Proffered papers 14.00–15.00 Tuesday 16 September 2003 8 Ampullary,pancreatic and biliary duct exfoliative brush cytology prepared by the thinprep method: performance and morphology

Background ERCP‐directed brush cytology is used to sample lesions of the pancreatic and biliary ducts and the ampulla of Vater. With conventional preparations, the sensitivity and specificity range from 44% to 63% and 80% to 98%, respectively, and increased N : C ratio, nuclear molding and loss of honeycombing are reliable features of malignancy. The performance and morphology of specimens prepared by ThinPrep, a liquid‐based cytology technique is mostly unknown. Methods The laboratory information system was searched for all cases prepared by ThinPrep. Patient disease classification of benign or malignant was determined by linkage with the provincial cancer registry and was the gold standard against which sensitivity, specificity, positive (PPV) and negative predictive values (NPV) were calculated. True positives and negatives were reviewed to identify predictive cytomorphologic features. Results Between 1996 and 2001, there were 149 ThinPrep specimens; 55 (37%) were reported as positive for malignancy and 94 (63%) as negative. Disease was classified as malignant in 86 (58%) patients and benign in 63 (42%). There were 42 false negative, 11 false positive, 52 true negative, and 44 true positive cytology results. Sensitivity was 51.2% (CI; 40.2 : 62.0), specificity 82.5% (CI; 70.5 : 90.6), and PPV and NPV 80.0% (CI; 66.6 : 89) and 55.3% (CI; 44.7 : 65.5), respectively. Cell groups with crowded, enlarged, irregular nuclei and nuclear features of vesicular chromatin and large, multiple irregular nucleoli correlated with malignant disease, while monolayered sheets of uniform columnar cells, regular nuclei and a finely granular chromatin correlated with benign disease. Conclusions The performance of ThinPrep brushings from this anatomic site equals conventional preparations. Cytomorphologic features of malignancy are more frequent and pronounced with ThinPrep.     

Predictive power of cytomorphological features in equivocal (C3, C4) breast FNAC

Definitive immediate diagnosis in breast fine needle aspiration cytology (FNAC) remains the aim for cytopathologists. We reviewed 72 consecutive equivocal (C3 and C4) aspirates with respect to 16 cytomorphological criteria. We assessed the power of each criterion at predicting either a malignant [positive predictive value (PPV)] or a benign [negative predictive value (NPV)] diagnosis by correlation with follow-up histology. Blind review led to 34% of cases being correctly definitively diagnosed. Eccentrically placed epithelial cell nuclei (PPV = 88%, sensitivity = 67%, specificity = 87%) and coarse nuclear chromatin (PPV = 81%, sensitivity = 72%, specificity = 83%) are the features that are most useful in predicting malignancy in this selected series. The presence of myoepithelial cells within epithelial groups is not a good indicator of a benign diagnosis (NPV = 24%, sensitivity = 80%, specificity = 53%).     

ERCP-directed brush cytology prepared by the Thinprep method: test performance and morphology of 149 cases.

Conventionally prepared endobiliary brushings are moderately (42%) sensitive and highly (98%) specific in detecting malignancy. The performance and morphological features of brushings prepared by Thinprep, a liquid-based method are mostly unknown. All brushings were retrieved from the laboratory files. Disease was classified as benign or malignant by linkage with the provincial cancer registry and sensitivity, specificity, positive (PPV) and negative predictive values (NPV) calculated. True positives and negatives were reviewed and predictive morphological features analysed by regression tree analysis. Out of 149 brushings, 55 (37%) were positive and 94 (63%) negative. Malignancy was identified in 86 (58%) and benign disease in 63 (42%) of the cases. The sensitivity was 51%, specificity 83%, PPV 80% and NPV 55%. Absolute discriminants of positive and negative brushings were not found, but nuclear cytoplasmic ratio was a useful feature. The performance of Thinprep-prepared brushings from this anatomical site was comparable with conventional preparations.     

Diagnostic value of fine needle aspirates processed by ThinPrep® for the assessment of axillary lymph node status in patients with invasive carcinoma of the breast

X. Jing, E. Wey and C. W. Michael Diagnostic value of fine needle aspirates processed by ThinPrep® for the assessment of axillary lymph node status in patients with invasive carcinoma of the breast Objective: To evaluate the utility of ThinPrep® as an optional specimen processing method for the detection of axillary lymph node metastasis of invasive breast carcinoma. Methods: A computer SNOMED search from the file at our institution between January 2003 and August 2011 retrieved a total of 209 fine needle aspiration (FNA) specimens of axillary lymph nodes prepared by ThinPrep and followed by axillary lymph node biopsy and/or dissection. Original cytological diagnoses and corresponding histological diagnoses were documented. Using the histological diagnoses as the gold standard, the diagnostic parameters including sensitivity, specificity, positive (PPV) and negative predictive values (NPV) and diagnostic accuracy were calculated. Both cytology and histology slides from cyto‐histologically discrepant cases were reviewed. Results: Out of a total of 209 specimens, 193 (92%) had adequate diagnostic material while the remaining 16 specimens (8%) were inadequate for cytological assessment. The diagnostic specimens included 168 invasive ductal carcinomas (IDC), 15 invasive lobular carcinomas (ILC) and 10 mixed carcinomas (IDC and ILC). Excluding 19 cases with malignant cells on FNA in which no residual tumour was found in fibrotic lymph nodes after neoadjuvant therapy (cytology and histology confirmed on review) ThinPrep detected nodal metastasis with an overall sensitivity of 77.5%, specificity of 100%, PPV of 100% and NPV of 53.7%. Diagnostic accuracy was 82.2%. There was no difference in Bloom–Richardson grade or the number or size of metastases between tumours with true‐positive and false‐negative cytology. Sampling error was the sole factor contributing to cyto‐histological discrepancy. Conclusions: ThinPrep is a good alternative to the conventional smear for cytological assessment of axillary lymph node status in patients with invasive breast carcinoma, particularly when specimens are collected at remote sites or when cytologists are not available for assistance during FNA.     

Could nuclear matrix protein 22 (NMP22) play a role with urine cytology in screening for bladder cancer? – experience at Kuwait University

Objectives:  This prospective study was undertaken to evaluate nuclear matrix protein (NMP22) compared to urine cytology in the detection of bladder cancer and also to determine whether indexing suspicious cytology to NMP22 could enhance the clinical utility of cytology.
Methods:  Cytological findings of voided urine collected prior to a cystoscopic biopsy were correlated with urine NMP22 assay in 46 patients attending the urology clinic in Mubarak Al-Kabeer Hospital. The patients were clinically categorized into newly diagnosed cases of transitional cell carcinoma (TCC), recurrent TCC, TCC in remission and controls.
Results:  Using histological diagnosis as the gold standard the sensitivity and specificity of NMP22 were 78% and 43% respectively and of cases with malignant urine cytology were 30% and 87% respectively. If suspicious and malignant cytology were combined as positive results the sensitivity increased significantly to 87% while the specificity decreased but not significantly to 74%. Suspicious or malignant cytology enhanced by positive NMP22 gave a sensitivity of 70% and specificity of 87% neither of which was significantly different from cytology alone. There were three false positive cases on cytology and 13 false positive cases on NMP22 assay. There were three false negative cytology and five false negative NMP22 cases but only one was false negative for both, resulting in a high sensitivity (96%) but low specificity (30%) if either positive NMP22 or malignant or suspicious cytology was taken as a positive result.
Conclusion:  Combining NMP22 with malignant or suspicious cytological result improved sensitivity for the detection of bladder cancer but with a major decrease in specificity, suggesting a potential role in screening rather than diagnosis.     

Fine needle aspiration outcomes of masses detected by positron emission tomography: correlation with standard uptake value

OBJECTIVES: To characterize the cytopathologic outcome of lesions detected on positron emission tomography (PET) scan. STUDY DESIGN: Cases with fine needle aspiration (FNA) performed because of a PET-positive lesion over an 18-month period were reviewed. Correlation with the standard uptake value (SUV) (using 2.5 as a cutoff value) was carried out. RESULTS: A total of 112 FNAs were found, of which 83 had adequate tissue for evaluation and available corresponding SUVs to be included in the final study. Fisher's exact test was carried out for correlation between FNA diagnosis and SUV Sixty-one (73.5%) lesions had an SUV > or = 2.5, 53 (87%) ofwhich were malignant and 8 (13%) benign on cytology. Twenty-two (26.5%) lesions had an SUV < 2.5, of which 12 (54.5%) showed benign and 10 (45.5%) showed malignant cytology. The overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and diagnostic accuracy of SUV were 84%, 60%, 87%, 56% and 78%, respectively. CONCLUSION: Our data show that FNA procedures performed for PET-positive lesions have high PPV, but low NPV. Therefore interpretation of PET SUV values < 2.5 as benign should be made with extreme caution.     

Role of urinary NMP-22 combined with urine cytology in follow-up surveillance of recurring superficial bladder urothelial carcinoma

OBJECTIVE: To determine efficacy and utility of NMP-22 in follow-up of bladder urothelial carcinoma (UC) and compare NMP-22 as a single evaluating test vs combination with cytology. STUDY DESIGN: Ninety-four consecutive urine cytology samples of bladder UC were identified. Patients received follow-up urine cytology, NMP-22 testing and cystoscopy with surgical biopsy. RESULTS: NMP-22 specificity was 100%, sensitivity 45%, positive predictive value (PPV) 100% and negative predictive value (NPV) 87%. NMP-22 showed lower sensitivity for high-grade lesions and higher for low-grade lesions. Cytologic diagnosis had a high inconclusive rate; when regarded as positive, it resulted in 75% sensitivity, 58% specificity, 33% PPV and 89% NPV. NMP-22 correctly classified 60% of false negative cases diagnosed by cytology with low-grade UC and clarified 27 inconclusive cytologic diagnoses. NMP-22 misclassified 9 cases as false negative, all with high-grade UC; all were correctly identified on cytology as true positive. Combined interpretation showed 90% sensitivity, 92% specificity, 75% PPV and 98% NPV. CONCLUSION: NMP-22 complements cytology by its higher sensitivity for low-grade lesions; its values are not affected by bacillus Calmette Guérin therapy changes, which are limiting in cytology. Combined interpretation of NMP-22 and cytology shows promise as an effective, noninvasive method for surveillance of UC.     

Higher diagnostic accuracy with the ThinPrep method in a simulated intraoperative environment

Objective:  To compare the accuracy of intraoperative fine needle aspiration cytology samples prepared by the ThinPrep method to conventional cytological methods. Specimen adequacy and turn around time (TAT) were also assessed.
Methods:  Fifty consecutive fresh tumours submitted for histological analysis were aspirated and each prepared as follows: (i) direct smear with H&E stain, (ii) direct smear with Pap stain, (iii) ThinPrep slide with H&E stain, and (iv) ThinPrep slide with Pap stain. The slides were randomly distributed to three cytopathologists for interpretation. The quality of the preparation, the diagnosis and the time needed for interpretation were recorded.
Results:  Accuracy was measured as the percentage of absolute agreement between the cytological and the histopathological diagnoses of the lesions. Histologically, there were 43 malignant and six benign lesions and one atypical lipoma. The TAT began when the slides/cytolyte specimens arrived at the lab and ended with the pathologist's diagnosis.
Conclusions:  In terms of accuracy and specimen adequacy, ThinPrep slides with Pap stain is the best procedure. This advantage however is offset by the longer testing time.     

Diagnostic value of the thin-layer, liquid-based Pap test in endometrial cancer: a retrospective study with emphasis on cytomorphologic features

OBJECTIVE: To examine the diagnostic value (sensitivity and specificity of the ThinPrep Pap test in the detection of endometrial cancer and assess the morphologic features of endometrial cancer in ThinPrep tests. STUDY DESIGN: In a retrospective, case-controlled study, we identified 60 Pap slides performed within 12 months of the tissue diagnosis of endometrial carcinoma and 56 comparison slides from patients without known carcinoma. The slides were blindly reviewed by the authors without knowledge of the surgical diagnosis. An independent diagnosis was given for the tests based on 14 diagnostic criteria from the 2001 Bethesda System and 6 additional criteria proposed by the authors. RESULTS: The sensitivity of detecting endometrial carcinoma was 88.3% (95% CI 77.8-94.2%) and specificity was 87.5% (95% CI 76.4-93.8%). The positive likelihood ratio was 7.067 (95% CI 3.513-14.217) and negative likelihood ratio was 0.133 (95% CI 0.066-0.269). Enlarged nuclei and the presence of nucleoli in endometrial cells were the most reliable indicators of endometrial cancer or atypical endometrial cells. CONCLUSION: The ThinPrep Pap test has high sensitivity and specificity in detecting or suggesting the presence of endometrial cancer. Certain cytomorphologic features are helpful in distinguishing benign and malignant endometrial lesions.     

The Contribution of Qualitative CEUS to the Determination of Malignancy in Adnexal Masses,Indeterminate on Conventional US – A Multicenter Study

The aim of this study is to evaluate the efficacy of qualitative analysis of contrast-enhanced ultrasound (CEUS) in discrimination of adnexal masses which were undetermined by conventional ultrasound (US). A total of 120 patients underwent transabdominal CEUS. The initial enhancement time and intensity compared with the uterine myometrium, contrast agent distribution patterns and dynamic changes of enhancement were assessed. The sensitivity (Sen), specificity (Spe), positive predictive value (PPV), negative predictive value (NPV), accuracy (ACC) and Youden’s index were calculated for contrast variables. The gold standard was the histological diagnosis. There were 48 malignant tumors and 72 benign tumors. The enhancement features of malignant masses were different from benign ones. Earlier or simultaneous enhancement with inhomogeneous enhancement yielded the highest capability in differential diagnosis, and Sen, Spe, PPV, NPV, ACC, Youden’s index was 89.6%, 97.2%, 93.2%, 95.6%, 93.3%, and 0.88, respectively. The qualitative evaluation of CEUS is useful in the differential diagnosis of adnexal masses where conventional US is indeterminate.     

IS MEASUREMENT OF CIRCULATING TUMOR DNA OF DIAGNOSTIC USE IN PATIENTS WITH THYROID NODULES?

Objective: Circulating tumor DNA (ctDNA), a subset of cell-free DNA (cfDNA), is a potential biomarker for thyroid cancer. We determined the performance of a ctDNA panel for detecting thyroid malignancy in patients with thyroid nodules.Methods: Sixty-six patients with thyroid nodules without a prior history of cancer enrolled in a prospective, 1-year study in which blood was drawn for ctDNA analysis prior to undergoing fine-needle aspiration biopsy (FNAB) of thyroid nodules. The ctDNA panel consisted of 96-mutations in 9 cancer driver genes. The primary outcome measures were the sensitivity, specificity, and positive and negative predictive values (PPV, NPV) of our ctDNA panel for the diagnosis of thyroid malignancy as determined by pathologic and/or molecular tissue examination.Results: Results from 10 subjects could not be determined due to inadequate volume or technical issues. The final classifications of the thyroid nodules were 13 malignant and 43 benign lesions. A KRAS G12V mutation was detected in the plasma of 1 patient with stage IVA papillary carcinoma whose tissue contained the same mutation. Two of the 43 patients with benign lesions also had ctDNA detected, giving a sensitivity of 7.7%, specificity of 95.35%, PPV of 33.33%, and NPV of 77.35%. There were no significant differences between benign or malignant lesions in cfDNA levels.Conclusion: Neither cfDNA measurements nor our panel of ctDNA mutations are sensitive or specific enough to provide valuable information over FNAB. An expanded panel and the inclusion of proteomics may improve sensitivity and specificity for thyroid cancer detection.Abbreviations: cfDNA = cell-free DNA; ctDNA = circulating tumor DNA; FNAB = fine-needle aspiration biopsy; NIFTP = noninvasive follicular thyroid neoplasm with papillary-like nuclear features     

Bile duct brushings cytology – improving sensitivity of diagnosis using the ThinPrep® technique: a review of 113 cases

OBJECTIVE: Common bile duct (CBD) brushings have been recognized as a technique of moderate sensitivity and high specificity in identifying carcinoma of the ampulla and pancreatico-biliary regions. This study evaluated the increase in sensitivity of this technique using the ThinPrep technique of specimen preparation when compared with conventional cytology smears. METHODS: A total of 113 bile duct brushings were included in the study (38 conventional smears and 75 slides prepared using the ThinPrep technique). All slides were reviewed by one cytologist. Five categories of reporting were used: inadequate, negative, atypia, suspicious and malignant. RESULTS: The inadequate category of reporting disappeared in the ThinPrep group with improved specimen fixation and preparation and hence reduced artefact. Sensitivity of diagnosis of malignancy increased from 39% in conventional smears to 53% in the ThinPrep group. Specificity, positive and negative predictive values and accuracy were 100%, 100%, 60% and 68% for conventional smears and were 100%, 100%, 60% and 72%, respectively, for ThinPrep specimens. CONCLUSIONS: ThinPrep technique was associated with increased sensitivity of diagnosis, in part due to improved specimen fixation and reduced artefact. Cytology of bile duct brushings is an important diagnostic tool for sites from which it can be difficult to obtain a histology biopsy. It may therefore provide the only opportunity for tissue diagnosis of carcinoma from these sites, hence the importance of optimizing sensitivity.     

Improving recognition of thyroid carcinoma in rapid-consultation specimens

OBJECTIVE: To improve recognition of thyroid carcinoma in rapid consultation on Diff-Quik-stained (Fisher Diagnostics, Middletown, Virginia, USA.) fine-needle aspiration (FNA) and rapid hematoxylin-eosin (H-E)-stained intraoperative scrape preparation (ISP) specimens by assessing 3 variables (anisokaryosis, nuclear overlap [NO] and scant/absent colloid) in cases of cellular follicular lesions (CFL), an indeterminate diagnostic category. STUDY DESIGN: Thirty-seven FNAs and 28 ISPs diagnosed as CFL, with histologic follow-up, were evaluated in blinded fashion by 3 cytopathologists assessing the 3 variables. RESULTS: Over 90% of the malignant cases showed NO in both FNA and ISP, while only 22% of the benign cases did; positive and negative predictive values (PPV and NPV) were 82% and 100%. All malignant cases showed significant anisokaryosis in both FNA and ISP in contrast to 24% of benign cases; PPV and NPV were 74% and 100%. Scant/absent colloid was seen in 87% and 39% of malignancies in FNA and ISP, respectively, as compared to 55% and 20% of the benign cases. PPV and NPV were 52% and 83% in FNA and 63% and 60% in ISP, respectively. CONCLUSION: Application of these variables improves recognition of thyroid carcinoma, particularly in fine needle aspirates, while additional material may be requested. With ISP, their absence supports recommending against further surgery. Together, optimal surgical planning and outcome may be obtained.     

Can the American Thyroid Association,K-Tirads,and Acr-Tirads Ultrasound Classification Systems Be Used to Predict Malignancy in Bethesda Category IV Nodules?

Objective: Management of thyroid nodules with Bethesda category III and IV cytology on fine needle aspiration (FNA) is challenging as they cannot be adequately classified as benign or malignant. Ultrasound (US) patterns have demonstrated the utility in evaluating the risk of malignancy (ROM) of Bethesda category III nodules. This study aims to evaluate the value of 3 well-established US grading systems (American Thyroid Association &lsqb;ATA], Korean Thyroid Imaging Reporting and Data System &lsqb;Korean-TIRADS], and The American College of Radiology Thyroid Imaging Reporting and Data System &lsqb;ACR-TIRADS]) in determining ROM in Bethesda category IV nodules.Methods: Ninety-two patients with 92 surgically resected thyroid nodules who had Bethesda category IV cytology on FNA were identified. Nodule images were retrospectively graded using the 3 systems in a blinded manner. Associations between US risk category and malignant pathology for each system were analyzed.Results: Of the 92 nodules, 56 (61%) were benign and 36 (39%) were malignant. Forty-seven per cent of ATA high risk nodules, 53% of K-TIRADS category 5 nodules, and 50% of ACR-TIRADS category 5 nodules were malignant. The ATA high-risk category had 25% sensitivity, 82% specificity, 47% positive predictive value (PPV) for malignancy. K-TIRADS category 5 had 25% sensitivity, 85% specificity, 53% PPV for malignancy. ACR-TIRADS category 5 had 25% sensitivity, 84% specificity, 50% PPV for malignancy. None of the 3 grading systems yielded a statistically significant correlation between US risk category and the ROM (P = .30, .72, .28).Conclusion: The ATA, Korean-TIRADS, and ACR-TIRADS classification systems are not helpful in stratifying ROM in patients with Bethesda category IV nodules. Clinicians should be cautious of using ultra-sound alone when deciding between therapeutic options for patients with Bethesda category IV thyroid nodules.     

Ultrasonographically and stereotactically guided fine-needle aspiration cytology of non-palpable breast lesions: cyto-histological correlation

Objective:  To determine the diagnostic accuracy of ultrasonographically (US) and stereotactically guided fine needle aspiration cytology (FNAC) for the diagnosis of non-palpable breast lesions.
Patients and methods:  From January 2002 to December 2004, 470 women with 478 mammographically detected non-palpable breast lesions had US or stereotactically guided FNAC of the breast lesion. Subsequent histological evaluation of the same lesion was performed at the Institute of Oncology Ljubljana, Slovenia. The correlation between the original cytological and histological diagnosis was assessed and the sensitivity, specificity, positive predictive value and negative predictive value were calculated.
Results:  Among US-guided FNACs, 144 (53.5%) were histologically verified benign lesions and 125 (46.5%) were carcinomas. Cytological diagnoses were: true positive (TP) in 63 cases (50.4%), suspicious in 35 (28%), false negative (FN) in eight (6.4%), and in 19 (15.2%) cases, the material was inadequate for diagnosis. In the stereotactically guided FNAC group, there were 209 women with 209 lesions, with 95 (45%) histologically proven carcinomas and 114 (55%) benign lesions. Cytological diagnoses were TP in 49 (51.6%) cases, true suspicious in 21 (22.1%), FN in nine (9.5%), and in 16 (16.8%) cases, the material was not adequate for the diagnosis.
Conclusion:  The sensitivity of 88.7% and specificity of 98.6% for US-guided FNAC and 84.5% and 100% for stereotactically guided FNAC, respectively, suggest that clinicians can rely upon cytological diagnosis for planning further management of women with non-palpable breast lesions.     

The role of cervical cytology and colposcopy in detecting cervical glandular neoplasia

Objective:  To determine the role of cervical cytology and colposcopy in the management of endocervical neoplasia.
Setting:  Colposcopy unit and cytology laboratory in a teaching hospital.
Sample:  Group 1 included 184 smears showing endocervical glandular neoplasia from 129 patients and group 2 included 101 patients with histology showing endocervical abnormalities in a 6-year period (1993–1998). Follow-up of 6–11 years to 2004 was available.
Methods:  Group 1 were identified from the cytology computer records. Group 2 were identified from histology records on the cytology database and a record of histology cases kept for audit purposes. The clinical records were examined retrospectively.
Results:  The positive predictive value (PPV) of abnormal endocervical cells in smears was 81.1% for significant glandular/squamous [cervical glandular intraepithelial neoplasia (CGIN)/cervical intraepithelial neoplasia grade2 (CIN2 or worse)] lesions. The PPV of colposcopy was 93.5% for significant glandular/squamous lesions of the cervix. The postcolposcopy probability of a significant lesion when colposcopy was normal was 87.5%. The sensitivity of colposcopy in detecting endocervical lesions was 9.8%. The sensitivity of cervical smears in detecting a significant endocervical abnormality (CGIN or worse) was 66.3%. The false negative rate for cytology of endocervical glandular lesions was 4.0%.
Conclusions:  Endocervical glandular neoplasia detected on cytology is predictive of significant cervical pathology even when colposcopy is normal, which supports excisional biopsy in the primary assessment of these smears. The high concomitant squamous abnormality rate justifies the use of colposcopy to direct biopsies from the ectocervix. Cervical cytology is the only current screening method for cervical glandular abnormalities but sensitivity is poor.     

Is CSF cytology a useful diagnostic procedure in staging paediatric CNS tumours?

Objectives:  To evaluate whether there are any factors that predict malignant cells being found in paediatric cerebrospinal fluid (CSF) samples. To determine whether CSF provides useful staging information not provided by magnetic resonance imaging (MRI) in paediatric patients with primary central nervous system (CNS) malignancy.
Methods:  We compared the CSF cytology and spinal MRI staging results in paediatric patients with primary CNS malignancy at a UK tertiary referral centre, over a decade.
Results:  Of 159 CSF samples, 72 samples were from 72 patients with primary CNS malignancy with spinal MRI available for comparison. Eight of these 72 had positive cytology (seven malignant and one suspicious). All had a high clinical suspicion of tumour at the time of sampling. Of the 72 patients, only two had evidence of CSF spread on MRI spinal staging and CSF cytology; ten had MRI without cytological evidence and six had cytological without MRI evidence.
Conclusions:  In paediatric patients with primary CNS tumours, CSF cytology provides useful staging information. Spinal MRI alone may miss some patients with CSF spread who would be identified with CSF cytology.     

3q26 amplification is an effective negative triage test for LSIL: a historical prospective study

Background

Women with low grade squamous intraepithelial lesions (LSIL) at cervical cancer screening are currently referred for further diagnostic work up despite 80% having no precancerous lesion. The primary purpose of this study is to measure the test characteristics of 3q26 chromosome gain (3q26 gain) as a host marker of carcinogenesis in women with LSIL. A negative triage test may allow these women to be followed by cytology alone without immediate referral to colposcopy.

Methods and Findings

A historical prospective study was designed to measure 3q26 gain from the archived liquid cytology specimens diagnosed as LSIL among women attending colposcopy between 2007 and 2009. 3q26 gain was assessed on the index liquid sample; and sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were measured at immediate triage and at 6–16 months after colposcopic biopsy. The sensitivity of 3q26 gain measured at immediate triage from automated and manually reviewed tests in 65 non-pregnant unique women was 70% (95% CI: 35, 93) with a NPV of 89% (95% CI: 78, 96). The sensitivity and NPV increased to 80% (95% CI: 28, 99) and 98% (95% CI: 87, 100), respectively, when only the automated method of detecting 3q26 gain was used.

Conclusions

3q26 gain demonstrates high sensitivity and NPV as a negative triage test for women with LSIL, allowing possible guideline changes to routine surveillance instead of immediate colposcopy. Prospective studies are ongoing to establish the sensitivity, specificity, PPV and NPV of 3q26 gain for LSIL over time.     

Reclassification of "atypical" diagnoses in endoscopic retrograde cholangiopancreaticography-guided biliary brushings

OBJECTIVE: To evaluate the usefulness of reclassifying "atypical" diagnoses in reporting biliary cytology using strict morphologic criteria. STUDY DESIGN: Cytologic specimens from 139 patients (direct, alcohol-fixed smears or cytocentrifuge preparations) were evaluated. Diagnoses were benign (70), atypical (36) and malignant (33). Using strict criteria--major (nuclear contour, chromatin pattern) and minor (polarity, cell types, nuclear size, nuclear grooves, nucleoli, mitosis, nuclear/cytoplasmic [N/C] ratio)--atypical cases were reevaluated and reclassified. Follow-up (F/U) was available on all cases. RESULTS: Atypical cases, (36) were reclassified as malignant (26), atypical favor benign (2)/reactive (3) and atypical, not otherwise specified (NOS) (5). Cases reclassified as malignant showed irregular nuclear contours, chromatin irregularities and rare mitosis. Nuclear enlargement, nucleoli and cellularity varied widely in all groups. N/C ratio was increased in most reclassified malignant cases. All 26 malignant reclassifications correlated with F/U of malignancy. Benign and reactive cases (5) were negative for malignancy on F/U (4), and in 1 case a metastatic carcinoma involving the biliary tree was found. In the 5 atypical (NOS) cases, F/U showed malignancy (3) and pancreatitis (2). Cytocentrifuge preparations made in our laboratory were of superior quality when compared to other methods of cell preparation. CONCLUSION: Irregularities in nuclear membrane and abnormal chromatin pattern were the most consistently useful features correlating with malignancy. The sensitivity and specificity of biliary brush cytology can be enhanced by using strict cytomorphologic criteria and proper collection and fixation, all of which decrease atypical diagnoses.