Using History and Colposcopy to Select Women for Endocervical Curettage: Results From 2,287 Cases

Abstract

In this study the authors conducted a retrospective review of the medical records of women who attended the dysplasia clinic at Cook County Hospital between July 1996, and December 1999 to investigate the frequency with which endocervical curettage (ECC) provides information that alters recommended treatment of cervical dysplasia. During this period 2825 patients had a colposcopic examination for evaluation of cervical abnormality. More than 80% of the women (N = 2287) underwent ECC at the time of colposcopy. In 209 patients the ECC specimen was unsatisfactory. The diagnosis was negative for disease in 1745 patients, dysplasia in 320 patients, and cancer in 13 patients. Compared with those with negative findings, women with a positive ECC specimen were three times more likely to have a diagnosis of cervical intraepithelial neoplasia grade 2 (CIN 2) or worse (P <.001) and nine times more likely to have a diagnosis of cancer (P <.001). For 105 of the 333 women with positive ECC specimens (105 of 2287, 4.6%), this diagnosis was the only indication for loop excision. Conization was done in 69 of these patients; 9 had no lesion, 8 had koilocytosis or squamous atypia, 2 had ungraded CIN, 17 had CIN 1, 3 had CIN 2, 17 had CIN 3, and 3 had cancer. Colposcopic examination in the three patients with cancer showed a benign lesion in two and a high-grade abnormality in one. In all, 33 (1.4%) who had no indication for loop excision except positive ECC were eventually diagnosed with CIN 2, CIN 3, or cancer. During the period of this study, loop excision or cone biopsy was recommended for all women with positive ECC, women with at least CIN 2 on colposcopic biopsy, women with an unsatisfactory colposcopic evaluation and cytology showing a low-grade squamous intraepithelial lesion, and those with cytologic evidence of cancer. In this series atypical squamous cells of unknown origin favoring dysplasia were found in 29 women of the 333 women with positive ECC results. These women are at low risk for serious disease, but adding this criterion to the indications for cone excision would have increased the percentage of patients whose treatment plan was altered by ECC. On the other hand, limiting the indications for cone biopsy to women with at least atypical glandular cells of undetermined significance or high-grade squamous intraepithelial lesion would have decreased the number of women whose treatment was changed by ECC. Also, if the patients with a colposcopic impression of CIN 2 or 3 were recommended for cone biopsy, the percentage of those whose treatment was changed by ECC results would have been lower. An analysis of possible characteristics associated with positive ECC results found higher age (P <.001), higher parity (P <.001), earlier age at first intercourse (P =.006), history of unsatisfactory colposcopy (P <.001), and history of colposcopic impression of CIN (P <.001) to be significant predictors of having a positive ECC result. However, in this series, using only these factors as indications for ECC would have missed several women with high-grade CIN or cancer. There would have been no missed cancers if ECC had been omitted in women with satisfactory colposcopy, a normal colposcopic impression, and/or nulliparity.