{"paper_id":"ca0b8dbb-4b21-45c5-9a11-6651257b5af2","body_text":"Open Journal of Obstetrics and Gynecology, 2025, 15(12), 2124-2137 \nhttps://www.scirp.org/journal/ojog \nISSN Online: 2160-8806 \nISSN Print: 2160-8792 \n \nDOI: 10.4236/ojog.2025.1512179  Dec.  24, 2025 2124 Open Journal of Obstetrics and Gynecology \n \n \n \n \nCorrelation between IOTA Simple Ultrasound \nRules and Histopathological Findings in \nAdnexal Masses: A Tertiary Hospital-Based \nStudy from Bangladesh \nFarhana Khatoon1* , Shirin Akter Begum1, Nasrin Akter1, Mohuwa Parvin1, Khairun Nahar1, \nFarzana Alam2, Md. Rifat Hassan3 \n1Gynecological Oncology Department, Bangladesh Medical University, Dhaka, Bangladesh \n2Department of Radiology and Imaging, Bangladesh Medical University, Bangladesh \n3Friendship Hospital Shyamnagar, Dhaka, Bangladesh \n \n \n \nAbstract \nBackground:  Preoperative differentiation of benign vs malignant adnexal \nmasses is essential for timely oncology referral and avoiding unnecessary sur-\ngery. We evaluated the diagnostic accuracy of the International Ovarian Tu-\nmor Analysis (IOTA) Simple Rules (SRs) in a tertiary oncology center in Bang-\nladesh. Methods:  In this prospective study, 94 consecutive patients underwent \nstandardized transvaginal/transabdominal ultrasonography and subsequent \nsurgery with histopathology as reference. Masses were classified per IOTA SRs \nas benign, malignant, or indeterminate. Diagnostic metrics were calculated for \nconclusive SR classifications.  Results:  Of 94 analyzable cases, histopathology \nshowed 53 benign (56.4%) and 41 malignant (43.6%). SR categorization \nyielded 48 benign (51.1%), 38 malignant (40.4%), and 8 indeterminate (8.5%). \nFor conclusive cases (n  = 86), IOTA SRs achieved sensitivity 84.2% (95% CI \n68.8 - 94.0), specificity 87.2% (74.3 - 95.1), PPV 84.2%, NPV 87.2%, and over-\nall accuracy 85.9% (76.6  - 92.5). The most frequent malignant features were \nM1 (irregular solid tumor) and M5 (very strong Doppler flow).  Conclusion:  \nIOTA SRs demonstrated robust, low- cost diagnostic performance in this ter-\ntiary oncology setting. Slightly lower estimates than some multicenter series \nlikely reflect single-center design, higher malignancy prevalence, and operator \nvariability. Larger multicenter studies in LMICs are warranted for broader val-\nidation. \nHow to cite this paper: Khatoon, F., \nBegum, S.A., Akter, N., Parvin, M., Nahar, \nK., Alam, F. and Hassan, Md.R. (2025) \nCorrelation between IOTA Simple Ultra-\nsound Rules and Histopathological Find-\nings in Adnexal Masses: A Tertiary Hospi-\ntal-Based Study from Bangladesh. Open \nJournal of Obstetrics and Gynecology, 15, \n2124-2137. \nhttps://doi.org/10.4236/ojog.2025.1512179 \n \nReceived:  November 27, 2025 \nAccepted: December 21, 2025 \nPublished: December 24, 2025 \n \nCopyright © 2025 by author(s) and  \nScientific Research Publishing Inc. \nThis work is licensed under the Creative \nCommons Attribution International  \nLicense (CC BY 4.0). \nhttp://creativecommons.org/licenses/by/4.0/   \n  \nOpen Access\n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2125 Open Journal of Obstetrics and Gynecology \n \nKeywords \nAdnexal Masses, IOTA Simple Rules, Ultrasonography, Diagnostic \nAccuracy, Bangladesh \n \n1. Introduction \nAdnexal masses are a frequent diagnostic and therapeutic challenge in gynecol-\nogy, encompassing conditions ranging from functional cysts to malignant ovarian \nneoplasms. Globally, ovarian cancer is the most lethal gynecological malignancy, \naccounting for over 295,000 new cases and 185,000 deaths annually, with five-year \nsurvival rates below 30% in many low - and middle-income countries. Early dif-\nferentiation of malignant from benign adnexal masses is therefore pivotal to ex-\npedite referral to gynecologic oncolo gy for malignancy and avoid unnecessary \nradical procedures for benign disease [1]-[5].  \nOver the last three decades, multiple strategies have been explored to improve \npreoperative diagnosis. Tumor markers such as CA -125, while widely used, have \nlimited specificity, particularly in premenopausal women [6]-[8]. Risk algorithms \nsuch as the Risk of Malignancy Index (RMI) and the OVA1 test combine CA-125, \nmenopausal status and imaging, but are either too resource -intensive or incon-\nsistently validated outside high-income settings [9]-[11]. In contrast, ultrasonog-\nraphy remains the first-line modality for adnexal mass evaluation due to its acces-\nsibility, non-invasiveness and affordability. However, its diagnostic reliability has \ntraditionally been undermined by operator dependence and lack o f standardized \ninterpretation criteria. \nTo address these limitations, the International Ovarian Tumor Analysis (IOTA) \ngroup developed standardized ultrasound terminology and the widely used IOTA \nSimple Rules (SRs) comprising five benign (B) and five malignant (M) features. \nIn a large multicenter prospective validation (~2,000 patients, 19 centers), SRs \nshowed excellent discrimination (sensitivity 92%, specificity 96%) with conclusive \nresults for ~77% of masses; adding expert judgment for inconclusive cases main-\ntained strong performance (91% sensi tivity, 93% specificity)  [12] [13]. As the \nIOTA framework matured, the three -step strategy— simple descriptors → SRs → \nexpert review— enabled non-expert sonographers to classify ~84% of masses us-\ning the first two steps, achieving sensitivity 95.2%, specificity 97.7%, and accuracy \n97.2% on ex ternal validation; a recent meta- analysis similarly reported pooled \nsensitivity and specificity near 94% [14] [15].  \nRegional evidence from Asia supports clinical utility. A large Indian study \nfound that IOTA SR and the ADNEX model achieved high diagnostic accuracy \n(~91%) and outperformed RMI 4; O -RADS showed the highest sensitivity (98%) \nwhile ADNEX provided the highest specificity (93%) [16]. \nAccording to GLOBOCAN 2022, Bangladesh recorded an estimated 2,846 new \novarian cancer cases and 1,857 deaths, underscoring the need for accurate, low -\n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2126 Open Journal of Obstetrics and Gynecology \n \ncost triage tools such as the IOTA Simple Rules [17]  [18]. Preliminary national \ndata from Chittagong Medical College Hospital (n  = 45) showed IOTA SRs out-\nperforming RMI 4 (accuracy 93.3% vs 73.3%) [19] [20]. \nTaken together, global validation, Asian data, and early Bangladeshi experience \nhighlight the diagnostic accuracy and feasibility of IOTA SRs. Yet adequately pow-\nered, tertiary oncology– based Bangladeshi studies remain scarce, particularly re-\ngarding the indeterminate group that challenges decision-making. Therefore, this \nstudy evaluated the diagnostic accuracy of IOTA Simple Ultrasound Rules against \nhistopathology in a tertiary oncology hospital in Bangladesh, estimating sensitiv-\nity, specificity, PPV, NPV, and outcomes of inconclusive cases, and describing the \nhistopathological spectrum of adnexal masses in this cohort. \n2. Materials and Methods \n2.1. Study Settings \nThis study was conducted at the Department of Gynecologic Oncology, Bangla-\ndesh Medical University, Shahbagh, Dhaka. Patients presenting with suspected \nadnexal masses to the outpatient and inpatient departments of Gynecologic On-\ncology and Obstetrics & Gynaecology units were included. \n2.2. Study Design \nA prospective, cross-sectional study design was employed to evaluate the diagnos-\ntic performance of the IOTA Simple Ultrasound Rules in distinguishing benign \nfrom malignant adnexal masses, using histopathological diagnosis as the reference \nstandard. \n2.3. Study Procedure \nEthical approval was obtained from the Institutional Review Board (IRB) of Bang-\nladesh Medical University before commencing the study. A total of 94 consecutive \npatients with clinically suspected adnexal masses, fulfilling inclusion criteria, were \nrecruited from outpatient and inpatient departments. Written informed consent \nwas obtained after explaining the study objectives, procedures, and voluntary par-\nticipation. \nEach participant underwent transvaginal ultrasonography (TVS) using either a \nVoluson P8 or Philips ultrasound machine. Ultrasound examinations followed the \nstandardized guidelines of the IOTA group (Timmerman et al., 2008). When TVS \nwas insufficient due to mass size or position, transabdominal ultrasonography was \nperformed. \nThe IOTA Simple Ultrasound Rules were applied to classify adnexal masses \nbased on specific sonographic features. A mass was classified as benign if one or \nmore benign features were present without any malignant features, malignant if \none or more malignant features were present without benign features, and incon-\nclusive if both benign and malignant features coexisted or if none of the defined \nfeatures were identified. \n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2127 Open Journal of Obstetrics and Gynecology \n \nMalignant features (M -Rules) included: irregular solid tumors (M1), ascites \n(M2), four or more papillary projections (M3), irregular multilocular solid tumors \n≥100 mm (M4), and very strong blood flow (color score 4, M5). Benign features \n(B-Rules) included: unilocular cysts (B1), solid components <7 mm (B2), acoustic \nshadows (B3), smooth multilocular tumors <100 mm (B4), and absent detectable \nblood flow (color score 1, B5). \nAll participants underwent surgical removal of the adnexal mass within 120 \ndays of ultrasound assessment. The surgical approach— laparoscopy or laparot-\nomy— was based on clinical indications. Excised specimens were examined histo-\npathologically, serving as the  definitive diagnostic reference. Ultrasound evalua-\ntions and interpretations were completed before surgery and before histopatho-\nlogical results were available to avoid bias. \n2.4. Sample Size \nThe sample size of 94 was calculated based on the estimated prevalence of ad-\nnexal masses, anticipated sensitivity and specificity of the IOTA rules, desired \nprecision, and confidence levels, providing adequate power to evaluate diagnos-\ntic accuracy.  \n2.5. Data Collection \nData were collected using a structured questionnaire capturing demographic details, \nclinical history, sonographic findings, serum CA-125 levels, and histopathological \noutcomes. Investigators conducted patient interviews and reviewed medical rec-\nords. Confidentiality was maintained by assigning unique identification numbers, \nwith data accessible only to the research team and used solely for this study. \n2.6. Data Analysis \nData were entered and analyzed using SPSS version 24. Quantitative variables \nwere expressed as mean ± standard deviation (SD), and categorical variables as \nfrequencies and percentages. Diagnostic performance metrics— sensitivity, speci-\nficity, PPV, and NPV — were calculated against histopathological diagnosis. The \nproportion of inconclusive cases by IOTA and their histopathological outcomes \nwere recorded. The distribution of final histopathological diagnoses was analyzed. \nStatistical significance was set at p < 0.05 with 95% confidence intervals. \n2.7. Ethical Considerations \nThe study adhered to the Declaration of Helsinki (1964) and its amendments. \nWritten informed consent was obtained after explaining the objectives, methods, \nrisks, and benefits. No experimental drugs or procedures beyond standard clinical \ncare were used. Co nfidentiality and anonymity were rigorously maintained via \nunique identifiers, with data access restricted to the research team. The study \nposed no additional risks or burdens beyond routine clinical management. No \nconflicts of interest were declared. \n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2128 Open Journal of Obstetrics and Gynecology \n \n3. Results \n3.1. Age and Menopausal Distribution  \nTable 1 summarizes malignancy increased with age, from 32–39% in younger groups \nto 63% among women older than 50 years. Postmenopausal women show higher \nmalignancy (43.9%) than premenopausal (37%), highlighting increased risk after \nmenopause. \n \nTable 1 . Age and menopausal distribution of respondents in benign and malignant tumor. \n(n = 94) \nCharacteristics Benign (%) Malignant (%) \nAge Group (in Years)   \n10 - 30 years 23 (67.6%) 11 (32.4%) \n30 - 50 years 20 (60.6%) 13 (39.4%) \n>50 years 10 (37.0%) 17 (63.0%) \nMenopausal status (n = 84) \nPremenopausal 17 (63.0%) 10 (37.0%) \nPostmenopausal 32 (56.1%) 25 (43.9%) \n3.2. Sonographic Categorization of Adnexal Masses \nThe sonographic assessment of adnexal masses using the IOTA Simple Ultra-\nsound Rules classified the masses into three categories. Among the 94 cases, 48 \n(51.1%) were categorized as benign, 38 (40.4%) as malignant, and 8 (8.5%) were \ninconclusive or indeterminate \n(Figure 1 ). \n \n \nFigure 1 . Sonographic categorization of adnexal mass by IOTA simple rules. \n3.3. Association of the Sonographic Findings and IOTA  \nClassification  \nThe majority of adnexal masses were unilateral across all histopathological \ngroups, found in 95.9% of benign, 97.4% of malignant, and all (100%) indetermi-\nnate cases, with no statistically significant difference (p = 0.8). Bilateral involve-\n\n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2129 Open Journal of Obstetrics and Gynecology \n \nment was rare. A significant association was observed in terms of locularity (p < \n0.001): unilocular cysts were most frequently seen in benign cases (89.8%), while \nmultilocular morphology predominated among malignant masses (71.1%). Re-\ngarding consistency, cystic structures were dominant in benign tumors (79.6%), \nwhereas malignant lesions more often exhibited mixed (50.0%) or solid (13.2%) \ncomponents (p < 0.001). The presence of free fluid was another strongly discrim-\ninating feature, significantly more common in malignant cases (71.1%) compared \nto benign (8.2%) and indeterminate groups (25.0%) (p < 0.001\n) (Table 2 ). \n \nTable 2 . Association of the sonographic findings and IOTA classification among the par-\nticipants. (n = 94) \nCharacteristics Benign (n = \n48) \nMalignant (n \n= 38) \nIndeterminate (n \n= 8) p-value \nNumber     \nUnilateral 47 (95.9%) 37 (97.4%) 8 (100.0%) 0.8 \nBilateral 1 (2.0%) 1 (2.6%) 0 (0.0%)  \nLocularity     \nUnilocular 44 (89.8%) 11 (28.9%) 6 (75.0%) <0.001 \nMultilocular 4 (8.2%) 27 (71.1%) 2 (25.0%)  \nConsistency     \nCystic 39 (79.6%) 14 (36.8%) 4 (50.0%) <0.001 \nMixed 8 (16.3%) 19 (50.0%) 2 (25.0%)  \nSolid 1 (2.0%) 5 (13.2%) 2 (25.0%)  \nFree Fluid     \nYes 4 (8.2%) 27 (71.1%) 2 (25.0%) <0.001 \nNo 43 (87.8%) 11 (28.9%) 6 (75.0%)  \n3.4. IOTA Sonography Diagnostic Performance \nThe IOTA Simple Sonographic categorization demonstrated high diagnostic ac-\ncuracy in differentiating benign and malignant adnexal masses when compared \nto histopathological diagnosis, with a sensitivity of 84.21% and specificity of \n87.23%. The positive predi ctive value (PPV) was 84.21%, while the negative pre-\ndictive value (NPV) was 87.23%, resulting in an overall accuracy of 85.88% \n(Ta-\nble 3). \n \nTable 3. Diagnostic accuracy of sonographic categorization compared to histopathological \ndiagnosis. (n = 86) \nDiagnosis Test  Sensitivity \n(%) \nSpecificity \n(%) PPV (%)  NPV (%)  Accuracy (%)  \nSonography vs \nHistopathology \n84.21 (68.75, \n93.98) \n87.23 \n(74.26, \n95.12) \n84.21 \n(68.75, \n93.98) \n87.23 \n(74.26, \n95.12) \n85.88 (76.64, \n92.49) \n3.5. Distribution of IOTA Ultrasound Features \nAmong malignant features, M1 (Irregular solid tumor) was the most frequently \nobserved, comprising 38.9% of malignant findings, followed by M5 (Very strong \n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2130 Open Journal of Obstetrics and Gynecology \n \nblood flow) at 27.8%. For benign features, B1 (Unilocular cyst) and B5 (No de-\ntectable blood flow (color score 1) were the predominant findings, each account-\ning for 32.9% of benign feature occurrences (Figure 2 ). \n \n \nFigure 2 . Distribution of IOTA ultrasound features. (n = 94) \n3.6. Association between Sonographic Categorization and  \nHistopathological Diagnosis \nThere was a statistically significant association between IOTA Simple Rules and \nhistopathology (p < 0.001). Of sonographically malignant masses, 84.2% were ma-\nlignant histologically; of sonographically benign, 87.2% were benign. Indetermi-\nnate cases (n = 8) showed 37.5% malignant and 62.5% benign outcomes (Table \n4). \n \nTable 4 . Association between sonographic categorization and histopathological diagnosis. \n(n = 94) \nSonographic Categorization Malignant n \n(%) Benign n (%) Total p-value \nMalignant 32 (84.2) 6 (15.8) 38 <0.001 \nBenign 6 (12.8) 41 (87.2) 47  \nIndeterminate 3 (37.5) 5 (62.5) 8  \nTotal 41 53 94  \n3.7. Distribution of Histopathological Diagnoses \nAmong the 94 participants included in the study, the distribution of histopatho-\nlogical diagnoses revealed that the majority of adnexal masses were benign. Spe-\ncifically, 53 cases (56.4%) were histologically confirmed as benign tumors, \nwhereas 41 cases (43.6%) were diagnosed as malignant (Figure 3 ). \n\n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2131 Open Journal of Obstetrics and Gynecology \n \n \nFigure 3 . Histological classification of Adnexal Mass. (n = 94) \n3.8. Sonographic Pattern Analysis of Inconclusive Cases  \nTable 5  details the specific sonographic rule combinations observed among the 8 \ncases categorized as indeterminate by IOTA and their corresponding histopatho-\nlogical outcomes. The most frequently observed combination was M2 + B2, found \nin 3 cases, two of which were benign and one malignant. Two cases exhibited the \npattern B3 + M4 + M5 (acoustic shadows with irregular multilocular solid tumors \nand very strong blood flow), and three cases showed B3 + M4, both patterns also \nresulting in benign histopathological diagnoses. \n \nTable 5 . Comparison between sonographic and histopathological findings of inconclusive \ncases. (n = 8) \nSonographic findings (IOTA \nrules) Frequency  \nHistopathology  \nBenign  Malignant  \nM2 + B2 3 2 1 \nB3 + M4 + M5 2  2 \nB3 + M4 3 3  \n3.9. Prevalence and Diagnostic Accuracy of Benign Sonographic  \nFeatures \nTable 6  summarizes benign ultrasound features (n = 57) and their diagnostic \nyield, where “Predicted” denotes correctly identified benign cases and “Result” the \ntotal instances of each feature. B2 (solid components <7 mm) was most frequent \n(13 cases) and most predictive (92.3%). B1 (unilocular cyst) and B3 (acoustic \nshadows) each appeared in 12 cases, with predictive values of 75.0% and 83.6%, \nrespectively. B5 (absent blood flow, color score 1) occurred in 12 cases with a \n75.0% predictive value. B4 (smooth multilocular tumors <100 mm) was least pre-\ndictive at 50.0% (8 cases). \n\n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2132 Open Journal of Obstetrics and Gynecology \n \nTable 6 . Prevalence and predictive power of benign factors. (n = 57) \nBenign  Predicted  Result  Percentage %  \nB1 9 12 75. 0 \nB2 12 13 92.3 \nB3 10 12 83.6 \nB4 4 8 50.0 \nB5 9 12 75.0 \n \nTable 7  summarizes the distribution and predictive accuracy of malignant ul-\ntrasound features. Among these, M5 (very strong blood flow, color score 4) was \nthe most commonly observed malignant criterion, found in 16 cases with a pre-\ndictive accuracy of 91.2%. M1 (irregular solid tumor) and M4 (irregular multiloc-\nular solid tumors ≥100 mm) were also frequent, observed in 9 and 7 cases respec-\ntively, with predictive accuracies of approximately 79% and 94%. Features such as \nM2 (ascites) showed strong predictive value exceeding 90%. Although M3 (pres-\nence of four or more papillary projections) was rare in this cohort (1 case), it \nachieved perfect predictive accuracy of 100%. \n \nTable 7 . Prevalence and predictive power of malignant factors. (n = 38) \nMalignant  Predicted  Results  Percentage %  \nM1 7 9 78.9 \nM2 4 5 92.9 \nM3 1 1 100 \nM4 6 7 94.4 \nM5 14 16 91.2 \n \nTable 8  presents the frequency and diagnostic distribution of combined benign \nand malignant ultrasound features per IOTA Simple Rules in adnexal masses. \nAmong the 94 cases analyzed, benign features B1 (unilocular cyst), B2 (solid com-\nponent <7 mm), B3 (acoustic sh adows), B4 (smooth multilocular tumors <100 \nmm), and B5 (absent blood flow, color score 1) were recorded alongside malig-\nnant features M1–M5. The most frequent benign features were B2 (13 cases) and \nB1/B3 (12 each), while malignant features M5 (16) and M4 ( 7) were most com-\nmon. Overall, 53/94 (56.4%) cases were benign and 41/94 (43.6%) were malignant. \n \nTable 8 . Observed combinations of benign and malignant ultrasound features of IOTA simple rules ranked by frequency (n = 94). \nApplicable B factors  Applicable M factors  FREQ Benign  Malignant  Rate of \nMalignancy  SN B1 B2 B3 B4 B5 M1 M2 M3 M4 M5 \n 12 13 12 8 12 9 5 1 7 16  53 41 43.6% \n4. Discussion \nThe study ’s findings indicate that applying the IOTA Simple Ultrasound Rules \n(SR) for diagnosing adnexal masses is a credible and effective method, aligning \naccurately with the histopathological results. Our findings indicate that IOTA SR \n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2133 Open Journal of Obstetrics and Gynecology \n \nis useful for differentiating benign from malignant adnexal masses and will help \nto manage these patients, especially in Bangladesh, a country with limited access \nto advanced diagnostic tools. The correlation between the ultrasound features and \nthe histopathology findings underscores the opportunity for ultrasound to be a \nnon-invasive, low-cost option for early diagnosis and pre-operative assessment. \nIn this tertiary oncology cohort, IOTA Simple Rules (SRs) showed robust diag-\nnostic performance (sensitivity 84.2%, specificity 87.2%, accuracy 85.9%) with an \nindeterminate rate of 8.5%. Malignancy prevalence was high (43.6%), consistent \nwith referral-center case mix. As high disease prevalence increases the proportion \nof true malignant cases, it can mathematically lower the negative predictive value \n(NPV), meaning even a negative test result carries a relatively higher residual risk \nof malignancy in such se ttings\n. Feature-level trends were coherent: multilocular-\nity, mixed/solid composition, and free fluid associated with malignancy, whereas \nunilocular and cystic morphology favored benign disease. Among single features, \nB2 (solid components <7 mm) and B3 (acoustic shadows) were strongly predictive \nof benignity, while M1 (irregular solid tumor) and M5 (very strong Doppler flow) \ndominated malignant findings. \nOur accuracy sits at the lower end of seminal IOTA validations (typically sen-\nsitivity/specificity ~90 - 96%) but remains clinically robust [12] [14] [15].The in-\ndeterminate proportion (8.5%) closely matches large external validations report-\ning ~7 - 11% [12] [14] [15]. Consistent with IOTA descriptors, acoustic shadows \n(B3) were benign-leaning— commonly seen in fibromas or dermoids— while asci-\ntes and solid/multilocular -solid architecture tracked with malignancy [12 ] [15] \n[21]. \nHead-to-head and synthesis studies from Asia report comparable or slightly \nhigher performance for SRs and related models. Indian and regional cohorts have \nshown mid -80s to mid -90s sensitivities/specificities for SRs and ADNEX, even \nwith non-expert examiners [16] [21] [22]. In particular, Khastgir \net al. found SRs \nand ADNEX to outperform RMI 4 overall; O-RADS tended to maximize sensitiv-\nity but sometimes at the expense of specificity— an effect echoed in systematic \ncomparisons [16] [21] [23]. Against this backdrop, our point estimates are plau-\nsible for a high- prevalence oncology setting and align with the directionality of \nfeature-outcome associations reported elsewhere [12] [14]-[16] [21] [23] [24]. \nSeveral context factors likely explain the modestly lower sensitivity/specificity \nthan multicenter benchmarks: (i) case mix/prevalence — our malignancy rate of \n43.6% can depress NPV and alter decision thresholds; (ii) operator variability —\nfeatures such as papillary projections (M3), vascular flow scoring (M5), and the \nrecognition of subtle solid components or septations are particularly prone to sub-\njective interpretation, affecting SR accuracy; (iii) equipment and protocol hetero-\ngeneity— different platforms and the need to alternate TVS/TAS in large or high-\nrising masses; and (iv) time to surgery (≤120 days) — interval change in lesion \ncharacteristics can introduce verification drift. These factors are typical in LMIC \noncology pathways and likely account for part of the gap from idealized multicen-\n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2134 Open Journal of Obstetrics and Gynecology \n \nter conditions [12] [14] [15] [22]. \nIndeterminate SR classifications were uncommon (8/94) yet clinically im-\nportant. In our series, mixed patterns yielded mixed outcomes ( e.g., M2 + B2 oc-\ncurred in three cases, two benign and one malignant ), whereas B3  + M4 + M5 \nskewed malignant and B3+M4 skewed benign. This reinforces standard escala-\ntion: apply SRs first; if indeterminate, proceed to a secondary risk tool (e.g., AD-\nNEX) or expert review, as recommended in the IOTA three-step strategy to reduce \nindeterminacy while preserving specificity [14] [15] [22]. In resource-constrained \nsettings, this tiered approach can foc us on oncologic referrals and limit unneces-\nsary laparotomies. \nAdnexal masses, particularly ovarian tumors, are a significant clinical challenge \nbecause of the difficulty in early diagnosis and the potential for malignancy, espe-\ncially in low -resource settings like Bangladesh. Many patients present with ad-\nvanced-stage disease, which is often associated with poorer prognosis and survival \nrates. For Bangladesh and similar contexts, SRs offer a low-cost, teachable frame-\nwork with performance that compares favorably to biomarker-heavy indices. Our \ndata echo prior Bangladeshi  experience where SRs outperformed RMI 4 against \nhistopathology [20]. Given the high malignancy prevalence in oncology centers, \nemphasizing structured SR acquisition, Doppler standardization, and brief up-\nskilling on feature recognition could lift sensitivity without sacrificing specificity. \nEmbedding a protocolized fallback (ADNEX or expert adjudication) for indeter-\nminate scans would further streamline triage. \nStratified analyses by menopausal status and histotype, plus decision -curve \nevaluation, would clarify where SRs add the most net clinical benefit and how to \noperationalize them across referral tiers. \nConsidering future research, IOTA SR should be evaluated alongside other in-\ndicators like serum CA-125 and contrast-enhanced ultrasound to further improve \nthe ability to differentiate benign from malignant masses. In addition, larger pro-\nspective multicenter studies with varied demographics would help validate IOTA \nSR across different clinical situations and its use in detecting early -stage malig-\nnancies. Finally, a review of borderline and uncommon malignancies would help \nrefine the rules and enhance the specificity of the system. \n5. Strengths and Limitations \nThis study has several strengths. Its prospective design with histopathology as the \ngold standard ensured a reliable assessment of diagnostic accuracy. The compre-\nhensive evaluation of all IOTA Simple Rule features, including indeterminate \ncases, allowed de tailed feature -level analysis. Conducted in a tertiary oncology \ncenter, the study captured a clinically relevant spectrum of adnexal masses and \nprovides contextually relevant evidence for Bangladesh and other LMIC settings, \naddressing a critical regional k nowledge gap. However, there are notable limita-\ntions. The single-center design and relatively small sample size may limit the gen-\neralizability of findings. Operator dependence of ultrasound, despite adherence to \n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2135 Open Journal of Obstetrics and Gynecology \n \nIOTA protocols, could introduce variability in feature detection. Referral bias to-\nward complex or high-risk cases may have led to an overrepresentation of malig-\nnant lesions, influencing sensitivity and specificity estimates. While the IOTA \nSimple Ultrasound Rules demonstrated robust diagnostic performance in our co-\nhort, the observed sensitivity (84.2%) and specificity (87.2%) were slightly lower \nthan the ranges reported in large multicenter validations (92  - 96%). This differ-\nence may reflect several factors, including the relatively small sample size, single -\ncenter design, and operator experience variability. Additionally, as a tertiary on-\ncology center, our cohort likely included a higher proportion of complex or ad-\nvanced masses, which could affect the performance metrics. These considerations \nshould be considered when interpreting the results and underscore the need for \nlarger, multicenter studies in Bangladesh to confirm generalizability.\n \n6. Conclusion \nThe IOTA Simple Ultrasound Rules demonstrated strong diagnostic performance \nin differentiating benign and malignant adnexal masses in a tertiary oncology set-\nting in Bangladesh, with high sensitivity, specificity, and overall accuracy. Feature-\nlevel analysi s confirmed that key sonographic markers such as multilocularity, \nsolid components, and ascites effectively discriminate malignancy, while indeter-\nminate cases highlight the need for careful evaluation. These findings support the \nfeasibility and clinical utility of implementing IOTA protocols in LMIC contexts, \npotentially improving early detection, guiding appropriate referrals, and reducing \nunnecessary surgical interventions. However, further large -scale, multicenter \nstudies are warranted to validate these results and ensure broader applicability \nacross diverse healthcare settings. \nAuthor Contributions \nAll authors contributed to the development of this work. \nConflicts of Interest \nThe authors declare no conflicts of interest regarding the publication of this paper. \nReferences \n[1] Phinyo, P., Patumanond, J., Saenrungmuaeng, P., Chirdchim, W., Pipanmekaporn, \nT., Tantraworasin, A., et al. (2020) Early-Stage Ovarian Malignancy Score versus Risk \nof Malignancy Indices: Accuracy and Clinical Utility for Preoperative Diagnosis of \nWomen with Adnexal Masses. Medicina, 56, Article No. 702.  \nhttps://doi.org/10.3390/medicina56120702 \n[2] Gui, J. (2025) Analysis of Global Ovarian Cancer Disease Burden and Its Changing \nTrend from 1990 to 2021. BMC Women’s Health, 25, Article No. 352.  \nhttps://doi.org/10.1186/s12905-025-03904-y \n[3] Yoeli-Bik, R., Longman, R.E., Wroblewski, K., Weigert, M., Abramowicz, J.S. and \nLengyel, E. (2023) Diagnostic Performance of Ultrasonography-Based Risk Models \nin Differentiating between Benign and Malignant Ovarian Tumors in a US Cohort. \n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2136 Open Journal of Obstetrics and Gynecology \n \nJAMA Network Open, 6, e2323289.  \nhttps://doi.org/10.1001/jamanetworkopen.2023.23289 \n[4] Koirala, P., Moon, A.S. and Chuang, L. (2020) Clinical Utility of Preoperative Assess-\nment in Ovarian Cancer Cytoreduction. Diagnostics, 10, Article No. 568.  \nhttps://doi.org/10.3390/diagnostics10080568 \n[5] Reinert, T., Nogueira-Rodrigues, A., Kestelman, F.P., Ashton-Prolla, P., Graudenz, \nM.S. and Bines, J. (2018) The Challenge of Evaluating Adnexal Masses in Patients \nwith Breast Cancer. Clinical Breast Cancer, 18, e587-e594.  \nhttps://doi.org/10.1016/j.clbc.2018.03.006 \n[6] Charkhchi, P., Cybulski, C., Gronwald, J., Wong, F.O., Narod, S.A. and Akbari, M.R. \n(2020) CA125 and Ovarian Cancer: A Comprehensive Review. Cancers, 12, Article \nNo. 3730. https://doi.org/10.3390/cancers12123730 \n[7] Al Musalhi, K., Al Kindi, M., Al Aisary, F., Ramadhan, F., Al Rawahi, T., Al Hatali, \nK., et al. (2016) Evaluation of HE4, CA-125, Risk of Ovarian Malignancy Algorithm \n(ROMA) and Risk of Malignancy Index (RMI) in the Preoperative Assessment of Pa-\ntients with Adnexal Mass. Oman Medical Journal, 31, 336-344.  \nhttps://doi.org/10.5001/omj.2016.68 \n[8] Shin, K., Kim, H., Kwon, B.S., Suh, D.S., Joo, J.K. and Kim, K.H. (2020) Clinical Use-\nfulness of Cancer Antigen (CA) 125, Human Epididymis 4, and CA72-4 Levels and \nRisk of Ovarian Malignancy Algorithm Values for Diagnosing Ovarian Tumors in \nKorean Patients with and without Endometriosis. Annals of Laboratory Medicine, \n40, 40-47. https://doi.org/10.3343/alm.2020.40.1.40 \n[9] Moolthiya, W. and Yuenyao, P. (2009) The Risk of Malignancy Index (RMI) in Di-\nagnosis of Ovarian Malignancy. Asian Pacific Journal of Cancer Prevention: APJCP, \n10, 865-868. \n[10] Tantipalakorn, C., Tinnangwattana, D., Lerthiranwong, T., Luewan, S. and Tong-\nsong, T. (2023) Comparisons of Effectiveness in Differentiating Benign from Malig-\nnant Ovarian Masses between Conventional and Modified Risk of Malignancy Index \n(RMI). International Journal of Environmental Research and Public Health, 20, Ar-\nticle No. 888. https://doi.org/10.3390/ijerph20010888 \n[11] Stephens, A.N., Hobbs, S.J., Kang, S., Bilandzic, M., Rainczuk, A., Oehler, M.K., et al. \n(2023) A Novel Predictive Multi-Marker Test for the Pre-Surgical Identification of \nOvarian Cancer. Cancers, 15, Article No. 5267.  \nhttps://doi.org/10.3390/cancers15215267 \n[12] Timmerman, D., Ameye, L., Fischerova, D., Epstein, E., Melis, G.B., Guerriero, S., et \nal. (2010) Simple Ultrasound Rules to Distinguish between Benign and Malignant \nAdnexal Masses before Surgery: Prospective Validation by IOTA Group. BMJ, 341, \nc6839. https://doi.org/10.1136/bmj.c6839 \n[13] Xie, W.T., Wang, Y.Q., Xiang, Z.S., Du, Z.S., Huang, S.X., Chen, Y.J., et al. (2022) \nEfficacy of IOTA Simple Rules, O-RADS, and CA125 to Distinguish Benign and Ma-\nlignant Adnexal Masses. Journal of Ovarian Research, 15, Article No. 15.  \nhttps://doi.org/10.1186/s13048-022-00947-9 \n[14] Alcázar, J.L., Vargas, F., Boscá, G., Salazar, B., Aguilar, J.C., Catalan, C., et al. (2025) \nIOTA Three-Step Strategy for Classifying Adnexal Masses: A Systematic Review and \nMeta-Analysis. Onco, 5, Article No. 22. https://doi.org/10.3390/onco5020022 \n[15] Hidalgo, J.J., Ros, F., Aubá, M., Errasti, T., Olartecoechea, B., Ruiz-Zambrana, Á., et \nal. (2019) Prospective External Validation of IOTA Three-Step Strategy for Charac-\nterizing and Classifying Adnexal Masses and Retrospective Assessment of Alternative \nTwo-Step Strategy Using Simple-Rules Risk. Ultrasound in Obstetrics & Gynecology, \n\nF. Khatoon et al. \n \n \nDOI: 10.4236/ojog.2025.1512179 2137 Open Journal of Obstetrics and Gynecology \n \n53, 693-700. https://doi.org/10.1002/uog.20163 \n[16] Khastgir, G., Siwatch, S., Singh, T., Jain, V., Sikka, P. and Srinivasan, R. (2025) Be-\nyond the Image: Performance of O-RADS, ADNEX, IOTA Simple Rules & RMI 4 in \nDifferentiating Benign and Malignant Adnexal Masses— An Indian Perspective. Eu-\nropean Journal of Obstetrics & Gynecology and Reproductive Biology, 313, Article \nID: 114585. https://doi.org/10.1016/j.ejogrb.2025.114585 \n[17] IARC. Bangladesh Fact Sheet.  \nhttps://gco.iarc.who.int/media/globocan/factsheets/populations/50-bangladesh-fact-\nsheet.pdf?utm_source=chatgpt.com  \n[18] Bray, F., Laversanne, M., Sung, H., Ferlay, J., Siegel, R.L., Soerjomataram, I., et al. \n(2024) Global Cancer Statistics 2022: GLOBOCAN Estimates of Incidence and Mor-\ntality Worldwide for 36 Cancers in 185 Countries. CA: A Cancer Journal for Clini-\ncians, 74, 229-263. https://doi.org/10.3322/caac.21834 \n[19] American Cancer Society. Global Cancer Facts & Figures 5. \n[20] Jabin, I., et al. (2022) Diagnostic Performance of IOTA-SRs and RMI 4 for Diagnosis \nof Malignancy in Ovarian Masses: Comparison with Histopathology. Journal of Chit-\ntagong Medical College Teachers’ Association, 33, 30-36. \n[21] Solanki, V., Singh, P., Sharma, C., Ghuman, N., Sureka, B., Shekhar, S., et al. (2020) \nPredicting Malignancy in Adnexal Masses by the International Ovarian Tumor Anal-\nysis-Simple Rules. Journal of Mid-life Health, 11, 217-223.  \nhttps://doi.org/10.4103/jmh.jmh_103_20 \n[22] Chankrachang, A., Lattiwongsakorn, W., Tantipalakorn, C. and Tongsong, T. (2025) \nDiagnostic Performance of ADNEX Model and IOTA Simple Rules in Differentiating \nMalignant from Benign Adnexal Masses When Assessed by Non-Expert Examiners. \nJournal of Clinical Medicine, 14, Article No. 2776.  \nhttps://doi.org/10.3390/jcm14082776 \n[23] Han, J., Wen, J. and Hu, W. (2024) Comparison of O-RADS with the ADNEX Model \nand IOTA SR for Risk Stratification of Adnexal Lesions: A Systematic Review and \nMeta-Analysis. Frontiers in Oncology, 14, Article ID: 1354837.  \nhttps://doi.org/10.3389/fonc.2024.1354837 \n[24] Su, N., Yang, Y., Liu, Z., Gao, L., Dai, Q., Li, J., et al. (2023) Validation of the Diag-\nnostic Efficacy of O-RADS in Adnexal Masses. Scientific Reports, 13, Article No. \n15667. https://doi.org/10.1038/s41598-023-42836-1","source_license":"CC0","license_restricted":false}