Abdominal masses in children pose a diagnostic challenge due to the diversity of their causes and the subtlety of their clinical manifestations. Computed tomography (CT), a powerful imaging tool, allows for the assessment of their location, nature, vascularisation and extent, thereby facilitating diagnosis, treatment planning and follow-up.

In resource-limited countries, such as those in sub-Saharan Africa, CT is an essential asset, particularly when access to histology is restricted or delayed [1] . The CT characteristics of abdominal masses vary according to their aetiology, and several studies have shown the good performance of CT in distinguishing tumour masses from non-tumour masses. A study in India reports that Wilms’ tumour (nephroblastoma) is the most common tumour in children, followed by neuroblastomas and lymphomas, reflecting a similar spectrum throughout Asia [2] . In West Africa, studies in Mali and Burkina Faso report a high incidence of nephroblastoma, followed by Burkitt’s lymphoma and neuroblastoma [3] [4] .

In Côte d’Ivoire, data on the CT scan characteristics of paediatric abdominal masses remain limited, despite high demand for imaging in this area. Hence the interest of this study, which aims to describe the epidemiological and CT characteristics of these masses and to evaluate their concordance with histological results, in order to improve diagnostic management in our context.

The aim of this study was to highlight the contribution of computed tomography (CT) in the aetiological identification of paediatric abdominal masses. Despite the obvious diagnostic value of CT, our study has certain limitations, notably the small sample size, which limits the generalisability of the results. This situation illustrates the economic and structural constraints of our local context, where access to histological examination remains limited.

With regard to epidemiological and clinical data, the average age of patients was 5.6 years, close to that reported by Hamidou (6.2 years) [3] . Abdominal tumours are rare before the age of 1, more common between the ages of 1 and 6, and then decrease between the ages of 7 and 14 [5] . This highlights the need for clinical vigilance at all paediatric ages [6] - [8] . The 0 - 6 age group was the most represented (58.3%), consistent with Hamidou [3] , reflecting increased vulnerability to certain diseases, particularly infectious diseases, at this age. The male predominance (58.3%) found in this study, also observed by other African authors [3] [6] , remains without clear explanation in the literature, but could reflect gender inequalities in access to healthcare. Clinically, abdominal mass was the main reason for CT scan (77.8%), with a preferential location in the left hypochondrium and flank (24.3%). These locations are classically associated with renal masses such as nephroblastoma or neuroblastoma [9] - [12] .

With regard to CT scan data, only the location, characterisation, nature of the mass and statistical analysis will be discussed. In this study, abdominal masses were mainly located in the retroperitoneum (58.3%), with a predominance of tissue density (41.7%), heterogeneity in 58.3% of cases, and irregular contours in 28.8%. These radiological features are frequently associated with malignant tumours, particularly heterogeneity and irregular contours, which indicate an infiltrative nature. The presence of calcifications, observed in 25% of cases, is also a sign suggestive of malignancy, particularly in pathologies such as neuroblastoma [13] . The masses identified in this study were mainly of tumour origin (72.2%), confirming the tumour tropism of this location, particularly in infants and older children. Nephroblastoma was the leading cause (30.6%), followed by lymphoma (16.7%) and neuroblastoma (5.6%), results comparable to those reported in the literature [13] - [16] . Nephroblastoma, a typical embryonic tumour in children, generally occurs between the ages of 1 and 5, with a peak around $3\frac{1}{2}$ years [16] [17] . Lymphoma, particularly Burkitt’s lymphoma, is common in sub-Saharan Africa in connection with Epstein-Barr virus infection and malaria [18] [19] . Neuroblastoma, the most common extracranial solid tumour in children, ranks third here, as also indicated in the international literature [13] [20] . Non-tumour masses (27.8%) included malformations (megaureter, megacolon) and infections (pyometrocolpos, tuberculosis, malaria). This etiological diversity can be explained by the endemic context of certain infections in Côte d’Ivoire, as well as by the prevalence of congenital anomalies in areas with poor access to healthcare [4] [13] [14] .

Statistical analysis revealed no significant link between age and location (p = 0.5087) or between age and type of mass (p = 0.4487). However, a significant association was observed between malignant masses and their retroperitoneal location (p = 0.00018), suggesting a particular affinity of these tumours for this anatomical region.

Histologically, nephroblastoma (or Wilms’ tumour), the most common tumour, appears macroscopically as a large mass, well defined by a pseudocapsule, with a heterogeneous appearance, combining solid, necrotic, haemorrhagic and sometimes cystic areas. It can be polar, central or multifocal. Microscopically, it is an embryonic tumour of renal origin, characterised by three components: epithelial, mesenchymal and blastemal, in varying proportions depending on the case [21] . Popov [19] describes the macroscopic features and the variability in the proportions of the three histological components (blastema, epithelial, stromal). In this study, nephroblastoma accounted for 42% of histological diagnoses.

With regard to radiological-histological concordance and the diagnostic performance of CT, the results confirm the predominance of nephroblastoma in our context, followed by lymphoma. The 100% radiological-histological concordance in biopsied cases demonstrates the reliability of CT, which must be reconciled with the absence of overall statistical correlation between imaging morphology and histology, the latter analysis taking into account cases without histological confirmation. The lack of overall statistical correlation between CT morphology and histology can be explained by the inclusion in this analysis of patients without histological confirmation and with morphologically non-specific lesions. On the other hand, the 100% radiological-histological concordance observed in the biopsy subgroup reflects the diagnostic reliability of CT when compared with anatomopathological evidence, which also reflects a selection bias towards the most suspicious or operable cases. The diagnostic performance is comparable to that reported by Inoussa (Senegal, 2022) and Wadhwani (India, 2023) [20] [21] . The high sensitivity (100%) of CT scans can be explained by their excellent ability to detect the characteristic signs of malignancy (large size, necrosis, local invasion), particularly for tumours such as nephroblastoma, which have well-defined characteristics [21] . The lower specificity (83.3%) stems from the fact that certain benign tumours can present complex radiological characteristics (e.g. vascularisation, heterogeneity) that mimic the appearance of malignant tumours, thus creating false positives [22] [23] . Despite this limitation, CT remains an essential first-line examination for characterising paediatric abdominal masses and guiding management towards histological confirmation.

CT is an indispensable tool for the initial diagnosis of paediatric abdominal masses, with excellent performance when compared to histology. Even in the absence of systematic histological confirmation, the use of a rigorous imaging protocol and clinical follow-up can reduce verification bias. The development of infrastructure, improved access to CT and histology, and appropriate multidisciplinary care appear to be priorities for optimising the management of these conditions and improving children’s chances of survival.

Written informed consent was obtained from the parents of the patients for the publication of this study.

The datasets used and analysed during the current study are available from the corresponding author upon request.

Bouassa Davy Melaine Kouakou coordinated the study, conducted the research and analysed the data. Malick Soro, Sara Carole Sanogo, Brou Lambert Yao, Ibrahiman Touré, Iburaima Alamun Akanji, Kouamé Paul Bonfils Kouassi, Ataa Ange Rebecca Kobenan, Akoli Eklou Baudouin Bravo-Tsri, Kesse Emile Tanoh, Allou Florent Kouadio, Celine Yao, and Konaté Issa contributed to the drafting and critical revision of the manuscript. All authors read and approved the final version of the manuscript.

The authors would like to thank all the staff of the Radiology Department of Bouaké Teaching Hospital for their support during the study.