SEARCH

SEARCH BY CITATION

Keywords:

  • p16INK4a;
  • CINtec;
  • ThinPrep;
  • liquid-based cytology;
  • cervical biomarker

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

BACKGROUND.

The aim of this study was to examine p16INK4a protein expression in ThinPrep (Cytyc Corporation, Marlborough, Mass) cervical specimens by using the CINtec p16INK4a Cytology Kit (Dako, Glostrup, Denmark). The ability of this assay to accurately identify underlying high-grade lesions was assessed by using follow-up biopsies and comparing these results with Hybrid Capture 2 (Digene, Gaithersburg, Md) high-risk HPV (hc2) results.

METHODS.

Three hundred ninety-eight residual ThinPrep samples were collected, and histological follow-up data were retrieved for abnormal cytology specimens. After preparation of a Papanicolaou-stained slide, a second slide was processed in preparation for p16INK4a immunostaining. High-risk human papillomavirus testing (hc2) was also performed.

RESULTS.

Of the 163 cytologically abnormal samples, 6-month biopsy follow-up data were available for 45% of the specimens. At initial blinded evaluation, 21 of the 26 cases with cervical intraepithelial neoplasia (CIN) II/III follow-up were positive for p16INK4a, yielding an overall diagnostic sensitivity of 81%; 29 of the 47 cases diagnosed as CIN I or less were p16INK4a negative, yielding a diagnostic specificity of 62%. In comparison, the hc2 test results indicated a diagnostic sensitivity of 100% with a diagnostic specificity of 15%. After review of selected cases with CIN II/III follow-up, 25 of 26 slides were deemed to be positive for p16INK4a, increasing the diagnostic sensitivity to 96%.

CONCLUSIONS.

The CINtec p16INK4a Cytology Kit, in combination with ThinPrep cervical samples, allowed clear evaluation of p16INK4a protein overexpression. Diagnostic specificity of the CINtec p16INK4a assay was significantly improved relative to hc2. To increase p16INK4a immunostaining in abnormal cells, a modified kit version with improved staining performance has been developed and is currently being evaluated. Cancer (Cancer Cytopathol) 2007. © 2007 American Cancer Society.

Carcinoma of the uterine cervix is the second most common neoplasm among women worldwide, and it is the fifth leading cause of all cancer-related deaths.1 Recent estimates indicate that approximately 500,000 new cases of cervical cancer are diagnosed worldwide annually. Since its introduction in the 1940s, the conventional Papanicolaou test has dramatically decreased the incidence and mortality rate associated with cervical cancer by identifying and classifying cellular and morphological changes associated with progression to cancer.2, 3 The implementation of this cytological test and screening paradigm for detection and treatment of precancerous lesions has, therefore, proven to be invaluable in reducing the incidence of this cancer. Furthermore, recent developments in sample collection, processing, and image-directed slide review4–7 have contributed to a 28% decline in cervical cancer within the United States during the past decade.8

Despite technological improvements, a number of studies have reported high interobserver variability and Papanicolaou test discordance with histological follow-up, particularly within the atypical squamous cells of undetermined significance (ASCUS) and low-grade squamous intraepithelial lesions (LSIL) diagnostic categories.9 Integrating the use of objective markers to determine the presence and severity of dysplastic cells may help overcome the limitations of cervical cytology. For example, high-risk human papillomavirus (HPV) has been shown to be present in 99% of all cervical cancers,10 and the concept that persistent viral infection is required for progression to cervical neoplasia is well accepted.11 However, while HPV DNA testing can provide an objective measurement, high-risk HPV testing cannot accurately discriminate between patients whose squamous intraepithelial lesions will persist or progress to invasive carcinoma and those whose lesions will regress spontaneously. It was reported in The ASCUS/LSIL Triage Study For Cervical Cancer (ALTS) trial that 83% of women with an LSIL Papanicolaou result were positive for high-risk HPV, a level too high to provide clinical utility in a patient triage strategy, as the associated risk of histologic cervical intraepithelial neoplasia (CIN) II/III is no more than 25%.12 Although a triage strategy that incorporated HPV detection within the ASCUS population was proven to be more sensitive for detecting underlying high-grade disease, decreased specificity was a primary concern.13

It is known that the human papillomavirus contributes to neoplastic progression predominantly through the action of 2 viral oncoproteins, E6 and E7, which interact with various host regulatory proteins to influence the function or expression levels of host gene products, eventually leading to the disruption of the cell cycle.14 It has been previously demonstrated that the E6 oncoprotein interacts with the p53 tumor suppressor protein resulting in the ubiquitination and premature degradation of p53, whereas E7 binds specifically to the retinoblastoma protein, pRb.15 As a consequence of E7 interaction with pRb, the E2F transcription factor is released, and progression through the G1/S phase transition checkpoint ensues.15, 16 p16INK4a is a cyclin-dependent kinase inhibitor that negatively regulates cell proliferation by inhibiting hyperphosphorylation of pRb via the cdk4/6 complex.17, 18 Overexpression of the p16INK4a protein has been well documented in cervical cancer19–21 and is a consequence of pRb targeted inactivation from E7.

It has been proposed that p16INK4a is a useful biomarker for the identification of cervical intraepithelial lesions because it is a measure of active HPV gene expression, rather than viral presence only.19, 20, 22, 23 However, its value in identifying those clinically relevant cases of HPV infection that have a more significant risk of underlying neoplasia still needs to be established. By using the CINtec p16INK4a Cytology Kit (Dako, Glostrup, Denmark) to evaluate ThinPrep (Cytyc Corporation, Marlborough, Mass) cervical specimens, this study was designed to further investigate p16INK4a as a potential marker to complement cytological diagnosis.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Specimen Preparation and Processing

ThinPrep vials (Cytyc Corporation, Marlborough, Mass) from Institutional Review Board-approved residual cervical specimens were obtained from Brigham and Women's Hospital (Boston, Mass) and the Medical University of South Carolina (Charleston, SC). All specimens were deidentified to maintain patient anonymity, however, unique vial identifiers were assigned in order to obtain patient follow-up information. Sample volume was also assessed to determine whether sufficient material would likely be available for the preparation of 2 ThinPrep slides as well as HPV testing. Two consecutive slides were prepared according to instructions detailed in the ThinPrep 2000 operator's manual, and all sample slides were prepared within 6 weeks of the specimen collection date.

One slide was stained with the ThinPrep Papanicolaou stain and served as the internal cytology reference slide. All stained reference slides were reviewed by a cytotechnologist. If disagreement in diagnosis occurred between the cytotechnologist and the clinical site, the slide was reviewed by a second cytotechnologist. The final specimen diagnosis reflected consensus by at least 2 of the 3 reviewers (clinical site and 2 cytotechnologists). In cases where consensus could not be reached, which included approximately 2% of the specimens, the sample was removed from the study. The second slide (p16INK4a slide) was processed in preparation for immunocytochemistry using the CINtec p16INK4a Cytology Kit for Manual Staining; code number K5340 (Dako, Glostrup, Denmark). Test slides were incubated in a 95% ethanol-fixative bath for at least 20 minutes, air dried for 5 seconds, dipped 10 times in CellFyx solution (Cytyc, Marlborough, Mass) and air dried overnight. Samples were coded to mask cytological diagnosis, and specimens were randomized during slide processing to eliminate potential bias during the various staining procedures performed throughout the course of this study.

In addition to the reference and p16INK4a slides, several additional control slides were prepared and incorporated into each immunostaining procedure. Two slides were prepared from a cell pool containing HeLa cells that had been mixed with normal buccal cells in roughly equivalent cell numbers and transferred to PreservCyt fixative (Cytyc, Marlborough, Mass) for at least 15 minutes before slide processing. The first slide was incubated with the p16INK4a antibody and was expected to yield positive (HeLa) and negative (buccal cells) chromogenic signals. The second slide, which served as the negative reagent control slide, was incubated with a primary antibody specific for a nonmammalian protein (glucose oxidase) from Aspergillus niger. In addition, a third control slide was prepared from a cell pool containing residual clinical samples with a cytological result of high-grade squamous intraepithelial lesion (HSIL). All 3 control slides were prepared simultaneously with the p16INK4a test slides. According to the protocol, test-slide results were excluded from analysis if unexpected staining results were observed with the control slides during an experiment.

Immunocytochemistry

Immunocytochemistry was performed according to the protocol provided with the CINtec p16INK4a Cytology Kit (K5340, Dako, Glostrup, Denmark) for detection of overexpression of the p16INK4a protein. The kit contains a collection of ready-to-use reagents optimized for immunocytochemical staining of p16INK4a in ThinPrep cervical cytology specimens. Immunostaining was completed within 3 weeks of slide preparation. Briefly, slides were immersed in 50% ethanol for 5 minutes, rinsed in distilled water, and placed in Wash Buffer to rehydrate cells. For the antigen retrieval process, slides were placed in a preheated 10 mM citrate solution (95°C to 99°C) and maintained at this temperature for 40 minutes followed by a 20-minute cool-down period at room temperature. After a brief rinse in distilled water, slides were immersed in Wash Buffer in preparation for immunostaining. Specimens were blocked with a peroxidase-blocking reagent for 5 minutes followed by a 30-minute incubation with p16INK4a antibody (clone E6H4). Slides were then incubated for 30 minutes with the visualization reagent followed by two 5-minute incubations with DAB substrate to allow visualization of the positive cells. Slides were subsequently counterstained with hematoxylin, dehydrated, mounted with permanent mounting medium, and coverslipped. All steps, with the exception of the epitope retrieval, were performed at room temperature. Immunoreactivity was visualized by light microscopy.

Hybrid Capture 2 High-Risk HPV Testing

Digene's Hybrid Capture 2 (hc2) (Digene, Gaithersburg, Md) high-risk HPV DNA Test was performed by using 2 mL of ThinPrep sample. All additional assay steps were performed according to the manufacturer's protocol for hc2 HR-HPV testing using ThinPrep cervical samples.

Evaluation of p16INK4a Immunoreactivity

A Data Collection Form (DCF) was used to capture information from each p16INK4a immunostained slide. This information included the staining intensity (range, 0–3; A score of 0 indicated complete lack of immunostaining, and scores of 1, 2, and 3 indicated increased immunostaining darkness.), the number and type of cells that exhibited staining (ie, dysplastic, atypical, metaplastic, endocervical), and p16INK4a signal localization (nuclear, cytoplasmic, or both). All p16INK4a-positive slides were subsequently reviewed by an external pathologist who made a final determination as to whether the slide was positive or negative for p16INK4a immunostaining. A p16INK4a-positive result was defined by immunostaining of at least 1 morphologically identified abnormal (dysplastic or atypical) cell. Staining intensity scores of 1, 2, or 3 and signal localization were not factors in determining a p16INK4a-positive result. A p16INK4a-negative result was defined as no cell staining or staining of morphologically identified metaplastic or endocervical cells. Staining of bacteria, seen in a small number of cases, was attributed to nonspecific binding and was not considered positive staining. Sensitivity and specificity of p16INK4a, as well as high-risk HPV detection, were determined relative to the cytology result and biopsy follow-up. The ability of the p16INK4a and hc2 tests to accurately identify underlying lesions of CIN II or CIN III was assessed for specimens that had follow-up data at 6 months. All sample information, as well as study results, were maintained in an Access (Microsoft, Redmond, Wash) database.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

A summary of the study design is presented in Figure 1 and described in more detail within Materials and Methods. A total of 398 residual ThinPrep samples were included in this study representing the following diagnostic categories: 235 negative for intraepithelial lesions or malignancy (NILM) (59%), 49 atypical squamous cells of undetermined significance (ASC-US) (12%), 6 atypical squamous cells–cannot exclude high-grade (ASC-H) (2%), 57 low-grade squamous intraepithelial lesions (LSIL) (14%), and 51 high-grade squamous intraepithelial lesions (HSIL) (13%). Because of the small number of specimens within the ASC-H category, these samples were combined with the HSIL diagnostic category. Table 1 summarizes the overall results for p16INK4a immunoreactivity and hc2 high-risk (HR) HPV status within each cytological category. Four HSIL samples were not included in this analysis because of inadequate sample volume for HPV testing; however, of these specimens, 3 were positive and 1 was negative for p16INK4a immunoreactivity. As described in Materials and Methods, p16INK4a positivity was defined as immunostaining of at least 1 morphologically abnormal dysplastic or atypical cell per ThinPrep slide. In 95% of the samples, p16INK4a immunoreactivity was observed within both the nuclear and cytoplasmic subcellular regions.

thumbnail image

Figure 1. Work flow and study design. Two slides were prepared from residual ThinPrep cervical specimens. The first slide was stained with the ThinPrep Papanicolaou stain and used as a diagnostic reference slide. A second slide was processed for immunostaining by using the CINtec p16INK4a Cytology Kit. The staining intensity, signal localization, number, and cell type (based upon morphology) of p16INK4a-positive cells were recorded for each slide. A selected number of cases were subsequently reviewed after study completion. Six-month follow-up data was obtained for cytologically abnormal cases and used to evaluate diagnostic sensitivity and specificity of p16INK4a and high-risk HPV (hc2).

Download figure to PowerPoint

Table 1. Summary of p16INK4a Immunocytochemistry and Hybrid Capture 2 (hc2) High-Risk HPV Test Results for ThinPrep Specimens Segregated by Cytological Diagnosis
TestCytological Diagnosis
p16INK4aHR-HPV (hc2)NILM No. (%)ASC-US No. (%)LSIL No. (%)HSIL No. (%)Total No. (%)
  • p16INKa positivity is defined as all dysplastic/atypical cell staining. HPV indicates human papilloma virus; NILM, negative for intraepithelial lesions or malignancy; ASC-US, atypical squamous cells of undetermined significance; LSIL, low-grade squamous intraepithelial lesions; HSIL, high-grade squamous intraepithelial lesions.

  • *

    53 of 57 HSIL cases in this study contained adequate sample volumes for HPV testing; of the 4 specimens not tested, 3 were positive, and 1 was negative for p16INKa.

++4 (2)6 (12)23 (40)42 (79)75 (19)
+17 (7)3 (6)1 (2)1 (2)22 (5)
+15 (6)20 (41)27 (47)8 (15)70 (18)
199 (85)20 (41)6 (11)2 (4)227 (58)
Total 235495753*394

p16INK4a positivity was assessed within each cytological category and was shown to increase from 9% (21 of 235) in NILM samples to 81% (43 of 53) in HSIL specimens. By comparison, positivity for high-risk HPV using the hc2 assay was 8% (19 of 235) for NILM specimens and 94% (50 of 53) in the HSIL category (Table 1 and Fig. 2). Within the NILM cytological category, p16INK4a and hc2 were comparable in the overall percentage of positive cases. Within the ASC-US category, positivity of HR-HPV (53%) was greater than p16INK4a (18%), a trend that continued for the LSIL and HSIL cytological categories with hc2 positivity of 88% and 94%, respectively, compared with 42% and 81% for p16INK4a. High-grade cells typically exhibited the highest levels of p16INK4a expression; approximately 80% of all p16INK4a-positive samples with a staining intensity of 3 were HSIL specimens. Figure 3 provides representative images of various specimens and staining intensities along with the numerical scores, which ranged from 0 to 3.

thumbnail image

Figure 2. Comparison of p16INK4a and Hybrid Capture 2 (hc2) high-risk HPV positivity according to cytological category.

Download figure to PowerPoint

thumbnail image

Figure 3. Representative images of p16INK4a immunoreactivity and staining intensity. (A) Staining intensity score of 0 in a specimen with a cytological result of NILM. (B) Light nuclear and cytoplasmic staining (intensity score 1) was observed in a specimen with a diagnosis of ASC-US. Moderate (intensity score 2) and strong (intensity score 3) staining were observed in (C) an LSIL specimen and (D) an HSIL specimen, respectively. Slides were counterstained with hematoxylin and imaged at a magnification of ×400.

Download figure to PowerPoint

To better evaluate the potential of p16INK4a as a cervical marker for preneoplasia, it is important to assess findings from cytology relative to follow-up histology data. Knowing the percentage of embedded CIN II/III within each cytological category would be one way to assess clinical performance. Available histological follow-up data was, therefore, obtained at each clinical site approximately 6 months following accrual of all study samples. As expected, follow-up data was not available for samples in the NILM cytological category. Of the 163 cytologically abnormal samples within the study, 6-month follow-up data was available for 73 (45%) specimens. More specifically, follow-up histology data were obtained for 15 of 49 (31%) of the cases defined cytologically as ASC-US, 28 of 57 (49%) LSIL, and 30 of 53 (57%) HSIL. Histologically, 47 of 73 (64%) study samples had a follow-up diagnosis of ≤CIN I; the remaining 26 of 73 (36%) had a follow-up diagnosis of either CIN II or CIN III.

Analysis of the histological follow-up data revealed that 21 of 26 specimens that were diagnosed as CIN II/III were p16INK4a positive, yielding an overall diagnostic sensitivity of 81% (95% confidence interval [CI], 62.1–91.5), whereas 29 of 47 cases diagnosed as CIN I or less were p16INK4a negative, yielding a diagnostic specificity of 62% (95% CI, 47.4–74.2) (Fig. 4A). The positive and negative predictive values (PPV and NPV) were 54% and 85%, respectively. Evaluation of the biopsy follow-up results revealed that 25 of 25 patients diagnosed as CIN II/III were HR-HPV positive, yielding a diagnostic sensitivity of 100% (95% CI, 86.7–100), whereas 7 of 47 samples with a follow-up diagnosis of ≤CIN I were HR-HPV negative, yielding an overall diagnostic specificity of 15% (95% CI, 7.4–27.7) (Fig. 4A). The positive and negative predictive values for the hc2 test were 38% and 100%, respectively. Figure 4B shows the distribution of follow-up data for ASC-US, LSIL, and HSIL cytological categories. The single CIN II/III biopsy specimen in the ASC-US category was negative for p16INK4a, yielding a diagnostic sensitivity of 0% (95% CI, 0.0–79.3). Cases diagnosed as ≤CIN I revealed that 13 of 14 were negative, yielding a diagnostic specificity of 93% (95% CI, 68.5–98.7). Follow-up biopsies for the LSIL category revealed that 4 of 5 CIN II/III cases were positive for p16INK4a, yielding a diagnostic sensitivity of 80% (95% CI, 37.6–96.4), and 12 of 23 cases with a follow-up biopsy of ≤CIN I were negative for p16INK4a, yielding a diagnostic specificity of 52% (95% CI, 33.0–70.8). Within the HSIL category, 17 of 20 of the CIN II/III biopsy specimens were positive for p16INK4a, yielding a diagnostic sensitivity of 85% (95% CI, 64.0–94.8). Cases diagnosed as ≤CIN I revealed that 4 of 10 were negative, yielding a diagnostic specificity of 40% (95% CI, 16.8–68.7). By comparison, diagnostic sensitivity of the hc2 test for the detection of CIN II/III was 100% in each of the ASC-US (95% CI, 20.7–100), LSIL (95% CI, 56.6–100) and HSIL (95% CI, 83.2–100) cytological categories. Specificity of the hc2 test for samples segregated by cytology diagnosis decreased from 21% in ASC-US (95% CI, 7.6–47.6) to 17% in LSIL (95% CI 7.0–37.1) and 0% in HSIL (95% CI, 0.0–27.8).

thumbnail image

Figure 4. (A) p16INK4a and Hybrid Capture 2 (hc2) high-risk HPV positivity compared with histological follow-up results for combined cytological categories. (B) Individual cytological categories (ASCUS, LSIL, and HSIL) compared with histological follow-up for p16INK4a positivity and Hybrid Capture 2 (hc2) high-risk HPV, respectively. *One specimen was excluded from hc2 analysis because of an insufficient volume for analysis.

Download figure to PowerPoint

Based upon the above results, a total of 24 cases were selected for additional slide review including 10 NILM, 2 ASC-US, 2 LSIL, and 10 HSIL. These cases included specimens with HSIL cytology that were p16INK4a negative, and NILM cases which were p16INK4a positive and HPV negative. A panel review was performed that included 2 cytotechnologists and 2 pathologists, and in 13 of 24 (54%) of the cases, the interpretation of p16INK4a immunostaining was modified. Discrepant results between primary and secondary review were attributed to weak p16INK4a-staining intensity, missed positive cells due to the small number of p16INK4a-immunostained cells per slide, or light hematoxylin counterstain that contributed to the misclassification of morphologically normal and abnormal cells. After slide review and reanalysis of the data, p16INK4a diagnostic sensitivity was shown to increase from 81% to 96%. The 2 p16INK4a-negative cases with a follow-up biopsy result of CIN II/III in the ASCUS and LSIL categories (Fig. 4B) were, upon review, found to contain 1–2 lightly stained dysplastic cells that were missed at initial evaluation.

In an effort to enhance the signal intensity of p16INK4a immunoreactivity in addition to the identification of abnormal cells, several modifications were made to the CINtec p16INK4a Cytology Kit, including optimization of the antigen retrieval step and incorporation of the ThinPrep nuclear stain. Results from internal and external studies confirmed that staining efficiency was significantly increased, giving more intense p16INK4a immunostaining and enhanced observation of nuclear morphological detail (data not shown).

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Although implementation of the Papanicolaou test has efficiently reduced morbidity and mortality of cervical cancer, there is still an opportunity for improvement. During the last decade, technological advances in sample collection and processing have been introduced and implemented in cervical screening programs. These initiatives have contributed to a significant decline in cervical cancer. Morphological interpretation of the cytological test, however, still has limitations. Stoler et al have demonstrated that misinterpretation and interobserver discrepancies are common, specifically within the ASC-US and LSIL cytology categories.9 Implementation of a new biomarker, such as p16INK4a, could assist in interpretation by facilitating identification of cytological intraepithelial lesions.

Several articles have proposed that overexpression of p16INK4a may be used as a marker for persistent high-risk HPV infection and detection of high-grade squamous intraepithelial lesions because of its relation to active HPV gene expression rather than viral presence only.19, 20, 22, 23 Furthermore, it has been suggested that p16INK4a can be useful in resolving ambiguous cases involving a differential diagnosis of cervical neoplasia.24 There are, however, still only limited data related to performance of p16INK4a in a clinical setting. In this study, which included approximately 400 specimens, p16INK4a overexpression was compared with cytological Papanicolaou results and histological diagnosis from the same patient.

Comparison of p16INK4a and HPV positivity from this study with the Q-Probe study by the College of American Pathologists provides a basis for evaluating p16INK4a as a potential marker to complement cytological diagnosis. Within the Q-Probe study, 22,439 cytology cases with associated histological follow-up information were reviewed to determine the correlation between cervical cytology diagnosis and the corresponding results obtained from surgical biopsies.25 This study demonstrated that 13% of ASCUS, 18% of LSIL, and 77% of HSIL cytology specimens were ≥CIN II by histology. Figure 5 compares results for p16INK4a and HR-HPV obtained in this study with the Q-Probe data in each of the ASC-US, LSIL, and HSIL cytology categories. Although p16INK4a positivity (ASC-US and LSIL cytology categories) in our study is higher than the percentage of embedded CIN II/III determined from the Q-Probe study, it is approximately half of the hc2 test, indicating that p16INK4a may be a better marker than HPV for identification of cases with embedded CIN II/III. As suggested in the ALTS trial, LSIL triage using hc2 is not effective because of the high positivity of the HPV test.12, 13 However, a p16INK4a positivity of 42% demonstrates the potential of using this marker for LSIL triage.

thumbnail image

Figure 5. p16INK4a and Hybrid Capture 2 (hc2) high-risk HPV positivity in ASCUS, LSIL, and HSIL cytological categories from this study is compared with the percentage of embedded CIN II/III determined from the Q-Probe study,25 which correlated diagnosis of cervical cytology specimens with biopsy follow-up data.

Download figure to PowerPoint

Numerous prospective studies using p16INK4a immunostaining on liquid-based cytology specimens24, 26–31 have observed good agreement between p16INK4a-positive staining and grade of intraepithelial lesion. Pientong et al31 reported p16INK4a staining in 57% of ASC-US, 33% of LSIL, and 93% of HSIL cases. The positivity in the ASC-US category is somewhat higher compared with our results. A study by Guo et al.29 investigated p16INK4a immunoreactivity and performed hc2 testing on 210 cytological specimens with Papanicolaou results of LSIL and HSIL. This study demonstrated that p16INK4a and hc2 positivity was 58% and 85%, respectively, for LSIL cases, and 97% and 86%, respectively, for HSIL. These numbers are in good concordance with our results. Subtle differences observed between these studies may be due to the prevalence of high-risk HPV in different populations or differences in the p16INK4a-staining protocol and/or slide interpretation. Results from the study by Guo et al29 were also compared with histological follow-up. In the LSIL group, the PPV for p16INK4a (33%) was higher than the PPV for hc2 (21%) for underlying lesions of ≥CIN II/III. These numbers are in agreement with results from our study in the LSIL diagnostic category, where PPVs of the p16INK4a and the hc2 tests were 27% and 21%, respectively. These results also emphasize that p16INK4a immunostaining for patients with LSIL cytology could be useful in a triage strategy.

Previous studies have reported that no detectable p16INK4a immunostaining was found in normal cervical squamous epithelial cells24–30 but occasionally observed in metaplastic and normal endocervical cells. However, this staining would not necessarily influence results when these cell types are identified by morphology.24 The experience from our study suggests that, in some cases, morphological assessment of p16INK4a-stained cells that are hematoxylin counterstained may be difficult. This was particularly the case for differentiation between p16INK4a-stained atypical and benign immature metaplastic cells. It is important to note that morphological assessment of p16INK4a-stained cells counterstained with hematoxylin is different from observations in Papanicolaou-stained slides, thus training and experience in assessing immunostained slides from cervical samples are very helpful. This was even apparent to our own evaluators as they gained experience in the interpretation of p16 INK4a immunostaining during the study.

The follow-up data that was available from 73 patients showed that 21 of 26 specimens diagnosed as CIN II/III were positive for p16INK4a, resulting in a diagnostic sensitivity of 81%. This is low compared with the diagnostic sensitivity reported in previous studies. For example, Nieh et al30 reported a diagnostic sensitivity for p16INK4a staining at 95% in a study that included 47 ASC-US and 19 ASC-H specimens. Likewise, Guo et al29 reported a diagnostic sensitivity of 98% for CIN II/III. Based upon the discrepancy between our data and the literature, selected cases were reviewed. During this review, 4 of the 5 cases discrepant between p16INK4a and CIN II/III histology were changed to concordant, resulting in a diagnostic sensitivity of 96% after reassessment. All 4 cases were called p16INK4a negative during the initial evaluation because of staining of only a small number of cells, and in 2 of these cases, the staining was weak. Upon review of the ThinPrep Papanicolaou-stained reference slides, all of the cases contained a sufficient number of dysplastic or atypical cells per slide (>10), suggesting that a low number of diagnostic cells per specimen was not a contributing factor in the initial classification of p16INK4a negativity for these particular cases. Reassessment strongly indicated the need to increase the staining intensity as well as the percentage of dysplastic cells stained. This has led to the development of a modified version of the kit that yields improved staining performance. Initial data obtained by using this modified kit indicates an increase in diagnostic sensitivity (data not shown).

In conclusion, this study demonstrates that use of the CINtec p16INK4a Cytology Kit, in combination with ThinPrep cervical samples, allows p16INK4a overexpression to be evaluated. p16INK4a expression provides a strong advantage by possessing a considerably higher diagnostic specificity compared with hc2. The diagnostic sensitivity, based upon initial slide evaluation, was relatively low; however, after reassessment, the sensitivity increased to 96%, a level similarly reported in other studies. Our results reinforce the potential value of using p16INK4a overexpression as a diagnostic tool in the management of cervical cancer screening. Further evaluation of p16INK4a overexpression for use in either ASCUS or LSIL triage of cytological specimens may be warranted because of the high diagnostic specificity of p16INK4a.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

We thank Berit Hoelund, Odense University Hospital, Denmark for her evaluation of the p16INK4a immunostained slides, Dr. Ed Cibas, Brigham and Women's Hospital for access to ThinPrep cervical cytology specimens, and Dr. Lawrence Burg, Cytyc Corporation, for critical review of the manuscript. We also acknowledge Dr. Marcus Trunk, MTM Laboratories, Germany for his participation in the review panel

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  • 1
    World Health Organization. Fact Sheet No. 297, 2006.
  • 2
    Papanicolaou GN. Science. 1942; 95: 438439.
  • 3
    Ostor AG. Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynecol Pathol. 1993; 12: 186192.
  • 4
    Grace A,McBrearty P,Troost S,Thornhill M,Kay E,Leader M. Comparative study: conventional cervical and ThinPrep Pap tests in a routine clinical setting. Cytopathol. 2002; 13: 200205.
  • 5
    Weintraub J,Morabia A. Efficacy of a liquid-based thin layer method for cervical cancer screening in a population with a low incidence of cervical cancer. Diagn Cytopathol. 2000; 22: 5259.
  • 6
    Dziura B,Quinn S,Richard K. Performance of an imaging system vs. manual screening in the detection of squamous intraepithelial lesions of the uterine cervix. Acta Cytol. 2006; 50: 309311.
  • 7
    Biscotti CV,Dawson AE,Dziura B, et al. Assisted primary screening using the automated ThinPrep Imaging System. Am J Clin Pathol. 2005; 123: 281287.
  • 8
    RiesLAG,EisnerMP,KosaryCL,HankeyBF,MillerBA,CleggL,MariottoA,FayMP,FeuerEJ,EdwardsBK (eds.) SEER Cancer Statistics Review, 1975–2000. National Cancer Institute, Bethesda, MD, 2003.
  • 9
    Stoler MH,Schiffmann M. Interobserver reproducibility of cervical cytologic and histologic interpretations: realistic estimates from the ASCUS-LSIL Triage Study. JAMA. 2001; 285: 15001505.
  • 10
    Walboomers JM,Jacobs MV,Manos MM, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol. 1999; 189: 1219.
  • 11
    Ho GY,Bierman R,Beardsley L,Chang CJ,Burk RD. Natural history of cervicovaginal papillomavirus infection in young women. N Eng J Med. 1998; 338: 423428.
  • 12
    The ASCUS-LSIL Triage Study Group. A randomized trial on the management of low-grade squamous intraepithelial lesion cytology interpretations. Am J Obstet Gynecol. 2003; 188: 13931400.
  • 13
    The ASCUS-LSIL Triage Study Group. Results of a randomized trial on the management of cytology interpretations of atypical squamous cells of undetermined significance. Am J Obstet Gynecol. 2003; 188: 13831392.
  • 14
    Galloway DA,McDougall JK. The disruption of cell cycle checkpoints by papillomavirus oncoproteins contributes to anogenital neoplasia. Semin Cancer Biol. 1996; 7: 309315.
  • 15
    Khleif SN,DeGregori J,Yee CL, et al. Inhibition of cyclin D-CDK4/CDK6 activity is associated with an E2F-mediated induction of cyclin kinase inhibitor activity. Proc Natl Acad Sci U S A. 1996; 93: 43504354.
  • 16
    Chellappal S,Kraus VB,Kroger B, et al. Adenovirus E1A, simian virus 40 tumor antigen, and human papillomavirus E7 protein share the capacity to disrupt the interaction between transcription factor E2F and the retinoblastoma gene product. Proc Nat Acad Sci U S A. 1992; 89: 45494553.
  • 17
    Serrano M,Hannon GJ,Beach D. A new regulatory motif in cell-cycle control causing specific inhibition of cyclin D/CDK4. Nature. 1993; 366: 704707.
  • 18
    Li Y,Nichols MA,Shay JW,Xiong Y. Transcriptional repression of the D-type cyclin-dependent kinase inhibitor p16 by the retinoblastoma susceptibility gene product pRb. Cancer Res. 1994: 54: 60786082.
  • 19
    Klaes R,Friedrich T,Spitkovsky D, et al. Overexpression of p16(INK4a) as a specific marker for dysplastic and neoplastic epithelial cells of the cervix uteri. Int J Cancer. 2001; 92: 276284.
  • 20
    Sano T,Oyama T,Kashiwabara K,Fukuda T,Nakajima T. Expression status of p16 protein is associated with human papillomavirus oncogenic potential in cervical and genital lesions. Am J Pathol. 1998; 153: 17411748.
  • 21
    Keating JT,Cviko A,Riethdorf S, et al. Ki-67, cyclin E, and p16INK4a are complementary surrogate biomarkers for human papilloma virus-related cervical neoplasia. Am J Surg Pathol. 2001; 25: 884891.
  • 22
    Agoff N,Lin P,Morihara J,Mao C,Kiviat N,Koutsky L. p16INK4a expression correlates with degree of cervical neoplasia: A comparison with Ki-67 expression and detection of high-risk HPV types. Mod Pathol. 2003; 16: 665673.
  • 23
    Stanley MA. Prognostic factors and new therapeutic approaches to cervical cancer. Virus Res. 2002; 89: 241248.
  • 24
    Saqi A,Pasha T,McGrath CM,Yu GH,Gupta P. Overexpression of p16INK4a in liquid-based specimens (SurePath) as marker of cervical dysplasia and neoplasia. Diagn Cytopathol. 2002; 27: 365370.
  • 25
    Jones BA,Novis D. Cervical biopsy–cytology correlation. A College of American Pathologists Q-Probes study of 22 439 correlations in 348 laboratories. Arch Pathol Lab Med. 1996; 120: 523553.
  • 26
    Nieh S,Chen SF,Chu TY, et al. Is p16(INK4A) expression more useful than human papillomavirus test to determine the outcome of atypical squamous cells of undetermined significance-categorized Pap smear? A comparative analysis using abnormal cervical smears with follow-up biopsies. Gynecol Oncol. 2005; 97: 3540.
  • 27
    Bose S,Evans H,Lantzy L,Scharre K,Youssef E. p16(INK4A) is a surrogate biomarker for a subset of human papilloma virus-associated dysplasias of the uterine cervix as determined on the Pap smear. Diagn Cytopathol. 2005; 32: 2124.
  • 28
    Nieh S,Chen SF,Chu TY,Lai HC,Fu E. Expression of p16INK4A in Pap smears containing atypical glandular cells from the uterine cervix. Acta Cytol. 2004; 48: 173180.
  • 29
    Guo M,Hu L,Baliga M,He Z,Hughson MD. The predictive value of p16INK4a and hybrid capture 2 human papillomavirus testing for high-grade cervical intraepithelial neoplasia. Am J Clin Pathol. 2004; 122: 894901.
  • 30
    Nieh S,Chen SF,Chu TY,Lai HC,Fu E. Expression of p16 INK4A in Papanicolaou smears containing atypical squamous cells of undetermined significance from the uterine cervix. Gynecol Oncol. 2003; 91: 201208.
  • 31
    Pientong C,Ekalaksananan T,Swadpanich U, et al. Immunocytochemical detection of p16INK4a protein in scraped cervical cells. Acta Cytol. 2003; 47: 616623.