Dear Sir, I thank Dr. Slevin and coauthors for writing an excellent account of the clinical and biologic relevance of hyaluronic acid (HA) and its fragments. In 1985, the seminal work from Dr. Kumar's group offered a glimpse of the role of HA fragments in angiogenesis. 1 In later years, the same group helped to elucidate some of the signal transduction mechanisms by which HA fragments stimulate bovine endothelial cell proliferation, migration and adhesion. 2 Consistent with the findings of Dr. Kumar's group, we previously demonstrated that HA fragments of 10 -15 disaccharide units induce proliferation and adhesion of human primary endothelial cells. 3 Furthermore, primary cultures of human pulmonary artery (HPAEC), lung microvessel (HMVEC-L) and umbilical vein-derived (HUVEC) endothelial cells show differential responsiveness to angiogenic HA fragments. While angiogenic fragments stimulate DNA synthesis in HMVEC-L and HPAEC, HUVEC do not respond to angiogenic HA fragments. 3 However, in contrast to the suggestion by Dr. Kumar's group that the biologic effects of angiogenic HA fragments and HA, observed in bovine endothelial cells, are mediated through cell surface CD44 receptors, 4 we found that RHAMM is the functional cell surface HA receptor expressed in HPAEC and HMVEC-L. RHAMMangiogenic HA fragment interaction induces tyrosine phosphorylation of p125FAK, paxillin and p42/44 ERK. 3 It is likely that the angiogenic HA fragments stimulate proliferation of primary human endothelial cells through the MAP kinase pathway. 3 Thus, it is possible that, in different cell types, angiogenic HA fragments mediate their biologic effects through different HA receptors.Regarding the findings of Dr. Kumar's group that, although HA levels are elevated in the sera of both Wilms' tumor patients and children with bone-metastasizing renal tumor of childhood (BMRTC), HA fragments are found only in the sera of BMRTC patients, 5 our findings in cancers of the bladder, prostate and head/neck are in agreement. 6 -8 We found that HA levels are elevated in the urine of bladder cancer patients regardless of tumor grade [i.e., low (G1), intermediate (G2) and high (G3)]. However, angiogenic HA fragments are found only in the urine of patients with G2 or G3 bladder tumors. 6 This finding could be explained by our observation that hyaluronidase levels are elevated only in G2 and G3 patients, which in turn could degrade tumor-associated HA into angiogenic fragments. We have also shown that in bladder tumor tissues HA is produced by both tumor stroma and tumor cells, whereas hyaluronidase (i.e., HYAL1 gene product) is produced only by tumor cells. 9,10 Also, we found that angiogenic HA fragments are present in high-grade prostate cancer tissues, which contain high levels of both HA and hyaluronidase. 7 In our previous article, we demonstrated that both high and low m.w. HA (i.e., HA fragments) species are present in the saliva of high-stage head-and-neck squamous cell carcinoma patients but not in the saliva of normal individuals. 8 Taken toge...