A CT-based radiomics classification model for the prediction of histological type and tumour grade in retroperitoneal sarcoma (RADSARC-R): a retrospective multicohort analysis

Preoperative Contrast-Enhanced CT-Based Deep Learning Radiomics Model for Distinguishing Retroperitoneal Lipomas and Well‑Differentiated Liposarcomas

RATIONALE AND OBJECTIVES

To assess the efficacy of a preoperative contrast-enhanced CT (CECT)-based deep learning radiomics nomogram (DLRN) for predicting murine double minute 2 (MDM2) gene amplification as a means of distinguishing between retroperitoneal well-differentiated liposarcomas (WDLPS) and lipomas.

METHODS

This retrospective multi-center study included 167 patients (training/external test cohort, 104/63) with MDM2-positive WDLPS or MDM2-negative lipomas. Clinical data and CECT features were independently measured and analyzed by two radiologists. A clinico-radiological model, radiomics signature (RS), deep learning and radiomics signature (DLRS), and a DLRN incorporating radiomics and deep learning features were developed to differentiate between WDLPS and lipoma. The model utility was evaluated based on the area under the receiver operating characteristic curve (AUC), accuracy, calibration curve, and decision curve analysis (DCA).

RESULTS

The DLRN showed good performance for distinguishing retroperitoneal lipomas and WDLPS in the training (AUC, 0.981; accuracy, 0.933) and external validation group (AUC, 0.861; accuracy, 0.810). The DeLong test revealed the DLRN was noticeably better than clinico-radiological and RS models (training: 0.981 vs. 0.890 vs. 0.751; validation: 0.861 vs. 0.724 vs. 0.700; both P < 0.05); however, no discernible difference in performance was seen between the DLRN and DLRS (training: 0.981 vs. 0.969; validation: 0.861 vs. 0.837; both P > 0.05). The calibration curve analysis and DCA demonstrated that the nomogram exhibited good calibration and offered substantial clinical advantages.

CONCLUSION

The DLRN exhibited strong predictive capability in predicting WDLPS and retroperitoneal lipomas preoperatively, making it a promising imaging biomarker that can facilitate personalized management and precision medicine.

New targeted therapies in liposarcoma: state of art and future perspectives

PURPOSE OF REVIEW

Liposarcomas (LPSs) represent the most common soft tissue sarcoma (STS) subtype, and exhibit distinct clinical molecular features according to histological subgroup. Chemotherapy (ChT), and in particular anthracycline-based schedules, still remains the standard of treatment for all LPS forms. However, given the increasing knowledge gained throughout last years about LPS molecular biology and their genomic profiling, new therapeutic alternatives with targeted drugs are now to be considered. In this review, we will highlight most promising ongoing and published clinical trials regarding targeted therapies in LPSs and provide some insights about future approaches and possible new treatment options for this rare disease.

RECENT FINDINGS

Among all the explored targets, mouse double minute 2 homolog amplification and CKD4-Rb axis inhibition seem to be the most promising target in well differentiated/dedifferentiated LPS subtype. On the other hand, myxoid LPS is known to have a particular sensitivity for trabectedin, which acts like a targeted drug due to its specific action on cellular DNA. In addition to these, multiple other strategies are now being evaluated in LPSs, including the administration of immune-checkpoint inhibitors (ICIs) and 'new-old' cytotoxic agents, such as cabazitaxel, in a continuously growing scenario.

SUMMARY

Although preliminary, results of recently published and ongoing examined clinical trials will hopefully be translated in clinical practice in the next future, leading the way to future research in this rare disease.

Sarculator: how to improve further prognostication of all sarcomas

PURPOSE OF REVIEW

Prognostication of soft tissue sarcomas is challenging due to the diversity of prognostic factors, compounded by the rarity of these tumors. Nomograms are useful predictive tools that assess multiple variables simultaneously, providing estimates of individual likelihoods of specific outcomes at defined time points. Although these models show promising predictive ability, their use underscores the need for further methodological refinement to address gaps in prognosis accuracy.

RECENT FINDINGS

Ongoing efforts focus on improving prognostic tools by either enhancing existing models based on established parameters or integrating novel prognostic markers, such as radiomics, genomic, proteomic, and immunologic factors. Artificial intelligence is a new field that is starting to be explored, as it has the capacity to combine and analyze vast and intricate amounts of relevant data, ranging from multiomics information to real-time patient outcomes.

SUMMARY

The integration of these innovative markers and methods could enhance the prognostic ability of nomograms such as Sarculator and ultimately enable more accurate and individualized healthcare. Currently, clinical variables continue to be the most significant and effective factors in terms of predicting outcomes in patients with STS. This review firstly introduces the rationale for developing and employing nomograms such as Sarculator, secondly, reflects on some of the latest and ongoing methodological refinements, and provides future perspectives in the field of prognostication of sarcomas.

UK guidelines for the management of soft tissue sarcomas

Soft tissue sarcomas (STS) are rare tumours arising in mesenchymal tissues and can occur almost anywhere in the body. Their rarity, and the heterogeneity of subtype and location, means that developing evidence-based guidelines is complicated by the limitations of the data available. This makes it more important that STS are managed by expert multidisciplinary teams, to ensure consistent and optimal treatment, recruitment to clinical trials, and the ongoing accumulation of further data and knowledge. The development of appropriate guidance, by an experienced panel referring to the evidence available, is therefore a useful foundation on which to build progress in the field. These guidelines are an update of the previous versions published in 2010 and 2016 [1, 2]. The original guidelines were drawn up by a panel of UK sarcoma specialists convened under the auspices of the British Sarcoma Group (BSG) and were intended to provide a framework for the multidisciplinary care of patients with soft tissue sarcomas. This iteration of the guidance, as well as updating the general multidisciplinary management of soft tissue sarcoma, includes specific sections relating to the management of sarcomas at defined anatomical sites: gynaecological sarcomas, retroperitoneal sarcomas, breast sarcomas, and skin sarcomas. These are generally managed collaboratively by site specific multidisciplinary teams linked to the regional sarcoma specialist team, as stipulated in the recently published sarcoma service specification [3]. In the UK, any patient with a suspected soft tissue sarcoma should be referred to a specialist regional soft tissues sarcoma service, to be managed by a specialist sarcoma multidisciplinary team. Once the diagnosis has been confirmed using appropriate imaging and a tissue biopsy, the main modality of management is usually surgical excision performed by a specialist surgeon, combined with pre- or post-operative radiotherapy for tumours at higher risk for local recurrence. Systemic anti-cancer therapy (SACT) may be utilised in cases where the histological subtype is considered more sensitive to systemic treatment. Regular follow-up is recommended to assess local control, development of metastatic disease, and any late effects of treatment.

Machine learning analysis of pathological images to predict 1-year progression-free survival of immunotherapy in patients with small-cell lung cancer

BACKGROUND

In small-cell lung cancer (SCLC), the tumor immune microenvironment (TIME) could be a promising biomarker for immunotherapy, but objectively evaluating TIME remains challenging. Hence, we aimed to develop a predictive biomarker of immunotherapy efficacy through a machine learning analysis of the TIME.

METHODS

We conducted a biomarker analysis in a prospective study of patients with extensive-stage SCLC who received chemoimmunotherapy as the first-line treatment. We trained a model to predict 1-year progression-free survival (PFS) using pathological images (H&E, programmed cell death-ligand 1 (PD-L1), and double immunohistochemical assay (cluster of differentiation 8 (CD8) and forkhead box P3 (FoxP3)) and patient information. The primary outcome was the mean area under the curve (AUC) of machine learning models in predicting the 1-year PFS.

RESULTS

We analyzed 100,544 patches of pathological images from 78 patients. The mean AUC values of patient information, pathological image, and combined models were 0.789 (range 0.571-0.982), 0.782 (range 0.750-0.911), and 0.868 (range 0.786-0.929), respectively. The PFS was longer in the high efficacy group than in the low efficacy group in all three models (patient information model, HR 0.468, 95% CI 0.287 to 0.762; pathological image model, HR 0.334, 95% CI 0.117 to 0.628; combined model, HR 0.353, 95% CI 0.195 to 0.637). The machine learning analysis of the TIME had better accuracy than the human count evaluations (AUC of human count, CD8-positive lymphocyte: 0.681, FoxP3-positive lymphocytes: 0.626, PD-L1 score: 0.567).

CONCLUSIONS

The spatial analysis of the TIME using machine learning predicted the immunotherapy efficacy in patients with SCLC, thus supporting its role as an immunotherapy biomarker.

Diagnostic accuracy and safety of percutaneous core needle biopsy of retroperitoneal tumours

BACKGROUND

Histologic subtype of cancer guides treatment sequencing and the extent of surgery for retroperitoneal tumours (RPTs) but concerns persist regarding percutaneous core needle biopsy (CNB).

OBJECTIVE

Endpoints were the incidence of early complications, needle tract seeding (NTS) after CNB, diagnostic accuracy.

METHODS

Between 2015 and 2022, data from patients with RPT who underwent a CNB and who operated on at Institut Curie were collected. We retrospectively reviewed the medical records and microscopic analysis of both CNB and surgical specimens to evaluate the diagnostic accuracy of CNB (quantified using positive and negative predictive values, PPV/NPV).

RESULTS

313 patients underwent CNB. In 10/326 (3 %) procedures, minor complications were observed. One of 212 (0.47 %) resected RPSs exhibited a local recurrence compatible with NTS. Microscopic analysis of CNB specimens allowed the classification of tumours between groups of cancers and benign/intermediate mesenchymal tumours in 307/313 (98 %) patients. Among the 204 patients with retroperitoneal sarcoma, the overall concordance between CNB and final pathology following resection was 178/204 (87.2 %). The respective PPVs of solitary fibrous tumour, dedifferentiated liposarcoma, leiomyosarcoma and well-differentiated liposarcoma were 100 %, 98 %, 97 % and 68 %, respectively. The diagnosis of a high-grade (G 2-3) sarcoma resulted in a high specificity (97 %) and PPV (98 %) but low sensitivity (76 %).

CONCLUSIONS

CNB allowed the classification of RPT in the vast majority of patients with a low morbidity rate. Concordance with final diagnosis was high for sarcomas with the exception of well-differentiated liposarcoma. As a result, CNB results should be integrated with imaging/radiomics by multidisciplinary tumour boards.