Artificial intelligence in histopathological image analysis of central nervous system tumours: A systematic review

The convergence of digital pathology and artificial intelligence could assist histopathology image analysis by providing tools for rapid, automated morphological analysis. This systematic review explores the use of artificial intelligence for histopathological image analysis of digitised central nervous system (CNS) tumour slides. Comprehensive searches were conducted across EMBASE, Medline and the Cochrane Library up to June 2023 using relevant keywords. Sixty-eight suitable studies were identified and qualitatively analysed. The risk of bias was evaluated using the Prediction model Risk of Bias Assessment Tool (PROBAST) criteria. All the studies were retrospective and preclinical. Gliomas were the most frequently analysed tumour type. The majority of studies used convolutional neural networks or support vector machines, and the most common goal of the model was for tumour classification and/or grading from haematoxylin and eosin-stained slides. The majority of studies were conducted when legacy World Health Organisation (WHO) classifications were in place, which at the time relied predominantly on histological (morphological) features but have since been superseded by molecular advances. Overall, there was a high risk of bias in all studies analysed. Persistent issues included inadequate transparency in reporting the number of patients and/or images within the model development and testing cohorts, absence of external validation, and insufficient recognition of batch effects in multi-institutional datasets. Based on these findings, we outline practical recommendations for future work including a framework for clinical implementation, in particular, better informing the artificial intelligence community of the needs of the neuropathologist.

Ovarian tissue cryopreservation and transplantation in patients with central nervous system tumours

STUDY QUESTION

Is there a possibility of reseeding cancer cells potentially present in frozen ovarian tissue from patients with central nervous system (CNS) tumours?

SUMMARY ANSWER

Malignancy reseeding in cryopreserved ovarian tissue from 20 patients with CNS tumours was not detected by histology, immunohistochemistry (IHC), molecular biology or xenotransplantation.

WHAT IS KNOWN ALREADY

Ovarian metastasis potential has been documented in patients with leukaemia, borderline ovarian tumours, advanced breast cancer and Ewing sarcoma. However, data on the safety of transplanting frozen-thawed ovarian tissue from cancer patients with CNS tumours are still lacking.

STUDY DESIGN, SIZE, DURATION

This prospective experimental study was conducted in an academic gynaecology research laboratory using cryopreserved ovarian cortex from 20 patients suffering from CNS tumours. Long-term (5 months) xenografting was performed in immunodeficient mice.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Subjects enrolled in the study were suffering from one of six types of CNS tumours including medulloblastoma, ependymoma, primitive neuroectodermal tumours, astrocytoma, glioblastoma and germinoma. The presence of malignant cells was investigated with disease-specific markers for each patient in cryopreserved and xenografted ovarian tissue by histology, IHC via expression of neuron-specific enolase (NSE) and glial fibrillary acidic protein (GFAP), and reverse transcription droplet digital polymerase chain reaction (RT-ddPCR) for quantification of GFAP and ENO2 gene amplification.

MAIN RESULTS AND THE ROLE OF CHANCE

Serial sections of cryopreserved and xenografted ovarian tissue from 20 patients showed no malignant cells by histology. All samples were negative for NSE and GFAP, although these neural markers were expressed extensively in the patients' primary tumours. Analysis by RT-ddPCR revealed no cancer cells detected in cryopreserved and xenografted ovarian fragments from subjects with astrocytoma, ependymoma, glioblastoma or medulloblastoma. Taken together, the study found no evidence of malignancy seeding in frozen-thawed and xenotransplanted ovarian tissue from patients affected by CNS cancers.

LIMITATIONS, REASONS FOR CAUTION

This analysis cannot guarantee complete elimination of disseminated disease from all cryopreserved ovarian cortex, since we are unable to examine the fragments used for transplantation.

WIDER IMPLICATIONS OF THE FINDINGS

This is the first study to be conducted in patients with CNS cancers undergoing ovarian tissue cryopreservation and transplantation, and clearly demonstrates no tumour seeding in their frozen-thawed and xenografted tissue. This information is vital for doctors to provide patients with meaningful and accurate advice on the possibilities and risks of ovarian tissue reimplantation.

STUDY FUNDING/COMPETING INTEREST(S)

This study was supported by grants from the Fonds National de la Recherche Scientifique de Belgique-the Excellence of Science (FNRS-EOS), number 30443682 awarded to M.-M.D. and T.Y.T.N., FNRS grant number 5/4/150/5 and FNRS-PDR Convention grant number T.0077.14 awarded to M.-M.D., grant 2018-042 from the Foundation Against Cancer awarded to A.C., and private donations (Ferrero, de Spoelberch). The authors declare no competing financial interests.

TRIAL REGISTRATION NUMBER

N/A.

Brain tumour diagnostics using a DNA methylation-based classifier as a diagnostic support tool

AIMS

Methylation profiling (MP) is increasingly incorporated in the diagnostic process of central nervous system (CNS) tumours at our centres in The Netherlands and Scandinavia. We aimed to identify the benefits and challenges of MP as a support tool for CNS tumour diagnostics.

METHODS

About 502 CNS tumour samples were analysed using (850 k) MP. Profiles were matched with the DKFZ/Heidelberg CNS Tumour Classifier. For each case, the final pathological diagnosis was compared to the diagnosis before MP.

RESULTS

In 54.4% (273/502) of all analysed cases, the suggested methylation class (calibrated score ≥0.9) corresponded with the initial pathological diagnosis. The diagnosis of 24.5% of these cases (67/273) was more refined after incorporation of the MP result. In 9.8% of cases (49/502), the MP result led to a new diagnosis, resulting in an altered WHO grade in 71.4% of these cases (35/49). In 1% of cases (5/502), the suggested class based on MP was initially disregarded/interpreted as misleading, but in retrospect, the MP result predicted the right diagnosis for three of these cases. In six cases, the suggested class was interpreted as 'discrepant but noncontributory'. The remaining 33.7% of cases (169/502) had a calibrated score <0.9, including 7.8% (39/502) for which no class indication was given at all (calibrated score <0.3).

CONCLUSIONS

MP is a powerful tool to confirm and fine-tune the pathological diagnosis of CNS tumours, and to avoid misdiagnoses. However, it is crucial to interpret the results in the context of clinical, radiological, histopathological and other molecular information.

Space-time clustering of childhood cancers in Switzerland: A nationwide study

The aetiology of childhood cancers remains largely unknown. It has been hypothesized that infections may be involved and that mini-epidemics thereof could result in space-time clustering of incident cases. Most previous studies support spatio-temporal clustering for leukaemia, while results for other diagnostic groups remain mixed. Few studies have corrected for uneven regional population shifts which can lead to spurious detection of clustering. We examined whether there is space-time clustering of childhood cancers in Switzerland identifying cases diagnosed at age <16 years between 1985 and 2010 from the Swiss Childhood Cancer Registry. Knox tests were performed on geocoded residence at birth and diagnosis separately for leukaemia, acute lymphoid leukaemia (ALL), lymphomas, tumours of the central nervous system, neuroblastomas and soft tissue sarcomas. We used Baker's Max statistic to correct for multiple testing and randomly sampled time-, sex- and age-matched controls from the resident population to correct for uneven regional population shifts. We observed space-time clustering of childhood leukaemia at birth (Baker's Max p = 0.045) but not at diagnosis (p = 0.98). Clustering was strongest for a spatial lag of <1 km and a temporal lag of <2 years (Observed/expected close pairs: 124/98; p Knox test = 0.003). A similar clustering pattern was observed for ALL though overall evidence was weaker (Baker's Max p = 0.13). Little evidence of clustering was found for other diagnostic groups (p > 0.2). Our study suggests that childhood leukaemia tends to cluster in space-time due to an etiologic factor present in early life.

The role of high-dose myeloablative chemotherapy with haematopoietic stem cell transplantation (HSCT) in children with central nervous system (CNS) tumours: protocol for a systematic review and meta-analysis

OBJECTIVES

The objective of the study is to conduct a systematic review to compare the effects of high-dose chemotherapy (HDCT) with autologous haematopoietic stem cell transplantation (HSCT) versus standard-dose chemotherapy (SDCT) in children with malignant central nervous system (CNS) tumours.

METHODS

Standard systematic review methods aimed at minimising bias will be employed for study identification, selection and data extraction. Ten electronic databases will be searched, along with citation searching and reference checking. Studies assessing the effects of HDCT with HSCT in children with CNS tumours will be included. The outcomes are survival (overall, progression-free, event-free, disease-free), response rates, short- and long-term adverse events and health-related quality of life (HRQoL). Two reviewers will independently screen and select randomised and non-randomised controlled trials and controlled and uncontrolled observational studies for inclusion. Quality assessment will be tailored to the different study designs. Where possible data will be summarised using combined estimates of effect for the hazard ratio for survival outcomes and the risk ratio for response rates. A fixed effect model will be used; sub-group analyses and meta-regression will be used to explore potential sources of heterogeneity between studies.

DISCUSSION

Given the poor prognosis of malignant brain tumours in children in terms of survival and quality of life, this review will help guide clinical practice by summarising the current evidence on the use of high-dose myeloblative chemotherapy with stem cell support in children with CNS tumours.

Papillary ependymoma of the endometrium

AIMS

To report an exceptional case of papillary ependymoma occurring in the endometrium.

METHODS AND RESULTS

A clinicopathological study was performed regarding a case of papillary ependymoma occurring in the endometrial cavity of a 61-year-old patient who had presented with a solid-type, stage III anaplastic ependymoma of the ovary, treated with cytoreductive surgery that included total abdominal hysterectomy. The uterus was enlarged and showed a dilated cavity, with broadly implanted papillary excrescences without myometrial invasion that were covered by tall, cylindrical cells. These cells had glial fibrillary acidic protein-expressing cytoplasm that was also positive for cytokeratins 7, 8, 18, and 34β-E12, epithelial membrane antigen, S100 protein, vimentin, and oestrogen and progesterone receptors.

CONCLUSIONS

Pathogenetically, the presence of this uterine ependymoma could represent either an example of multicentricity or a phenomenon of transtubal implantation of the ovarian tumour. Exceptionally, papillary ependymoma can occur in the endometrium, and may prompt differential diagnoses with other papillary endometrial tumours. Pathologists should consider this rare possibility in the differential diagnosis of papillary ovarian and endometrial lesions.

The United Kingdom Childhood Cancer Study of exposure to domestic sources of ionising radiation: 2: gamma radiation

This article reports measurements of household levels of gamma and cosmic rays at the addresses of children with cancer at the time of diagnosis and six months before, and of similar data at the addresses of control children. There is no indication of increased risk with increasing dose rates either in matched or unmatched analyses, with or without adjustment for deprivation. Sub-division by diagnostic group did not reveal any association with any specific types of malignancy. Studies of the relationship between household gamma rays and radon concentration show no evidence of any interactions.