PURPOSE: To identify radiographic signs of mucosal damage by comparing hyst
erosalpingography with salpingoscopy in a prospective study.
MATERIALS AND METHODS: Forty-one candidates for laparoscopy underwent hyste
rosaipingography and peroperative salpingoscopy; at both, tubal patency was
noted. Radiographic criteria for mucosal abnormality were rounded filling
defects (ie, the cobblestone pattern) and the absence of longitudinal radio
lucent bands in the ampullary tract. At salpingoscopy, tubal mucosa was cat
egorized by means of inspection into five classes of fold pattern: classes
I and ii, normal; classes Ill-V, abnormal. Hysterosalpingographic and salpi
ngoscopic results were compared by means of a two-by-two table and kappa st
atistics.
RESULTS: Seventy-four tubes were evaluated. At hysterosalpingography, 31 tu
bes were distally nonpatent. Of these, 26 showed a distal obstruction at sa
lpingoscopy. None of the patent tubes at hysterosalpingography showed obstr
uction at salpingoscopy. The agreement between hysterosalpingography and sa
lpingoscopy in detecting abnormal mucosal pattern was 89.2% (kappa, 6.73; P
< .001). The cobblestone pattern always corresponded to intraluminal adhes
ions at salpingoscopy. The absence of radiolucent bands corresponded to abn
ormal mucosa at salpingoscopy in four of six cases. The cobblestone pattern
was found only in hydrosalpinges and never in patent tubes. Six normal pat
ent tubes at hysterosalpingography showed intraluminal adhesions at salping
oscopy.
CONCLUSION: Results indicate that the cobblestone pattern is an effective r
adiographic sign of intraluminal adhesions in hydrosalpinges and suggest th
at intraluminal disease in patent tubes might not always be excluded on nor
mal hysterosalpingograms.