A 66-year-old non-smoker Chinese female was hospitalised after her first episode of haemoptysis (approximate volume, 100 mL) in February 2012. She reported being in good health, except for an episode of right lower lobe pneumonia about 6 months before presentation, which was treated with antibiotics. Despite radiological recovery, she complained of occasional dry cough, malaise, and weight loss of approximately 10 pounds in the subsequent months.

At admission, she was afebrile, without dyspnoea, and with unremarkable physical findings. Apart from mild anaemia (haemoglobin, 107 g/L) and slightly elevated erythrocyte sedimentation rate of 47 mm/h, other laboratory tests including white cell and platelet counts, C-reactive protein, coagulation profile, and liver and renal function tests were within normal limits. Culture, acid-fast staining, and cytological examination of the sputum did not reveal any abnormality. Radiological examination of the chest revealed a peripheral shadow measuring 1.5 cm in diameter in the right middle zone. Thoracic computed tomography (CT) revealed a huge contrast-enhancing mass with low, non-homogeneous attenuation filling the right pulmonary artery (PA), extending up to the right upper and lower segmental pulmonary arteries, as well as into the mediastinum. Patchy foci were also noted in the pulmonary trunk (Fig 1). Peripheral soft tissue densities were noted in the right middle lobe, which may have been either infarcts or tumour deposits. Bronchoscopy and endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) were performed to examine the airways and to obtain histological samples from the mediastinal extension of the mass. No bleeding source or significant endobronchial lesions were identified during bronchoscopy. A heterogeneous mass with distinct margins extending from the lower end of trachea to the right hilar regions along the right main bronchus was revealed by EBUS. Endobronchial ultrasound-guided transbronchial needle aspiration of the mass was performed at the lower end of trachea (Fig 2a). Histological examination revealed tumour fragments with spindle cells lying in myxoid stroma with no definite differentiation (Fig 2b). Immunohistochemistry revealed strongly positive smooth muscle actin expression, while CD-31, EMA, CK, calretinin, and S-100 were all negative. These findings were consistent with a diagnosis of pulmonary artery sarcoma (PAS). Echocardiogram revealed heterogeneous masses inside the main pulmonary trunk, and between the aorta and the anterior aspect of left atrium, with no evidence of valvular dysfunction or pulmonary hypertension. Due to the presence of extensive disease and central location of the tumour, aggressive surgery was not considered appropriate by cardiothoracic surgeons. The patient declined palliative chemotherapy or radiotherapy and defaulted from follow-up.

Pulmonary artery sarcoma is a rare condition that was first reported in 1923 from an autopsy.1 Apart from a few case series and reviews,2 3 4 5 6 most cases were only isolated case reports. Although it is sometimes divided into intimal and mural types, and further histological subtypes, the adherent nature of the tumour and the frequent lack of tissue differentiation sometimes do not allow such classifications, such as in our patient. As both the clinical and radiological features can resemble closely that of pulmonary embolism (PE), some reported cases were initially treated with anticoagulation therapy but without response.3 4 A definite diagnosis is usually delayed and made either at autopsy or intra-operatively with frozen sections.5 While dyspnoea, cough, and chest pain are the commonest presenting symptoms,2 3 4 5 haemoptysis has also been reported.3 7 Lung involvement is commonly found,3 4 although the lung lesion in our patient could represent either tumour deposit or infarct, and both could have led to haemoptysis. On CT scan, the characteristics of PAS often mimic PE. Other features that support a diagnosis of PAS include a low-attenuation defect filling the whole lumen of main or proximal PA and extravascular extension,6 both of which were present in our patient. Considering the CT scan findings, along with the absence of clinical features and risk factors of vascular thrombosis, malignancy was suspected in our patient. Apart from CT scan, both magnetic resonance imaging (MRI; with gadolinium contrast enhancement)4 and fluorodeoxyglucose–positron emission tomography (FDG-PET)7 have been described as useful for the diagnosis of PAS. Imaging like CT and MRI may have enabled us to determine the exact nature of peripheral lung lesion in our patient.

The prognosis of PAS is poor, with a median survival of 1.5 months without any treatment.2 If possible, radical resection of the tumour should be considered since median survival has been demonstrated to be significantly better with this option (36.5 ± 20.2 months) than that with incomplete resection such as tumour debulking and thromboendarterectomy (median survival, 11 ± 3 months).4 Radical resections are usually major operations that might involve extensive resections and subsequent reconstructions and/ or prosthetic replacement of PA, pneumonectomy or thromboendarterectomy, and often requiring cardiopulmonary bypass.4 5 Multimodality treatment including neoadjuvant or adjuvant chemotherapy and radiation, together with surgery, might further improve survival.3 4 Owing to the rarity of PAS, currently there are no prospective data to guide the best management practices for PAS. Despite the apparent prognostic superiority of curative resection, an adequate cardiopulmonary reserve for the major surgery and the absence of extensive disease that precludes radical resection would be necessary. In view of the extensive disease, radical tumour resection was not considered in our patient. A definite diagnosis at an early stage would improve the chance of survival of PAS patients. However, in the past, such an opportunity was usually not available unless a patient underwent surgery.

Endobronchial ultrasound-guided transbronchial needle aspiration is a minimally invasive diagnostic procedure. The efficacy and safety of EBUS-TBNA, via the bronchoscopic route, in mediastinal and hilar lymph node staging of lung cancer have been well documented, with high sensitivity, specificity, and diagnostic accuracy of 92.3%, 100% and 98%, respectively.8 This treatment modality can offer a more cost-effective and less invasive diagnostic option than mediastinoscopy.9 It is also useful for diagnosing central pulmonary lesions adjacent to the bronchus10 and non-malignant conditions.11 Performed in an endoscopic suite, the procedure involves using a special endoscopic instrument incorporating a 7.5-MHz curvilinear ultrasonic transducer at the tip of a flexible bronchoscope, which allows needle aspiration of target lesions under real-time ultrasound guidance via a 22-gauge needle. The procedure is very safe and can be performed under local anaesthesia and conscious sedation.8 9 10 11 Our case is the second report in medical literature which illustrates the potential applicability of EBUS-TBNA in a histological diagnosis for PAS through a non-surgical route. In contrast to our case, in the first report describing two patients in literature, the TBNA involved puncturing of the pulmonary arterial wall in one of the patients undergoing the procedure.12 Although PA puncture in EBUS-TBNA has been reported to be safe,13 self-limiting intramural haematoma and haemopneumomediastinum have been described after an accidental EBUS-TBNA puncture,14 and so, the safety of EBUS-TBNA in the routine differential diagnosis of PAS and pulmonary thromboembolism have been doubted.15 In contrast, our patient had suspicious radiological features of PAS and absence of clinical features and risks for PE. We believe using EBUS-TBNA to approach extravascular extension of the lesion in such patients offers a relatively non-invasive means for early diagnosis of PAS. However, the safety of PA punctures in EBUS-TBNA needs to be clarified in large studies. It must be remembered that performing EBUS-TBNA in patients with acute PE and pulmonary hypertension may be associated with a high risk of complications such as respiratory distress and bleeding.15 Thus, the routine use of EBUS-TBNA as a tool to distinguish PE from the rarer PAS is not encouraged. While MRI or FDG-PET may be considered before performing invasive diagnostic investigations,4 the present report suggests that EBUS-TBNA can potentially provide clinicians a diagnostic alternative to surgical exploration for patients suspected of having PAS.

Pulmonary artery sarcoma is a rare disease with poor prognosis, which is often diagnosed late and frequently mimics PE. Endobronchial ultrasound-guided transbronchial needle aspiration may be a safe and easy option for the early diagnosis of the condition. Extravascular extension of the lesion on thoracic CT provides a diagnostic clue for PAS, and also offers a safe approach for performing EBUS-TBNA.

No conflicts of interest were declared by authors.

1. Mandelstamm M. Uber primare neubildungen des herzens [in German]. Virchows Arch 1923;245:43-54. CrossRef

2. Krüger I, Borowski A, Horst M, de Vivie ER, Theissen P, Gorss-Fengels W. Symptoms, diagnosis and therapy of primary sarcomas of pulmonary artery. Thorac Cardiovasc Surg 1990;38:91-5. CrossRef

3. Huo L, Moran CA, Fuller GN, Gladish G, Suster S. Pulmonary artery sarcoma: a clinicopathologic and immunohistochemical study of 12 cases. Am J Clin Pathol 2006;125:419-24. CrossRef

4. Blackmon SH, Rice DC, Correa AM, et al. Management of primary pulmonary artery sarcomas. Ann Thorac Surg 2009;87:977-84. CrossRef

5. Mayer E, Kriegsmann J, Gaumann A, et al. Surgical treatment of pulmonary artery sarcoma. J Thorac Cardiovasc Surg 2001;121:77-82. CrossRef

6. Yi CA, Lee KS, Choe YH, Han D, Kwon OJ, Kim S. Computed tomography in pulmonary artery sarcoma: distinguishing features from pulmonary embolic disease. J Comput Assist Tomogr 2004;28:34-9. CrossRef

7. Tueller C, Fischer Biner R, Minder S, et al. FDG-PET in diagnostic work-up of pulmonary artery sarcomas. Eur Respir J 2010;35:444-6. CrossRef

8. Yasufuku K, Nakajima T, Motoori K, et al. Comparison of endobronchial ultrasound, positron emission tomography, and CT for lymph node staging of lung cancer. Chest 2006;130:710-8. CrossRef

9. Navani N, Lawrence DR, Kolvekar S, et al. Endobronchial ultrasound-guided transbronchial needle aspiration prevents mediastinoscopies in the diagnosis of isolated mediastinal lymphadenopathy: a prospective trial. Am J Respir Crit Care Med 2012;186:255-60. CrossRef

10. Tournoy KG, Rintoul RC, van Meerbeeck JP, et al. EBUS-TBNA for the diagnosis of central parenchymal lung lesions not visible at routine bronchoscopy. Lung Cancer 2009;63:45-9. CrossRef

11. Wong M, Yasufuku K, Nakajima T, et al. Endobronchial ultrasound: new insight for the diagnosis of sarcoidosis. Eur Respir J 2007;29:1182-6. CrossRef

12. Park JS, Chung JH, Jheon S, et al. EBUS-TBNA in the differential diagnosis of pulmonary artery sarcoma and thromboembolism. Eur Respir J 2011;38:1480-2. CrossRef

13. Vincent B, Huggins JT, Doelken P, Silvestri G. Successful real-time endobronchial ultrasound-guided transbronchial needle aspiration of a hilar lung mass obtained by traversing the pulmonary artery. J Thorac Oncol 2006;1:362-4. CrossRef

14. Botana-Rial M, Nú-ez-Delgado M, Pallarés-Sanmartín A, et al. Intramural hematoma of the pulmonary artery and hemopneumomediastinum after endobronchial ultrasound-guided transbronchial needle aspiration. Respiration 2012;83:353-6. CrossRef

15. Montani D, Jaïs X, Sitbon O, Dartevelle P, Simonneau G, Humbert M. EBUS-TBNA in the differential diagnosis of pulmonary artery sarcoma and thromboembolism. Eur Respir J 2012;39:1549-50. CrossRef

A 10-year-old boy was referred from another hospital for malaise and on-and-off fever for 2 weeks in August 2010. There were no respiratory, urinary, gastro-intestinal, or neurological symptoms. Physical examination revealed a febrile child with pallor. There were multiple bruises over the lower limbs, multiple cervical and inguinal lymph nodes, and hepatosplenomegaly. The potassium level checked in the referring hospital was 4.3 mmol/L.

Laboratory investigations in our hospital revealed a white blood cell (WBC) count of 340.6 x 10⁹ /L with blast cell count of 282.69 x 10⁹ /L (83%). The neutrophil count was 45.41 x 10⁹ /L (13.3%), with 1.14 x 10⁹ /L (0.3%) metamyelocytes. The haemoglobin level was 71 g/L and platelet count was 68 x 10⁹ /L. Blood biochemistry revealed the following serum levels: potassium 6.0 mmol/L, sodium 137 mmol/L, urea 5.0 mmol/L, creatinine 61 μmol/L, calcium 2.33 mmol/L, phosphate 1.38 mmol/L, urate 0.4 mmol/L, and the serum lactate dehydrogenase concentration was 3800 IU/L. The potassium assay was repeated with blood specimen drawn from a venepuncture into a serum specimen bottle. The result was 7.4 mmol/L. It was suspected to be a factitious result related to clotting of the specimen or haemolysis. Another venous sample was collected into a heparinised bottle for plasma potassium level assay and was sent to laboratory by pneumatic tube for urgent analysis. The result was 8.2 mmol/L.

In view of the rapidly rising potassium level in a patient at risk of tumour lysis syndrome, despite cytotoxic chemotherapy not having been commenced, the patient was transferred to the paediatric intensive care unit. Electrocardiography revealed a sinus tachycardia of 125 beats/min with no features to suggest hyperkalaemic change. An arterial specimen drawn from an arterial line into a heparinised bottle was sent to the laboratory by special messenger. Treatment of hyperkalaemia was commenced before availability of the result in view of the perceived rapidly rising potassium level from 6.0 mmol/L to 8.2 mmol/L in less than 5 hours. He was given a dose of calcium gluconate, insulin with dextrose, and a dose of calcium polystyrene sulphonate resin. Hyperhydration with intravenous fluid 2.5 L/m²/day together with allopurinol, ceftazidime and amikacin were also commenced. The plasma potassium level in the pre-treatment arterial specimen was 4.2 mmol/L.

Potassium level was measured again after dextrose-insulin using an arterial heparinised sample. The specimen was sent to laboratory by pneumatic tube for urgent analysis. The result was 9.2 mmol/L, which was suspected to be factitious. Therefore paired serum and plasma specimens were sent to the laboratory by special messenger for careful, urgent processing. The resulting plasma potassium level was 4.1 mmol/L and the serum sample showed interference from potassium leakage. As the condition of pseudohyperkalaemia became evident, no further therapy for hyperkalaemia was given. The patient was then transferred to a paediatric oncology centre, where the diagnosis of acute T-cell lymphoblastic leukaemia was confirmed. His electrocardiogram and blood glucose level were normal all along; the plasma potassium level checked before transfer was 4.6 mmol/L. The Table shows a summary of serum and plasma potassium level results with reference to the sampling time after hospital admission and processing techniques. Arterial line samples drawn into heparinised bottles with delivery by messenger showed more reliable and consistent results.

Hyperkalaemia is a potentially life-threatening condition for which emergency treatment may be necessary. However, pseudohyperkalaemia is a common laboratory artefact and, if unrecognised, may lead to a dilemma.1 2 Workup for falsely elevated potassium levels may lead to delay in treatment and waste of resources. On the other hand, aggressive treatment including renal dialysis may be unnecessarily commenced.3 Therefore, clinicians should be aware of the conditions that may give rise to falsely elevated potassium levels.

In most cases, pseudohyperkalaemia occurs during the collection process, transport, or storage of specimens.4 The artefact is due to leaching of potassium from cytosols during clotting or storage of the sample.1 Leakage of potassium from blood cells can occur as an in-vitro phenomenon during blood coagulation.5 6 7 This phenomenon is usually encountered in serum and not plasma. Serum potassium levels have been reported to be higher than plasma levels, with a mean difference of 0.36 ± 0.18 mmol/L in samples with a normal number of blood cells.8 Therefore anticoagulated plasma samples provide more accurate measurement of the true potassium level.1 Lithium heparin is the recommended anticoagulant for this purpose.4

Potassium release from red blood cells by in-vitro haemolysis is a well-recognised cause of spurious results. Conditions that induce in-vitro haemolysis include fist clenching during phlebotomy, drawing blood into an evacuated tube, use of small-gauge needles, use of tourniquets, cold storage, delay in sample processing, mechanical trauma during vigorous mixing, or hard centrifugation.1 3 There is a rare genetic condition, familial pseudohyperkalaemia, which is an autosomal dominant disorder associated with excessive leakage of potassium across red cell membranes.9

Pseudohyperkalaemia can also be the result of pre-existing pathological conditions resulting in cellular potassium leakage. Such conditions include acute leukaemia, chronic myeloproliferative disorders (chronic myeloid leukaemia, polycythaemia vera, essential thrombocythaemia), chronic lymphocytic leukaemia, and reactive thrombocytosis.1 10 The large number of blood cells may exaggerate the effects of potassium leakage from coagulation and further increase the discrepancy between serum and plasma potassium levels. Unphysiological conditions and shortage of metabolic fuels leading to impaired sodium/potassium adenosine triphosphatase activity may contribute to release of potassium from large numbers of white cells.7 The abnormal fragility of malignant leukocytes also makes them susceptible to mechanical stress. Colussi11 reported that the minor mechanical stress of drawing blood into vacuum tubes or syringe shaking induced lysis of leukaemic lymphocytes that appeared in blood smears as lymphocytic ghosts called “basket cells”. Kellerman and Thornbery3 reported the occurrence of pseudohyperkalaemia due to pneumatic tube transport in a leukaemic patient with a WBC count of 290 x 10⁹ /L, but they did not find any significant differences in potassium values between walked and tube-transported specimens in control patients with normal WBC counts. The effect on potassium resulting from pneumatic tube transport is likely due to both WBC number and fragility.3 Chawla et al12 also reported a case of pseudohyperkalaemia due to mechanical disruption of leukocytes in a patient with chronic lymphocytic leukaemia and proposed to designate this phenomenon as pneumatic tube “pseudo tumour lysis syndrome”. Ruddy et al13 also reported a chronic lymphocytic leukaemia patient with venous potassium levels spuriously higher than arterial potassium levels. The authors hypothesised that this was likely due to a greater opportunity for lysis of white blood cells in the venous blood related to differences in mechanical stressors between venous and arterial blood draw techniques.13 In our patient, the possible contributing factors for pseudohyperkalaemia include high WBC count, fragile leukaemic blast cells, clotting of serum specimens, and mechanical trauma secondary to pneumatic tube transport.

Pseudohyperkalaemia is characterised by an elevation of serum potassium levels in the absence of clinical evidence of electrolyte imbalance,10 and should be suspected when there are no other clinical features of hyperkalaemia, such as peaked T waves and QRS widening on the electrocardiogram.3 The potassium level should therefore be interpreted together with the clinical context and other investigation results. Hyperkalaemia is exceptionally unlikely if renal indices are normal and there are no predisposing factors, such as intake of potassium supplements and/or drugs that raise potassium levels.4 In our patient, the possibility of a spurious result was suspected at an early stage. Treatment was commenced because potassium levels showed a rising trend in a patient at risk of tumour lysis syndrome. Ultimately, the pseudohyperkalaemia was confirmed by elimination of the possible causes of measurement error. In this patient, there were hints to remind clinicians to consider the possibility of factitious results as the electrocardiogram was normal all along and the other expected biochemical changes of tumour lysis syndrome were absent.

Besides pseudohyperkalaemia, factitious hypokalaemia may also be encountered in patients with leukaemia with WBC counts higher than 100 x 10⁹ /L when blood samples are allowed to stand at room temperature.1 This phenomenon is related to transcellular potassium shift into leukaemic cells.1 Even in normal specimens, pseudohypokalaemia may also be noted if sample analysis is delayed, and is believed to be mediated by sodium-potassium-exchanging ATPase,14 while specimen deterioration due to long storage can lead to pseudohyperkalaemia.4

In conclusion, when considering investigation and treatment, clinicians should be aware of the potential causes of pseudohyperkalaemia in leukaemic patients. Extreme care in handling blood samples is very important. The use of pneumatic tube transport for potassium analysis should be avoided in leukaemic patients.

1. Dalal BI, Brigden ML. Factitious biochemical measurements resulting from haematologic conditions. Am J Clin Pathol 2009;131:195-204. CrossRef

2. Brigden ML, Dalal BI. Spurious and artifactual test results, II: morphologic abnormalities, pseudosyndromes and spurious test results. Lab Med 1999;30:397-405.

3. Kellerman PS, Thornbery JM. Pseudohyperkalaemia due to pneumatic tube transport in a leukaemic patient. Am J Kidney Dis 2005;46:746-8. CrossRef

4. Smellie WS. Spurious hyperkalaemia. BMJ 2007;334:693-5. CrossRef

5. Hartmann RC, Auditore JV, Jackson DP. Studies on thrombocytosis. Hyperkalaemia due to release of potassium from platelet during coagulation. J Clin Invest 1958;37:699-707. CrossRef

6. Bronson WR, DeVita VT, Carbone PP, Cotlove E. Pseudohyperkalaemia due to release of potassium from white blood cells during clotting. N Engl J Med 1966;274:369-75. CrossRef

7. Colussi G, Cipriani D. Pseudohyperkalaemia in extreme leukocytosis. Am J Nephrol 1995;15:450-2. CrossRef

8. Nijsten MW, de Smet BJ, Dofferhoff AS. Pseudohyperkalemia and platelet counts. N Engl J Med 1991;325:1107. CrossRef

9. Iolascon A, Stewart GW, Ajetunmobi JF, et al. Familial pseudohyperkalaemia maps to the same locus as dehydration hereditary stomatocytosis (hereditary xerocytosis). Blood 1999;93:3120-3.

10. Sevastos N, Theodossiades G, Efstathiou S, Papatheodoridis GV, Manesis E, Archimandritis AJ. Pseudohyperkalaemia in serum: the phenomenon and its clinical magnitude. J Lab Clin Med 2006;147:139-44. CrossRef

11. Colussi G. Pseudohyperkalaemia in leukaemias. Am J Kidney Dis 2006;47:373. CrossRef

12. Chawla NR, Shapiro J, Sham RL. Pneumatic tube "pseudo tumour lysis syndrome" in chronic lymphocytic leukaemia. Am J Hematol 2009;84:613-4. CrossRef

13. Ruddy KJ, Wu D, Brown JR. Pseudohyperkalaemia in chronic lymphocytic leukaemia. J Clin Oncol 2008;26:2781-2. CrossRef

14. Sodi R, Davison AS, Holmes E, Hine TJ, Roberts NB. The phenomenon of seasonal pseudohypokalemia: effects of ambient temperature, plasma glucose and role for sodium-potassium-exchanging-ATPase. Clin Biochem 2009;42:813-8. CrossRef

A 10-year-old Chinese girl presented to a paediatric clinic with epilepsy. She had normal intelligence and social interaction ability, and no relevant family history. She had been having recurrent sleep seizures, with generalised twitching of all four limbs, cyanosis, up-rolling of eyeballs, drooling, and urinary incontinence. Her head circumference was 61 cm (5 cm > the 97th percentile). Physical examination was otherwise normal. The electroencephalogram showed an abnormal focus at the right temporo-occipital region. Goldmann perimetry revealed no abnormality. Magnetic resonance imaging (MRI) with contrast showed pachygyria (total loss of sulcation) with hyperplastic white matter and disorganisation of the grey and white matters in the right occipital lobe. A 2.5 cm x 1.8 cm T2 hyperintense area around the right occipital horn was consistent with gliosis. The right occipital lobe was mildly enlarged. Repeated MRI 5 and 7 years after presentation showed no interval changes. Her findings were compatible with focal cortical dysplasia involving the right occipital lobe (Fig 1). The epilepsy was under good control with carbamazepine treatment. She developed a goitre at the age of 16 years, but was euthyroid and ultrasonography showed a multinodular goitre. Hypertrichosis was noted when she was 17 years old. Multiple tiny papules were noted at the perinasal region, of which the patient regarded as coarse skin. At the age of 19 years, she was incidentally found to have iron deficiency anaemia (haemoglobin 96 g/L [reference range, 117-149 g/L], mean corpuscular volume 68.5 fL [82-97 fL], serum iron 2 μmol/L [5.0-30.4 μmol/L], total iron binding capacity 77.1 μmol/L [44.8-76.0 μmol/L]). She had no gastro-intestinal symptoms, but did receive iron supplements.

She had a left hemithyroidectomy at the age of 22 years for increasing size of a dominant left-sided thyroid nodule that expanded from 1.7 × 1.4 × 0.8 cm to 3.6 × 2.9 × 4.2 cm over 10 months. Fine-needle aspiration showed lymphocytic thyroiditis, and excisional biopsy revealed adenomatous hyperplasia. At the age of 23 years, enlargement of a thyroid nodule in the right thyroid lobe (3 cm in diameter) was noted; fine-needle aspiration suggested it was an adenomatous nodule. Total thyroidectomy was performed, and excisional biopsy revealed an atypical nodule with atypical enlarged vesicular nuclei and small distinct nucleoli. The patient also developed a 3-cm scalp papilloma, shown by excisional biopsy to be fibrous papule. At the age of 24 years, she also experienced coffee ground vomiting; oesophagogastroduodenoscopy showed diffuse glycogen deposits at the lower oesophagus and multiple gastric polyps. Polypectomy yielded lymphoid hyperplasia, and oesophageal mucosa biopsy was reported as showing glycogenic acanthosis.

In view of her multiple benign tumours, the possibility of a hamartomatous polyposis syndrome was suspected. Subsequent clinical examination yielded multiple papillomas over the face and tongue, but there was no pigmentation of the lips. These mucocutaneous features were pathognomonic criteria of Cowden syndrome. Notably, macrocephaly, thyroid adenoma, gastro-intestinal hamartomas (oesophageal glycogenic acanthosis and gastric polyps), and skin fibromas fulfilled one major and three minor clinical criteria of Cowden syndrome. Mutational analysis of the PTEN gene (Fig 2) showed a heterozygous thymine deletion in exon 8 at position 1023 (NM_000314.4(PTEN_ v001):c.1023del). This deletion creates a frame shift starting at codon Phe341. The new reading frame ends in a stop codon 2 position downstream (NM_000314.4(PTEN_i001):p.(Phe341Leufs*3). This mutation was not detected in the mother. The patient’s father was deceased 10 years earlier due to lung cancer, but he did not have macrocephaly or a history of any other tumours.

Cowden syndrome was first described by Lloyd and Dennis in 1963.1 It is characterised by macrocephaly, mucocutaneous lesions, acral (extremity/limb) keratosis, papillomatous papules, and high risk of development of cancer in the breast, thyroid, and endometrium. It is a rare autosomal dominant disease, with an estimated prevalence of 1 in 1 000 000 to 250 000 based on genetic identification.2 Lesions can occur in tissues derived from all three embryonic germ cell layers. The subtle and variable clinical manifestations contribute to the difficulty in making a clinical diagnosis. In 1996, the international Cowden Consortium identified germline PTEN (phosphatase and tensin homologue on chromosome 10) mutations as a cause of Cowden syndrome.2 This is a tumour suppressor gene encoding a major lipid phosphatase that functions in the phosphoinositide 3-kinase signalling cascade.3 It regulates cellular processes crucial for normal development, including cell proliferation, soma growth, cell death, and cell migration.4 The importance on PTEN mutations in making a diagnosis, genetic counselling, and clinical surveillance for the development of malignancies is well recognised. Apart from cancer susceptibility, the PTEN mutation was also implicated as candidate gene in developmental disorder like autism and mental retardation.4

Our patient was first suspected to have Cowden syndrome because of the interesting endoscopic findings, as multiple gastric polyps are rarely encountered in young adults. Patients with such polyps should be examined for various genetic syndromes associated with gastro-intestinal polyps, including Peutz-Jeghers syndrome, Cowden syndrome, and Cronkhite-Canada syndrome.5 Histology shows Helicobacter pylori–negative, non-specific lymphoid hyperplasia. Diffuse oesophageal glycogenic acanthosis has seldom been seen in young adults. The concurrence of oesophageal glycogenic acanthosis and multiple gastric polyps is associated with Cowden syndrome.6 7 8

Adult-onset Lhermitte-Duclos disease (LDD), which presents clinically with progressive cerebellar sign and increased intracranial pressure, is also a feature. A dysplastic cerebellar gangliocytoma is now considered to be one of the central nervous system (CNS) pathognomonic criteria by the revised Cowden Syndrome Consortium. The other overt CNS manifestations of Cowden syndrome included macrocephaly, autism, developmental delay, brain tumour, and LDD. The unique feature in our patient was the presentation with generalised epilepsy in childhood and the identification of focal cortical dysplasia in the right occipital region. To our knowledge, such brain MRI imaging has not been described previously in Cowden syndrome.9 Apart from regulating cell growth, the PTEN gene also plays a role in cell migration. It interacts with focal adhesion kinase, which results in the inhibition of cell migration and spread. Focal cortical dysplasia is a kind of neuronal migration disorder. Our patient was noted to have a right occipital focal cortical dysplasia and right megalencephaly. These features suggest CNS manifestations of the PTEN mutation. Inclusion of this feature in Cowden syndrome may facilitate early diagnosis.

Cowden syndrome represents a late-onset phenotype of the PTEN mutation. Early presentation of PTEN mutations include Bannayan-Riley-Ruvalcaba syndrome and autism spectrum disorder with macrocephaly. Bannayan-Riley-Ruvalcaba syndrome is a congenital disorder characterised by macrocephaly, hamartomatous intestinal polyps, lipomas, and pigmented macules on the penis. To our knowledge, our patient represents a new PTEN mutation phenotype, with macrocephaly and focal cortical dysplasia at the occipital region and subsequent full-blown presentation of Cowden syndrome.

In conclusion, we believe we have identified a new phenotype of Cowden syndrome and a new PTEN indel mutation. This PTEN mutation appears to cause megalencephaly, epilepsy, focal cortical dysplasia in the occipital region in childhood, and multiple hamartomatous lesions in late adolescence and early adulthood. The inclusion of focal cortical dysplasia in the occipital region as a CNS feature of the PTEN mutation could facilitate early diagnosis of PTEN mutation syndromes, and clinicians to undertake early surveillance for possible malignancies. In the latest published prospective study of the Ohio cohort of patients with germline PTEN mutation, the penetrance of breast cancer was found to begin at around the age of 30 years rising to an estimated 85% lifetime risk.10 The lifetime risk of breast cancer in females with PTEN mutations was even higher than the best estimates for individuals with BRCA1 or BRCA2 mutations.11 The PTEN-related endometrial cancer risk begins at the age of 25 years rising to 30% by the age of 60 years.10 The thyroid cancer risk begins at birth and continues lifelong.10 Risks of colorectal and kidney cancers begin at around the age of 40 years, with a lifetime risk of 9% and 34%, respectively.10 The earliest reported age of onset of melanoma was in a 3-year-old patient.10

1. Lloyd KM 2nd, Dennis M. Cowden's disease. A possible new symptom complex with multiple system involvement. Ann Intern Med 1963;58:136-42. CrossRef

2. Farooq A, Walker LJ, Bowling J, Audisio RA. Cowden syndrome. Cancer Treat Rev 2010;36:577-83. CrossRef

3. Shen WH, Balajee AS, Wang J, et al. Essential role for nuclear PTEN in maintaining chromosomal integrity. Cell 2007;128:157-70. CrossRef

4. Orrico A, Galli L, Buoni S, Orsi A, Vonella G, Sorrentino V. Novel PTEN mutations in neurodevelopmental disorders and macrocephaly. Clin Genet 2009;75:195-8. CrossRef

5. Hizawa K, Iida M, Matsumoto T, et al. Gastrointestinal manifestations of Cowden's disease. Report of four cases. J Clin Gastroenterol 1994;18:13-8. CrossRef

6. McGarrity TJ, Wagner Baker MJ, Ruggiero FM, et al. GI polyposis and glycogenic acanthosis of the esophagus associated with PTEN mutation positive Cowden syndrome in the absence of cutaneous manifestations. Am J Gastroenterol 2003;98:1429-34. CrossRef

7. Vasovcak P, Krepelova A, Puchmajerova A, et al. A novel mutation of PTEN gene in a patient with Cowden syndrome with excessive papillomatosis of the lips, discrete cutaneous lesions, and gastrointestinal polyposis. Eur J Gastroenterol Hepatol 2007;19:513-7. CrossRef

8. Ha M, Chung JW, Hahm KB, et al. A case of Cowden syndrome diagnosed from multiple gastric polyposis. World J Gastroenterol 2012;18:861-4. CrossRef

9. Lok, C, Viseux V, Avril MF, et al. Brain magnetic resonance imaging in patients with Cowden syndrome. Medicine (Baltimore) 2005;84:129-36. CrossRef

10. Tan MH, Mester JL, Ngeow J, Rybicki LA, Orloff MS, Eng C. Lifetime cancer risks in individuals with germline PTEN mutations. Clin Cancer Res 2012;18:400-7. CrossRef

11. Metcalfe K, Lubinski J, Lynch HT, et al. Family history of cancer and cancer risks in women with BRCA1 or BRCA2 mutations. J Natl Cancer Inst 2010;102:1874-8. CrossRef

Metastases from gastric adenocarcinoma to scrotal structures are rare, most being intra-testicular. Extra-testicular metastases are even rarer. It is extremely rare for such a metastasis to be the first manifestation of an occult primary tumour. Herein we report on a patient who presented with a solid extra-testicular mass, which was later confirmed to be a scrotal wall metastasis from an occult gastric adenocarcinoma. To our best knowledge, there has been no previous report of a scrotal wall metastasis from gastric adenocarcinoma.

A 66-year-old man, who enjoyed unremarkable past health, presented with a 2-week history of painless ‘right testicular’ swelling in May 2011. Examination yielded a 4-cm hard, irregular, and non-tender right scrotal mass.

An urgent ultrasound revealed an 18 x 13 x 21-mm solid extra-testicular mass with heterogeneous echogenicity in the right scrotum (Figs 1a-c), which was separate from the normal-looking right testis. Equivocal involvement of the right epididymis was noted on the ultrasound at that time. Increased vascularity of the mass lesion was noted on colour Doppler study. A small right hydrocele was also evident.

In view of a possible malignancy, the patient then underwent surgical scrotal exploration, which revealed a markedly thickened scrotal wall. The dartos muscles could not be well delineated and considered a probable sight of invasion by tumour, though the fat plane between the scrotal wall and the tunica was preserved. The testes could be well separated from the thickened scrotal wall.

A full-thickness scrotal wall incisional biopsy revealed diffuse infiltration of the subcutaneous tissue and dartos muscle by aggregates and islands of adenocarcinoma, more heavily on the right side. Evidence of a desmoplastic reaction was noted. The scrotal skin was unremarkable (Figs 2a-b).

Immunological studies showed that the tumour cells were positive for CK7, carcinoembryonic antigen, p53, and CDX2; focally positive for CK20 and negative for prostate-specific antigen (Figs 2c-f). These features favoured a tumour of the upper gastro-intestinal or genital origin. Biopsies of the urethra and urinary bladder from the same operation were negative of malignancy.

Subsequent computed tomography (CT) of the abdomen and pelvis to search for any underlying malignancy showed bilateral scrotal soft tissue thickening (Figs 1d-e), and oesophagogastroduodenoscopy revealed a hard, irregular, and circumferential ulcerative tumour over the lower part of body of stomach (Fig 1f). Biopsy of this ulcerative tumour revealed a poorly differentiated adenocarcinoma.

Palliative chemotherapy with the XELOX regimen (capecitabine plus oxaliplatin) was started. The chemotherapy regimen was subsequently changed to the FOLFOX regimen (folinic acid, fluorouracil, and oxaliplatin) because of progression of the gastric malignancy. He remained otherwise asymptomatic 10 months following the initial presentation, when this report was submitted for publication.

The scrotum is a musculocutaneous sac composed of two compartments, divided by a midline septum. Each compartment contains a testis, epididymis, spermatic cord, and associated fascial coverings.1 The scrotal wall is composed of pigmented skin, subcutaneous tissue, and the closely related dartos fascia and dartos muscle.

Metastases to scrotal structures are rare. Among these, the testis is the most frequently involved site. Less than 3% of testicular malignancies are secondaries; the lung, prostate, and gastro-intestinal tract are the most common primary sites.2 3 Metastases to extra-testicular scrotal structures such as the epididymis, spermatic cord, and scrotal wall are even rarer. Metastases account for 8.1% of malignancies in the epididymis and spermatic cord,3 4 mostly reported as single case reports.3

Metastasis to the scrotal wall involving the subcutaneous tissue and dartos muscle with normal scrotal skin (as in our patient) is extremely rare. A previous review could only identify sporadic cases of scrotal wall metastases; the primaries being from malignant melanoma, anal carcinoma, and lung carcinoma.5 To our best knowledge, no scrotal wall metastasis from gastric adenocarcinoma has ever been reported in the literature. On the other hand, cutaneous metastases involving the scrotal skin alone are relatively more common. In contrast to scrotal wall metastases which present as a scrotal swelling, scrotal skin metastases present as cutaneous polypoid lesions, ulcers, or papules.6 7 8

Several pathways by which primary tumours metastasise to the scrotal structures have been proposed. They include retrograde venous embolism, retrograde lymphatic extension, arterial embolism, direct invasion along the testicular cord, and transperitoneal seeding through a congenital hydrocele.9 Although the exact pathway of the metastasis in our patient is not clear, absence of intra-abdominal, pelvic lymphadenopathy, and testicular cord thickening all favour embolism or transperitoneal seeding as the route.

Clinically, a scrotal wall metastasis usually presents as a painless scrotal swelling with normal overlying skin, with a firm-to-hard mass being evident on physical examination. Ultrasonography usually reveals a solid extra-testicular mass, that is mostly hypoechoic but can have variable echogenicity.2 Previously described sonographic features of a scrotal wall metastasis from a lung primary also entailed increased peripheral vascularity and poor delineation with the epididymis,2 in which the features were also observed in our patient.

Ultrasonography is usually the first imaging modality for evaluating patients presenting with scrotal swelling. Its high spatial resolution provides nearly 100% sensitivity in identifying a scrotal mass and a 98 to 100% sensitivity in differentiating intra-testicular versus extra-testicular lesions.1 The two most important factors to consider during evaluation of a scrotal mass are whether it is intra- or extra-testicular in location, and whether it is cystic or solid in nature.1 This is because more than 95% intra-testicular masses are malignant while most that are extra-testicular are benign.10

Most extra-testicular masses are cystic and almost all are benign.1 Solid extra-testicular masses are uncommon and 97% of them are also benign.1 11 Among these solid extra-testicular masses, lipoma and adenomatoid tumours are most frequently encountered, and account for about 45% and 33% of all extra-testicular masses, respectively.11

Malignancies account for the remaining 3% of extra-testicular masses. It was estimated that 8.1% of malignancies in the epididymis and spermatic cord are due to metastases.3 4 Rhabdomyosarcoma is the most common primary malignant tumour, accounting for 40% of all malignant extra-testicular tumours and are most common in infants and children.11 In adults, examples of primary malignant tumours include leiomyosarcoma, liposarcoma, fibrosarcoma, malignant fibrous histiocytoma, and primary adenocarcinoma of epididymis.5

Secondary extra-testicular malignant tumours usually occur against a background of known advanced malignancy. Common primary sites include prostate, kidney, gastro-intestinal tract, and pancreas.11 Epididymis and spermatic cord are the usual sites of involvement, while scrotal wall metastases are extremely rare. Only sporadic case reports can be identified in the literature for metastases to the scrotal wall (from melanoma, anal carcinoma, and lung carcinoma).2 12

Lesion diameter, volume, and presence of vascularity on Doppler ultrasonography assist differentiation of malignant from benign lesions.13 Multifocal lesions and heterogeneity have also been suggested as supporting metastatic disease.14 15 However, in most instances, considerable overlap in sonographic appearances of many solid extra-testicular masses preclude a specific diagnosis.1

Magnetic resonance imaging (MRI) is useful in characterising certain extra-testicular lesions such as lipoma, haematoma, and fibrous pseudotumour. Enhancement patterns in gadolinium-enhanced MRIs are also useful in differentiating malignant from benign lesions. In such cases, MRI findings may obviate the need for surgery or change the surgical approach.14

In patients without a known history of malignancy such as ours, diagnosis of an extra-testicular metastasis on the initial ultrasound is difficult. The sonographic differential diagnosis includes other extra-testicular benign and malignant tumours. Clinical correlation is essential to enable better differentiation of malignant from benign lesions. The clinical finding of a hard scrotal mass in our patient raised concerns of malignancy, and hence surgical exploration was undertaken. The final diagnosis still depends on biopsy and pathological study. In our patient, histology and immunological studies of the scrotal wall biopsy hinted at the final diagnosis of occult gastric adenocarcinoma. Although positron emission tomography–CT may be useful for seeking an occult primary malignancy, it is not commonly used as a first-line imaging modality in patients presenting with a scrotal wall lesion. However, it could be offered to search for an underlying primary when histology shows adenocarcinoma but immunological study results are pending.

Extra-testicular metastases are rare and have non-specific sonographic features, which always pose difficulties in diagnosis, particularly in patients without a known primary malignancy. We hereby report the first case of a gastric adenocarcinoma presenting as scrotal wall metastasis. This case also demonstrates the importance of radiological, clinical, and pathological correlations in making the final diagnosis.

1. Woodward PJ, Schwab CM, Sesterhenn IA. From the archives of the AFIP: extratesticular scrotal masses: radiologic-pathologic correlation. Radiographics 2003;23:215-40. Crossref

2. Dogra V, Saad W, Rubens DJ. Sonographic appearance of scrotal wall metastases from lung adenocarcinoma. AJR Am J Roentgenol 2002;179:1647-8. Crossref

3. Dutt N, Bates AW, Baithun SI. Secondary neoplasms of the male genital tract with different patterns of involvement in adults and children. Histopathology 2000;37:323-31. Crossref

4. Algaba F, Santaularia JM, Villavicencio H. Metastatic tumor of the epididymis and spermatic cord. Eur Urol 1983;9:56-9.

5. Dogra VS, Gottlieb GH, Oka M, Rubens DJ. Sonography of the scrotum. Radiology 2003;227:18-36. Crossref

6. Aridogan IA, Satar N, Doran E, Tansug MZ. Scrotal skin metastases of renal cell carcinoma: a case report. Acta Chir Belg 2004;104:599-600.

7. McWeeney DM, Martin ST, Ryan RS, Tobbia IN, Donnellan PP, Barry KM. Scrotal metastases from colorectal carcinoma: a case report. Cases J 2009;2:111. Crossref

8. Wang SQ, Mecca PS, Myskowski PL, Slovin SF. Scrotal and penile papules and plaques as the initial manifestation of a cutaneous metastasis of adenocarcinoma of the prostate: case report and review of the literature. J Cutan Pathol 2008;35:681-4. Crossref

9. Hanash KA, Carney JA, Kelalis PP. Metastatic tumors to the testicles: routes of metastasis. J Urol 1969;102:465-8.

10. Moghe PK, Brady AP. Ultrasound of testicular epidermoid cysts. Br J Radiol 1999;72:942-5.

11. Akbar SA, Sayyed TA, Jafri SZ, Hasteh F, Neill JS. Multimodality imaging of paratesticular neoplasms and their rare mimics. Radiographics 2003;23:1461-76. Crossref

12. Ferguson MA, White BA, Johnson DE, Carrington PR, Schaefer RF. Carcinoma en cuirasse of the scrotum: an unusual presentation of lung carcinoma metastatic to the scrotum. J Urol 1998;160(6 Pt 1):2154-5.

13. Alleman WG, Gorman B, King BF, Larson DR, Cheville JC, Nehra A. Benign and malignant epididymal masses evaluated with scrotal sonography: clinical and pathologic review of 85 patients. J Ultrasound Med 2008;27:1195-202.

14. Cassidy FH, Ishioka KM, McMahon CJ, et al. MR imaging of scrotal tumors and pseudotumors. Radiographics 2010;30:665-83. Crossref

15. Souza FF, Di Salvo D. Sonographic features of a metastatic extratesticular gastrointestinal stromal tumor. J Ultrasound Med 2008;27:1639-42.