Women with abnormal Papanicolaou smear results and other cervical abnormalities usually are investigated further by colposcopy and punch biopsy of suspicious areas. Biopsies can be painful and cause significant bleeding, and even multiple biopsies can provide inadequate sampling of the endocervix and miss significant lesions. Monk and colleagues tested a new device to sample the endocervix. The spiral-shaped brush with stiff bristles was designed to collect samples rich in basal and parabasal cells that can be evaluated cytologically. The authors compared the specimens obtained using the brush with those obtained by punch biopsy in 52 women scheduled for large loop excision of the transformation zone because of abnormal results on a previous cervical biopsy.
All of the patients underwent sampling with the brush followed by punch biopsy, and both were performed by the same investigator using standardized techniques. Punch biopsy was taken of the same cervical areas only after all bleeding had resolved. Each patient scored discomfort using a scale from 0 to 4 and the investigators estimated bleeding on a scale from 0 to 3 following each procedure. A local anesthetic was then administered, and large loop excision of the transformation zone was performed. Cytologic findings on the three samples were compared.
Complete data were available for 47 of the 52 patients. Biopsy specimens taken with the brush were inadequate for assessment in eight women; in five of these, the punch biopsy specimens also were unsuitable for diagnosis. Only one of the remaining three women had a lesion identified by punch biopsy that was not detected on the brush specimen. Samples taken by large loop excision showed significant pathology in 39 women, of whom 31 had high-grade lesions and one had invasive cervical cancer. The punch biopsy correlated with high-grade disease in 53.1 percent of cases, whereas the brush biopsy specimens correlated in 79.3 percent when cell block techniques were used in preparation and in 76.7 percent when liquid cytology techniques were used (see the accompanying table on page 1081).
Patients scored the punch biopsy as significantly more painful than the brush technique. Pain scores of two or more were given by 17 women for punch biopsy but only by three for the brush sampling. Bleeding was also significantly less with the brush technique. Following punch biopsy, 28 of 47 women bled for more than one minute and required repeated operator intervention to control bleeding. Only two patients bled for more than one minute following brush sampling.
The authors conclude that the brush technique provides adequate sampling for the identification of cervical pathology with reduced patient discomfort and bleeding. With the brush technique, a larger proportion of the cervical surface area can be sampled, and the results are comparable to those of punch biopsy. They recommend that the brush technique be more widely used to evaluate women with suspicious cervical pathology, except when samples of the subepithelial stroma are needed to assess invasion.
editor's note: A device that offers comparable results to punch biopsy with less pain and bleeding and is easier to perform is very welcome. Innovations usually result in procedures becoming restricted to use by the sub-specialist; this one could mean that many more women would be investigated by their family physicians instead of being immediately referred for punch biopsy. Nevertheless, much depends on the skill of the operator, and we must not embrace a promising technology without making sure that the risk of missing the patient with invasive carcinoma has been clarified. We need large studies of the use of this technique in primary care. If a simpler and more acceptable technique becomes available, it could play an important role in reducing the continuing death rate of about 4,000 women in the United States every year from cervical cancer. Many of these cases result from women not following up after suspicious Pap smears because the next stage of investigation was unacceptable or inaccessible for financial or other reasons.—a.d.w.