- Email: [email protected]
Kimura's disease: imaging patterns on computed tomography
Clinical Radiology (2009) 64, 994e999
Kimura’s disease: imaging patterns on computed tomography Anil Gopinathan*, T.Y. Tan Department of Radiology, Changi General Hospital, Singapore Received 11 March 2009; received in revised form 24 June 2009; accepted 7 July 2009
AIM: To define the role of computed tomography (CT) in identifying and classifying the imaging patterns of diagnostic value in Kimura’s disease of the head and neck. METHODS: A retrospective study was undertaken comprising 13 patients with histopathological evidence of Kimura’s disease. The patients’ clinical and pathological records were reviewed against a detailed analysis of their CT images performed from the base of the skull to the arch of the aorta. RESULTS: Both well-defined, nodular masses, as well as ill-defined plaque-like infiltrative masses were seen in the subcutaneous tissue of the head and neck region. All patients had lesions adjacent to the major salivary glands. The parotid gland was affected in 10 of the 13 cases and the submandibular gland was affected in the rest. Contrast enhancement was variable. More than half of the cases had associated lymphadenopathy. Some of them showed atrophy of the skin and subcutaneous fat overlying the subcutaneous masses. Blood eosinophilia was a consistent feature in all the cases. CONCLUSION: The patterns of distribution, morphology, and enhancement of the lesions in Kimura’s disease that can be demonstrated at CT, enables a confident, non-invasive diagnosis of this condition, in an appropriate clinical context. ª 2009 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction Kimura’s disease is a rare chronic inflammatory disorder of unknown origin that occurs mainly in young Asian males.1,2 The first report of this disease was made in Chinese literature in 1937, by Kim and Szeto. It is still named after Kimura who reported in 1948 that the lesion was an unusual granulation combined with hyperplastic change of lymphatic tissue. The head and neck region is typically involved with painless masses affecting the salivary glands, adjacent subcutaneous tissues, regional lymph nodes, and other sites. Clinically, the lesion is often misdiagnosed as salivary gland malignancy.3 Pathologically it is characterized by eosinophilic lymphofolliculoid granuloma with lymphoid hyperplasia, remarkable infiltration * Guarantor and correspondent: Anil Gopinathan, Department of Radiology, Changi General Hospital, 2 Simei Street 3, Singapore 529889, Singapore. Tel.: þ65 96735410; fax: þ65 6850 4859. E-mail address: [email protected] (A. Gopinathan).
of eosinophils and proliferation of capillaries. There is an associated increase in immunoglobulin (IgE) levels and eosinophilia on peripheral blood smear.4e6 This study attempts to classify the different patterns of lesions seen in Kimura’s disease, which can be effectively demonstrated at CT, a readily accessible and first-line diagnostic mode for head and neck disease entities. In the appropriate clinical context, such pattern recognition can greatly facilitate the diagnosis and spare the patient from potentially harmful and unnecessary invasive diagnostic procedures.
Materials and methods The clinical and pathology records of 13 patients with histological evidence of Kimura’s disease, accumulated over a period of 7 years were reviewed retrospectively. Plain and post-intravenous contrast medium CT images from the skull base to the arch of the aorta were available for
0009-9260/$ - see front matter ª 2009 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.crad.2009.07.003
Kimura’s disease
995
10 patients. Three patients had contrast-enhanced images with the same z-axis coverage and unenhanced images obtained up to the angle of the mandible. The age, sex, and distribution of lesions were also analysed from the available data, as well as the imaging characteristics.
Results All the patients were male. The age at presentation ranged from 26e61 years with a mean age of 43 years. All of them had presented with a mass in the head and neck region. No history of any concurrent renal or autoimmune disease was elicited. Each of them had blood eosinophilia at peripheral blood smear examination. IgE levels were not measured in any of the patients. All the patients had soft-tissue masses related to the major salivary glands (10 in relation to parotid, whereas three were in relation to the submandibular glands). An entirely intraparotid lesion was noted only in one patient (Fig. 1), who had a well defined, homogeneously enhancing nodule within the superficial lobe of the left parotid. This could be distinguished from a parotid neoplasm because of the concurrent subcutaneous masses in the neck. A subcutaneous mass along the medial canthus of the eye (distant from a salivary gland) was noted in one patient (Fig. 2). However, this patient had other lesions that were more typical of Kimura’s disease and adjacent to the
Figure 1 Well-defined homogeneously enhancing, left, intraparotid nodular lesion mimicking a parotid neoplasm.
Figure 2 A poorly enhancing subcutaneous mass near the left medial canthus. Also noted is the concurrent right post-auricular, subcutaneous, plaque-like lesion.
parotid, thereby avoiding ambiguity in diagnosis. The lesions were multiple in all the cases. Bilateral involvement was seen in only one of the 13 cases included in this study (Fig. 3). All the subcutaneous masses seen in the present case series could be classified into one of the two following morphological patterns (1) type 1 masses, which are well-defined and nodular in configuration (Fig. 4); and (2) type 2 masses, which are ill-defined with a plaque-like configuration (Figs. 5 and 6). The lesions of the latter variety were more frequently encountered (in 11 of the
Figure 3 Bilateral, ill-defined soft-tissue masses seen in both the parotid regions.
996
A. Gopinathan, T.Y. Tan
Figure 4 Type 1 lesion characterized by the welldefined, intensely enhancing, subcutaneous nodule overlying the left sternomastoid. An enhancing left level 2 lymph node is also seen.
Figure 6 Type 2 lesion with intense heterogeneous enhancement as demonstrated in the subcutaneous mass in the left parotid region.
13 cases), whereas the former were seen in fewer cases (five of the 13 cases). Some patients also had a simultaneous occurrence of both type 1 and type 2 lesions. The enhancement properties were classified by considering enhancement similar to skeletal muscle as poor, greater than skeletal muscle as moderate, and similar to thyroid gland as intense. All the type 1 masses were iso- to hypodense to
skeletal muscle on non enhanced scans with homogenous and intense post contrast enhancement (Figs. 4 and 7); whereas the type 2 masses were heterogeneous in density with variable enhancement characteristics. Only two out of the 11 type 2 lesions showed intense enhancement (Fig. 6), with five out of the 11 showing moderate enhancement and poor enhancement was demonstrated in four. It was observed that the plaques in those patients who had concurrent type 1 lesion were more enhancing than those who had only
Figure 5 Type 2 lesion with poor enhancement as demonstrated by this ill-defined, plaque-like, subcutaneous lesion infiltrating the underlying superficial lobe of parotid gland with atrophy of the overlying skin and subcutaneous fat.
Figure 7 A well-defined, homogeneously enhancing, subcutaneous nodule adjacent to the right submandibular gland.
Kimura’s disease
type 2 lesions. Another consistent observation was the atrophic changes in the overlying skin and subcutaneous fat in seven of the 11 type 2 masses. This was more severe with the poorly enhancing lesions. Lymph nodes were considered abnormal if they had long axis diameter of more than 1.5 cm when located at level -1 or level -2, while for rest of the cervical lymph node stations, 1 cm was considered as the upper limit for normal. Abnormal enhancement (i.e., greater than skeletal muscle) was the only other observed lymph nodal abnormality in these patients.7,8 Eight of the 13 patients showed enlarged ipsilateral lymph nodes, homogeneous in appearance, and round to oval in shape. Their enhancement was comparable to the subcutaneous masses (Fig. 8). All the patients with type 1 lesions showed nodal enhancement greater than the skeletal muscles in the vicinity.
Discussion Kimura’s disease is also known as subcutaneous eosinophilic lymphoid granuloma, eosinophilic lymphoid follicular hyperplasia, and eosinophilic folliculosis of the skin. The histological pattern is similar in all the lesions of Kimura’s disease irrespective of their location. Lymphoid hyperplasia is invariably present. Foci of closely packed lymphocytes are admixed with areas showing diffuse infiltration of lymphocytes, plasma cells,
Figure 8 Well-defined, homogeneously enhancing, type 1, subcutaneous lesions are seen overlying the left sternomastoid along with left level 1B and level 2 lymph nodes with similar enhancement characteristics. This patient had a concurrent type 2 lesion adjacent to the left parotid as depicted by Fig. 6.
997
and mast cells. Prominent germinal centres containing cellular, vascular, and fibrous components is seen in all the involved lymph nodes.2,9,10 The cause of the disease is unknown, but it is presumed to be a self-limiting allergic or autoimmune response triggered by an unknown stimulus.9 This hypothesis is supported by the increased levels of IgE immunoglobulins and raised eosinophil count in the peripheral blood smear of these patients.4,9,11 The disease has a male predominance (87%) with most cases occurring in patients in their third and fourth decades of life.2,3,12 The reported sex ratio is variable between 3:16 to 10:1.4 In the present series, all the patients were men and most of them were in the fifth decade of life (six out of 13). The typical triad of presentation includes the subcutaneous masses with predilection for head and neck region, enlargement of the draining nodes, and involvement of major salivary glands. Apart from occasional pruritus over the site of the subcutaneous mass, there are few other physical findings.13 The lesions can be solitary or multiple, localized or disseminated. The subcutaneous masses are variable in size from 2e11 cm.14 There are many reports of concurrent nephropathy, including nephrotic syndrome,15 and one report described proteinuria in 12e16% of the cases.16
Imaging features There is limited radiological literature available on Kimura’s disease due to its rarity and restricted epidemiological distribution. The lesions in Kimura’s disease show certain patterns of distribution, morphology, and enhancement after the administration of contrast media, which is well demonstrated on CT. Most of the lesions in Kimura’s disease are seen in the vicinity of major salivary glands.1e4,12 They are either located within the gland or present as subcutaneous masses that imperceptibly merge with it. There are few case reports of other sites such as the forearm,17 groin,2 oral cavity,3 and abdomen18 being involved. The lesions were classified into two specific morphological subtypes, as described earlier. Several features of the granulomatous masses in Kimura disease has been described in previous case series,11,14,19e21 but we could not find any attempt at a classification being made before. The type 1 lesions are relatively well-defined, nodular masses with homogeneous enhancement. Similar lesions have been described by Takeishi et al.4 at magnetic resonance imaging (MRI) and by Ahuja et al.19 at ultrasonography. The type 2 lesions
998
are ill-defined, plaque-like in configuration, and heterogeneous in appearance with variable enhancement. These are more frequently encountered both in the present study as well as in other documented case series and case reports. These lesions tend to infiltrate the adjacent salivary gland with atrophy of overlying skin. A similar observation has been made earlier by Takeishi et al.4 and Oguz et al.16 Intense contrast enhancement has been classically described by Som et al.14 However, this is not a consistent feature as poorly and moderately enhancing lesions are frequently seen.4,11,21,22 The relatively well-defined, homogeneously enhancing type 1 lesion and the ill-defined, poorly enhancing type 2 lesion probably represent the two opposite ends of the spectrum in an inflammatory disease continuum. This assumption is supported by a) well defined lesions seen in younger and ill defined ones in the older patients; b) better enhancement of type 2 lesions when they coexist with type 1 lesions in the same patient as compared to the enhancement seen in patients with all their lesions of type 2 morphology; c) poorly enhancing lesions show more prominent atrophy of the overlying subcutaneous fat, presumably reflecting the chronicity. Takahashi et al.21 had attributed the variable contrast enhancement to differing degrees of fibrosis and vascular proliferation. Lim et al.22 found diminishing post contrast enhancement of Kimura’s lesions in the follow-up images of the same patient. They attributed this phenomenon to progressive fibrosis and sclerosis around the postcapillary venules, which eventually get obliterated and thus the lesions fail to enhance. Ahuja et al.19 also states that in chronic cases, fibrosis within the lesion is a common feature and this may explain the lack of contrast enhancement on CT and the paucity of vessels on colour flow imaging. The involvement of lymph nodes has been reported to range from 42e100%2 (61% in the present series). Well-defined, round to oval draining lymph nodes with high vascularity have been demonstrated4,11,19 by using different imaging methods, such as sonography, CT, and MRI. In the present series, there was a more frequent association of nodes with enhancing subcutaneous masses rather than with non-enhancing masses. A reference to this observation could not be found in earlier published literature. The post-contrast enhancement of nodes is comparable with that of the co-existing subcutaneous lesions. Kimura’s disease needs to be differentiated from parotid tumours, lymphoma, tuberculosis and ALHE (angiolymphoid hyperplasia with eosinophilia). The presence of an associated
A. Gopinathan, T.Y. Tan
subcutaneous component differentiates it from primary parotid neoplasms. The parotid tumours are usually encapsulated and are limited to the parotid gland. Lymphoma may have similar presentation but the distribution of lesions and long clinical course as in Kimura’s disease is unusual. Inflammatory conditions, such as tuberculosis, need to be excluded. Tuberculous nodes are generally necrotic with rim enhancement and a tendency for matting unlike Kimura’s disease where they are usually solid and more homogeneous and circumscribed. ALHE is a distinct entity different from Kimura’s disease. The differentiating features include lack of predilection to a any particular race or sex, smaller and more superficial lesions of ALHE with a tendency to bleed when irritated (as against the indolent masses seen in Kimura’s disease) and less frequent association with lymphadenopathy. The inflammatory masses in ALHE are mostly restricted to the dermis. At histopathology the lack of fibrosis and sclerosis is distinct in ALHE, as fibrosis is prominent at all stages in Kimura’s disease.2,22 Correlation with these clinicopathological findings is essential as differentiation on imaging alone may be difficult. In conclusion, the patterns of distribution, morphology, and enhancement of the lesions in Kimura’s disease demonstrated at CT make it an effective tool in the diagnosis of this condition. The salient diagnostic features of Kimura’s disease may be summarized as (1) predominance in young oriental males; (2) granulomatous masses adjacent to or infiltrating the major salivary glands; (3) subcutaneous masses with two distinct morphological patterns: type 1 being well-defined enhancing, homogeneous masses and type 2 being heterogeneous, ill-defined, plaque-like infiltrating masses with variable enhancement. Overlap between these patterns, as well simultaneous occurrence of both the types of lesions, is not unusual; (4) Frequent presence of associated, well-defined, enlarged draining lymph nodes with enhancement similar to the subcutaneous masses; (5) Blood eosinophilia and increased levels of serum IgE.
References 1. Kase Y, Ikeda T, Yamane M, et al. Kimura’s disease: report of 4 cases with a review of 130 reported cases. Otolaryngol Head Neck Surg 1990;63:413e8. 2. Li TJ, Chen XM, Wang SZ, et al. Kimura’s disease: a clinicopathologic study of 54 Chinese patients. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;82:549e55. 3. Iwai H, Nakae H, Ikeda K, et al. Kimura’s disease diagnosis and prognostic factors. Otolaryngol Head Neck Surg 2007; 137:306e11.
Kimura’s disease
4. Takeishi M, Makino Y, Nishioka H, et al. Kimura disease: diagnostic imaging findings and surgical treatment. J Craniofac Surg 2007;18:1062e6. 5. Urabe A, Tsuneyoshi M, Enjoji M. Epithelioid hemangioma versus Kimura’s disease: a comparative clinicopathologic study. Am J Surg Pathol 1987;11:758e66. 6. Hui PK, Chan JKC, Ng CS, et al. Lymphadenopathy of Kimura’s disease. Am J Surg Pathol 1989;13:177e86. 7. Stevens MH, Harnsberger HR, Mancuso AA, et al. Computed tomography of cervical lymph nodes. Staging and management of head and neck cancer. Arch Otolaryngol 1985;111:735e9. 8. Som PM. Lymph nodes of the neck. Radiology 1987;165: 593e600. 9. Shetty AK, Beaty MW, Mc Guirt WF, et al. Kimuras disease: a diagnostic challenge. Pediatrics 2002;110(3):e39. 10. Chun S, Ji HG. Kimura’s disease and angiolymphoid hyperplasia with eosinophilia: clinical and histopathological differences. J Am Acad Dermatol 1992;27:954e8. 11. Hiwatashi A, Hasuo K, Shiina T, et al. Kimura’s disease with bilateral auricular masses. AJNR Am J Neuroradiol 1999;20: 1976e8. 12. Som PM, Brandwein MS. Salivary glands: anatomy and pathology. In: Som PM, Curtin HD, editors. 4th ed., Head and Neck Imaging, vol. 2 St Louis: Mosby; 2003. p. 2121e2.
999
13. Nyrop M. Kimura’s disease: case report and brief review of the literature. J Laryngol Otol 1994;108:1005e7. 14. Som PM, Biller HF. Kimura disease involving parotid gland and cervical nodes: CT and MR findings. J Comput Assist Tomogr 1992;16:320e2. 15. Sud K, Saha T, Das A, et al. Kimura’s disease and minimal change nephritic syndrome. Nephrol Dial Transplant 1996; 11:1349e51. 16. Oguz KK, Ozturk A, Clia A. Magnetic resonance imaging findings in Kimura’s disease. Neuroradiology 2004;46:855e8. 17. Choi J, Lee G, Kong KY, et al. Imaging findings of Kimura’s disease in the soft tissue of the upper extremity. Am J Radiol 2005;184:193e9. 18. Jeong YY, Song SK, Heo SH, et al. Imaging of Kimura’s disease involving the abdomen. AJR 2006;187:W131e8. 0361e803X/ 06/1871eW131. doi: 10.2214/AJR.05.0840. 19. Ahuja AT, Loke TKL, Mok CO, et al. Ultrasound of Kimura’s disease. Clin Radiol 1995;50:170e3. 20. Smith JRG, Hadgis C, Van Hasselt A, et al. CT of Kimura disease. AJNR Am J Neuroradiol 1989;10(Suppl. 5):S34e6. 21. Takahashi S, Ueda J, Furukawa T, et al. Kimura disease: CT and MR findings. AJNR Am J Neuroradiol 1996;17:382e5. 22. Lim WEH, Tan NG, Tan KP. Radiological features in a patient with Kimura’s disease. Singapore Med J 1997;38:125e8.
Kimura’s disease: imaging patterns on computed tomography Anil Gopinathan*, T.Y. Tan Department of Radiology, Changi General Hospital, Singapore Received 11 March 2009; received in revised form 24 June 2009; accepted 7 July 2009
AIM: To define the role of computed tomography (CT) in identifying and classifying the imaging patterns of diagnostic value in Kimura’s disease of the head and neck. METHODS: A retrospective study was undertaken comprising 13 patients with histopathological evidence of Kimura’s disease. The patients’ clinical and pathological records were reviewed against a detailed analysis of their CT images performed from the base of the skull to the arch of the aorta. RESULTS: Both well-defined, nodular masses, as well as ill-defined plaque-like infiltrative masses were seen in the subcutaneous tissue of the head and neck region. All patients had lesions adjacent to the major salivary glands. The parotid gland was affected in 10 of the 13 cases and the submandibular gland was affected in the rest. Contrast enhancement was variable. More than half of the cases had associated lymphadenopathy. Some of them showed atrophy of the skin and subcutaneous fat overlying the subcutaneous masses. Blood eosinophilia was a consistent feature in all the cases. CONCLUSION: The patterns of distribution, morphology, and enhancement of the lesions in Kimura’s disease that can be demonstrated at CT, enables a confident, non-invasive diagnosis of this condition, in an appropriate clinical context. ª 2009 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction Kimura’s disease is a rare chronic inflammatory disorder of unknown origin that occurs mainly in young Asian males.1,2 The first report of this disease was made in Chinese literature in 1937, by Kim and Szeto. It is still named after Kimura who reported in 1948 that the lesion was an unusual granulation combined with hyperplastic change of lymphatic tissue. The head and neck region is typically involved with painless masses affecting the salivary glands, adjacent subcutaneous tissues, regional lymph nodes, and other sites. Clinically, the lesion is often misdiagnosed as salivary gland malignancy.3 Pathologically it is characterized by eosinophilic lymphofolliculoid granuloma with lymphoid hyperplasia, remarkable infiltration * Guarantor and correspondent: Anil Gopinathan, Department of Radiology, Changi General Hospital, 2 Simei Street 3, Singapore 529889, Singapore. Tel.: þ65 96735410; fax: þ65 6850 4859. E-mail address: [email protected] (A. Gopinathan).
of eosinophils and proliferation of capillaries. There is an associated increase in immunoglobulin (IgE) levels and eosinophilia on peripheral blood smear.4e6 This study attempts to classify the different patterns of lesions seen in Kimura’s disease, which can be effectively demonstrated at CT, a readily accessible and first-line diagnostic mode for head and neck disease entities. In the appropriate clinical context, such pattern recognition can greatly facilitate the diagnosis and spare the patient from potentially harmful and unnecessary invasive diagnostic procedures.
Materials and methods The clinical and pathology records of 13 patients with histological evidence of Kimura’s disease, accumulated over a period of 7 years were reviewed retrospectively. Plain and post-intravenous contrast medium CT images from the skull base to the arch of the aorta were available for
0009-9260/$ - see front matter ª 2009 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.crad.2009.07.003
Kimura’s disease
995
10 patients. Three patients had contrast-enhanced images with the same z-axis coverage and unenhanced images obtained up to the angle of the mandible. The age, sex, and distribution of lesions were also analysed from the available data, as well as the imaging characteristics.
Results All the patients were male. The age at presentation ranged from 26e61 years with a mean age of 43 years. All of them had presented with a mass in the head and neck region. No history of any concurrent renal or autoimmune disease was elicited. Each of them had blood eosinophilia at peripheral blood smear examination. IgE levels were not measured in any of the patients. All the patients had soft-tissue masses related to the major salivary glands (10 in relation to parotid, whereas three were in relation to the submandibular glands). An entirely intraparotid lesion was noted only in one patient (Fig. 1), who had a well defined, homogeneously enhancing nodule within the superficial lobe of the left parotid. This could be distinguished from a parotid neoplasm because of the concurrent subcutaneous masses in the neck. A subcutaneous mass along the medial canthus of the eye (distant from a salivary gland) was noted in one patient (Fig. 2). However, this patient had other lesions that were more typical of Kimura’s disease and adjacent to the
Figure 1 Well-defined homogeneously enhancing, left, intraparotid nodular lesion mimicking a parotid neoplasm.
Figure 2 A poorly enhancing subcutaneous mass near the left medial canthus. Also noted is the concurrent right post-auricular, subcutaneous, plaque-like lesion.
parotid, thereby avoiding ambiguity in diagnosis. The lesions were multiple in all the cases. Bilateral involvement was seen in only one of the 13 cases included in this study (Fig. 3). All the subcutaneous masses seen in the present case series could be classified into one of the two following morphological patterns (1) type 1 masses, which are well-defined and nodular in configuration (Fig. 4); and (2) type 2 masses, which are ill-defined with a plaque-like configuration (Figs. 5 and 6). The lesions of the latter variety were more frequently encountered (in 11 of the
Figure 3 Bilateral, ill-defined soft-tissue masses seen in both the parotid regions.
996
A. Gopinathan, T.Y. Tan
Figure 4 Type 1 lesion characterized by the welldefined, intensely enhancing, subcutaneous nodule overlying the left sternomastoid. An enhancing left level 2 lymph node is also seen.
Figure 6 Type 2 lesion with intense heterogeneous enhancement as demonstrated in the subcutaneous mass in the left parotid region.
13 cases), whereas the former were seen in fewer cases (five of the 13 cases). Some patients also had a simultaneous occurrence of both type 1 and type 2 lesions. The enhancement properties were classified by considering enhancement similar to skeletal muscle as poor, greater than skeletal muscle as moderate, and similar to thyroid gland as intense. All the type 1 masses were iso- to hypodense to
skeletal muscle on non enhanced scans with homogenous and intense post contrast enhancement (Figs. 4 and 7); whereas the type 2 masses were heterogeneous in density with variable enhancement characteristics. Only two out of the 11 type 2 lesions showed intense enhancement (Fig. 6), with five out of the 11 showing moderate enhancement and poor enhancement was demonstrated in four. It was observed that the plaques in those patients who had concurrent type 1 lesion were more enhancing than those who had only
Figure 5 Type 2 lesion with poor enhancement as demonstrated by this ill-defined, plaque-like, subcutaneous lesion infiltrating the underlying superficial lobe of parotid gland with atrophy of the overlying skin and subcutaneous fat.
Figure 7 A well-defined, homogeneously enhancing, subcutaneous nodule adjacent to the right submandibular gland.
Kimura’s disease
type 2 lesions. Another consistent observation was the atrophic changes in the overlying skin and subcutaneous fat in seven of the 11 type 2 masses. This was more severe with the poorly enhancing lesions. Lymph nodes were considered abnormal if they had long axis diameter of more than 1.5 cm when located at level -1 or level -2, while for rest of the cervical lymph node stations, 1 cm was considered as the upper limit for normal. Abnormal enhancement (i.e., greater than skeletal muscle) was the only other observed lymph nodal abnormality in these patients.7,8 Eight of the 13 patients showed enlarged ipsilateral lymph nodes, homogeneous in appearance, and round to oval in shape. Their enhancement was comparable to the subcutaneous masses (Fig. 8). All the patients with type 1 lesions showed nodal enhancement greater than the skeletal muscles in the vicinity.
Discussion Kimura’s disease is also known as subcutaneous eosinophilic lymphoid granuloma, eosinophilic lymphoid follicular hyperplasia, and eosinophilic folliculosis of the skin. The histological pattern is similar in all the lesions of Kimura’s disease irrespective of their location. Lymphoid hyperplasia is invariably present. Foci of closely packed lymphocytes are admixed with areas showing diffuse infiltration of lymphocytes, plasma cells,
Figure 8 Well-defined, homogeneously enhancing, type 1, subcutaneous lesions are seen overlying the left sternomastoid along with left level 1B and level 2 lymph nodes with similar enhancement characteristics. This patient had a concurrent type 2 lesion adjacent to the left parotid as depicted by Fig. 6.
997
and mast cells. Prominent germinal centres containing cellular, vascular, and fibrous components is seen in all the involved lymph nodes.2,9,10 The cause of the disease is unknown, but it is presumed to be a self-limiting allergic or autoimmune response triggered by an unknown stimulus.9 This hypothesis is supported by the increased levels of IgE immunoglobulins and raised eosinophil count in the peripheral blood smear of these patients.4,9,11 The disease has a male predominance (87%) with most cases occurring in patients in their third and fourth decades of life.2,3,12 The reported sex ratio is variable between 3:16 to 10:1.4 In the present series, all the patients were men and most of them were in the fifth decade of life (six out of 13). The typical triad of presentation includes the subcutaneous masses with predilection for head and neck region, enlargement of the draining nodes, and involvement of major salivary glands. Apart from occasional pruritus over the site of the subcutaneous mass, there are few other physical findings.13 The lesions can be solitary or multiple, localized or disseminated. The subcutaneous masses are variable in size from 2e11 cm.14 There are many reports of concurrent nephropathy, including nephrotic syndrome,15 and one report described proteinuria in 12e16% of the cases.16
Imaging features There is limited radiological literature available on Kimura’s disease due to its rarity and restricted epidemiological distribution. The lesions in Kimura’s disease show certain patterns of distribution, morphology, and enhancement after the administration of contrast media, which is well demonstrated on CT. Most of the lesions in Kimura’s disease are seen in the vicinity of major salivary glands.1e4,12 They are either located within the gland or present as subcutaneous masses that imperceptibly merge with it. There are few case reports of other sites such as the forearm,17 groin,2 oral cavity,3 and abdomen18 being involved. The lesions were classified into two specific morphological subtypes, as described earlier. Several features of the granulomatous masses in Kimura disease has been described in previous case series,11,14,19e21 but we could not find any attempt at a classification being made before. The type 1 lesions are relatively well-defined, nodular masses with homogeneous enhancement. Similar lesions have been described by Takeishi et al.4 at magnetic resonance imaging (MRI) and by Ahuja et al.19 at ultrasonography. The type 2 lesions
998
are ill-defined, plaque-like in configuration, and heterogeneous in appearance with variable enhancement. These are more frequently encountered both in the present study as well as in other documented case series and case reports. These lesions tend to infiltrate the adjacent salivary gland with atrophy of overlying skin. A similar observation has been made earlier by Takeishi et al.4 and Oguz et al.16 Intense contrast enhancement has been classically described by Som et al.14 However, this is not a consistent feature as poorly and moderately enhancing lesions are frequently seen.4,11,21,22 The relatively well-defined, homogeneously enhancing type 1 lesion and the ill-defined, poorly enhancing type 2 lesion probably represent the two opposite ends of the spectrum in an inflammatory disease continuum. This assumption is supported by a) well defined lesions seen in younger and ill defined ones in the older patients; b) better enhancement of type 2 lesions when they coexist with type 1 lesions in the same patient as compared to the enhancement seen in patients with all their lesions of type 2 morphology; c) poorly enhancing lesions show more prominent atrophy of the overlying subcutaneous fat, presumably reflecting the chronicity. Takahashi et al.21 had attributed the variable contrast enhancement to differing degrees of fibrosis and vascular proliferation. Lim et al.22 found diminishing post contrast enhancement of Kimura’s lesions in the follow-up images of the same patient. They attributed this phenomenon to progressive fibrosis and sclerosis around the postcapillary venules, which eventually get obliterated and thus the lesions fail to enhance. Ahuja et al.19 also states that in chronic cases, fibrosis within the lesion is a common feature and this may explain the lack of contrast enhancement on CT and the paucity of vessels on colour flow imaging. The involvement of lymph nodes has been reported to range from 42e100%2 (61% in the present series). Well-defined, round to oval draining lymph nodes with high vascularity have been demonstrated4,11,19 by using different imaging methods, such as sonography, CT, and MRI. In the present series, there was a more frequent association of nodes with enhancing subcutaneous masses rather than with non-enhancing masses. A reference to this observation could not be found in earlier published literature. The post-contrast enhancement of nodes is comparable with that of the co-existing subcutaneous lesions. Kimura’s disease needs to be differentiated from parotid tumours, lymphoma, tuberculosis and ALHE (angiolymphoid hyperplasia with eosinophilia). The presence of an associated
A. Gopinathan, T.Y. Tan
subcutaneous component differentiates it from primary parotid neoplasms. The parotid tumours are usually encapsulated and are limited to the parotid gland. Lymphoma may have similar presentation but the distribution of lesions and long clinical course as in Kimura’s disease is unusual. Inflammatory conditions, such as tuberculosis, need to be excluded. Tuberculous nodes are generally necrotic with rim enhancement and a tendency for matting unlike Kimura’s disease where they are usually solid and more homogeneous and circumscribed. ALHE is a distinct entity different from Kimura’s disease. The differentiating features include lack of predilection to a any particular race or sex, smaller and more superficial lesions of ALHE with a tendency to bleed when irritated (as against the indolent masses seen in Kimura’s disease) and less frequent association with lymphadenopathy. The inflammatory masses in ALHE are mostly restricted to the dermis. At histopathology the lack of fibrosis and sclerosis is distinct in ALHE, as fibrosis is prominent at all stages in Kimura’s disease.2,22 Correlation with these clinicopathological findings is essential as differentiation on imaging alone may be difficult. In conclusion, the patterns of distribution, morphology, and enhancement of the lesions in Kimura’s disease demonstrated at CT make it an effective tool in the diagnosis of this condition. The salient diagnostic features of Kimura’s disease may be summarized as (1) predominance in young oriental males; (2) granulomatous masses adjacent to or infiltrating the major salivary glands; (3) subcutaneous masses with two distinct morphological patterns: type 1 being well-defined enhancing, homogeneous masses and type 2 being heterogeneous, ill-defined, plaque-like infiltrating masses with variable enhancement. Overlap between these patterns, as well simultaneous occurrence of both the types of lesions, is not unusual; (4) Frequent presence of associated, well-defined, enlarged draining lymph nodes with enhancement similar to the subcutaneous masses; (5) Blood eosinophilia and increased levels of serum IgE.
References 1. Kase Y, Ikeda T, Yamane M, et al. Kimura’s disease: report of 4 cases with a review of 130 reported cases. Otolaryngol Head Neck Surg 1990;63:413e8. 2. Li TJ, Chen XM, Wang SZ, et al. Kimura’s disease: a clinicopathologic study of 54 Chinese patients. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;82:549e55. 3. Iwai H, Nakae H, Ikeda K, et al. Kimura’s disease diagnosis and prognostic factors. Otolaryngol Head Neck Surg 2007; 137:306e11.
Kimura’s disease
4. Takeishi M, Makino Y, Nishioka H, et al. Kimura disease: diagnostic imaging findings and surgical treatment. J Craniofac Surg 2007;18:1062e6. 5. Urabe A, Tsuneyoshi M, Enjoji M. Epithelioid hemangioma versus Kimura’s disease: a comparative clinicopathologic study. Am J Surg Pathol 1987;11:758e66. 6. Hui PK, Chan JKC, Ng CS, et al. Lymphadenopathy of Kimura’s disease. Am J Surg Pathol 1989;13:177e86. 7. Stevens MH, Harnsberger HR, Mancuso AA, et al. Computed tomography of cervical lymph nodes. Staging and management of head and neck cancer. Arch Otolaryngol 1985;111:735e9. 8. Som PM. Lymph nodes of the neck. Radiology 1987;165: 593e600. 9. Shetty AK, Beaty MW, Mc Guirt WF, et al. Kimuras disease: a diagnostic challenge. Pediatrics 2002;110(3):e39. 10. Chun S, Ji HG. Kimura’s disease and angiolymphoid hyperplasia with eosinophilia: clinical and histopathological differences. J Am Acad Dermatol 1992;27:954e8. 11. Hiwatashi A, Hasuo K, Shiina T, et al. Kimura’s disease with bilateral auricular masses. AJNR Am J Neuroradiol 1999;20: 1976e8. 12. Som PM, Brandwein MS. Salivary glands: anatomy and pathology. In: Som PM, Curtin HD, editors. 4th ed., Head and Neck Imaging, vol. 2 St Louis: Mosby; 2003. p. 2121e2.
999
13. Nyrop M. Kimura’s disease: case report and brief review of the literature. J Laryngol Otol 1994;108:1005e7. 14. Som PM, Biller HF. Kimura disease involving parotid gland and cervical nodes: CT and MR findings. J Comput Assist Tomogr 1992;16:320e2. 15. Sud K, Saha T, Das A, et al. Kimura’s disease and minimal change nephritic syndrome. Nephrol Dial Transplant 1996; 11:1349e51. 16. Oguz KK, Ozturk A, Clia A. Magnetic resonance imaging findings in Kimura’s disease. Neuroradiology 2004;46:855e8. 17. Choi J, Lee G, Kong KY, et al. Imaging findings of Kimura’s disease in the soft tissue of the upper extremity. Am J Radiol 2005;184:193e9. 18. Jeong YY, Song SK, Heo SH, et al. Imaging of Kimura’s disease involving the abdomen. AJR 2006;187:W131e8. 0361e803X/ 06/1871eW131. doi: 10.2214/AJR.05.0840. 19. Ahuja AT, Loke TKL, Mok CO, et al. Ultrasound of Kimura’s disease. Clin Radiol 1995;50:170e3. 20. Smith JRG, Hadgis C, Van Hasselt A, et al. CT of Kimura disease. AJNR Am J Neuroradiol 1989;10(Suppl. 5):S34e6. 21. Takahashi S, Ueda J, Furukawa T, et al. Kimura disease: CT and MR findings. AJNR Am J Neuroradiol 1996;17:382e5. 22. Lim WEH, Tan NG, Tan KP. Radiological features in a patient with Kimura’s disease. Singapore Med J 1997;38:125e8.