To date, the IOTA study is the largest study in the literature on ultrasound diagnosis of ovarian pathology. It started years ago in 1999 and included nine European centers. The 5-year survival rate associated with ovarian cancer is less than 30%. The limiting factor for early diagnosis of ovarian tumour is lack of standardised terms and procedures in gynaecological sonography. A standardized technique for preoperative classification of adnexal masses was defined by IOTA group. Major highlight of the study were 10 simple ultrasound rules that had high sensitivity and specificity and were applicable to a large number of tumours. On application of one or more M-rules in the absence of a B-rule, or one or more B-rules in the absence of a M-rule, the mass is classified as malignant or benign respectively. If both M-rules and B-rules apply, or if no rule applies, the mass could not be classified [1], and was labeled as inconclusive. Although large no. of studies are available proving the efficacy of these rules, however a prospective study directly applying these diagnostic rules to the patient was lacking. This study was performed to truly establish the diagnostic utility of these rules in our country and to estimate and compare the sensitivity and specificity of given rules with histological diagnosis and establish their use as a tool in early diagnosis of ovarian malignancy.
Materials and Methods
The present study was a hospital based case control prospective study carried out in department of radiodiagnosis, Govt. medical college and Rajindra hospital, Patiala, Punjab, India. Case control study was done in 50 women with suspected ovarian pathology attending gynaecology clinic at department of Obstetrics and Gynaecology during the period of December 2014 to October 2016. Mandatory written informed consent was taken from all the patient for participation in study. Patient with suspicion of adnexal mass on pelvic examination or discovered during previous sonographic examination were included and pregnant patient and patient not willing for surgery in our hospital were excluded. Transvaginal sonography was done on all the patients. Examination was limited to Transabdominal sonography in virgins and where it was not possible to completely visualize the mass by a transvaginal probe. Permission from the hospital authority for conducting the study was obtained. Sonographic assessment of the given adnexal masses was done using wide band intracavitary transducer having frequency 5-7.5 MHz for transvaginal sonography and 2-5 MHz curved transducer for transabdominal sonography which was available in our department. Ultrasound was done by both experienced (experience of seven years or more in abdominopelvic sonography) and new sonographers and each patient were classified as benign or malignant on basis of IOTA simple rules [Table/Fig-1]. In cases of bilateral mass, the larger mass was considered. During the sonographic examination both the sonographers were blind folded for their results as well as the histopathological finding. All patients included in the study underwent surgery (within 120 days of sonographic examination). Histopathological diagnosis was obtained in all cases and used as the gold standard. Collected data was statistically analysed using chi square test and kappa statistical method.
Simple IOTA rules for predicting benign or malignant ovarian tumour.
| Rules for predicting a malignant tumour (M-rules) | Rules for predicting a benign tumour (B-rules) |
|---|
| M1 Irregular solid tumour | B1 Unilocular cyst |
| M2 Presence of ascites | B2 Presence of solid components where the largest solid component is <7 mm in largest diameter; |
| M3 At least four papillary structures | B3 Presence of acoustic shadows |
| M4 Irregular multilocular solid tumour with largest diameter ≥100 mm | B4 Smooth multilocular tumour with largest diameter <100 mm |
| M5 Very strong blood flow (color score 4) | B5 No blood flow (color score 1) |
Results
During the study 55 women were eligible but five were excluded as they deny for surgery. Out of 50 patients eight had bilateral masses (larger one was considered for study). The youngest patient in the present study was 18-year-old and the eldest was 77-year-old female, the mean age (SD) was 42.5 years. Malignancy was more common in 6th decade age group [Table/Fig-2]. Malignancy was more common in postmenopausal patients (71.43%) [Table/Fig-3]. Out of total 50 cases studied, IOTA Rules classified 45 patients, 29 as benign and 16 as malignant. Five cases where rules could not be applied or where both B and M rules were applicable were labelled as indeterminate or inconclusive cases [Table/Fig-4].
Age distribution in benign and malignant ovarian masses (n=50).
| Age Group (in years) | Benign | Malignant |
|---|
| Number | Percentage | Number | Percentage |
|---|
| 11-25 | 6 | 16.67 | 0 | 0.00 |
| 26-40 | 23 | 63.89 | 1 | 7.14 |
| 41-55 | 4 | 11.11 | 3 | 21.43 |
| 56-70 | 2 | 5.56 | 9 | 64.29 |
| 71-85 | 1 | 2.77 | 1 | 7.14 |
| Total | 36 | 100.00 | 14 | 100.00 |
Relation of menopausal status with ovarian tumours.
| Menopausal Status | Benign | % | Malignant | % |
|---|
| Postmenopausal | 5 | 13.89 | 10 | 71.43 |
| Premenopausal | 31 | 86.12 | 4 | 28.57 |
| Total | 36 | 100 | 14 | 100 |
Classification of cases as per IOTA simple ultrasound rules.
| Nature of the Mass as per IOTA Rules | No. |
|---|
| Benign | 29 |
| Malignant | 16 |
| Indeterminate | 5 |
| Total | 50 |
In the present study, 23 different combinations of Benign and Malignant simple ultrasound features were observed as tabulated in [Table/Fig-5]. Out of all, most common was combination of Rules B1 and B5 (eight out of 23 combinations) showing no risk of malignancy on final histopathological analysis. Out of eight combinations having 100% rate of malignancy, most frequent combination was application of Rule M2 alone or in combination with M4 (two each out of eight combinations).
Observed combinations of benign and malignant ultrasound features of IOTA simple rules ranked by frequency.
| Sr. N o. | Applicable B-Factors | Applicable M-Factors | Frequency | Benign | Malignant | Rate of malignancy (%) |
|---|
| B1 | B2 | B3 | B4 | B5 | M1 | M2 | M3 | M4 | M5 |
|---|
| 1 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 8 | 8 | 0 | 0.00 |
| 2 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 6 | 6 | 0 | 0.00 |
| 3 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 6 | 5 | 1 | 16.67 |
| 4 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 4 | 4 | 0 | 0.00 |
| 5 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 3 | 2 | 1 | 33.33 |
| 6 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 3 | 2 | 1 | 33.33 |
| 7 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 2 | 0 | 2 | 100.00 |
| 8 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 0 | 2 | 0 | 2 | 100.00 |
| 9 | 1 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 2 | 2 | 0 | 0.00 |
| 10 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 0 | 1 | 0 | 1 | 0 | 1 | 100.00 |
| 11 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 1 | 0 | 1 | 100.00 |
| 12 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 100.00 |
| 13 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 1 | 1 | 0 | 1 | 100.00 |
| 14 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 1 | 100.00 |
| 15 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 1 | 1 | 0 | 1 | 100.00 |
| 16 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 1 | 0 | 1 | 100.00 |
| 17 | 0 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0.00 |
| 18 | 1 | 1 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0.00 |
| 19 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 1 | 0 | 0.00 |
| 20 | 0 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0.00 |
| 21 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0.00 |
| 22 | 0 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0.00 |
| 23 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 1 | 1 | 0 | 0.00 |
B5 was the most common occurring factor in all benign lesions, followed by B1. But B2 and B4 predicted the result most correctly (100%) followed by B5 and B1 [Table/Fig-6].
Prevalence and predictive power of benign factors.
| Benign | Predicted | Result | Percentage |
|---|
| B1 | 17 | 16 | 94% |
| B2 | 2 | 2 | 100% |
| B3 | 6 | 5 | 83% |
| B4 | 6 | 6 | 100% |
| B5 | 20 | 19 | 95% |
Rules M1, M2, M4 were all equally common factors each present in eight patients with suspected malignancy. Out of these best M factor was M2 (presence of ascites) which correctly predicted malignancy in all the eight patients in which it was found [Table/Fig-7].
Prevalence and predictive power of malignant factors.
| Malignant | Predicted | Result | Percentage |
|---|
| M1 | 8 | 4 | 50% |
| M2 | 8 | 8 | 100% |
| M3 | 1 | 1 | 100% |
| M4 | 8 | 5 | 63% |
| M5 | 4 | 3 | 75% |
Kappa statistics showed high level of agreement between USG and histopathological findings and it was statistically significant (K=0.323) [Table/Fig-8]. The sensitivity for the detection of malignancy in cases where IOTA simple rules were applicable was 91.66% and the specificity was 84.84%. Accuracy was 88.88% [Table/Fig-9]. Classifying inconclusive cases as malignant the sensitivity and specificity was 93% and 80% respectively.
Comparision of results of IOTA simple rules with histopathological findings.
| Nature of the Mass as per IOTA Rules | No. | Histopathological Result |
|---|
| Benign | Malignant |
|---|
| Benign | 29 | 28 | 1 |
| Malignant | 16 | 5 | 11 |
| Indeterminate | 5 | 3 | 2 |
Efficacy of IOTA simple rules.
| Efficacy of IOTA Simple Rules |
|---|
| Sensitivity | 91.66% |
| Specificity | 84.84% |
| PPV | 68.75% |
| NPV | 96.55% |
| Accuracy | 88.88% |
There was one false negative case in our study which on sonography had unilocular cyst and histopathological diagnosis of epithelial ovarian carcinoma, patient had history of contralateral ovarian carcinoma which was operated two years back.
On screening [Table/Fig-10], it shows Comparison between sonographic and histopathological findings of inconclusive cases.
Comparison between sonographic and histopathological findings of inconclusive cases.
| Sonographic findings(IOTA RULES) | Histopathology |
|---|
| M4 + B5 | Mucinous cystadenoma |
| M4+M2+B5 | Epithelial ovarian carcinoma |
| M1+B3+B5 | Tubo-ovarian abscess |
| Smooth, regular, solid tumour(>100 mm), no ascites and moderate blood flow(no rule applied) | Dysgerminoma |
| M1+B3+B5 | Endometrioma |
Considering new residents in our institution as inexperienced sonographers in comparison to professors as experienced sonographers, no difference was found in interpretation of IOTA simple rules proving IOTA rules to be highly reproducible, easy to train and apply.
Discussion
In previous published studies the IOTA ultrasound rules were not directly applied during sonographic examination, the sonographic data was later collected from patients and was evaluated as per prediction models. Till date only few studies which applied this diagnostic test directly to patient have been performed. Our study overcomes this limitation by directly applying IOTA simple ultrasound rules on the patients. A total of 55 patients with suspected ovarian pathology were evaluated using transvaginal ultrasonography and transabdominal ultrasonography when transvaginal approach was not feasible. All the cases were examined first by an inexperienced sonographer (new residents with less than one year training in ultrasound) with prior knowledge of IOTA rules, the findings were noted. Same patient was later examined by an experienced sonographer and mass was again classified strictly as per IOTA rules. No expert opinion was taken into consideration. Both the rules were later compared with each other and 100% agreement was found between two results. Findings were correlated with histopathological findings. Out of 55 patients initially evaluated in the study, 50 patients were included in the final analysis who underwent surgery. Thus rate of inconclusive result was 10%. On histopathology, out of total 50 cases studied 14(28%) were malignant and 36 (72%) were benign. Out of 45 cases where IOTA simple rules were applicable, 16(35.5%) were malignant and 29(64.4%) were benign. Out of the five cases classified as inconclusive two were malignant and three were benign.
The sensitivity and specificity of present study most closely related to study by Hartman CA et al., who reported a sensitivity and specificity of 91% and 87% respectively. The specificity of our study was lower as compared to these seven studies. This variation may be due to limited number of patients studied in the present study as compared to other studies [Table/Fig-11] [2-8].
Comparison of results of present study with the published data is shown.
| Author and year of study | No. of Patients | Patients in whom rules are applicable | Malignant Tumours | Benign Tumours | Prevalence (%) | True Positive | False Positive | False Negative | True Negative | Sensitivity (%) | Specificity (%) |
|---|
| Timmerman D et al., [2] (2008) | 507 | 386 | - | - | 29 | 106 | 25 | 6 | 249 | 95 | 91 |
| Timmerman D et al., [3] (2010) | 1938 | 1501 | 542 | 1396 | 25 | 340 | 49 | 29 | 1083 | 92 | 96 |
| Fathallah K et al., [4] (2011) | 122 | 109 | 14 | 108 | 10.10 | 8 | 3 | 3 | 95 | 73 | 97 |
| Hartman CA et al., [5] (2012) | 103 | 91 | 30 | 73 | 24.20 | 20 | 9 | 2 | 60 | 91 | 87 |
| Sayasneh A et al., [6] (2013) | 255 | 214 | 74 | 181 | 24.80 | 46 | 3 | 7 | 158 | 87 | 98 |
| Alcazar JL et al., [7] (2013) | 340 | 270 | 55 | 285 | 12.20 | 29 | 6 | 4 | 231 | 88 | 97 |
| Nunes N et al., [8] (2012) | 303 | 237 | 135 | 168 | 44.30 | 101 | 15 | 4 | 117 | 96 | 89 |
| Present study (2016) | 50 | 45 | 14 | 36 | 28.00 | 11 | 5 | 1 | 28 | 91.66 | 84.84 |
Sub analyses was done for pre- and postmenopausal women in our study and compared with the literature [Table/Fig-12] [3,5-8].
Comparative sensitivity and specificity of published data with the current study in pre and postmenopausal women [3,5-8].
| Author and year of study | Patients with Rules Applicable | Prevalence of Malignancy (%) | Sensitivity | Specificity |
|---|
| ALL WOMAN |
| Timmerman D et al., [3] (2010) | 796 | 22.2 | 96 | 96 |
| Hartman CA et al., [5] (2012) | 91 | 24.2 | 87 | 87 |
| Sayasneh A et al., [6] (2013) | 214 | 24.8 | 98 | 98 |
| Alcazar JL et al., [7] (2013) | 270 | 12.2 | 97 | 97 |
| Nunes N et al., [8] (2012) | 237 | 44.3 | 89 | 89 |
| Present Study (2016) | 45 | 28 | 91.66 | 84.84 |
| PREMENOPAUSAL |
| Timmerman D et al., [3] (2010) | 526 | 9.5 | 90 | 97 |
| Hartman CA et al., [5] (2012) | 39 | 23.1 | 89 | 90 |
| Sayasneh A et al., [6] (2013) | 143 | 15.4 | 82 | 100 |
| Alcazar JL et al., [7] (2013) | 217 | 7.4 | 88 | 97 |
| Nunes N et al., [8] (2012) | 130 | 22.3 | 90 | 89 |
| Present Study (2016) | 30 | 3.33 | 100 | 86.2 |
| POSTMENOPAUSAL |
| Timmerman D et al., [3] (2010) | 270 | 47 | 91 | 94 |
| Hartman CA et al., [5] (2012) | 36 | 30.6 | 91 | 80 |
| Sayasneh A et al., [6] (2013) | 71 | 43.7 | 90 | 93 |
| Alcazar JL et al., [7] (2013) | 53 | 32.1 | 88 | 100 |
| Nunes N et al., [8] (2012) | 107 | 71 | 99 | 87 |
| Present Study (2016) | 15 | 73.33 | 90.9 | 75 |
It shows that sensitivity was higher in premenopausal women (100%) than in postmenopausal women (90.9%) while specificity was slightly lower (86.2% and 75% respectively). The increased sensitivity and specificity in premenopausal women compared to the postmenopausal women in present study may be explained by increased number of inconclusive cases in the premenopausal patients in present study where the simple rules could not be applied (10%).
The rate of inconclusive result in this study was 10%. Using a strategy classifying the inconclusive cases as malignant was adopted by many already published studies in this regard as shown in [Table/Fig-13] [3-5,7-9].
Test performance of different simple rules (sr) strategies in the IOTA studies.
| Study | Sensitivity If SR Applicable | Specificity If SR Applicable | Sensitivity With SR+MA | Specificity With SR+MA |
|---|
| Timmerman D et al., [3] | 91 | 96 | 94 | 80 |
| Fathallah K et al., [4] | 73 | 97 | 79 | 88 |
| Hartman CA et al., [5] | 91 | 87 | 94 | 76 |
| Alcazar JL et al., [7] | 88 | 97 | 93 | 81 |
| Nunes N et al., [8] | 96 | 89 | 97 | 70 |
| Tantipalakorn C et al., [9] | 83 | 95 | 87 | 81 |
| Present study | 92 | 85 | 93 | 80 |
SR-simple rules; SR+MA- simple rules+ inconclusive result considered as malignant.
Applying this strategy in present study the sensitivity of the index test increased from 92% to 93% whereas specificity decreased from 85% to 80% respectively. This result correlated well with the published data and the sensitivity and specificity of present study was close to sensitivity and specificity as per study by Timmerman et al., [3].
Limitation
The major limitation of this study was small sample size.
Conclusion
The sensitivity for the detection of malignancy in cases where IOTA simple rules were applicable was 91.66% and the specificity was 84.84%. Accuracy was 88.88%. Classifying inconclusive cases as malignant the sensitivity and specificity were 93% and 80% respectively. The most important factor in ovarian malignancy is time of detection. With early detection, it is possible to improve the survival of patient significantly. USG offers inherent advantages of easy availability, low cost and lack of radiation exposure but being more subjective than other modalities. IOTA simple ultrasound rules can eliminate this problem as they are highly sensitive and specific in predicting ovarian malignancy yet being reproducible, easy to train and use.
SR-simple rules; SR+MA- simple rules+ inconclusive result considered as malignant.