Aneurysmal subarachnoid haemorrhage (SAH) is an uncommon but fatal clinical event. The reported incidence in a recent international population-based epidemiological study ranged between 2 and 16 per 100 000 population,1 and the mortality rate ranged from 8% to 67%.2 Among all SAH cases, less than 1% are considered to originate from the spine.3 We present a case of isolated spinal artery aneurysm as a rare culprit of spinal and intracranial SAH to illustrate the diagnostic challenge, its relevant aetiopathological features, and implications for clinical management.

A 74-year-old Chinese man with a history of hypertension and ischaemic heart disease presented to the Accident and Emergency Department in February 2013 for sudden onset of chest pain radiating to the neck and back. Clinical examination and electrocardiogram on admission showed no evidence of myocardial infarction. Initial working diagnosis of acute aortic dissection was also excluded with urgent computed tomography (CT) aortogram. On retrospective analysis, subtle but definite hyperdensities within the thecal sac at upper thoracic levels were noted, suggestive of acute spinal SAH (Fig 1).

The patient developed gradual progressive headache and confusion requiring intubation on the fourth day of admission. Plain CT of the brain at that time revealed further diffuse acute SAH with predominance over the posterior aspect, as well as a small amount of acute intraventricular haemorrhage. No skull vault fracture was observed. Concurrent CT cerebral arteriogram and venogram did not demonstrate any aneurysms or venous sinus thrombosis. Subsequent digital subtraction angiogram (DSA) of cerebral arteries performed on the next day was unremarkable.

In view of posterior predominance of the intracranial SAH with negative cerebral angiograms, an urgent magnetic resonance imaging (MRI) of the brain and cervical spine (Fig 2) was performed for suspected subtle intracranial pathology and a spinal origin of SAH (observed rarely). A lobulated 8-mm intradural, extramedullary lesion with internal hypointense signal was seen at the T1/2 level abutting the lateral surface of the spinal cord. Trace contrast enhancement was noted within the lesion. Adjacent hyperintense signal and susceptibility artefacts within the dural sac were also observed, suggestive of sentinel clot formation. The spinal cord was displaced slightly by the lesion with associated mild cord oedema. Radiological differential diagnosis at this juncture included cavernoma, largely thrombosed spinal aneurysm, other exophytic haemorrhagic intramedullary tumour or slow-flow arteriovenous malformation (AVM).

Subsequent laminectomy and surgical exploration were performed for definitive diagnosis, which revealed a 7-mm saccular aneurysm surrounded by fibrin and an old haematoma from C7 to T2 level. The aneurysm arose from and was incorporated with the radicular artery at T1 level, showing internal partial thrombosis. The aneurysmal sac was trapped and carefully excised without jeopardising the rest of the arterial supply to the cord. The patient had good clinical recovery in postoperative rehabilitation with no further neurological complaints.

Compared with more common causes of spinal SAH like AVM, dural arteriovenous fistula (dAVF) and haemorrhagic spinal cord tumours, spinal artery aneurysm remains a rare entity with a reported incidence of less than 1 in 3000 spinal angiograms, according to a large-scale review by Pia and Djindjian.4 Ever since the first case report on spinal aneurysm back in 1930,5 experience from various studies remains limited mainly to isolated case reports and case series with small sample sizes.

Spinal aneurysms are distinct from intracranial counterparts in several ways. First, they occur mostly along the course of their parent arteries which have a small calibre and are less affected by atherosclerosis.6 On the contrary, intracranial aneurysms are well known for their predilection for branching points of large-sized arteries which are more haemodynamically challenged. Second, spinal aneurysms tend to be small in size, whereas giant aneurysms are exclusively found intracranially, especially in patients with connective tissue disease. Third, many spinal aneurysms are found to be dissecting aneurysms in histology, which explains their fusiform shape and lack of a surgical neck.5 This hinders direct surgical clipping during treatment planning.

For aetiology, isolated spinal aneurysms, as in our case, are rare lesions. More commonly, they are associated with concomitant vascular lesions or occlusions which recruit the spinal artery as a collateral route for supplying and increasing local blood flow. This, in turn, imposes haemodynamic stress and induces aneurysm formation along the parent arteries. Reported associations include spinal cord AVM,7 dAVF,8 aortic coarctation,5 Moyamoya disease,9 and bilateral vertebral artery occlusion.10 Another group of spinal aneurysms is related to underlying vasculopathies. Common examples include collagen vascular disease such as rheumatoid arthritis,11 mycosis,12 and syphilis.5

The diagnosis of spinal artery aneurysm could be challenging and delayed due to its rarity. Most patients present with headache and backache related to aneurysmal rupture and SAH. Neurological deficits including paraparesis, quadriplegia, and cord compression have also been reported.13 As for investigations, MRI and DSA of the spine are sufficient to demonstrate the location and vascular anatomy of a spinal artery aneurysm on most occasions which, in turn, are valuable for treatment planning by neurosurgeons. The position of the aneurysm with respect to the spinal cord determines whether the anterior approach for transthoracic vertebrectomy or posterior laminectomy is most appropriate. In a large-scale literature review by Geibprasert et al,14 32 spinal artery aneurysms were studied, in which the majority (62.5%) arose from the anterior spinal artery. Origin from the radicular artery, as in our case, was scarcely seen (n=2). The authors also observed that posterior spinal artery aneurysms were predominantly isolated dissecting aneurysms. On the contrary, those from the anterior spinal artery are more diverse in aetiology, eg, related to candidiasis and connective tissue disease, which implies a non-surgical management approach with medical treatment for the underlying disorder. Even for isolated dissecting aneurysms from the anterior axis, surgical ligation or endovascular embolisation is still limited by increased risk of postoperative complications including spinal cord infarction due to possible compromise of the dominant arterial supply to the cord. Options of surgical approach include resection and wrapping of the aneurysmal sac, with possible microvascular reconstruction in individual cases. Rare cases of spontaneous complete healing of the dissecting aneurysm have been reported,12 but prompt treatment should not be delayed due to the associated severe complications.

Spinal artery aneurysms are rare culprits of SAH, with different morphological features of their intracranial counterparts. Associations with concurrent vascular lesions and vasculopathy are frequent, with a minority of cases being isolated in aetiology. Magnetic resonance imaging and DSA of the spine show their merit in delineating the location and vascular anatomy of a spinal artery aneurysm, which are key determinants in management planning. Surgical treatment should be prompt and performed cautiously in view of possible substantial neurological deficit when the arterial supply of the cord is jeopardised.

1. Feigin VL, Lawes CM, Bennett DA, Barker-Collo SL, Parag V. Worldwide stroke incidence and early case fatality reported in 56 population-based studies: a systematic review. Lancet Neurol 2009;8:355-69. Crossref

2. Nieuwkamp DJ, Setz LE, Algra A, Linn FH, de Rooij NK, Rinkel GJ. Changes in case fatality of aneurysmal subarachnoid haemorrhage over time, according to age, sex, and region: a meta-analysis. Lancet Neurol 2009;8:635-42. Crossref

3. van Gijn J, Kerr RS, Rinkel GJ. Subarachnoid haemorrhage. Lancet 2007;369:306-18. Crossref

4. Pia HW, Djindjian R. Spinal angiomas: advances in diagnosis and therapy. Berlin: Springer Verlag; 1987.

5. Rengachary SS, Duke DA, Tsai FY, Kragel PJ. Spinal arterial aneurysm: case report. Neurosurgery 1993;33:125-9; discussion 129-30. Crossref

6. Leech PJ, Stokes BA, ApSimon T, Harper C. Unruptured aneurysm of the anterior spinal artery presenting as paraparesis. Case report. J Neurosurg 1976;45:331-3. Crossref

7. Konan AV, Raymond J, Roy D. Transarterial embolization of aneurysms associated with spinal cord arteriovenous malformations. Report of four cases. J Neurosurg 1999;90(1 Suppl):148-54.

8. Malek AM, Halbach VV, Phatouros CC, et al. Spinal dural arteriovenous fistula with an associated feeding artery aneurysm: case report. Neurosurgery 1999;44:877-80. Crossref

9. Walz DM, Woldenberg RF, Setton A. Pseudoaneurysm of the anterior spinal artery in a patient with Moyamoya: an unusual cause of subarachnoid hemorrhage. AJNR Am J Neuroradiol 2006;27:1576-8.

10. Kawamura S, Yoshida T, Nonoyama Y, Yamada M, Suzuki A, Yasui N. Ruptured anterior spinal artery aneurysm: a case report. Surg Neurol 1999;51:608-12. Crossref

11. Toyota S, Wakayama A, Fujimoto Y, Sugiura S, Yoshimine T. Dissecting aneurysm of the radiculomedullary artery originating from extracranial vertebral artery dissection in a patient with rheumatoid cervical spine disease: an unusual cause of subarachnoid hemorrhage. Case report. J Neurosurg Spine 2007;7:660-3. Crossref

12. Berlis A, Scheufler KM, Schmahl C, Rauer S, Götz F, Schumacher M. Solitary spinal artery aneurysms as a rare source of spinal subarachnoid hemorrhage: potential etiology and treatment strategy. ANJR Am J Neuroradiol 2005;26:405-10.

13. Yahiro T, Hirakawa K, Iwaasa M, Tsugu H, Fukushima T, Utsunomiya H. Pseudoaneurysm of the thoracic radiculomedullary artery with subarachnoid hemorrhage. Case report. J Neurosurg 2004;100(3 Suppl Spine):312-5.

14. Geibprasert S, Krings T, Apitzsch J, Reinges MH, Nolte KW, Hans FJ. Subarachnoid hemorrhage following posterior spinal artery aneurysm. A case report and review of the literature. Interv Neuroradiol 2010;16:183-90.

We report the case of a 5-year-old boy, who presented with recurrent episodes of unexplained iron-deficiency anaemia in February 2010 since the age of 27 months. Serial chest X-rays (CXRs) showed bilateral reticulonodular haziness. Bronchoalveolar lavage and lung biopsy confirmed the diagnosis of idiopathic pulmonary hemosiderosis (IPH).

The patient first presented at 27 months of age in Mainland China with malaise, loss of appetite, and shortness of breath for 10 days. He did not have fever, cough, or haemoptysis. He received two doses of H1N1 vaccination before pallor was noted. There was no history of drug or herb intake.

The presenting haemoglobin (Hb) level was 65 g/L (reference range [RR], 115-145 g/L), mean corpuscular volume (MCV) 84.4 fL (RR, 76-90 fL), mean corpuscular haemoglobin (MCH) level 25.8 pg (RR, 25-31 pg), and reticulocyte count 10.4% (reference level [RL], <2%). White cell count and platelet count were unremarkable. Blood smear showed moderate anisopoikilocytosis with polychromasia. Bone marrow aspiration and trephine biopsy revealed active marrow with erythroid preponderance.

Other investigations were performed for anaemia. Serum lactate dehydrogenase (LDH) level was elevated to 980 U/L (RL, <615 U/L). Haptoglobin level was reduced (<0.09 g/L; RR, 0.3-2.7 g/L). Direct Coombs test was negative. Bilirubin was normal. Urine for bilirubin and urobilinogen was negative. Haemoglobin pattern, serum vitamin B12, and G6PD activity were unremarkable. Serum ferritin level was 137 pmol/L (RR, 45-449 pmol/L). Antinuclear antibody assay was negative. Flow cytometry showed normal expression of CD55 and CD59 which ruled out paroxysmal nocturnal haemoglobinuria. Donath-Landsteiner antibody assay was negative, which excluded the diagnosis of paroxysmal cold haemoglobinuria.

Blood work for viral infection including antibodies to parvovirus, Epstein-Barr virus (EBV), mycoplasma, and viral titres was unremarkable.

Screening for blood loss was negative. Stool for occult blood and urine for Hb were negative. Red blood cell scan showed no evidence of haemorrhage.

Blood transfusion was given to correct the anaemia. Microcytic hypochromic anaemia was noted at 2 months after presentation. Haemoglobin level was 54 g/L. Mean corpuscular volume dropped from 84.4 fL to 68 fL. Mean corpuscular haemoglobin level was 21.3 pg. Iron profile showed a low serum iron level of 3 µmol/L (RR, 5-20 µmol/L), elevated total iron-binding capacity (TIBC) of 70 µmol/L (RR, 37-68 µmol/L), and iron saturation of 4% (RR, 20-55%). However, the serum ferritin level was normal at 220 pmol/L (RR, 45-449 pmol/L). Iron supplement was started as a therapeutic trial for suspected iron-deficiency anaemia.

His Hb level remained stable in the next 2 years. However, in the subsequent 8 months, there were three intermittent episodes of anaemia (lowest Hb level, 60 g/L). The anaemia occurred with fever and cough which were considered symptoms of pneumonia and upper respiratory tract infection. There was no history of haemoptysis.

At 36 months after the presentation, he was admitted again for pallor, fatigue, and fever. This time, he needed oxygen therapy. Chest was clear with mild subcostal insucking. Hepatomegaly of 4 cm below the costal margin was noted. Chest X-ray showed bilateral reticulonodular haziness (Fig 1). Haemoglobin level was 61 g/L. Both MCV (78 fL) and MCH level (25 pg) were in the lower normal limits. The serum ferritin level was 879 pmol/L. A diagnosis of pneumonia was made; he was started on oral co-amoxiclav (amoxicillin and clavulanic acid) and azithromycin and given blood transfusion. Fever subsided and oxygen was weaned off. Ultrasonography of abdomen performed 1 month later showed no hepatomegaly; mild hepatic coarsening was suggestive of parenchymal disease. Liver function tests were normal all along. However, the patient tested positive for EBV immunoglobulin (Ig) M antibodies. Thus, he was diagnosed to have pneumonia with EBV infection.

Review of old CXRs showed similar reticulonodular shadows. In view of his history of recurrent iron-deficiency anaemia and CXR findings, pulmonary hemosiderosis was suspected. Flexible bronchoscopy was performed; bronchoscopic lavage over left lingular and right middle lobe showed blood-stained fluid. Bronchoalveolar lavage yielded abundant hemosiderin-laden macrophages (HLM index, 92%). High-resolution computed tomography of thorax showed extensive ground glass opacities and reticular shadows, suggestive of interstitial lung disease. Diffuse visceral pleural brownish deposits were noted over the entire left lung on video-assisted thoracoscopy. Lung biopsy was performed to rule out systemic disorders with pulmonary capillaritis, which could cause diffuse alveolar haemorrhage (DAH). These included Goodpasture’s syndrome, IgA nephropathy, Wegener’s granulomatosis, systemic lupus erythematosus, and antiphospholipid syndrome. Biopsy of lung tissue showed numerous HLM. There was no evidence of capillaritis or vasculitis. Absence of fibrosis and negative exposure history also excluded hypersensitivity pneumonitis. Immunostaining for IgG, IgM, and IgA was negative. The overall picture was compatible with pulmonary hemosiderosis.

Further blood investigations were performed to exclude systemic causes of pulmonary haemorrhage as stated above. Antiglomerular basement membrane antibodies, rheumatoid factor, anti-neutrophil cytoplasmic antibodies, antinuclear antibodies, anti-extractable nuclear antibodies, and anti-cardiolipin IgG antibodies were not detected. Furthermore, the patient was negative for anti-transglutaminase antibody for coeliac disease. He tested weakly positive for IgE antibodies against cow’s milk. Immunoglobulin pattern was unremarkable apart from mildly raised IgA antibody level at 2.28 g/L (range, 0.5-1.92 g/L). Renal function and urinalysis were normal.

A diagnosis of IPH was made based on the above findings. Oral prednisolone was started after the diagnosis for disease control. Chest X-ray performed after 3 months of prednisolone showed improvement in reticulonodular densities (Fig 2). We plan to monitor the disease with clinical symptoms, Hb levels, LDH levels, CXRs, and spirometry.

Idiopathic pulmonary hemosiderosis is a rare disease in children with an unknown aetiology. The estimated yearly incidence among Swedish children from 1960 through 1979 was 0.24 per 1 000 000 children.1 A retrospective review of records from a tertiary paediatric hospital in northern Taiwan noted five cases over 25 years.2 Patients classically presented with a triad of recurrent or chronic pulmonary symptoms (cough, dyspnoea, wheeze, haemoptysis), pulmonary infiltrates on CXR, and iron-deficiency anaemia. Our patient had only anaemia without obvious underlying causes. Subsequent CXR changes led to the suspicion of IPH.

Serum ferritin has been traditionally taken as a reliable surrogate marker of body iron stores. Hypoferritinaemia is commonly used as a diagnostic marker for iron deficiency.3 However, as it is an acute-phase reactant, abnormally raised serum ferritin level may be seen during acute infection or liver disease even in the presence of iron deficiency.4 In IPH, iron study usually shows low serum iron with low iron saturation, and microcytosis and hypochromia in the blood picture. However, plasma ferritin level can be normal or elevated in IPH because of alveolar synthesis and release into the circulation and does not reflect the iron deposits in the body.5 This makes the diagnosis of iron-deficiency anaemia in IPH difficult. We recommend the use of serum iron and transferrin saturation (serum iron/TIBC) instead to evaluate suspected iron-deficiency anaemia.4

Diagnosis of IPH is based on exclusion of other causes of intrapulmonary haemorrhage and systemic diseases. In the absence of systemic disease, findings of HLM in bronchoscopic lavage or gastric aspirate/sputum along with chronic pulmonary symptoms lead to a diagnosis of IPH. Lung biopsy is the gold standard for diagnosis. We performed lung biopsy to exclude pulmonary capillaritis, which is one of the causes of DAH. Pulmonary capillaritis is a small-vessel vasculitis, which can occur as an isolated condition or in association with multiple systemic vasculitides. Isolated DAH without identifiable causation or associated disease is referred to as IPH.6

Daily oral corticosteroids or weekly intravenous pulse methylprednisolone is commonly used in the induction treatment of IPH. Other immunosuppressive agents such as azathioprine, cyclophosphamide, and hydroxychloroquine have also been used alone or in combination with oral corticosteroids.7 8 9 10 11 Low-dose oral corticosteroids, azathioprine, or methotrexate are used in maintenance phase. As there is lack of large patient series and inadequate follow-up in previous studies, the prognosis of IPH remains unclear. However, aggressive treatment with the use of corticosteroids and immunosuppressive agents are associated with a prolonged survival and improved prognosis.12 Long-term low-dose corticosteroid therapy was also reported to result in a milder disease course and prevent bleeding crisis.13

In conclusion, iron-deficiency anaemia results from poor dietary intake of iron in infants and toddlers. However, every child older than 24 months presenting with iron-deficiency anaemia should be evaluated for chronic blood loss. In this report, we have illustrated that anaemia without any respiratory symptoms can be the sole presenting feature of IPH, preceding other signs and symptoms, especially in young children. Haemoptysis may not be present in young children with IPH, as they tend to swallow their sputum. We recommend that when children present with unexplained anaemia and bilateral lung infiltrations, pulmonary haemorrhage should be suspected.

1. Kjellman B, Elinder G, Garwicz S, Svan H. Idiopathic pulmonary haemosiderosis in Swedish children. Acta Paediatr Scand 1984;73:584-8. Crossref

2. Yao TC, Hung IJ, Wong KS, Huang JL, Niu CK. Idiopathic pulmonary haemosiderosis: an Oriental experience. J Paediatr Child Health 2003;39:27-30. Crossref

3. Rybo E. Diagnosis of iron deficiency. Scand J Haematol Suppl 1985;43:5-39.

4. Li CH, Lee AC, Mak TW, Szeto SC. Transferrin saturation for the diagnosis of iron deficiency in febrile anaemic children. Hong Kong Pract 2003;25:363-6.

5. Ioachimescu OC, Sieber S, Kotch A. Idiopathic pulmonary haemosiderosis revisited. Eur Respir J 2004;24:162-70. Crossref

6. Fullmer JJ, Langston C, Dishop MK, Fan LL. Pulmonary capillaritis in children: a review of eight cases with comparison to other alveolar hemorrhage syndromes. J Pediatr 2005;146:376-81. Crossref

7. Milman N, Pedersen FM. Idiopathic pulmonary haemosiderosis. Epidemiology, pathogenic aspects and diagnosis. Respir Med 1998;92:902-7. Crossref

8. Rossi GA, Balzano E, Battistini E, et al. Long-term prednisolone and azathioprine treatment of a patient with idiopathic pulmonary hemosiderosis. Pediatr Pulmonol 1992;13:176-80. Crossref

9. Colombo JL, Stolz SM. Treatment of life-threatening primary pulmonary hemosiderosis with cyclophosphamide. Chest 1992;102:959-60. Crossref

10. Zaki M, al Saleh Q, al Mutari G. Effectiveness of chloroquine therapy in idiopathic pulmonary hemosiderosis. Pediatr Pulmonol 1995;20:125-6. Crossref

11. Bush A, Sheppard MN, Warner JO. Chloroquine in idiopathic pulmonary haemosiderosis. Arch Dis Child 1992;67:625-7. Crossref

12. Saeed MM, Woo MS, MacLaughlin EF, Margetis MF, Keens TG. Prognosis in pediatric idiopathic pulmonary hemosiderosis. Chest 1999;116:721-5. Crossref

13. Kiper N, Göçmen A, Ozçelik U, Dilber E, Anadol D. Long-term clinical course of patients with idiopathic pulmonary hemosiderosis (1979-1994): prolonged survival with low-dose corticosteroid therapy. Pediatr Pulmonol 1999;27:180-4. Crossref

Synthetic fibre granuloma of the conjunctiva, sometimes known as ‘teddy bear granuloma’, was first described by Weinberg et al in 1984.1 It is a rare granulomatous foreign body reaction of the conjunctiva to synthetic fibres. It occurs most commonly in children, and usually presents as a unilateral, inferior conjunctival mass of the lower eyelid. The lesion is known as ‘teddy bear granuloma’ because some cases were caused by materials used in stuffed toy animals.2

Seventeen cases of conjunctival synthetic fibre ‘teddy bear granuloma’ have been reported in the literature. To the best of our knowledge, this is the first reported case of this condition in Hong Kong.

A 7-year-old girl with good health presented to the ophthalmology clinic of United Christian Hospital, Hong Kong, in December 2012 with a left lower eyelid conjunctival mass for 1 month. There was no history of trauma. It was an incidental finding by the girl’s mother and the girl did not complain of any pain or discomfort. There was no change in visual acuity.

Examination showed a 3.5 mm x 1.5 mm conjunctival mass in the inferior fornix of her left lower eyelid (Fig a). It was embedded with a bunch of hair-like material. The lesion prolapsed easily with gentle pressure over the lower eyelid but could not be removed during slit-lamp examination. The rest of her ophthalmological examination was normal. The girl’s mother was very keen on removal of the mass. Excisional biopsy of the mass was performed under general anaesthesia. The mass was excised and sent for histopathological analysis.

Microscopic examination revealed a piece of conjunctival mucosa with stromal granulation tissue showing heavy chronic inflammation, mild activity, and aggregates of foreign body consistent with synthetic fibres, associated with giant cell reaction (Figs b to d). The fibres were refractile and colourless. In another section, scanty hair was seen in the stroma. The picture was compatible with a diagnosis of synthetic fibre granuloma of the conjunctiva.

Postoperatively, the wound healed well and there was no recurrence of the lesion at 1.5 years after excision.

Protective mechanisms of the eye including blinking and tearing normally remove any foreign body that comes into contact with the ocular surface. Occasionally, foreign body may be retained in the eyelid fornix, encapsulated by mucous, embedded in the underlying stroma, and, subsequently, induces a local inflammatory response.2 Synthetic fibre granuloma of the conjunctiva occurs when synthetic fibres are inoculated in the conjunctiva of the eyelid fornix leading to an inflammatory reaction. The lesion is also commonly known as ‘teddy bear granuloma’ because some cases were caused by materials used in stuffed toy animals.2 Various other objects have been suggested as the source of the lesion, including blankets, beddings, and pullover sweaters.2 3 4

The majority of patients were brought in by parents or caretakers who identified a mass in the child’s eyelid. The patients were usually asymptomatic, without a history of trauma. Affected children may rarely present with symptoms of ocular irritation and foreign body sensation.5 Synthetic fibre conjunctival granuloma is usually unilateral, and mainly occurs in the inferior eyelid fornix, except in one reported case where it presented superiorly.1

Differential diagnoses of synthetic fibre conjunctival granuloma include chalazion, pyogenic granuloma, papillary hyperplasia, sarcoidosis, dermoid, or neoplasm including rhabdomyosarcoma.2 6 7 It has been proposed that the most reliable clinical sign to suggest this diagnosis was the presence of a unilateral inferior conjunctival mass in a child or adolescent.2 In addition, the histological features of synthetic fibre conjunctival granuloma are characteristic and diagnostic. Microscopic examination reveals granulomatous inflammatory cell response with lymphocytes, plasma cells and eosinophils, and foreign-body giant cells surrounding the exogenous synthetic fibres.4 8

Treatment of synthetic fibre conjunctival granuloma involves surgical removal of the foreign body and excision of the granuloma.2 Should the granuloma present early and the patient be compliant, it has been suggested to remove the lesion during slit-lamp examination under topical anaesthesia with minimal bleeding and discomfort.9 However, since the granuloma is usually present for a long duration before being noticed, the lesion could be deeply embedded. As a result, excision in the operating theatre under general anaesthesia is often needed, particularly when patients are very young and anxious. Prognosis following surgical excision is excellent.6

Although the entity of synthetic fibre conjunctival granuloma was recognised more than two decades ago, clinicians, including ophthalmologists and pathologists, are unfamiliar with this condition.4 While the number of reports in the literature is limited, accurate reporting may actually reveal a higher incidence of this entity.9 An awareness of this condition will allow early and accurate diagnosis and treatment, which subsequently spare the risks and expense associated with general anaesthesia.3

No conflicts of interests were declared by authors.

1. Weinberg JC, Eagle RC Jr, Font RL, Streeten BW, Hidayat A, Morris DA. Conjunctival synthetic fiber granuloma. A lesion that resembles conjunctivitis nodosa. Ophthalmology 1984;91:867-72. CrossRef

2. Schmack I, Kang SJ, Grossniklaus HE, Lambert SR. Conjunctival granulomas caused by synthetic fibers: report of two cases and review of literature. J AAPOS 2005;9:567-71. CrossRef

3. Enzenauer RW, Speers WC. Teddy bear granuloma of the conjunctiva. J Pediatr Ophthalmol Strabismus 2002;39:46-8.

4. Ferry AP. Synthetic fiber granuloma. ‘Teddy bear’ granuloma of the conjunctiva. Arch Ophthalmol 1994;112:1339-41. CrossRef

5. Farooq MK, Prause JU, Heegaard S. Synthetic fiber from a teddy bear causing keratitis and conjunctival granuloma: case report. BMC Ophthalmol 2011;11:17. CrossRef

6. Shields JA, Augsburger JJ, Stechschulte J, Repka M. Synthetic fiber granuloma of the conjunctiva. Am J Ophthalmol 1985;99:598-600. CrossRef

7. Lueder GT. Synthetic fiber granuloma. Arch Ophthalmol 1995;113:848-9. CrossRef

8. Batta B, Robin A, George JL, Angioi K. “Teddy bear granuloma”, a rare condition: a case report of a 3-year-old child [in French]. J Fr Ophtalmol 2012;35:117-20. CrossRef

9. Resnick SC, Schainker BA, Ortiz JM. Conjunctival synthetic and nonsynthetic fiber granulomas. Cornea 1991;10:59-62. CrossRef

Xanthogranulomatous inflammation (XGI) is a rare but well-defined disease, first reported by Oberling in 1935.1 The disease process was most frequently reported in the kidney and gallbladder. Rare occurrence in the gastro-intestinal tract was illustrated in only one recently reported case in the terminal ileum,2 four reported cases in the colon,3 4 5 6 eight cases in a series of interval appendicectomy specimens,7 and eight cases with gastric involvement.8 9 10 11 12 13 Of the four cases with colonic involvement, two involved the sigmoid colon,3 4 one involved the caecum,5 and one involved the ascending colon.6 Most of these colonic lesions presented with a mass-forming lesion with predominant submucosal involvement, while primary mucosal involvement was only reported in the last case involving the ascending colon. We, herein, report the second case of XGI in the terminal ileum with mucosal ulceration and full-thickness involvement of the bowel wall, presenting as a painful right-lower-quadrant abdominal mass.

A 51-year-old Chinese male presented to the Emergency Department on 2 December 2012. He was a chronic smoker and alcoholic. He complained of right-sided abdominal pain for the past 2 weeks. The pain was not associated with nausea, vomiting, constipation, or diarrhoea. There was no anorexia or weight loss. There was no history of melaena. His medical history included diabetes, hypertension, and gout. There was no history of tuberculosis.

He was admitted to a hospital in Mainland China 10 days before the index admission for the same problem. While in that hospital, he had raised white cell count of 16.2 x 10⁹ /L, and an ultrasound of abdomen revealed a gallstone and a renal stone. There was no hydronephrosis. A course of antibiotics was given, but the symptoms persisted. The patient returned from Mainland China on 2 December 2012, and attended our Emergency Department for further management.

On admission, the physical examination of the respiratory, cardiovascular, and central nervous systems was unremarkable. Abdominal examination revealed tenderness in the right lower quadrant. Per rectal examination revealed no blood, melaena, or mass. Abdominal X-ray showed no specific abnormalities.

His blood tests revealed mildly elevated white cell count of 10.6 x 10⁹ /L (reference range [RR], 3.7-9.2 x 10⁹ /L), haemoglobin level of 137 g/L (RR, 134-171 g/L), normal amylase level of 57 U/L (RR, 30-128 U/L), and normal electrolytes and liver enzymes. An urgent ultrasound of the abdomen and pelvis raised suspicion of acute appendicitis, with a tubular, non-peristaltic, non-compressible structure measuring 1.55 cm in diameter at the appendicular region, with small amount of loculated fluid around the lesion (Fig 1), corresponding with the site of maximum tenderness. The distal end of the lesion was obscured by bowel gas; it could not be ascertained whether it was blind-ended. The same study also revealed the presence of a gallstone and a right lower pole renal stone.

With a preliminary diagnosis of acute appendicitis, diagnostic laparoscopy was performed on 3 December 2012. There was an ileocaecal mass fixed to the posterior abdominal wall, which was difficult to mobilise despite an open approach via gridiron incision. Upon conversion to midline laparotomy, a large inflammatory mass was found at the ileocaecal junction, compatible with an infiltrative tumour. The mass demonstrated through-and-through invasion into the ileal mesentery, involving several loops of the ileum. Ileocaecal lymph node enlargement was noted. The appendix was not identified. There was no gross cavity or pus. With the suspicion of carcinoma of appendix, limited right hemicolectomy with en-bloc resection of the mass together with 65 cm of the terminal ileum, caecum, and proximal ascending colon was performed. Primary sub-end to sub-end, side-to-side anastomosis was fashioned.

The postoperative course of the patient was complicated by on-and-off fever with elevated white cell count of up to 22.5 x 10⁹ /L. Blood culture taken on postoperative day 2 yielded no bacterial growth. Erythema and serous discharge were noted in the paraumbilical region of the laparotomy and gridiron wounds, which were managed with dressing and packing. Wound swab yielded scanty growth of Escherichia coli. The fever responded to a 10-day course of cefuroxime and metronidazole. He was discharged on postoperative day 10.

A colonoscopy done 4 months after the operation in April 2013 revealed no abnormalities.

Gross examination of the ileocolectomy specimen revealed a 0.5-cm ileal mucosal ulcer, which was 1.5 cm proximal to the ileocaecal valve. The periappendicular mass was haemorrhagic and covered with exudate, within which was a retrocaecal appendix, measuring 5 cm in length and 1 cm in diameter, surrounded by necrotic and yellowish tissue. Cut surface of the appendix was unremarkable. A few lymph nodes were found in the ileocaecal fossa.

Microscopic examination of the appendicular mass showed abscess, haemorrhage, and XGI which consisted predominantly of foamy histiocytes, scattered neutrophils, lymphoplasmacytic cells, a few multinucleated giant cells, and congested capillaries with surrounding fibrosis (Fig 2). The foamy histiocytes were positive for CD68 on immunostaining, confirming their histiocytic origin. No Michaelis-Gutmann bodies were detected. The terminal ileum ulcer revealed similar XGI which extended through the bowel wall to involve the mesentery. The appendicular mucosa showed no neutrophilic infiltrate. A few reactive lymph nodes were noted. There was no evidence of malignancy or granuloma.

Xanthogranulomatous inflammation is a form of chronic inflammatory condition characterised macroscopically by mass-forming golden yellow tumours and microscopically by aggregation of lipid-laden foamy histiocytes including multinucleated giant cells, with a minor component of chronic and acute inflammatory cells and fibrous reaction. It was first described by Oberling in 1935 in three cases of retroperitoneal xanthogranulomas.1 Its occurrence in the endometrium, ovary, fallopian tubes, vagina, testis, epididymis, stomach, bone, skin (as fistulation secondary to inflammation primarily involving another internal organ),14 appendix,7 15 urinary bladder, thyroid, and adrenal glands has been reported, with the highest prevalence reported in the kidney and gallbladder. A majority of XGI cases present as a mass-like lesion with an extension of fibrosis and inflammation to the surrounding tissues, leading to diagnostic difficulties in differentiating them from infiltrative malignant tumours.

Pathological differential diagnoses bearing similar histological features include malakoplakia, which is characterised by an inflammatory and destructive xanthomatous proliferation with the presence of Michaelis-Gutmann bodies, which are intracytoplasmic laminated concretions usually positive for periodic acid–Schiff, von Kossa, and Prussian blue stains. Macrophages known as von Hansemann cells are more granular and eosinophilic and have less vacuolated cytoplasm than ordinary histocytes. Other differential diagnoses include localised xanthoma deposits without parenchymal destruction or xanthomas with prominent foam cell features.

Although the pathological features of XGI are well described, its exact pathogenesis is not well established.

Various proposed mechanisms include chronic recurrent infection, obstruction, immunological disorders, and defective lipid transport. It is generally believed that the localised proliferation of lipid-laden foamy histiocytes in XGI represents chronic suppurative inflammation secondary to interaction between the host and micro-organisms. Examples of immunological disorders include disrupted chemotaxis of polymorphs and macrophages, which is a specific immune response toward Proteus and Escherichia infections. A recently reported case2 involving the terminal ileum proposed a possible mechanism of perforation due to an ingested foreign body. However, none of the above hypotheses were able to fully explain the anatomical distribution of the condition, which is most common in the appendix where neither perforation due to ingested foreign bodies nor chronic suppurative inflammation is most often found. In this reported case, infected laparotomy wound swab yielded E coli, while there were no symptoms to suggest pre-existing chronic suppurative inflammation. There was no evidence of a penetrating foreign body on history or gross examination of the pathological specimen.

Rare occurrence of XGI in the lower gastro-intestinal tract is illustrated by only one reported case in the terminal ileum,2 four reported cases in the colon with two involving the sigmoid colon,3 4 one involving the caecum,5 and one involving the ascending colon.6 However, a histopathological review of 22 interval appendicectomy specimens by Guo and Greenson7 in 2003 reported presence of XGI in eight cases (36.4%) of interval appendicectomy versus none in the 44 matched patients receiving acute appendicectomy, suggesting XGI may be underreported as a delayed consequence of acute inflammation and that these histological changes are secondary to the time interval of inflammation rather than intrinsic factors specific to the patient or disease.

Due to endoscopic, radiological, and the intra-operative macroscopic resemblance to infiltrative malignant neoplasms, these lesions warrant excision with a wide margin, similar to treatment of locally advanced malignancies.

However, there has been inadequate evidence to suggest association between XGI and gastro-intestinal malignancies. Of the eight cases of XGI of the stomach reported in the literature,8 9 10 11 12 13 co-existence of XGI with gastric cancer was reported in three cases. Histological examination of these cases did not support continuity between the xanthogranuloma and adenocarcinoma. In the other case of XGI in the terminal ileum reported by Yoon et al,2 preoperative endoscopy and biopsy showed ulcers with acute and chronic inflammation only. However, surgical resection was considered unavoidable in view of the radiological findings highly suggestive of appendiceal cancer. While preoperative endoscopic biopsy may not be helpful to exclude malignancy as most of the lesions are submucosal, intra-operative frozen section may be helpful to avoid unnecessary radical surgery.

We report a case of XGI with full-thickness involvement of the terminal ileum presenting with a tender intraperitoneal mass. This report aimed to emphasise ileal involvement of XGI, although rare, as one of the differential diagnoses of mass lesions in the small bowel mimicking malignant neoplasms.

No conflicts of interest were declared by the authors.

1. Oberling C. Retroperitoneal xanthogranuloma. Am J Cancer 1935;23:477-89. CrossRef

2. Yoon JS, Jeon YC, Kim TY, et al. Xanthogranulomatous inflammation in terminal ileum presenting as an appendiceal mass: case report and review of the literature. Clin Endosc 2013;46:193-6. CrossRef

3. Lo CY, Lorentz TG, Poon CS. Xanthogranulomatous inflammation of the sigmoid colon: a case report. Aust N Z J Surg 1996;66:643-4. CrossRef

4. Oh YH, Seong SS, Jang KS, et al. Xanthogranulomatous inflammation presenting as a submucosal mass of the sigmoid colon. Pathol Int 2005;55:440-4. CrossRef

5. Anadol AZ, Gonul II, Tezel E. Xanthogranulomatous inflammation of the colon: a rare cause of cecal mass with bleeding. South Med J 2009;102:196-9.

6. Dhawan S, Jain D, Kalhan SK. Xanthogranulomatous inflammation of ascending colon with mucosal involvement: report of a first case. J Crohns Colitis 2011;5:245-8. CrossRef

7. Guo G, Greenson JK. Histopathology of interval (delayed) appendectomy specimens: strong association with granulomatous and xanthogranulomatous appendicitis. Am J Surg Pathol 2003;27:1147-51. CrossRef

8. Zhang L, Huang X, Li J. Xanthogranuloma of the stomach: a case report. Eur J Surg Oncol 1992;18:293-5.

9. Guarino M, Reale D, Micoli G, Tricomi P, Cristofori E. Xanthogranulomatous gastritis: association with xanthogranulomatous cholecystitis. J Clin Pathol 1993;46:88-90. CrossRef

10. Lespi PJ. Gastric xanthogranuloma (inflammatory malignant fibrohistiocytoma). Case report and literature review [in Spanish]. Acta Gastroenterol Latinoam 1998;28:309-10.

11. Lai HY, Chen JH, Chen CK, et al. Xanthogranulomatous pseudotumor of stomach induced by perforated peptic ulcer mimicking a stromal tumor. Eur Radiol 2006;16:2371-2. CrossRef

12. Kubosawa H, Yano K, Oda K, et al. Xanthogranulomatous gastritis with pseudosarcomatous changes. Pathol Int 2007;57:291-5. CrossRef

13. Kinoshita H, Yamaguchi S, Sakata Y, Arii K, Mori K, Kodama R. A rare case of xanthogranuloma of the stomach masquerading as an advanced stage tumor. World J Surg Oncol 2011;9:67. CrossRef

14. Rogers S, Slater DN, Anderson JA, Parsons MA. Cutaneous xanthogranulomatous inflammation: a potential indicator of internal disease. Br J Dermatol 1992;126:290-3.

15. Chuang YF, Cheng TI, Soong TC, Tsou MH. Xanthogranulomatous appendicitis. J Formos Med Assoc 2005;104:752-4. CrossRef