Printer Friendly

The utility of UroVysion fluorescence in situ hybridization in pancreatic fine-needle aspiration samples directed and obtained by endoscopic ultrasonography.

Pancreatic cancer is the fourth leading cause of cancer-related death in the United States. In 2010, it was estimated that 6% of cancer deaths in males and 7% of cancer deaths in females were due to pancreatic cancer. The 5-year survival rate for all stages of pancreatic cancer is the lowest of all the major cancer sites. (1)

The management of a patient with a pancreatic lesion can be challenging. Owing to the location and the nature of the organ, it is difficult to obtain tissue samples. Physicians have increasingly used a number of techniques to obtain tissue samples for diagnosis. Endoscopy, percutaneous image-guided, fine-needle aspiration (FNA), brushings, washings, the use of a spy glass to obtain tissue samples, and endoscopic ultrasound-guided FNA (EUSFNA) are all utilized techniques. (2-6) Once the clinician obtains the sample, the pathology examination of the tissue can be just as challenging. Microscopic examination findings of the pancreas can be very difficult to interpret owing to extensive desmoplastic fibrosis associated with neoplasia, acute and chronic pancreatitis, and tissue necrosis. Since EUS-FNA can be used to obtain tissue and is generally felt to have fewer complications, such as pancreatitis and peritonitis, than large-core needle biopsies and wedge biopsies of the pancreas, this modality has gained increased use. (7)

Some studies have shown that the microscopic diagnostic sensitivity and specificity of samples obtained by EUS-FNA is from 64% to 94% and 71% to 95%, respectively. (8) Advanced cytologic and molecular markers have been described for patients with bile duct strictures that are felt to increase the sensitivity and specificity in diagnosing malignancy over routine cytology. Some articles have reported the use of UroVysion fluorescence in situ hybridization (UFISH) (Abbott Molecular Inc, Des Plaines, Illinois) to help confirm the diagnosis of malignancy in patients with biliary strictures. Luna et al (9) studied 233 consecutive patients undergoing endoscopic retrograde cholangiopancreatography for pancreatic strictures with cytology, digital image analysis (DIA), and UFISH. They reported that polysomy UFISH results had 100% specificity and considered UFISH polysomy to be equivalent to a positive cytology finding for the diagnosis of malignant pancreatobiliary strictures. They caution that their population of patients did have a high pretest probability of cancer and the performance of these tests in a screening patient population was unclear. Some of their cases exhibited trisomy for chromosome 7 and 3. Since trisomy 7 has been observed in both neoplastic and nonneoplastic conditions, they concluded that trisomy UFISH results in pancreatobiliary strictures must be interpreted with caution and placed into the clinical context.

Levy et al (10) also studied the use of intraductal ultrasonography (IDUS), DIA, and UFISH in patients with biliary strictures. They examined 86 patients with indeterminate biliary strictures and concluded that DIA, UFISH, and IDUS enhance the accuracy by allowing diagnosis of malignancy in a substantial number of patients with false-negative cytology and histology findings. They felt that their findings support the routine use of these modalities for patients with biliary strictures. Barr Fritcher et al (11) studied UFISH and DIA and concluded that a positive DIA result was not a significant independent predictor of carcinoma in the presence of UFISH and routine cytology on pancreatobiliary brushing specimens. They discontinued the use of DIA and performed routine cytology and UFISH for all patients being evaluated for pancreatobiliary malignancies. Other studies (12-14) have reported the utility and value of UFISH primarily in biliary brushings. Levy et al (15) published results of EUS FNA cytology with DIA and UFISH of 19 patients with pancreatic masses, combined with 23 other patients with lymphadenopathy and esophageal, gastric, or thyroid masses. Their series included patients with squamous cell carcinoma (esophagus), adenocarcinoma (stomach, pancreatic, esophagus), pancreatic mucinous cystic neoplasm, sarcoma, lung carcinoma, thyroid carcinoma, malignant gastrointestinal stromal tumor, melanoma, adenocarcinoma of unknown primary, and lymphoma. With these diverse lesions, they reported a sensitivity of 77% and a specificity of 100% with UFISH. To our knowledge, no one has reported the use of the UFISH test exclusively in pancreatic EUS-FNA specimens, using only UFISH.

Our center primarily serves referred patients and our gastroenterologists specialize in the evaluation of pancreatic lesions. We examined our experience with pancreatic specimens obtained by EUS-FNA during a 2-year period. The purpose of this study is to investigate if a positive UFISH test result on pancreatic FNA specimens would have the same implications in the diagnosis of pancreatic adenocarcinoma as a positive UFISH test result has for the diagnosis of adenocarcinoma in biliary strictures.

MATERIALS AND METHODS

Samples were obtained by clinicians using EUS-FNA. During the EUS-FNA procedures, the pathologist may have been asked to do a rapid evaluation of the specimen and triage the specimen for further analysis. When the pathologist was present, the material was directly placed on a sterile glass slide. Any small fragments of tissue were removed and placed in formalin. The liquid supernatant material on the slide was then poured into PerservCyt (Hologic, Bedford, Massachusetts). The remaining material on the slide was used for squash preparation. One slide was stained with modified Romanowsky stain (Three Step Stain, Richard-Allan Scientific, Kalamazoo, MI) for the rapid evaluation. The other slide was rapidly placed in alcohol for a Papanicolau staining, to be examined later. In addition, cytospins were made from the supernatant material in PerservCyt for examination.

If a pathologist was not available for immediate triage of the specimen, the entire material was placed in PreservCyt. The specimen was later examined for tissue fragments for a cell block, and cytospins were made from a portion of the fluid.

Samples for UFISH were most commonly obtained from the specimen placed in PreservCyt after blood clot, and small tissue samples were removed for cell block. Sometimes the atypical cells were only seen on the squash preparations and in these cases the smears were decolorized and then submitted for UFISH. This had the advantage of determining whether cytologically atypical cells also had UFISH chromosome abnormalities. In other cases, the atypical cells were only seen on the cell block. In these cases, the cell block was recut at 8 mm and then submitted for the UFISH analysis. Sixty FNA cases were submitted for UFISH. In 46 cases, UFISH was done on the specimen placed in PreservCyt, 7 were done from the cell block, and 4 were done on smears. Three of the 4 smears were destained Papanicolau stains that were originally fixed in alcohol. One of the 4 was on a destained air-dried modified Romanowsky stain preparation. Three samples did not show at least 50 epithelial cells and were insufficient for UFISH evaluation.

The specimens were examined by at least 1 of 4 different pathologists in our practice. Most cases were examined by more than 1 pathologist, particularly in difficult cases. We did not use a standardized reporting format, but all interpretations could be classified into 4 different categories:

1. Malignant: Cases with definitive carcinoma or material consistent with malignancy.

2. Suspicious for adenocarcinoma: Cases with significant cellular atypia suggestive of, but not conclusive for, malignancy.

3. Reactive: Cases with some atypia, but reactive changes were favored as the most likely cause.

4. Negative: Cases with definite reactive changes or negative for malignancy.

Our study only used samples from patients who had lesions that were negative, suggestive of, or diagnostic for pancreatic adenocarcinoma. Other lesions, such as intraductal papillary mucinous neoplasm, islet cell tumors, and other mucinous lesions, were found but the results were not included.

The UFISH test was run similarly as for the specimens collected from urine or the biliary tract. (14) The probe set consisted of directly labeled chromosome enumeration probes (CEPs) to the pericentromeric regions of chromosome 3 (CEP 3), 7 (CEP 7), and 17 (CEP 17), and to chromosomal band 9p21. The samples were manually counted. At least 50 epithelial cells were needed for a sample to be considered "adequate," but in most negative cases, more than 100 cells were examined if they were present. All samples were analyzed manually by a technologist, and target cells were saved and counted with the BioView Duet (BioView, Nes Ziona, Israel).

Our UFISH test was internally validated against known cases that were positive and negative for pancreatic adenocarcinoma. Our criteria were based on suggested criteria in the literature (9) for biliary specimens. A patient's specimen was considered "positive for polysomy" (and suspicious for malignancy) if 5 or more cells showed gains of 2 or more of the 4 probes. If there were 10 or more cells that showed 3 or more copies of chromosome 7 (or 3) and 2 or fewer copies of the 3 other probes, the specimen was said to be "positive for trisomy." Trisomy 7 or trisomy 3 has a higher cutoff because signal splitting can lead to false-positive trisomic signals being observed at low numbers even in normal specimens.

For patient follow-up, we examined our hospital electronic medical records on the cases that did not have definitive FNA results. The length of time from the original biopsy ranged from 10 months to 2 years and 7 months. Patient records were examined for imaging studies, CA19-9 laboratory tests, and follow-up visits with additional tissue biopsies. If patients had a subsequent biopsy from the pancreas or a metastatic site after the original biopsy, which was consistent with pancreatic adenocarcinoma, that patient was categorized as "biopsy-positive for pancreatic adenocarcinoma." If the imaging studies at the time of the original biopsy or after showed invasion into the portal vein or evidence of metastatic disease, the patient was categorized as "clinically positive for pancreatic malignancy." Patients with a pancreatic mass on endoscopic ultrasonography and a CA19-9 of greater than 5000 U/mL were also categorized as "clinically positive for pancreatic malignancy." If the patient had no mass and a clinical diagnosis of autoimmune, acute, or chronic pancreatitis, the patient was categorized as "pancreatitis." Fourteen patients did not have further follow-up. Six of 14 patients with no further follow-up did not have a mass described. Eight patients without further follow-up were noted to have pancreatic mass in the original endoscopic ultrasound report. We did not feel a mass lesion alone was sufficient to categorize these patients as having pancreatic adenocarcinoma. Reports have shown that mass lesions of the pancreas can be caused by diseases other than pancreatic adenocarcinoma, including chronic pancreatitis, autoimmune pancreatitis, and other pancreatic tumors. (16-18) All 14 patients were categorized as "no further follow-up."

RESULTS

In the 2-year period, 189 samples from EUS-FNA were submitted. Of the 189 EUS-FNA samples, 129 did not have UFISH testing. One hundred thirteen of the samples were clinically and ultrasonographically benign cysts or pseudocysts that were cytologically very hypocellular or acellular. Sixteen cases were positive for malignancy on FNA but were not submitted for UFISH, since additional confirmation of the lesion was not felt to be necessary. Sixty cases were submitted for UFISH. Three of these samples did not show at least 50 epithelial cells and were called "quantity not sufficient." One of these was in a "suspicious" case and 2 were in cases with negative FNA results.

Of the 57 EUS-FNA cases with test results that were submitted for UFISH, the cytologic diagnosis was matched with the UFISH diagnosis in Table 1. Cases submitted for UFISH were categorized as "positive" or "negative." Like urinary samples, we defined a positive pancreatic or biliary UFISH result as a significant number of cells with polysomy, trisomy, or loss of the 9p21 locus (0 gold). In samples from biliary strictures, false-positive cases due to trisomy 7 and trisomy 3 have been described in patients without cancer, usually in association with patients with primary sclerosing cholangitis. (9) No previous cases with a loss of the 9p21 locus have been described in the literature when using UFISH on pancreatic or biliary samples. In our clinical reports, the results stated the abnormality identified, rather than simply using "positive" or "negative." We felt it was important to give clinicians a description of the actual abnormality so they might have as much information as possible to assess the patient's condition.

Twenty-seven patients had positive FNA results; 25 of these were UFISH positive. All positive UFISH cases on pancreatic EUS-FNA with definitive (positive) FNA results showed polysomy (Figure 1, A through C). One of these polysomic cases also had trisomy 7 (Figure 2, A and B). The remainder of the pancreatic EUS-FNA cases that tested positively for UFISH were also polysomic, except for 1 positive case that occurred in a sample that was felt to be cytologically reactive. This case exhibited a "0-gold pattern" (Figure 3, A through C) without polysomy or trisomy.

In the cases analyzed for UFISH and obtained from EUS-FNA, the sensitivity of UFISH testing, when compared to the samples that had a definitive FNA diagnosis of adenocarcinoma, was 93% (25 of 27 cases). Review of the 2 UFISH false-negative cases showed 1 case that had a very limited number of cancer cells on the cell block specimen. Insufficient material was left in the cell block to run the UFISH test; therefore, the UFISH test was run on the PreservCyt supernatant. The supernatant was hypocellular, having only 66 cells, but 3 polysomic cells were present. The criteria for a positive sample were not met. We know that the tumor was polysomic, since this patient underwent a Whipple resection, and a sample was submitted for UFISH. The Whipple resection sample was positive for polysomy; therefore, this false negative was due to the limited number of tumor cells in the original sample. The second case was also felt to be due to limited numbers of tumor cells in the sample submitted for UFISH. The specificity of pancreatic EUS-FNA UFISH, when compared to the samples that had a definitive negative FNA diagnosis, was 100% (8 of 8 cases).

Table 2 shows the results for patients with FNA results suspicious for adenocarcinoma and the corresponding UFISH test results. For 9 patients with suspicious FNA results and a positive UFISH finding, 2 subsequently had a tissue diagnosis of pancreatic adenocarcinoma. Two patients had EUS imaging studies at the time of the FNA biopsy that showed evidence of portal vein invasion. One patient had a subsequent chest imaging study showing metastatic malignancy. The last 4 patients had imaging masses seen during the EUS procedure but further follow-up was not available. There were no patients with "suspicious for adenocarcinoma" FNA results and a positive UFISH finding that fit the above criteria for pancreatitis. Among the 4 patients who had UFISH-negative results, no patients fit the criteria for biopsy positive for pancreatic adenocarcinoma or clinically positive for pancreatic malignancy. Two patients had a clinical diagnosis of pancreatitis on clinical follow-up. There were 6 patients with no further follow-up. Four patients with no follow-up had a positive UFISH result and 2 patients had a negative UFISH result.

[FIGURE 1 OMITTED]

The follow-up information from hospital records for patients who had an FNA classified as reactive or negative is shown in Table 3. As mentioned previously, 1 patient had an FNA sample that was thought to have reactive features and was UFISH positive with a 0-gold pattern. This patient had a reported CA19-9 greater than 5000 U/mL, new-onset diabetes, and an imaging mass in the pancreas suggestive of malignancy. Clinically, this patient was thought to have either biliary or pancreatic adenocarcinoma. Three of the 16 patients with a negative UFISH finding on initial EUS-FNA had subsequent tissue diagnoses of pancreatic carcinoma. Two of these were retested with UFISH and the tumor results were positive. One patient was not retested with UFISH. Five patients had a diagnosis of pancreatitis. Eight patients had no follow-up.

If we assume that patients who had EUS-FNA positive results (16 positive without UFISH testing and 27 positive with UFISH testing), and patients who were later categorized as biopsy-positive for pancreatic adenocarcinoma (5 patients) or clinically positive for pancreatic malignancy (4 patients), are our true-positive patient group (52 patients), then FNA definitively identified 83% (43 of 52) of patients with pancreatic adenocarcinoma. UFISH in conjunction with FNA would have been able to increase the diagnostic sensitivity to 94% (49 of 52) of patients if polysomic and 0-gold UFISH results were considered positive. If we define the patients who had EUS-FNA negative results and patients with a diagnosis of pancreatitis as our true-negative group, then the specificity of UFISH is 100% (13 of 13). Fine-needle aspiration correctly identified as negative or reactive 85% (11 of 13) of true-negative patients but did not accurately classify 2 patients with suspicious FNA results who turned out to have clinical pancreatitis. Therefore, the diagnostic sensitivity and specificity of UFISH is greater than that of the examination of tissues in our study. Adding the UFISH test to the FNA results would have increased the true diagnostic sensitivity from 83% to 94%, considering polysomy and "0 gold" as positive UFISH test results, while increasing the specificity from 85% to 100% and helping to appropriately categorize patients with suspicious FNA results who had pancreatitis.

COMMENT

The evaluation of a patient with a pancreatic mass suggestive of adenocarcinoma can be difficult. EUS-FNA has become an important test in this process. Once pathologists receive the specimen, they are also faced with the challenge of providing the diagnosis, usually with a limited amount of tissue sample. A number of ductal lesions must be considered in the differential diagnosis, including normal ductal cells, reactive changes in ductal cells due to acute and chronic pancreatitis, autoimmune pancreatitis, intraductal papillary mucinous neoplasm, benign pancreatic ductal tumors, and ductal adenocarcinoma. In addition, other benign and malignant pancreatic lesions including endocrine tumors, solid-pseudopapillary tumor, bile duct carcinoma, and metastatic malignancies may be present.

[FIGURE 2 OMITTED]

Our series consisted of 76 patients with clinical findings suspicious for pancreatic adenocarcinoma. Sixteen patients did not have UFISH testing and 57 were successfully tested with UFISH. Three patient samples did not have sufficient cells present for UFISH evaluation. Eighteen percent of 76 patient samples had atypical cells suggestive of ductal adenocarcinoma. In these cases, an additional supporting test would have assisted us in making a definitive diagnosis. Since well-differentiated pancreatic adenocarcinomas may be difficult to distinguish from reactive pancreatic tissue, an adjuvant test would also be helpful in lesions suspected of being reactive or inflammatory when there is a significant clinical concern for malignancy. Our study shows that the UFISH test fits the criteria as an important adjuvant test to FNA in the diagnosis of pancreatic adenocarcinoma.

The true sensitivity and specificity of UFISH for patients with pancreatic adenocarcinoma was difficult to determine owing to a number of factors. One factor was the ability of the aspirator to sample the malignancy. In some patients with a malignancy, malignant cells are not obtained on the FNA. This can be due to technical difficulties with the aspirator in getting to the lesion, as well as the physical and biological characteristics of the tumor. In addition, some tumors may have a significant amount of surrounding inflammation, which may obscure the main lesion.

In our practice, most of our patients were referred to our hospital specifically for the biopsy of a pancreatic lesion, using the endoscopic ultrasound procedure. Once the procedure was done, they frequently went back to the referring physician and we did not have access to their further evaluations or care. We were able to obtain follow-up information on 17 of 31 patients (55%) with suspicious, reactive, or negative cytology results.

We were interested in determining the utility of UFISH as an aid in the diagnosis of pancreatic adenocarcinoma, particularly in cases for which the diagnosis is equivocal. In this scenario, we were most interested in the specificity of the test. We were able to assess this aspect of the test by comparing the UFISH results to cases with a definitive FNA result as the gold standard. On the basis of definitive FNA results, (ie, FNA results positive or negative for malignancy), our series of 57 patients who underwent EUS-FNA and had UFISH testing showed a 93% sensitivity and 100% specificity. These findings are similar to results described by other researchers on biliary brushings and washings associated with biliary strictures. (9-14)

Our patient follow-up studies helped us to look at the true sensitivity and specificity of UFISH as an adjuvant test to FNA. Of the 30 cases with FNA results that were suspicious, reactive, or negative, we feel we were able to definitively classify the condition of 7 patients as benign and of 9 as malignant, on the basis of our hospital medical records. For the remaining 14, even though 8 were thought to have a mass on EUS, we did not feel there were sufficient additional criteria to support a definitive diagnosis. With the use of patient follow-up information, FNA and UFISH were able to increase the sensitivity from 83% to 94% and increase the specificity from 85% to 100%. In 100% of patients with benign diagnoses, the UFISH test result was negative. Two of these cases involved patients who had FNA results that were suspicious for malignancy. UFISH was also able to support a diagnosis of malignancy in 6 of the 9 patients (67%) who had a malignancy but did not have definitive FNA results. Five of these cases were for patients with suspicious FNA results. One case involved a patient with FNA results thought to be reactive. For 7 patients with suspicious FNA results who had sufficient follow-up to be categorized as true-positive or true-negative cases, UFISH was 100% sensitive and 100% specific.

The sensitivity of combined UFISH and FNA procedure may actually be greater or less than 94% because we were not able to find patient follow-up information on 14 of our cases that had negative FNA results or did not have definitive FNA results. Since our patients were referred for the EUS procedure because they had symptoms suggesting pancreatic malignancy, this group of patients had a much higher probability of pancreatic malignancy when compared to a random population. In the 14 cases for which we did not have follow-up, we feel there were probably additional patients who had pancreatic cancer that was not detected by the FNA or UFISH procedure. We also feel that the negative UFISH and FNA result was due to insufficient numbers of malignant cells obtained by the aspiration biopsy procedure in a very high percentage of those cases. This belief is supported by the fact that all of the patients diagnosed as having pancreatic adenocarcinoma in our series, who originally had UFISH negative results, had positive results when UFISH was repeated with samples containing sufficient malignant cells. Our experience indicates that when a good tumor sample is obtained, a very high percentage of pancreatic adenocarcinomas have positive test results by UFISH (either polysomy, trisomy, or 0 gold). It is not inconceivable, however, that a small percentage of tumors may not be UFISH positive.

[FIGURE 3 OMITTED]

The follow-up data support the use of UFISH as a valuable adjuvant test to EUS-FNA in the diagnosis of pancreatic adenocarcinoma. It is particularly helpful in cases for which the FNA diagnosis of malignancy is not absolute and for which changes were present that were suggestive of, but not diagnostic, for malignancy. Levy et al (15) also demonstrated this in a series in which 5 of their patients had cytologic results that were interpreted as benign, atypical, or suspicious, but the DIA/UFISH results suggested malignancy. Follow-up of these patients showed all died owing to their malignancies.

A positive UFISH test result may give the examining physician an additional tool to support a definitive diagnosis for a patient with a cytological suspicious finding with worrisome clinical and imaging features. Other authors (9,10) have found that when only polysomic results were considered in biliary brushing or washing specimens, the specificity was 100%. Barr Fritcher et al (11) state that polysomic UFISH finding is indicative of carcinoma in the context of a biliary stricture. Our study indicates that a positive UFISH test result should at least suggest that the patient be considered for close follow-up and additional sampling. A negative UFISH test result, in contrast, may also prompt the search for other factors that may not have been previously considered, such as autoimmune pancreatitis or obstructing stones.

Of particular interest is the finding of trisomy and the loss of 9p21 in biliary and pancreatic specimens. Our case associated with trisomy on EUS-FNA involved chromosome 7 and occurred with polysomic cells. We did see "isolated" trisomy 7 with 2 other cases during the 2-year period but both of these cases involved biliary specimens. Trisomy has been correlated with malignancy in urine samples and biliary brushing or washing (BBW) specimens in some investigations. In the series by Luna et al, (9) the authors indicate that although trisomy 7 or trisomy 3 in BBW specimens was more frequently found in malignant cells, the lesion may also be found in some nonmalignant inflammatory lesions and should be interpreted with caution and placed into the clinical context. Our sole pancreatic sample of trisomy 7 was associated with polysomy and involved a patient who had malignancy. One of the 2 remaining "isolated trisomy" cases in BBW specimens involved a patient who eventually underwent an open excision and was diagnosed with malignancy. No definite diagnosis was obtained for the second isolated BBW trisomy case. All of these cases involved trisomy 7. No trisomy 3 cases were discovered.

Loss of 9p21 is sometimes called 0 gold owing to the absence of the 2 gold-colored probes. In urine samples, 0-gold specimens, like trisomy specimens, have a higher threshold number of cells for diagnosis and are felt to be consistent with a specimen with a positive finding. Twelve 0-gold cells are needed in urine samples to consider the sample positive for this marker. The loss of 9p21 has not been described, to our knowledge, in EUS-FNA or BBW samples. Taniai et al (19) reported that functional point mutations in the p16INK4a promoter region likely contribute to the initiation/progression of cholangiocarcinoma in primary sclerosing cholangitis. Ahrendt et al (20) described the loss of heterozygosity (9p21 loss) and promoter methylation of p16INK4a in primary sclerosing cholangitis-associated cholangiocarcinoma, which may result in gene inactivation. The mutations in these studies were not analyzed with UFISH, but they deal with the same chromosome region as that of the 9p21 UFISH probe. Our case with loss of the 9p21 UFISH probe (0 gold) involved a patient with findings that evoked a clinical suspicion for cholangiocarcinoma, and this patient was categorized in our study as clinically positive for pancreatic malignancy, by imaging studies and a markedly increased CA 19-9.

Further studies are necessary to see if this UFISH abnormality is more suggestive of cholangiocarcinoma than pancreatic adenocarcinoma.

CONCLUSION

Our study indicates that UroVysion FISH testing is an appropriate and useful adjuvant technique in the detection and diagnosis of pancreatic adenocarcinoma. We advocate the UFISH test as a supplement test to FNA, not as a replacement for it. Although UFISH appears to have higher sensitivity and specificity than FNA, we are not comfortable with making the diagnosis of pancreatic adenocarcinoma without at least having cytologically atypical cells suggestive of carcinoma. In particular, UFISH test results that are polysomic are highly specific for a pancreatic adenocarcinoma. Previous studies done on biliary strictures corroborate this finding. The high specificity of UFISH may help increase the ability of the pathologist to make a definitive diagnosis for patients who have clinical, imaging, and cytologic features that all suggest malignancy. A polysomic UFISH test result in an FNA sample with atypical cells is virtually diagnostic of malignancy. A negative UFISH result does not rule out malignancy but may prompt the search for other causes of a sonographic or radiographic mass. In our series, the most common cause for a negative UFISH result for a patient with pancreatic adenocarcinoma was insufficient numbers of malignant cells for the UFISH test. It should also be noted that polysomic UFISH finding is not specific for pancreatic adenocarcinoma and can be seen in a number of malignant lesions. (15) Another point to be stressed is that we have not seen a positive polysomic UFISH result in a patient who did not have pancreatic adenocarcinoma.

We think the test will have the greatest utility in cases for which the FNA diagnosis of malignancy is not absolute and changes are present that are suggestive of, but not diagnostic for, malignancy. Our studies show that UFISH increases the sensitivity and specificity for the detection of pancreatic adenocarcinoma and helps to appropriately categorize patients with FNA results that are suspicious for, but not diagnostic of, pancreatic adenocarcinoma. Since our sample size is somewhat small, we would recommend that additional studies be performed to validate our results and the clinical use of the test.

References

(1.) American Cancer Society. Cancer facts and figures 2010. http://www. cancer.orgacs/groups/content/@epidemiologysurveilance/documents/document/ acspc-026238.pdf. Accessed August 1, 2011.

(2.) Lightwood R, Reber HA, Way LW. The risk and accuracy of pancreatic biopsy. Am J Surg. 1976; 132(2):189-194.

(3.) Paulsen SD, Nghiem HV, Negussie E, et al. Clinical observations: evaluation of imaging-guided core biopsy of pancreatic masses. Am J Roentgenol. 2006; 187(3):769-772.

(4.) Henke AC, Jensen CS, Cohen MB. Cytologic diagnosis of adenocarcinoma in biliary and pancreatic duct brushings. Adv Anat Pathol. 2002; 9(5):301-308.

(5.) Roorda AK, Kupec JT, Sundaram U. I spy biliary and pancreatic ducts: The Spyglass single-operator peroral cholangiopancreatoscopy system. Pract Gastroenterol. 2009; 32(4):15-32.

(6.) Gress F, Gottlieb K, Sherman S, Lehman G. Endoscopic ultrasonography-guided fine-needle aspiration of suspected pancreatic cancer. Ann Intern Med. 2001; 134(6):459-464.

(7.) Geisinger KR, Stanley, MW, Raab SS, Silverman JF, Abati A. Pancreas, Extrahepatic BileDucts, and Gallbladder. In: Modern Cytopathology. Philadelphia, PA: Churchill Livingstone; 2004:543.

(8.) Noh KW, Wallace MB. EUS-FNA in diagnosis and staging of pancreatic adenocarcinoma. Medscape Gen Med. 2005; 7(2):15.

(9.) Luna LM, Kipp BR, Halling KC, et al. Advanced cytologic techniques for the detection of malignant pancreatobiliary strictures. Gastroenterology. 2006; 131(4):1064-1072.

(10.) Levy MJ, Baron TH, Clayton AC, et al. Prospective evaluation of advanced molecular markers and imaging techniques in patients with indeterminate bile duct strictures. Am J Gastroenterol. 2008; 103(5):1263-1273.

(11.) Barr Fritcher EG, Kipp BR, Halling KC, et al. A multivariable model using advanced cytologic method. Gastroenterology. 2009; 136(7):2180-2186.

(12.) Halling KC, Kipps BR. Fluorescence in situ hybridization in diagnostic cytology. Hum Pathol. 2007; 38 (8):1137-1144.

(13.) Barr Fritcher EG, Kipp BR, Slezak JM, et al. Correlating routine cytology, quantitative nuclear morphometry by digital image analysis, and genetic alterations by fluorescence in situ hybridization to assess the sensitivity of cytology for detecting pancreatobiliary malignancy. Am J Clin Pathol. 2007; 128(2):272 279.

(14.) Kipp BR, Karnes RJ, Brankley SM, et al. A comparison of routine cytology and fluorescence in situ hybridization for the detection of malignant bile duct strictures. Am J Gastroenterol. 2004; 99(9):1675-1681.

(15.) Levy MJ, Clain JE, Clayton A, et al. Preliminary experience comparing routine cytology results with the composite results of digital image analysis and fluorescence in situ hybridization in patients undergoing EUS-guided FNA. Gastrointest Endosc. 2007; 66(3):483-490.

(16.) Wenig BM, Heffess CS. Inflammatory, Infectious, and Non-Neoplastic Disorders of the Pancreas. In: Odze RD, Goldblum JR, eds. Surgical Pathology of the GI Tract, Liver, Biliary Tract, and Pancreas. 2nd ed Philadelphia, PA: Saunders Elsevier; 2009:877-960.

(17.) Law R, Bronner M, Vogt D, Stevens T. Autoimmune pancreatitis: a mimic of pancreatic cancer. Cleve Clin J Med. 2009; 76(10):607-615.

(18.) Kim T, Murakami T, Takamura M, et al. Pancreatic mass due to chronic pancreatitis: correlation of CT and MR imaging features with pathologic findings, Am J Roentgenol. 2001; 177(2):367-371.

(19.) Taniai M, Higuchi H, Burgart LJ, et al. p16INK4a promoter mutations are frequent in primary sclerosing cholangitis (PSC) and PSC-associated cholangio carcinoma. Gastroenterology. 2002; 123(4):1090-1098.

(20.) Ahrendt AS, Eisenberger CF, Yip L, et al. Chromosome 9p21 loss and p16 inactivation in primary sclerosing cholangitis-associated cholangiocarcinoma. J Surg Res. 1999; 84(1):88-93.

David N. Henkes, MD; Sandeep N. Patel, DO; Laura A. Rosenkranz, MD; Jose L. Escobedo, CT(ASCP)

Accepted for publication March 1, 2012.

From the Department of Pathology and Laboratory, CHRISTUS Santa Rosa Centre City Hospital, San Antonio, Texas (Dr Henkes); and the Division of Molecular Diagnostics, Pathology Reference Laboratory, San Antonio, Texas (Dr Henkes and Mr Escobedo); and the Department of Medicine/Gastroenterology, University of Texas Health Science Center at San Antonio, San Antonio, Texas (Drs Patel and Rosenkranz).

The authors have no relevant financial interest in the products or companies described in this article.

No compensation was received from any company for Abbott UroVysion (Abbott Molecular Inc, Des Plaines, Illinois), BioView Duet (BioView, Nes Ziona, Israel), or PerservCyt (Hologic, Bedford, Massachusetts) for this study.

The images on BioView Duet have never been published previously.

Reprints: David N. Henkes, MD, Division of Molecular Diagnostics, Pathology Reference Laboratory, 9600 Datapoint Dr, San Antonio, TX 78229 (e-mail: david.henkes@christushealth.org).
Table 1. Fine-Needle Aspiration (FNA) Diagnosis
Versus UroVysion (a) Fluorescence In Situ Hybridization
(UFISH) Results On Pancreatic Endoscopic Ultrasound-
Guided FNA Specimens (b)

UFISH Results, No.

FNA Diagnosis        Negative   Positive

Negative                8          0
Consistent with
  reactive changes      8          1
Suspicious              4          9
Positive                2         25

(a) Abbott Molecular Inc, Des Plaines, Illinois.

(b) Three of 60 cases did not have sufficient cells
for analysis.

Table 2. Patient Follow-up Correlation for 13 Patients
With "Suspicious" Cytology Findings and UroVysion (a)
Fluorescence In Situ Hybridization (UFISH)

                               UFISH Result, No.

                               Positive   Negative

Biopsy result positive for
  pancreatic adenocarcinoma       2          0
Clinically positive for
  pancreatic malignancy           3          0
Pancreatitis                      0          2
No follow-up information          4          2

(a) Abbott Molecular Inc, Des Plaines, Illinois.

Table 3. Patient Follow-up Correlation for 17 Patients
With "Reactive" or "Negative" Cytology Findings and
UroVysion (a) Fluorescence In Situ Hybridization (UFISH)

                               UFISH Result, No.

                              Positive   Negative

Biopsy result positive for
  pancreatic adenocarcinoma      0          3
Clinically positive for
  pancreatic malignancy          1          0
Pancreatitis                     0          5
No follow-up information         0          8

(a) Abbott Molecular Inc, Des Plaines, Illinois.
COPYRIGHT 2013 College of American Pathologists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Henkes, David N.; Patel, Sandeep N.; Rosenkranz, Laura A.; Escobedo, Jose L.
Publication:Archives of Pathology & Laboratory Medicine
Article Type:Report
Date:Jan 1, 2013
Words:5668
Previous Article:Selected reviews in oncologic surgical pathology including advances in laboratory technology: excerpts from the course "The Surgical Pathology of...
Next Article:Papillary syncytial metaplasia of fallopian tube endometriosis: a potential pitfall in the diagnosis of serous tubal intraepithelial carcinoma.
Topics:

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters