Download 21 ...
, . 183: 126 (1997)
INTERPRETATION OF THE SIGNAL PATTERNS PRODUCED BY NISH IN CERVICAL NEOPLASIA HARBOURING HPV The signal patterns produced by non-isotopic in situ hybridization (NISH) in cervical neoplasia harbouring human papillomavirus (HPV) have been validated by three independent laboratories.1–3 These NISH signals have been interpreted as follows: type 1: diffuse intranuclear signal representing episomal virus; type 2: punctate intranuclear signal representing integrated virus; and type 3: a combination of types 1 and 2 physical states of HPV DNA. Although the significance of NISH HPV signal patterns in anal carcinoma has not yet been elucidated, the signal morphology has been clearly described. It is therefore perplexing to note the interpretation of NISH signal patterns in a recent publication in this journal.4 Williams et al. have misinterpreted a type 3 NISH signal (Figs 1 and 2) and a type 2 signal (Fig. 3) as a diffuse nuclear pattern. If these signal patterns are to contribute to our understanding of HPV carcinogenesis, then it is crucial that we are uniform in our interpretation. The second curious observation made by Williams et al. is that of the HPV cytoplasmic signal. Although no supporting evidence in the form of photomicrographs is provided to corroborate this observation, it would have
been interesting to have read their interpretation of this novel signal pattern. To the best of our knowledge, such a NISH signal pattern has not been previously described using HPV DNA probes. K. C W. G Department of Anatomical Pathology, School of Pathology, South Africa Institute for Medical Research and University of Witwatersrand, PO Box 1038, Johannesburg 2000, South Africa REFERENCES 1. 2. 3. 4.
Cooper K, Herrington CS, Stickland JE, Evans MF, MdGee JO’D. Episomal and integrated human papillomavirus in cervical neoplasia shown by non-isotopic in situ hybridisation. Clin Pathol 1991; 44: 990–996. Kristiansen E, Jenkin A, Holm R. Coexistence of episomal and integrated HPV 16 DNA in squamous cell carcinoma of the cervix. J Clin Pathol 1994; 47: 253–256. Berumen J, Unger ER, Casas L, et al. Amplification of human papillomavirus types 16 and 18 n invasive cervical cancer. Hum Pathol 1995; 26: 676–681. Williams GR, Lu QL, Love SB, Talbot IC, Northover JMA. Properties of HPV-positive and HPV-negative anal carcinomas. J Pathol 1996; 180: 378–382.
AUTHORS’ REPLY While we recognize that studies on HPV DNA in the cervix have led to the characterization of the three types of NISH signal referred to by Dr Cooper and Dr Grayson, the signal patterns have not been correlated with the physical state of HPV DNA in anal epithelium. Also, in practice, it is often difficult to distinguish types 1 and 3; i.e., to see punctate positivity on a background of diffuse signal. For these reasons, we limited our observations to punctate and diffuse patterns, with no further subdivision. We therefore interpreted Figs 1 and 2 as representing diffuse staining. We do, however, agree that Fig. 3 illustrates punctate (type 2) signal, rather than diffuse as suggested in the text. Cytoplasmic HPV DNA signal using NISH techniques has been documented in association with epithelial maturation in viral warts.1 The present paper is the first to describe cytoplasmic signal for HPV by ISH in anal carcinoma. The signal is usually weak and difficult to demonstrate by photography and the sig-
? 1997 by John Wiley & Sons, Ltd.
nificance has yet to be determined, it being unclear whether these signals represent HPV DNA or mRNA. However, the presence of cytoplasmic signal, which is always associated with strong nuclear signal and well-differentiated epithelial tumour cells, may represent viral activation or an early state of transition from latent to lytic infection and is thus worthy of further investigation. G. R. W, Q. L. L I. C. T Imperial Cancer Research Fund Histopathology Unit, 44 Lincoln’s Inn Fields, London WC2A 3PX REFERENCES 1.
Kankatsu Y, Sherwood MJ. In situ hybridisation at light and electron microscopic levels: identification of human papillomavirus nucleic acids. Pathology 1992; 24: 91–98.