Dear Editor

We read with interest the case report from Karim and colleagues in the Journal.[1] This case raises an important issue.

Our study [2] has been quoted as suggesting that we have described HAE with normal C4. We do not entirely accept this interpretation. Although there were HAE patients with normal C4 this was only achieved whilst on adequate treatment. All 20 HAE patients in whom we were able to identify samples either pre-treatment or when off treatment had a low C4 (sensitivity 100%).

This study prompted a review of 44 cases with a presumed diagnosis of HAE and we have shown that misdiagnosis is not uncommon.[3] Of relevance to the case of Karim and colleagues, two patient samples had apparently low immunochemical C1 inhibitor levels but this was shown to be a method- dependent phenomenon and levels were normal when rechecked at a second laboratory. A further confounding factor was that one laboratory was using an inappropriate reference range for C4, based on historical data. Recent UK National Quality Assessment Scheme returns have highlighted the importance of having current local reference ranges, as there are substantial differences between some user groups.

Patients with normal C4 and putative HAE have been described previously. Frank and colleagues [4] describe 4 angioedema cases with normal C4 but it is unclear from the text whether this is at presentation or on treatment. Importantly they note that C4 is never normal during an attack. Sánchez Palacios and colleagues describe an unusual patient with chronic angioedema, low C1 inhibitor and normal C4 but the diagnosis is unclear[5]. On balance, HAE (Type I or Type II) seems unlikely. Our study [3] confirmed that ‘in unequivocal HAE, at presentation or when off treatment, C4 is invariably below 40% of the normal mean level’.

The case of Karim and colleagues, describing angioedema with a normal C4, is interesting. In these circumstances there may be alternative causative mechanisms, other than HAE Type I or Type II, such as activation of the of the kallikrein-bradykinin pathway. Reporting of such cases is essential for further clarification. However, without further information (i.e. genetic and functional studies on the family) we cannot conclude that this represents a case of HAE Type I or Type II.

Serum C4, as with any pathological investigation, must be interpreted in the clinical context of the patient. Serum C4, with a locally validated reference range, is a very reproducible and reliable screening test for HAE Type I and Type II.[2] However, in the face of a strong family history or clinical history it should not be the sole test. This does not invalidate the use of C4 as a screening test, especially in a condition with a low prior probability and demanding immunochemical and functional assays.

References

1. Karim, Y., Griffiths, H., Deacock, S. Normal complement C4 values do not exclude hereditary angioedema. J Clin Pathol 2004;57:213-4.

2. Gompels, M. M., Lock, R. J., Morgan, J. E., Osborne, J., Brown, A., Virgo, P. F. A multi-centre evaluation of the diagnostic efficiency of serological investigations for C1 inhibitor deficiency. J Clin Pathol 2002;55:145-7.

3. Gompels, M. M., Lock, R. J., Unsworth, D. J., Johnston, S. L., Archer, C. B., Davies, S. V. Misdiagnosis of hereditary angioedema (Type 1 and Type 2). Br J Dermatol 2003;148:719-23.

4. Frank, M. M., Gelfand, J. A., Atkinson, J. P. Hereditary angioedema: the clinical syndrome and its management. Ann Intern Med 1976;84:580-93.

5. Sánchez Palacios, A., Schamann Medina, F., García Marrero, J. A. Chronic angioedema. Three relevant cases. Allergol et Immunopathol 1998;26:195-8.

Dear Editor

We appreciated the comments of Dr Zardawi [1] and agree that actin is a ubiquitous cytoskeletal protein of microfilaments and demonstrable in a variety of cells and tumor types. As we described, all leiomyosarcomas are smooth muscle actin (SMA)-positive, and desmin, muscle specific actin (MSA) and h-caldesmon are positive in a great majority of these tumors. However, none of these is absolutely specific for smooth muscle, and positivity for two or more of these markers is more supportive of leiomyosarcoma than positivity for one alone.

Anti-SMA monoclonal antibody, 1A4, detects mainly with an alpha smooth muscle actin isoform. This marker exhibits a more restricted pattern of staining for smooth muscle, but this may be expressed in rhabdomyosarcomas and in cells with a myofibroblastic or myoepithelial phenotype. As another anti-MSA monoclonal antibody, HHF35, recognizes all alpha actins (skeletal, smooth, and cardiac) and gamma smooth muscle actin, its specificity for smooth muscle is therefore quite limited. In this regard, recognition of the fact that non-muscle lesions exhibiting so-called myoid differentiation are also SMA-positive will prevent overdiagnosis as leiomyosarcoma.

Reference

(1) Zardawi IM. Re: Expression of smooth muscle markers in so called malignant fibrous histiocytoma [electronic response to Hasegawa et al Expression of smooth muscle markers in so called malignant fibrous histiocytomas] jclinpath.com 2003http://jcp.bmjjournals.com/cgi/eletters/56/9/666#45

Dear Editor

I read with interest the article on the expression of smooth muscle markers in so called malignant fibrous histiocytomas by Hasegawa et al. in the Journal.[1]

While I agree with Hasegawa and colleagues that pleomorphic malignant fibrous histiocytoma can be regarded as an undifferentiated sarcoma, I feel Hasegawa et al are placing too much reliance on so called smooth muscle markers. Most of these markers are ubiquitous, present in many cells. Many of the so called muscle specific markers stain microfilaments which form the cytoskeletal framework of most cells. However, these markers are particularly expressed in muscle cells, myoepithelial cells cell and myofibroblasts. In neoplasia, these markers indicate myogenic differentiation which can be seen in a raft of tumours which make them unreliable in tumour classification.

A search of the Internet (www.ipox.org) shows smooth muscle actin (SMA) immunoreactivity in 150 types of tumour from different parts of the body with 100% staining in 32 tumour types, between 50% and 99% staining in 45 tumour types, and between 4% and 49% staining in 73 tumour types. The same source cites 0% staining in 80 tumour types. The question that needs to be asked therefore is not which tumours are SMA positive but rather which tumours do not express this marker.

SMA is very much like neuron specific enolase (NSE) which can be found in neurons, neuroendocrine cells, striated and smooth muscle cells, megakaryocytes, T cells and platelets to name a few. NSE reactivity has been reported in over 140 tumour types (www.ipox.org).

Reference

(1) Hasegawa T, Hasegawa F, Hirose T, Sano T, Mastuno Y. Expression of smooth muscle markers in so called malignant fibrous histiocytoma. J Clin Pathol 2003; 56:666-671.

Dear Editor

We read the article by Horny et al. describing bone marrow mast cell (MC) specific protease expression patterns in cases of systemic mastocytosis and myelodysplastic syndromes (MDS) with great interest.[1] Increase in bone marrow MC is a known feature of various hematological diseases including myeloproliferative disorders and acquired severe aplastic anemia (SAA). Although the MC increase is clonal in mastocytosis and benign in acquired SAA, its nature has not been fully understood in myeloproliferative and myelodysplastic disorders.

Acquired SAA and hypoplastic MDS share several clinical and bone marrow features and often difficult to distinguish. Both conditions respond to immune suppressive therapies. Is the increased MC in these conditions simply, an innocent consequence of hemopoietic cell injury sparing MC or alternatively, does it contribute to the development of severe bone marrow hypoplasia/aplasia in return? Mast cells have long life spans and it is likely that they were not directly affected by the attack against stem cell compartment resulting in relative MC increase in the bone marrow. Low to normal stem cell factor (SCF) levels has been shown in SAA unlike the increased levels of other hemopoietic growth factors.[2] This may be explained by greater dependency of MC survival and growth on SCF than other growth factors and by a negative feedback control mechanism in a population that is already supplied by an autocrine pathway. In support of this explanation, a reaction mimicking systemic mastocytosis was observed in an aplastic anemia patient treated with SCF accompanied by a partial and transient hemopoietic recovery.[3]

Mast cells with various enzyme expression patterns may mediate different functions in certain tissues that they exist in. These patterns may also be related to the maturational stage of MC. Nevertheless, the predominant MC type in certain tissues may be determined by the environmental needs. We think that coexistence of chymase-expressing MC (MCC) and chymase and tryptase-expressing MC (MCCT) in physiological conditions reflects naturally occurring balance contributing to tissue homeostasis. It is known that MC can also act as antigen presenters as well as effector elements of human immune system. Mast cells can kill target cells through secretion of cytokines such as TNF-a and serine proteases and potentially through direct cell-to-cell interaction. Granzyme H, one of the MC serine proteases, has chymase activity [4] and chymase is known to induce apoptosis in target cells.[5] It was also demonstrated that mast cell-derived cell line P815 contains granzyme B RNA.[6] On the other hand, tryptase, another MC protease, is a well-known mitogen that could induce growth of certain cells such as airway smooth muscle cells, fibroblasts, neuronal cells.[7] Tryptase-expressing MC (MCT) are often found in tissue repair sites characterized by fibrosis.

The predominance of MCT in systemic mastocytosis and MDS patients was consistent with the typical presence of hypercellular marrow in these conditions.[1] Although authors did not provide the number of cases with hypoplastic MDS in their series, the frequency has been 5-10% in adult literature suggesting that the majority, if not all of their cases had normocellular or hyperplastic MDS. The autocrine production of SCF with increased tryptase activity might have contributed to the extremely hypercellular bone marrow in those cases. The authors also described hypocellular bone marrow associated with a focal increase in MC with strong chymase expression in a case of indolent systemic mastocytosis that suggests a possible MCC contribution to hypocellularity. We recently showed an association between MC persistence and poor outcome in childhood SAA following immune suppression.[8] In another study, we demonstrated long- term liquid culture-grown human bone marrow MC cytotoxicity against human leukemia cells.[9] It is possible that those MC had strong chymase expression. Regardless of the mechanisms involved, MC, preferentially MCC increase may contribute to hypocellularity in acquired SAA and hypoplastic MDS. This explanation is also consistent with the lack of fibrosis in acquired SAA and hypoplastic MDS that could be secondary to specific MCC increase.

References

(1) Horny H-P, Greschniok A, Jordan J-H, Menke DM, Valent P : Chymase expressing bone marrow mast cells in mastocytosis and myelodysplastic syndromes: an immunohistochemical and morphometric study. J Clin Pathol 2003;56:103-106.

(2) Kojima S, Matsuyama T, Kodera Y. Plasma levels and production of soluble stem cell factor by marrow stromal cells in patients with aplastic anaemia. Br J Haematol 1997;99:440-6.

(3) Jordan JH, Schernthaner GH, Fritsche-Polanz R, et al. Stem cell factor- induced bone marrow mast cell hyperplasia mimicking systemic mastocytosis (SM): histopathologic and morphologic evaluation with special reference to recently established SM-criteria. Leuk Lymphoma 2002;43:575-82.

(4) Edwards KM, Kam CM, Powers JC, Trapani JA. The human cytotoxic T cell granule serine protease granzyme H has chymotrypsin-like (chymase) activity and is taken up into cytoplasmic vesicles reminiscent of granzyme B-containing endosomes. J Biol Chem 1999;274:30468-73.

(5) Leskinen M, Wang Y, Leszczynski D, Lindstedt KA, Kovanen PT. Mast cell chymase induces apoptosis of vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 2001;21:516-22.

(6) Garcia-Sanz JA, MacDonald HR, Jenne DE et al. Cell specificity of granzyme gene expression. J Immunol 1990;145:3111-8.

(7) Ruoss SJ, Hartmann T, Caughey GH. Mast cell tryptase is a mitogen for cultured fibroblasts. J Clin Invest 1991;88:493-9.

(8) Chien M, Abella E, Rabah R, Ravindranath Y, Savasan S. Mast cell persistence is associated with poor outcome in childhood severe aplastic anemia following immune suppression. Presented in the 17th Annual Meeting of ASPH/O, Seattle, Washington, May 3- May 6, 2003. Pediatr Res 2003;53:292A, Abstract #1663.

(9) Özdemir Ö, Ravindranath Y, Savasan S. Evaluation of long-term liquid culture grown human bone marrow mast cell cytotoxicity against human leukemia cells. (44th annual meeting of the American Society of Hematology, Philadelphia, Pennsylvania, December 6-10, 2002.) Blood 2002; 100:45b, Abstract #3642.