Pathology (January 2014) 46(1), pp. 69–92

CORRESPONDENCE Detection of hepatitis B virus DNA in serum is rare in the absence of hepatitis B surface antigen: impacts for detection and monitoring of chronic hepatitis B

Sir, Hepatitis B (HBV) is one of the most common infectious diseases worldwide, with 2 billion people having evidence of past or present infection worldwide and over 200,000 chronically infected in Australia alone.1,2 It has been estimated that in 2010 there were 312,400 deaths due to HBV-associated cirrhosis worldwide.3 Serological testing is the gold standard for diagnosis, with detection of hepatitis B surface (HBsAg) and envelope antigen (HBeAg) consistent with active replicating virus, and antihepatitis B surface antibody (anti-HBs Ab) found following successful immunisation or control of natural infection. Over time it has become clear that loss of HBsAg with or without the development of anti-HBs Ab does not represent ‘clearance’ of HBV as previously thought. In some patients, low level viral replication within the liver may persist associated with, or in the absence of, low level detectable HBV DNA in the serum.4 The prevalence of so called ‘occult HBV’ varies geographically but was found in 5–10% of European blood donors with isolated anti-hepatitis B core Ab.5 The long-term clinical impacts have not yet been determined but these patients are at risk of HBV reactivation in the setting of immunosuppression.6 Quantification of HBV DNA is used to confirm the presence of circulating HBV and to monitor and guide treatment of chronic HBV, but its ability to predict HBsAg re-emergence and active hepatitis in the setting of immunosuppression remains undefined.7 Due to a perceived increase in requests for HBV DNA testing in patients traditionally thought at low risk for active HBV (HBsAg negative, anti-HBs Ab positive) a retrospective review of HBV DNA testing was performed to provide local data as to the reliability of serological and biochemical markers as predictors of HBV DNA positivity, and thus guide its ongoing use. All patients who had HBV DNA quantification at the Alfred Hospital from September 2008 to September 2010 were identified from the laboratory records and included. The Alfred Hospital is an adult tertiary referral centre with large hepatology, haematology and infectious diseases units. It is the state referral centre for the management of HIV and performs a number of solid organ and haematological transplants each year. Study information was collected from the electronic pathology system and included age, gender, serological testing and the result of HBV DNA including viral load when positive. Only serological testing performed at the Alfred Hospital and ordered on the same day or prior to the HBV DNA was recorded (i.e., the results that would be available to the laboratory or physician when ordering the test). The most recent liver function test, specifically albumin, gamma-glutamyl transferase (GGT) and alanine aminotransferase (ALT), were recorded, as was the patient’s hepatitis C (HCV) and HIV sero-status and the medical specialty of the Print ISSN 0031-3025/Online ISSN 1465-3931

#

requesting physician. In patients with multiple HBV DNA requests over the 2 year study period each was considered a new event and recorded separately. HBV DNA requests were processed at the state reference laboratory using the Abbott RealTime HBV assay (Abbott Molecular, USA), detection limit 15 IU/mL.8 All serological testing is performed on site using the Abbott ARCHITECT i2000SR system. Anti-HBs Ab was considered positive when the titre was 10 IU/mL or greater. Results were summarised using chi-squared tests to evaluate differences in proportions and the Mann–Whitney U test for continuous data. Statistical significance was regarded as p < 0.05, two tailed and all statistical analyses were performed using Stata 11.0/IC (Stata, USA). The project received approval from the Alfred Hospital Ethics committee. From 443 patients, 1101 samples were identified; seven were subsequently excluded because the sample had not been processed, hence 1094 samples (436 patients) were included. Patient characteristics are shown in Table 1. Ninety-nine (22.7%) were HIV positive; 65 (14.9%) HCV positive and 27 (6.2%) were co-infected with HIV and HCV. Twenty patients were concurrently HIV, HCV and anti-HBc Ab positive; 227 (52.1%) and 132 (30.3%) had not been screened for HIV or HCV, respectively. In total, 473 (43.2%) samples had detectable HBV DNA; 192 (44.0%) of all patients at some time during the study period. Samples from HIV positive patients had lower HBV viral loads and were less likely to have detectable HBV DNA (see Table 1). A HBsAg result was not available in 43 (10.3%) patients. Anti-HBs Ab and anti-HBc Ab data were unavailable in 111 (25.4%) and 113 (25.9%), respectively. All patients with detectable HBV DNA were HBsAg positive. That is, there were no cases of occult hepatitis B detected by HBV DNA testing over the study period. Of the samples that were HBsAg positive [865 (87%)], 426 (49.2%) had detectable HBV DNA. Of the anti-HBs Ab positive samples [103 (13.0%)], 25 had detectable HBV DNA (all 25 concurrently HBsAg positive). The mean anti-HBs Ab titre in samples with detectable HBV DNA and anti-HBs Ab was 18.1 mIU/mL (range 10–74), compared with 285 mIU/mL (range 11–1000) in samples with detectable anti-HBs Ab but negative HBV DNA ( p < 0.001). There were 37 samples from 26 patients with isolated anti-HBc Ab positivity (HBsAg and anti-HBs Ab negative); all were HBV DNA negative, 10 of these patients were HIV positive and the remainder were under the care of a haematologist. The relationship between HBV DNA detectability and liver function tests are shown in Table 2. Overall samples with detectable HBV DNA had higher ALT and lower GGT levels. This result was not true for the subset of patients who were HIV positive in whom a detectable HBV DNA was associated with lower albumin levels but no differences in ALT or GGT. Hepatitis B is a major cause of morbidity and mortality worldwide. This study confirms the importance of serology in diagnosing and monitoring HBV and, strikingly, over a 2 year period found no instances of detectable HBV DNA in serum in the absence of concurrent HBsAg. While the role of HBV DNA measurement in treatment initiation and monitoring in HBsAg positive infections has been well established,9 it may be a poor option in patients

2013 Royal College of Pathologists of Australasia

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

70

Pathology (2014), 46(1), January

CORRESPONDENCE

Table 1

Patient characteristics HIV status Negative*

Positive

436 (100) 302 (69.3) 49.1 (19 – 82)

337 (77.3) 207 (61.4) 49.9 (19 – 82)

99 (22.7) 95 (95.9) 46.3 (26 – 70)

269 117 24 8 4 14 65 473 2482 865 104 40 742 146 595

261 30 20 8 4 14 38 398 3860 650 78 17 546 101 531

8 87 4 0 0 0 27 75 594 215 26 244 196 45 64

Total n (%) Male Age, mean years (range) Speciality of requesting physician Gastroenterology Infectious diseases Haematology Transplant Oncology Miscellaneous HCV Ab positive HBV DNA detectable HBV DNA titre (IU/mL),{ median (range) HBsAg detectable anti-HBs Ab detectable anti-HBsAb titrez (mIU/mL), median (range) anti-HBcAb detectable HBeAg detectable anti-HBe Ab detectable

(61.7) (26.8) (5.5) (1.8) (0.9) (3.2) (21.3) (43.2) (17–5.32 109) (79.0) (9.5) (10–1000) (67.8) (13.3) (54.4)

(77.4) (8.9) (5.9) (2.3) (1.1) (4.1) (18.1) (47.6) (19–5.32 109) (77.7) (9.3) (10–1000) (96.9) (12.1) (63.5)

p value – – –

(8.0) (87.8) (4.0) (0.0) (0.0) (0.0) (28.7) (29.1) (17–3.51 109) (83.3) (10.1) (12–1000) (85.9) (17.4) (24.8)

– – – – – – 256 >256 >256 >256 >256 >256

>256 16–64 1.5–32 16–>256 32–64 256 0.5 48–64 24 32 192 >256 1

Allelic profile

ST

CC

7-1-1-1-5-1-1 5-3-1-9-1-1-5 4-1-1-1-1-1-1 9-1-1-1-12-1-1 4-1-1-1-1-1-1 15-1-1-1-1-20-1

18 137 555 80 555 203 ND 18 ND 80 809 412 446

17 5 17 17 17 17

7-1-1-1-5-1-1 9-1-1-1-12-1-1 15-13-9-17-8-49-6 15-1-1-44-1-20-1 5-2-6-6-1-7-1

17 17 Non 17 17 9

CC, clonal complex; MIC, minimum inhibitory concentration; MLST, multilocus sequence type; ND, not done; PFGE, pulsed-field gel electrophoresis; ST, sequence type; TEC, teicoplanin; VAN; vancomycin.

total, 87% of isolates were multi-resistant to vancomycin, teicoplanin, penicillin and ampicillin. Twenty-one (68%) and 14 (45%) isolates were resistant to quinupristin/dalfopristin and high level gentamicin, respectively. All isolates were linezolid susceptible. The PFGE pulsotypes included: pulsotype A (1 isolate), B (6), C (1), D (13), E (3), F (1), H (2), I (1), K (1), L (1) and M (1). One patient carried three pulsotypes (C, D and H). Four of the six pulsotypes with single isolates (pulsotypes A, I, K and M) were cultured from patients who had either been admitted as a direct transfer, or had been hospitalised in the previous 12 months, from an overseas hospital. The initial cases in pulsotype F (3 cases) and pulsotype H (2 cases) had also been hospitalised overseas in the previous 12 months. No overseas hospital risk factors could be identified for five pulsotypes: C, E and L (single isolates); and B and D (6 and 13 patients, respectively). Increased travel, particularly when associated with healthcare contact overseas, compounded with the emergence of epidemic clones carrying resistance determinants, has been shown to contribute to the global spread of multi-resistant organisms.11 Epidemiological data on the 31 Royal Perth Hospital isolates suggests importation of vanA E. faecium, including the epidemic CC 17, from overseas (where vanA VRE predominates), and may partially explain the recent local increase in vanA VRE incidence. This hypothesis is supported by recent molecular evidence of the introduction of other multi-resistant organisms into WA from overseas including methicillin-resistant Staphylococcus aureus.12 Although clinical infections due to VRE in WA are infrequently seen, the increasing isolation of vanA E. faecium from rectal and faecal surveillance specimens is of concern. Compared to vanB VRE, the therapeutic options for vanA VRE are limited due to higher teicoplanin MICs. This combined with demonstrated capacity for clonal dissemination and horizontal gene transfer13 is worrisome and should prompt other Australian healthcare facilities to have a heightened awareness for detection and proactive infection control responses to this emerging threat. Acknowledgements: We gratefully acknowledge the scientific and technical staff from the Australian Collaborating Centre for Enterococcus and Staphylococcus Species (ACCESS) Typing and Research; the Department of

Microbiology and Infectious Diseases, PathWest Laboratory Medicine – WA, Royal Perth Hospital; and the WA Genomic Resource Centre, Department of Clinical Immunology and Biochemical Genetics, PathWest Laboratory Medicine – WA, Royal Perth Hospital, for the scientific and technical expertise. In addition we thank the public and private medical laboratories in Western Australia for referring the isolates. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Geoffrey W. Coombs1,2,3 Denise Daley1 Julie C. Pearson1 Paul R. Ingram1,3 1

Department of Microbiology and Infectious Diseases, PathWest Laboratory Medicine - WA, Royal Perth Hospital, Perth, 2Australian Collaborating Centre for Enterococcus and Staphylococcus Species (ACCESS) Typing and Research, School of Biomedical Sciences, Curtin University, Perth, and 3 Department of Pathology and Laboratory Medicine, University of Western Australia, Perth, WA, Australia Contact Dr G. Coombs. E-mail: [email protected] 1. Bearman GM, Wenzel RP. Bacteremias: a leading cause of death. Arch Med Res 2005; 36: 646–59. 2. Cetinkaya Y, Falk P, Mayhall CG. Vancomycin-resistant enterococci. Clin Microbiol Rev 2000; 13: 686–707. 3. Klare I, Heier H, Claus H, Reissbrodt R, Witte W. vanA-mediated high-level glycopeptide resistance in Enterococcus faecium from animal husbandry. FEMS Microbiol Lett 1995; 125: 165–71. 4. Kamarulzaman A, TFA, Boquest AL, Geddes JE, Richards MJ. Vancomycin-resistant Enterococcus faecium in a liver transplant recipient. Aust N Z J Med 1995; 25: 560 (Abstr). 5. Bell JM, Paton JC, Turnidge J. Emergence of vancomycin-resistant enterococci in Australia: phenotypic and genotypic characteristics of isolates. J Clin Microbiol 1998; 36: 2187–90. 6. Cooper E, Paull A, O’Reilly M. Characteristics of a large cluster of vancomycin-resistant enterococci in an Australian hospital. Infect Control Hosp Epidemiol 2002; 23: 151–3. 7. Christiansen KJ, Tibbett PA, Beresford W, et al. Eradication of a large outbreak of a single strain of vanB vancomycin-resistant Enterococcus faecium at a major Australian teaching hospital. Infect Control Hosp Epidemiol 2004; 25: 384–90.

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

CORRESPONDENCE

8. Johnson PD, Ballard SA, Grabsch EA, et al. A sustained hospital outbreak of vancomycin-resistant Enterococcus faecium bacteremia due to emergence of vanB E. faecium sequence type 203. J Infect Dis 2010; 202: 1278–86. 9. Government of Western Australia, Department of Health. Infection Prevention and Control of Vancomycin-Resistant Enterococci (VRE) in Western Australian Acute Care Healthcare Facilities. Version 2, December 2011. Perth: Government of Western Australia, Department of Health, 2011. http://www.health.wa.gov.au/CircularsNew/attachments/629.pdf 10. Clinical and Laboratory Standards Institute (CLSI). Performance Standards for Antimicrobial Susceptibility Testing. Twenty-second Informational Supplement. M100-S22. Wayne, PA: CLSI, 2012. 11. Rogers BA, Kennedy KJ, Sidjabat HE, Jones M, Collignon P, Paterson DL. Prolonged carriage of resistant E. coli by returned travellers: clonality, risk factors and bacterial characteristics. Eur J Clin Microbiol Infect Dis 2012; 31: 2413–20. 12. Monecke S, Ehricht R, Slickers P, Tan HL, Coombs G. The molecular epidemiology and evolution of the Panton-Valentine leukocidin-positive, methicillin-resistant Staphylococcus aureus strain USA300 in Western Australia. Clin Microbio Infect 2009; 15: 770–6. 13. Werner G, Coque TM, Hammerum AM, et al. Emergence and spread of vancomycin resistance among enterococci in Europe. Euro Surveill 2008; 13: 19046.

DOI: 10.1097/PAT.0000000000000038

Chronic myeloid leukaemia masquerading as primary myelofibrosis Sir, Chronic myeloid leukaemia (CML) is a myeloproliferative neoplasm characterised by the BCR-ABL fusion gene as a result of translocation between chromosomes 9 and 22. The natural history of untreated CML is usually triphasic: an initial indolent chronic phase (CP) is followed by an accelerated phase (AP) and a blast phase (BP). Approximately 85% of CML patients are diagnosed in the chronic phase. We report a patient who had atypical features of CML at presentation mimicking primary myelofibrosis. A 51-year-old Caucasian male patient was referred to haematology by his family doctor for evaluation of symptomatic anaemia and thrombocytopenia. His main symptoms were lethargy, easy bruising and fever of 4 weeks duration. There was no palpable organomegaly in the abdomen and no palpable peripheral lymphadenopathy. His performance status on ECOG scale was 1. Initial blood tests showed a low haemoglobin [77 g/L, reference range (RR) 130–170 g/L] and a low platelet count (33  109/L, RR 150–450). His total white cell count was 10  109/L (RR 4–11) with a normal neutrophil count, but the basophil count was raised (1.2  109/L). The blood film demonstrated tear drop poikilocytes, granulocytic left shift, basophilia (12%) and a leuko-erythroblastic picture with a peripheral blast count of 1%. Liver function tests, serum creatinine, electrolytes and uric acid levels were within normal range, but the lactate dehydrogenase (LDH) level was high (1213 U/L, RR 125–243 U/L). Since the picture was suggestive of a bone marrow infiltrative process, a bone marrow biopsy was performed from the posterior iliac crest. Then bone marrow aspirate was unsuccessful due to a dry tap. Histopathological examination of the bone marrow trephine revealed sclerotic trabecular bone with diffuse coarse collagen fibrosis and absence of fat spaces. The cellular elements were predominantly small mononuclear cells distributed throughout the marrow with their morphology distorted

75

due to the marrow fibrosis (Fig. 1A). Atypical small megakaryocytes were increased, the majority were scattered with a few small clusters and these cells were highlighted by CD61 immunohistochemical stain. Blasts were not increased in the marrow and a very few CD34 positive cells were highlighted by immunohistochemistry. Reticulin fibres were increased diffusely and densely with extensive intersections on silver nitrate staining (Fig. 1B). The grade of marrow fibrosis was MF3 according to the semi-quantitative grading of marrow fibrosis.1 Based on these morphological features a provisional diagnosis of primary myelofibrosis was made but there was no splenomegaly or hepatomegaly clinically and on abdominal computed tomography (CT) scans. Since the marrow aspirate was a dry tap, tests for JAK2 V617F and BCR-ABL mutation analyses were performed on the blood samples by polymerase chain reaction (PCR) technique. The result of the JAK2 V617F mutation was negative, but the BCR-ABL fusion gene was positive (b2a2 and b3a2 transcripts) and when measured by quantitative real time PCR was 50% on the international scale. These reports were received after 3 weeks and the diagnosis was revised to chronic myeloid leukaemia (CML). After the positive BCR-ABL result, cytogenetic tests were performed on a peripheral blood sample; however, the culture did not yield sufficient cells for analysis. Interphase fluorescent in situ hybridisation (FISH) analysis using Vysis LSI BCR-ABL dual fusion probe (Abbott Molecular, USA) on unstained bone marrow trephine imprints showed a BCR-ABL rearrangement in 60% of the cells. Only a

Fig. 1 (A) Bone marrow trephine biopsy showing hypercellular marrow, marked collagen fibrosis and osteosclerosis. (B) Bone marrow trephine biopsy with silver nitrate staining showing marked increase in reticulin fibres (MF3).

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

76

Pathology (2014), 46(1), January

CORRESPONDENCE

single fusion signal was noted in all those cells indicating that there was deletion of the derivative chromosome 9. Initially our patient was managed supportively and after CML was diagnosed dasatinib, a tyrosine kinase inhibitor (TKI) was started at the dose of 100 mg daily. He tolerated the treatment well without any side effects and blood counts were monitored periodically. Four weeks after starting dasatinib, his basophil counts dropped to normal values, which was the first indication of a haematological response. After 10 weeks the total white cell count, differential count and platelet count normalised; however, the haemoglobin remained mildly reduced (102 g/L). A repeat bone marrow biopsy after 4 months of treatment demonstrated recovery of normal haematopoiesis, normal megakaryocyte morphology, resolution of bone sclerosis (Fig. 2A) and significant reduction of marrow fibrosis from MF3 to MF1 on silver nitrate staining (Fig. 2B). The BCR-ABL quantification test on peripheral blood at 4 months showed no detectable transcripts (0%) which was consistent with a rapid achievement of molecular remission. Twelve months after the diagnosis, the patient continued to take dasatinib 100 mg daily and he remained healthy with normal blood counts and undetectable BCR-ABL transcript on blood tests until the last review. Up to 30% of CML patients can present with marrow fibrosis, commonly mild (MF1) or moderate (MF2); however, advanced marrow fibrosis (MF3) is rare, occurring in less than 5% of patients.2 Marked reticulin and collagen fibrosis may

also be present in the accelerated phase of CML.3 The bone marrow in accelerated phase is usually hypercellular with large clusters or sheets of small abnormal megakaryocytes.4 An increase in myeloblasts may be noted, more readily appreciated with CD34 immunohistochemical stain performed on the trephine biopsy.5 Studies have shown that patients with CML and marrow fibrosis on TKI therapy demonstrated a significant resolution of fibrosis, which is due to the inhibition of pro-fibrotic cytokines such as transforming growth factor b (TGFb) and platelet derived growth factor (PDGF).6–8 Marrow fibrosis is associated historically with a poor prognosis in patients with CML. With TKI therapy the prognosis of newly diagnosed CML patients with severe marrow fibrosis is better, however 15% of patients with advanced marrow fibrosis still had a worse outcome.9 Our patient presented with advanced marrow fibrosis (MF3), marked anaemia and thrombocytopenia which were highly suggestive of primary myelofibrosis and these features are uncommon in newly diagnosed CML patients. Testing for the BCR-ABL fusion gene in his peripheral blood was very helpful in the early diagnosis of CML. The clinical presentation along with the cytogenetic abnormality (deletion of the derivative chromosome 9) suggested that the patient presented in the accelerated phase of CML.10 However, other typical features of accelerated phase such as high blast percentage in marrow and splenomegaly were absent. A few similar cases with combined Philadelphia chromosome positivity and features of primary myelofibrosis have been described so far.11,12 However, in a similar case, the disease response to imatinib was reported to be poor.12 Another case report describes a patient who initially presented with primary myelofibrosis, but developed CML 7 years later. In this patient imatinib therapy led to disappearance of CML clone, but the myelofibrosis persisted.13 Within 4 months of starting TKI therapy our patient had recovery of normal haematopoiesis, resolution of marrow fibrosis and rapid achievement of molecular remission, which support the diagnosis of CML alone in our patient. Such a rapid resolution of marrow fibrosis and normalisation of blood counts is unlikely in patients with primary myelofibrosis with the currently available treatment modalities. In conclusion, CML can rarely mimic primary myelofibrosis and BCR-ABL molecular analysis should be performed on the blood samples or trephine imprints of patients with a presumptive diagnosis of myeloproliferative neoplasm if the marrow aspirate is dry or unsuccessful. Acknowledgement: We thank Professor Tim Hughes, Department of Haematology, SA Pathology, Adelaide, Australia, for his expert opinion and helping us with molecular studies for this patient. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Muhajir Mohamed1,2 Michael Beamish1 Julian Grabek1 Launceston General Hospital, Launceston, and 2Launceston Clinical School, University of Tasmania, Launceston, Tas, Australia 1

Fig. 2 (A) Bone marrow trephine biopsy after 4 months of dasatinib treatment showing resolution of fibrosis and osteosclerosis and normal haematopoiesis. (B) Bone marrow trephine biopsy and silver nitrate staining after 4 months of dasatinib treatment showing resolution of reticulin fibrosis.

Contact Dr M. Mohamed. E-mail: [email protected]

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

CORRESPONDENCE

1. Thiele J, Kvasnicka HM, Facchetti F, et al. European consensus on grading bone marrow fibrosis and assessment of cellularity. Haematologica 2005; 90: 1128–32. 2. Thiele J, Kvasnicka HM. Myelofibrosis in chronic myeloproliferative disorders–dynamics and clinical impact. Histol Histopathol 2006; 21: 1367–78. 3. Buesche G, Hehlmann R, Hecker H, et al. Marrow fibrosis, indicator of therapy failure in chronic myeloid leukemia – prospective long-term results from a randomized-controlled trial. Leukemia 2003; 17: 2444–53. 4. Muehleck SD, McKenna RW, Arthur DC, et al. Transformation of chronic myelogenous leukemia: clinical, morphologic, and cytogenetic features. Am J Clin Pathol 1984; 82: 1–14. 5. Orazi A, Neiman RS, Cualing H, et al. CD34 immunostaining of bone marrow biopsy specimens is a reliable way to classify the phases of chronic myeloid leukemia. Am J Clin Pathol 1994; 101: 426–8. 6. Bueso-Ramos CE, Cortes J, Talpaz M, et al. Imatinib mesylate therapy reduces bone marrow fibrosis in patients with chronic myelogenous leukemia. Cancer 2004; 101: 332–6. 7. Buchdunger E, Cioffi CL, Law N, et al. Abl protein-tyrosine kinase inhibitor STI571 inhibits in vitro signal transduction mediated by c-kit and platelet-derived growth factor receptors. J Pharmacol Exp Ther 2000; 295: 139–45. 8. Daniels CE, Wilkes MC, Edens M, et al. Imatinib mesylate inhibits the profibrogenic activity of TGF-beta and prevents bleomycin-mediated lung fibrosis. J Clin Invest 2004; 114: 1308–16. 9. Kantarjian HM, Bueso-Ramos CE, Talpaz M, et al. Significance of myelofibrosis in early chronic-phase, chronic myelogenous leukemia on imatinib mesylate therapy. Cancer 2005; 104: 777–80. 10. Turhan AG. Chronic myelogenous leukaemia (CML). Atlas Genet Cytogenet Oncol Haematol. August 2008; cited June 2013. http://Atlas GeneticsOncology.org/Anomalies/CML.html 11. Koshy J, Alperin J, Jana B, et al. A case of Philadelphia chromosome positive myeloproliferative neoplasm in a pregnant woman with unusual primary myelofibrosis features. Case Rep Hematol 2013; 2013: 702831. 12. Chen F, Zhang C, Wang W, et al. A case of Philadelphia-chromosome positive chronic idiopathic myelofibrosis. Leuk Res 2008; 32: 665–7. 13. Laibe S, Tadrist Z, Arnoulet C, et al. A myeloproliferative disorder may hide another one. Leuk Res 2009; 33: 1133–6.

DOI: 10.1097/PAT.0000000000000030

Plasmablastic lymphoma presenting as a colonic stricture in Crohn’s disease Sir, Plasmablastic lymphoma (PBL) is a variant of diffuse large B-cell lymphoma (DLBCL) originally described in the oral cavity of human immunodeficiency virus (HIV) positive patients.1 It was subsequently found to occur in extra-oral sites and in the absence of immunosuppression.2–4 We present a case of PBL arising in a patient with longstanding Crohn’s disease. Our patient was a 54-year-old man admitted for surgical management of a colonic stricture. He had a long (approximately 30 years) history of Crohn’s disease, previously treated with sulphasalazine and prednisone, but he was not on any medications at time of admission. He was never treated with biological agents. There were no other comorbidities and no history of haematological malignancies. In the subtotal colectomy specimen an area of stricture was noted, but no other lesions were seen on the mucosal surface or on sectioning. Microscopic examination showed a malignant infiltrate measuring approximately 20 mm at the site of the stricture, undermining and ulcerating the mucosa. It had a discohesive sheet-like growth pattern with associated delicate fibrosis. The cells had round uniform nuclei and a small amount of amphophilic cytoplasm. Chromatin pattern was coarse and nucleoli were variably prominent, ranging from small and multiple (‘centroblast-like’) to large and centrally located

77

(‘immunoblast-like’) (Fig. 1A). Focal plasmacytic differentiation was seen and mitotic activity was brisk. Active Crohn’s colitis was present in the surrounding large bowel. The neoplastic cells were positive for CD138, MUM1, BOB1 and Oct-2, weakly positive for CD45 and EMA, and negative for CD20, CD30, CD56, CD79a, PAX5, ALK1, S100 and AE1/AE3 (Fig. 1B–E). Ki-67 index was >90% (Fig. 1F). Epstein–Barr virus (EBV) in situ hybridisation (ISH) was negative. No other disease focus was identified on positron emission tomography (PET) scan, and other investigations including computed tomography (CT), full blood examination, plasma biochemistry and renal function were unremarkable. A decision was made for active monitoring only and no chemotherapy was given. The patient remained disease free 20 months post-diagnosis. PBL is defined in the 2008 WHO classification as a neoplasm composed of B immunoblast-like cells having the immunophenotype of plasma cells.5 It was first described in 1997 by Delecluse et al.,1 where a set of common clinical associations were also identified. It had a predilection for the oral cavity of HIV positive patients, and the cells were also often infected by EBV. PBL was soon identified in a variety of extra-oral sites, including skin, lymph node, bone, gastrointestinal tract and soft tissue.2–4 It also occurred in patients with immunosuppression due to other causes, such as solid organ transplantation, and in patients without any identifiable immunosuppression.2–4 Morphologically, PBL can assume a number of appearances. The usual pattern is sheet-like and cohesive, composed of large cells with abundant amphophilic to basophilic cytoplasm, sometimes having an epithelioid appearance. Nuclei contain coarse chromatin and variably prominent nucleoli, and may resemble centroblasts or immunoblasts. Definite plasmacytic differentiation ranges from not identifiable to prominent. Mitotic activity and apoptotic debris are easily seen. The immunophenotype of the neoplastic cells is identical to plasma cells. The cells are positive for CD38, CD138, VS38c and MUM1, variably positive for CD30, CD79a, cytoplasmic immunoglobulin and EMA, with little to no expression of CD20, CD45 and PAX5.5 Ki-67 index is very high (>90%).5 EBV positivity, as detected by EBER ISH, is seen in 82% of HIV positive patients but only 46% of HIV negative patients.6 Translocations involving the MYC gene can be identified in a subset of cases.7 Regardless of site or immunocompetence, in most studies PBL is seen to present late (Ann Arbor stage III or IV) and pursue an aggressive course.1,4,6 However, this aggressiveness appears to be dependent on immunosuppression; for instance, prognosis in HIV positive patients was very poor in earlier studies but improved significantly in later studies, presumably due to effective management of HIV.1,8 Long-term survival is possible even in patients with advanced disease.4 The development of lymphoma in inflammatory bowel disease is associated with treatment by biological agents, notably thiopurines, and to a lesser extent anti-TNF-a antibodies.9 This increase is not seen in patients not on any therapy or those on other forms of immunosuppression, such as corticosteroids.9 A broad range of lymphomas can occur, including B-cell and T-cell types,10 and prognosis is dependent on the type of lymphoma, degree of immunosuppression, and comorbidities associated with the inflammatory bowel disease. Our case is unusual in that PBL presented within a colonic stricture in an immunocompetent patient. The relationship between the lymphoma and stricture is uncertain; possibilities

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

78

Pathology (2014), 46(1), January

CORRESPONDENCE

A

B

C

D

E

F

Fig. 1 (A) The tumour cells have centroblastic and immunoblastic features. (B) They are positive for CD138, (C) MUM1 and (D) BOB1, (E) negative for CD56, and (F) have a Ki-67 index of >90%.

include lymphoma inducing local fibrosis and stricture formation, or the local immunological abnormalities and persistent inflammation associated with the stricture resulting in the development of lymphoma. This atypical presentation raises a number of differential diagnoses and illustrates the potential diagnostic difficulties. When morphology appears undifferentiated and plasmacytic differentiation is minimal, the differential diagnosis includes metastatic carcinoma/melanoma, B-cell lymphomas with plasmablastic features (such as plasmablastic Burkitt lymphoma and ALK1 positive DLBCL), and plasmablastic myeloma/plasmacytoma. Metastatic carcinoma/melanoma and B-cell lymphomas with plasmablastic features are readily excluded with standard immunohistochemical panels, although an erroneous diagnosis of metastatic carcinoma is possible due to lack of B-cell markers and EMA positivity. Newer B-cell markers such as MUM1, BOB1 and Oct-2 may prove useful in this distinction, as in our case. The most important differential is plasmablastic myeloma/ plasmacytoma, as this distinction has significant treatment implications. In the absence of clinical and radiological information, distinction is sometimes not possible, as the neoplastic cells in both conditions display the morphological and immunohistochemical features of plasma cells. Some studies have shown CD56 negativity3,11 and a ‘high’ Ki-67 index (>60%)1,12 is in favour of PBL; however, this distinction is not clear cut. EBV infection, if present, is suggestive of PBL, but positivity is occasionally seen in myeloma/plasmacytoma.11 Furthermore, PBL in HIV negative patients is EBER negative in more than 50% of cases,6 so the absence of EBV antigens does not exclude the diagnosis. When these studies are non-contributory, the distinction is made by clinicopathological correlation. The absence of bony lesions, anaemia, hypercalcaemia, renal dysfunction and serum/urinary paraprotein rules out myeloma, although in some studies the presence of paraprotein was considered acceptable for a diagnosis of PBL.3 In the case of a solitary lesion, distinction from plasmablastic plasmacytoma may be impossible; close follow-up is required in these patients due to the risk of

progression to overt myeloma in solitary plasmacytoma. Our case did not display any clinical or biochemical features of myeloma, while the neoplastic cells were negative for CD56 and had a Ki-67 index of >90%. Therefore, PBL was the favoured diagnosis, rather than plasmablastic plasmacytoma. In summary, we report a case of PBL presenting as a colonic stricture in Crohn’s disease, an association not previously reported in the literature. PBL is an aggressive variant of DLBCL which may occur in a variety of sites, with or without immunosuppression. It should be included as a differential diagnosis in poorly differentiated large cell malignancies negative for cytokeratins, melanocyte antigens, and standard lymphoid markers. Acknowledgements: We would like to thank Dr Sophia Otto at SA Pathology for providing us with one of the references. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Cheng Liu1 Winny Varikatt1,2 Chow Heok P’ng1,2 1

ICPMR, Westmead Hospital, Westmead, and 2University of Western Sydney, Richmond, NSW, Australia Contact Dr C. Liu. E-mail: [email protected] 1. Delecluse HJ, Anagnostopoulos I, Dallenbach F, et al. Plasmablastic lymphomas of the oral cavity: a new entity associated with the human immunodeficiency virus infection. Blood 1997; 89: 1413–20. 2. Castillo JJ, Reagan JL. Plasmablastic lymphoma: a systematic review. Scientific World Journal 2011; 11: 687–96. 3. Hsi ED, Lorsbach RB, Fend F, et al. Plasmablastic lymphoma and related disorders. Am J Clin Pathol 2011; 136: 183–94. 4. Teruya-Feldstein J, Chiao E, Filippa DA, et al. CD20-negative large-cell lymphoma with plasmablastic features: a clinically heterogenous spectrum in both HIV-positive and -negative patients. Ann Oncol 2004; 15: 1673–9.

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

CORRESPONDENCE

5. Stein H, Harris NL, Campo E. Plasmablastic lymphoma. In: Swerdlow SH, Campo E, Harris NL, et al, editors. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissue. Lyon: IARC Press, 2008; 256–7. 6. Castillo JJ, Winer ES, Stachurski D, et al. Clinical and pathological differences between human immunodeficiency virus-positive and human immunodeficiency virus-negative patients with plasmablastic lymphoma. Leuk Lymphoma 2010; 51: 2047–53. 7. Valera A, Balague´ O, Colomo L, et al. IG/MYC rearrangements are the main cytogenetic alteration in plasmablastic lymphomas. Am J Surg Pathol 2010; 34: 1686–94. 8. Castillo JJ, Winer ES, Stachurski D, et al. Prognostic factors in chemotherapy-treated patients with HIV-associated plasmablastic lymphoma. Oncologist 2010; 15: 293–9. 9. Subramaniam K, D’Rozario J, Pavli P. Lymphoma and other lymphoproliferative disorders in inflammatory bowel disease: a review. J Gastroenterol Hepatol 2013; 28: 24–30. 10. Beaugerie L, Brousse N, Bouvier AM, et al. Lymphoproliferative disorders in patients receiving thiopurines for inflammatory bowel disease: a prospective observational cohort study. Lancet 2009; 374: 1617–25. 11. Colomo L, Loong F, Rives S, et al. Diffuse large B-cell lymphomas with plasmablastic differentiation represent a heterogeneous group of disease entities. Am J Surg Pathol 2004; 28: 736–47. 12. Kane S, Khurana A, Parulkar G, et al. Minimum diagnostic criteria for plasmablastic lymphoma of oral/sinonasal region encountered in a tertiary cancer hospital of a developing country. J Oral Pathol Med 2009; 38: 138–44.

DOI: 10.1097/PAT.0000000000000039

BRAF V600E mutation specific immunohistochemistry with clone VE1 is not reliable in pituitary adenomas Sir, BRAF mutations occur frequently in many neoplasms including papillary thyroid carcinoma, melanoma, colorectal cancer, low grade ovarian serous neoplasms, hairy cell leukaemia and Langerhans cell histiocytosis, and at a low but clinically significant incidence in many other tumours including lung cancer and cerebral neoplasia, summarised by Pakneshan et al.1 The presence of a BRAF mutation can be highly significant both in establishing a definitive tissue diagnosis and in guiding therapy given the proven efficacy of BRAF inhibition in certain tumours harbouring these mutations such as hairy cell leukaemia and melanoma and emerging evidence of its potential efficacy in other neoplasms such as lung cancer and certain brain tumours.1 The BRAFV600E mutation which results in the substitution of glutamic acid (E) for valine (V) at codon 600 is the most common and clinically significant BRAF mutation.1 The identification of BRAF mutation in the routine diagnostic setting has been greatly facilitated by the development of mutation specific immunohistochemistry, particularly using clone VE1 which identifies the BRAFV600E mutation but not other BRAF mutations.2 Several groups have now demonstrated that immunohistochemistry with clone VE1 is highly sensitive and specific for the presence of BRAFV600E mutation in many tumour types including cerebral neoplasia, melanoma, papillary thyroid carcinoma, hairy cell leukaemia, colorectal carcinoma and lung cancer.2–10 Occasionally VE1 does produce patchy but clearly non-specific positive staining in otherwise negative tumours or non-neoplastic tissue; for example in colons focal non-specific staining has been reported

79

in smooth muscle or non-neoplastic epithelium.8 However, this does not seem to cause difficulty in interpretation in the diagnostic laboratory. In fact most studies have demonstrated that mutation specific IHC may be more sensitive and specific than molecular techniques in the routine clinical setting.5–9 Therefore, mutation specific immunohistochemistry for BRAFV600E with clone VE1 is currently used routinely in many diagnostic surgical pathology laboratories. It would be reasonable to assume that if the antibody is highly sensitive and specific in so many neoplasms it could be deployed for use in all other settings without the need for specific validation studies for each tumour type. After all, this has been the model traditionally employed for other immunohistochemical markers. However, recently Sperveslage et al. reported that BRAFV600E immunohistochemistry with clone VE1 frequently demonstrated false positive staining in pituitary adenomas.11 The authors performed BRAFV600E mutation specific immunohistochemistry with clone VE1 on 78 pituitary adenomas, of which 18 were positive: 12 weakly, five moderately and one strongly positive. However all 10 positive cases analysed by two different molecular methods (clamped probe assay and Sanger sequencing) were negative for the BRAFV600E mutation and also for mutations of the putative homologs ARAF F453E and CRAF V429E. We sought to investigate the specificity of BRAFV600E immunohistochemistry using the same clone in a large cohort of pituitary tumours. We constructed a tissue microarray (TMA) containing duplicate cores from 294 pituitary adenomas comprising all available pituitary adenomas resected at the Royal North Shore Hospital, Sydney, Australia, during the period 1998– 2012. IHC was performed with clone VE1 using identical methods to that which we have previously employed in papillary thyroid carcinoma and colorectal carcinoma with extremely high sensitivity and specificity (apparently outperforming molecular analysis).6,8 Cases were scored by a pathologist with experience in interpreting VE1 immunohistochemistry, blinded to all clinical and other data. Positive cytoplasmic staining was observed in 24 cases (8.2%), 19 of which were moderately or focally positive and five (2%) of which were diffusely strongly positive (Fig. 1). The intensity of the staining was greater than any we had previously encountered in papillary thyroid carcinoma, colorectal carcinoma or melanoma and was reproduced on whole mount sections. Interestingly, we did not observe any significant nuclear staining (a finding we have come to associate with non-specific staining at other sites) in these strongly positive cases. We then macrodissected the five most strongly positive cases to ensure a neoplastic cellularity of greater than 80% and performed matrix assisted laser desorption/ionisation time of flight (MALDI-TOF) polymerase chain reaction (PCR) using an assay specifically validated to detect the BRAF V600E, V600R, V600K, and V600 M mutations. All five cases were negative for BRAF mutation. Taken together with the results of the study by Sperveslage et al.,11 our findings clearly confirm that even diffuse strong positive staining with the VE1 clone is not associated with the presence of the BRAFV600E mutation in pituitary adenomas. Therefore, we caution that BRAFV600E mutation specific immunohistochemistry should not be used diagnostically in pituitary neoplasia. Furthermore, we recommend that VE1 immunohistochemistry should be validated for each specific tumour type before it is deployed clinically.

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

80

Pathology (2014), 46(1), January

CORRESPONDENCE

Fig. 1 Serial sections stained with (A) H&E and (B) BRAFV600E immunohistochemistry with clone VE1. The intense and diffuse cytoplasmic staining for VE1, greater than we have encountered in other tumours, is emphasised.

The cause of this apparent false positive staining for VE1 which at this stage appears to only occur in pituitary adenomas is unknown. Interestingly, Sperveslage et al. found that VE1 staining was common in growth hormone and ACTH producing pituitary adenomas. When we independently performed unsupervised binary logistic regression modelling to assess the correlation between VE1 staining and various clinicopathological features, the only variables associated with VE1 positivity were ACTH and growth hormone expression ( p < 0.05). Therefore, we suspect that an as yet unknown homologous protein particularly associated with growth hormone and ACTH positive pituitary adenomas may be the cause of false positive staining for VE1. We postulate that because this protein is sufficiently homologous to cause positive staining with clone VE1, there may yet be a role for BRAF inhibitors which target the same epitope in treatment resistant acromegaly or pituitary Cushing disease, even in the absence of BRAF mutation.

5. Koperek O, Kornauth C, Capper D, et al. Immunohistochemical detection of the BRAF V600E-mutated protein in papillary thyroid carcinoma. Am J Surg Pathol 2012; 36: 844–50. 6. Bullock M, O’Neill C, Chou A, et al. Utilization of a MAB for BRAF V600E detection in papillary thyroid carcinoma. Endocr Relat Cancer 2012; 19: 779–84. 7. Long GV, Wilmott JS, Capper D, et al. Immunohistochemistry is highly sensitive and specific for the detection of V600E BRAF mutation in melanoma. Am J Surg Pathol 2013; 37: 61–5. 8. Toon CW, Walsh MD, Chou A, et al. BRAFV600E immunohistochemistry facilitates universal screening of colorectal cancers for Lynch syndrome. Am J Surg Pathol 2013; 37: 1592–602. 9. Preusser M, Capper D, Berghoff AS, et al. Expression of BRAF V600E mutant protein in epithelial ovarian tumours. Appl Immunohistochem Mol Morphol 2013; 21: 159–64. 10. Ilie M, Long E, Hofman V, et al. Diagnostic value of immunohistochemistry for the detection of the BRAFV600E mutation in primary lung adenocarcinoma Caucasian patients. Ann Oncol 2013; 24: 742–8. 11. Sperveslage J, Gierke M, Capper D, et al. VE1 immunohistochemistry in pituitary adenomas is not associated with BRAF V600E mutation. Acta Neuropathol 2013; 125: 911–2.

Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose.

DOI: 10.1097/PAT.0000000000000033

Mahtab Farzin1,2 Christopher W. Toon2,3 Adele Clarkson1,2 Loretta Sioson1,2 Anthony J. Gill1,2,4 1 Department of Anatomical Pathology, Royal North Shore Hospital, St Leonards, 2Cancer Diagnosis and Pathology Research Group, Kolling Institute of Medial Research, St Leonards, 3Histopath Pathology, North Ryde, and 4Sydney Medical School, University of Sydney, Sydney, NSW, Australia

Contact Dr A. J. Gill. E-mail: [email protected] 1. Pakneshan S, Salajegheh A, Smith RA, Lam AK-Y. Clinicopathological relevance of BRAF mutations in human cancer. Pathology 2013; 45: 346– 56. 2. Capper D, Preusser M, Habel A, et al. Assessment of BRAF V600E mutation status by immunohistochemistry with a mutation-specific monoclonal antibody. Acta Neuropathol 2011; 122: 11–9. 3. Skorokhod A, Capper D, von Deimling A, et al. Detection of BRAF V600E mutations in skin metastases of malignant melanoma by monoclonal antibody VE1. J Am Acad Dermatol 2012; 67: 488–91. 4. Andrulis M, Penzel R, Weichert W, von Deimling A, Capper D. Application of a BRAF V600E mutation-specific antibody for the diagnosis of hairy cell leukemia. Am J Surg Pathol 2012; 36: 1796–1800.

Osseous metaplasia in polypoid rectal mucosal prolapse Sir, Osseous metaplasia (OM) is rare in the gastrointestinal tract and has been reported in inflammatory rectal polyps and in solitary rectal ulcer syndrome in only a few single case reports. Herein described is a case of polypoid rectal mucosal prolapse with OM. It is proposed that the earlier cases of inflammatory polyps may all share the aetiological factor of mucosal prolapse. Rectal mucosal prolapse is a common cause of non-neoplastic polyps and prolapsing polypoid mucosal polyps can occur in the absence of a history of prolapse. Many of the histological changes typical of mucosal prolapse are the same as those in which OM is described also, although the two are rarely described together. Causes of OM and the spectrum of mucosal changes in mucosal prolapse are discussed. A 17-year-old female presented with a history of passing mucous and blood per-rectum and reported increased stool frequency. On questioning, she admitted frequent visits to the lavatory with strenuous, fruitless attempts at defecation for periods of an hour at a time because of unresolved tenesmus.

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

CORRESPONDENCE

Fig. 1 Line of OM between glands and extravasated mucin (H&E, low power).

Proctoscopy revealed a ring of raised, ‘villous’ appearing mucosa at the squamocolumnar junction associated with circumferential mucosal prolapse. Clinical photographs were not taken. Two biopsies of polypoid rectal mucosa were obtained. The laboratory received two purple coloured mucosal nodules, each 14 mm in maximum dimension. They were fixed in 10% buffered formalin, each bisected at cut up and routinely processed. Microscopically, each piece appeared similar and comprised hyperplastic rectal mucosa with crypt elongation and dilatation, and fibromuscular hypertrophy within expanded lamina propria. There were ‘caps’ of ulcerated granulation tissue at the surface and mucin extravasation into the stroma. OM was present in both pieces, within peri-crypt lamina propria and often intimately associated with extravasated mucin as well as haemosiderin at the interface of crypts and granulation tissue (Fig. 1 and 2). The appearance was consistent with the effects of mucosal prolapse and each piece resembled an inflammatory cap polyp. No viral inclusions, granulomata, dysplasia or malignancy were seen. OM is a rare pathological finding in the gastrointestinal tract and has been found in a variety of circumstances including benign and malignant conditions.1 It has been described in less than 25 benign polyps (neoplastic and non-neoplastic).2 OM has been specifically reported in inflammatory polypoid rectal mucosa on only six prior occasions,1–6 including in documented rectal mucosal prolapse,1 summarised in Table 1. The histological features are all similar with crypt dilatation,

Fig. 2 OM adjacent to gland, extravasated mucin and haemosiderin deposition (H&E, high power).

81

inflamed and ulcerated granulation tissue and OM related to crypts and/or extravasated mucin. OM often presents as an incidental finding lacking clinical significance, unless misinterpreted as a sinister finding in the setting of malignancy, e.g., carcinosarcoma1 or by falsely implying bone involvement by a pathological process. The precise aetiology remains obscure, but theories abound. Osteoblasts, which produce collagen and ground substance, may differentiate from proliferating stromal mesenchymal cells or fibroblasts, for which epithelial cells and mesenchymal cells may provide the stimulus by expressing alkaline phosphatase, bone morphometric proteins, transforming growth factors and other transcription factors. Adequate concentrations of calcium, phosphate and hydroxyl groups are required to allow subsequent mineralisation. Features such as necrosis, and mucin extravasation, which may interact with calcium salts, may be amongst a number of instigating factors. Necrosis, inflammation, pre-existing calcification, increased vascularity, mucin extravasation and ulceration are often described histologically when OM is present in gastrointestinal lesions. Repeated local trauma, including ischaemia due to prolonged straining (relevant to rectal lesions), may contribute to these changes.1–3,5 The morphological appearance of the inflammatory polyps described in Table 1 suggest they may have the unifying aetiology of mucosal prolapse. Mucosal prolapse can occur anywhere in the gastrointestinal tract, but is most often seen in the recto-sigmoid region. It also occurs at the edges of diverticular openings, at stoma sites, haemorrhoids, adjacent to tumours and polyps, and in cap polyposis. Polyps may occur in the various forms of prolapse and they may be single or multiple. Classic histological features of prolapse are crypt architectural distortion with often diamond shaped glands, serrated architecture, and fibromuscular hypertrophy of the lamina propria but the histology varies. The expansive terminology for lesions with considerable overlap of histological features—rectal mucosal prolapse/solitary rectal ulcer syndrome (SRUS), inflammatory cloacogenic polyp, inflammatory myoglandular polyp, inflammatory cap polyps, polypoid prolapsing folds of diverticular disease/prolapsing mucosal polyps and proctitis cystica profunda—are explained by the unifying pathological mechanism of mucosal prolapse. Features that are characteristic to many of these entities, e.g., ulceration, pseudomembranes, inflammation, haemorrhage, vascular congestion and ectasia, submucosal fibrosis, crypt displacement and cystic dilatation (including colitis cystica profunda) occur to varying degrees following repeated mucosal injury from prolapse, occurring on top of the basic changes, and may reflect polyps at different stages of development and contribute to diagnostic confusion as they may mask or divert attention from the classical changes.7,8 Symptoms of rectal mucosal prolapse/SRUS include per rectal passage of blood and mucus, tenesmus and fruitless straining. Prolapse may be associated with either constipation or diarrhoea, but individuals may also be asymptomatic. Therefore, the absence of a clinical history of prolapse type symptoms does not preclude that an individual polyp is prolapse related. Incorrect diagnosis continues despite recognition of the condition over a century ago.7,8 Because the histological features described in prolapse are not unique to that concept, prolapsing mucosa may mimic a number of entities.8 Diagnosis may be markedly delayed and

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

82

Pathology (2014), 46(1), January

CORRESPONDENCE

Table 1

Inflammatory polypoid rectal mucosa with osseous metaplasia in the literature

Sex

Age, years

Site

Presentation

Size

Histology Inflamed and ulcerated granulation tissue overlying mucin filled dilated glands adjacent to which there was OM Inflamed and ulcerated granulation tissue and cystically dilated glands with an island of bone at the polyp base Active chronic inflammation, crypt distortion, haemosiderin deposition, fibromuscular hypertrophy of lamina propria, OM superficial lamina propria. Diagnosis of SRUS Inflamed and ulcerated granulation tissue with OM associated with distorted glands Inflamed and ulcerated granulation tissue and elongated, branching dilated mucin filled crypts. Small focus of OM at the junction of intact crypts and granulation tissue Elongated, dilated or tortuous hyperplastic crypts superficially covered by a cap of ulcerated granulation tissue, fibrin and acute inflammatory exudate, mucin extravasation and haemosiderin. Foci of OM in proximity to mucin lakes Inflammatory granulation tissue cap overlying elongated, dilated, mucin filled crypts, mucin extravasation. Focal fibromuscular hypertrophy. OM associated with crypts at junction of mucin and granulation tissue

M

25

Posterior rectum at 10cm

Rectal bleeding

10mm

F

22

Rectum

10mm

M

17

Rectum at 12cm

Low abdominal pain, constipation alternating with occasional diarrhoea, excess mucus in stool 5 year history intermittent diarrhoea and passage of blood and mucus

M

39

Lower Rectum

Melaena

12mm

M

74

Rectum

Surveillance following prior tubular adenoma

10mm

M

21

Posterior low rectum

10 months rectal bleeding

60 mm, occupied >50% of lumen

F

17

Ano-rectal junction

Passage of mucus and blood, tenesmus

Circumferential

Polypoid, friable, ulcerated rectal mucosa

Year (reference) 1981 (3)

1992 (4)

2001 (1)

2010 (5)

2012 (2)

2012 (6)

2013 (current case)

OM, osseous metaplasia; SRUS, solitary rectal ulcer syndrome.

patients may even be commenced on steroids for presumed inflammatory bowel disease because of symptoms of diarrhoea and inflammation seen on biopsy.1 Villiform polypoid prolapse may mimic villous adenoma, serrated polyps and adenocarcinoma.7,8 Inflamed granulation tissue containing polyps can perfectly mimic almost any ulcerated polyp, including juvenile (retention) polyps. Occurrence of a polyp at the anal verge should warrant consideration that it is prolapse related.8 Like juvenile polyps (JP), Peutz–Jeghers polyps (PJP) may also bear striking resemblance to polypoid mucosa in mucosal prolapse. There is hypothetically a unifying reason for this similarity between these hamartomatous polyps and the changes of mucosal prolapse. It has been postulated from modelling, that a genetic predisposition to mucosal prolapse occurs in Peutz–Jeghers syndrome patients due to germ-line disruption of a cell polarity pathway, which is proposed to underlie the formation of the characteristic hamartomatous polyps in this condition. In this hypothesis, it is proposed that germ-line mutation of LKB1, in time leads to epithelial displacement (into the faecal stream) and secondary changes due to mucosal prolapse such as smooth muscle arborisation and gland displacement. So according to this as yet unproven hypothesis, PJP polyps would really be prolapse related polyps arising in mucosa prone to prolapse.9 Also of note, utilising a mouse knockout gene model of juvenile polyposis, disruption of bone morphometric protein (BMP) signalling was shown to induce crypts to bud off or grow perpendicular to the cryptvillous axis, architectural features which are typical in the hamartomatous polyps of this condition.9,10 Whether the

germ-line mutations in juvenile polyposis might also predispose to mucosal prolapse and contribute to polyp formation may also warrant investigation. Both a PJP11 and at least six JP2 have been reported with OM. Histological features often encountered in mucosal prolapse such as ulceration, mucin extravasation, haemosiderin deposition and inflammation are exactly the setting in which OM is so often described, so it is surprising that the two are rarely encountered or (perhaps more likely) reported together. The current case is of OM arising in inflamed polypoid rectal mucosal prolapse, and is histologically similar to the earlier reported cases of inflammatory polyps with OM, which may well also be prolapse induced. Accurate diagnosis of prolapse requires a detailed accurate history, endoscopic and histological examination.8 The possibility of mucosal prolapse should always be borne in mind as in retrospect it is often the most obvious explanation for a hitherto difficult to classify rectal polyp. Conflicts of interest and sources of funding: The author states that there are no conflicts of interest to disclose. Mark James Wilsher Douglass Hanly Moir Pathology, Department of Histopathology, Macquarie Park, NSW, Australia Contact Dr M. J. Wilsher. E-mail: [email protected]

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

CORRESPONDENCE

1. Kaushik VY, Moriarty KJ, Lipscomb GR, Bissett DL, Wells C. Osseous metaplasia associated with solitary rectal ulcer syndrome. CME Gastroenterology 2001; 4: 37–9. 2. Odium BR, Bechtold ML, Diaz-Arias A. Osseous metaplasia in an inflammatory polyp of the rectum: case report and review of the literature. Gastroenterol Res 2012; 5: 74–8. 3. Sperling MH, Friedman CJ. Osseous metaplasia in a benign colon polyp. Gastrointest Endosc 1981; 27: 198–9. 4. Castelli MF, Roberts J. Ossification is a benign rectal polyp. Am J Gastroenterol 1992; 87: 543–4. 5. Oono Y, Fu K-I, Kakamura H, et al. World J Gastrointest Surg 2010; 2: 104–6. 6. Papaconstantinou I, Karakatsanis A, Benia X, Polymeneas G, Kostopoulou E. Solitary rectal cap polyp: case report and review of the literature. World J Gastrointest Surg 2012; 4: 157–62. 7. De Petris G, Leung ST. Pseudoneoplasms of the gastrointestinal tract. Arch Pathol Lab Med 2010; 134: 378–92. 8. Singh B, McC Mortensen NJ, Warren BJ. Histological mimicry in mucosal prolapse. Histopathology 2007; 50: 97–102. 9. Jansen M, de Leng WWJ, Baas AF, et al. Mucosal prolapse in the pathogenesis of Peutz-Jehgers polyposis. Gut 2006; 55: 1–5. 10. Haramis AP, Begthel L, van den Born M, et al. De novo crypt formation and juvenile polyposis on BMP inhibition in mouse intestine. Science 2004; 303: 1604–6. 11. Narita T, Ohnuma H, Yokoyama S. Peutz-Jeghers syndrome with osseous metaplasia of the intestinal polyps. Pathol Int 1995; 45: 388–92.

DOI: 10.1097/PAT.0000000000000031

Myxoid variant of adrenocortical carcinoma: a report of two illustrative cases and a brief review of the literature Sir, Adrenocortical carcinomas represent about 0.02% of all malignant tumours. They can be either functional or non-functional and are usually diagnosed in the late stages, weighing between 100 g and 4.7 kg.1 Several subtypes of adrenal carcinomas, such as oncocytic or myxoid carcinomas, and adrenocortical carcinomas with sarcomatous areas have been described.1 However, it is very difficult to perform their differential diagnosis and thus a proper correlation between their clinical, histological, and immunohistochemical features is necessary. In this paper, we present two myxoid adrenocortical carcinomas (MACs) with lipomatous metaplasia and their clinical, histopathological, molecular and immunohistochemical features. To date, only 27 cases of MAC have been published;1–11 one of them was also associated with lipomatous metaplasia,2 and the cases described here are the second and third case, respectively. To our knowledge, no data about K-ras mutations and/or RET expression in these myxoid tumours have been published yet. Case 1 was a 36-year-old previously healthy female who presented with moderate weight gain (5 kg in 5 months) and amenorrhoea for 6 months. Her serology showed slight anaemia (haemoglobin 10.7 g/dL, haematocrit 31.6%). The physical examination evidenced moderate obesity (body weight 95 kg, height 1.70 m) with particular predisposition of fat on the trunk and the back of the neck (buffalo hump). Extreme dryness of skin and facial and truncal acne were also present. The serum cortisol level was moderately high (130 mg/dL). These aspects, correlated with amenorrhoea, suggested Cushing syndrome. A non-tender abdomen with voluntary guarding and a smooth

83

solid mass was palpated in the left hypochondrium. The abdominal ultrasound examination revealed a 16.7  9.81 cm sized inhomogeneous hypoechogenic mass at the upper pole of the left kidney. Abdominal computed tomography confirmed the suspicion of a left adrenal tumour without involving the left kidney; the right adrenal gland was in normal limits. The renal function was normal. Left adrenalectomy was performed. Case 2 was 41-year-old previously healthy male who presented with moderate fatigue, abdominal pain, and minimal weight gain (3 kg in 4 months). His serology also showed a slight anaemia, without other alterations; both serum cortisol and ACTH levels were within normal limits (75 mg/dL and 42.4 pg/mL). The abdominal ultrasound and computed tomography examinations revealed a 12.8  7 cm sized right adrenal tumour without involving the right kidney. Right adrenalectomy was performed. Macroscopic examination of the surgical specimens showed a 17  11  4 cm sized tumour weighing 1.2 kg (Case 1) and a 540 g, 13  7  4 cm sized tumour (Case 2). Both tumours were encapsulated and presented similar features. The surface was partially bosselated but the capsule was intact. On the cut section, the tumour had a nodular aspect, with grey gelatinous and yellowish solid areas admixed with haemorrhages and necrotic zones (Fig. 1). No capsular invasion and/or crossing were observed in either case. In Case 1, histopathological examination showed a heterogeneous aspect of the well-vascularised tumour. About 20% of the tumour consisted of tumour clusters and cords, whereas the other 80% showed a characteristic diffuse architecture. The tumour cells were mostly large, polygonal-shaped, with eosinophil cytoplasm and large nuclei, some of them pallisading around the blood vessels. The vacuolar cells were very rarely observed, comprising less than 25% of the tumour. Necrosis and haemorrhages were expanded, large fibrotic septa were noticed, and capsular, sinusoidal and vascular invasion was seen, without crossing the tumour capsule. High nuclear pleomorphism, giant bizarre nuclei, atypical mitoses, and a mitotic rate >10/50 high power fields (HPF) were also noticed. A characteristic feature was described in the tumour stroma; 20% of tumour had a myxoid aspect, being Alcian blue positive and negative for periodic acid-Schiff (PAS) stain and floating clusters of tumour cells were present in this acellular myxoid background. Among the tumour cells, focal metaplastic lipomatous zones were evidenced (Fig. 1). In Case 2, microscopic examination revealed similar features with some differences; the myxoid stroma was present in about 40% of tumour and vascular/sinusoidal invasion was not seen. In both cases, the tumour cells displayed diffuse positivity for vimentin, inhibin, CD56, and Melan A. Focal expression of synaptophysin, NSE, and RET oncoprotein was also observed. Chromogranin, keratin AE1/AE3, EMA, CEA, calretinin, HMB45 and S100 were negative (Fig. 1). The Ki-67 proliferative index was about 10% and the p53 nuclear positivity was higher than 80%. Molecular examination did not detect mutations in K-ras (codon 12 and 13) in exons 18 through 21 of epidermal growth factor receptor (EGFR) genes. The final diagnosis was MAC with lipomatous metaplasia in both of the cases based on the macro- and microscopic features, as correlated with the clinical aspects and immunohistochemical profile. The tumour was of the functional type in the first case and the non-functional type in the second case. No adjuvant therapy was performed.

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

84

CORRESPONDENCE

Pathology (2014), 46(1), January

Fig. 1 (A,B) Myxoid adrenocortical carcinoma is an encapsulated tumour with large haemorrhagic and necrotic areas on the cut section. (C) The histopathological examination shows an inhomogeneous aspect, the tumour cells floating in an Alcian blue-positive myxoid stroma; (D) some of the cells present a diffuse arrangement and (E) metaplastic lipomatous areas are admixed through the tumour cells. The tumour cells are marked by (F) inhibin, (G) RET oncoprotein, and (H) CD56.

In Case 1, no post-operative complications, metastases or recurrences were reported 6 months following surgery. In Case 2, solitary liver metastasis occurred 2 years after surgery and liver metastasectomy was performed. Four years after the first diagnosis, local recurrence was reported and the residual tumour was surgically removed. The patient is still alive 5 years after the initial diagnosis. MAC was first described by Tang et al. in 1979 based on the light and electron microscope.3 Since then, another 26 cases have been described1–11 but the diagnosis of this tumour remains a challenge. Differential diagnosis should be especially based on light-microscopic features associated with the clinicopathological features. The immunohistochemical stains are necessary to confirm the diagnosis. Weiss’s scoring system and its revised index are used in clinical practice to differentiate adrenocortical carcinomas from their benign and borderline counterparts.1 The nine Weiss’s histological criteria used to appreciate the malignancy are as follows: diffuse architecture in at least one-third of the tumour; clear cells in less than 25% of the tumour; the presence of necrosis, vascular invasion, capsular invasion, sinusoidal invasion, high Fuhrman’s nuclear grade, atypical mitoses, and a mitotic count of >5/50 HPF.1 The malignancy involves the presence of at least three criteria, two of them indicating a borderline neoplasm. However, in the present cases,

establishing the malignancy was easily done (8 of 9 criteria and 6 of 9 criteria were evidenced in Cases 1 and 2, respectively). However, the differentiation between MAC, malignant phaeochromocytoma, and a metastatic tumour was challenging. Tumour size and weight indicated a primary tumour and the immunohistochemical profile excluded metastasis. The last tumour that was taken into account for a complete differential diagnosis was phaeochromocytoma. Clinically, functional MAC can be associates with feminisation with gynaecomastia in males/hyperoestrogenism or androgenism in females, Conn syndrome1 or, as in one of our cases, Cushing syndrome. At the same time, Cushing syndrome can also be accompanied by hypertension, which is similar to phaeochromocytoma. In case of non-functional MAC the clinical features are non-specific. Microscopically, the architecture of the tumour cells can be similar in haematoxylin and eosin staining in both MAC and phaeochromocytoma. However, the myxoid areas are characteristic for MAC, being present in 5–100% of the tumour.1–11 Establishment of a proper targeting therapy is mandatory as the MAC rapidly progresses and leads to lung, liver, and/ or abdominal metastases. To date, a specific chemotherapeutic protocol does not exist. Making a synthesis of the literature data, Singh et al.12 concluded that, in the case of adrenocortical

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

CORRESPONDENCE

carcinomas, independent of their histological subtype, surgery is the therapy of choice in resectable tumours and the residual disease can be destroyed by using cisplatin, doxorubicin, etoposide, streptozocine, and mitotane, but their gastrointestinal and neurological toxicities have been frequently reported.1 One of the most recent studies takes into account the possibility of antiEGFR therapy in MAC with promising results and minimal side effects in other tumours, such as colorectal and pulmonary carcinomas.4 Although the present cases show the absence of K-ras mutation, which is an inclusion criteria for this therapy, the absence of EGFR mutation, which is in line with the Chinese study,4 again proves that these drugs cannot be an alternative therapy for MAC in the future. This confirms that the EGFR/Kras pathway is not involved in the carcinogenesis of MAC. However, the RET positivity, yet to be reported in the literature, could indicate a possible response of these tumours to regorafenib, a novel oral drug that seems to be an anti-angiogenic and anti-oncogenic kinase inhibitor that acts against KIT, RET, RAF, and VEGF genes.13 In conclusion, MAC is rarely observed but its diagnosis remains difficult; immunohistochemical stains are useful only to confirm the diagnosis established by using the optical microscope. Despite the aggressive nature of MACs and high rate of recurrence, survival seems improved by metastasectomy. Other molecular pathways and BRAF mutations should be studied in these tumours to identify new molecular targets that can be used to treat them. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Simona Gurzu1 Zoltan Szentirmay3 Tivadar Bara2 Tivadar Bara Jr2 Ioan Jung1 1 Department of Pathology, 2Department of Surgery, University of Medicine and Pharmacy of Tirgu-Mures, Romania; and 3Center of Tumors, National Institute of Oncology, Budapest, Hungary

Contact Dr S. Gurzu. E-mail: [email protected] 1. De Krijger R, Papathomas TG. Adrenocortical neoplasia: evolving concepts in tumorigenesis with an emphasis on adrenal cortical carcinoma variants. Virchows Arch 2012; 460: 9–18. 2. Izumi M, Serizawa H, Iwaya K, Takeda K, Sasano H, Mukai K. A case of myxoid adrenocortical carcinoma with extensive lipomatous metaplasia. Arch Pathol Lab Med 2003; 127: 227–30. 3. Tang CK, Harriman BB, Toker C. Myxoid adrenal cortical carcinoma: a light and electronic microscopic study. Arch Pathol Lab Med 1979; 103: 635–8. 4. Zhang J, Sun J, Liang Z, Gao J, Zeng X, Liu T. A study of the clinicopathologic features and EGFR gene status of ten chinese cases. Am J Clin Pathol 2011; 136: 783–92. 5. Suresh B, Kishore TA, Albert AS, Joy A. Myxoid adrenal cortical carcinoma – a rare variant of adrenocortical carcinoma. Indian J Med Sci 2005; 59: 505–7. 6. Brown FM, Gaffey TA, Wold LE, Lloyd RV. Myxoid neoplasm of the adrenal cortex: a rare histologic variant. Am J Surg Pathol 2000; 24: 396–401. 7. Papotti M, Volante M, Duregon E, et al. Adrenocortical tumors with myxoid features: a distinct morphologic and phenotypical variant exhibiting malignant behavior. Am J Surg Pathol 2010; 34: 973–83. 8. Hsieh MS, Chen JH, Lin LW. Myxoid adrenal cortical carcinoma presenting as primary hyperaldosteronism: case report and review of the literature. Int J Surg Pathol 2011; 19: 803–7.

85

9. Raparia K, Ayala AG, Sienko A, Zhai QJ, Ro JY. Myxoid adrenal cortical neoplasms. Ann Diagn Pathol 2008; 12: 344–8. 10. Karim RZ, Wills EJ, McCarthy SW, Scolyer RA. Myxoid variant of adrenocortical carcinoma: report of a unique case. Pathol Int 2006; 56: 89–94. 11. Forsthoefel KF. Myxoid adrenal cortical carcinoma. A case report with differential diagnostic considerations. Arch Pathol Lab Med 1994; 118: 1151–3. 12. Singh O, Gupta SS. Giant adrenal cortical carcinoma. J Kidney Dis Transpl 2011; 22: 153–5. 13. Grothey A, Van Cutsem E, Sobrero A, et al. CORRECT study groupRegorafenib monotherapy for previously treated metastatic colorectal cancer (CORRECT): an international, multicentre, randomised, placebocontrolled, phase 3 trial. Lancet 2013; 381: 303–12.

DOI: 10.1097/PAT.0000000000000035

Bilateral, multicentric low-grade adenosquamous carcinomas of the breast, each arising within a separate radial scar/ complex sclerosing lesion Sir, Low grade adenosquamous carcinoma (LGASC) is a rare form of mammary carcinoma. Only a handful of reports of this tumour type have been published since Rosen and Ernsberger first described the entity.1 Herein we describe a case of bilateral, multicentric LGASCs occurring in a 45-year-old woman, each arising in a radial scar or complex sclerosing lesion (CSL), necessitating mastectomy. A 45-year-old woman presented with a 2-year history of a small nodule in the right breast at the 9.30 o’clock position, 3 cm from the nipple. She was undergoing biennial screening because of a family history of a second degree relative with breast cancer in her fifth decade. In 2011 this nodule was shown to correspond on ultrasound scan to a subtle area of hypoechoic change 7  4 mm, which was slightly irregular in outline and sonographically indeterminate (Fig. 1A). Ultrasound examination of the left breast was normal. Bilateral mammograms showed no abnormality. Fine needle aspiration (FNA) was performed and slides prepared from three passes. The smears were highly cellular and contained biphasic groups of cohesive epithelial and myoepithelial cells, abundant bare nuclei and pink loose stroma and as a result were interpreted as originating from a fibroadenoma (Fig. 1B). However, on more recent review there were also variably discohesive monomorphic epithelial cells (Fig. 1C), and very occasional compact elongated groups of cells were present, that were somewhat squamoid and spindled (Fig. 1C,D). Repeat ultrasound and mammography were performed in 2013. The lesion in the right breast at 9.30 o’clock, 3 cm from the nipple was again seen on ultrasound, measured at 6  4 mm and an irregular outline noted with ill-defined margins. It was essentially unchanged. Bilateral mammograms were again normal. Core biopsy was performed because of the indistinct margin. The specimen comprised two core biopsies of breast tissue measuring 17  1 mm and 18  1 mm. In each there were benign ducts, some radiating within fibroelastotic stroma as well as excess stromal mucin, the appearance suggesting a CSL. However, in each, there was a subtly different population of compact ducts with vague squamous metaplasia and surrounding spindle cells forming a ‘lamellar cuff’. These ducts and cords extended almost imperceptibly between the benign ducts.

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

86

Pathology (2014), 46(1), January

CORRESPONDENCE

A

B

C

D

Fig. 1 (A) Ultrasound from 2011 with indeterminate lesion. This remained essentially unchanged by 2013 (not shown). (B) Fine needle aspiration (FNA) cytology 2011 with biphasic epithelium, bare nuclei and pink loose stroma mimicking a fibroadenoma (May-Gru¨nwald Giemsa). (C) FNA cytology 2011 showing some disassociation, pale blue epithelial cells and a compact group of elongated cells (May-Gru¨nwald Giemsa). (D) FNA cytology 2011 showing compact group of elongated cells (May-Gru¨nwald Giemsa).

Immunohistochemistry was performed. Myoepithelial cells were confirmed around the majority of the ducts with p63 and calponin immunohistochemical (IHC) markers. The infiltrating process showed an incomplete myoepithelial component. P63, 34bE12 and CK5/6 IHC markers highlighted the squamous component. Calponin decorated the spindle cells around the abnormal ducts and cords. Tumour cells almost all lacked oestrogen (ER) and progesterone (PR) receptor expression. The features were reported as highly suspicious for LGASC. A diagnostic excision biopsy of the right breast was performed, the specimen measuring 30  17  10 mm and weighing 3 g. A CSL was present (Fig. 2A) with characteristic radiating benign ducts within fibroelastotic vascularised stroma intermixed within which LGASC was confirmed (Fig. 2B,C), extending to several margins. It measured 25 mm in maximum dimension and two satellite foci were also present, each comprising only several neoplastic ducts/cords, one residing within an additional margin specimen. The tumour foci showed the same immunohistochemical profile as seen in the core biopsy, including p63 expression by the myoepithelial and squamous components (Fig. 2D), lamellar cuffs of spindle cells were highlighted around tubules by calponin (Fig. 2E) and expression of high molecular weight cytokeratins by the squamous elements (Fig. 2F). A small intracanalicular fibroadenoma was present at a margin. On review this was not thought to be the source of the 2011 FNA. Three weeks after the excision biopsy, a skin and nipplesparing completion right mastectomy was performed. The specimen measured 105  100  25 mm, weighing 110 g. Changes attributable to the recent surgery were seen in the lateral and upper outer quadrant (9–11 o’clock position) extending near to the nipple bed. In the lower outer quadrant at the 5 o’clock position, 2 cm from the nipple bed, a poorly circumscribed, irregularly shaped, pale yellow lesion was seen, measuring 9  9  8 mm. Histologically, at the site of recent

surgery there was a collapsed cavity lined by metaplastic squamous epithelium and there was florid reparative response, and no convincing residual previously sampled tumour. Two consecutive sagittal slices from the 5 o’clock position showed three if not four small radial scars over a 15 mm span, each displaying within them characteristic features of LGASC. Each was completely excised. There was a small fibroadenoma, possibly the remainder of that seen in the first excision specimen. A sentinel node biopsy was negative for metastatic carcinoma examined at four levels through the node and with an AE1/AE3 IHC marker performed on one section parallel to the second level. Four weeks later the patient underwent a contralateral (left) skin and nipple sparing prophylactic mastectomy. The specimen measured 130  100  25 mm, but was not weighed. The cut surface was described as ‘fibrofatty and rubbery with numerous pink lobules’. Microscopically, there were at least six radial scars, each with convincing foci of LGASC, resembling that seen in the right breast. They ranged from 0.5 mm to 3 mm and were present both medial and lateral to the nipple site over a distance of 50 mm. A 6  4 mm fibroadenoma was present at the nipple site. Margins were clear of tumour. LGASC is a rare and unique form of relatively indolent, rarely metastasising mammary carcinoma. Characterised by distinct glandular and variable squamous components it is regarded as a metaplastic carcinoma. It most often presents as a palpable mass with a wide range of age at diagnosis.1–3 Lesions are typically less than 5 cm at diagnosis2 and may be smaller if detected by screening mammogram.3 The cut surface of LGASC has been described as white or pale yellow.1 The imaging appearance of LGASC was most recently reported in a series of 10 cases.3 The tumours were most commonly in the periareolar region and presented as a solid mass with a variety of appearances on mammogram and with

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

CORRESPONDENCE

A

B

C

D

E

F

87

Fig. 2 Excision specimen. (A) Complex sclerosing lesion with low grade adenosquamous carcinoma (LGASC) (H&E). (B) LGASC (H&E). (C) LGASC with mantle of spindle cell stroma (H&E). (D) p63 highlighting variable myoepithelial layer, and highlighting the squamous component also. (E) Calponin highlighting normal ‘garland’ of myoepithelial cells below, whereas LGASC showed incomplete myoepithelial layer and a ‘lamellar cuff’ of spindle cells around ducts above. (F) 34bE12 high molecular weight keratin highlighting squamous cells.

ultrasound, and there were ‘no unique imaging features’. The imaging appearance of LGASC overlaps with other neoplasms, both benign and malignant so pathological examination is essential for diagnosis.3 Definitive diagnosis is rarely made prior to excision due to the inherent limitations of FNA and core biopsy and the subtle histological features of the lesion, which mimics more common benign and low grade lesions such as sclerosing adenosis, CSL/radial scar, nipple adenoma and tubular carcinoma.2 As in the current case, immunohistochemical markers significantly aid the diagnosis. An inconsistent myoepithelial layer is present and highlighted with calponin and p63 IHC markers, the latter also marking the squamous component. p63 also helps delineate the boundaries of the tumour by highlighting the stromal metaplastic spindle cell component. Additionally, surrounding neoplastic ducts, a ‘lamellar cuff’ of calponin and smooth muscle actin expressing stromal spindle cells is a helpful clue,2 as opposed to the characteristic garland-like pattern staining of myoepithelial cells investing benign ducts. ER and PR expression is typically negative, a helpful clue as benign or other low grade mimics such as tubular carcinoma express these hormone receptors. The FNA cytology characteristics of LGASC have been described.2,4 Difficulty lies in the fact that LGASC lacks the characteristic features of malignancy and mimics benign lesions, in that the lesion is biphasic and myoepithelial cells are present. Smears may be variably cellular, with angulated sheets and tubular structures of epithelial cells, which may show hints at central squamous differentiation. There are background naked nuclei which may be atypical as well as spindle cells representing the stromal component. Single atypical epithelial cells may also be seen. Immunohistochemistry with high molecular weight cytokeratin may aid in diagnosis.2,4 Sironi et al.4 described the presence of fragments of amorphous elastic tissue. Elastosis was present in each of the tumours in the current case as they each resided within a radial scar/CSL, although elastic fragments were not noted in the FNA sample.

Distinction from benign elements is made more challenging because as well as mimicking benign entities, LGASC may itself arise within benign processes such as CSL (as in the current case), as well as intraduct papilloma, nipple adenoma and adenomyoepithelioma, lesions which themselves may show squamous metaplasia.2,5,6 Denley and colleagues5 described four cases of LGASC each arising within a CSL, the LGASC component lacking ER expression. Of the 33 cases of metaplastic spindle cell breast tumours associated with fibrosclerotic lesions that Goggi and colleagues described,6 this included 11 cases of LGASC, five of which were recurrent lesions not recognised in the initial biopsies, which were reported as a CSL or papilloma alone. One case of LGASC with a prominent glandular component metastasised to a lymph node presenting as a solitary squamous cyst. Other lesions such as tubular carcinoma are more commonly described within or associated with radial scars/CSL. Whether they are direct precursors or only associated with increased risk of malignancy is debated and as LGASC is so rare, how many of these arise without the features of a radial scar/CSL is also unclear. Carcinomas may also alternatively produce a stromal reaction that mimics a radial scar due to central sclerosis and elastosis.5 A CSL with prominent adenosquamous proliferation (ASP) was regarded as LGASC by Agrawal et al.7 In the current case, the diagnosis of LGASC in the right breast was relatively straightforward. In the left breast the foci of ASP in each radial scar were of varying extent, but they were permeative with variable insinuation between ducts extending out from a central nidus. Moreover, the basic morphological and immunohistochemical profile was of LGASC, so they were regarded as such. The distinction between sclerosing lesions with prominent ASP and LGASC is highly subjective and moreover, probably artificial. The laboratory has received four other excision biopsy specimens with bona fide LGASC in the last decade. Of these, one arose within a CSL and another emanated from an intraduct papilloma and was an incidental 0.9 mm lesion in a case with

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

88

CORRESPONDENCE

invasive ductal carcinoma. The current case is unique as disease is bilateral and multicentric arising in multiple radial scars and a CSL. Complete excision is the required form of treatment as recurrence occurs relatively frequently, presumably due to incomplete or marginal excision. The current case suggests another reason for a high recurrence rate may be tumour multifocality. Although LGASC is rarely metastasising, mastectomy may be required due to locally aggressive behaviour1–3 and in the instance of multiple tumours. Mastectomy was clearly the correct choice in the current case. As the histology of LGASC is so subtle, careful assessment of margins is required. Adjuvant therapies such as chemotherapy and radiotherapy may have no role, following adequate local excision.3 Acknowledgements: Great thanks are owed to Drs Elizabeth Sinclair, Ivan Burchett and Ian Burgess. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Mark James Wilsher1 Kylie Snook2 1

Douglass Hanly Moir Pathology, Macquarie Park, and Consultant Breast Surgeon, The Poche Centre, North Sydney, NSW, Australia 2

Contact Dr M. J. Wilsher. E-mail: [email protected] 1. Rosen PP, Ernsberger D. Low-grade adenosquamous carcinoma. A variant of metaplastic mammary carcinoma. Am J Surg Pathol 1987; 11: 351–8. 2. Soo K, Tan PH. Low-grade adenosquamous carcinoma of the breast. J Clin Pathol 2013; 66: 506–11. 3. Scali EP, Ali RH, Hayes M, Tyldesley S, Hassell P. Low-grade adenosquamous carcinoma of the breast: imaging and histopathologic characteristics of this rare disease. Can Assoc Radiol J 2013; Feb 13: (Epub ahead of print). 4. Sironi M, Lanata S, Pollone M, Saro F. Fine-needle aspiration cytology of low-grade adenosquamous carcinoma of the breast. Diagn Cytopathol 2012; 40: 713–5. 5. Denley H, Pinder SE, Tan PH, et al. Metaplastic carcinoma of the breast arising within complex sclerosing lesion: a report of five cases. Histopathology 2000; 36: 203–9. 6. Gobbi H, Simpson JF, Jensen RA, Olson SJ, Page DL. Metaplastic spindle cell breast tumors arising within papillomas, complex sclerosing lesions, and nipple adenomas. Mod Pathol 2003; 16: 893–901. 7. Agrawal A, Saha S, Ellis IO, et al. Adenosquamous carcinoma of breast in a 19 years old woman: a case report. World J Surg Oncol 2010; 8: 44.

DOI: 10.1097/PAT.0000000000000034

A low grade PIN-like neoplasm of the transition zone immunohistochemically negative for basal cell markers: a possible example of low grade adenocarcinoma with stratified epithelium Sir, In 2003 Amin et al. reported in a conference abstract a potential pitfall in the recognition of invasive prostatic adenocarcinoma,

Pathology (2014), 46(1), January

i.e., circumferential perineural invasion seen in adenocarcinoma with the features of micropapillary high grade prostatic intraepithelial neoplasia (HGPIN).1–3 In 2006 Hameed and Humphrey were the first to fully describe an unusual subset of prostatic adenocarcinoma composed of single glands lined by stratified non-cribriform epithelium and to highlight that this lesion mimics HGPIN.4 In 2008 Tavora and Epstein termed this pattern as ‘PIN-like prostatic duct adenocarcinoma’.5 We report herein a case of low grade PIN-like lesion of the transition zone of the prostate associated with benign prostatic hyperplasia (BPH) and immunohistochemically negative for basal cell markers. We believe that it could be an example of low grade adenocarcinoma with stratified epithelium. The case was a 73-year-old man with a clinical diagnosis of bladder outlet obstruction and with progressive worsening of the urinary symptoms. The total serum prostate specific antigen (PSA) was 4.21 ng/mL with a free-to-total ratio of 20%. Digital rectal examination (DRE) was negative. Transrectal ultrasound (TRUS) evaluation did not show hypoechoic lesions. The patient refused any pharmacological treatment for fear of side effects, such as impotency. Suprapubic simple prostatectomy was performed to relieve the patient from the urinary symptoms. The patient was followed up for 5 years after the operation. The urinary symptoms improved. The PSA decreased to 1.2 ng/mL and remained stable until he was last seen by the clinicians. DRE remained negative. TRUS did not show changes in the residual prostate. The diameters of the surgical specimen were 7  5  4 cm and the weight was 80 g (the volume of the corresponding ellipsoid was 87.92 cc). The specimen was totally sliced at an interval of 4 mm. On the cut surface the parenchyma was whitish in colour, nodular in appearance with microcystic areas. A well demarcated yellow nodule measuring 6  4 mm (volume 0.4 cc) was seen in one of the slices. The area was 5 mm away from external surface of the specimen. Initial generous sampling of the specimen was carried out, including the entire yellow area. Additional sampling was performed before signing out the pathology report. Histologically, the initial and additional tissue samples showed the features of BPH with focal cystic dilatation of the glands. Mild focal chronic inflammation was present in the periglandular stroma. The yellow nodule seen macroscopically was composed of tightly packed glands, separated by little stroma, and with a well-defined peripheral border (Fig. 1A). The mean greatest external diameter of the glands was 269 mm, the diameter being below 300 mm in 80% of the glands, smaller than that of the non-cystically dilated non-neoplastic glands of the surrounding parenchyma (mean greatest external diameter 526 mm; greater than 300 mm in all glands). The glands, none of them dilated, were lined by epithelium 3-to-4 cells thick, the cells showing crowded nuclei and scant cytoplasm (Fig. 1B,C). The epithelial architecture was flat, tufted and micropapillary in 60%, 30% and 10% of the glands, respectively (some with more than one pattern). The nuclei were mostly round, with a mean greatest diameter of 9.28 mm (the diameter of the nuclei in the non-neoplastic nuclei being 7.29 mm, and 8.57 mm in a HGPIN case not part of this study) (see Table 1 and Fig. 2A–F for additional comparisons) and showed an irregularly distributed chromatin. The nucleoli, visible in approximately 40% of the nuclei, were small and prominent in 3% of the nuclei (Fig. 1B,C). The lumen of 10% of the glands contained pink dense secretion. Necrosis was not present. Rare mitoses were seen (Fig. 1C).

Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.

CORRESPONDENCE

A

B

D

89

C

E

F

Fig. 1 (A) Histological section showing the lesion that macroscopically appeared of yellow colour (dotted circle). The parenchyma surrounding the circled area shows the features of benign prostatic hyperplasia (BPH) with cystically dilated glands. (B) Glands separated by scant stroma and lined by stratified epithelium with tufted and flat patterns. (C) Gland lined by stratified epithelium with a mitotic figure (arrowed). The nuclei are mostly round in shape with chromatin irregularly distributed and small nucleoli. Pink secretion is present in the lumen. (D) p63 immunohistochemistry. The glands are negative, i.e., a basal cell layer is not present. p63 positive cells are present in the normal gland adjacent to the negative cells (positive internal control). (E) 34bE12 immunohistochemistry. The glands are negative, i.e., a basal cell layer is not present. 34bE12 positive cells are present in the normal gland adjacent to the negative cells (positive internal control). (F) AMACR immunohistochemistry. The gland shows strong cytoplasmic staining.

The immunohistochemical investigation showed that the epithelial lining was PSA positive and it was negative for the basal cell markers p63, high molecular weight cytokeratin (34bE12) (Fig. 1D,E) and cytokeratin 5/6. The normal glands in the surrounding parenchyma or entrapped within the lesion showed variable expression of the basal cell markers in cells in basal position. The cells were alpha-methylacyl coenzyme A racemase (AMACR) positive in a dot-like fashion in 90% of glands (Fig. 1F), the surrounding BPH tissue being negative. S100 and smooth muscle actin were negative. Table 1

The initial differential diagnosis was low grade PIN versus basal cell hyperplasia. However, the results of the immunohistochemical studies did not support the diagnosis of low grade PIN and excluded that of basal cell hyperplasia for the lack of basal cell marker expression. A diagnosis of ‘low grade PIN-like adenocarcinoma of the transition zone, occupying less than 5% of the specimen’ was made, based on the literature on the differential diagnosis and mimickers of PIN.2 – 5 The Gleason score was not included in the pathology report, even though a comment

Quantitative analysis in the current case and in the surrounding BPH tissue

Type of lesion BPH (from the current case) (Fig. 2A) Lesion in current case (Fig. 2B) Low grade PIN (Fig. 2C) High grade PIN (Fig. 2D) Acinar adenocarcinoma, Gleason score 2 þ 2 ¼ 4 (Fig. 2E) Ductal adenocarcinoma (Fig. 2F)

Mean greatest external diameter of the glands, mm 526 269 397 384 158 Not calculated

Mean nuclear greatest diameter*, mm (SD) 7.29 9.28 7.74 8.57 9.72 12.63

(0.78) (0.85) (0.7) (0.96) (1.05) (1.19)

% of nuclei with prominent nucleoli

Myxoid variant of adrenocortical carcinoma: a report of two illustrative cases and a brief review of the literature.

Myxoid variant of adrenocortical carcinoma: a report of two illustrative cases and a brief review of the literature. - PDF Download Free
2MB Sizes 0 Downloads 0 Views