Interpretation of hysterosalpingography

The advent of assisted reproduction techniques such as in vitro fertilisation (IVF) has led to a significant increase in the demand for hysterosalpingography (HSG), a fluoroscopic imaging procedure involving the introduction of iodinated water-soluble contrast medium into the female genital tract to delineate the endo-cervical canal, uterine cavity, fallopian tubes and pattern of intraperitoneal spill. NICE guidance for the investigation of subfertility and a common perception is that HSG is primarily recommended for the detection of tubal disease. Performed correctly it is also invaluable in detecting and monitoring uterine pathology. The technique was first described in 1910 by Rindfleish using a bismuth solution; by the early 1920s the oily contrast medium lipiodol was being used when, amazingly, one of the indications for the technique was to diagnose pregnancy. By the 1990s water soluble contrast media had largely replaced lipiodol for HSGs. Current indications for the procedure are numerous (table 1). HSG typically forms part of subfertility investigations that also include an ultrasound, hormonal profiles and semen analysis. Other imaging modalities are usually supplementary (such as MRI) but can be helpful to further delineate abnormalities detected at HSG. It is important to appreciate that findings considered abnormal at HSG may or may not be of significance in regards to the patient’s symptoms and subsequent management. This article will focus on factors that influence the interpretation of hysterosalpingograms. A comprehensive description of the technique, contraindications and complications can be found elsewhere. A crucial factor upon which an accurate interpretation can be made is image quality; achieving the highest possible image quality is dependent on numerous other factors including clinical background, patient anxiety, imaging sequence, procedure, technique and the experience of the radiographer and operator (table 2). The paper or electronic request form for HSG should provide the patient’s gynaecological and obstetric history, other significant medical history and results of investigations to date. Past history affects the conduct and interpretation of the study. For example, previous uterine instrumentation should alert the radiologist to the possibility of uterine scar tissue, Asherman’s syndrome. A history of ectopic pregnancy increases the index of suspicion for tubal pathology such as salpingitis isthmica nodosa (SIN). An empathetic, calm environment is essential; women who have researched HSGs on the internet or had a poor smear experience will be anxious. Anxiety predisposes to pain which in turn can cause vaginismus or tubal spasm and an erroneous diagnosis of tubal occlusion. A simple preprocedural checklist allows the operator to confirm all relevant information, establish rapport, provide a full explanation of the procedure, record the result of the pregnancy test and any drugs given and obtain written informed consent. Confirmation of demographic details, date of last menstrual period together with the outcome of any previous pregnancy, ie live birth, miscarriage, termination or ectopic, is essential. We recommend performing a urine pregnancy test in all women as well as ascertaining no intercourse in that menstrual cycle – positive tests can occur in women who have experienced apparently cyclical bleeding in an early pregnancy. History of post-partum complications should be sought, eg manual removal of placenta or post-partum haemorrhage; pelvic inflammatory disease, which may indicate tubal disease; previous pelvic or abdominal surgery such as myomectomy, tubal surgery, C-section or evacuation retained products of conception, which can all affect interpretation. Ideally the procedure is undertaken and supervised by experienced staff with an interest in gynaecology. The procedure must be conducted gently and slowly, as a rushed speculum insertion or forceful contrast medium injection will cause unnecessary pain and diminish image quality. A wide range of catheters must be available, because the external cervical os may range in size from pinpoint to large and patulous. A 5Fr balloon inflated gently in the upper cervical canal is our catheter of choice. Once the catheter is in place the pelvic cavity is examined fluoroscopically and a control image acquired if any radio-opaque lesions are observed. Infusing contrast gently, a minimum of four image exposures, early filling AP, RAO, LAO and AP to show spill, are acquired supplemented by frame grabs as desired. Frame grab, fluoroscopic images alone are insufficient for accurate image interpretation. A low dose programme must be used and screening kept to a minimum. Women will frequently voice concerns about dose to the ovaries and question the operator regarding potential harm to future pregnancies. The national diagnostic reference level (DRL) for HSG is 2Gycm2; 9 our local DRL for HSG is 0.7Gycm2. It is essential to obtain a true en-face view of the uterine cavity (figure 1); views that only show a foreshortened view down the long axis of the uterus are non-diagnostic. Oblique images are essential (figures 2a-b) and can be obtained using a C-arm, patient rotation or a combination of both. Air bubbles can be distinguished from polyps by use of patient rotation, and it should be noted that small air bubbles can lodge at the cornu and mimic cornual occlusion. If a balloon catheter is used and is inflated in the lower uterine segment a view following deflation (figures 2c-d) must be obtained to avoid missing lower uterine segment pathology. The range of uterine and /or tubal abnormalities that can be identified are shown in tables 3 and 4.