Heather Mitchell

Detection of Laboratory False Negative Smears by the PAPNET Cytologic Screening System

OBJECTIVE To evaluate the ability of PAPNET-assisted screening to identify abnormal slides that had originally been reported as negative on manual screening. STUDY DESIGN One hundred ninety-five abnormal slides were seeded into 20,000 slides that had been assessed as showing no abnormality on two occasions by manual screening. All slides were submitted for PAPNET review and the tiles assessed by trained cytotechnologists. RESULTS With a single assessment of the PAPNET tiles and when the prevalence of seeded abnormality was around 1%, only 44% of the seeded abnormal slides were recognized as abnormal. With multiple independent assessments and by increasing the prevalence of abnormality to 81%, 83% of the seeded abnormalities were recognized by at least one of three reviewing cytotechnologists. This increase in sensitivity appeared to be associated with a state of relative hyperalertness in the reviewing cytotechnologists. CONCLUSION The sensitivity of a PAPNET-assisted review was < 100% for the detection of seeded abnormal slides. Altering the format of presentation of information to the cytotechnologist may result in improved sensitivity.

A study of women who appear to default from management of an abnormal Pap smear.

EDITORIAL COMMENT: Most readers of the journal will have patients who have had minor abnormalities in a Pap smear and who have passed the statutory time for repetition of cervical cytology. Is the expense of a central register of abnormal pap smears worthwhile? ‐ do the reminder letters to practitioners result in diagnosis of significant pathology when the defaulters are rounded up? (i.e. what proportion have CIN III or worse). This paper caused the reviewer to ponder that it may be better to spend our dwindling resources on encouraging never‐smeared women over the age of 30 to attend for cytology, especially those over 50 years of age ‐ should general practitioners seek to enrol these women from their practices rather than be worried about defaulters who have had a previous smear indicative of dysplasia? How many general hospitals have a policy of performing cervical cytology on all adult female patients in their medical and surgical wards? ‐ surely it would be a simple administrative matter to do this as the women were admitted to hospital. Our reviewer also made the comment, suggested by the authors in their penultimate sentence, that patients referred to a dysplasia clinic should be returned to their general practitioner for continued follow‐up after initial assessment and treatment.

Detection of Unsuspected Abnormalities by PAPNET-Assisted Review

OBJECTIVE To determine the positive predictive value of abnormalities detected by PAPNET-assisted review of slides considered to show no abnormality on two manual screenings and to evaluate the repeatability of technical codes assigned by the PAPNET scanner. STUDY DESIGN PAPNET-assisted review was performed on 19,805 slides that had been assessed as showing no abnormality on two occasions by manual screening plus 195 slides with abnormal cells seeded at random. Abnormalities detected by cytotechnologists were graded by cytopathologists and compared with the findings of later histology/cytology. RESULTS On PAPNET-assisted review, the cytotechnologists identified 212 slides as containing unsuspected abnormalities; cytopathologists agreed with 76% (162/212) of these predictions. Later histology/cytology confirmed 54% (14/26) and 32% (33/102) of the predictions of high and low grade abnormality, respectively. The PAPNET scanner gave concordant technical codes for 94.8% of 2,690 slides that were submitted twice for scanning. CONCLUSION Some additional abnormalities will be detected by a PAPNET-assisted review, even among slides considered negative on two manual screenings. However, many of these abnormalities will not be confirmed on later investigation and will thus appear to represent false positive cytology. The assigning of technical codes by the PAPNET machine is subject to a degree of variation on repeat evaluation of the same slides.

Outcome after a cytological prediction of glandular abnormality

Background:  The National Health and Medical Research Council of Australia Guidelines for the Management of Women with Screen Detected Abnormalities are under review. The availability of population‐based Australian data on the outcome after a cytology prediction of glandular abnormality was considered relevant to revising the recommended investigations for women with these abnormalities.

Recent Negative Cytology Prior to Histologically Confirmed Carcinoma in Situ of the Cervix

Summary: Women with histologically confirmed carcinoma in situ of the cervix were studied within the records of the Victorian Cervical Cytology Registry. The prevalence of histologically confirmed carcinoma in situ during 1992 was 2.67 per 1,000 women screened. Thirty per cent (401 of 1,327) of the women with carcinoma in situ had negative cervical cytology reported during the 2 years prior to the diagnosis of carcinoma in situ and at least 49% (648 of 1,327) had negative cytology during the preceding 5 years. Adenocarcinoma in situ comprised 3.4% of all cases; these women were significantly older and more likely to have had a recent negative smear report than women with squamous carcinoma in situ. Fifty‐five per cent of the women with squamous carcinoma in situ had HPV reported on the biopsy compared with only 27% of the women with adenocarcinoma in situ. Ten per cent of the women with carcinoma in situ had a past history of cytological or histological abnormality; this proportion did not vary by type of carcinoma in situ. This relatively high proportion of negative cytology in close proximity to a diagnosis of carcinoma in situ is to be expected if there is active treatment of lesser lesions and frequent screening of members of the community.

DIFFERENCES BETWEEN FALSE NEGATIVE AND TRUE POSITIVE PAPANICOLAOU SMEARS ON PAPNET ASSISTED REVIEW

This study explored whether there were differences between false‐negative and true‐positive Papanicolaou (Pap) smears in the number of abnormal images on a Papnet‐assisted review. The degree of agreement between cytotechnologists over Papnet tile status (normal/abnormal) was assessed. False‐negative and true‐positive Pap smears preceding a histologic diagnosis of carcinoma in situ were scanned by Papnet, and the resulting digital images and slides were assessed independently by three cytotechnologists. The median number of abnormal tiles was 7.3 for false‐negative slides and 29.7 for true‐positive slides. Three‐way agreement between cytotechnologists was better for true‐positive slides (kappa, 0.60) than for false‐negative slides (kappa, 0.47). These results confirm intrinsic differences between false‐negative and true‐positive Pap smears. The fair to good agreement beyond chance on tile status indicates that a cautious approach should be adopted if digital image review is performed in a judgemental manner. Diagn. Cytopathol. 1998;19:138–140.

AN AUDIT OF THE WOMEN WHO DIED DURING 1994 FROM CANCER OF THE CERVIX IN VICTORIA, AUSTRALIA

An audit of 73 Victorian women who died from cervical cancer during 1994 is presented. Seventy per cent of the deaths occurred in women who were 50 years at the time of cancer diagnosis. Younger women had significantly better screening histories than older women, but overall only 10% of the deaths occurred among women who were adequately screened. There was no excess of deaths in rural women or in women of non‐English speaking background. Four of the 6 deaths in women who were diagnosed with cancer while <35 years of age occurred in adequately screened women. If this finding is confirmed in a larger series, it raises questions about the ability of Papanicolaou screening to control disease in this age group.

Evidence Against Diathermy as a Beneficial Treatment for Human Papillomavirus Infection of the Cervix

EDITORIAL COMMENT: In recent years there have been numerous reports regarding human papillomavirus infection of the female genital tract and its possible association with preinvasive and invasive carcinoma of these sites (cervix, vagina, vulva). The most alarming information comes from studies of DNA hybridization techniques which indicate that a very high proportion of patients with apparently normal cervices on colposcopy and cytology harbour this virus. Many studies indicate that only a small proportion of patients who show clinical evidence of the papillomavirus have evidence of continuing infection or go on to develop intraepithelial or invasive carcinoma. The proportion of patients who spontaneously eradicate signs of the disease is not established, neither do we know why some patients cure themselves of the disease and why others do not. There is also confusion in the literature as to whether local treatment can eradicate this disease. This paper gives a very good account of the current state of knowledge about papillomavirus infection. It seems difficult to understand how diathermy to the cervix can eradicate this viral infection when many studies have shown that papillomavirus infection is usually present in the vagina and vulva when it involves the cervix, unless the cellular damage induced by the cautery enhances the patients immune response. There is evidence in numerous reports that biopsy of papillomavirus lesions and early dysplasias of the cervix modifies the natural history of these lesions, possibly by alteration of local immune function. The present study also indicates that treatment of the cervix for papillomavirus infection alone does not affect the incidence of subsequent development of intraepithelial neoplasia. These authors have stressed the fact that further long‐term follow‐up is required of patients with evidence of papillomavirus infection.

Pap Smear Results of Victorian Teenagers, 1980–1987

EDITORIAL COMMENT: There are times when appraisal of the medical literature seems as difficult as that of articles in the lay press — whose story do you believe? This careful paper reporting a large series of patients is reassuring — there is no epidemic of human papillomavirus in Victorian teenagers since the prevalence in the last decade has remained at about 5%. For cervical intraepithelial neoplasia the figures are less reassuring but not alarming. Next we must ask our cytologists to provide similar comparative data for the other age groups in Victoria. Certainly this paper would indicate that the pessimism of the colposcopists due to investigation of patients referred to them is not the last word on human papillomavirus infection.