The value of ultrasonography in the diagnosis of adnexal masses.
Seventy patients with palpable adnexal masses were subjected to ultrasonographic examination. Three patients with negative ultrasound and no disease were excluded from the final analysis. Correct diagnosis was obtained in 58.2% patients; contributory information in 16.4%. Ultrasonography is valuable in diagnosing functional and benign ovarian neoplasms. It is also useful in suspecting malignant ovarian neoplasms and confirming diagnosis of ectopic pregnancy, if correlated with the clinical findings.
Diagnosis of adnexal masses in a female patient presents diverse possibilities. These range from an ectopic pregnancy requiring immediate surgery to an ovarian malignancy or an inflammatory mass, requiring planned surgery or appropriate drug therapy. Ultrasonography has been used as a diagnostic modality in this situation,,,,. The aim of this study is to evaluate the role of ultrasonography in the diagnosis of adnexal masses.
Seventy patients having a palpable adnexal mass were selected for the study. After a detailed history and physical examination, complete blood count, pregnancy test and relevant biochcmical investigations, the probable clinical diagnosis was made. All patients were then subjected to ultrasound scanning of the pelvic organs using Aloka SSD Real Time Scanner with 3.5 and 5 mHz mechanical sector and 3.5 mHz linear transducers. Serial longitudinal and transverse scans of the pelvic organs obtained. The mass was then studied in detail with angled and additional scans. Where indicated, liver, para-aortic nodes, kidneys and other areas such as Morrison's pouch were also scanned. The diagnosis was confirmed by laparoscopy, laparotomy or by follow- up examinations when surgery was not indicated. The tissue obtained at surgery was subjected to histopathological examination.
At operation, the presence of dense adhesions, matted intestines or adherent omentum around the mass was noted. Flimsy adhesions, if present, were ignored. Ultrasonography findings were then correlated with the final diagnosis.
Ultrasonography provided a variable amount of information which was classified as diagnostic, contributory or erroneous as modified from Walsh et al. The establishment of a specific histologic diagnosis or the distinction of benign from malignant ovarian neoplasms was considered "diagnostic". The correct delineation of location, size, texture and consistency of a mass without a specific histologic diagnosis was classified as "contributory" information. Information, which did not assist in diagnosis or was misleading was classified as "erroneous".
In those patients who underwent operation, diagnostic accuracy using ultrasound was compared between two groups of patients one with adhesions and the other without adhesions.
Though seventy patients were selected for the study, 3 of thcm were excluded from the final analysis, because in these 3 patients ultrasonography was negative and no pathology was detected on investigation. So data of 67 patients were analysed.
Forty-nine of 67 patients (73%) were in the reproductive age group (20-40 years), while 10 (16%) were menopausal. Eighty per cent of all patients attended the out-patient or infertility clinic while the remaining presented with acute symptoms in the casualty department. Variable pathological conditions encountered on ultrasonography in the study are shown in [Table - 1].
The accuracy of ultrasonography results is shown in [Table - 2]. Of 67 patients, ultrasonography was diagnostic in 39 (58.2%) and yielded contributory data in 11 (16.4%). Thus in 74.6% of all patients, the information obtained was useful.
All functional ovarian cysts were diagnosed correctly, while 94.4% correlation was obtained in benign ovarian neoplasms (See [Figure - 1] and [Figure - 2]). Although haemorrhage in an ovarian cyst could be identified by the presence of low level internal echoes, torsion was diagnosed in only one out of six cases. However, ultrasonography was valuable in identifying an ovarian cyst as a possible cause of acute abdominal pain.
Malignancy was correctly suspected in 8 of 9 ovarian tumours. At laparotomy, consular invasion and fixity of the neoplasm were seen in 5 patients, out of whom only one was diagnosed ultrasonography.
In 15 patients of proved ectopic pregnancy, a correlation of 73.3% was obtained, ultrasound being diagnostic in 40% and contributory in 33.3%.
Hydro-salpinges could be diagnosed as retort shaped cystic masses on ultrasonography (See [Figure - 3]). Of 9 cases of pelvic inflammatory disease, only 2 could be correctly diagnosed sonographically.
In all, 62 patients underwent either laparoscopy or laparotomy. The correlation of the presence or absence of significant adhesions around the mass with the accuracy of the ultrasonography examination in these patients is shown in [Table - 3].
In early studies, bi-stable ultrasound imaging was used to distinguish cystic from solid masses. Correct localisation and description of masses could be achieved in 79-95% of patients in various studies,,,. With the advent of grey scale ultrasound, detailed tissue characterisation and histopathological diagnosis were attempted on ultrasound,.
In 74.6% of our patients, the ultrasonography diagnosis correlated with the final diagnosis. Walsh et a obtained diagnostic results in 56% and contributory results 23%, with an overall correlation in 79% out of 204 pelvic masses. Their slightly better results are attributable to the inclusion in their study of uterine masses such as fibroids, vesicular mole and intrauterine pregnancy in addition to adnexal masses.
All unil-ocular ovarian cysts in this study were benign on histopathological examination irrespective of size. Meire et a1 have described a 10.5% incidence of malignancy in unil-ocular tumours more than 5 cm in diameter. Multiloculation, thick septa and solid nodules are reliable indicators of malignancy on ultrasonography. Using these criteria, we could correctly suspect malignancy in 88.9% of the patients.
An unexpected sonographic diagnosis of malignancy in a young, infertile patient with a unilateral, mobile ovarian cyst enabled us to explain the possibility of radical surgery to the patients pre-operatively. Histopathology of this cyst revealed a mucinous cystadeno-carcinoma. Cystadenocarcinoma has 'diagnostic' appearances of solid tissue lining the inside wall of the cyst or papillary tumour excrescences protruding from the septa. However, we cases could not predict the histopathological diagnosis on ultrasonography.
An ovarian dermoid can be diagnosed by its typical appearance of a complex adnexal mass with clusters of highly reflective dense echoes within the lesion. This was the only ovarian neoplasm in which we could predict the histopathological diagnosis confidently.
In an old patient with post-menopausal bleeding, a solid ovarian mass was predicted to, be malignant but turned out to be a Brenner tumour.
The prognosis of cases with malignant ovarian tumours depends upon complete surgical resection. This is facilitated by the preoperative knowledge of actual tumour extent and staging. Requard et al have described 48% accuracy in correct staging of ovarian tumours on sonography. However, we did not find ultrasonography reliable in pre-operative staging.
In 15 patients with proved ectopic pregnancy, 73.3% correlation was obtained, which is comparable to other studies in which 71.090% cases have been correctly diagnosed on ultrasound,,,.
Ultrasonography could rule out an intrauterine pregnancy in all the patients. However, the demonstration of a viable extra-uterine pregnancy was not possible in most patients. Using a combination of criteria viz., a complex adnexal mass adjacent to an enlarged empty uterus, a pseudo-gestational sac with a single echogenic rim  and free fluid in the pouch of Douglas, we obtained correlation in 73.3% of the patients (See [Figure - 4]).
In one patient with abdominal pain, spotting and a palpable pelivic mass lateral to the uterus, ectopic pregnancy was ruled out by noting a pelvic kidney and normal adnexa on ultrasonography. Intravenous pyclography confirmed the diagnosis, obviating a laparoscopy. A similar case was reported by Cochrane and Thomas.
The early diagnosis of ectopic pregnancy is now possible with the use of serial ? human chorionic gonadotropin determinations and transvaginal ultrasonography examination,. However, these facilities are not easily accessible in India and transabdominal ultrasonography still plays an important role.
Endometrioid cysts of the ovaries present as diffuse internal echoes mimicking solid masses on ultrasonography. In our study, one patient diagnosed on sonography as a solid ovarian neoplasm was found to have endometrioid cysts at laparotomy (See [Figure - 5]).
Lawson and Albarelli attributed errors to an over interpretation of loops of bowel, technically poor examinations, misinterpretation of ectopic pregnancy or small lesions at the lower limit of resolution. The improved resolution and technological superiority of modern machines has improved the accuracy of ultrasonography. We found that the presence of adhesions around the mass is significant factor for diagnostic errors, possibly by the inter-position of bowel loops, latera anchoring of the adnexal or the presence of omentum surrounding the adnexal mass. Bowel gas is known to mimic hyper-echoeic zones of a dermoid Cyst.
Thus we conclude that, ultrasonography provides a definite diagnosis in most of the cases functional ovarian cysts and benign ovarian neoplasms and is helpful in suspecting ovarian malignancy. When combined with a careful history, examination and available laboratory tests, it is an important aid in the noinvasive investigations of ectopic pregnancy.