ABSTRACTS FROM RECENT CYTOLOGY JOURNALS

Pulmonary hamartoma (PH) in FNA; Wood, Frost et al, page 412

Apart from re-stating the very well known characteristics of pulmonary hamartoma, this article shows that the fibromyxoid stroma of PH stains strongly with S-100, which other background elements do not.

Specific MTSS antibody in TB FNA; Goel and Budhwar, page 424.

This article from Lucknow on 340 cases stained with specific monoclonal antibody MTSS raised against a 38- kDa immunodominant protein antigen was positive in 96 and 97% of archival and fresh FNA smears, far better than ZN positivity, observed in 44% and 42% of archival and fresh FNA samples. The antibody is available from Novocastra. The pictures are totally hopeless, but the text is good.

Three methods for quality control in gynecologic cytology; Utagawa et al, page 439.

This article from Portugal used three methods to determine effectiveness of quality assurance. Their selection of cases by high risk criteria for rescreening selected 20.7% of samples with atypia. In contrast, 100% rapid review selected only 2.5%. The currently accepted 10% rescreening selected only 0.06%. Mounting evidence is accumulating that the 10% rescreening of samples is less effective than other options.

Lack of utility of CK5/6 in mesothelioma vs adenocarcinoma effusions; Dejmek, page 579.

Annika Dejmek, Lund University, Sweden applied CK5/6 immunostaining to ethanol-fixed cytospin preparations from 74 benign and malignant effusions. Contrary to experience with histologic sections, the conclusion was that whereas CK5/6 was positive in 7 of 8 mesotheliomas (and in 9 of 11 benign mesothelial proliferations), it was also positive in 11 of 17 pulmonary adenocarcinomas and 12 of 31 adenocarcinomas of non-pulmonary origin. Further, reactivity was also found in 3 of 5 non-small lung cancers and 1 squamous carcinoma. This article quotes the only apparent prior study of CK5/6 in effusions, by Han, which was performed on formalin-fixed cell blocks, and which also reported reactivity of CK5/6 in 25% of adenocarcinomas tested. These results indicate that CK5/6 is a non-specific marker in effusion cytology. (speak to Scott before this is distributed)

Brush cytology in the diagnosis of Helicobacter; Mostaghni et al, page 597.

EUS of the upper GI has resulted in almost complete disappearance of gastric brushings! An article from Iran on 109 patients with dyspepsia is a timely reminder that gastric brushing cytology provides a sensitive, inexpensive, easy and accurate technique for rapid detection of Helicobacter pylori infections. The sensitivity of brush cytology (95%) was higher than that of histology (81%), and rapid urease testing (72%). This article utilized Giemsa staining, but the very characteristic organisms can just as easily be seen on Papanicolaou. The morphologic features are sufficiently specific, and immunostaining is unnecessary.

INCIs in cervical cytology!; Hashi et al, page 535.

This article from Japan is likely to be of interest to surgical pathologists, and to those seeing conventional smears. It compares lobular endocervical glandular hyperplasia (LEGH) with the very lethal minimal deviation adenocarcinoma (MDA), a diagnosis difficult even on surgical pathology. The article has good black and white images, and for the first time describes the presence of intranuclear cytoplasmic inclusions (INCIs) in 22 of 24 cases of LEGH; these were not found in MDA.

HBME-1 and CK19 in papillary thyroid carcinoma (PTC); Nga et al, page 550.

I have avoided articles on immunocytochemistry in thyroid FNA, as there is a well known lack of sensitivity and/or specificity in most studies. However, this article was fairly convincing for the dual positivity of HBME-1 and CK19 in papillary thyroid carcinoma, which stained all of 9 PTC. These were compared with Hurthle cell neoplasms (HCN) and benign nodules. Staining for HBME-1 or CK19 was occasionally seen in HCN and benign nodules, but the dual staining was never seen in a non-PTC or benign nodule. Stains were done on cell block material. PTCs showed cytoplasmic and membranous staining with luminal accentuation. The authors also insisted on greater than 25% of cells with at least moderate (2+) staining intensity for a slide to be read as positive. It is not stated whether any of the cases were follicular variants of PTC.

KOC and S100A4 immunoreactivity in biliary brushings; Ligato et al, page 561.

Diagnosis of malignancy in biliary brushings is known to lack sensitivity by cytology alone. This article describes improvement in the sensitivity for the diagnosis of pancreaticobiliary malignancies from 83% (which is in itself unusually high) to 100% using positivity for KOC or S100A4 antibodies. This was done on alcohol-fixed cytology slides. Specificity was reasonable at 95%.

Ciliated tufts in an endodermal cyst of the spinal canal; Naran et al, page 605.

We have seen ciliated columnar cells and recognized their importance in the diagnosis of hepatic ciliated foregut cysts and in duplication cysts seen in other sites. This Letter illustrates and describes ciliated columnar cells obtained from aspiration of an spinal canal endodermal cyst level L1-L2 vertebral body.

Ciliocytopththoria, multiflagellated protozoa, or both?; Martinez-Giron et al, page 609

Two pages on after the ciliated endodermal spinal cyst, you will see a lovely letter with excellent black and white pictures, submitted from several authors in Spain. They depict degenerating columnar cells, with terminal bars and regularly arranged cilia, and compare those with multi-flagellated protozoal organisms. The latter have no terminal bar, and their cilia lack the regularity of those seen in bronchial epithelial cells. I'm not sure of the practicality of this observation, but the pictures are really good.

Tissue Microarrays of Cervical Intraepithelial Neoplasia, pages 123-129.

With a fanfare editorial (pages 121-122), Volker Schneider of a private lab in Freiberg, Germany, has created tissue microarrays of cervical squamous and glandular lesions to test the feasibility of studying molecular markers. The pictures are excellent, and the diagnosis represented in the biopsy itself in 82% of blocks cored. I see somewhat limited value for these cervical fragments but microarrays are a wide open field for cytopathology research and could easily be performed on archival blocks of cytological fluids and FNAs, in which its use would have far more potential.

Acetic Acid Recovery of Low Cellularity Cervical Smears, pages 136-140..

This paper, from the University of Toronto, will be of interest to all involved in technical aspects of ThinPrep processing. Adequate samples containing greater than 5,000 cells were retrieved from 147 Pap tests requiring reprocessing. Recovery was almost 90% in cloudy samples, down to 72% in blood-stained samples. Further, abnormal cells were identified in 9% of reprocessed samples. (See later article, page 2, on same topic)

Assessment of Cytologic Preparations Made From 19 Gauge Core Needles, pages 141-146.

We are occasionally presented with core biopsies (usually after failed FNA) and asked to make a rapid cytologic diagnosis on them. A group from Boston University Medical Center assessed 40 slides from 10 cases, prepared by four different methods - touching, rolling, dragging or pushing the material along a glass slide. The overall adequacy of the cytoslide was 88%, but their accuracy only 50% in comparison with later histologic diagnoses. I take these figures from their table, rather than their abstract or text, which seems to make these figures higher ?!? They have identified the touch method as superior (cytologic accuracy 5/10), but their table shows the roll and push methods each had an accuracy of 7/10. This is an unsatisfying article, but the figures could be used to inform clinicians that the likelihood of getting an accurate diagnosis from a core is only in the region of 50%, which I am sure everybody would agree is poor.

CDX-2 _Expression in Pulmonary FNA, pages 191-195.

TTF-1, CK7 and CK20 may be controversial in the differential diagnosis of mucinous lung and colonic adenocarcinomas in pulmonary FNAs. Bronchioloalveolar carcinoma (BAC) may be negative for TTF, and aberrantly positive for CK-20! CDX-2 has been used in histopathology, being a good transcription factor in normal adult colonic mucosa. This group from NYU showed CDX-2 to have a 75% sensitivity, 100% specificity for colonic tumors in pulmonary FNA. However, others have reported CDX2 in ovarian, ancreaticobiliary, gastric and esophageal mucinous tumors. Importantly, in several studies all lung carcinomas have been negative for CDX2. The addition of CDX2 (available in this Lab) may be valuable if TTF-1, CK7 and CK20 yield aberrant results. As with TTF-1, CDX2 is a nuclear stain which varies in intensity, but does not cause back ground staining. (Having said all that, there has been a report of a rare pulmonary tumor, the goblet cell variant of primary mucinous/colloid carcinoma, which has been positive for CDX2 as well!). Please note that it is only the mucinous tumors of lung, colon, ovary, pancreas which behave in an aberrant fashion with TTF-1 and CDX2.

Polyoma Virus Infection as a Risk Factor for Bladder Cancer, pages 201-203.

To date, no study has statistically linked polyoma virus (PV) including BK, JC and SV40 virus types, to bladder cancer development. This group from Mount Sinai School of Medicine, New York, has reviewed the subsequent urinary history in 3782 patients having urine cytology. (Almost 70% of tumors which developed were of low grade: ie polyoma virus was not being mistaken for cancer). Over 3 years, 15.8% of PV positive patients developed bladder cancer, compared with 3.75% of those without PV (p <0.001). Transplant patients were excluded, as were those being followed for prior bladder cancer. All patients were immunocompetent. I am left wondering why such a momentous finding is relegated to the middle pages of Diagnostic Cytopathology?

Protocol for Blood-Stained ThinPrep Paps Using Glacial Acetic Acid, pages 210-213.

Uncannily, in concert with the prior article in Acta Cytologica, is a study from Madison, Wisconsin, in which a detailed protocol is presented for the use of glacial acetic acid in minimal or major blood-staining of ThinPrep tests. 94% of glacial acetic processed (GAP) cases were thereafter adequate for cytologic interpretation, and amongst them, 11% were found to have a lesion.

Intracytoplasmic Lumina in Micropapillary Carcinoma of the Lung, pages 224-226.

In 2002, Mahul Amin described a poor prognostic variant of lung adenocarcinoma which he termed micropapillary. Only two cytologic descriptions exist, both from Japan, and this is the second one. It describes the presence of micropapillary fragments in a pleural fluid of a patient after treatment for micropapillary lung cancer. In addition to the micropapillary formations, distinctive intracytoplasmic lumina, positive for Alcian Blue and PAS, were noted. Have we ever had an FNA from micropapillary lung cancer?

High Risk HPV Testing in the AGC Category, pages 235-239.

According to the Bethesda System, diagnoses of AGC are referred for colposcopic evaluation and/or endocervical curettage. A combined paper by Dr. Prabodh Gupta and three New York labs has performed HC2 testing on 144 Pap tests with a result of AGC, and in particular those 60 patients in whom follow-up was available. Of 43 HPV-positive patients with AGC, HSIL was found in 27, LSIL in 6, and ASCUS in 3. One patient had AIS, and one had microglandular hyperplasia. Five patients had no abnormality visualized on colposcopic evaluation. Of 60 AGC positive, HPV negative women with follow-up, the most common diagnoses were endometrial polyps (9), endometrial cancer (2), microglandular hyperplasia (2), endocervical polyp (4), no AIS and 1 HSIL. The conclusion is that most cases of AGC result in a squamous abnormality being diagnosed, and that HPV testing would be valuable in directing attention to the squamous rather than to the endocervical or endometrial epithelia. Of 103 patients diagnosed with AGC, who had follow-up, 43 showed no abnormality.

Cytotechnologist-Cytopathologist Discrepancy in Non-Gyn Cytopathology, pages 265-269

CLIA '88 does not require the cytotechnologist (CT) to screen non-GYN cases, and there are few guidelines for quality assessment and quality improvement. This article from the University of Pittsburgh by Drs. Ohori and Raab attempt to define errors of CT-screened non-GYN samples numerically, classifying them as minor or major discrepancies. It promotes review of discrepant cases at consensus, to improve and monitor CT performance in non-GYN cytopathology, and to keep the CT informed of complexities of this form of diagnosis. I can't recall ever having read something as negative as the subsequent article by Gary Gill, called "Counterpoint". He concludes totally the reverse viewpoint, stating flatly that CLIA does not require CT screening of non-GYN preparations, that this effort is not reimbursable, that harmonizing interpretations of technologists and pathologists is impossible, that cytotechnologists' interpretations do not influence patient management, that false negative non-GYN interpretations have not resulted in law suits, and that monitoring discrepancies diverts precious human resources from more effective utilization elsewhere in the lab! Which side do you choose?

Deep-Seated Lymphoma/Leukemia Diagnosis by EUS FNA, pages 703-709

Dr. David Chhieng's group describes 13 cases of deep-seated lymphoma, 10 primary and 3 recurrent, together with a case of hairy cell leukemia, diagnosed on ultrasound-guided endoscopically directed FNA of deep-seated lymph nodes. Ancillary studies were performed, obviating the need for tissue diagnosis.

Repeat Thyroid FNA in Patients with Initial Benign Results, pages 698-702.

A group from the has examined characteristics of 402 patients undergoing thyroid surgery during a 22 month period. Their focus was the value of repeat FNA when initial cytodiagnosis was negative. In 70 patients, they found that one repeat FNA increased sensitivity for malignancy from 81.7% to 90.4%, and decreased the false negative rate from 17.1% to 11.4%. There was no further improvement in performance characteristics with more than one repeated FNA. The indication for repeat FNA was almost always given as the persistence of a dominant thyroid nodule.

TTF-1 and HepPar-1 in Hepatocytes, pages 722-726.

This article, from our own department, includes in its authorship Yijun Pang, Jeannette Mitchell, Sharon Mount and Kum Cooper. They studied HepPar-1 and TTF-1 in normal human liver tissue by iImmunoelectron microscopy, using renal tubules as a control sample. Both antibodies were found to label mitochondria of hepatocytes, but not those of renal tubular epithelium. The specific binding of TTF-1 in hepatocytic mitochondria suggests a potential usefulness for identifying HCC.

Medicolegal Aspects of Error in Pathology, pages 617-619.

From the School of Public Health at the University of California, Berkeley, David Troxel investigated claims made to The Doctors Company in Napa, California, between 1998 and 2003. Of all claims, 57% involved five categories of specimen type/diagnosis. False negative diagnosis of melanoma was the single most common reason. Breast biopsy claims were a close second, but when combined with breast FNA and breast frozen section, breast specimens were the most common cause of pathology malpractice claims. Cervical Pap tests were third in frequency, and 98% of these involved false negatives. Of gynecologic surgical pathology claims 42% involved misdiagnosis of ovarian tumors. Many other facts and figures are quoted in this article, which is recommended reading for all concerned in quality assurance.

Rapid Rescreening of Cervical Smears in a High Risk Population, pages 110-115

Pam Michelow, Grace McKee and Felicity Hlongwane have reported from the Cytopathology Unit in Johannesburg on their results of six years' rapid review conventional Pap tests. Of almost a quarter of a million cases, 62866 or 26% underwent rapid review. Amended reports were issued in 0.59% of these cases. The amendments included 101 cases with HSIL and ASC-H (classified together), 143 LSILs, 54 ASCUS cases, and 33 ACG cases. False negative proportion for HSIL and ASC-H combined was 5.76. No squamous cell carcinomas were diagnosed on rapid review, but one patient with HSIL/ASC-H on review had squamous cell carcinoma on biopsy. Three technologists with a lower sensitivity of primary screening were identified and required retraining.

EUS-FNA of Pancreatic Neuroendocrine Tumors, page 10

Nine cases with immunocytochemistry on cell blocks, mainly CD56, synaptophysin, chromogranin and PGP9.5 are described. Nothing new in the way of morphology.

Impact of a Cytopathologist Expert on 106 EUS procedures, page 18

I would love to know the story behind this study, which emanates from the Pasteur Hospital in France. Cases (n=46) causing difficulty in interpretation by the local cytopathologist were mailed to an expert cytopathologist; sensitivity for malignancy was 72% (local) vs 88% (expert). This series is very odd because they appear to have 19 benign serous cystadenomas! These tumors are incredibly rarely reported in FNA. They have four mucinous cystadenomas and four pseudocysts. I cannot imagine that they were dealing with what we would term microcystic serous adenomas.

EUS Sampling by FNA and Trucut (ie core biopsy), page 27

This appears to be a combined study by several London hospitals. A total of 159 patients underwent FNA alone, or combination of both sampling modalities. The findings were that for lesions greater than 2.0 cm, combining FNA/trucut (core) biopsy improved the diagnostic accuracy. Overall accuracy for EUS FNA alone was 77%, for EUS-core 73%, and for the combined sampling 91%. Core attempts failed to diagnose 14% (6/43) of pancreatic malignancies diagnosed by EUS FNA. It was more useful in peripancreatic tumors such as lymphoma, GIST, small cell tumors, and neuroendocrine tumors. Two complications occurred, one case of abdominal pain and one of a minor bile leak. (I don't like 19 gauge needles used in any form of FNA, especially in the pancreas. With the possibility of leak of digestive pancreatic enzymes, I would never be happy with a core needle procedure.)

EUS FNA of the bile duct and gallbladder, page 42

A group from the University of Alabama, including David Chhieng, has reported 53 cases of bile duct and gallbladder aspirates in this UK journal. I wonder why?

IPEX of Peritoneal Mesothelioma and Carcinoma of the Peritoneum, page 67

Here is the second article on this topic. It is concentrated specifically on 20 mesotheliomas, 14 primary peritoneal carcinomas, and 14 metastatic serous ovarian carcinomas, applying a panel of 16 antibodies. Of 20 mesotheliomas, 17 (85%) stained for calretinin, all being negative for ER and PR. ER was positive in 93% of 28 carcinomas and PR was positive in 29% of 28 carcinomas. Positive staining for BER-EP4 was identified in 96% of 28 carcinomas and 2 (10%) of 20 mesotheliomas. For this particular diagnostic problem, a combination of calretinin, BER-EP4, estrogen and progesterone proved useful.

Endometrial Cell Clusters - Immunocytochemistry, p 77

A group from Harvard, included Ed Cibas, has studied endometrial cell clusters in ThinPrep slides using a stromal marker (CD10) and an epithelial marker (AE1/AE3). Of 59 clusters, 37% were purely stromal, 29% purely glandular, and 34% mixed. Specifically, as we suspected always, the classic "exodus" pattern has a stromal core with an epithelial envelope!

Immunodistinction of peritoneal mesotheliomas and serous carcinomas, page 34

Here is the third paper (of a current inexplicable glut) on immunohistochemistry, specifically targeting epithelioid mesotheliomas vs serous carcinomas. The single author, Nelson Ordoñez, regards BER-EP4 and MOC31 (epithelial markers), together with calretinin and D2-40 (mesothelial markers) as best discriminators. .

IPEX in epithelioid mesothelioma: analysis of published data p 223

This paper by King, Thatcher et al of Manchester, UK, is important in that it is an analysis of 88 published papers on the subject, expressing sensitivity and specificity of 15 antibodies specifically to distinguish pulmonary adenocarcinoma from pleural mesothelioma. No single antibody made the distinction in this review. For pulmonary adenoca, they advise MOC-31 and BG-8 (both 93% sensitivity); monoclonal CEA and TTF (both close to 100% specific). For mesotheliomas, CK5/6 and HBME-1 were 85% sensitive, CK5/6 and WT1 85% and 96% specific respectively. Is this the final word? I doubt it.

Three Preparation Methods and Immunocytochemistry in Effusions, page 6

There seems to be a glut of papers recently on the distinction of adenocarcinoma, reactive mesothelial cells and mesothelioma in serous effusions. I will quote all of them, as I know the subject is of interest and importance. This first article found a panel of MOC-31, Ber-EP4, CA19-9 and CEA to be suitable to distinguish between adenocarcinoma and reactive mesothelial cells. They found ethanol smears show stronger immunoreactivity to all antibodies tested than ethanol or formalin-fixed cell blocks!

FISH to distinguish follicular lymphoma gr. I and II from reactive hyperplasia, page 11

Follicular lymphomas may present difficulties in diagnosis as there is a range of lymphoid size (traditionally regarded as a "benign or hyperplastic" morphologic feature). In such circumstances, FISH analysis for chromosomal translocation t(14;18) is critical for final cytologic diagnosis.

Vegetable Cells in Pap Smears, page 45

The authors themselves advise going to the online version at www.interscience.wiley.com to see the color pictures of examples of vegetable cells in Pap smears, the main importance being the resemblance of cells from various vaginal pessaries to koilocytes.

Dysbacteriosis in Cervical Smears of Dutch-Moroccan Immigrants, page 56

Dr. Tilde Boon, who has written so much of note in the past, here surmises that the difference she has noticed between the vaginal flora of Moroccan immigrants and native born Dutch women was due to cultural differences in vaginal hygiene, in particular the practice of douching, by the Moroccan immigrant women. The article appears under my name as Section Editor. Please know that I rejected this article on the basis that the sexual practices of the two groups have not been taken into account, and that the promotion of douching was not something I wished to be associated with. I can send you my letter of rejection if you wish to see it.

Effect of Bethesda 2001 on Categories of ASC and Particularly ASC-H, page 62

This combined authorship from NYU and the University of Maryland has written a very learned article on what Bethesda 2001 has done to the stratification of women with various types of ASC. In essence, they confirm that Bethesda 2001 reduces the ASC rate and better identifies women at risk for high grade SIL. They add, and I feel this is very important and not previously proven, that ASC-H requires very close clinical follow up, as a significant % of women with ASC-H require more than one colposcopy to identify the high grade lesion.

Editorial: Modern Approach to Lymphoma Diagnosis by FNA, page 429.

Ruth Katz has written an editorial introducing a number of articles in this and the upcoming issue of Cancer Cytopathology attempting to "restore respect to what was recently called an inaccurate and unnecessary procedure that may delay or misguide treatment" - ie FNA of lymphomas. Ruth has mustered the troops in answer to the damaging article by Hehn et al (J Clin Oncol 2004; 22:3046-3052), which she mentioned at our Cyto Update in May 2005. If you missed her talk, this Editorial and the following two articles will get you up to speed.

Strategies to Diagnose Lymphomas by FNA Using Ancillary Studies, page 432.

Nancy Caraway, from Ruth Katz's department at MD Anderson, has written an excellent and beautifully illustrated article on an approach to lymphoma diagnosis by FNA, with particular reference to ancillary studies such as immunophenotyping, ploidy analysis, cytogenetics and molecular studies. Her approach is that of monotonous lymphoid populations of small cells, monomorphous populations of medium size cells, monomorphous populations of large cells, a polymorphous lymphoid population and, finally, Hodgkin's lymphoma. The article is well worth reading.

Flow Cytometry in the Diagnosis of Lymphomas by FNA, page 443.

This article completes the triplet of lymphoma articles in this State of the Art Symposium. It is well illustrated and highly worth reading.

ASC-H: HPV Testing Rather Than Colposcopy? page 457.

A group from Rochester raises the interesting proposition of reflex HPV testing possibly supplanting the current recommendation of colposcopic evaluation for ASC-H. They studied all ASC-H Pap smears over a two year period which had HPV DNA and colposcopic biopsy verification. All patients found to have HSIL had positive HPV results (100% NPV). Among patients with negative follow up, 50% had positive HPV, 50% negative HPV results. It is this latter group of 50% ASC-H with negative HPVs that the group feels could be spared colposcopy without any loss of sensitivity or specificity. The approach would reduce medical costs.

p16 in Liquid-Based Cytology, page 461.

A combined authorship from Germany and France has produced a very thoughtful study of p16 in cervical cytology, confirming its use in ThinPrep specimens, indicating that not only positivity, but degree of staining, together with nuclear morphology, might provide useful stratification in patients with abnormal Pap smears. This article will be of great help to us in our own study of p16 in ASC-H patients.

Atypical Urine Cytology in a Tertiary Care Center, page 468

Please read this article and see the pictures included. I have no doubt that it is a very useful article. However(!), I have never seen so many "howevers" in one article! Apparently, transitional cell clusters in voided urine are relevant clinically, particularly with a previous history of TCC and/or urinary calculi. However(!), there was no statistical correlation between cluster architecture and outcome. If you find a clear message in this article, please let me know what it is.

Applications of FISH in Cytopathology: a review, page 587

For anyone still wondering about the basics of FISH, but afraid to ask, this review by Oliveira et al from Harvard and Mayo Clinic will tell you everything you need to know, in four easy pages.

Papanicolaou Stain-Induced Fluorescence in TB Diagnosis, page 600.

This study from Nagpur, India, mirrors (but does not quote !) a SAIMR study we did on the demonstration of tuberculosis organisms by autofluorescence in FNA. These authors confirm our findings that Cytology, with fluorescent microscopy of the original Papanicolaou slide, without rhodamine or auramine, will permit visualization by fluorescence of the organisms in extra- pulmonary TB. It is superior to, and quicker than, conventional ZN methods, and provides ready identification within hours of FNA.

Glut-1 in Serous Fluids, page 621.

This paper by Afify et al from UC-Davis confirms findings we made four years ago, and also doesn't quote us! Identification of adenocarcinoma cells in a sea of reactive mesothelial cells, and their distinction from mesothelioma, is the setting in which Glut-1 may be very useful. Apart from adenocarcinomas of breast origin (low sensitivity), Glut-1 had excellent sensitivity in ovarian (100%) & lung (91%) tumors, 72% overall. This compared with CEA(74%) Ber-Ep4(85%) and B72.3(62%).

Normal Parathyroid Gland Cells, page 627. ***NEW***

This is apparently the first description of normal parathyroid cytomorphology, studied in 47 patients at the University of Zagreb. The authors describe light and dark chief cells and their differences in Papanicolaou and Giemsa staining. The majority of samples contained both. Oxyphilic cells were seen in 17% cases. (Please don't tell anyone about this, otherwise we will be invited to do intraoperative assessments of parathyroid presence in thyroidectomy specimens, which we would prefer not to do!)

100% Rapid Rescreening for QA, page 639.

For those labs using conventional smears, who do not have access to automated screening devices, it would appear that 100% rapid rescreening is better than 10% full rescreening currently imposed. In this study of almost 3,000 cases from the Lutz Institute, Portugal, 100% rapid rescreening yielded 386 (13%) cases, including 35 HSILs and 2 SCCs. In comparison, random 10% rescreening of these cases detected no lesions, but did find five unsatisfactory samples.

I commend to all who are interested in more than just morphology to read the Introduction/Editorial by Carlos Bedrossian, on page 289. It heralds a new section in the journal, concentrating on Geographic Cytopathology. It is a broad ranging, punchy overview of Cytopathology since the time of Papanicolaou, but without the dry historical data with which so many others have peppered the literature. Dr. Bedrossian indicates how 1988 was a watershed year in Cytopathology, launching not only the many ramifications of the infamous Wall Street Journal article on Pap mills, but also spurring the introduction of new technologies, new nomenclature system, more litigation, and other issues which have completely altered the face of cytopathology in the last 15 years. Dr. Bedrossian then goes on to wonder whether we have not gone too molecular, too regulation-based, too litigation conscious at the expense of more clinical, social, experimental, etiologic and other types of creative endeavors. He deplores the overused clichés, the copycat study designs, and the near identical stereotypical observations in the published literature and says "If imitation is the grandest form of flattery, one can almost hear the cacophony of patting each other on the back, drowning the music of scientific originality"! He makes a leap from there to the subject of the new Geographic Cytopathology segment and introduces the first articles on the subject which, sadly in my view, really do not live up to the fanfare which he composed. The journal further contains a number of research articles on serous effusions under the title of "The Norwegian Radium Hospital Symposium Issue 2005". Of these, I found one interesting.

ß-catenin in Mesothelial Proliferation and Mesothelioma, page 320. ***NEW*** in Cytopathology

Carlos Bedrossian and Claire Michael investigated ß-catenin staining in 54 serous effusions and 11 pleural biopsies, in which the diagnostic decision was mesothelioma, mesothelioma in-situ and reactive mesothelial cells. A distinct membranous or sub-membranous staining pattern was seen in all 23 cases of normal and reactive mesothelium. In contrast, reduced membranous staining with markedly increased cytoplasmic and nuclear staining was seen in 33 of 42 mesotheliomas and mesothelioma in-situ cases. ß-catenin is obtainable from Biogenex, and may be useful in this particular diagnostic area.

Thyroglossal Duct Cyst, page 365 (?back to the past).

There has been a trend recently to publish series of articles on conditions which I thought had been fully evaluated in previous literature, to the point of being unquestioned textbook fare. Authors from the Johns Hopkins Hospital, Baltimore, have analyzed 26 cases of thyroglossal duct cyst. They showed a diagnostic sensitivity of only 62% and a PPV of only 69% for the diagnosis of TDC. Abundant colloid, most often with ciliated columnar epithelium, was the predominant cytopathologic finding, but metaplastic squamous or mature squamous cells could occur. Thyroid epithelium is only rarely identified. Differential diagnosis includes lymphoepithelial cyst, thyroid gland lesions and various lymphadenopathies.

Paired Ductal Lavage (DL) and FNA in Breast Cancer, page 370.

Britt-Marie Ljung is one of the authors of this multi-institutional study of paired DL and FNA samples from six patients with breast malignancy. This is an exact replica of the study we did with Marie Wood and Mary Stanley. In the current publication, FNA diagnosed malignancy in all, while two DLs showed marked atypia, one mild atypia, and two were benign. The article has a very figure of all six cases in which side-by-side images of cells from the DL and the FNA are shown. In the four cases called marked or mild atypia, it is obvious looking at the pictures that the cells seen in the DL are in fact malignant cells, but the call was not made due to scarce cell groups, degeneration, or very good differentiation. In this, the study differs from our own in which only one of the cases had malignant cells in the DL, even after review with the sections of the breast at hand. The article concludes that the interpretation of DL samples is more challenging that that of FNA material. The authors state that to increase the sensitivity, the number of epithelial cells required for a diagnosis should be higher than that previously set. However, looking at the pictures, it would seem to me that the diagnostic criteria in DL might need down-grading, as the categorization "mild or moderate atypia" is, certainly in their pictures, an underdiagnosis of malignancy.

Her-2/Neu by Chromogenic ISH in Cytology, page 376. ***NEW in Cytopathology***

If only we received more breast samples, and more fluids from breast patients, this is something beautifully geared towards Cytopathology. It is an in-situ hybridization (ISH) study, but one which is labelled by a chromogen for light microscopy, as opposed to fluorescence for fluorescent microscopy. As such, it is cheaper, quicker and easier to interpret than routine FISH. The authors studied 25 cases of FNAs and touch imprints from infiltrating duct carcinomas by both CISH and FISH for the presence of Her-2 DNA. Of these, three were amplified and twenty non-amplified. Pictures of the cases show apparently easy distinction between non-amplification and amplification. This might be a very useful technique to study on archival cytologic material.

Gastric and Duodenal Epithelium and B72.3, page 381.

Martha Pitman from Harvard usually writes the most outstanding, useful and novel articles, but I'm afraid I don't "get" this one. She and Dave Wilbur have studied brushings from the gastric and duodenal mucosa of 14 Whipple specimens and have stained these with B72.3. The purpose was to document the morphology of normal gastric and duodenal epithelium (which I thought was sorted out decades ago) to permit recognition as background material in endoscopic ultrasound-guided pancreatic FNAs. Perhaps they had a case in which someone was lead astray by profuse gastro-duodenal epithelium. Further, they have demonstrated that goblet cells stain positively for B72.3. Their pictures are good.

The Thin Layer CellSlide Device, page 387.

A firm in Florence, Italy called Menarini Diagnostics has produced a device (which looks very small in the pictures) called the CellSlide, the purpose of which is to create a liquid-based sample. Their early studies indicate that their liquid-based samples are less sensitive but more specific than conventional slides.

Lymphangioma of Posterior Mediastinum, page 412.

This single case report on non-diagnostic FNA from a lymphangioma of the posterior mediastinum in an adult, is mentioned only because it comes from Coffing and Gutmann at Dartmouth.

Resolving ASCUS by Manual Reprocessing, page 434.

This group under John Maksem has used their particular manual liquid-based method of reprocessing and showed that about two thirds of diagnoses of ASC could be resolved to negative or to LSIL. They quote Daniel Hoerl having proven a similar degree of resolution using reprocessing by ThinPrep. They regard this as more cost beneficial than sending the residual sample for HPV. It is something we should discuss!

Toluidine Positivity in Chromophobe Renal Cell Carcinoma ***NEW***

This is a completely original observation, as far as I can tell. It emanates from the Tata Memorial Hospital in Mumbai, India and is written by Drs. Tanuja Shet and S. Desai. They observed, both in FNAs and in frozen sections stained with Toluidine Blue, very dramatic positive pink staining granules filling the cytoplasm of cells from chromophobe renal carcinomas. Conventional renal carcinomas, oncocytomas, and angiomyolipomas did not stain. In addition, one eosinophilic variant of chromophobe renal carcinoma did not stain. They attempted to repeat this finding in paraffin sections, but found that sections from chromophobe renal carcinomas were negative. They feel that Toluidine Blue positivity was labile and lost with tissue processing, dehydration, and even in formalin fixation. Their pictures are very dramatic and this may be a highly useful observation for a wary cytopathologist or frozen section pathologist on site, as well as a nephron-saving observation for the patient concerned.

FNA Liver

For anyone who missed Dr. Aileen Wee's lecture, she has published a full article on the algorithmic approach to current diagnosis of hepatocellular carcinomas online in CytoJournal. Her text is clear and her pictures are excellent.

Computerized Nuclear Morphometry in Thyroid Follicular Neoplasms

Contrary to previously published data, authors from Taiwan have concluded in a study of 36 cases of follicular cancer and 36 cases of follicular adenoma that there is indeed a statistically significant difference in the following parameters between carcinoma and adenoma: mean nuclear area, mean nuclear perimeter, largest to smallest diameter ratio, coefficient of variation of the nuclear area. This will always be a research tool.

ASCUS by Screening, Downgraded to Benign Reactive by the Pathologist ***NEW***

The authorship of this article from Allegheny General in Pittsburgh includes Stephen Raab. Cases interpreted by technologists as ASCUS/AGUS, downgraded to benign by pathologists, have not been studied. This study finds the % of downgraded cases compared with all ASCUS/AGUS cases per pathologist ranged from 4.8% to 43.7%. When followed up over time, these downgraded cases had the same SIL rates as cases of repair, ranging from 8-11%, seemingly justifying the downgrading. The authors suggest that this rate of ASCUS/AGUS cases downgraded by the pathologist might be used to monitor cytotechnologists and pathologists as a measure of laboratory quality.

Cytologic Grade in Pancreatic FNA

This article from Dr. Chhieng et al in Alabama evaluated group architecture, single cells, nuclear grade, mucus, bizarre cells and necrosis in 116 patients, designating the grade as low or high. Six-month survival was 12 vs 6 mths respectively. They feel their grading system is useful, particularly in those patients who do not undergo surgery.

Epithelial Displacement in Breast Lesions ***NEW association***

The displacement of epithelial cells by procedural investigations of the breast was initially documented years ago by Paul Peter Rosen. Authors from Mount Sinai Medical Center in New York have reviewed ten years of cases, yielding 53 with epithelial displacement. Of note is that with the exception of three cases, DE was associated with one or more underlying papillary lesions, including pure intraductal papilloma, DCIS involving intraductal papilloma, micropapillary DCIS, and invasive carcinoma of papillary type. The other three cases involved mucinous carcinoma and cystic lesions.

It is ten years since the first observation was made on the occurrence of primary non-Hodgkin's large cell lymphomas in serous cavities, without any other apparent involvement of the lymphoreticular system. This entirely new entity was first described in 1995, and its delineation, with KSHV or HHV-8 details, was summarized in the above Advances in 2006 by Jonathan Said.

HPV negative LSIL: An extension of the ALTS trial, page 253.

LSIL cases without confirmatory HPV positivity from the ALTS trial (11% of all LSIL) were re-evaluated. Follow-up parameters indicate conversion to HPV positive status in up to 30% of cases, and development H-SIL in some cases up to three years later. Authors for the ALTS group thus do not regard HPV-neg LSIL as a distinct entity, but rather the result of cytologic misinterpretation or false negative HPV tests.

Multiple-slide-blinded review of Pap slides in the context of litigation, page 263.

Andrew Renshaw and Blair Holladay have summarized the blinded review program of the Center for Cytopathology and Molecular Research at the Medical University of South Carolina, from 1998-2004. A total of 135 litigation slides were studied. Each litigation slide was "secreted" among 59 other validated cases, the resulting set of 60 cases being reviewed by cytotechnologists; cytopathologists are not involved in the re- screening. The methodology is meant to recapitulate usual practice as far as possible. The results are extremely interesting; only 10% of litigation cases are regularly and reliably identified as abnormal by the above method. In contrast, if a litigation case is reviewed as a single case by an expert cytologist, it will be reclassified as abnormal in 56% of cases. Reliance on a single review by a so-called expert will inevitably over- estimate by a large margin (more than four times greater) the number of litigation cases they state "should have been reliably identified by the original cytotechnologist". [Not stated in the article is the fact that this methodology is strongly advocated for any defendant lab facing litigation. Nor that the exercise is very expensive, costing over $13,000; this, however, would be far less than an unfavorable judgment. On the other hand, the plaintiff would do better with a single expert reviewer!].

Ampullary tumors encountered in EUS-FNA, page 289.

This is apparently the first series (35 patients) in which EUS- FNA targeted lesions ampulla of Vater. The sensitivity, specificity, PPV and NPV were 82%, 100%, 100% and 77%, with a diagnostic accuracy of 89%. Usual cytologic criteria were applied; the images suggest some cases are extremely bizarre and readily recognized, whilst others are very well-differentiated, presenting the usual challenges of recognition.

FNA of IPMT, page 298.

This study from Barcelona will be reviewed by Michael Harvey at the next Journal Club. It is a series of 11 cases of IPMT (with unfortunately poor color photographs), which outlines the difficulties which may be encountered in cytodiagnosis.

HepPar-1 in Barrett's esophagus, page 304.

This study confirms a recent finding from histopathology, now in cytologic preparations - i.e. that HepPar-1 is a moderately sensitive and highly specific immuno- marker of intestinal metaplasia in Barrett's esophagus. Conversely, it does not stain cardiac-type metaplasia, which apparently has no malignant potential.

p63 and TTF-1 in lung cancers, pages 223.

The take-home message of this paper from Mount Sinai in New York is that the addition of p63 to TTF-1 as a panel for primary vs metastatic lung tumors in both the lung and other sites may be extremely useful, to give some confirmation of poorly differentiated squamous-cell carcinomas of the lung, which form a significant proportion of the total of lung cancers, and which do not stain with TTF-1.

Ultrasound-assisted transthoracic biopsy; cells or sections?, page 233.

This large study of 97 consecutive patients, from Tygerberg Hospital in South Africa, includes Colleen Wright in the authorship. The authors recruited 97 consecutive patients with thoracic tumors involving the chest wall, pleura, peripheral lung or anterior mediastinum. All had ultrasound (US)-guided FNA, and the majority were considered suitable for safe cutting needle biopsy as well. Adequacy was achieved in 81% of FNAs, 80% of cutting needle biopsies, with a combined yield of 90%. In terms of definitive final diagnosis, FNA performed best for epithelial pulmonary tumors, whereas cutting biopsy was more often required for sarcomas, lymphomas and other tumors of the chest wall requiring sections/ancillary investigations. The authors comment that US is often forgotten in the workup of thoracic tumors, but represents a very low tech, low cost method for those lesions involving chest wall. If intervening lung tissue is present between the mass and the skin surface, US will not be of value.

FNA of large-cell neuroendocrine carcinoma of the lung, page 238.

I approach every paper on large-cell neuroendocrine carcinoma (LC-NEC)as the one that is finally going to permit me to make this diagnosis with assurance. This study of 11 cases from four departments in Israel is unfortunately not the one! This is possibly due to very poor black and white photographs, as well as the fact that cytological features are promised in Table II; however that table turns out to show immunocytochemistry. LC-NEC is a distinct clinical entity which still has the capacity to create confusion, both in cytodiagnosis and surgical pathology.

FISH for Her-2/neu in archival FNA specimens, page 471.

Our own prospective study indicated that FISH on FNA samples correlated well with FISH on biopsies of breast cancers, and our article is referenced here. This paper from the University of California at Irvine goes one step further by performing Her-2 FISH on archival, destained, alcohol-fixed FNA slides of breast cancer. Their methodology is provided in detail and seems standard. The authors review seven previous studies in which FISH was performed on cytologic preparations and conclude that previously stained FNA smears are suitable targets for FISH analysis, with good correlation to histopathology. [Not discussed is the possibility that this might also be valuable in destained slides from metastases, eg fluids, lymph node aspirates, in which we are often asked to evaluate Her-2 status. Such an extension study would make a good, quick study for an interested resident].

Cystic salivary gland FNA: analysis of 21 cases, page 489.

This study from Long Island Jewish Medical Center specifically targets the challenges existing in correct interpretation of salivary cysts. The pitfalls of cytodiagnosis of cystic salivary lesions are outlined. The topic is not new, but another paper from a good institution implies that problems remain in this area. One conclusion worth a cautionary note: whenever mucus is seen, unless there is very obvious evidence of a specific benign entity, the referring clinician should be made aware that low grade muco-epidermoid ca is a possibility to be excluded by other means.

Conventional and liquid-based (Autocyte) cytology in low-resource settings in Latin America, page 500.

This article may be of interest to the South African readers, who are often asked whether liquid-based cytology would be of value in that low resource setting. In this multi-site study, conventional Paps had higher sensitivity for high grade lesions, Autocyte had higher sensitivity for ASCUS and low grade lesions. [I think studies and figures such as these are so important. I believe the findings are inevitable, as low resource communities are essentially previously unscreened, and their lesions therefore of large size, with exfoliation of numerous cells into the sample. Liquid-based cytology on the other hand, reaches its greatest potential in previously screened communities, in whom new lesions are likely to be small, exfoliating few cells, which will be more easily missed on conventional smears. Where ASCUS/LSIL are unlikely to prompt colposcopy, there seems little point in identifying more of them by expensive means.

Balloon cell melanoma, page 543.

The recognition, or failure thereof, of balloon cell melanoma, is the potential nightmare of all in Cytopathology! This case report is of balloon-cell melanoma metastatic to neck nodes, without prior history. The cells are mononuclear, binucleated or multinucleated, with highly vacuolated or ballooned cytoplasm. Intranuclear inclusions are seen, but honestly, the cells resemble renal cancer more than melanoma!

From the Archives - why don't we do this more often?

TTF-1, HepPar, pCEA, CK7 & 20 are all used to distinguish adenocarcinoma from hepatocellular carcinoma (HCC) in liver FNA. But the distinction between benign and malignant liver groupings cannot be made in this way and may present a real problem. From fairly old literature, way back in Acta Cytologica 1995;39;596-8, Gagliano proposed the use of ordinary non-immuno, inexpensive reticulin staining of FNA slides in which the issue of concern is regenerating nodule vs HCC. The former would contain reticulin, the latter would not. This finding was confirmed in later work from Boston. Why have we forgotten about it? CD34 and Factor VIII immunostaining are proposed in this situation, but add nothing to good old-fashioned histochemical staining for reticulin.

FNA of hepatoblastoma, pages 355-364.

This article is from the Cytopathology Department I always visit in New Delhi at the All India Institute for Medical Sciences (AIIMS). The authors are Venka Iyer, Kusum Kapila and Kusum Verma, who have analyzed the largest series of hepatoblastomas (HB) to date, comprising 26 cases. They found trabecular arrangements and acinar formation in all, and extramedullary hemopoiesis in 20 cases. They claim to recognize fetal subtypes (abundant cytoplasm, small rounded nucleus, single central nucleolus), and to distinguish it from embryonal subtypes (scant cytoplasm, pleomorphic nucleus, two to four angulated nucleoli, and mitoses). When the two patterns were combined in a single tumor, both were recognizable. None of the cases showed bile pigment or the prominent eosinophilic central macronucleoli, which would be seen in hepatocellular carcinoma. Only one HB demonstrated intranuclear inclusions, and only one showed sinusoidal wrapping. At their institution, patients with HB are given chemotherapy on the basis of FNA, serum AFP and supportive radiology, without initial surgical sampling.

Nucle**OLAR** positivity for CD20 in T-cell lymphomas, pages 365-372.

T-cell lymphomas are far less frequent than B-cell lymphomas, except in the lymphoblastic category, in which 50-80% are derived from T-cells. This is an initial observation by Dilip Das of Kuwait University, on the unexpected and previously undescribed staining of nucleoli of T-lymphoblastic lymphomas (2) and large T-cell lymphoma (1) by CD20. Although this was only seen in 3 of 25 T-cell lymphomas, the staining by CD20, when it occurred, was prominent and occurred in up to 90% of cells. Please note that this nucleolar staining is quite distinct from the expected CD3 cytoplasmic staining of T-cell lymphomas and CD20 cytoplasmic staining of B-cell lymphomas. The author has no definitive explanation for this observation, but suggests that as nucleolar genes code for ribosomal RNA, the staining reaction may reflect RNA protein assembly which takes place in the surrounding dense fibrillar component of the nucleolus. This finding needs corroboration by other studies, but may be helpful on occasion, when immunophenotyping is used to classify lymphomas.

Increase in incidence of peritoneal collagen balls over a 10 year period, pages 387-390.

To confirm an initial observation within their department, these authors reviewed peritoneal washings and ascitic fluid seen between 1993 and 2002, confirming their observation that the incidence of intralperitoneal collagen balls had increased from 3.2% to 28.5% over this period. No procedural modifications were made in specimen processing during this interval. Please note that this is not merely increased recognition, as all the fluid specimens were actually rescreened to obtain these figures. The significance of this finding is unknown at this time.

Reflex HPV testing and age, pages 194-198.

In this study from the University of Alabama, the authorship (which includes David Chhieng) reviewed the contribution of reflex HPV testing to the diagnosis of CIN II-III on the basis of age. Their findings were that of 246 patients with CIN II-III on biopsy, 15% reached biopsy on the basis of positive reflex HPV testing after a cytodiagnosis of ASCUS. The remaining 85% reached biopsy by cytodiagnosis of SIL and ASCUS-H. They found no significant difference in the proportion of CIN II-III detected through reflex HPV testing dependent on age.

Benign endometrial cells in women >40, pages 207-216.

In this study from the University of Minnesota, a year-long analysis of almost 30,000 women >40 with Pap tests showing benign endometrial cells (866), was compared with a control group (597) of women >40, who had Pap tests without endometrial cells, but who had endometrial sampling for other reasons. The current study found the study group did not have more endometrial hyperplasias or malignancies than the control group. In fact, there were only 9 endometrial hyperplasias, and no adenocarcinomas, in 159 women with endometrial cells. Findings such as this could be used to influence changes in Bethesda. The authors stated that, on occasion, reporting endometrial cells may lead to the diagnosis of an otherwise occult malignancy; however, the added cost to the medical system, elevated patient anxiety in all the patients who do not have any endometrial abnormality, and even the occasional potential complication of the endometrial biopsy or curettage procedure itself, should be weighed carefully against this slight possibility.

Small papillary thyroid cancer, no longer occult, pages 217-219.

In this paper from the Baptist Hospital, Miami, Andrew Renshaw reviewed thyroid resection specimens obtained over a 65 month period, correlating the findings with cytologic information. In 720 resections performed, 210 papillary carcinomas were identified; the mean size decreased from 2.1 cm to 1.7 cm over the period of the study. The percentage of tumors measuring less than 1.0 cm increased from 21% to 39% (highly statistically significant). The vast majority of these small tumors were identified by prior FNA. Therefore, the finding of papillary carcinomas measuring less than 1.0 cm in diameter can no longer be called clinically insignificant or occult.

Core needle biopsy and FNA of musculoskeletal lesions, pages 229-239.

The authorship of this paper from Lund, Sweden, includes Mans Akerman, inarguably the doyen of soft tissue FNA. They investigated 130 patients by FNA and core needle biopsy performed by the cytopathologist, and found that the combination of modalities correctly identified 77/78 malignant lesions, and 50/52 benign lesions. Tumor subtype was determined in 77% and malignant grade determined in 90% of primary soft tissue tumors. They concluded that obtaining FNA as well as core needle biopsy at the same clinic visit by the cytopathologist made preliminary diagnosis on the day of referral possible and increased the number of correct diagnoses, enabling subtyping and grading of sarcomas. This applied particularly to the heterogeneous group of spindle cell tumors, notoriously difficult to diagnose accurately on FNA. Initial FNA provides diagnostic information enabling immediate discussion of management with the patient; cores are evaluated 2 days after sampling, with immunohistochemistry, as determined by FNA. This approach has obviated the need for open surgical biopsy in the vast majority of patients. Very few centers have historically adopted FNA as a diagnostic modality to be acted upon in primary soft tissue tumors. This supposed added "death knell" for FNA is thus not unexpected, although it is a shock coming from Mans Akerman himself!

Editorial: Changing face of cervical screening, pages 63-64

Despite the title, this editorial, by Andrea Dawson of the Cleveland Clinic, concerns the introduction of the ThinPrep Imaging System (TIS) at her institution. The mechanism of the Imager test is very well known to us, but of interest is her documentation that at the end of 2004, 158 test systems had been placed in laboratories across the USA, involving around 1000 cytotechnologists. The Cleveland Clinic findings are improved turnaround time (reduced from 3-21 days down to 2-3 days), increased detection of particularly low-grade abnormalities, and improvement in laboratory false negative rates; she notes that in almost all cases, the correct diagnosis could have been rendered after review of the 22 "fields of view", although obviously manual screening of slides showing abnormalities was performed. Dr. Dawson raises the implications of widespread automated screening for proficiency testing.

FNA of clear cell sarcoma of the kidney (CCSK), pages 83-89.

This analysis of 10 FNAs from 8 patient with CCSK is again by Drs Iyer and Verma from AIIMS in New Delhi. Again, it is likely to be of more interest to the South African readers of this report than to the USA readers. Distinction of CCSK from Wilm's tumor (WT) is critical, as the former (the so-called bone seeking or bone metastasizing tumor) has a far worse prognosis, requires different chemotherapy, and is more often inoperable; it requires preoperative chemotherapy without initial surgery, and therefore needs precise FNA recognition. The authors describe a triphasic appearance for this tumor (as in WT). They identify cord cells, septal cells, and small pyknotic cells, the last being degenerative changes, identified in 9/10 FNAs. Stromal fragments (which also occur in WTs) were seen in 8/10 FNAs. Those dealing with pediatric tumors are well advised to read this article. Unfortunately, the pictures are in black and white, but the text is well written and the cytologic appearance well described.

Anal cytology: Role for reflex HPV? Pages 152-156.

In a study of 158 patients, Ann Walts and associates from Cedars Sinai in LA found very similar percentages of HPV positive cases in anal ASCUS, LSIL and HSIL as were originally obtained in the ALTS trial study for comparable cervical diagnoses. Their findings suggest that reflex HPV testing might be helpful in triaging patients with anal ASCUS. However, as cytology frequently underestimates the final diagnosis in anal lesions, ASCUS HPV positive cases should be sent directly to anoscopy, as should all LSILs and above. ASCUS HPV negative cases could be followed, provided it is realized that these lesions occur in immune-deficient patients who may progress rapidly from minor to major anal lesions. (Our local Anal SIL study had similar HPV findings, but we had no subsequent verification by histopathology in our cases)

Cytopathologic analysis of paraspinal masses, pages 157-161

Dr S Z Ali of Johns Hopkins, and associates, analyzed 59 cases of paraspinal mass aspirates. The paraspinal space is a loosely defined anatomic zone extending from the intraspinal contents to paravertebral soft tissues. Paraspinal masses are uncommon lesions, presenting a wide spectrum of diagnoses. However, their potential for permanent damage to the spinal cord makes them an important, often emergency diagnosis, in cytopathology. In this particular study, metastatic adenocarcinomas (9), non-small cell carcinomas (NOS) (9), and lymphomas/myelomas (7), were the most common of the malignant lesions. However, non-neoplastic lesions were encountered, including chronic inflammation (NOS), blastomycosis, mycobacteria, and abscess.

Trichomonas vaginalis immunoreactivity for p16, pages 210-213

This is such a neat study! Liron Pantanowitz, previously from the SAIMR, currently at Tufts, has written on a previously unrecognized phenomenon - the potential for trichomonads to stain with p16. p16 is emerging as a useful biomarker for cervical dysplasia and malignancy. (Its particular potential in the category ASCUS-H is to be investigated in our department by Michael Harvey in coming months). While evaluating p16 at Tufts, an observation was made On a particular case that Trichomonas vaginalis stained positively. The authors then undertook an investigation of 10 consecutive satisfactory liquid-based Papanicolaou tests, diagnosed as negative for SIL or malignancy, but in which Trichomonas vaginalis was present. Trichomonads were consistently p16 positive. This must be recognized when studying p16 in small-celled dysplasias and neoplasias, so as not to make false positive calls. The images show that even the flagella stained positively in well-preserved samples!

From the Archives - Whatever happened to?

Cytopathology (UK) 1995, Vol 6, pp149-5. Chorionic villi in cervical smears

Ten years ago, a group from St Mary's in Manchester described and illustrated full chorionic villi in post-partum cervical smears taken 5-8 weeks after delivery. They were often seen in association with trophoblastic cells. Has anyone ever seen one?

Urine and serum cytology and molecular biology for BK virus nephropathy in renal transplant patients, page 235.

After a fabulous historic treatise by Leopold Koss on "decoy cells" (page 233), the above lead article from University of East Piedmont, Italy, reports on the use of 328 urines, 53 serum samples and 24 renal biopsies from 106 renal transplant patients. Polyoma virus (PV) infection is frequent in transplant patients. But PV nephropathy occurred only in patients with BKV viremia and a high number and frequency of decoy cells in the urine. In the absence of decoy cells, PV nephritis could be excluded. Thus, in transplant patients, molecular biology on either urine or serum to confirm PV nephritis need only be done if decoy cells are present in the urine!

Collagen balls in cervical smears, page 262.

A group at Mount Sinai School of Medicine in New York describes seven women with collagen balls, not in peritoneal washings where they were first described, and where we have learned to identify them, but in cervical smears. The structures looked identical to those we recognize in fluid, being three dimensional hyalinized cores, covered by benign-appearing cells. They did not correlate with any current Pap smear abnormality and were thought to arise on the surface of the ovary, being transported to the cervical smear along the fallopian tube and through the endometrial cavity.

Proficiency testing, editorials by William J. Frable and Andrew A. Renshaw.

These insightful editorials look at the method by which Cytopathology has arrived at the point being the first medical specialty whose regulatory activity has been assumed by the federal government. The consequences of this move for cytopathologists and other medical professionals is discussed in detail.

Angiosarcoma after breast conserving therapy.

This is the seventh publication on FNA of angiosarcoma of the breast skin, an issue which has been presented to the dept previously. The development of highly aggressive vasoformative lesions of the skin of the breast after breast conserving therapy and radiation is rare. This is a single case report, beautifully illustrated by Papanicolaou and immunochemical stains, and substantiated by histology. As expected, the low power impression is that of a spindle-shaped, vascular-based lesion, but at higher magnification, the cells are unusually rounded and "epithelioid", thus closely mimicking carcinoma, melanoma and other epithelioid soft tissue tumors. The authors tabulate the six prior publications on both classic and epithelioid variants of angiosarcoma. The earliest report dates from 1989 and not all cases were correlated at the time to prior radiation therapy. The average age of patients was 66 years, but three have been aged 50-53, contradicting the prior thought that this uncommon effect of radiation affected women over 60 only.

Proficiency in Cytology, page 1.

In this journal, Diane Davey has authored the editorial on the recently introduced proficiency programs, urging that organizations with more experience, such as CAP and ASCP, take over this federally mandated exercise.

Small cell tumors of the liver: a study of 91 cases, page 8.

Dr. Elhosseiny is part of the authorship of this article from the Departments of Cytology in Jacksonville, Florida and Burlington, Vermont. By far the majority of small-celled tumors encountered were small cell undifferentiated carcinomas and neuroendocrine carcinomas. Ten other tumor types were included in this paper, which is illustrated in color, and which advises on immunocytochemical panels.

Diagnostic accuracy of imprint cytology in Hodgkin's disease, page 20.

This article contains magnificent pictures of Hodgkin's disease (two of which appear on the cover), as well as a summary of the immunophenotyping of Hodgkin's cells.

Sampling of adrenal glands by endoscopic ultrasound FNA, page 26.

This paper described the author's experience in the aspiration of 24 cases of adrenal gland masses, most encountered in the course of workup of patients for metastatic carcinoma of other sites. Surprisingly, cortical adenomas (19) far outnumbered metastatic carcinomas (3). The article is a multi-institutional article from the University of Virginia and four hospitals in the Minneapolis region. Sadly, the few color pictures which are included are of not of good quality. Apparently, the left adrenal particularly is easily accessed during endoscopic ultrasound-guided procedures. I wonder if Dr. Zubarik would be interested in this?

Endometriosis of the kidney, diagnosis by FNA, page 60.

There is a letter to the Editor containing what appears to be the first FNA report of the diagnosis of endometrioma in a kidney, not a frequent occurrence, and not one which would immediately spring to mind when looking at benign-appearing columnar cells admixed with benign-appearing tubular cells! Once again, the illustrations are not good, but the letter does bring an unusual diagnosis to light.

Significance of E-Cadherin, N-Cadherin and CD44 in serous effusions using liquid based cytology.

It is very unreliable to distinguish reactive mesothelial cells, malignant mesothelioma and metastatic adenocarcinoma based on morphology alone. Cadherin cell adhesion molecules are tumour suppressor genes. Loss of cadherin function is associated with more aggressive tumour behaviour. E-cadherin and N-cadherin show specific expression for epithelial and mesenchymal tissue respectively. CD44 plays a role in cell matrix adhesion. The aim of this study was to assess the diagnostic utility of these markers in serous effusions. A total of 120 serous effusions were obtained ( 58 pleural, 62 peritoneal) from 22 reactive mesotheliums, 6 malignant mesotheliomas, 92 metastatic adenocarcinoma (15 lung, 15 breast, 9 stomach, 6 pancreas, 5 colon, 40 ovary, 2 urianry bladder). All cases were immunostained with monoclonal antibodies specific for E- cadherin, N-cadherin and CD44. 90% of metastatic adenocarcinoma expressed E-cadherin compared with 1/6 malignant mesothelioma and 1/22 reactive mesothelium. All 6 cases of malignant mesothelioma expressed N-cadherin whereas metastatic adenocarcinoma proved negative. CD44 immunoreactivity was seen in 81% benign effusions and in 23% metastatic adenocarcinoma. The combination of these 3 markers appears to be a useful panel:Pos E-cad and neg CD44- metastatic adenocarcinoma, Neg E-cad and pos CD44- mesothelial cells, N-cad positive in malignant mesothelioma

D Malle, R Valeri, C Photiou, K Kaplanis, C Andreadis, D Tsavdaridis, C Destouni. Acta Cytol 2005; 49: 11-16.

Fine needle aspiration cytology of endocrine neoplasms of the pancreas.

Most tumours of the pancreas are primary unresectable malignant neoplasms of exocrine origin and have a poor prognosis. In contrast, endocrine tumours are less commonand have a better prognosis, even when metastatses are present. The aim of this study was to analyze the role of FNA in the preoperative diagnosis of endocrine neoplasms. A total of 20 patients with cytology of pancreatic endocrine tumours were selected, 13 with biopsy and immunochemical confirmation and 7 with immunocytochemical confirmation alone. The main cytologic features were a predominantly single cell pattern with small poorly cohesive groups, intermediate to large cells with scanty cytoplasm, frequent binucleation, variable nuclear pleomorphism and finely granular distribution of chromatin. The differential diagnosis includes acinar cell carcinoma ( greater cellular cohesion and prominent nucleoli compared with endocrine tumours) and solid-pseudopapillary tumour (myxoid stroma, greater vascular component, dirty background and nuclear grooves compared with endocrine tumours. However, these tumours may be synaptophysin positive but chromamgranin A negative). FNA offers the possibility of a precise diagnosis. The cytologic features differ considerably from those of pancreatic adenocarcinoma. In difficult cases, immunocytochemistry is very helpful.

J Jimenez- Heffernan, B Vicandi, P Lopez- Ferrer, P Gonzalez- Peramato, A Perez- Campos, J Viguer. Acta Cytol 2004; 48: 295-301.

Extranodal presentation in patients with acquired immunodeficiency syndrome non-Hodgkin’s lymphoma (AIDS-NHL).

NHL is the second most common AIDS-associated malignancy. There is an increased risk of NHL in HIV patients. The pattern of presentation of NHL may differ between HIV positive and HIV negative patients. Patients may present with asymptomatic lymphadenpathy or extranodal disease. More than 80% of lymphomas in HIV positive patients are high grade B-cell lymphomas compared to 10-15% in HIV negative populations. Most of systemic AIDS-NHL are Burkitt’s or Burkitt-like and diffuse large B-cell NHL. Primary central nervous system lymphoma is a lymphoma arising within and confined to the brain, leptomeninges, eye and spinal cord. This entity is found in about 19% of patients with AIDS-NHL and is the most common brain tumour in AIDS patients. Systemic AIDS-NHL treatment involves modifications of conventional chemotherapy in association with antiretroviral therapy. The article provides 3 case studies of AIDS-NHS (primary central nervous system lymphoma, AIDS-NHL of the liver AIDS-NHL of the tongue).

P Barnardt. Extranodal presentation in patients with acquired immunodeficiency syndrome non-Hodgkin’s lymphoma (AIDS-NHL). The Southern African Journal of HIV Medicine. March 2005, p37-40