Radiotherapy modification based on artificial intelligence and radiomics applied to (18F)-fluorodeoxyglucose positron emission tomography/computed tomography.

Over the last decades, the refinement of radiation therapy techniques has been associated with an increasing interest for individualized radiation therapy with the aim of increasing or maintaining tumor control and reducing radiation toxicity. Developments in artificial intelligence (AI), particularly machine learning and deep learning, in imaging sciences, including nuclear medecine, have led to significant enthusiasm for the concept of "rapid learning health system". AI combined with radiomics applied to (18F)-fluorodeoxyglucose positron emission tomography/computed tomography ([18F]-FDG PET/CT) offers a unique opportunity for the development of predictive models that can help stratify each patient's risk and guide treatment decisions for optimal outcomes and quality of life of patients treated with radiation therapy. Here we present an overview of the current contribution of AI and radiomics-based machine learning models applied to (18F)-FDG PET/CT in the management of cancer treated by radiation therapy.

Correlation between fluorodeoxyglucose hotspots on preradiotherapy PET/CT and areas of cancer local relapse: Systematic review of literature.

The aim of the present paper is to systematically review all available literature on preradiotherapy high uptake areas (hotspots) as a potential target for dose escalation in different tumour sites, and to understand the potential role and limitations of fluorodeoxyglucose (FDG)-positron-emission tomography (PET)/computed tomography (CT) in this context. An electronic database (Medline) search was conducted to identify articles reporting on a correlation between high tracer uptake on pretreatment PET and preferential sites of local recurrence after radiotherapy. Search was limited to English language. No date range limitation was applied. Among 45 studies initially identified, nine series matching with inclusion criteria have finally been retained from the literature after reviewing (5 retrospective and 4 prospective). Primary tumour locations were head-neck (n=2), lung (n=4), oesophageal (n=2) and rectal (n=1) areas. Overlaps between FDG hotspot on preradiotherapy PET/CT and site of local recurrence on post-treatment scan showed good to excellent agreement. Only studies on head-neck cancer reported moderate agreement probably explained by the lack of reproducibility of the patients positioning between pre- and post-treatment FDG-PET/CT; and by the rigid registration process of images limited by post-therapeutic changes that highly affect anatomical landmarks. FDG hotspot-guided radiotherapy may allow dose escalation in respecting a robust methodology (treatment position, co-registration method, four-dimensional PET).

Relationship between type of unprovoked venous thromboembolism and cancer location: An individual patient data meta-analysis.

BACKGROUND: Unprovoked venous thromboembolism (VTE) may be the first manifestation of an underlying cancer. We aimed to assess the period prevalence of occult cancer detection stratified by VTE location (deep vein thrombosis [DVT], pulmonary embolism [PE] or both) and the anatomical relationship between occult cancer and VTE. METHODS: Post-hoc analysis of a systematic review and individual patient data meta-analysis of adults with unprovoked VTE with at least 12 months of follow-up. Cancer types were grouped according to thoracic, abdomino-pelvic, or other locations. RESULTS: A total of 2300 patients were eligible including 1218 with DVT only (53%), 719 with PE only (31%), and 363 with both PE and DVT (16%). The pooled 12-month period prevalence of cancer in DVT only, PE only, and DVT + PE was 5.6% (95% CI, 4.4 to 7.2), 4.3% (95% CI, 2.7 to 6.9), and 5.6% (95% CI, 1.7 to 15.5), respectively. Most occult cancers were located in the abdomen (68.4%). The proportion of patients with an abdomino-pelvic cancer was not different in patients with DVT + PE (81%; 95% CI, 54 to 96) than in those with DVT (68%; 95% CI, 57 to 78) or PE alone (65%; 95% CI, 48 to 79). CONCLUSION: The 12-month prevalence of occult cancer was similar in patients with DVT only, PE only, or both. Most cancers were located in the abdomen, and there was no relationship between VTE type and cancer location.

Predicting the risk of cancer after unprovoked venous thromboembolism: external validation of the RIETE score.

Essentials Patients at high-risk of occult cancer may benefit from extensive screening. We validated the RIETE cancer score in the MVTEP study. One in three patients were classified as high-risk, 10% of whom had cancer diagnosed. The RIETE score identifies a subgroup at high risk for cancer. SUMMARY: Background Most recent trials evaluating extensive screening strategies for occult cancer in patients with unprovoked venous thromboembolism have failed, because, among other reasons, of an overall low rate of occult cancer. The RIETE investigators recently proposed a score aimed at identifying a subgroup at higher risk. Methods We retrospectively computed the RIETE score for all patients included in the MVTEP study, which evaluated the accuracy of [¹8F]fluorodeoxyglucose-positron emission tomography in the screening of occult cancer in patients with unprovoked venous thromboembolism. Performance of the RIETE score was assessed according to the proportion of patients classified in each risk group, and the corresponding rates of cancer diagnosis. Results Among the 386 patients included in the analysis, 136 patients (35.3%) were classified as high risk by the RIETE score. Cancer was diagnosed in 16 (11.8%) of them, whereas it was diagnosed in nine (3.6%) of the 250 patients with a low RIETE cancer score: odds ratio of 3.6 (95% confidence interval [CI] 1.53-8.32). The area under the receiver operating characteristic curve was 0.63 (95% CI 0.51-0.74). Conclusion The RIETE score seems to be able to identify a subgroup at high risk for cancer (10%) in our specific dataset of patients with unprovoked venous thromboembolism.

Value of (18)F-FDG PET/CT for therapeutic assessment of patients with polymyalgia rheumatica receiving tocilizumab as first-line treatment.

PURPOSE: To evaluate the use of (18)F-FDG PET/CT for the assessment of tocilizumab (TCZ) as first-line treatment in patients with polymyalgia rheumatica (PMR). METHODS: Patients with PMR were prospectively enrolled in a multicentre clinical trial assessing TCZ therapy (the TENOR trial). The patients underwent FDG PET/CT at baseline, after the first infusion of TCZ (TCZ 1) and after the last infusion of TCZ (TCZ 3). Responses to treatment were evaluated in terms of the PMR activity score (PMR-AS), and the C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) laboratory tests. Maximal standardized uptake value (SUVmax) was used for assessment of FDG uptake in regions usually affected in PMR (spinous processes, hips, shoulders, sternoclavicular region and ischial tuberosities). The Wilcoxon test was applied to evaluate the changes in parameters after the infusions and Spearman's rank correlation test was applied to assess the correlations between SUVmax and PMR-AS, CRP and ESR. RESULTS: Of 21 patients included in the trial, 18 were evaluated. The median bioclinical parameter values decreased after TCZ 1 (PMR-AS from 38.2 to 15.7, CRP from 65.2 to 0.4 mg/l and ESR from 49 to 6.5 mm; all p < 0.05) as did the median SUVmax (from 5.8 to 5.2; p < 0.05). All values also decreased after TCZ 3 (PMR-AS from 38.2 to 3.9, CRP from 65.2 to 0.2, ESR from 49 to 2, and SUVmax from 5.8 to 4.7; p < 0.05). In a region-based analysis, all SUVmax were significantly reduced after TCZ 3, except the values for the cervical spinous processes and shoulder regions. With regard to correlations, few significant differences were found between ∆SUVmax and the other parameters including ∆PMR-AS, ∆CRP and ∆ESR in the patient-based and region-based analysis. CONCLUSION: FDG uptake decreased significantly but moderately after TCZ therapy in PMR patients, and might reflect disease activity.

Interpretation of suspect head and neck fixations seen on PET/CT in lung cancer.

OBJECTIVES: Smoking is the major risk factor for lung and head and neck cancer. The purpose of the present study was to determine the clinical impact of serendipitously revealed head and neck fixation on PET/CT in patients undergoing investigation for lung cancer. MATERIAL AND METHODS: The reports from PET/CT studies for patients with lung cancer from September 2005 and April 2012 were retrospectively reviewed. Head and neck incidentaloma was interpreted as suggestive of second primary malignancy. These incidental findings were compared with the final diagnosis obtained from clinical and histological investigation. RESULTS: Five hundred and ninety-two patients were investigated on PET/CT for lung cancer in the study period. PET/CT-positive head and neck lesions suggestive of second primary malignancy were found in 65 (11%) patients. Nasoendoscopy was performed in 23 patients and biopsy in 10. In 4 patients (17.4% of those explored), a second primary malignant lesion was proved on histology: 2 squamous cell carcinomas (larynx and oral cavity), 1 undifferentiated carcinoma (parotid), and 1 osteosarcoma (mandible). At a median 13 months' follow-up, 3 of the 4 patients with a second primary had died from disease-related causes and 1 was free of recurrence. Metastases from lung adenocarcinoma were found in 2 patients (0.34%). CONCLUSIONS: PET/CT detected incidental head and neck malignant tumors in at least 0.68% of lung cancer patients, but in 6.4% of those with suspect head and neck fixation.

Initial staging of squamous cell carcinoma of the oral cavity, larynx and pharynx (excluding nasopharynx). Part 2: Remote extension assessment and exploration for secondary synchronous locations outside of the upper aerodigestive tract. 2012 SFORL guidelines.

OBJECTIVES: This report presents the French Society of ORL (SFORL) guidelines for exploration for remote metastasis and synchronous second cancer in initial staging of head and neck squamous cell carcinoma. MATERIALS AND METHODS: An exhaustive literature review was analyzed by a multidisciplinary work-group. RESULTS: The thorax is the most frequent location of remote metastases and synchronous second cancer outside of the upper aerodigestive tract. Thoracic CT is recommended as first-line examination in all cases (grade B). 18-FDG PET/CT is recommended when the thoracic CT image is doubtful or in case of high metastatic risk (grade B), for the detection of non-pulmonary remote metastasis. Esophageal exploration is recommended in case of significant risk of synchronous esophageal cancer (hypopharyngeal or oropharyngeal tumor, chronic alcohol intoxication) (grade B). The reference examination is flexible endoscopy of the upper digestive tract (grade B). CONCLUSION: The present grade B recommendations rationalize the roles of the various first-line radiological and endoscopic examinations for remote metastasis and synchronous second cancer, so as to limit the number of examinations performed, thereby reducing the time needed for initial staging.

Initial staging of squamous cell carcinoma of the oral cavity, larynx and pharynx (excluding nasopharynx). Part I: Locoregional extension assessment: 2012 SFORL guidelines.

OBJECTIVES: To set out good practice guidelines for locoregional extension assessment of squamous cell carcinoma of the head and neck (excluding nasopharynx, nasal cavities and sinuses). MATERIALS AND METHODS: A critical multidisciplinary review of the literature on locoregional extension assessment of squamous cell carcinoma of the head and neck was conducted, applying levels of evidence in line with the French health authority's (HAS) literature analysis guide of January 2000. CONCLUSION: Based on the levels of evidence of the selected articles and on work-group consensus, graded guidelines are set out for clinical, endoscopic and imaging locoregional extension assessment of head and neck cancer.