Logo of amjpatholAmerican Journal of Pathology For AuthorsAmerican Journal of Pathology SubscribeAmerican Journal of Pathology SearchAmerican Journal of Pathology Current IssueAmerican Journal of Pathology About the JournalAmerican Journal of Pathology
Am J Pathol. 2009 Nov; 175(5): 1831–1847.
PMCID: PMC2770709

Detection of Genomic Amplification of the Human Telomerase Gene TERC, a Potential Marker for Triage of Women with HPV-Positive, Abnormal Pap Smears

Abstract

The vast majority of invasive cervical carcinomas harbor additional copies of the chromosome arm 3q, resulting in genomic amplification of the human telomerase gene TERC. Here, we evaluated TERC amplification in routinely collected liquid based cytology (LBC) samples with histologically confirmed diagnoses. A set of 78 LBC samples from a Swedish patient cohort were analyzed with a four-color fluorescence in situ hybridization probe panel that included TERC. Clinical follow-up included additional histological evaluation and Pap smears. Human papillomavirus status was available for all cases. The correlation of cytology, TERC amplification, human papillomavirus typing, and histological diagnosis showed that infection with high-risk human papillomavirus was detected in 64% of the LBC samples with normal histopathology, in 65% of the cervical intraepithelial neoplasia (CIN)1, 95% of the CIN2, 96% of the CIN3 lesions, and all carcinomas. Seven percent of the lesions with normal histopathology were positive for TERC amplification, 24% of the CIN1, 64% of the CIN2, 91% of the CIN3 lesions, and 100% of invasive carcinomas. This demonstrates that detection of genomic amplification of TERC in LBC samples can identify patients with histopathologically confirmed CIN3 or cancer. Indeed, the proportion of TERC-positive cases increases with the severity of dysplasia. Among the markers tested, detection of TERC amplification in cytological samples has the highest combined sensitivity and specificity for discernment of low-grade from high-grade dysplasia and cancer.

Cervical carcinoma is the second most common malignancy among women world-wide.1 The introduction of screening programs based on cytological examination of cervical smears resulted in a significant decrease in both incidence and mortality rates.2 The most common cervical cancer type is squamous cell cervical carcinoma.3 Infections with high-risk human papillomavirus (HR-HPV) are almost invariably found in women with neoplastic disease and progression from low-grade to high-grade dysplasia, and invasive disease is very rare in the absence of HR-HPV.4 However, transient HPV infections that do not result in the development of high-grade dysplasia are common in young, sexually active women.

The causal relationship between HR-HPV infection and cervical cancer has made the detection of the virus an attractive approach to identify women at risk of developing cervical cancer.5 However, it has also become increasingly evident that other factors, in addition to HR-HPV per se, are required for cervical carcinogenesis, since only rarely will women infected with HR-HPV eventually develop cervical cancer. Therefore, biomarkers strongly associated with the propensity of low-grade lesions to progress to high-grade lesions and cancer would be of great value for the individualization of the clinical management of women with abnormal Pap smears.6

The expression of the HPV protein E7 triggers the activation of the cyclin-dependent kinase inhibitor p16INK4a, which has been explored for its diagnostic potential.7 In normal cells, p16 expression is essentially undetectable by immunocytochemistry. It is possible that such staining will not only improve the chances of detecting premalignant cells, but also aid in the distinction between premalignant and reactive atypia.8

In addition to infection with HPV and its consequent expression of p16, the majority of cervical carcinomas carry extra copies of the long arm of chromosome 3 and with it genomic amplification of the RNA component of the human telomerase gene (TERC), which resides on cytoband 3q26.9,10 Therefore, genomic amplification of this gene is likely to play a central role in progression from low-grade dysplasia to high-grade cervical intraepithelial neoplasia (CIN) and invasive cancer. In a previous study, we showed that progression is never observed in the absence of genomic amplification, and, inversely, extra copies of this gene are not present in lesions that spontaneously regress.11

The aim of the present study was to validate these findings among women undergoing testing for cervical cancer following an abnormal Pap smear at population-based screening, and to correlate the amplification status of TERC with the histological evaluation of concordant biopsies. Genomic copy numbers of TERC and the oncogene MYC were determined on liquid based cytology (LBC) slides using fluorescence in situ hybridization (FISH), and the findings were systematically compared with HPV status, p16 protein expression, and the results of cytological screening and histopathological diagnoses.

Materials and Methods

Patients and Tissue Samples

We consecutively enrolled 78 women with any grade of cytological abnormalities detected at a population-based screening. The women were referred for extended testing at the Department of Gynecology at the Karolinska University Hospital Huddinge in Stockholm, Sweden. The extended testing took place during 2005, 2 to 6 months after the initial population-based screening. The mean age of the women was 35.3 years (median 33 years, range 23 to 60 years).

Clinical Evaluation

Enrollment, clinical characterization, and sampling of all women have been previously reported in detail.12 From each woman we collected material for a Pap smear and LBC, and performed colposcopy and punch biopsies for histopathological evaluation. In many cases, a subsequent loop electrosurgical excision procedure (LEEP) histopathology and/or Pap smear follow-up were available.

Cytological and histopathological samples were re-evaluated by an experienced pathologist (A.H.). Samples were diagnosed in a blinded fashion, ie, the respective assessment was disclosed only after both the cytology and histology samples had been evaluated. Cytological categories were defined according to the Bethesda nomenclature,13 including the subgroups within normal limits (WNL), atypical squamous cells—undetermined significance (ASCUS), low-grade squamous intraepithelial lesions (LSIL), atypical squamous cells—cannot exclude high-grade squamous intraepithelial lesions (HSIL) or high-grade changes (ASC-H), and HSIL. According to Swedish recommendations, cases of koilocytosis without signs of dysplasia are reported as WNL. Therefore, the LSIL group contains samples corresponding to cytological CIN grade 1 (CIN1) only. Histopathological diagnosis followed the World Health Organization criteria14 and included the subclasses benign (WNL), CIN1, CIN2, CIN3, and carcinoma. Women with CIN2 or CIN3 were treated by LEEP.

Follow-up data were collected until 2007 (Table 1), including Pap smears in all cases and colposcopy, plus biopsy when indicated. LBC samples were used for all of the HPV analyses, the FISH analyses of TERC, MYC, and HPV, and for the p16 analysis. Additionally, for a subset of cases, 4-μm sections of the biopsies, and/or LEEPs were also hybridized and analyzed. All samples used in the study were obtained with informed consent and permission from the local ethical committee. The flow chart in Figure 1 summarizes the study design.

Figure 1
Flow chart of the study design.
Table 1A
Comprehensive Table of All Analyses and Clinical Data for All 78 Cases Studied
Table 1B
Continued
Table 1C
Continued
Table 1D
Continued
Table 1E
Continued
Table 1F
Continued

HPV Status

HPV status was evaluated by four different assays: 1) a prototype line blot analysis from Roche; 2) a quantitative real-time PCR method specific for 10 high-risk HPV types (HPV 16, 18/45 group, 31 and 33/35/39/52/58/67 group (Quantovir AB, Uppsala, Sweden); 3) detection of E6/E7 mRNA expression of HPV 16, 18, 31, 33, and 45 HR-HPV with real-time nucleic acid sequence-based amplification assay method (PreTect HPV-Proofer; NorChip AS, Oslo, Norway); and 4) HPV-FISH (see also Figure 1). The Roche line blot assay was performed as described earlier.15,16,17 Details of the method and real-time PCR results obtained with the samples analyzed in this study, have been reported.18,19 Results from the E6/E7 mRNA with the samples used in this communication have been described in a previous publication.12 These previously published data were used only as reference points for performance comparison with the previously unpublished line blot and FISH data (Table 1 and Figure 2).12,18,20

Figure 2
Percentage of positive cases within the categories WNL, CIN1, CIN2, CIN3, and cancer for the different tests applied. Data for HPV Quantovir, Norchip mRNA, and p16 have been obtained from previous publications.12,18,20

In addition HPV was detected by FISH (see examples in Figure 3, A–E) as described below.

Figure 3
Examples for HPV-positive (A, B, C, D) and negative (E) nuclei, with and without chromosomal aberrations. The images show the 4,6-diamidino-2-phenylindole staining of the cell nucleus (Blue), the HPV staining (Green) in infected cells, the CEP7 probe ...

Status of p16 Expression

Results from previously published p16 immunocytochemical analyses12 are summarized in Table 1. In short, p16 immunoreactivity was determined on LBC slides and scored as negative (−; less than three stained reactive cells per slide) or positive (+, ++, and +++) depending on the staining intensity.

FISH

Slides for FISH analyses were prepared from samples of LBC specimens according to standard procedures.21 Four-color FISH analysis was performed on each case using a probe set targeting the centromere of chromosome 7 (CEP7), the TERC locus at chromosome band 3q26 (LSI TERC), the MYC locus at 8q24.2 (LSI MYC), and an HPV probe that contained DNA of the types 16, 18, 30, 45, 51, and 58. The probes were provided by Abbott Molecular, Inc. (Des Plaines, IL) through a Cooperative Research and Development Agreement (CRADA #001039). CEP7 was labeled with Spectrum Aqua, the TERC contig with Spectrum Gold, MYC with Spectrum Red, and the HPV probe was biotin-labeled. The HPV probe was previously shown to detect types 16, 18, 26, 31, 33, 35, 39, 45, 52, 53, 56, 58, 59, 66, and 82,22,23 indicating that the DNA of the HPV types contained in the mix cross-reacts with other types because of sequence homology. There were no samples with HPV 30 or HPV51 infection included in the studies cited.22,23

Hybridization, posthybridization washes, and detection of the HPV probe were performed according to the probe manufacturer’s recommendations.22 In brief, the LBC slides were incubated for 2 minutes in 2× standard saline citrate at room temperature followed by pepsin (0.3 mg/ml in 10 mmol/L HCl) at 37°C for 5 to 10 minutes, fixed in 1% neutral-buffered formalin at room temperature for 5 minutes, and then washed in 1× PBS for 5 minutes. Slides were dehydrated in an ethanol series and air-dried. Co-denaturation at 72°C for 2 minutes and overnight hybridization at 37°C were done on a ThermoBrite (Statspin). Slides were then washed in 2× SSC/0.3% Nonidet P-40 for 2 minutes at 48°C and in 2× SSC/0.1% Nonidet P-40 for 1 minute at room temperature. The HPV probe was detected using Alexa Fluor 488 Tyramide Signal Amplification kit number 22 (Invitrogen, Carlsbad, CA) according to the manufacturer’s directions. Finally, antifade solution (VectaShield, Vector Laboratories, Burlingame, CA) containing the nuclear counterstain 4,6-diamidino-2-phenylindole was applied and coverslips were added.

Scoring of FISH Results

Image acquisition and analyses were performed using a Leica DM-RXA fluorescence microscope (Leica, Wetzlar, Germany) equipped with custom optical filters for 4,6-diamidino-2-phenylindole, Spectrum Aqua, Spectrum Gold, Spectrum Red, and Spectrum Green (Chroma Technologies, Rockingham, VT) and a ×40 Plan Apo (NA 1.25) objective. Scoring of FISH results was done without knowledge of cytological or histopathological evaluation or any of the other markers. HPV fluorescence patterns and TERC and MYC signals were enumerated by screening and counting the entire slide visually using the Spectrum Green filter for HPV, with the Spectrum Gold filter for TERC, and with the Spectrum Red filter for MYC. All nuclei on a slide were evaluated. On average, 2320 nuclei per slide with a range of 232 to 4996 nuclei were counted. For the HPV probe the results were classified as negative (0), or positive with episomal (1), episomal and integrated (2), or integrated (3) patterns depending on the appearance of the fluorescence signals (see Figure 3 as an example). CEP7, TERC, and MYC were evaluated by signal counts per nucleus. The results were registered as patterns for the probe panel (CEP7-TERC-MYC-HPV) for each nucleus. For example, a nucleus with two signals for CEP7, three signals for TERC, two signals for MYC, accompanied by an HPV-FISH episomal pattern was recorded as a “2321” pattern. Accordingly, cells with normal signal numbers and HPV-FISH negative result were reported as “2220.” Their number was recorded on a manual counter and used to establish the total cell count by adding it to the number of nuclei with aberrant patterns. Nuclei with more than two TERC signals and/or more than two MYC signals and/or HPV-FISH positivity were imaged using a CoolSnap camera and multifocus imaging with the Leica CW4000-FISH imaging software for each of the optical filters. This software also allows for registering the imaged nuclei in relocation charts with their positions and their signal patterns. This enabled us to revisit the few cases for which ambiguous signal constellations were observed, and to arrive to a consensus count among three observers (P.S., K.H.-H., and T.R.). The complete FISH results comprised of the total cell count per slide and a list of all of the cells deviating from the FISH pattern of normal cells (see column “Signal patterns observed” in Table 1). All of the other columns provided for the FISH data in Table 1 were calculated from these raw data (eg, number of cells with more than two TERC signals excluding or including cells with four CEP7 and four TERC signals, etc). Cells with four signals for each probe, “444” were categorized as “tetraploid,” acknowledging the fact that other chromosomal aneuploidies could exist in these cells.

Statistical Methods

Receiver operating characteristics (ROC) curves were used to establish optimal thresholds and to identify FISH count parameters that best distinguish Thinprep samples with underlying CIN2, CIN3, and cancer histology, from Thinprep samples with underlying WNL and CIN1 histology. ROC curves were generated by plotting the sensitivity for detecting CIN2, CIN3, and cancer versus 1 minus the specificity for detecting WNL and CIN1, obtained at percent cell thresholds ranging from 0% to 10% (0.05% increments). Curves were generated based on the number of tetraploid cells, the number of cells with TERC gain (excluding “tetraploid” cells and cells that have four CEP7 signals and four TERC signals), the number of cells with either tetraploidy or TERC gain (ie, all gains of TERC), the number of cells with MYC gain (excluding “tetraploid” cells and cells that have four CEP7 signals and four MYC signals), and the number of cells with either tetraploidy or MYC gain (ie, all gains of MYC). Curves that come closest to the ideal values of 100% sensitivity and 100% specificity (Figure 4A, top left corner of ROC graph) provide the best combination of sensitivity and specificity (assuming equal importance of each) and optimal thresholds are typically selected from points near the “breaks” in the curves (region closest to top left corner; curve slope near 45°). A better view of the dependence of sensitivity and specificity on threshold can be obtained by plotting the distance from ideal (DFI) versus threshold (Figure 4B). DFI is defined here as the distance from the idealpoint (0,1) on the ROC plot (100% sensitivity, 100% specificity), and is calculated as [(1 − sensitivity)2 + (1 − specificity)2]1/2. DFI is smallest for the best combined sensitivity and specificity (giving equal weight to each) and varies from 0 for thresholds providing 100% sensitivity and 100% specificity, to 1.414 for thresholds providing 0% sensitivity and 0% specificity.

Figure 4
ROC and DFI curves for TERC and MYC gain. A: ROC plot of sensitivity versus 1-specificity at thresholds ranging between 0% and 100% abnormal cells. B: Plot of DFI versus threshold (refer to Materials and Methods for details). The yellow diamonds in both ...

Results

A four-color FISH probe set for detection of CEP7, TERC (3q26), MYC (8q24), and HPV was hybridized to 78 cervical LBC samples of a Swedish patient cohort. Examples of the FISH analyses are shown in Figure 3. For all of the patients a biopsy was taken at a colposcopy performed at the same time as the LBC sampling, and for most of the patients a subsequent LEEP histology and Pap smear follow-up were available. When no visible lesion was observed, a biopsy was taken close to the squamo-columnar junction.

The LBC samples were also analyzed by line blot and Quantovir for HPV infection, detection of E6/E7 mRNA of HPV 16, 18, 31, 33, and 45 high-risk types of HPV expression with the real-time nucleic acid sequence-based amplification assay and for p16 expression by immunocytochemistry. The study design is presented in Figure 1. Table 1 summarizes the FISH, HPV, and p16 data available for each sample together with the cytologic screening results, the histological diagnoses from biopsy and LEEP, and the clinical follow-up data. The results of the primary Pap smears taken at the population-based screenings at the referral clinics for the 78 women enrolled showed ASCUS in 14 cases, 34 smears had LSIL, 22 had HSIL, and six had ASC-H. In two cases we were unable to retrieve the initial diagnosis.

The samples in Table 1 are arranged by the highest grade of histological diagnosis (when biopsy and LEEP histology differed from each other, the higher grade was used).

When sorting according to the highest grade histology, 14 of the 78 samples were WNL, 17 samples were CIN1, 22 samples were CIN2, 23 samples were CIN3, and two samples were squamous cell cervical carcinoma, maintaining a good balance among the CIN groups. The number of TERC-positive cases increased with severity of the cervical lesion. Applying a threshold of more than eight TERC-positive cells before a case was considered TERC positive, 7% (one of 14) of WNL cases, 24% (4/17) of CIN1, 64% (14/22) of CIN2, 91% (21/23) of CIN3, and 100% (2/2) of the carcinomas were positive for genomic amplification of TERC. Figure 212,18,20 shows the percentages of positive cases in lesions assessed as WNL, CIN1, CIN2, CIN3, and cancer comparing the different markers and tests used in this study. The TERC test achieved the highest combined specificity and sensitivity for discernment of low-grade from high-grade lesions and cancer. This can also be appreciated in Table 2, which shows the sensitivity and specificity for the different tests used. When distinguishing between benign and CIN1 lesions on the one side and CIN2, CIN3, and cancers on the other, the TERC test has a sensitivity of 79% and a specificity of 84%. When distinguishing between benign lesions versus CIN3 lesions the sensitivity of the test is 91% and the specificity is 93%. In comparison, as expected from a referral population, the HPV test (line blot) has a very high sensitivity (97% high-risk-HPV, 100% low-risk and high-risk-HPV), but an extremely low specificity (36% HR-HPV, 19% LR and HR-HPV). Expression of p16 and MYC copy number are both associated with a slightly better specificity than TERC, but their sensitivities are around 10% points lower than the sensitivity of the TERC test. Noteworthy, in the distinction of WNL samples versus CIN3 samples, TERC performs better both in specificity and sensitivity, with 7% points and 9% points, respectively, as compared with p16. HPV-FISH alone has a low specificity, but also a relatively low sensitivity, especially in comparison with the line blot HPV results. This can be mostly attributed to the inability of the HPV-FISH test to detect HPV type 31 in our Swedish material, though the probe has been shown to detect type 31 in other cohorts.22 The HPV-FISH test did not detect 20 cases that were positive by HPV line blot, of which seven cases were type 31, two cases were type 68, and two cases were type 73. Types 68 and 73 are not expected to be detected by HPV-FISH since there is only 35% homology between these types and the types in the probe mix. The rest of the cases not detected by HPV-FISH involved types 16, 18, 33, 39, 51, 53, 56, and 66. When using a threshold that has been proposed and applied by Sokolova et al22 (ie, four or more cells positive for both TERC and/or MYC and HPV-FISH), we achieve good specificity but very low sensitivity (see Table 2), which can be partly explained by the suboptimal performance of the HPV-FISH probe in our material. However, even CIN3 cases that were HPV-FISH positive and had chromosomally aberrant nuclei, did not necessarily have four and more double-positive nuclei and were therefore negative when using this criterion (see cases 181, 240, 227, see Table 1).

Table 2
Sensitivities and Specificities of the Different Tests Detecting WNL and CIN1 Lesions Versus CIN2, CIN3, and Cancer Lesions and Detecting WNL Versus CIN3 Lesions from a Sample of Women Referred to Colposcopy and Punch Biopsies Based on a Prior Abnormal ...

Both, the Quantovir and the line blot HPV tests have low specificity (42.9% and 35.5%, respectively), however, the line blot has a higher sensitivity (97% vs. 72.5%). The discrepancy in sensitivity can be partly explained by the fact that the Quantovir test does not cover all HPV types tested for in the line blot assay. Disagreements between types found by line blot and types found by Quantovir are most likely due to binding of the Quantovir primers to HPV types other than the one they were designed for because of sequence homology.

We did not observe a clear increase in the incidence of HR-HPV oncogene mRNA expression with increasing severity of the lesion by using nucleic acid sequence-based amplification assay (NorChip). The HPV mRNA test (NorChip) shows good specificity, but lacks sensitivity (only 50% of CIN2, CIN3, and cancer lesions are detected), which can be mainly attributed to its inability to detect CIN2 lesions. Its sensitivity for detecting CIN3 lesions is better with 68.4%, but still far below the TERC marker (91.3%).

Figure 5 demonstrates that the number of TERC-positive cells is positively correlated with the severity of the lesions. Interestingly, TERC-positive cells often display similar or even identical patterns within the same lesions (Table 1, signal patterns observed), which is a strong indication for expansion of an initial clone following the acquisition of extra copies of 3q. Especially in the more severe lesions, it is evident that the high numbers of TERC-positive cells are often due to the expansion of one or a few specific clones. This is exemplified in, eg, case 227, with aberrant nuclei displaying only one pattern, ie, “364” in 131 of 133 cells with aberrant signal counts.

Figure 5
Average number of TERC-positive cells observed in WNL, CIN1, CIN 2, and CIN3 lesions. Error bars indicate SEM.

We also analyzed sections of the biopsies and the LEEPs of the two CIN3 cases that were not positive for TERC gain in the LBC samples. The LBC samples of both of these cases were scored as LSIL. Case 230 had a CIN3 biopsy and a CIN1 LEEP and case 118 had a CIN1 biopsy and a CIN3 LEEP. The CIN3 biopsy of case 230 showed a small nest of TERC-positive cells while the CIN1 LEEP turned out to be negative, indicating that the biopsy very possibly removed the TERC positive CIN3 cells completely. The CIN1 biopsy of case 118 was TERC-negative, however, the CIN3 LEEP taken 6 months later showed TERC positive cells (see Figure 6, A–D).

Figure 6
Histopathology and corresponding FISH images (TERC marker in yellow) of the biopsies and LEEPs of cases 118, 230, 250, and 90. A: Biopsy 230 (CIN3) shows several nuclei with more than three TERC signals. However, the corresponding LBC sample was below ...

Classification Algorithms

Figure 4A shows ROC curves for discriminating CIN2, CIN3, and cancer, from lesions diagnosed as WNL and CIN1. Based on the curve lying closest to the top left corner of the graph, TERC gain including tetraploidy provides the best method for distinguishing CIN2, CIN3, and cancer from WNL and CIN1. Figure 4B shows DFI curves, and as expected from the ROC curves, TERC gain including tetraploidy shows the lowest DFI. The minimum of this curve indicates the threshold with the best combined specificity and sensitivity, which in our dataset are 8.6 TERC-positive cells. Based on these analyses, we selected the optimal threshold to determine TERC positivity as a case with nine and more cells containing more than two signals of TERC including “tetraploid” cells and cells with four signals for TERC and four signals for CEP7. Using this threshold the TERC test reached a specificity of 83.9% and a sensitivity of 78.7% to distinguish benign and CIN1 lesions from CIN2, CIN3, and cancer. The specificity and sensitivity with which WNL could be distinguished from CIN3 was 92.9% and 91.3%, respectively.

Discussion

We have previously demonstrated that a FISH assay for the visualization of genomic amplification of the human telomerase RNA gene TERC on chromosome band 3q26.3 can distinguish Pap smears with high-grade cytology from those with normal or low-grade cytology with high sensitivity and specificity.24 In a second retrospective study, we have expanded on these results by showing that detection of TERC gain can help stratifying women with abnormal Pap smears according to the risk of progression from low-grade lesions to higher-grade lesions and cancer. In addition, applying the TERC test to routinely collected Pap smears increased the sensitivity of cytological screening.11

In the current study, we were able to determine the presence of TERC-positive cells in LBC samples collected on the same day as the diagnostic biopsy. In addition, we were able to correlate the TERC-positive LBC results with follow-up Pap, LEEP, and/or biopsy samples. By using all available histopathological data from baseline and follow-up, we were able to enhance the diagnostic accuracy of our case definition by defining case status based on the highest grade histology results observed. For example, combining histology results of the biopsy and LEEP has the advantage that possible sampling errors will be reduced, especially as the LEEP excision often results in larger, more representative material. However, having the LEEP excision taken at a later time point can, in certain cases, mean that the lesion advanced to a higher degree than it was actually at the LBC sampling. In contrast, the diagnostic biopsy might have presented a cure with complete removal of dysplastic cells.

In this study, we also included a MYC probe, a CEP7 probe that served as a control for the ploidy of the cells, and a probe for the detection of HPV genomes in the FISH analysis, allowing the evaluation of test performance using TERC alone and in combination with these markers.

According to these evaluation parameters, TERC positivity was observed in 7% (one of 14) WNL cases, 24% (4/17) CIN1, 64% (14/22) CIN2, 91% (21/23) CIN3 lesions, and 100% (2/2) carcinomas. Interestingly, the two TERC-negative cases with CIN3 histopathology (cases 118 and 230, Table 1), had cytological screening results and histological diagnoses of the biopsy and LEEP that differed from each other. The LBC cytology of both cases was assessed as LSIL. Case 118 presented with a CIN1 biopsy and a CIN3 LEEP 6 months later. TERC analysis on the histological samples showed negativity for the biopsy and positivity for the LEEP, indicating that the TERC gain was not present at the time of the LBC sampling, but developed later during the progression from CIN1 to CIN3 (see Figure 6C, D).

In contrast, case 230 displayed a CIN3 biopsy, while the subsequent LEEP was diagnosed as a CIN1 lesion. TERC analysis of these samples showed positivity for the CIN3 biopsy and negativity for the CIN1 LEEP. This case might present an example for the scenario that the diagnostic biopsy can be curative with complete removal of all dysplastic cells. This is further supported by the observation that the area of TERC-positive cells within the biopsy was very small and contained (see Figure 6A, B). The small size of the lesion is most likely also the reason that only very few TERC-positive cells were detected in the LBC sample (ie, four, which is below the threshold of more than eight positive cells). Interestingly, the original referral Pap smear was judged as ASC-H, indicating that this first cytology sampling possibly yielded a more representative material than the sampling for the LBC. Examples like this one underscore the importance of representative sampling, but also highlight one of the recognized limitations of cervical screening.

The hybridization results of the tissue sections of these two CIN3 cases show that so far we have not observed a cervical lesion with a histological diagnosis greater than CIN2 that does not show TERC positivity. There are, however, TERC-negative lesions diagnosed histologically as CIN2. This corroborates that CIN3 can be considered the true endpoint lesion while CIN2 lesions rather present one link in the continuous progression chain that can be further stratified by TERC positivity as being associated with significant risk of becoming a CIN3 lesion with further progression to invasive disease.

The diagnostic potential of genomic amplification of TERC as a powerful biomarker is further emphasized by the following cases: Case 201 had LBC LSIL cytology with underlying CIN3 histology and was clearly detected with 51 TERC-positive cells in the LBC sample. Case 204 cytology was recorded as ASC-H, with 25 TERC-positive cells, but the biopsy result only recorded CIN1. However, the LEEP histology 1 month later showed a CIN3 lesion. Case 156 scored normal for cytology, however, we found 20 TERC-positive cells. The biopsy was classified as CIN3, and a subsequent LEEP excision 2.5 months later revealed a microinvasive carcinoma. The cytological samples of case 112 were initially assessed as LSIL, which was confirmed in the biopsy, but the LBC sample was positive for TERC amplification. LEEP excision as much as 17 months later revealed a CIN3. Arguably, it is not possible to establish whether a sampling error during the biopsy precluded the detection or, if that is not the case, how soon after the biopsy sampling the lesion progressed to CIN3. However, this case shows the full potential of the TERC test, as this patient could have been treated according to the true severity of the lesion before further progression.

It is also interesting to mention case 182, which presented with a HSIL LBC cytology that was strongly positive for TERC and MYC. However, biopsy, subsequent LEEP and follow-up Pap smears were all benign. Looking at the main pattern of aberration, the TERC gene presents with four copies on a tetraploid background, while the MYC gene shows a relative gain with five copies. We have observed in a previous study that cases that were positive for TERC as a consequence of tetraploidy can either progress or regress.11 Case 182 seems to be an example for a case that regressed. Interestingly, the MYC gene presented a relative gain within the tetraploid nucleus. If this case truly regressed by itself, this would indicate that relative gains for the MYC gene are less specific for cervical cancer progression than relative gains for the TERC gene, which would be consistent with the fact that gain of chromosome 8q and with it genomic amplification of the MYC oncogene is rarely observed in cervical tumorigenesis.

The LBC samples were also analyzed for alternative disease markers including high-risk and low-risk HPV types by HPV line blot, HPV viral load by the Quantovir assay, HPV mRNA expression by nucleic acid sequence-based amplification assay, and p16 expression using immunocytochemistry.

When comparing the different assays used in this study, the line blot HPV test had the best sensitivity, detecting all CIN2 and CIN3 lesions when including both low-risk and high-risk HPV infections. These data corroborate that HPV testing is an excellent first screen to identify women with a higher risk of developing cervical cancer. However, as known before and observed also in this study with a specificity of the line blot test of 19.4% (low-risk and high-risk HPV) or 35.5% (high-risk HPV only), HPV testing has only limited power to stratify low-grade from high-grade disease and can therefore not be used to efficiently triage patients further. Therefore additional markers that can triage patients to avoid overtreatment and not to overlook relevant lesions are needed. Potential triage markers tested in this study are TERC, MYC, p16, and HPV mRNA expression.

With a specificity of 87.1%, which is slightly better than the TERC test (83.9%) and the HPV mRNA expression test (78.7%), p16 and MYC performed well. However, their sensitivities with 68.1% and 66.0%, respectively, are more than 10% points lower than the sensitivity of the TERC test (78.7%). In the stratification between benign lesions and CIN3 lesions (rather than benign, CIN1, and CIN2 versus CIN3 and cancer), the TERC test sensitivity and specificity increased to 91.3% and 92.9%, respectively. The sensitivity of p16 and MYC for that stratification increased only to 82.6%, which is 10% lower than the TERC test. HPV mRNA showed a very low sensitivity of 50% for the distinction between benign and CIN1 vs. CIN2, CIN3, and cancer, and a slightly better sensitivity of 68.4% for the distinction between benign versus CIN3 lesions. These data show that the TERC test has the best combined specificity and sensitivity in the panel of markers tested in this study.

p16 is a cyclin-dependent kinase inhibitor involved in cell cycle control. Current data suggest that inactivation of pRB through HPV E7 results in enhanced expression of p16, which might therefore represent a specific and sensitive biomarker for cells with active expression of HPV oncogenes.25,26 Several investigators applied p16 as an immunochemical assay to Pap smears, LBC slides or histological samples of benign, dysplastic, and cancerous lesions. Klaes et al7 reported that all benign lesions and all CIN1 lesions associated with low-risk HPV had no detectable expression of p16, while all CIN2 and CIN3 lesions and all CIN1 lesions with high-risk HPV showed marked overexpression. Other papers show a less stratified and less sensitive outcome, often observing a few positive lesions in the normal and CIN1 entities, and detecting around 70% to 80% of the high-grade lesions.27,28,29 These latter observations are more consistent with our data that show few positive lesions in the WNL and CIN1 category, and 55% and 83% positivity for CIN2 and CIN3 lesions, respectively. Interestingly, p16 was not positive in one of the two cancers in our material. P16-negative cervical cancers have been previously observed by other investigators, and Masoudi et al30 speculate that the decrease or negativity in some of the carcinomas could be due to a loss of expression during progression, perhaps as a result of a promoter methylation. Interestingly, they found that p16 negativity in squamous cell carcinoma was an adverse prognostic marker.

Recently, two other studies applying the TERC FISH probe as a diagnostic marker on cervical Thinprep samples were published.22,31 The studies are similar in that they reduce the amount of technician time required to evaluate the specimens. Both studies are in agreement with our data that TERC positivity can detect high-grade lesions with high sensitivity. However, both studies observe a rather low specificity of the test, in contrast to our findings. On careful evaluation, the possible explanation for this discrepancy seems to be mostly due to different evaluation methods. Caraway et al31 applied an automated evaluation approach based on a commercial imaging system (MetaSystems, Waltham, MA) and on an algorithm similar to the one used for automated evaluation of UroVysion samples.32 The enumeration strategy included a pre-selection of cells of interest based on nuclear size and shape. They also drastically reduced the number of evaluated nuclei to only the 25 most atypical nuclei per slide. In this approach, the threshold was put to two and more cells with more than two TERC signals and all cytological HSIL lesions were detected (100% sensitivity), which equals our results. However, due to the low threshold, also 70% of the LSIL lesions were scored as positive. Another drawback of this method is that TERC aberrations do not always happen in the largest and most irregular nuclei, but do present in morphologically unsuspicious nuclei.11

Sokolova et al22 used a combined evaluation approach of TERC and/or MYC positivity within HPV-FISH−positive cells with a threshold of four and more double-positive cells. By this method, the specimen is scanned for HPV-stained cells and only those cells are enumerated. In the absence of HPV staining a representative number of cells are enumerated. The immediate drawback that the authors experienced with this approach is that not all TERC-positive cells show HPV-FISH positivity, partly because of sensitivity issues of the HPV-FISH test, and partly because of the possibility that cells that have acquired chromosomal aneuploidies can become negative for HPV. The authors mention that two of their high-grade lesions show a substantial number of TERC-positive cells, but cannot be scored positively, due to the fact that these cells are not positive for HPV-FISH, which compromises the sensitivity of the TERC test to a certain degree. Notably, also the specificity of their study is much lower than in our study. Again, it is very likely that the combined evaluation of HPV and TERC and MYC data within the same cell is responsible for this decrease in specificity. Most LSIL lesions are HPV positive, which is the reason why HPV testing has a very low specificity. If one combines HPV positivity with a relatively low threshold for TERC-positive cells (four cells and more), the stratification power of the TERC test will be diminished. If we were to apply a threshold of four or more TERC and/or MYC-positive cells (irrespective of HPV-FISH) to our dataset, we would achieve very similar specificity values as Sokolova et al, 22 ie, 53% (compared with 56%) for CIN1 lesions while increasing the sensitivity to 96% (compared with 86%). This could be interpreted as evidence that HPV-FISH does not contribute specificity to the TERC data while reducing sensitivity, although this approach may permit a more rapid means of specimen evaluation that would be beneficial in a routine screening application.

In summary, our data show that the TERC test is the most sensitive and specific test for the stratification of benign and low-grade cervical lesions from high-grade lesions and cancer when compared with other tests used for stratification. It also becomes clearly evident that the evaluation method used for genetic FISH markers like the TERC test needs to be adapted to the clinical samples analyzed. As screening entire LBC slides for FISH markers is tedious and time-consuming, alternative approaches, such as the two approaches referred to above, should be tested. While it would be desirable to combine FISH markers with tests more easily amenable to a high throughput format, our study shows that such combinations often diluted the performance of detecting genomic amplification of TERC as a powerful biomarker for disease progression. These results are supported by a large body of literature demonstrating the central role of this specific aneuploidy for cervical tumorigenesis.9,10,11,22,24,31,33,34,35,36,37,38,39,40,41

HPV testing is increasingly used as a screening method, either alone or in conjunction with Pap smear testing and its high sensitivity and its ease and reproducibility of analysis makes it a very attractive first screen.6 However, especially in younger women, HPV infections are often transient and overtreatment of these women is a serious problem. Our study showed that TERC is a powerful marker to discern histologically confirmed low-grade from high-grade cervical disease. In the future, one could envision a screening strategy in which HPV-positive Pap smears are triaged with the help of the TERC marker that would contribute to a more evidence-based individualized clinical management. This strategy is supported through biological causality, because both HPV and gain of 3q act in concert in disease initiation and progression.

Acknowledgments

We thank Buddy Chen for the preparation of figures and editorial assistance and Joseph Cheng for information technology and microscope support. We appreciate valuable discussions with Dr. Elizabeth Unger, Centers for Disease Control and Prevention. We also thank Carmen Flores-Staino and Ingrid Norman for their technical assistance.

Footnotes

Address reprint requests to Thomas Ried, Genetics Branch, Center for Cancer Research, National Cancer Institute, 50 South Drive, Building 50, Rm 1306, Bethesda, MD 20892. E-mail: vog.hin.liam@tdeir.

Supported by The Swedish Cancer Foundation (CAN 2007/1044), the Swedish Research Council, the Medical Research Council, the Cancer Society in Stockholm, the Göran Gustafsson Foundation, the Stockholm County Council, and the Swedish Labor Market Insurance and the Intramural Research Program of the NIH, National Cancer Institute. P. Gravitt and P. Sowjanya were funded via the NIH SPORE P50 CA98252.

S.A. and P.S. contributed equally to this work.

CRADA agreement with Abbott Molecular, Inc., Des Plaines, IL (CRADA #001039).

Patent and Patent Application: United States Patent 5,919,624 issued Jul.6, 1999. Ried T, Auer G, Schroeck E, Heselmeyer K, Macville M: Methods for detecting cervical cancer.

International Application Number PCT (Patent Cooperation Treaty)/US 2006/006116. Ried T, Heselmeyer-Haddad K, Steinberg W, Auer G, Andersson S, Larsson C: Methods for detecting progression of low grade cervical dysplasia and for detecting adenocarcinomas.

References

  • World Health Organization Geneva, Switzerland: WHO; Comprehensive cervical cancer controla guide to essential practice. 2006:p 3.
  • Koss LG. The Papanicolaou test for cervical cancer detection. A triumph and a tragedy. JAMA. 1989;261:737–743. [PubMed]
  • Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, Snijders PJ, Peto J, Meijer CJ, Munoz N. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol. 1999;189:12–19. [PubMed]
  • Khan MJ, Castle PE, Lorincz AT, Wacholder S, Sherman M, Scott DR, Rush BB, Glass AG, Schiffman M. The elevated 10-year risk of cervical precancer and cancer in women with human papillomavirus (HPV) type 16 or 18 and the possible utility of type-specific HPV testing in clinical practice. J Natl Cancer Inst. 2005;97:1072–1079. [PubMed]
  • Cuzick J, Arbyn M, Sankaranarayanan R, Tsu V, Ronco G, Mayrand MH, Dillner J, Meijer CJ. Overview of human papillomavirus-based and other novel options for cervical cancer screening in developed and developing countries. Vaccine. 2008;26 Suppl 10:K29–K41. [PubMed]
  • Gravitt PE, Coutlee F, Iftner T, Sellors JW, Quint WG, Wheeler CM. New technologies in cervical cancer screening. Vaccine. 2008;26 Suppl 10:K42–K52. [PubMed]
  • Klaes R, Friedrich T, Spitkovsky D, Ridder R, Rudy W, Petry U, Dallenbach-Hellweg G, Schmidt D, von Knebel Doeberitz M. Overexpression of p16(INK4A) as a specific marker for dysplastic and neoplastic epithelial cells of the cervix uteri. Int J Cancer. 2001;92:276–284. [PubMed]
  • Norman I, Brismar S, Zhu J, Gaberi V, Hagmar B, Hjerpe A, Andersson S. p16(INK4a) immunocytochemistry in liquid-based cervical cytology: is it feasible for clinical use? Int J Oncol. 2007;31:1339–1343. [PubMed]
  • Heselmeyer K, Schrock E, du Manoir S, Blegen H, Shah K, Steinbeck R, Auer G, Ried T. Gain of chromosome 3q defines the transition from severe dysplasia to invasive carcinoma of the uterine cervix. Proc Natl Acad Sci USA. 1996;93:479–484. [PMC free article] [PubMed]
  • Heselmeyer K, Macville M, Schrock E, Blegen H, Hellstrom AC, Shah K, Auer G, Ried T. Advanced-stage cervical carcinomas are defined by a recurrent pattern of chromosomal aberrations revealing high genetic instability and a consistent gain of chromosome arm 3q. Genes Chromosomes Cancer. 1997;19:233–240. [PubMed]
  • Heselmeyer-Haddad K, Sommerfeld K, White NM, Chaudhri N, Morrison LE, Palanisamy N, Wang ZY, Auer G, Steinberg W, Ried T. Genomic amplification of the human telomerase gene (TERC) in pap smears predicts the development of cervical cancer. Am J Pathol. 2005;166:1229–1238. [PMC free article] [PubMed]
  • Andersson S, Hansson B, Norman I, Gaberi V, Mints M, Hjerpe A, Karlsen F, Johansson B. Expression of E6/E7 mRNA from ‘high risk’ human papillomavirus in relation to CIN grade, viral load and p16INK4a. Int J Oncol. 2006;29:705–711. [PubMed]
  • Solomon D, Davey D, Kurman R, Moriarty A, O'Connor D, Prey M, Raab S, Sherman M, Wilbur D, Wright T, Jr, Young N. The 2001 Bethesda System: terminology for reporting results of cervical cytology. JAMA. 2002;287:2114–2119. [PubMed]
  • Richart RM. A modified terminology for cervical intraepithelial neoplasia. Obstet Gynecol. 1990;75:131–133. [PubMed]
  • Peyton CL, Gravitt PE, Hunt WC, Hundley RS, Zhao M, Apple RJ, Wheeler CM. Determinants of genital human papillomavirus detection in a US population. J Infect Dis. 2001;183:1554–1564. [PubMed]
  • Gravitt PE, Peyton CL, Alessi TQ, Wheeler CM, Coutlee F, Hildesheim A, Schiffman MH, Scott DR, Apple RJ. Improved amplification of genital human papillomaviruses. J Clin Microbiol. 2000;38:357–361. [PMC free article] [PubMed]
  • Gravitt PE, Peyton CL, Apple RJ, Wheeler CM. Genotyping of 27 human papillomavirus types by using L1 consensus PCR products by a single-hybridization, reverse line blot detection method. J Clin Microbiol. 1998;36:3020–3027. [PMC free article] [PubMed]
  • Andersson S, Safari H, Mints M, Lewensohn-Fuchs I, Gyllensten U, Johansson B. Type distribution, viral load and integration status of high-risk human papillomaviruses in pre-stages of cervical cancer (CIN). Br J Cancer. 2005;92:2195–2200. [PMC free article] [PubMed]
  • Moberg M, Gustavsson I, Gyllensten U. Real-time PCR-based system for simultaneous quantification of human papillomavirus types associated with high risk of cervical cancer. J Clin Microbiol. 2003;41:3221–3228. [PMC free article] [PubMed]
  • Andersson S, Wangsa D, Flores-Staino C, Safari H, Mints M, Hjerpe A, Hagmar B, Johansson B. Expression of p16(INK4a) in relation to histopathology and viral load of ‘high-risk’ HPV types in cervical neoplastic lesions. Eur J Cancer. 2006;42:2815–2820. [PubMed]
  • Linder J, Zahniser D. ThinPrep Papanicolaou testing to reduce false-negative cervical cytology. Arch Pathol Lab Med. 1998;122:139–144. [PubMed]
  • Sokolova I, Algeciras-Schimnich A, Song M, Sitailo S, Policht F, Kipp BR, Voss JS, Halling KC, Ruth A, King W, Underwood D, Brainard J, Morrison L. Chromosomal biomarkers for detection of human papillomavirus associated genomic instability in epithelial cells of cervical cytology specimens. J Mol Diagn. 2007;9:604–611. [PMC free article] [PubMed]
  • Algeciras-Schimnich A, Policht F, Sitailo S, Song M, Morrison L, Sokolova I. Evaluation of quantity and staining pattern of human papillomavirus (HPV)-infected epithelial cells in thin-layer cervical specimens using optimized HPV-CARD assay. Cancer. 2007;111:330–338. [PubMed]
  • Heselmeyer-Haddad K, Janz V, Castle PE, Chaudhri N, White N, Wilber K, Morrison LE, Auer G, Burroughs FH, Sherman ME, Ried T. Detection of genomic amplification of the human telomerase gene (TERC) in cytologic specimens as a genetic test for the diagnosis of cervical dysplasia. Am J Pathol. 2003;163:1405–1416. [PMC free article] [PubMed]
  • Keating JT, Ince T, Crum CP. Surrogate biomarkers of HPV infection in cervical neoplasia screening and diagnosis. Adv Anat Pathol. 2001;8:83–92. [PubMed]
  • Wang JL, Zheng BY, Li XD, Angstrom T, Lindstrom MS, Wallin KL. Predictive significance of the alterations of p16INK4A, p14ARF, p53, and proliferating cell nuclear antigen expression in the progression of cervical cancer. Clin Cancer Res. 2004;10:2407–2414. [PubMed]
  • Meyer JL, Hanlon DW, Andersen BT, Rasmussen OF, Bisgaard K. Evaluation of p16INK4a expression in ThinPrep cervical specimens with the CINtec p16INK4a assay: correlation with biopsy follow-up results. Cancer. 2007;111:83–92. [PubMed]
  • Duncan L, Jacob S, Hubbard E. Evaluation of p16INK4a as a diagnostic tool in the triage of Pap smears demonstrating atypical squamous cells of undetermined significance. Cancer. 2008;114:34–48. [PubMed]
  • Zhang Q, Kuhn L, Denny LA, De Souza M, Taylor S, Wright TC., Jr Impact of utilizing p16INK4A immunohistochemistry on estimated performance of three cervical cancer screening tests. Int J Cancer. 2007;120:351–356. [PubMed]
  • Masoudi H, Van Niekerk DJ, Gilks CB, Cheang M, Bilek K, Fischer U, Ehlen T, Miller D, Horn LC. Loss of p16 INK4 expression in invasive squamous cell carcinoma of the uterine cervix is an adverse prognostic marker. Histopathology. 2006;49:542–545. [PubMed]
  • Caraway NP, Khanna A, Dawlett M, Guo M, Guo N, Lin E, Katz RL. Gain of the 3q26 region in cervicovaginal liquid-based pap preparations is associated with squamous intraepithelial lesions and squamous cell carcinoma. Gynecol Oncol. 2008;110:37–42. [PubMed]
  • Sokolova IA, Halling KC, Jenkins RB, Burkhardt HM, Meyer RG, Seelig SA, King W. The development of a multitarget, multicolor fluorescence in situ hybridization assay for the detection of urothelial carcinoma in urine. J Mol Diagn. 2000;2:116–123. [PMC free article] [PubMed]
  • Kirchhoff M, Rose H, Petersen BL, Maahr J, Gerdes T, Lundsteen C, Bryndorf T, Kryger-Baggesen N, Christensen L, Engelholm SA, Philip J. Comparative genomic hybridization reveals a recurrent pattern of chromosomal aberrations in severe dysplasia/carcinoma in situ of the cervix and in advanced-stage cervical carcinoma. Genes Chromosomes Cancer. 1999;24:144–150. [PubMed]
  • Hopman AH, Theelen W, Hommelberg PP, Kamps MA, Herrington CS, Morrison LE, Speel EJ, Smedts F, Ramaekers FC. Genomic integration of oncogenic HPV and gain of the human telomerase gene TERC at 3q26 are strongly associated events in the progression of uterine cervical dysplasia to invasive cancer. J Pathol. 2006;210:412–419. [PubMed]
  • Rao PH, Arias-Pulido H, Lu XY, Harris CP, Vargas H, Zhang FF, Narayan G, Schneider A, Terry MB, Murty VV. Chromosomal amplifications, 3q gain and deletions of 2q33–q37 are the frequent genetic changes in cervical carcinoma. BMC Cancer. 2004;4:5. [PMC free article] [PubMed]
  • Matthews CP, Shera KA, McDougall JK. Genomic changes and HPV type in cervical carcinoma. Proc Soc Exp Biol Med. 2000;223:316–321. [PubMed]
  • Dellas A, Torhorst J, Jiang F, Proffitt J, Schultheiss E, Holzgreve W, Sauter G, Mihatsch MJ, Moch H. Prognostic value of genomic alterations in invasive cervical squamous cell carcinoma of clinical stage IB detected by comparative genomic hybridization. Cancer Res. 1999;59:3475–3479. [PubMed]
  • Allen DG, White DJ, Hutchins AM, Scurry JP, Tabrizi SN, Garland SM, Armes JE. Progressive genetic aberrations detected by comparative genomic hybridization in squamous cell cervical cancer. Br J Cancer. 2000;83:1659–1663. [PMC free article] [PubMed]
  • Umayahara K, Numa F, Suehiro Y, Sakata A, Nawata S, Ogata H, Suminami Y, Sakamoto M, Sasaki K, Kato H. Comparative genomic hybridization detects genetic alterations during early stages of cervical cancer progression. Genes Chromosomes Cancer. 2002;33:98–102. [PubMed]
  • Hidalgo A, Monroy A, Arana RM, Taja L, Vazquez G, Salcedo M. Chromosomal imbalances in four new uterine cervix carcinoma derived cell lines. BMC Cancer. 2003;3:8. [PMC free article] [PubMed]
  • Narayan G, Pulido HA, Koul S, Lu XY, Harris CP, Yeh YA, Vargas H, Posso H, Terry MB, Gissmann L, Schneider A, Mansukhani M, Rao PH, Murty VV. Genetic analysis identifies putative tumor suppressor sites at 2q35–q36.1 and 2q36.3-q37.1 involved in cervical cancer progression. Oncogene. 2003;22:3489–3499. [PubMed]

Articles from The American Journal of Pathology are provided here courtesy of American Society for Investigative Pathology
PubReader format: click here to try

Formats:

Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Cited in Books
    Cited in Books
    NCBI Bookshelf books that cite the current articles.
  • Gene
    Gene
    Gene records that cite the current articles. Citations in Gene are added manually by NCBI or imported from outside public resources.
  • GEO Profiles
    GEO Profiles
    Gene Expression Omnibus (GEO) Profiles of molecular abundance data. The current articles are references on the Gene record associated with the GEO profile.
  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles
  • Substance
    Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...