Matters arising 145

4 Reynolds JEF (Ed) Martindale. The Extra Phar- There are a number of small points in With regard to the points raised by Drs macoepia, 29th ed. London: Pharmaceutical respect ofthe data they present which require Evans and Kell. Patients attending our clinic Press. 1989. 5 National Institute of Occupational Safety and clarification: the indications for taking a are offered cervical cytology if (a) they have Hazards. Recommendations for occupational cervical smear are actually not given and it is not had a smear within the last 3 years or (b) safety and health standards. 1988. Morbidity not clear whether the 185 patients represent they or their sexual partners have genital

and mortality weekly report (suppl) and they have not had a smear within Genitourin Med: first published as 10.1136/sti.68.2.145-b on 1 April 1992. Downloaded from 1988;37:8. the total number smeared over the 5 month 6 Ethyl chloride BP. Production Information. period of study. It is really quite important to one year. The 185 women in the study were Berkshire UK. Bengue and Company Ltd know who was invited to participate and who drawn from 191 women having smears dur- 1982. declined. ing the study period. No patients declined to 7 Nordin C, Rosenguist M, Hollstedt C. Sniffing of ethyl chloride-an uncommon form of The proportion of abnormal smears was answer the life-style questions, but six abuse with serious mental and neurological much lower in the non- group (7 of 55) patients, all from the warts/warts contact symptoms. Int J Addict 1988;23:623-7. than in the wart group (52 of 117). However, group were not offered appoint- 8 Hes JP, Cohn DF, Strefler M. Ethyl chloride ments as they were about to leave the area sniffing and cerebellar dysfunction. Israel Ann the wart group is twice the size of the non- Psychiatry 1979;17:122-5. wart group, which may not be representative and thus were not included in the study. of women patients as a whole. All patients in the study with abnormal Although it is clearly stated that 59 patients smears, except two, who defaulted from had a cervical biopsy, it is less clear how follow up, were colposcoped, as were all, many were colposcoped. Surely some except 3 from the warts/warts contact group patients with abnormal smears showed no who had normal smears. Increased Incidence of cervical cyto- abnormality on colposcopy and therefore did Table 3 in the paper should have been logical abnormalities in women with not have a biopsy. If these patients are headed "Abnormal cytology results com- genital warts included in table 3, it is not clear from the pared with colposcopy results" and "Biopsy legend, but 65 (117-52) patients seem to proven CIN". Thus the NO CIN column I was interested to read the study by Rowen et have gone missing. represents those whose biopsies were neg- al' showing a higher rate of smear abnor- While the authors' conclusions appear ative and those who had a normal colposcopy malities in women with or contacts of genital valid from the data they present, the rele- and were therefore not biopsied. We apol- warts. Their observations agree with my own vance of mildly abnormal smears is called ogise for the confusion this may have ( Griffiths M, MD thesis, University of into question. Their biopsy results show that caused. London), where I found abnormal smears in cervical intra-epithelial neoplasia (CIN) was In the small number of women with 28% of women with warts and only 9% of present in 30% (13 of 43) of patients with abnormal smears but no history of warts or controls. Both studies effectively repeat the warts, 11% (1 of 9) of patients in contact wart contact, we would agree that there was a findings of Franceschi and colleagues2 who with warts, but in 43% (3 of 7) of patients high rate of CIN. They did however differ found an excess of abnormal smears (largely without warts or wart contact. From this it from other groups by virtue of having sig- of "supefficial dyskaryosis") in women with could well be concluded that genital warts are nificantly more sexual partners and it is warts compared with other STD clinic not related to CIN. possible some may have been infected with attenders, though a review of their paper B A EVANS HPV without developing warts. What is not demonstrates that high grade abnormalities P D KELL known is the natural history of sub-clinical Department of Genitourinary Medicine, were more common in controls. Charing Cross Hospital, HPV infection and whether such lesions However, we have shown no difference in London W6 8Rl, UK ultimately develop into frank warts or aceto- the risk of cervical epithelial disease between white lesions and if not, whether these sub- the two groups,3 when judged by colposcopy clinical lesions are also associated with and histology. We hypothesised that the abnormal cytology in the absence of warts. reason for this apparent discrepancy might, Rowen et al reply: We also agree that we did not find a at least in part, be due to more cautious The letters from Drs Griffiths, Evans and significant incidence of CIN in the warts/ examination and reporting of smears coming Kell concerning our recent paper are read warts contact groups, a point alluded to in which from women known to have warts, resulting with interest and raise some points the discussion. We did find differences in http://sti.bmj.com/ in a relative over-reporting of (particularly merit discussion. In our study the screeners rates of cytological abnormalities between minor) abnormalities by cytologists. This were aware of clinical details. It is of course the warts/warts contact group and the non hypothesis was supported by the findings of a possible to over report smears. However, we wart/wart contact group and forward the pilot study in which smears from women with do not feel that significant numbers are over notion that these abnormalities may be the warts were sent to cytology with clinical reported as several safeguards are in place to result of an acute reaction to HPV infection details of either "warts" or "routine" accord- prevent this. Firstly, a relatively junior which had settled by the time colposcopy was ing to prior randomisation. The study screener cannot send out an abnormal report performed. showed an excess of "abnormal" smears without the smear being reviewed by a senior D ROWEN on October 1, 2021 by guest. Protected copyright. a smear deemed to show CA CARNE among "warts" patients but this difference screener. Secondly, C SONNEX just failed to reach statistical significance mild dyskaryosis must be reviewed by a P COPPER owing to sample size. pathologist. Thirdly, follow-up smears from I believe that cytologists are more likely to women with borderline abnormalities on report abnormal smears if the clinical infor- previous smears are screened by a senior mation given refers to a history of warts, and MLSO. If any abnormality is found on that therefore would be interested to know wheth- smear, it and the previous smear are then er the cytologist in this particular study was reviewed by the cytopathologist. Further- blind to clinical information concerning the more, if there were significant over reporting patients' history of warts. one might suspect that the "current smear Increased evidence of cervical cytolog- MALCOLM GRIFFITHS normal, previous smear abnormal" group in ical abnormalities In women with genital Department of Obstetrics and Gynaecology, Wexham Park Hospital, our study would be larger than we found. warts Slough, Berks SL2 4HL, UK Dr Griffiths' results from his pilot study in which screeners were blinded to the real We read with great interest Dr Rowen et al's' 1 Rowen D, Came CA, Sonnex C, Cooper P. There or paper examining the need for increased Increased incidence of cervical cytological clinical details are of interest. may abnormalities in women with genital warts or may not be an excess of smears reported as cytological vigilance in women with genital contact with warts: a need for increased abnormal in the "warts" group. However his warts or contact with genital warts, and agree vigilance? Genitourin Med 1991;67:460-3. conclusion that the failure to demonstrate a that this group should also be offered colpo- 2 Franceschi S, Doll R, la Vecchia C. et al. Genital warts and cervical neoplasia: an epidemio- statistically significant difference in rates of scopic examination of the irrespective logical study. BrJ Cancer 1983;48:621-8. smears reported as abnormal in the two of their cervical cytology result. Our results 3 Griffiths M, Sanderson D, Penna LK. Cervical groups was simply due to sample size cannot and experience are in agreement with the epithelial abnormalities among women with above conclusion. We present figures from vulval warts-no more common than among be justified at this stage. If a full scale study, controls. Int J Gynaecol Cancer (In Press). with sufficient numbers in each group subse- our department on women with genital warts quently demonstrates a significant difference and negative cytology. In the period May in rates, then one may draw the conclusion 1987 to June 1988, 248 women with genital Rowen et al have looked at an important issue that the pilot study failed to demonstrate warts and 12 with wart contacts, attending regarding the relationship between genital significant differences because of sample the genitourinary medicine out-patient warts and . sizes. clinic, Royal Liverpool University Hospital