Urgent cerebrospinal fluid analysis: is it necessary?

Pathology (December 2014) 46(7), pp. 649–670

CORRESPONDENCE Urgent cerebrospinal fluid analysis: is it necessary? Sir, Rapid diagnosis and treatment is imperative to improve morbidity and mortality in bacterial meningitis. In general, history and physical examination alone lack sufficient sensitivity and specificity to confirm or exclude the diagnosis. Hence, patients are often hospitalised and treated empirically until additional laboratory results are available. Cerebrospinal fluid (CSF) biochemistry and microscopy improve the diagnostic yield to a certain extent. Predictive models which incorporate a combination of history, examination findings and laboratory results are available to improve the diagnostic accuracy of bacterial meningitis.1–3 Although these predictive models have good positive and negative predictive values, no test reliably excludes bacterial meningitis. Thus, physicians rely on additional tests such as culture and molecular testing, which increasingly can be performed in a real time manner, before making a definitive diagnosis. Pending these definitive results, most clinicians would continue empiric therapy regardless of initial CSF biochemistry and microscopy results. Data regarding the role or need for urgent CSF cell count and its influence on patient management in the current molecular era are lacking. Urgent processing of samples continues to be expected from diagnostic laboratories despite considerable inter and intrapersonal variability in CSF cell count results, changing cell count dynamics in early compared to later samples (especially in viral meningitis) and the influence of host immune responses on counts such that normal cell counts fail to rule out meningitis.4 Current automated systems are unable to perform counts accurately,5 necessitating ongoing manual processing. In laboratories that do not operate 24 hours / 7 days, this leads to scientific staff call-backs after-hours resulting in disruption to the routine work flows and additional laboratory costs. Following ethics approval, we conducted a retrospective study at the Royal Prince Alfred Hospital (RPAH), Sydney, to determine whether these CSF call-backs after-hours influenced clinical management. All CSF call-backs (i.e., between 10pm and 6am) for an 8 month period between January and August 2013 were identified through the laboratory information system, as were CSF parameters [gram stain, white cell counts, subsequent culture and polymerase chain reaction (PCR) results]. After excluding traumatic taps, a leukocyte count of 5 106/L was regarded as normal, except in patients under the age of one month in whom 20 106/L was considered normal. Demographic data, location [e.g. Emergency Department (ED) versus ward] of the patient at time of lumbar puncture (LP), indication for LP and final discharge diagnosis were retrieved from the medical records. Similarly, management decisions were retrieved and defined as (1) definite when direct documentation linked CSF cell count to alteration in prescribed antimicrobials, patient discharge or forming an alternative diagnosis; and (2) probable when the aforementioned actions occurred in the absence of a direct documented link. In total, 71 CSF call-backs occurred during the 8 month study period, equating to at least two call-backs weekly. Of these, Print ISSN 0031-3025/Online ISSN 1465-3931

#

68 (96%) LPs were performed in ED. Three (4%) were performed on the ward following two direct ward admissions and one request from neurosurgical intensive care. Patients’ ages ranged from 0 to 79 years with a male predominance (62%). Seventeen paediatric (24%) after-hours LPs were performed, of which the majority (n ¼ 15) were under the age of one. Overall, normal CSF parameters were observed in 61 (86%) patients. Two of these patients (2/61; 3%) had a confirmed microbiological diagnosis based on a positive PCR result for enterovirus. In the remaining LPs (10/71; 14%), the CSF leukocyte counts ranged from 12 to 5508 106/L. Three of these patients (3/10; 30%) had an established microbiological diagnosis; positive enterovirus and herpes simplex virus PCR one each, and a positive Streptococcus pneumoniae culture (n ¼ 1). A provisional diagnosis of a central nervous system (CNS) infection requiring a LP was made in only 72% of cases with overall management altered based on the CSF cell count in 36 (51%) patients (Table 1). In most cases, CSF cell count assisted in ruling out a CNS infection (n ¼ 34) with 13 of the 34 patients subsequently discharged directly from ED. Alternatively, an abnormal cell count was used to rule in CNS infection in two patients. In contrast, 35 (49%) patients did not have an alteration to management based on CSF leukocyte count. A total of 26 patients were continued on empiric treatment until further results became available (i.e., negative CSF culture and PCR). This included 14 paediatric patients who had LPs performed as part of ‘septic work-up’. In addition, in patients with a provisional diagnosis of subarachnoid haemorrhage (SAH; n ¼ 10), further testing was based on the presence of CSF xanthochromia rather than red cell count in 80% of cases. One patient from the neurosurgical intensive care unit had no indication for cell count as the LP was performed to improve CNS pressure only. Call-backs of microbiology laboratory scientists were helpful in the majority (51%) of patients. In patients with a high index of suspicion for a CNS infection (based on the documented LP indication), a normal CSF cell count facilitated an alternate diagnosis, despite two microbiologically confirmed infections having normal CSF parameters. Similarly, in patients with a low pre-test probability for meningitis, although the need for LP could be questioned, a normal CSF cell count facilitated early patient discharge and prevented unnecessary antibiotic usage. On the other hand, 37% of patients were admitted for empiric treatment regardless of the initial LP indication or microbiological parameters. No clinical or microbiological features were identified which would allow us to distinguish these patients from those whose therapy was altered. In contrast, when the provisional diagnosis was SAH, xanthochromia alone determined patient management as microbiological testing including red cell clearance is unable to reliably distinguish between SAH and traumatic tap.6,7 Thus if the laboratory could ensure accurate ordering, call-backs could be averted for this indication. Similarly, LPs as part of ‘routine septic work-up’ in paediatrics could be processed during normal hours as all patients received antibiotics for 24 to 48 hours irrespective of initial CSF cell count. However, given the sensitive nature of medical management of children, a

2014 Royal College of Pathologists of Australasia

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

650 Table 1

Pathology (2014), 46(7), December

CORRESPONDENCE

Number of patients in whom the management decision was altered based on initial cerebrospinal fluid cell count Normal CSF microbiological parameters

CSF pleocytosis (WCC 12–5508 106/L)

Total

33 28 61z

3 7 10§

36* (51%) 35{ (49%) 71

Management decision altered by initial CSF parameters Yes No Total *

Appropriate treatment commenced based on CSF results (n ¼ 2), cell count used to rule out CNS infection (n ¼ 21) and discharged from emergency department based on CSF results (n ¼ 13). { Treatment continued until PCR/culture results were available (n ¼ 26), only xanthochromia was necessary (n ¼ 8), and no real indication for CSF (n ¼ 1). z Confirmed microbiological diagnosis (n ¼ 2; both positive for enterovirus PCR). § Confirmed microbiological diagnosis (n ¼ 3; one case each positive for enterovirus PCR, herpes simplex virus PCR and S. pneumoniae culture). CNS, central nervous system; CSF, cerebrospinal fluid; PCR, polymerase chain reaction; WCC, white cell count.

delay in processing the CSF until working hours is unlikely to be widely acceptable, especially as some concern exists about degradation of CSF leukocyte counts with delayed processing.8 Given the heterogeneity of indications for LP and clinical management decisions, we were unable to construct a call-back algorithm for triaging CSF requests. However, scope for improvement is possible with education of need for LP and appropriate testing requests likely to result in a reduced number of call-backs. There are several limitations of our study which include the retrospective nature of data collection, being a single centre study, and having a small sample size, with an even smaller proportion (15%) of patients having a CNS infection. Whether these results hold true for other infections such as tuberculous, fungal or bacterial meningitis is unknown. There is also a valid concern that CSF cell parameters degrade over time.8,9 However, methods including the addition of serum containing medium can be used to preserve the cells from degradation,10–12 thus allowing a delay in CSF processing and eliminating the need for callbacks. It is likely in the future that the nature of testing will evolve such that definitive testing is performed in a real-time fashion and management decisions are based on microbiological results as opposed to an interpreted surrogate, namely CSF cell count. This will, as with CSF cell count, require optimisation of ordering and diagnostic work flows. Despite common belief, this study highlights that in a large number of patients (49%), urgent CSF results did not influence initial clinical decision making. However, in this study, we were unable to determine the factors that differentiated when CSF cell count altered patient management. As a result, we continue to offer this service after-hours at RPAH. Further multicentre prospective studies (including paediatric-only centres) are necessary in order to determine the true utility of initial CSF microbiological parameters in this molecular era. Conflicts of interest and sources of funding: This work in part was presented at Pathology Update 2014, poster number 179. The authors state that there are no conflicts of interest to disclose. Shobini Sivagnanam Raymond C. Chan Catherine Janto Sebastiaan J. Van Hal Department of Microbiology, Royal Prince Alfred Hospital, Camperdown, Sydney, NSW, Australia

Contact Dr S. Sivagnanam. E-mail: [email protected] 1. McKinney WP, Heudebert GR, Harper SA, et al. Validation of a clinical prediction rule for the differential diagnosis of acute meningitis. J Gen Intern Med 1994; 9: 8–12. 2. Hoen B, Viel JF, Paquot C, et al. Multivariate approach to differential diagnosis of acute meningitis. Eur J Clin Microbiol Infect Dis 1995; 14: 267–74. 3. Spanos A, Harrell FE Jr, Durack DT. Differential diagnosis of acute meningitis. An analysis of the predictive value of initial observations. JAMA 1989; 262: 2700–7. 4. Brouwer MC, Thwaites GE, Tunkel AR, van de Beek D. Dilemmas in the diagnosis of acute community-acquired bacterial meningitis. Lancet 2012; 380: 1684–92. 5. Sandhaus LM, Ciarlini P, Kidric D, et al. Automated cerebrospinal fluid cell counts using the Sysmex XE-5000: is it time for new reference ranges? Am J Clin Pathol 2010; 134: 734–8. 6. Dupont SA, Wijdicks EF, Manno EM, Rabinstein AA. Thunderclap headache and normal computed tomographic results: value of cerebrospinal fluid analysis. Mayo Clinic Proc 2008; 83: 1326–31. 7. Heasley DC, Mohamed MA, Yousem DM. Clearing of red blood cells in lumbar puncture does not rule out ruptured aneurysm in patients with suspected subarachnoid hemorrhage but negative head CT findings. Am J Neuroradiol 2005; 26: 820–4. 8. Rajesh NT, Dutta S, Prasad R, Narang A. Effect of delay in analysis on neonatal cerebrospinal fluid parameters. Arch Dis Child 2010; 95: F25–9. 9. Steele RW, Marmer DJ, O’Brien MD, et al. Leukocyte survival in cerebrospinal fluid. J Clin Microbiol 1986; 23: 965–6. 10. de Graaf MT, van den Broek PD, Kraan J, et al. Addition of serumcontaining medium to cerebrospinal fluid prevents cellular loss over time. J Neurol 2011; 258: 1507–12. 11. Fry MM, Vernau W, Kass PH, Vernau KM. Effects of time, initial composition, and stabilizing agents on the results of canine cerebrospinal fluid analysis. Vet Clin Pathol 2006; 35: 72–7. 12. Veerman AJ, Huismans L, van Zantwijk I. Storage of cerebrospinal fluid samples at room temperature. Acta Cytol 1985; 29: 188–9.

DOI: 10.1097/PAT.0000000000000174

Multidrug resistant, blaVEB positive Pseudomonas aeruginosa causing high mortality among haematology patients Sir, We recently described an outbreak of multidrug resistant Pseudomonas aeruginosa (MDR-PAE) involving three patients in a haematology unit in 2009 that was associated with contaminated sink drains.1 In retrospect, P. aeruginosa with the same antimicrobial susceptibility profile (resistant to


CORRESPONDENCE

Table 1

Sex

651

Characteristics of patients with bacteraemia with blaVEB positive P. aeruginosa

Age, years

Underlying clinical condition

Chemotherapy

M

48

Acute lymphoblastic leukaemia

HyperCVAD Imatinib

M

51

Acute myeloid leukaemia

HiDAC

M

54


7þ3

M

36

Acute lymphoblastic leukaemia

HyperCVAD

M

26

Acute biphenotypic leukaemia

Fludarabine and busulfan

M

52


MIDAM

F

53


HiDAC

F

17


HiDAC

Antibiotics received Imipenem Amikacin Vancomycin Meropenem Amikacin Linezolid Cefepime Amikacin Vancomycin Metronidazole Meropenem Amikacin Vancomycin Meropenem Vancomycin Polymyxin B Caspofungin Meropenem Vancomycin Polymyxin B Meropenem Vancomycin Polymyxin B Meropenem Amikacin Polymyxin B

Length of admission before bacteraemia, days

Outcome

25

Died 5 days later

18

Died 3 days later

21

Died 8 days later

18

Died 2 days later

10

Died 1 day later

15

Died 1 day later

21

Died 2 days later

19

Discharged 41 days later

7 þ 3, cytarabine, daunorubicin; HiDAC, high-dose cytarabine; HyperCVAD, cyclophosphamide, vincristine, doxorubicin, dexamethasone; MIDAM, cytarabine, mitoxantrone, gemtuzumab ozogamicin, G-CSF.

ceftazidime, cefepime, meropenem, imipenem, amikacin, piperacillin-tazobactam, aztreonem, ciprofloxacin, and susceptible only to polymixin B) had been isolated from haematology patients over a period of 2 years going back to 2007. The first cluster of MDR-PAE was isolated from eight patients between July 2007 and April 2008. However, as the patients were scattered among different locations with no obvious epidemiological link between them, the isolates were archived but not typed. The second cluster of MDR-PAE was isolated from five patients between December 2008 and March 2009. This time as three patients were in different rooms on the same ward, an outbreak investigation was carried out which led to contaminated sinks being implicated.1 Retrospective molecular typing of archived MDR-PAE with the same antimicrobial susceptibility pattern was carried out by pulsed-field gel electrophoresis after restriction digestion of genomic DNA with SpeI.2 All 13 MDR-PAE isolated from patients belonged to the same clone as that isolated from two contaminated sinks. Multilocus sequence typing showed that the outbreak isolates belonged to ST357, a sequence type that has been associated with outbreaks of blaIMP-1 carrying P. aeruginosa in Japan.3,4 A representative strain DB82445/08 was positive for primers specific for blaVEB5 (Genbank Accession number HM246150) but negative for blaTEM, blaSHV, blaGES, blaPER, blaOXA-2, blaOXA-10, blaOXA-18, blaOXA-48, blaKPC, blaGIM, blaVIM, blaIMP, blaSIM, and blaSPM. The expression of intrinsic resistance determinants was determined by quantitative real-time reverse transcription-polymerase chain reaction. Transcription levels of the chromosomally encoded ampC cephalosporinase and oprD were similar to the pansensitive ATCC strain 27853.6 The expression of mexXY efflux pump systems showed a 5-fold increase compared to the ATCC strain, whereas that of mexAB-oprM, mexCD-oprJ, mexEFoprN, and mexJK operons remained similar.7

This is the first description of the blaVEB gene in Singapore. blaVEB is an Ambler class A extended-spectrum b-lactamase gene that may be found as gene cassettes in integrons, and on plasmids thereby facilitating its transfer between different species of bacteria.8 The first description of the VEB-1 enzyme in P. aeruginosa was that of a strain isolated from a French patient who had been hospitalised in southeast Asia, and it can be found in other Gram negative bacilli in the region.9,10 We are not sure why blaVEB has suddenly emerged in Singapore but the haemotology unit does receive foreign patients from surrounding countries and it may possibly have been imported. Though not a carbapenemase, VEB in combination with other mechanisms may give rise to a very drug resistant phenotype.11 This outbreak strain was associated with a high mortality. Of eight patients with bacteraemia, all of whom were neutropenic, seven died despite treatment with a combination of polymixin B and other antimicrobials (Table 1). We were unable to determine the causes of the first outbreak as the limited environmental screening performed at the time was negative, although notably did not include sampling the sink drains. This experience argues for a low threshold to initiate molecular epidemiology studies of MDR organisms to identify clonal outbreaks even when the epidemiological links between patients may not initially be apparent. This is especially so when the infections are associated with high patient morbidity and mortality. Acknowledgements: We acknowledge Miss Khoo Cheng Teng for technical assistance. We also acknowledge Molly Howe, RN, for her efforts. Conflicts of interest and sources of funding: This work was supported by a Singhealth Foundation Grant, SHF/FG385S/ 2008 and a Singapore General Hospital DCR grant. The authors state that there are no conflicts of interest to disclose.


652

CORRESPONDENCE

Thuan Tong Tan1 Moi Lin Ling2 Ban Hock Tan1 Tse Hsien Koh3 Department of Infectious Diseases, 2Director of Infection Control, and 3Department of Pathology, Singapore General Hospital, Singapore 1

Contact Dr T. H. Koh. E-mail: [email protected] 1. Ling ML, How KB. Pseudomonas aeruginosa outbreak linked to sink drainage design. Healthcare Infect 2013; 18: 143–6. 2. Talon D, Capellier G, Boillot A, et al. Use of pulsed-field gel electrophoresis as an epidemiologic tool during an outbreak of Pseudomonas aeruginosa lung infections in an intensive care unit. Intensive Care Med 1995; 21: 996–1002. 3. Curran B, Jonas D, Grundmann H, et al. Development of a multilocus sequence typing scheme for the opportunistic pathogen Pseudomonas aeruginosa. J Clin Microbiology 2004; 42: 5644–9. 4. Kouda S, Ohara M, Onodera M, et al. Increased prevalence and clonal dissemination of multidrug-resistant Pseudomonas aeruginosa with the blaIMP-1 gene cassette in Hiroshima. J Antimicrob Chemother 2009; 64: 46–51. 5. Dallenne C, Da Costa A, Decre´ D, et al. Development of a set of multiplex PCR assays for the deteection of genes encoding important b-lactamases in Enterobacteriaceae. J Antimicrob Chemother 2009; 65: 490–5. 6. Savli H, Karadenizli A, Kolayli F, Gundes S, Ozbek U, Vahaboglu H. Expression stability of six housekeeping genes: A proposal for resistance gene quantification studies of Pseudomonas aeruginosa by real-time quantitative RT-PCR. J Med Microbiol 2003; 52: 403–8. 7. Hocquet D, Nordmann P, El Garch F, et al. Involvement of the MexXYOprM efflux system in emergence of cefepime resistance in clinical strains of Pseudomonas aeruginosa. Antimicrob Agents Chemother 2006; 50: 1347–51. 8. Poirel L, Naas T, Guibert M, et al. Molecular and biochemical characterization of VEB-1, a novel class A extended-spectrum beta-lactamase encoded by an Escherichia coli integron gene. Antimicrob Agents Chemother 1999; 43: 573–81. 9. Naas T, Poirel L, Karim A, et al. Molecular characterization of In50, a class 1 integron encoding the gene for the extended-spectrum beta-lactamase VEB-1 in Pseudomonas aeruginosa. FEMS Microbiol Lett 1999; 176: 411–9. 10. Naas T, Benaoudia F, Massuard S, et al. Integron-located VEB-1 extendedspectrum beta-lactamase gene in a Proteus mirabilis clinical isolate from Vietnam. J Antimicrob Chemother 2000; 46: 703–11. 11. Woodford N, Zhang J, Kaufmann ME, et al. Detection of Pseudomonas aeruginosa isolates producing VEB-type extended-spectrum beta-lactamases in the United Kingdom. J Antimicrob Chemother 2008; 62: 1265–8.

DOI: 10.1097/PAT.0000000000000177


Dracunculiasis itself is rarely associated with increased mortality; however, emergence of the mature female worm may cause complications such as secondary bacterial infection which can lead to temporary or permanent disability.2 Although the global Guinea Worm Eradication Program called for dracunculiasis elimination in 1986, the disease remains endemic in South Sudan, Ethiopia, Ghana and Chad.3 It is uncommon to find cases of dracunculiasis outside of these four countries, with one other case reported in Australia in a Sudanese immigrant.4 Here, we report a recent case of D. medinensis seen in Melbourne, Australia, which was presumed to be acquired in South Sudan. A 38-year-old male South Sudanese migrant presented with a 1 year history of pain in the sole of his left foot with no other reported symptoms. The patient migrated to Australia 4 years prior to being admitted and had not left Australia since initial arrival. Patient consent was obtained during admission. An ultrasound of his foot was performed to investigate a suspected embedded foreign body (FB), with a linear measurement of 6.4 mm (length), positioned 4.7 mm laterally and 7.3 mm medially deep to the skin. An X-ray of his left foot and ankle revealed two areas of soft tissue calcification: one in the dorsal aspect of his lower calf, medial to the Achilles tendon and the second in the plantar aspect of his foot (Fig. 1), which were both described as tubular in configuration and curling upon themselves. Magnetic resonance imaging (MRI) was arranged for further evaluation and confirmed the calcific areas seen on plain X-ray. The two FBs (FB1 and FB2) were surgically removed and appeared to be skeletonised specimen encased in dense inflammatory tissue. Histopathology findings described subcutaneous abscesses with some multinucleated giant cells and eosinophils, with pieces of tegument (Fig. 2) consistent with helminth. FB1 and FB2 contained partly calcified degenerative worm. A polymerase chain reaction (PCR) targeting the nematode rDNA region spanning the second internal transcribed spacer (ITS-2) was attempted from the FB tissue samples, but no DNA could be extracted most likely due to the nematode being degenerate. Due to its characteristic appearance on X-ray, histological results, and the patient’s background, a diagnosis of D. medinensis was made. Overall, post-operative outcome was satisfactory with the patient representing to hospital on two separate occasions with haemoserous discharge from the surgical wound which was managed conservatively with wound dressings. Since the global campaign to eradicate dracunculiasis began in the mid-1980s,5 the number of reported cases has reduced

An experience with dracunculiasis in Melbourne, Australia Sir, Dracunculiasis (Guinea worm disease) is caused by Dracunculus medinensis, a parasitic nematode. These parasites are transmitted to humans via ingestion of pond water containing copepods (a subclass of crustaceans) infested with D. medinensis larvae. Although it is believed that no animal reservoir exists for this worm, recent observations in changes to the epidemiological profile of the disease suggest that a paratenic host, namely fish, may exist, and that domesticated animals such as dogs may be involved in passing the infection onto humans.1

Fig. 1 X-ray of the left foot showing the two foreign bodies (arrows).


CORRESPONDENCE

653

and histopathology results provide an effective means of identifying calcified forms of the disease. Acknowledgements: The authors would like to acknowledge Professor Wayne Morrison (Plastic and Reconstructive Surgery, St Vincent’s Hospital Melbourne), A/Professor John Slavin (Pathology, St Vincent’s Hospital Melbourne) and Dr Jaco Verweij (St Elisabeth Hospital, Tilburg, Netherlands) for their contributions.

Fig. 2 Histology of the sectioned foreign body showing a calcified worm with sections of broken tegument (inset).

from 900,000 in 20 countries in 1989 to 148 in the remaining four endemic countries in 2013.3 Provision of safe drinking water to affected communities either via filtration, use of insecticides, or accessing water from bore-hole or hand-dug wells has provided the most effective means of preventing the disease.6 Although dracunculiasis meets scientific criteria for eradication and the program to eradicate it has seen a vast decrease in the number of reported cases, the true extent and burden of dracunculiasis among at-risk populations is largely unknown because cases of the disease are most likely under-reported.7 Interestingly, rediscovery of D. medinensis in Chad after a 10 year absence has posed the question of whether paratenic hosts may play a role in the life cycle of the worm and thus be involved in altering the epidemiological pattern of the disease.8 Evidence such as this points to the question of whether a disease that dates back 3600 years can be fully eradicated in a few short decades.1,9 In this case report, diagnosis was made primarily through appearance on X-ray and patient background. Given the calcified degenerative nature of the surgically removed worm, it is likely that the patient had acquired it many years prior to migrating to Australia. Previous reports state that radiological diagnosis of dracunculiasis is straightforward given its characteristic serpentine appearance with linear calcification,4 and the fact that mature D. medinensis grow to such a large size (up to 100 centimetres for females).10 Other filarial nematodes such as Onchocerca volvulus and Loa loa also calcify in similar fashion and occur mainly in the hands and feet of affected individuals, however these are usually much smaller in size.11 It has been reported that most patients with calcified D. medinensis remain asymptomatic;12 however, our patient presented with significant pain and discomfort that was evident for approximately 12 months prior to admission. Although the burden of dracunculiasis remains small worldwide, the disease has been difficult to fully eradicate. This has mainly been attributed to the long incubation period (up to 1 year) and the lack of curative treatment.7 It may also be due to recently described zoonotic components of the disease.8 South Sudan appears to account for the majority of acute cases, but it is important to note that cases are still being reported, albeit sporadically, in countries certified as being dracunculiasis free. Increased refugee and migration movement from endemic as well as recently certified free countries may lead to cases of imported dracunculiasis being reported, and hence clinical vigilance in non-endemic countries remains important. As described in this case report, use of radiological techniques

Conflicts of interest and sources of funding: This case was presented at the 24th European Congress of Clinical Microbiology and Infectious Diseases (ECCMID) 2014. The authors state that there are no conflicts of interest to disclose. Menino O. Cotta1 Maria V. Pereira E Cotta2 Jonathan Darby2 Tom R. Sutherland2 Harsha Sheorey2 Melbourne Health, and 2St Vincent’s Hospital, Melbourne, Vic, Australia 1

Contact M. O. Cotta. E-mail: [email protected] 1. Cairncross S. Mother Nature’s surprises. Am J Trop Med Hyg 2014; 90: 3–4. 2. Spring M, Spearman P. Dracunculiasis: report of an imported case in the United States. Clin Infect Dis 1997; 25: 749–50. 3. World Health Organization. Monthly report on dracunculiasis cases, January 2014. Weekly Epidemiological Record 2014; 11: 115–6. 4. Menon T. Incidental finding of Dracunculus medinensis in Australia. Med J Aust 2005; 183: 51–2. 5. Hopkins DR, Ruiz-Tiben E. Strategies for dracunculiasis eradication. Bull World Health Organ 1991; 69: 533–40. 6. Ruiz-Tiben E, Hopkins DR. Dracunculiasis (Guinea worm disease) eradication. Adv Parasitol 2006; 61: 275–309. 7. Hopkins DR. Disease eradication. N Engl J Med 2013; 368: 54–63. 8. Eberhard ML, Ruiz-Tiben E, Hopkins DR, et al. The peculiar epidemiology of dracunculiasis in Chad. Am J Trop Med Hyg 2014; 90: 61–70. 9. Cairncross S, Tayeh A, Korkor AS. Why is dracunculiasis eradication taking so long? Trends Parasitol 2012; 28: 225–30. 10. Cairncross S, Muller R, Zagaria N. Dracunculiasis (Guinea worm disease) and the eradication initiative. Clin Microbiol Rev 2002; 15: 223–46. 11. Gulanikar A. Dracunculiasis: two cases with rare presentations. J Cutan Aesthet Surg 2012; 5: 281–3. 12. Muller R. Dracunculus and dracunculiasis. Adv Parasitol 1971; 9: 73–151.

DOI: 10.1097/PAT.0000000000000169

Renal cell carcinoma with rhabdoid features and loss of INI1 expression in an individual without sickle cell trait Sir, Renal medullary carcinoma (RMC), first described by Davis et al. in 1995,1 occurs most frequently among the young, with average age in the third decade and marked male predominance. Evidence of sickle cell trait, disease, or related haemoglobinopathy, is seen in all cases.


654


CORRESPONDENCE

Morphologically, RMC is characterised by the high grade, poorly differentiated appearance of glands, which often show a reticular and cribriform appearance, infiltrative growth with desmoplasia and stromal inflammation. A subset of cases show a rhaboid morphology. Metastases are both lymphatic and vascular with lymph nodes, liver and lungs most often involved. The RMC cases show loss of nuclear expression of the transcriptional regulator SMARCB1 (i.e., INI1).2–4 The prognosis is poor, with only a minor subset surviving more than a year. Chemotherapy has been known to prolong survival by a few months, whereas radiotherapy does not seem to alter the course of the disease. Discussion of whether RMC and collecting duct carcinoma (CDC) are a spectrum, with the RMC variant at the more aggressive end, has been extensive.1,5,6 A 33-year-old male of Caucasian ancestry presented with gross haematuria. Weight loss and palpable mass were present. This prompted a computed tomography (CT) scan of the chest and abdomen. There were pulmonary nodules, with hilar lymph nodes significantly enlarged. There was also a 5.8 5.2 5.0 cm irregular left lower pole renal mass, with adjacent para-aortic lymphadenopathy. Radical nephrectomy was performed. Due to the presence of pulmonary nodules and lymphadenopathy, considered to be secondary deposits, the patient received adjuvant chemotherapy. Six months after surgery he was alive, however with worsening conditions and no response to chemotherapy. Preoperative laboratory tests showed mild anaemia, with a haemoglobin level of less than 12 g/dL. Since there was no evidence of sickle cell trait or other haemoglobinopathy by haemoglobin electrophoresis, it was considered cancer-related anaemia. Macroscopically, this was a poorly circumscribed tumour, whitish in colour, 6 cm in greatest diameter, occupying chiefly the renal medulla (Fig. 1A). The histology was that of a high grade carcinoma with rhabdoid features (Fig. 1B,C), based in the collecting system but widely infiltrative and with invasion of the sinus fat (pT3a), with variable, tubular, nested, and

cribriform architecture and sclerosing to myxoid stromal reaction. The cells were eosinophilic with clear nuclei and with prominent nucleoli. Neutrophils were admixed with the tumour. Drepanocytes were not identified histologically. The lesion showed immunohistochemical expression of pancytokeratin Oscar, as well as EMA, arguing against a translocation carcinoma or epithelioid mesenchymal lesion. The neoplasm was carbonic anhydrase 9 entirely negative, thus excluding a high grade clear cell renal cell carcinoma with rhabdoid features. Expression of INI1 was uniformly lost (Fig. 1D), while OCT3/4 and ALK-1 were negative. The neoplasm was patchy positive for PAX8, supporting a renal histogenesis for this lesion and arguing against a neoplasm in the rhabdoid tumour or proximal type epithelioid sarcoma class of lesions, which would have to be considered given the loss of INI1. With few exceptions RMCs are seen in young black patients with sickle cell trait between ages 10 and 40 (mean age 22 years) and chiefly in males by 2:1. The common symptoms are gross haematuria and flank or abdominal pain. Weight loss and palpable mass are common. Metastatic deposits such as cervical nodes or brain tumour may be the initial evidence of disease. In the clinical setting of a young black patient with sickle cell trait it is often possible to anticipate the correct diagnosis with imaging studies.1,7 Centrally located tumours with an infiltrative growth pattern, invading renal sinus, are typical for RMC. These are poorly circumscribed tumours occupying chiefly the renal medulla, usually with satellite nodules in the cortex and in the perinephric and peripelvic soft tissue. Size ranges from 4 to 12 cm with a mean of 7 cm. Most show haemorrhage and necrosis. Histopathologically, a reticular growth pattern and a more compact adenoid cystic or cribriform morphology are the common features. Most cases also have poorly differentiated areas consisting of solid sheets of cells with a rhabdoid or squamoid quality. The cells are eosinophilic with clear nuclei, usually with prominent nucleoli. Neutrophils are often admixed

A

B

C

D

Fig. 1 (A) Gross appearance of the radical nephrectomy specimen with a tumour, whitish in colour and 6 cm in greatest diameter, occupying chiefly the renal medulla. (B) High grade carcinoma. (C) Rhabdoid features. (D) Loss of INI1 expression (the inflammatory cells in the stroma are positive, i.e., positive internal control).


CORRESPONDENCE

with the tumour and the advancing margins are often bounded by lymphocytes. Oedematous or collagenous stroma forms a considerable component of many tumours. A majority of cases show droplets of cytoplasmic mucin and sickled erythrocytes. Pancytokeratin immunohistochemistry is nearly always positive as is EMA but typically less strongly so. CEA is usually positive. One study found strong expression of low molecular weight cytokeratin (CAM 5.2) but negative high molecular weight cytokeratin.8 Basically, RMC shows loss of nuclear expression of the transcriptional regulator SMARCB1 (i.e., INI1), encoded on chromosome 22.2–5 Molecular studies have correlated this finding to loss of heterozygosity5 or hemizygous deletions2 at the SMARCB1 locus, although loss of chromosome 22 has also been observed.8 Recent data have suggested that acquisition of expression of the stem cell marker POU5F1 (i.e., OCT3/4) may also be diagnostically helpful.9 The Heidelberg Classification of renal neoplasia of 1996 noted RMC as a variant of CDC.6 In contrast, the argument to separate RMC from CDC and define it strictly by clinical setting dates back to the original Davis et al.,1 who described the entity literally as a ‘sickle cell nephropathy’ with a markedly poor prognosis. This observation continues through larger series and interinstitutional clinicopathological comparisons of RMC and CDC.3,5,8 Against definition of RMC by INI1 status, at least one recent experience identifies complete loss of expression of INI1 in 15% and focal/weak expression of INI1 in up to 30% of CDC, with no difference in survival by INI1 status.5 For that matter, infrequent cases of sickle cell trait-associated RMC have shown retained INI1. While further studies are necessary to understand the relationship between these cancers, CDC should be defined as proposed in the ISUP Vancouver Classification and outlined above.10,11 For renal medullary carcinoma, it should be defined on the basis of both appropriate histomorphology and evidence of sickle cell trait or disease, as originally proposed by Davis et al.1 The term ‘unclassified renal cell carcinoma with medullary phenotype’ has been proposed by Amin et al.12 for cases of high grade carcinoma of the distal nephron showing morphology, immunophenotypic, or molecular features characteristic of medullary carcinoma but in a patient without evidence of haemoglobinopathy. With this diagnosis, it has been recommend that comments are included with respect to the clinical setting, i.e., lack of evidence of sickle cell disease, trait, or related haemoglobinopathy, and description of morphological and imununohistochemical features that are present. In conclusion, a renal cell carcinoma is described, showing the rhabdoid features and loss of nuclear expression of the transcriptional regulator INI1 usually seen in RMC, in a patient confirmed not to harbour sickle cell trait or other haemoglobinopathy. Pathologists should be aware of the existence of such tumours, meanwhile awaiting further clarification on the nature of this entity. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Marina Scarpelli1* Roberta Mazzucchelli1* Antonio Lopez-Beltran2 Liang Cheng3 Michele De Nictolis4

655

Matteo Santoni5 Rodolfo Montironi1 1

Section of Pathological Anatomy, Polytechnic University of the Marche Region, School of Medicine, United Hospitals, Ancona, Italy; 2Department of Surgery, Cordoba University Medical School, Cordoba, Spain; 3Department of Pathology and Laboratory Medicine, Indiana University School of Medicine, Indianapolis, IN, USA; 4Section of Pathological Anatomy, Azienda Ospedaliera Marche Nord, Pesaro; and 5 Medical Oncology, Polytechnic University of the Marche Region, School of Medicine, United Hospitals, Ancona, Italy; *these authors contributed equally to this work Contact Professor R. Montironi. E-mail: [email protected] 1. Davis CJ Jr, Mostofi FK, Sesterhenn IA. Renal medullary carcinoma. The seventh sickle cell nephropathy. Am J Surg Pathol 1995; 19: 1–11. 2. Calderaro J, Moroch J, Pierron G, et al. SMARCB1/INI1 inactivation in renal medullary carcinoma. Histopathology 2012; 61: 428–35. 3. Cheng JX, Tretiakova M, Gong C, et al. Renal medullary carcinoma: rhabdoid features and the absence of INI1 expression as markers of aggressive behavior. Mod Pathol 2008; 21: 647–52. 4. Hollmann TJ, Hornick JL. INI1-deficient tumors: diagnostic features and molecular genetics. Am J Surg Pathol 2011; 35: e47–63. 5. Elwood H, Chaux A, Schultz L, et al. Immunohistochemical analysis of SMARCB1/INI-1 expression in collecting duct carcinoma. Urology 2011; 78: e471–5. 6. Kovacs G, Akhtar M, Beckwith BJ, et al. The Heidelberg classification of renal cell tumours. J Pathol 1997; 183: 131–3. 7. Schaeffer EM, Guzzo TJ, Furge KA, et al. Renal medullary carcinoma: molecular, pathological and clinical evidence for treatment with topoisomerase-inhibiting therapy. BJU Int 2010; 106: 62–5. 8. Rao P, Tannir NM, Tamboli P. Expression of OCT3/4 in renal medullary carcinoma represents a potential diagnostic pitfall. Am J Surg Pathol 2012; 36: 583–8. 9. Carvalho JC1, Thomas DG, McHugh JB, et al. p63, CK7, PAX8 and INI-1: an optimal immunohistochemical panel to distinguish poorly differentiated urothelial cell carcinoma from high-grade tumours of the renal collecting system. Histopathology 2012; 60: 597–608. 10. Delahunt B, Egevad L, Montironi R, et al. International Society of Urological Pathology (ISUP) consensus conference on renal neoplasia: rationale and organization. Am J Surg Pathol 2013; 37: 1463– 8. 11. Srigley JR, Delahunt B, Eble JN, et al. ISUP Renal Tumor PanelThe International Society of Urological Pathology (ISUP) Vancouver Classification of Renal Neoplasia. Am J Surg Pathol 2013; 37: 1469–89. 12. Amin BM, Smith SC, Agaimy A, et al. Collecting duct carcinoma versus renal medullary carcinoma. An appeal for nosologic and biological clarity. Am J Surg Pathol 2014; 38: 871–4.

DOI: 10.1097/PAT.0000000000000175

Adenocarcinoma arising within ectopic prostatic tissue in the urinary bladder Sir, Ectopic prostate tissue has been reported in many sites throughout the body, most commonly in the urethra, seminal vesicles, epididymis, testis and urinary bladder.1–6 Outside of the genitourinary system, ectopic prostatic tissue is a distinctly uncommon phenomenon, although it has occasionally been identified in males, and surprisingly also in females, in a wide range of sites including the spleen, rectum and uterine cervix, as well as


656


CORRESPONDENCE

in ovarian teratomas.7–10 Adenocarcinoma arising in ectopic prostate tissue is a rare finding which is important to differentiate from metastatic prostate carcinoma. We describe a case of acinar adenocarcinoma arising in ectopic prostate tissue in the dome of the bladder. A 70-year-old man presented to the urologist with a 3 year history of obstructive urinary symptom with elevated prostate specific antigen (PSA) of 4.72 mg/L which increased to 5.67 mg/L after 3 months. The patient described a restricted urinary stream with some hesitancy and he also experienced nocturia averaging once per night. Examination of the prostate revealed a moderate to large, smooth and firm, but benign feeling gland. No haemoglobin or leukocyte esterase was detected with a urine test strip, and a urinary tract ultrasound reported a slightly enlarged prostate and a normal urinary bladder. Cystoscopy revealed a cystic looking lesion in the dome of the bladder which was biopsied (Fig. 1A). The bladder had a normal capacity with mild to moderate trabeculation. Routine ultrasound guided prostate biopsies were performed at the time of the cystoscopy. The bladder dome lesion contained several small cysts lined by GATA binding protein 3 (GATA3) positive urothelium and

columnar epithelium surrounded by fibrous tissue and smooth muscle consistent with urachal remnants (Fig. 1B). In one area of this lesion there were a few ectopic prostate glands along with numerous small atypical glands, some containing pink amorphous secretion, lined by cuboidal to columnar epithelium with nuclear enlargement, prominent nucleoli and straight luminal borders (Fig. 1C,D). These atypical glands did not have a basal layer and fused glands were not seen. The glandular morphology was consistent with acinar adenocarcinoma of prostatic type. The prostatic nature of the benign and malignant glands seen within the bladder lesion was confirmed by positive immunostaining of secretory cells for PSA, prostatic specific membrane antigen (PSMA) and prostatic acid phosphatase (PAP). The basal cells in the non-neoplastic prostatic glands stained for p63, cytokeratin 34bE12 and GATA3 (Fig. 1E,F). The urachal remnants did not stain for PSA, PSMA or PAP. A follow-up re-biopsy of this bladder dome site 4 months later showed residual urachal cyst surrounded by scar tissue. Residual adenocarcinoma was not seen. Five of eight prostatic needle core biopsy specimens taken at the time of the bladder biopsy contained acinar adenocarcinoma with a Gleason score of 7 (3þ4), with the grade 4 tumour

A

B

C

D

E

F

Fig. 1 (A) Cystoscopic image of the 5 mm diameter cystic lesion at the dome of the urinary bladder. The lesion had a reddish brown variegated appearance not typical of the usual bland urachal cyst. (B) Urachal remnants with some cystic dilatation and eosinophilic secretion. Adenocarcinoma is indicated by arrow (H&E). (C) Nonneoplastic ectopic prostate glands (arrow) with urachal remnants (bottom) and acinar adenocarcinoma (top and left) (H&E). (D) Acinar adenocarcinoma with amorphous pink secretion, straight luminal borders, enlarged nuclei and prominent nucleoli. (H&E). (E) Cytokeratin 34bE12 staining of urachal remnants (bottom) and nonneoplastic prostate glands (arrow). (F) PSMA IHC showing strong intensity staining of adenocarcinoma. Urachal remnants (bottom) are PSA negative.


CORRESPONDENCE

occupying approximately 5% of the tumour present. These biopsies did not show evidence of perineural, lymphovascular or extraprostatic tissue invasion and the patient was treated with radiotherapy. The first case of adenocarcinoma arising within ectopic prostatic tissue was reported in a 72-year-old man by Adams in 1993.8 He reported an encapsulated firm nodule of tissue adjacent to the rectum but separate from the prostate measuring 3.2 cm in diameter. The prostate contained adenocarcinoma with Gleason score 5 (3þ2), as did the nodule, which also contained benign prostatic tissue. Gardner et al. recently described a second case of adenocarcinoma arising in ectopic prostate tissue, in this instance in the urinary bladder dome of a 62-year-old male.6 A Gleason score of 6 (3þ3) was given to this focus found incidentally after a radical cystoprostatectomy for invasive high grade papillary urothelial carcinoma. Two incidental foci of prostatic adenocarcinoma were identified within and confined to the prostate with a Gleason score of 7 (3þ4). The possibility of metastatic disease must be considered when prostatic adenocarcinoma is seen outside the prostate, but computed tomography and bone scans of our patient did not show evidence of distant metastases. This combined with the finding of normal prostate glands adjacent to the adenocarcinoma is strongly suggestive of origin from the ectopic tissue. In the present case, the positive staining of both the benign glandular tissue and adenocarcinoma for PSA, PSMA and PAP provides confirmation of the prostatic nature of this tissue. Various hypotheses have been proposed to account for the presence of ectopic prostate tissue: in the prostatic urethra and bladder trigone the ectopic tissue may represent a vestigial remnant from embryogenesis, while for other sites metaplasia and aberrant embryogenesis, i.e., ‘migration’ or ‘misplacement’ of prostate glands, have been postulated.6,9 Nowels et al. suggested that the frequently positive staining (35% of their 40 cases) of cystitis cystica and glandularis for PSA and PSAP supports metaplasia as a possible mechanism for the development of ectopic prostate tissue.11 In another study of bladder biopsy specimens, Cohen et al. found prostatic cytoplasmic secretions in 19 of 34 randomly biopsied cases of cystitis cystica and glandularis. In six of their 34 cases, ‘large numbers’ of morphologically normal prostate glands were identified, suggesting this is a not uncommon occurrence.12 Metaplasia may also account for the presence of prostate glands amongst normal endocervical glands, although the close alignment of the uterus and urogenital sinus during embryogenesis raises the possibility that periurethral glands (which have some similarities both histologically and immunohistochemically to prostate glands) may be misplaced during this process.9 Alternatively, in some cases, aberrantly located prostate glands represent a component of a teratoma rather than true ectopia, for instance in ovarian and testicular teratomas.10 In conclusion, we report a rare occurrence of prostatic acinar adenocarcinoma arising from ectopic prostate tissue in the bladder dome. Awareness of the numerous sites in which ectopic prostatic tissue may occur, along with immunohistochemical analysis, is important in ensuring this phenomenon is not misdiagnosed and appropriate clinical management instituted. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose.

657

Anastasia Backhouse1 Arthur Vasilaras2 Teresa Yung3 James G. Kench4,5 1

PathWest Laboratory Medicine, QEII Medical Centre, Perth, WA, 2Department of Urology, Royal Prince Alfred Hospital, Camperdown, NSW, 3Healthscope Pathology, Bella Vista, NSW, 4Department of Tissue Pathology and Diagnostic Oncology, Royal Prince Alfred Hospital, Camperdown, NSW, and 5Sydney Medical School, University of Sydney, NSW, Australia Contact Professor J. G. Kench. E-mail: [email protected] 1. Leifert S, Lurie A, Kellner J. Ectopic prostatic tissue in urethra. Urology 1985; 26: 509–10. 2. Lau SK, Chu PG. Prostatic tissue ectopia within the seminal vesicle: a potential source of confusion with seminal vesicle involvement by prostatic adenocarcinoma. Virchows Arch 2006; 449: 600–2. 3. Bromberg WD, Kozlowski JM, Oyasu R. Prostate-type gland in the epididymis. J Urol 1991; 145: 1273–4. 4. Milburn JM, Bluth EI, Mitchell WT Jr. Ectopic prostate in the testicle: an unusual cause of a solid testicular mass on ultrasonography. J Ultrasound Med 1994; 13: 578–80. 5. Morey AF, Kreder KJ, Wikert GA, Cooper G, Dresner ML. Ectopic prostate tissue at the bladder dome. J Urol 1989; 141: 942–3. 6. Gardner JM, Khurana H, Leach FS, Ayala AG, Zhai J, Ro JY. Adenocarcinoma in ectopic prostatic tissue at dome of bladder: a case report of a patient with urothelial carcinoma of the bladder and adenocarcinoma of the prostate. Arch Pathol Lab Med 2010; 134: 1271–5. 7. Vogel U, Negri G, Bultmann B. Ectopic prostatic tissue in the spleen. Virchows Arch 1996; 427: 543–5. 8. Adams JR Jr. Adenocarcinoma in ectopic prostatic tissue. J Urol 1993; 150: 1253–4. 9. Nucci MR, Ferry JA, Young RH. Ectopic prostatic tissue in the uterine cervix: a report of four cases and review of ectopic prostatic tissue. Am J Surg Pathol 2000; 24: 1224–30. 10. McLachlin CM, Srigley JR. Prostatic tissue in mature cystic teratomas of the ovary. Am J Surg Pathol 1992; 16: 780–4. 11. Nowels K, Kent E, Rinsho K, Oyasu R. Prostate specific antigen and acid phosphatase-reactive cells in cystitis cystica and glandularis. Arch Pathol Lab Med 1988; 112: 734–7. 12. Cohen RJ, Garrett K, Golding JL, Thomas RB, McNeal JE. Epithelial differentiation of the lower urinary tract with recognition of the minor prostatic glands. Hum Pathol 2002; 33: 905–9.

DOI: 10.1097/PAT.0000000000000171

Thyroid-like follicular renal cell carcinoma: an emerging morphological variant Sir, The classification working group of the International Society of Urological Pathology (ISUP) was entrusted to make recommendations regarding additions, changes, and refinements to the current World Health Organization (WHO) classification system for renal tumours.1 Based on available literature, three rare primary renal tumours were considered as emerging or provisional new entities. Thyroid-like follicular renal cell carcinoma (TLF-RCC) was one such tumour in the provisional category with only a few cases reported in the English literature. Herein we describe a case of primary TLF-RCC.


658


CORRESPONDENCE

A 33-year-old male patient presented with right sided flank pain on and off for one year. However, the patient did not have any history of haematuria or oliguria. Ultrasound of the abdomen revealed a well defined, echogenic, space occupying lesion in the mid-pole of the right kidney along the lateral border. Subsequently, contrast enhancing magnetic resonance imaging was carried out which showed a well defined, partially exophytic lesion measuring 5.3 4.3 4.0 cm in the lateral cortex of the right kidney in the interpolar region. The tumour showed moderate intensity with focal cystic areas. No regional or distant metastatic lesions were noted. Tc99M DTPA scan showed relative function of 47% with a photopenic defect in the mid-pole of the right kidney. Based on these investigations, a nephron sparing partial nephrectomy was performed. The thyroid function tests carried out subsequently were within normal range. On gross examination the partial nephrectomy specimen measured 6 5 4 cm with a bulge on the lateral border. The cut surface showed a well defined mass measuring 5.0 4.0 3.8 cm with yellow tan cut surface and focal haemorrhagic areas, however no necrosis was seen (Fig. 1A). The tumour did not show capsular invasion or extension into perinephric fat. The surgical resection margin was free of tumour. Microscopically, a well circumscribed tumour was seen showing diffuse follicular architecture composed of microfollicles as well as macrofollicles filled with inspissated colloid-like material (Fig. 1B–D). Focal areas showed papillary architecture accounting for 5% of the tumour. The follicles were placed back to back, while lymphoid aggregates, haemorrhages and cholesterol crystals were seen in intervening areas focally. The tumour cells lining the follicles were cuboidal with moderate eosinophilic cytoplasm, round to oval nuclei, fine chromatin and inconspicuous nucleoli. The cells lining the cystically dilated macrofollicles were flattened with small nuclei. The tumour cells did not show any nuclear features of papillary thyroid carcinoma. There was no significant

nuclear pleomorphism, mitotic activity or necrosis and the Fuhrman nuclear grade was 2. No capsular invasion or extension into perinephric fat was seen microscopically. On immunohistochemistry (Fig. 2A–D), the tumour cells showed immunoreactivity for Pax2, cytokeratin 7 (CK7), vimentin and epithelial membrane antigen (EMA). The tumour cells were negative for thyroglobulin (TG), thyroid transcription factor 1 (TTF1), cytokeratin 20 (CK20), alpha-methylaminoacyl CoA racemase (AMACR), Pax8, high molecular weight cytokeratin (CK34bE12), Wilms tumour antigen 1 (WT1), CD10, CD15 and CD56. Based on the morphology and immunohistochemical profile, a diagnosis of thyroid-like follicular renal cell carcinoma was made. Since the first report of thyroid-like follicular renal cell carcinoma by Jung et al.,2 13 additional cases have been reported in the English literature.3–9 However, due to the limited number of cases and uncertainty about maximum allowable papillary component in this tumour, it was not recognised as an established entity in the 2013 Vancouver classification of renal neoplasms. The clinicopathological details of the reported cases are described in Table 1. Morphologically, previously reported TLF-RCCs showed predominantly micro- and macrofollicular architecture with focal papillary pattern in some,3,7,9 like the present case. Ohe et al.10 reported a tumour that showed predominantly papillary architecture with a secondary follicular component amounting to about 30%. A consensus about relative proportion of thyroid follicular-like component and papillary component is needed to definitely categorise a tumour as primary TLF-RCC or papillary carcinoma with thyroid-like differentiation. This distinction may have clinical implication in prognosis, although there are limited data on the prognosis of TLF-RCC. The follicles filled with eosinophilic secretions can be seen in conventional renal tumours as well as other less common tumours such as metanephric adenoma. The acinar structures in clear cell carcinoma can dilate producing a micro- as well as macrocystic

A

B

C

D

Fig. 1 (A) Tumour with yellow-tan cut surface with focal haemorrhages. (B) Normal renal parenchyma interface. The tumour is showing diffuse follicular architecture with interspersed cholesterol clefts. (C) Peripheral areas showing dense lymphoid aggregates. (D) Micro- and macrofollicles filled with colloid-like material.


CORRESPONDENCE

A

B

C

D

659

Fig. 2 Tumour cells showing strong immunoreactivity for (A) Pax2 and (B) cytokeratin 7, and negativity for (C) TTF1 and (D) thyroglobulin.

pattern. In tubular pattern of papillary carcinoma, small- to medium-sized tubules are lined by cuboidal or columnar cells and may contain secretions. Such features are only seen in focal areas, although a case of papillary renal cell carcinoma with large, cystically dilated, thyroid-like follicles comprising a 20% area has been reported.11 In such cases ancillary tests such as immunohistochemistry, cytogenetics and molecular studies might be helpful. Before making a definitive diagnosis of TLF-RCC a detailed work-up is needed to exclude the possibility of metastatic follicular carcinoma or follicular variant of papillary thyroid carcinoma. Metastasis of these primary thyroid tumours to kidney is rare and only a few cases have been reported in literature. However, all of these tumours retain the thyroid specific immunoprofile with reactivity to TTF1 and thyroglobulin at the metastatic site, and therefore can be easily distinguished from TLF-RCC. The present case showed negativity for TTF1 and thyroglobulin and positivity for CK7, vimentin, EMA and relatively renal specific nuclear transcription factor

Table 1

Pax2, thus ruling out possibility for metastatic thyroid malignancy. There are reports that raise the issues about nomenclature of this entity. Angell et al.12 reported a case with features of papillary carcinoma of thyroid and labelled it as ‘primary thyroid like carcinoma of the kidney’. Khoja et al.6 labelled a case with similar morphology as papillary thyroid carcinoma (follicular variant)-like tumour of kidney. These cases highlight the need for unifying definition and nomenclature, although terminology used may be the authors’ own views. In conclusion, we have described a thyroid-like follicular renal cell carcinoma, a rare but emerging tumour entity under the 2013 Vancouver classification of renal neoplasia. Metastasis from primary thyroid malignancy must be excluded before making a definitive diagnosis of TLF-RCC. Caution is necessary before labelling any tumour TLF-RCC, as conventional renal tumours can show variable follicular architecture. Additional cases as well as more data on prognosis are needed

Clinical profile of all cases of thyroid-like follicular renal cell carcinoma reported in the English literature to date

Case Jung et al.2 Sterlacci et al.3 Amin et al.4

Dhillon et al.5 Khoja et al.6 Alessandrini et al.7 Malde et al.8 Volavsek et al.9 Present case

Age/Sex 32/F 29/F 53/F 29/F 45/M 83/M 35/M 50/F 34/M 31/F 76/M 41/F 29/F 34/M 33/M

Presenting symptoms

Location

Incidental

Right lower pole Left mid pole Right mid pole Right upper pole Right lower pole Left lower pole Right mid pole Right mid pole Right mid pole Left upper pole Left upper pole Right lower pole Left lower pole Left lower pole Right mid pole

Incidental Incidental Incidental Incidental Incidental Incidental Haematuria, flank pain Haematuria, weight loss, flank pain Haematuria Incidental Abdominal pain Pain Right flank pain

Size, cm 11.8 5.0 2.1 1.9 3.5 2.1 3.0 4.0 6.2 4.0 4.5 4.3 6.5 5.5 5.0

Stage

Follow-up

pT2N0M0 pT1bN0M1 pT1aNx pT1aNx pT1aN1 pT1aNx pT1aNx pT1aN0 pT1bN2M1 pT1aN0 pT1bN0 pT1bNx pT1bN0 pT1bNx pT1bN0M0

6 months 5 years 54 months 84 months 17 months, lost to follow-up 48 months 20 months 7 months 3 months 21 months 11 months 4 months 4 months 6 months 3 months


660

CORRESPONDENCE

before TLF-RCC can be included as a distinctive entity in the WHO classification of renal tumours in the future. Abhijit Chougule1 Amanjit Bal1 Ashim Das1 Brusabhanu Nayak2 Departments of 1Histopathology, and 2Urology, Post Graduate Institute of Medical Education and Research (PGIMER), Sector – 12, Chandigarh 160012, India Contact Dr A. Bal. E-mail: [email protected] 1. Srigley JR, Delahunt B, Eble JN, et al. The International Society of Urological Pathology (ISUP) Vancouver Classification of Renal Neoplasia. Am J Surg Pathol 2013; 37: 1469–89. 2. Jung SJ, Chung JI, Park SH, Ayala AG, Ro JY. Thyroid follicular carcinoma-like tumor of kidney: a case report with morphologic, immunohistochemical, and genetic analysis. Am J Surg Pathol 2006; 30: 411–5. 3. Sterlacci W, Verdorfer I, Gabriel M, Mikuz G. Thyroid follicular carcinoma-like renal tumor: a case report with morphologic, immunophenotypic, cytogenetic, and scintigraphic studies. Virchows Arch 2008; 452: 91–5. 4. Amin MB, Gupta R, Ondrej H, et al. Primary thyroid-like follicular carcinoma of the kidney: report of 6 cases of a histologically distinctive adult renal epithelial neoplasm. Am J Surg Pathol 2009; 33: 393–400. 5. Dhillon J, Tannir NM, Matin SF, Tamboli P, Czerniak BA, Guo CC. Thyroid-like follicular carcinoma of the kidney with metastases to the lungs and retroperitoneal lymph nodes. Hum Pathol 2011; 42: 146–50. 6. Khoja HA, Almutawa A, Binmahfooz A, Aslam M, Ghazi AA, Almaiman S. Papillary thyroid carcinoma-like tumor of the kidney: a case report. Int J Surg Pathol 2012; 20: 411–5. 7. Alessandrini L, Fassan M, Gardiman MP, Guttilla A, Zattoni F, Galletti TP. Thyroid-like follicular carcinoma of the kidney: report of two cases with detailed immunohistochemical profile and literature review. Virchows Arch 2012; 461: 345–50. 8. Malde S, Sheikh I, Woodman I, Fish D, Bilagi P, Sheriff MK. Primary thyroid-like follicular renal cell carcinoma: an emerging entity. Case Rep Pathol 2013; 2013: 687427. 9. Volavsek M, Strojan-Flezar M, Mikuz G. Thyroid-like follicular carcinoma of the kidney in a patient with nephrolithiasis and polycystic kidney disease: a case report. Diagn Pathol 2013; 8: 108. 10. Ohe C, Kuroda N, Pan CC, et al. A unique renal cell carcinoma with features of papillary renal cell carcinoma and thyroid-like carcinoma: a morphological, immunohistochemical and genetic study. Histopathology 2010; 57: 494–7. 11. Fadare O, Lam S, Rubin C, Renshaw IL, Nerby CL. Papillary renal cell carcinoma with diffuse clear cells and thyroid-like macrofollicular areas. Ann Diagn Pathol 2010; 14: 284–91. 12. Angell SK, Pruthi R, Freiha FS. Primary thyroidlike carcinoma of the kidney. Urology 1996; 48: 632–5.

DOI: 10.1097/PAT.0000000000000176

Lipid storing pyloric metaplasia of the gall bladder Sir, Cholesterolosis and pyloric metaplasia are common findings in cholecystectomy specimen removed for cholelithiasis.1,2 Cholesterolosis is an accumulation of cholesterol esters and triglycerides in macrophages in the lamina propria of the gallbladder. Glands akin to pyloric mucosa forming small lobules are seen in mucosa of the biliary tree in pyloric metaplasia. The following is a case of widespread cholesterolosis and pyloric metaplasia in the gall bladder of a 19-year-old female. The


metaplastic epithelium showed extensive lipid storage. The mechanism of cholesterolosis and accumulation of lipids within the metaplastic epithelium is explored. The patient was 19-year-old female who presented with longstanding abdominal pain, bloating, diarrhoea and occasional bouts of biliary colic. Upper and lower gastrointestinal endoscopic examination failed to reveal any abnormality. Ultrasound examination showed small polyps in the gall bladder mucosa. No calculi were present. Her serum cholesterol was 6.8 mmol/L [reference range (RR) 3.9–5.5] and low density lipoprotein (LDL) cholesterol was 4.7 mmol/L (RR 1.7–3.5). Iron studies showed low serum iron at 3.9 mmol/L (RR 5–30) and low saturation at 5% (RR 10–45). Other investigations including full blood count, liver and thyroid function test were unremarkable. Cholecystectomy was performed. The specimen was a partially opened gallbladder which measured 90 mm in length and up to 25 mm in diameter. The serosa was congested and oedematous and the mucosa appeared yellow and rough. No calculi were received. Microscopically, the gall bladder showed extensive pyloric metaplasia involving the entire gallbladder mucosa. Severe cholesterolosis with small cholesterol polyps was also seen but no significant inflammatory changes were detected and there was no mural fibrosis (Fig. 1). No Helicobacter-like organisms were detected. The metaplastic epithelium was voluminous and pale (Fig. 2A,B). These showed small amounts of neutral mucin by the Alcian blue-periodic acid Schiff (PAS) stain with and without diastase. The cytoplasm of these cells showed light eosinophilic to magenta discolouration as depicted in Fig. 2A,B. These cells appeared to be storing large amounts of lipids, hence the light staining with mucin stains. Immunostains for broad spectrum cytokeratin AE1/AE3 and for the macrophage marker CD68 were also performed. The histiocytic cells in the stroma were strongly CD68 positive (Fig. 2C) and the glandular epithelium was AE1/AE3 positive (Fig. 2D). Reduced cytokeratin staining was seen in the lipid laden epithelium (Fig. 2D, upper), whereas the nonaffected epithelium showed strong staining (Fig. 2D, lower). Cholesterolosis in the gall bladder is usually a diffuse process but a polypoid form can also occur. The latter often produces cholesterol polyps. Cholesterolosis is frequently observed in gall bladders removed for cholelithiasis or chronic cholecystitis. A prevalence of up to 8% has been noted.1 Terada found cholesterolosis and cholesterol polyps in 11% and 6% of gall bladders, respectively.2 Others like Tyagi et al. have found cholesterolosis to be less frequent and in a study of 415 cholecystectomy specimens they observed cholesterolosis in 2.7% of the specimens.3 In cholesterolosis there is an accumulation of cholesterol esters and triglycerides in macrophages in the gallbladder mucosa. An increase in the activity of the cholesterol ester enzyme has been observed in patients with cholesterolosis. It has been suggested that alteration in bile composition in the gall bladder is responsible for cholesterolosis, but this hypothesis has not been fully substantiated.1 Cholesterolosis is usually an incidental finding and it is unclear whether it causes clinical symptoms. However, the patient in this report had symptoms of gall bladder disease for up to 4 years. Although the patient’s symptoms may not have been directly related to cholesterolosis, she had a number of small cholesterol polyps (Fig. 1F) and detachment and passage of one or more of the polyps may have contributed to her symptoms. The absorption of cholesterol in the gall bladder which is akin to that in the small intestine is predominantly passive, related to


CORRESPONDENCE

661

Fig. 1 The histological attributes of the gall bladder in the routine H&E stained sections. (A) Expansion of the mucosal fold and replacement of the normal glandular tissue by pyloric type glands. (B) Higher power view of the mucosa with pyloric type glands. The cytoplasm of most of the epithelial cell appears clear due to lipid accumulation. (C) Almost complete replacement of the normal cytoplasmic structures by lipid. (D) Unaffected cells (right upper) and cells completely replaced by lipid. (E) Lipid rich metaplastic epithelium and foamy (xanthomatous) macrophages in the stroma. (F) Cholesterol polyp with massive lipid deposition. Thin arrows show the lipid laden cryptal epithelial cell in D and E and thick arrows show the lipid laden stromal macrophages in C and E.

the high luminal concentration of cholesterol, bile salts and phospholipids. This is supported by the similarity in histochemical demonstration of lipids in gall bladder epithelium at cholecystectomy and in intestinal epithelium after lipid challenge.4 An increase in the biliary content of dihydroxy bile salts has been shown to increase gall bladder permeability to both hydrophobic and hydrophilic molecules. This may lead to the accumulation of lipids in the mucosa and cholesterolosis.4 Pyloric metaplasia is an adaptive process and like all metaplastic processes it represents a response to chronic injury. The frequency of pyloric metaplasia in cholecystectomies has not been fully elucidated, but in a study by Terada pyloric gland metaplasia was found in 54% of the patients,2 Seretis and colleagues found metaplastic features in 25.6% of resected gallbladder specimens,5 and a study from Paraguay found the prevalence of pyloric metaplasia and intestinal metaplasia in cholecystectomies in 22.6% and 2.1%, respectively.6 In this patient there was no apparent injury such as cholelithiasis or chronic cholecystitis but the gall bladder showed extensive cholesterolosis. In the absence of other aetiological factors, high luminal cholesterol levels may have contributed to this metaplastic change. This is a likely explanation as the metaplastic change involved the entire mucosal surface of the gall bladder. Tayagi et al. studied 415 cholecystectomy specimens and found cholesterolosis in only 2.7% of the specimens.3 The

patients with cholesterolosis in that study were multiparous females and of slightly younger age than the average of 43.6 years in the cohort.3 The patient in this study was 19 years old at the time of surgery and had no gall stones or cholecystitis and was nulliparous. Interestingly Zhou and others have found an association between Helicobacter pylori and metaplasia in the gall bladder. They appear to suggest a possible relationship between H. pylori infection and precancerous and even cancerous lesions of the gall bladder.7 This patient did not have H. pylori-like organisms either in her stomach in a previous biopsy or in the gallbladder in the cholecystectomy specimen. Minor degrees of cholesterolosis and pyloric metaplasia are commonly encountered in the gall bladder. These conditions may coexist as they are both relatively common. Extreme degrees of both, in the abscess of chronic inflammation, are unusual. Cholesterol storage in metaplastic epithelium has not been reported. Pathologists are familiar with pyloric gland metaplasia as this is a common finding in gall bladders removed for cholelithiasis or other gall bladder disease. The process is benign and attracts little interest. However, lipid accumulation in metaplastic glands can produce an abnormal morphology which can mimic primary or metastatic carcinoma, particularly from the kidney. Therefore, awareness of this condition can prevent an erroneous diagnosis of cancer.


662

CORRESPONDENCE


Fig. 2 Histochemical and immunohistochemical characteristics of the gall bladder. (A) Diastase treated Alcian blue-periodic acid Schiff (ABDPAS) stained section showing eosinophilic to magenta coloured neutral mucin in the metaplastic cells. (B) ABDPAS stained section showing the neutral mucin in the metaplastic cells (upper) and almost complete lack of staining in the lipid laden epithelial cells. (C) Immunoperoxidase stained section for the macrophage marker CD69 (diaminobenzidine as chromogen) showing strong staining (left upper) and no staining (right lower) in the epithelial cell. Thin arrow shows the macrophages. (D) Immunoperoxidase stained section for the broad spectrum cytokeratin AE1/AE3 (diaminobenzidine as chromogen) showing strong staining in the metaplastic cell unaffected by lipid accumulation (lower) and almost no staining the lipid laden epithelial cell (upper). Thick arrow shows the epithelial cell.

Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Ibrahim Zardawi Douglass Hanly Moir Pathology, and University of Newcastle, NSW, Australia Contact Professor I. Zardawi. E-mail: [email protected], [email protected] 1. Sandri L, Colecchia A, Larocca A, et al. Gallbladder cholesterol polyps and cholesterolosis. Minerva Gastroenterol Dietol 2003; 49: 217–24. 2. Terada T. Histopathologic features and frequency of gall bladder lesions in consecutive 540 cholecystectomies. Int J Clin Exp Pathol 2013; 6: 91–6. 3. Tyagi SP, Tyagi N, Maheshwari V, Ashraf SM, Sahoo P. Morphological changes in diseased gall bladder: A study of 415 cholecystectomies at Aligarh. J Indian Med Assoc 1992; 90: 178–81. 4. Ross PE, Butt AN, Gallacher C. Cholesterol absorption by the gall bladder. J Clin Pathol 1990; 43: 572–5. 5. Seretis C, Lagoudianakis E, Gemenetzis G, Seretis F, Pappas A, Gourgiotis S. Metaplastic changes in chronic cholecystitis: implications for early diagnosis and surgical intervention to prevent the gallbladder metaplasiadysplasia-carcinoma sequence. J Clin Med Res 2014; 6: 26–9. 6. Segovia Lohse HA, Cuenca Torres OM. Prevalence and sequence of metaplasia-dysplasia-carcinoma of the gallbladder. A single centre retrospective study. Cir Esp 2013; 91: 672–5. 7. Zhou D, Guan WB, Wang JD, Zhang Y, Gong W, Quan ZW. A comparative study of clinicopathological features between chronic cholecystitis patients with and without Helicobacter pylori infection in gallbladder mucosa. PLoS One 2013; 8: e70265.

DOI: 10.1097/PAT.0000000000000170

Cytological features of pancreatic intraductal tubulopapillary neoplasm and an unexpected immunohistochemical profile

Sir, Pancreatic intraductal tubulopapillary neoplasm (ITPN) is a recently recognised intraductal proliferation of presumably pancreatobiliary epithelium without evidence of mucin. ITPN is histologically distinct from the usual mucinous intraductal papillary neoplasms and shows positive immunostaining for MUC-1. The present case highlights cytological features of ITPN and compares them with the resected specimen cytohistology. In addition, an unexpected finding regarding immunohistochemical profile of ITPN is described. A 48-year-old male with no previous medical history presented with abdominal pain. He was found to have an approximately 2.0 cm hypoechoic mass in the mid-body of the pancreas to the left of the portal vein. The mass was extending into the pancreatic tail and was obstructing the main pancreatic duct of that region. Ectatic side branches as well as associated chronic pancreatitis were noted. Solid pseudopapillary neoplasm, cystic neuroendocrine tumour and intraductal papillary mucinous neoplasm were considered. An endoscopic ultrasound guided fine needle aspiration was performed. A Thin Prep was prepared from the fresh sample. A cell block was made after routine processing, and embedded in paraffin. Sections 4 mm thick were stained with haematoxylin and eosin (H&E). Immunostains were performed using


CORRESPONDENCE

commercially available antibodies: CK7 (Dako, Denmark), CK20 (Dako), chromogranin (Ventana Medical Systems, USA), synaptophysin (Leica NCL, Germany), CDX2 (Leica NCL), progesterone receptor (Leica NCL), CD10 (Leica NCL), beta-catenin (BD, USA), PAX-8, P504S (Biocare, USA), vimentin, MUC1 (Leica NCL), and MUC5AC (Leica NCL). Appropriate positive and negative controls were used throughout. The cytology Papanicolaou slide showed a highly cellular sample. Clusters of cohesive cells showing papillary (Fig. 1A) and tubulopapillary architecture were observed (Fig. 1B). Numerous single cells and rare stromal acellular material were present. The neoplastic cells, including single cells, were cytologically bland, most with elongated nuclei, occasional nuclear grooves, inconspicuous nucleoli and rare pseudoinclusions (Fig. 1C). The cytoplasm was scant to moderate. Rare mitotic figures were noted. Some cells showed vacuolated cytoplasm; however, no obvious mucin was present. A corresponding cell block showed similar cells admixed with benign acinic cells. Immunohistochemical stains performed on cell block sections showed that the tumour cells were strongly and diffusely positive for cytokeratin CK7 and vimentin (Fig. 1D) and negative for CK20, chromogranin, synaptophysin, CDX2, progesterone receptor (PR), CD10, beta-catenin, PAX-8 and P504S. In the absence of mucin and an unusual immunohistochemical profile of coexpression of vimentin and CK7, a diagnosis of ‘papillary epithelial neoplasm, favour pancreatic origin’ was made.

663

Two months later the patient underwent distal pancreatectomy with splenectomy. The specimen showed a side-branch type intraductal tumour 1.3 cm in greatest dimension, filling the expanded duct and showing tubulopapillary architecture (Fig. 2A). The cells were cuboidal with round or elongated nuclei, occasional grooves and unsuspicious nucleoli, similar to the previous cytology specimen. However, the histology specimen showed more frequent mitotic figures and focal necrosis (Fig. 2B). No evidence of invasion or metastatic disease was identified. Immunostains were performed on the resection specimen, showing results identical to those performed earlier on the cell block specimen. In addition, the tumour was positive for MUC-1 and negative for MUC-5AC and antichymotrypsin. Interestingly, all benign ducts in the vicinity of the tumour were negative for vimentin. These findings were consistent with pancreatic branch-duct intraductal tubulopapillary neoplasm. The coexpression of vimentin and CK7 has not been reported for pancreatic intraductal neoplasms. Therefore, three cases of conventional intraductal papillary mucinous neoplasms (IPMN) of pancreatobiliary, gastric and intestinal subtypes were also stained for CK7 and vimentin. While these all expressed CK7, none was positive for vimentin. Intraductal tubulopapillary neoplasm of the pancreas was first introduced by Yamaguchi and colleagues in 20091 and is a new class of pancreatic tumour in the World Health Organization’s Classification of Tumors of the Digestive System.2 Since the first publication, similar neoplasms have been described in the bile ducts.3 By definition tubulopapillary neoplasm of the

A

B

C

D

Fig. 1 (A) The aspirate from the pancreas showing single cells and cohesive clusters in papillary arrangement with noticeable fibrovascular cores (Papanicolaou). (B) Tubulopapillary arrangement of the neoplastic cells (Papanicolaou). (C) The cytologically bland neoplastic cells with round and elongated nuclei, occasional nuclear grooves and inconspicuous nucleoli (Papanicolaou). (D) Neoplastic cells strongly positive for vimentin.


664


CORRESPONDENCE

A

B Fig. 2 (A) Resection specimen showing nodular proliferation of neoplastic cells with tubulopapillary architecture expanding a side-branch duct (H&E). (B) A solid area of tumour with mitotic figures and focal necrosis. The neoplastic cells show cytological features similar to the cytology specimen (H&E).

pancreas is characterised by a macroscopically visible solid nodular tumour occupying dilated ducts; no visible mucin; tubulopapillary growth; uniform high grade atypia throughout the neoplasm; easily recognisable necrotic foci; evident ductal differentiation, as indicated by the expression of cytokeratin 7 and/or cytokeratin 19; lack of acinar differentiation, as indicated by the absence of trypsin; absence of expression of MUC 2, MUC 5AC, and fascin.3 ITPN is also characterised by the absence of KRAS and BRAF mutations commonly found in other types of intraductal neoplasms such as pancreatic intraepithelial neoplasia (PanIN) and intraductal papillary mucinous neoplasm (IPMN).1,4 The most recent molecular findings include mutations in PIK3CA that were strongly associated with expression of phosphorylated AKT. Although the finding was present in most ITPNs, few IPMNs also showed the expression of AKT. The cytological features of ITPN have not been well studied. Guan et al.5 reported a case in a 41-year-old woman who was found to have a 2.2 cm cyst in the head of the pancreas. A fineneedle aspiration was performed; a cytospin preparation showed a few clusters of highly atypical cells from which a broad differential could be drawn. No unique or characteristic cytological features of ITPN were described in this case. In contrast, our case had a highly cellular cytology specimen with well-preserved cytoarchitecture features comparable with those in the resection specimen. The ample material allowed further

immunohistochemical work-up, which facilitated exclusion of several differential diagnoses as described below. In our cytology specimen the cells were rather bland with round and elongated nuclei and rare nuclear grooves, occasional pseudonuclear inclusions and rare mitotic figures. Of note is that the cytological specimen of the present case did not show significant atypia which could be attributed to the sampling issue. Based on those cytological features and the presence of papillary and tubular arrangements of the cells, a diagnosis of solid pseudopapillary neoplasm was considered. However, no intra- or extracellular hyaline globules, slender cytoplasmic processes or naked capillaries were observed. Acinar cell carcinoma was unlikely due to the above architectural and cytoplasmic features. Usually acinar cell carcinoma is well differentiated and the cells are similar to the non-neoplastic counterpart, with round nuclei with prominent nucleoli and granular cytoplasm, which the present case did not show. Architecturally, the case was similar to intraductal papillary mucinous neoplasm, except for the lack of distinct mucinous epithelium and extracellular mucin. The immunohistochemical profile excluded neuroendocrine and acinar differentiation. The tumour was negative for beta-catenin (nuclear), CD10 and PR. There was strong and diffuse expression of cytokeratins which excluded solid pseudopapillary neoplasm. However, the unexpected immunoreactivity to vimentin, which was confirmed on the resection specimen, complicated the picture at the time of biopsy and widened the differential diagnosis. The unusual coexpression of vimentin and CK7 as well as the architectural and cytological features raised the possibility of metastatic papillary renal cell carcinoma type 1 although the absence of renal mass and negative PAX8 and P504S made this differential diagnosis unlikely. In view of this unusual immunohistochemical finding, three cases of IPMN of different epithelial differentiation were stained immunohistochemically for vimentin, all of which were negative. The expression of vimentin is of unknown origin and is rarely present in pancreatic neoplasms, except in solid pseudopapillary tumour for which a cell of origin is still a mystery. An interesting feature of the current case is that all neoplastic ducts are lined by vimentin-positive epithelium and are clearly distinct from adjacent cytologically similar but vimentin negative non-neoplastic ductal epithelium. Since no myoepithelial cells are found in the pancreas no other plausible explanation for this phenomenon is available at this time. A similar question of origin is found in most vimentin positive renal cell carcinomas, where the tubular origin of renal cell carcinoma is postulated, although the only vimentin positive structure in the normal kidney is the glomerulus. No myoepithelial cells are described in the kidney either. This is the first report that provides a detailed cytological description of pancreatic ITPN. The unexpected immunoreactivity to vimentin may be helpful in differentiating ITPN from other intraductal papillary neoplasms of the pancreas and is worth further evaluation in larger scale studies. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Lei Zhao John Hart Shu-Yuan Xiao Tatjana Antic


CORRESPONDENCE

Department of Pathology, The University of Chicago, Chicago, IL, USA Contact Dr T. Antic. E-mail: [email protected] 1. Yamaguchi H, Shimizu M, Shinichi B, et al. Intraductal tubulopapillary neoplasm of the pancreas distinct from pancreatic intraepithelial neoplasia and intraductal papillary mucinous neoplasms. Am J Surg Pathol 2009; 33: 1164–72. 2. Adsay NV, Fukushima N, Furukawa T, et al. Intraductal neoplasms of the pancreas. In Bosman FT, Hruban RH, Carneiro F, et al., editors. WHO Classification of Tumours of the Digestive System. Lyon: IARC, 2010; 304–13. 3. Katabi N, Torres J, Klimstra DS. Intraductal tubular neoplasms of the bile ducts. Am J Surg Pathol 2012; 36: 1647–55. 4. Yamaguchi H, Kuboki Y, Hatori T, et al. Somatic mutations in PIK3CA and activation of AKT in intraductal tubulopapillary neoplasms of the pancreas. Am J Surg Pathol 2011; 35: 1812–7. 5. Guan H, Gurda G, Lennon AM, et al. Intraductal tubulopapillary neoplasm of the pancreas on fine needle aspiration: case report with differential diagnosis. Diagn Cytopathol 2014; 42: 156–60.

DOI: 10.1097/PAT.0000000000000172

Endosalpingiosis in an axillary lymph node with synchronous micro-metastatic mammary carcinoma Sir, Ectopic epithelial inclusions are extremely rare in axillary lymph nodes, but if present may generate a false positive diagnosis of metastatic carcinoma during intraoperative or routine assessment.1 Herein presented is a case of endosalpingiosis (ES) presenting in an axillary sentinel lymph node in a patient with breast carcinoma and concurrent nodal micrometastasis. The patient, a 51-year-old female, presented with a 3 week history of thickening in the upper pole of the right breast. Mammogram showed suspicious calcification throughout the upper pole of the right breast which correlated with a 4 cm area of distortion on ultrasound from 10 to 12 o’clock. Magnetic resonance imaging (MRI) demonstrated an at least 6.4 cm area of suspicious enhancement extending from the 9 to 2 o’clock position with extension to the nipple. Right axillary lymph nodes appeared normal on both ultrasound and MRI scans. Core biopsies made at initial and repeat ultrasounds from the 10 and 12 o’clock positions revealed low nuclear grade ductal carcinoma in situ (DCIS). Subsequently, a nipple sacrificing, skin-sparing right mastectomy was performed. Intraoperative imprint cytological examination was performed on three right axillary sentinel nodes, which were thought clinically suspicious, but no epithelial cells were seen microscopically. The specimens were fixed in 10% neutral buffered formalin. Macroscopically, in the lateral breast, concentrated in the upper outer quadrant, there was an area of granularity over 10 cm in area. Microscopically, this corresponded to extensive mass forming DCIS from the 8 to 1 o’clock position including deep to the nipple. It had mixed morphology but prominent cribriform and micropapillary growth patterns. At the 11 o’clock position, there was an at least 13 mm span of invasive carcinoma with an encapsulated

665

papillary type appearance, with cribriform and micropapillary architecture within expansive nodules devoid of a surrounding myoepithelial layer that was demonstrated elsewhere in multiple blocks with calponin and p63 immunohistochemical markers (Fig. 1A). It was strongly oestrogen receptor (ER) and progesterone receptor (PR) positive, Her2 negative by SISH method and the MIB-1/Ki-67 proliferation index was less than 10%. Sentinel lymph node one displayed a 0.3 mm micro-metastasis (Fig. 1B). Sentinel lymph node two contained a collapsed duct at least 2 mm in diameter and further smaller ducts within the node and in the subcapsular region (Fig. 2A). Ciliated columnar cells with intervening intercalated peg cells resembling fallopian tube epithelium lined the ducts (Fig. 2B). No endometrial type stroma surrounded the ducts. The epithelial cells sat on a defined basement membrane. They expressed cytokeratin 7, PAX8 (weak), ER and PR immunomarkers (Fig. 2C,D). They did not express CK20, GATA3, calretinin or TTF1 immunomarkers and no investing myoepithelial layer was seen with calponin and p63. In view of its mu¨llerian or fallopian type epithelium-like appearance, it was reported as axillary nodal ES. The micro-metastasis in sentinel node one appeared different and a diagnosis of micro-metastasis was upheld. The latter was present in only one haematoxylin and eosin (H&E) stained level, and could not be further assessed immunohistochemically. Sentinel node three was uninvolved by either process. Benign heterotopic inclusions in lymph nodes represent foci of ectopic non-neoplastic tissue. They typically are either

A

B

Fig. 1 (A) Right breast carcinoma with papillary architecture (H&E). (B) Right axillary sentinel node 1, with a 0.3 mm micrometastasis in the subcapsular region (H&E).


666


CORRESPONDENCE

A

B

C

D

Fig. 2 (A) Sentinel node 2 with endosalpingiosis (H&E). (B) Endosalpingiosis detail showing ciliated columnar cells and intercalated peg cells (H&E). (C) PAX-8 expression by endosalpingiosis. (D) Oestrogen receptor expression by endosalpingiosis.

epithelial, naevomelanocytic, decidual or mesothelial and are a potential source for confusion with metastatic disease. They have been reported at various sites including pelvis, groin, abdomen, mediastinum, axilla and neck. Depending upon site, epithelial cell types may have a variety of origins or phenotypes including mu¨llerian, mammary, pancreas, salivary, squamous and thyroid. Axillary lymph nodes more commonly show mammary and naevomelanocytic inclusions than other types.2,3 Benign epithelial inclusions, including glandular, are extremely rare in axillary lymph nodes; in a recent autopsy study, none were found in 3904 carefully examined lymph nodes from bilateral axillary dissections of 80 female and male individuals. This result suggests that they are far rarer than the 5% toted in the literature. Their presence may lead to a false diagnosis of metastatic carcinoma during intraoperative examination by various modalities including with PCR based methods.1 Glandular inclusions in axillary nodes are typically located in the capsule, septa or subcapsular region of the involved node. They may be mammary, mu¨llerian, squamous or mixed glandular and squamous. Mammary inclusions typically comprise an epithelial and myoepithelial layer akin to normal breast epithelium. However, they may show a variety of changes as seen in the breast including apocrine metaplasia and sclerosing adenosis, and malignant transformation may also occur.4 Mu¨llerian inclusions (or ES) are the most common reported nodal inclusions and are found almost exclusively in pelvic lymph nodes with an incidence ranging from 2% to 34%.2,3 They are extremely rare above the diaphragm. Exceptionally, ES has been described in an intramammary lymph node

presenting as microcalcification on mammogram5 and has been described in axillary nodes on only a few occasions.4,6–12 It is important to be aware of the possibility of the phenomenon outside of the pelvis as it may be misdiagnosed as nodal metastasis not only from breast but also from papillary tumours of the ovary or thyroid.10 In this instance there was papillary carcinoma within the breast, interpreted as an admixture of (predominantly) in situ and (focal) invasive carcinoma, the latter with encapsulated papillary features. Although controversy exists as to the true nature of encapsulated papillary carcinoma of breast, lymph node metastasis has been previously documented.13 Histologically, mu¨llerian inclusions or ES comprises fallopian tube type epithelium including ciliated columnar cells, intercalated peg cells and mucinous cells, presenting outside of a fallopian tube. It may form branching glands, the lining may have a papillary configuration and psammomatous calcification may be seen.4 The calcification may on occasion be the reason for its detection.5 Axillary lymph node examples are typically detected during their examination in breast cancer patients.4,6 – 12 Endometrial-like stroma is not noted. Atypical features include cytological atypia and mitotic activity.11 ‘Florid ES’ involved 22 axillary nodes in one case study.10 In several instances ES been described coexisting in lymph nodes with metastatic breast carcinoma.6,10 The immunoprofile of ES is that of tubal epithelium and includes expression of CK7, WT1, PAX8, HBME1, oestrogen and progesterone receptors. It lacks expression of mammary markers such as mammaglobin and GATA3 and lacks a surrounding myoepithelial layer with markers such as p63,


CORRESPONDENCE

calponin and smooth muscle actin.4,9,11–12 TTF1 is not expressed and the Ki-67 proliferation index is low.10 ES has an unclear aetiology. In the abdomen, ES may be the result of mu¨llerian metaplasia of pelvic peritoneal mesothelium or submesothelium, the most commonly accepted theory. Alternatively, it could arise from peritoneal implantation of sloughed or surgically displaced fallopian tube epithelium. Lymphatic dissemination of shed tubal epithelium or of metaplastic peritoneal mesothelium may explain its presence within lymph nodes.3,9 In this case, the ES did not transfer on imprint cytology and it was recognised in routine histological examination. Although the epithelium within the other lymph node could not be further assessed immunohistochemically because it cut out in deeper levels, the histological appearance was not typical of ES. Knowledge of the rare possibility of ES in axillary lymph nodes is important as it may potentially lead to unnecessary axillary lymph node dissection with its inherent morbidity. Acknowledgements: With thanks to Merran McKesser. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Mark Wilsher1 Kylie L. Snook2 1

Douglass Hanly Moir Pathology, Macquarie Park, and Breast and Surgical Oncology at the Poche Centre, North Sydney, NSW, Australia 2

Contact Dr M. Wilsher. E-mail: [email protected] 1. Iken S, Schmidt M, Braun C, et al. Absence of ectopic epithelial inclusions in 3,904 axillary lymph nodes examined in sentinel technique. Breast Cancer Res Treat 2012; 132: 621–4. 2. Sawicki MP, Howard TJ, Passaro E. Heterotopic tissue in lymph nodes. An unrecognized problem. Arch Surg 1990; 125: 1394–9. 3. Spinardi JR, Goncalves IRD, La Falce TS, et al. Benign inclusions in lymph nodes. Int J Morphol 2007; 25: 625–9. 4. Fellegara G, Carcangiu ML, Rosai J. Benign epithelial inclusions in axillary lymph nodes: report of 18 cases and review of the literature. Am J Surg Pathol 2011; 35: 1123–33. 5. Henley JD, Michael HB, English GW, et al. Benign mu¨llerian lymph node inclusions. An unusual case with implications for pathogenesis and review of the literature. Arch Pathol Lab Med 1995; 119: 841–4. 6. Piana S, Asioli S, Cavazza A. Benign Mu¨llerian inclusions coexisting with breast metastatic carcinoma in an axillary lymph node. Virchows Arch 2005; 446: 467–9. 7. Norton LE1, Komenaka IK, Emerson RE, et al. Benign glandular inclusions a rare cause of a false positive sentinel node. J Surg Oncol 2007; 95: 593–6. 8. Peng Y, Ashfaq R, Ewing G, et al. False-positive sentinel lymph nodes in breast cancer patients caused by benign glandular inclusions: report of three cases and review of the literature. Am J Clin Pathol 2008; 130: 21–7. 9. Corben AD, Nehhozina T, Garg K, et al. Endosalpingiosis in axillary lymph nodes: a possible pitfall in the staging of patients with breast carcinoma. Am J Surg Pathol 2010; 34: 1211–6. 10. Stolnicu S, Preda O, Kinga S, et al. Florid, papillary endosalpingiosis of the axillary lymph nodes. Breast J 2011; 17: 268–72. 11. Sarode VR, Euhus D, Thompson M, et al. Atypical endosalpingiosis in axillary sentinel lymph node: a potential source of false-positive diagnosis of metastasis. Breast J 2011; 17: 672–3. 12. Salehi AH, Omeroglu G, Kanber Y, et al. Endosalpingiosis in axillary lymph nodes simulating metastatic breast carcinoma: a potential diagnostic pitfall. Int J Surg Pathol 2013; 21: 610–2.

667

13. Lakhani SR, Ellis IO, Schnitt SJ, et al. WHO Classification of Tumours of the Breast. 4th ed. Lyon: IARC, 2012; Chapter 7, Intraductal papillary lesions.

DOI: 10.1097/PAT.0000000000000173

Mammary analogue secretory carcinoma of minor salivary glands Sir, Mammary analogue secretory carcinoma (MASC) is a recently described distinctive salivary gland neoplasm with similar morphological and immunohistochemical attributes to secretory carcinoma (SC) of the breast.1 Prior to its description in 2010, MASC used to be regarded as one of a number of known benign and malignant salivary gland tumours.2,3 MASC is usually low grade but high grade transformation has recently been described.4 A case of MASC in the upper lip is presented and the differential diagnosis of this intriguing tumour is discussed. A 66-year-old man presented with a 6 month history of thickening associated with swelling on the inner surface of the upper lip. The swelling which was subject to accidental biting fluctuated in size. On examination, a mass which was thought to represent minor salivary gland tumour was found. The mass was excised and the patient is being clinically followed up with no evidence of disease recurrence. The resected specimen consisted of a mucosal covered piece of tissue 12 10 8 mm. The cut surface showed a partly cystic, brown, 10 mm nodule. An encapsulated partly cystic tumour was seen microscopically (Fig. 1A). The tumour was made up of eosinophilic cells with glandular areas and microcystic spaces (Fig. 1B,C). The central part of the tumour showed degeneration and necrosis (Fig. 1A). Haemosiderin deposition attributed to occasional accidental biting by the patient was seen in parts of the tumour. The tumour cells appeared ‘oncocytic’ with round to oval nuclei and granular eosinophilic cytoplasm (Fig. 1D,E). Mild nuclear atypia was seen but mitoses were very rare. Globular eosinophilic material was present in the cytoplasm of some of the tumour cells and slightly basophilic material was noted in most of the microcystic spaces (Fig. 1E). Neutral and acidic mucinous material was noted in a few cells. Most of the cells also showed periodic acid Schiff (PAS) positivity, accounting for the ‘oncocytic’ appearance in the routine sections. The adjacent minor salivary glands and the overlying surface epithelium were normal. The lines of excision were clear. The tumour cells were positive for gross cystic disease fluid protein 15 (GCDFP-15; Fig. 2A), S100 (Fig. 2B), mammaglobin A (Fig. 2C) and MUC4 (Fig. 2D). Carcinoembryonic antigen (CEA), epithelial membrane antigen (EMA), vimentin, P63, b-catenin and cyclin D1 were also positive. Oestrogen, progesterone, human epidermal growth factor 2 and epidermal growth factor 1 were not expressed. Cytokeratin 7 was positive but cytokeratin 20 was negative. The tumour had a low proliferative index with up to 5% of the cells expressing MIB-1. Therefore, the tumour showed all the attributes of MASC. MASC was previously most often diagnosed as adenocarcinoma NOS, acinic cell carcinoma (ACC), mucoepidermoid carcinoma (MEC), polymorphous low grade adenocarcinoma and cystadenocarcinoma or cystadenoma.2,3 Cystadenoma and


668


CORRESPONDENCE

A

H&E ×1

B

H&E ×4

C

D

H&E ×40

E

H&E ×20

H&E ×60

Fig. 1 Histological and histochemical properties of the lesion. (A) Three cross sections of the tumour. The tumour is circumscribed with cystic change and eosinophilic luminal contents. (B) Part of an encapsulated lesion (right) and normal minor salivary glandular tissue (upper left). The tumour displays a microcystic growth pattern. (C) Higher magnification of the tumour showing eosinophilic (right) and slightly basophilic cells (left). (D,E) Large eosinophilic cells with pseudoglandular growth pattern. The cells have oval to spindle shaped mildly atypical nuclei. An eosinophilic globule is seen in an oval space (centre, E) and slightly basophilic material is present within other spaces.

ACC were the initial differential diagnoses in this case. Indeed, Williams and Chiosea estimate that as many as 8% of salivary cystadenomas may represent cases of MASC.5 The majority of cases in the initial report were located in the parotid gland.1 Subsequent reports have shown that while the parotid gland is the most common site of origin, MASC may also arise in the other major salivary glands and the oral cavity, including the lip, buccal mucosa and soft palate. MASC usually shows a number of growth patterns including cystic, tubular, solid or papillary architecture. Eosinophilic cytoplasm, intraluminal or intracytoplasmic colloid-like secretions that stain positive for PAS and are diastase resistant are usually seen.2,3 The morphological spectrum of MASC has grown to include tumours with both macrocystic and microcystic areas, intracystic papillary patterns and solid areas, and the morphological differential diagnosis of MASC may be broader than previously considered, especially on small biopsy material.3 Other tumour types that may show similar architectural patterns include polymorphous low grade adenocarcinoma and adenoid cystic carcinoma.4 Immunohistochemically MASC shows positivity with numerous antibodies; the most useful being mammaglobin, MUC4 and S100. Mammaglobin immunoreactivity has been found in 55% of all salivary gland tumours.6 Mammaglobin positivity alone is not diagnostic of MASC. MUC4 has been reported in >80% of cases of MASC.7 MUC4 has also

been shown to be positive in MEC and MUC4 positivity in MEC has been associated with increased risk of recurrence and death.7 The prognostic value of MUC4 in MASC is unclear. S100 is variably expressed in many salivary gland tumours, particularly those with a myoepithelial element. Co-expression of S100 with mammaglobin has been regarded as a very useful marker in identifying MASC.8 However, these markers have also been shown to be expressed by the morphological mimics of MASC.2 Most studies have detected a low proliferation index by Ki-67 staining with 5–28% of MASC tumour cells expressing this marker.4 This tumour also had a low proliferative index with not more than 5% of the tumour cells expressing MIB1. MASC usually presents as a slow-growing, painless mass in adults.3 The tumour in our patient, in line with the majority of previously described cases, was circumscribed and completely encapsulated. Poor circumscription, infiltrative growth or invasion has been noted in about one-quarter of the cases.5 MASC has an overall favourable outcome but like ACC and other low-grade salivary gland carcinomas, MASC does have the capacity to occasionally behave aggressively in the form of recurrences, regional lymph node metastases, and even disease-related deaths.4 Currently there is no way to definitively predict the behaviour of MSAC. This is largely because MASC is such a rare tumour, accounting for

Cerebrospinal fluid analysis by flow cytometry in acute lymphoblastic leukemia: is it all that it is cracked up to be?

Cerebrospinal fluid analysis.

Cerebrospinal fluid analysis.

Continuing education--is it necessary?

Lower gastrointestinal bleeding: is urgent colonoscopy necessary for all hematochezia?

Cognitive rehabilitation for schizophrenia: is it possible? Is it necessary?

Neutral mechanical alignment: Is it Necessary?

NASOGASTRIC DECOMPRESSION IN CHOLECYSTECTOMY, IS IT NECESSARY?

Alteration of baricity: is it necessary?

Preoperative fasting - is it really necessary?

Government intervention: is it necessary for nursing?

Endometrial Assessment: When is it Necessary?

Lumbar puncture-related cerebrospinal fluid leakage on magnetic resonance myelography: is it a clinically significant finding?

Cerebrospinal fluid analysis after unprovoked first seizure.

Higher energy: is it necessary, is it worth the cost for radiation oncology?

Correcting for multiple-testing in multi-arm trials: is it necessary and is it done?

Is It Necessary to Prepare the Enamel before Dental Bleaching?

Surveillance colonoscopy in low-risk postpolypectomy patients: Is it necessary?

Routine discharge examination of babies: is it necessary?

Cerebrospinal fluid flow. III. Pathological cerebrospinal fluid pulsations.

White diet: is it necessary during tooth whitening?

Is it necessary to rename gestational diabetes mellitus?

Colonoscopy after CT-diagnosed acute diverticulitis: Is it really necessary?

Vaccination Against Hepatitis A for Hemophilic Patients: Is It Necessary?