1
|
Lovinfosse P, Ferreira M, Withofs N, Jadoul A, Derwael C, Frix AN, Guiot J, Bernard C, Diep AN, Donneau AF, Lejeune M, Bonnet C, Vos W, Meyer PE, Hustinx R. Distinction of Lymphoma from Sarcoidosis on 18F-FDG PET/CT: Evaluation of Radiomics-Feature-Guided Machine Learning Versus Human Reader Performance. J Nucl Med 2022; 63:1933-1940. [PMID: 35589406 PMCID: PMC9730930 DOI: 10.2967/jnumed.121.263598] [Citation(s) in RCA: 13] [Impact Index Per Article: 6.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/26/2021] [Revised: 05/10/2022] [Indexed: 01/11/2023] Open
Abstract
Sarcoidosis and lymphoma often share common features on 18F-FDG PET/CT, such as intense hypermetabolic lesions in lymph nodes and multiple organs. We aimed at developing and validating radiomics signatures to differentiate sarcoidosis from Hodgkin lymphoma (HL) and diffuse large B-cell lymphoma (DLBCL). Methods: We retrospectively collected 420 patients (169 sarcoidosis, 140 HL, and 111 DLBCL) who underwent pretreatment 18F-FDG PET/CT at the University Hospital of Liege. The studies were randomly distributed to 4 physicians, who gave their diagnostic suggestion among the 3 diseases. The individual and pooled performance of the physicians was then calculated. Interobserver variability was evaluated using a sample of 34 studies interpreted by all physicians. Volumes of interest were delineated over the lesions and the liver using MIM software, and 215 radiomics features were extracted using the RadiomiX Toolbox. Models were developed combining clinical data (age, sex, and weight) and radiomics (original and tumor-to-liver TLR radiomics), with 7 different feature selection approaches and 4 different machine-learning (ML) classifiers, to differentiate sarcoidosis and lymphomas on both lesion-based and patient-based approaches. Results: For identifying lymphoma versus sarcoidosis, physicians' pooled sensitivity, specificity, area under the receiver-operating-characteristic curve (AUC), and accuracy were 0.99 (95% CI, 0.97-1.00), 0.75 (95% CI, 0.68-0.81), 0.87 (95% CI, 0.84-0.90), and 89.3%, respectively, whereas for identifying HL in the tumor population, it was 0.58 (95% CI, 0.49-0.66), 0.82 (95% CI, 0.74-0.89), 0.70 (95% CI, 0.64-0.75) and 68.5%, respectively. Moderate agreement was found among observers for the diagnosis of lymphoma versus sarcoidosis and HL versus DLBCL, with Fleiss κ-values of 0.66 (95% CI, 0.45-0.87) and 0.69 (95% CI, 0.45-0.93), respectively. The best ML models for identifying lymphoma versus sarcoidosis showed an AUC of 0.94 (95% CI, 0.93-0.95) and 0.85 (95% CI, 0.82-0.88) in lesion- and patient-based approaches, respectively, using TLR radiomics (plus age for the second). To differentiate HL from DLBCL, we obtained an AUC of 0.95 (95% CI, 0.93-0.96) in the lesion-based approach using TLR radiomics and 0.86 (95% CI, 0.80-0.91) in the patient-based approach using original radiomics and age. Conclusion: Characterization of sarcoidosis and lymphoma lesions is feasible using ML and radiomics, with very good to excellent performance, equivalent to or better than that of physicians, who showed significant interobserver variability in their assessment.
Collapse
Affiliation(s)
- Pierre Lovinfosse
- Division of Nuclear Medicine and Oncological Imaging, CHU of Liège, Liège, Belgium
| | - Marta Ferreira
- GIGA-CRC In Vivo Imaging, University of Liège, Liège, Belgium
| | - Nadia Withofs
- Division of Nuclear Medicine and Oncological Imaging, CHU of Liège, Liège, Belgium
| | - Alexandre Jadoul
- Division of Nuclear Medicine and Oncological Imaging, CHU of Liège, Liège, Belgium
| | - Céline Derwael
- Division of Nuclear Medicine and Oncological Imaging, CHU of Liège, Liège, Belgium
| | - Anne-Noelle Frix
- Department of Respiratory Medicine, CHU of Liège, Liège, Belgium
| | - Julien Guiot
- Department of Respiratory Medicine, CHU of Liège, Liège, Belgium
| | - Claire Bernard
- Division of Nuclear Medicine and Oncological Imaging, CHU of Liège, Liège, Belgium
| | - Anh Nguyet Diep
- Biostatistics Unit, Department of Public Health, University of Liège, Liège, Belgium
| | | | - Marie Lejeune
- Department of Hematology, CHU of Liège, Liège, Belgium
| | | | - Wim Vos
- Radiomics SA, Liège, Belgium; and
| | - Patrick E. Meyer
- Bioinformatics and Systems Biology Lab, University of Liège, Liège, Belgium
| | - Roland Hustinx
- Division of Nuclear Medicine and Oncological Imaging, CHU of Liège, Liège, Belgium
| |
Collapse
|
2
|
Wang W, Mai B, Ali H, Chen L. Uncommon histiocyte-rich pseudotumor after chemotherapy in peripheral T-cell lymphoma. Leuk Lymphoma 2022; 63:2013-2015. [PMID: 35357259 DOI: 10.1080/10428194.2022.2056176] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/18/2022]
Affiliation(s)
- Wei Wang
- Department of Pathology and Laboratory Medicine, The University of Texas Health Science Center at Houston, Houston, TX, USA
| | - Brenda Mai
- Department of Pathology and Laboratory Medicine, The University of Texas Health Science Center at Houston, Houston, TX, USA
| | - Haval Ali
- Department of Pathology and Laboratory Medicine, The University of Texas Health Science Center at Houston, Houston, TX, USA
| | - Lei Chen
- Department of Pathology and Laboratory Medicine, The University of Texas Health Science Center at Houston, Houston, TX, USA
| |
Collapse
|
3
|
Postchemotherapy Histiocyte-rich Pseudotumor Mimicking Residual Lymphoma: A Report of 11 Cases Correlating Clinicopathologic and Radiologic Findings. Am J Surg Pathol 2021; 45:160-168. [PMID: 32769427 DOI: 10.1097/pas.0000000000001547] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/25/2022]
Abstract
Postchemotherapy histiocyte-rich pseudotumor is a rare event in lymphoma patients and can cause elevated metabolic activity on positron emission tomography-computed tomography scan mimicking residual tumor. Here, we reported 11 lymphoma cases showing mass-like lesions with increased fluorodeoxyglucose uptake after chemotherapy. These postchemotherapy lesions occurred in various anatomic sites including spleen, mediastinum, lymph node, and other tissue locations, concerning for refractory or residual lymphoma. Their median size was 2.7 cm (range, 1.4 to 7.7 cm) and the median standardized uptake value on positron emission tomography-computed tomography was 10.6 (range, 5.2 to 13.8). Histologic examination of these lesions demonstrated reactive changes mainly composed of histiocyte-rich proliferation without viable lymphoma. Fat necrosis, cholesterol cleft, and calcium deposit were also commonly observed. After biopsies, 3 patients received additional chemotherapy, 2 had stem cell transplant with adjuvant chemotherapy or radiation, 1 had surgical excision, and the remaining 5 patients did not receive any further treatment. Follow-up imaging studies showed the resolved or decreased fluorodeoxyglucose activities in all patients including those without additional treatments, consistent with benign/reactive nature of these pseudotumor lesions. This study illustrates postchemotherapy mass-like lesions with elevated metabolic activity do not always represent residual disease and provides awareness of correlation between radiologic and histologic features of these lesions to avoid misinterpretation and overtreatment of lymphoma patients after chemotherapy.
Collapse
|