells of Undetermined Significance, AGUS = Atypical Glandular Cells of Undetermined Significance, LSIL = Low-grade Squamous Intraepithelial Lesion, HSIL = High-grade Squamous Intraepithelial Lesion, SQ CA = Squamous Cell Carcinoma, GL CA = Glandular Cell Adenocarcinoma

Table 3: Three Category Diagnostic Classification Table Conventional ThinPrep

NEG

ASCUS/AGUS+

LSIL+

TOTAL

NEG

5224

298

71

5593

ASCUS/ AGUS+

331

132

54

1154

LSIL+

125

99

413

637

TOTAL

5680

529

538

6747

Table 4: Two Category Diagnostic Classification Table, LSIL and More Severe Diagnoses Conventional

ThinPrep NEG/ASCUS/ AGUS+ LSIL+ TOTAL

NEG/ASCUS/ AGUS+

LSIL+

TOTAL

5985

125

6110

224

413

637

6209

538

6747

MAN-04008-001 Rev. 001

Page 7 of 27

Table 5: Two Category Diagnostic Classification Table, ASCUS/AGUS and More Severe Diagnoses Conventional ThinPrep

NEG ASCUS/ AGUS+ TOTAL

NEG

ASCUS/AGUS+

TOTAL

5224

369

5593

456

698

1154

5680

1067

6747

The diagnostic data analysis from the sites is summarized in Table 6 and 7. When the p-value is significant (p < 0.05), the method favored is indicated in the tables. Table 6: Results by Site, LSIL and More Severe Lesions Site

Cases

ThinPrep LSIL+

Convent. LSIL+

Increased Detection*

p-Value

Method Favored

S1

1,336

46

31

48%

0.027

ThinPrep

S2

1,563

78

45

73%

<0.001

ThipPrep

S3

1,058

67

40

68%

<0.001

ThinPrep

H1

971

125

96

30%

<0.001

ThinPrep

H2

1,010

111

130

(15%)

0.135

Neither

H3

809

210

196

7%

0.374

Neither

®

*Increased detection = ThinPrep LSIL+ - Conventional LSIL+ x 100% Conventional LSIL+

For LSIL and more severe lesions, the diagnostic comparison statistically favored the ThinPrep® method at four sites and was statistically equivalent at two sites.

MAN-04008-001 Rev. 001

Page 8 of 27

Table 7: Results by Site, ASCUS/AGUS and More Severe Lesions Site

Cases

ThinPrep ASCUS+

Convent. ASCUS+

Increased Detection*

p-Value

Method Favored

S1

1,336

117

93

26%

0.067

Neither

S2

1,563

124

80

55%

<0.001

ThinPrep

S3

1,058

123

81

52%

<0.001

ThinPrep

H1

971

204

173

18%

0.007

ThinPrep

H2

1,010

259

282

(8%)

0.360

Neither

H3

809

327

359

(9%)

0.102

Neither

*Increased detection = ThinPrep ASCUS+ - Conventional ASCUS+ x 100% Conventional ASCUS+

For ASCUS/AGUS and more severe lesions, the diagnostic comparison statistically favored the ThinPrep method at three sites and was statistically equivalent at three sites. One pathologist served as an independent reviewer for the six clinical sites, receiving both slides from cases where the two methods were either abnormal or discrepant. Since a true reference cannot be determined in such studies and therefore true sensitivity cannot be calculated, the use of an expert cytologic review provides an alternative to histologic confirmation by biopsy or human papillomavirus (HPV) testing as a means for determining the reference diagnosis. The reference diagnosis was the more severe diagnosis from either of the ThinPrep or conventional Pap slides as determined by the independent pathologist. The number of slides diagnosed as abnormal at each site, compared to the reference diagnosis of the independent pathologist, provides the proportion of LSIL or more severe lesions (Table 8) and the proportion of ASCUS/AGUS or more severe lesions (Table 9). The statistical analysis allows a comparison of the two methods and a determination of which method is favored when using the independent pathologist for expert cytologic review as the adjudicator of the final diagnosis.

MAN-04008-001 Rev. 001

Page 9 of 27

Table 8: Independent Pathologist Results by Site, LSIL and More Severe Lesions Cases Site

Positive by Independent Pathologist

ThinPrep Positive

Conventional Positive

p-Value

Method Favored

S1

50

33

25

0.170

Neither

S2

65

48

33

0.042

ThinPrep

S3

77

54

33

<0.001

ThinPrep

H1

116

102

81

<0.001

ThinPrep

H2

115

86

90

0.876

Neither

H3

126

120

112

0.170

Neither

For LSIL and more severe lesions, the diagnostic comparison statistically favored the ThinPrep method at three sites and was statistically equivalent at three sites.

Table 9: Independent Pathologist Results by Site, ASCUS/AGUS and More Severe Lesions Cases Site

Positive by Independent Pathologist

ThinPrep® Positive

Conventional Positive

p-Value

Method Favored

S1 S2 S3

92 101 109

72 85 95

68 59 65

0.900 0.005 <0.001

Neither ThinPrep ThinPrep

H1 H2 H3

170 171 204

155 143 190

143 154 191

0.237 0.330 1.000

Neither Neither Neither

For ASCUS/AGUS and more severe lesions, the diagnostic comparison statistically favored the ThinPrep method at two sites and was statistically equivalent at four sites.

MAN-04008-001 Rev. 001

Page 10 of 27

Table 10 below shows the summary for all sites of the descriptive diagnosis for all Bethesda System categories. Table 10: Summary of Descriptive Diagnosis ThinPrep

Descriptive Diagnosis

Conventional

Number of Patients: 6747

N

%

N

%

Benign Cellular Changes: Infection: Trichomonas Vaginalis Candida spp. Coccobacilli Actinomyces spp. Herpes Other Reactive Cellular Changes Associated with: Inflammation Atrophic Vaginitis Radiation Other Epithelial Cell Abnormalities: Squamous Cell: ASCUS favor reactive favor neoplastic undetermined LSIL HSIL Carcinoma Glandular Cell: Benign Endometrial cells in Postmenopausal Women Atypical Glandular Cells (AGUS) favor reactive favor neoplastic undetermined Endocervical Adenocarcinoma

1592

23.6

1591

23.6

136 406 690 2 3 155

2.0 6.0 10.2 0.0 0.0 2.3

185 259 608 3 8 285

2.7 3.8 9.0 0.0 0.1 4.2

353 32 2 25

5.2 0.5 0.0 0.4

385 48 1 37

5.7 0.7 0.0 0.5

1159

17.2

1077

16.0

501 128 161 213 469 167 1

7.4 1.9 2.4 3.2 7.0 2.5 0.0

521 131 140 250 367 167 3

7.7 1.9 2.1 3.7 5.4 2.5 0.0

7

0.1

10

0.1

21

0.3

9

0.1

9 0 12

0.1 0.0 0.2

4 3 2

0.1 0.0 0.0

0

0.0

1

0.0

Note: Some patients had more than one diagnostic subcategory.

MAN-04008-001 Rev. 001

Page 11 of 27

Table 11 shows the rates of detection for infection, reactive changes, and the total benign cellular changes for both the ThinPrep® and conventional methods at all sites. Table 11: Benign Cellular Changes Results

Benign Cellular Changes

ThinPrep

Conventional

N

%

N

%

Infection

1392

20.6

1348

20.0

Reactive Changes

412

6.1

471

7.0

Total*

1592

23.6

1591

23.6

* Total includes some patients that may have had both an infection and reactive cellular change.

Tables 12, 13, and 14 show the specimen adequacy results for the ThinPrep method and conventional smear method for all of the study sites. Of the 7,360 total patients enrolled, 7,223 are included in this analysis. Cases with patient’s age less than 18 years or patients with a hysterectomy were excluded from this analysis. Two additional clinical studies were conducted to evaluate specimen adequacy results when samples were deposited directly into the PreservCyt® vial, without first making a conventional Pap smear. This specimen collection technique is the intended use for the ThinPrep 2000 system. Tables 15 and 16 present the split sample and direct to vial results.

MAN-04008-001 Rev. 001

Page 12 of 27

Table 12: Summary of Specimen Adequacy Results Specimen Adequacy Number of Patients: 7223 Satisfactory Satisfactory for Evaluation but Limited by: Air-Drying Artifact Thick Smear Endocervical Component Absent Scant Squamous Epithelial Component Obscuring Blood Obscuring Inflammation No Clinical History Cytolysis Other Unsatisfactory for Evaluation: Air-Drying Artifact Thick Smear Endocervical Component Absent Scant Squamous Epithelial Component Obscuring Blood Obscuring Inflammation No Clinical History Cytolysis Other

ThinPrep

Conventional

N

%

N

%

5656

78.3

5101

70.6

1431

19.8

2008

27.8

1 9

0.0 0.1

136 65

1.9 0.9

1140

15.8

681

9.4

150

2.1

47

0.7

55 141 12 19 10 136 0 0

0.8 2.0 0.2 0.3 0.1 1.9 0.0 0.0

339 1008 6 119 26 114 13 7

4.7 14.0 0.1 1.6 0.4 1.6 0.2 0.1

25

0.3

11

0.2

106

1.5

47

0.7

23 5 0 0 31

0.3 0.1 0.0 0.0 0.4

58 41 0 4 9

0.8 0.6 0.0 0.1 0.1

Note: Some patients had more than one subcategory.

Table 13: Specimen Adequacy Results Conventional ThinPrep

SAT

SBLB

UNSAT

TOTAL

SAT

4316

1302

38

5656

SBLB

722

665

44

1431

UNSAT

63

41

32

136

TOTAL

5101

2008

114

7223

SAT=Satisfactory, SBLB=Satisfactory But Limited By, UNSAT=Unsatisfactory

MAN-04008-001 Rev. 001

Page 13 of 27

Table 14: Specimen Adequacy Results by Site ThinPrep Convent. ThinPrep Convent. ThinPrep Convent. SAT SAT SBLB SBLB UNSAT UNSAT Cases Cases Cases Cases Cases Cases

Site

Cases

S1

1,386

1092

1178

265

204

29

4

S2

1,668

1530

1477

130

178

8

13

S3

1,093

896

650

183

432

14

11

H1

1,046

760

660

266

375

20

11

H2

1,049

709

712

323

330

17

7

H3

981

669

424

264

489

48

68

All Sites

7,223

5656

5101

1431

2008

136

114

The Satisfactory But Limited By (SBLB) category can be broken down into many subcategories, one of which is the absence of Endocervical Component. Table 15 shows the Satisfactory But Limited By category “No ECC’s” for ThinPrep® and conventional slides. Table 15: Specimen Adequacy Results by Site, SBLB Rates for no Endocervical Component SBLB Due to No ECC’s ThinPrep ThinPrep Conventional Conventional SBLBSBLBSBLB- no SBLB- no no ECC’s (%) no ECC’s (%) ECC’s ECC’s

Site

Cases

S1

1,386

237

17.1%

162

11.7%

S2

1,668

104

6.2%

73

4.4%

S3

1,093

145

13.3%

84

7.7%

H1

1,046

229

21.9%

115

11.0%

H2

1,049

305

29.1%

150

14.3%

H3

981

120

12.2%

97

9.9%

All Sites

7,223

1140

15.8%

681

9.4%

For the results of the clinical study involving a split-sample protocol, there was a 6.4 percent difference between conventional and ThinPrep methods in detecting endocervical component. This is similar to previous studies using a split sample methodology.

Direct-to-vial Endocervical Component (ECC) Studies For the intended use of the ThinPrep® 2000 system, the cervical sampling device will be rinsed directly into a PreservCyt® vial, rather than splitting the cellular sample. It was expected that this would result in an increase in the pick-up of endocervical cells and metaplastic cells. To verify this hypothesis, two studies were performed using the direct-to-vial method and are summarized in Table 16. Overall, no difference was found between ThinPrep and conventional methods in these two studies.

MAN-04008-001 Rev. 001

Page 14 of 27

Table 16: Summary of Direct-to-vial Endocervical Component (ECC) Studies Study

Number of Evaluable Patients

SBLB due to No Endocervical Component

Comparable Conventional Pap Smear Percentage

299

9.36%

9.43%1

484

4.96%

4.38%2

Direct-to-Vial Feasibility Direct-to-Vial Clinical Study

1. Direct-to-Vial Feasibility study compared to overall clinical investigation conventional Pap smear SBLB-No Endocervical Component rate. 2. Direct-to-Vial Clinical study compared to site S2 clinical investigation conventional Pap smear SBLB-No Endocervical Component rate.

Direct-to-Vial HSIL+ Study Following initial FDA approval of the ThinPrep system, Hologic conducted a multi-site direct-tovial clinical study to evaluate the ThinPrep 2000 system versus conventional Pap smear for the detection of High Grade Squamous Intraepithelial and more severe lesions (HSIL+). Two types of patient groups were enrolled in the trial from ten (10) leading academic hospitals in major metropolitan areas throughout the United States. From each site, one group consisted of patients representative of a routine Pap test screening population and the other group made up of patients representative of a referral population enrolled at the time of colposcopic examination. The ThinPrep specimens were collected prospectively and compared against a historical control cohort. The historical cohort consisted of data collected from the same clinics and clinicians (if available) used to collect the ThinPrep specimens. These data were collected sequentially from patients seen immediately prior to the initiation of the study. The results from this study showed a detection rate of 511 / 20,917 for the conventional Pap smear versus 399 / 10,226 for the ThinPrep slides. For these clinical sites and these study populations, this indicates a 59.7% increase in detection of HSIL+ lesions for the ThinPrep specimens. These results are summarized in Table 17. Table 17: Summary of Direct-to-Vial HSIL+ Study Total CP (n)

HSIL+

Percent (%)

Total TP (n)

HSIL+

Percent (%)

Percent Change (%)

2,439

51

2.1

1,218

26

2.1

+2.1

S2

2,075

44

2.1

1,001

57

5.7

+168.5

S3

2,034

7

0.3

1,016

16

1.6

+357.6

S4

2,043

14

0.7

1,000

19

1.9

+177.3

S5

2,040

166

8.1

1,004

98

9.8

+20.0

S6

2,011

37

1.8

1,004

39

3.9

+111.1

S7

2,221

58

2.6

1,000

45

4.5

+72.3

S8

2,039

61

3.0

983

44

4.5

+49.6

S9

2,000

4

0.2

1,000

5

0.5

+150.0

S10

2,015

69

3.4

1,000

50

5.0

+46.0

20,917

511

2.4

10,226

399

3.9

59.7(p<0.001)

Site S1

Total

Percent Change (%) = ((TP HSIL+/TP Total)/(CP HSIL+/CP Total)-1) *100 MAN-04008-001 Rev. 001

Page 15 of 27

Glandular Disease Detection – Published Studies The detection of endocervical glandular lesions is an essential function of the Pap test. However, abnormal glandular cells in the Pap sample may also originate from the endometrium or from extrauterine sites. The Pap test is not intended to be a screening test for such lesions. When suspected glandular abnormalities are identified, their accurate classification as true glandular versus squamous lesions is important for proper evaluation and subsequent treatment (e.g. choice of excisional biopsy method versus conservative follow-up). Multiple peer-reviewed publications4-9 report on the improved ability of the ThinPrep 2000 system to detect glandular disease versus the conventional Pap smear. Although these studies do not consistently address sensitivity of different Pap testing methods in detecting specific types of glandular disease, the reported results are consistent with more frequent biopsy confirmation of abnormal glandular findings by the ThinPrep Pap test compared to conventional cytology. Thus, the finding of a glandular abnormality on a ThinPrep Pap test slide merits increased attention for definitive evaluation of potential endocervical or endometrial pathology.

ThinPrep 5000 Processor Compared to ThinPrep 2000 System A study was conducted to estimate the Positive Percent Agreement (PPA) and Negative Percent Agreement (NPA) for specimens processed on the ThinPrep 5000 processor as compared with processing using the ThinPrep 2000 System.

Clinical Study Design The study was a prospective, multi-center, split-sample, blinded evaluation of ThinPrep slides of known diagnoses generated from residual cytological specimens. The study was conducted at Hologic, Inc., Marlborough, MA and at two external laboratories in the United States. One thousand two hundred sixty (1260) specimens were procured for and selected from Hologic’s Residual Specimen Inventory for Hologic’s laboratory. At the external study sites specimens were from residual cytological specimens from the clinical laboratory (after the laboratory has prepared a slide from the vial and has signed-out the case per standard practice). The laboratory’s specimens were only supplemented from Hologic’s inventory with the rarest cytologic diagnostic categories (AGUS and Cancer), if needed. Slides prepared for the study were from specimens processed within 6 weeks of specimen collection. All study specimens were processed both on a ThinPrep 5000 processor and a ThinPrep 2000 system. The order in which the slides were processed was alternated in blocks of 20. All slides were stained, coverslipped, and read manually following standard laboratory procedures; all slides prepared at a site were reviewed independently by each of the three (3) pairs of cytotechnologists/pathologists. All cytologic diagnoses were determined in accordance with the Bethesda System 2001 criteria for all slides1.

MAN-04008-001 Rev. 001

Page 16 of 27

Table 18: Laboratory ThinPrep 5000 Diagnosis vs. Laboratory ThinPrep 2000 Diagnosis for First Pair of Cytotechnologist/Pathologist (Combined Sites) Lab ThinPrep 5000 Diagnosis UNSAT NILM ASC-US AGUS LSIL ASC-H HSIL Cancer Total

Lab ThinPrep 2000 Diagnosis UNSAT

NILM

31 9 3 1

9 624 23 5 6 6

44

673

ASC-US 32 59 19 7 2 119

AGUS

LSIL

1 2 3 7 1 2

1 4 33

16

ASC-H

HSIL

111 9 12

3 10 1 9 27 16

170

66

2 1 3 14 12 109 3 144

Cancer

3

2 23 28

Total 42 676 132 20 160 63 141 26 1260

Reference Diagnosis by Adjudication Review After all slides in the study were reviewed, all ThinPrep 2000 and ThinPrep 5000 slides were subject to an adjudication review. Adjudication was done at a facility that was not one of the study sites conducting the study. Slides for adjudication were evenly divided between three (3) adjudication panels each consisting of one (1) cytotechnologist and three (3) independent pathologists. Each adjudication panel was blinded to the original review diagnosis for all slides and each independent pathologist within each panel was also blinded to other adjudicator’s diagnoses for all slides. Adjudication consensus agreement was obtained for each slide reviewed. Consensus agreement was achieved when at least two (2) of the three (3) pathologists from a panel rendered an identical diagnosis. In cases where consensus agreement was not achieved the panel members were brought together at a multi-head microscope to review the slides together and come to a consensus diagnosis. For each specimen, an adjudicated diagnosis for the ThinPrep 2000 slide and an adjudicated diagnosis for the ThinPrep 5000 slide were obtained. Table 19: Adjudicated ThinPrep 5000 Diagnosis vs. Adjudicated ThinPrep 2000 Diagnosis (Combined Sites) Adjudicated ThinPrep 5000 Diagnosis UNSAT NILM ASC-US AGUS LSIL ASC-H HSIL Cancer Total

Adjudicated ThinPrep 2000 Diagnosis UNSAT

NILM

14 12

8 696 33 4 12 7

26

760

ASC-US

AGUS

LSIL

ASC-H

39 48 1 20 4 7

8 4 6

9 26

1 2 7

2 1

135 6 9

3 7 8

119

21

185

28

HSIL 4 4 4 10 11 66 2 101

Cancer

3

1 16 20

Total 23 770 122 18 180 37 92 18 1260

For each specimen, the Reference Diagnosis (RD) was considered as the most abnormal diagnosis from the adjudicated diagnoses of the ThinPrep 2000 and ThinPrep 5000 slides. In the study, there were 22 Cancer, 124 HSIL, 39 ASC-H, 202 LSIL, 23 AGUS, 120 ASC-US, and 696 NILM specimens. Thirty-four (34) specimens had UNSAT either with ThinPrep 2000 or with ThinPrep 5000 or with both. Clinical sensitivity and specificity (e.g., with reference to a histological diagnosis) cannot be measured in this study which relied on cytological examination alone. Instead, laboratory positive and negative diagnoses by both methods, ThinPrep 5000 and MAN-04008-001 Rev. 001

Page 17 of 27

ThinPrep 2000, for the specimens with Reference Diagnosis of ASC-US+ (combined ASC-US, AGUS, LSIL, ASC-H, HSIL, and Cancer), LSIL+ (combined LSIL, ASC-H, HSIL, and Cancer), ASC-H+ (combined ASC-H, HSIL, and Cancer) and HSIL+ (combined HSIL and Cancer) were compared.

Clinical Study Results Tables 20 through 23 present the comparison of Laboratory true positive and negative rates for ASC-US+, LSIL+, ASC-H+, and HSIL+. Table 20: Laboratory ThinPrep 5000 Results vs Laboratory ThinPrep 2000 Results for the Specimens with Reference Diagnosis of ASC-US+

In the study, there were 530 specimens with Reference Diagnosis of ASC-US+ (combined ASC-US, AGUS, LSIL, ASC-H, HSIL, and Cancer) and 696 specimens with Reference Diagnosis of NILM. In this table, “Positive” means ASC-US+ or UNSAT, and “Negative” means NILM. All percentages are rounded to the nearest 0.1%. ASC-US+ Lab CT/ Pathologist #1

Positive Percent Agreement ThinPrep 5000 ThinPrep 2000

Negative Percent Agreement

Difference

ThinPrep 5000 ThinPrep 2000

Difference

(95% CI)

(95% CI)

(95% CI)

(95% CI)

(95% CI)

(95% CI)

90.9%

89.4%

1.5%

89.1%

87.9%

1.1%

(482/530) (474/530) (8/530) (620/696) (612/696) (8/696) (88.2% to 93.1%) (86.5% to 91.8%) (-0.7% to 3.8%) (86.5% to 91.2%) (85.3% to 90.1%) (-1.1% to 3.5%)

87.0%

#2

86.6%

0.4%

88.6%

90.7%

-2.0%

(461/530) (459/530) (2/530) (617/696) (631/696) (-14/696) (83.8% to 89.6%) (83.4% to 89.2%) (-2.7% to 3.4%) (86.1% to 90.8%) (88.3% to 92.6%) (-4.4% to 0.3%)

87.5%

#3

88.5%

-0.9%

87.6%

88.1%

-0.4%

(464/530) (469/530) (-5/530) (610/696) (613/696) (-3/696) (84.5% to 90.1%) (85.5% to 90.9%) (-3.7% to 1.8%) (85.0% to 89.9%) (85.5% to 90.3%) (-2.9% to 2.0%)

Table 21: Laboratory ThinPrep 5000 Results vs Laboratory ThinPrep 2000 Results for the Specimens with Reference Diagnosis of LSIL+

In the study, there were 387 specimens with Reference Diagnosis of LSIL+ (combined LSIL, ASC-H, HSIL, and Cancer) and 839 specimens with Reference Diagnosis of (combined NILM, ASC-US, and AGUS). In this table, “Positive” means LSIL+ or UNSAT, and “Negative” means NILM or ASC-US/AGUS. All percentages are rounded to the nearest 0.1%. LSIL+ Lab CT/ Pathologist #1 #2

#3

Positive Percent Agreement ThinPrep 5000 ThinPrep 2000

Negative Percent Agreement

Difference

ThinPrep 5000 ThinPrep 2000

Difference

(95% CI)

(95% CI)

(95% CI)

(95% CI)

(95% CI)

(95% CI)

84.8%

86.8%

-2.1%

90.3%

89.5%

0.8%

(328/387) (336/387) (-8/387) (758/839) (751/839) (7/839) (80.8% to 88.0%) (83.1% to 89.8%) (-5.9% to 1.7%) (88.2% to 92.2%) (87.3% to 91.4%) (-1.1% to 2.8%)

84.0%

83.5%

0.5%

91.7%

91.4%

0.2%

(325/387) (323/387) (2/387) (769/839) (767/839) (2/839) (80.0% to 87.3%) (79.4% to 86.8%) (-3.6% to 4.6%) (89.6% to 93.3%) (89.3% to 93.1%) (-1.7% to 2.2%)

84.0%

87.3%

-3.4%

88.6%

89.4%

-0.8%

(325/387) (338/387) (-13/387) (743/839) (750/839) (-7/839) (80.0% to 87.3%) (83.7% to 90.3%) (-7.4% to 0.6%) (86.2% to 90.5%) (87.1% to 91.3%) (-2.9% to 1.2%) MAN-04008-001 Rev. 001

Page 18 of 27

Table 22: Laboratory ThinPrep 5000 Results vs Laboratory ThinPrep 2000 Results for the Specimens with Reference Diagnosis of ASC-H+

In the study, there were 185 specimens with Reference Diagnosis of ASC-H+ (combined ASC-H, HSIL, and Cancer) and 1,041 specimens with Reference Diagnosis of (combined NILM, ASC-US/AGUS, and LSIL). In this table, “Positive” means ASC-H+ or UNSAT, and “Negative” means NILM, ASC-US/AGUS, or LSIL. All percentages are rounded to the nearest 0.1%. ASC-H+ Lab CT/ Pathologist #1

Positive Percent Agreement ThinPrep 5000 ThinPrep 2000

Negative Percent Agreement

Difference

ThinPrep 5000 ThinPrep 2000

Difference

(95% CI)

(95% CI)

(95% CI)

(95% CI)

(95% CI)

(95% CI)

81.6%

84.3%

-2.7%

90.6%

90.6%

0.0%

(151/185) (156/185) (-5/185) (943/1041) (943/1041) (0/1041) (75.4% to 86.5%) (78.4% to 88.9%) (-8.6% to 3.2%) (88.7% to 92.2%) (88.7% to 92.2%) (-1.6% to 1.6%)

81.6%

#2

81.1%

0.5%

91.7%

91.1%

0.7%

(151/185) (150/185) (1/185) (955/1041) (948/1041) (7/1041) (75.4% to 86.5%) (74.8% to 86.1%) (-6.0% to 7.1%) (89.9% to 93.3%) (89.2% to 92.7%) (-1.0% to 2.3%)

85.4%

#3

84.9%

0.5%

89.8%

90.6%

-0.8%

(158/185) (157/185) (1/185) (935/1041) (943/1041) (-8/1041) (79.6% to 89.8%) (79.0% to 89.3%) (-5.4% to 6.5%) (87.8% to 91.5%) (88.7% to 92.2%) (-2.5% to 0.9%)

Table 23: Laboratory ThinPrep 5000 Results vs Laboratory ThinPrep 2000 Results for the Specimens with Reference Diagnosis of HSIL+

In the study, there were 146 specimens with Reference Diagnosis of HSIL+ (combined HSIL and Cancer) and 1,080 specimens with Reference Diagnosis of (combined NILM, ASC-US/AGUS, LSIL, and ASC-H). In this table, “Positive” means HSIL+ or UNSAT, and “Negative” means NILM, ASC-US/AGUS, LSIL, or ASC-H. All percentages are rounded to the nearest 0.1%. HSIL+ Lab CT/ Pathologist #1 #2

#3

Positive Percent Agreement ThinPrep 5000 ThinPrep 2000

Difference

Negative Percent Agreement ThinPrep 5000 ThinPrep 2000

Difference

(95% CI)

(95% CI)

(95% CI)

(95% CI)

(95% CI)

(95% CI)

77.4%

80.1%

-2.7%

93.2%

93.2%

0.0%

(113/146) (117/146) (-4/146) (1007/1080) (1007/1080) (0/1080) (70.0% to 83.4%) (72.9% to 85.8%) (-9.8% to 4.3%) (91.6% to 94.6%) (91.6% to 94.6%) (-1.4% to 1.4%)

69.9%

74.7%

-4.8%

94.3%

94.7%

-0.5%

(102/146) (109/146) (-7/146) (1018/1080) (1023/1080) (-5/1080) (62.0% to 76.7%) (67.0% to 81.0%) (-11.8% to 2.3%) (92.7% to 95.5%) (93.2% to 95.9%) (-1.9% to 1.0%)

78.1%

82.9%

-4.8%

91.9%

92.3%

-0.5%

(114/146) (121/146) (-7/146) (992/1080) (997/1080) (-5/1080) (70.7% to 84.0%) (75.9% to 88.1%) (-12.6% to 3.1%) (90.1% to 93.3%) (90.6% to 93.8%) (-2.1% to 1.2%)

In the study, there were 2.06% (26/1260) ThinPrep 2000 slides with UNSAT results by Adjudication and 1.83% (23/1260) ThinPrep 5000 slides with UNSAT results by Adjudication.

MAN-04008-001 Rev. 001

Page 19 of 27

Agreement among Laboratory Cytotechnologists/Pathologists The following tables indicate the extent to which the laboratory cytotechnologists / pathologists at a given site agreed amongst themselves on the diagnosis, comparing the ThinPrep 5000 processor to the ThinPrep 2000 system. Tables are provided for ASC-US+ and ASC-H+. In Table 24 for ASC-H+, the number of specimens is shown for which various levels of agreement among the CTs occurred. Either all three CTs rated the slide as positive (ASC-H+), two out of three rated it positive, one out of three, or none of them. Table 24: Laboratory Cytotechnologist/Pathologist Agreement, All Results, ASC-H+ ThinPrep 2000 System Three lab CTs have read the same ThinPrep 2000 slide from a vial Three CTs Three ASC-H+ Two CTs had ASC-H+ One CT had ASC-H+ ThinPrep 5000 Processor Three lab CTs have read the same ThinPrep 5000 slide from a vial

had ASC-H+

& one had <ASC-H

& two had <ASC-H

CTs had <ASC-H

Totals

Three CTs had ASC-H+

111

21

6

0

138

Two CTs had ASC-H+ and one had <ASC-H

32

30

21

7

90

One CT had ASC-H+ and two had <ASC-H

7

9

43

28

87

Three CTs had <ASC-H

2

8

37

898

945

Totals

152

68

107

933

1260

ThinPrep 2000 System Three lab CTs have read the same ThinPrep 2000 slide from a vial

ThinPrep 5000 Processor Three lab CTs have read the same ThinPrep 5000 slide from a vial

ASC-H+

Three or two CTs had ASC-H+

Three or two CTs had <ASC-H

Totals

Three or two CTs had ASC-H+

194

34

242

Three or two CTs had <ASC-H

26

1006

1032

220

1040

1260

Totals

MAN-04008-001 Rev. 001

Page 20 of 27