Purpose
Lymphoma lesion detection and segmentation on whole-body FDG-PET/CT are a challenging task because of the diversity of involved nodes, organs or physiological uptakes. We sought to ...investigate the performances of a three-dimensional (3D) convolutional neural network (CNN) to automatically segment total metabolic tumour volume (TMTV) in large datasets of patients with diffuse large B cell lymphoma (DLBCL).
Methods
The dataset contained pre-therapy FDG-PET/CT from 733 DLBCL patients of 2 prospective LYmphoma Study Association (LYSA
)
trials. The first cohort (
n
= 639) was used for training using a 5-fold cross validation scheme. The second cohort (
n
= 94) was used for external validation of TMTV predictions. Ground truth masks were manually obtained after a 41% SUVmax adaptive thresholding of lymphoma lesions. A 3D U-net architecture with 2 input channels for PET and CT was trained on patches randomly sampled within PET/CTs with a summed cross entropy and Dice similarity coefficient (DSC) loss. Segmentation performance was assessed by the DSC and Jaccard coefficients. Finally, TMTV predictions were validated on the second independent cohort.
Results
Mean DSC and Jaccard coefficients (± standard deviation) in the validations set were 0.73 ± 0.20 and 0.68 ± 0.21, respectively. An underestimation of mean TMTV by − 12 mL (2.8%) ± 263 was found in the validation sets of the first cohort (
P
= 0.27). In the second cohort, an underestimation of mean TMTV by − 116 mL (20.8%) ± 425 was statistically significant (
P
= 0.01).
Conclusion
Our CNN is a promising tool for automatic detection and segmentation of lymphoma lesions, despite slight underestimation of TMTV. The fully automatic and open-source features of this CNN will allow to increase both dissemination in routine practice and reproducibility of TMTV assessment in lymphoma patients.
This review considers the potential utility of positron emission tomography (PET) tracers in the setting of response monitoring in breast cancer, with a special emphasis on glucose metabolic changes ...assessed with 18F‐fluorodeoxyglucose (FDG). In the neoadjuvant setting of breast cancer, the metabolic response can predict the final complete pathologic response after the first cycles of chemotherapy. Because tumor metabolic behavior highly depends on cancer subtype, studies are ongoing to define the optimal metabolic criteria of tumor response in each subtype. The recent multicentric randomized AVATAXHER trial has suggested, in the human epidermal growth factor 2‐positive subtype, a clinical benefit of early tailoring the neoadjuvant treatment in women with poor metabolic response after the first course of treatment. In the bone‐dominant metastatic setting, there is increasing clinical evidence that FDG‐PET/computed tomography (CT) is the most accurate imaging modality for assessment of the tumor response to treatment when both metabolic information and morphologic information are considered. Nevertheless, there is a need to define standardized metabolic criteria of response, including the heterogeneity of response among metastases, and to evaluate the costs and health outcome of FDG‐PET/CT compared with conventional imaging. New non‐FDG radiotracers highlighting specific molecular hallmarks of breast cancer cells have recently emerged in preclinical and clinical studies. These biomarkers can take into account the heterogeneity of tumor biology in metastatic lesions. They may provide valuable clinical information for physicians to select and monitor the effectiveness of novel therapeutics targeting the same molecular pathways of breast tumor.
This review considers the potential utility of positron emission tomography (PET) tracers in the setting of response monitoring in breast cancer, with a special emphasis on glucose metabolic changes assessed with 18F‐fluorodeoxyglucose. It discusses the different PET timing and metabolic criteria to define response that have been evaluated in previous studies. The development of new radiotracers of specific molecular pathways of breast cancer cells is also a challenging and promising research area.
To investigate the respective influence of software tool and total metabolic tumor volume (TMTV0) calculation method on prognostic stratification of baseline 2-deoxy-2-18Ffluoro-D-glucose positron ...emission tomography (18FFDG-PET) in newly diagnosed Hodgkin lymphoma (HL).
59 patients with newly diagnosed HL were retrospectively included. 18FFDG-PET was performed before any treatment. Four sets of TMTV0 were calculated with Beth Israel (BI) software: based on an absolute threshold selecting voxel with standardized uptake value (SUV) >2.5 (TMTV02.5), applying a per-lesion threshold of 41% of the SUV max (TMTV041) and using a per-patient adapted threshold based on SUV max of the liver (>125% and >140% of SUV max of the liver background; TMTV0125 and TMTV0140). TMTV041 was also determined with commercial software for comparison of software tools. ROC curves were used to determine the optimal threshold for each TMTV0 to predict treatment failure.
Median follow-up was 39 months. There was an excellent correlation between TMTV041 determined with BI and with the commercial software (r = 0.96, p<0.0001). The median TMTV0 value for TMTV041, TMTV02.5, TMTV0125 and TMTV0140 were respectively 160 (used as reference), 210 (28;154 p = 0.005), 183 (-4;114 p = 0.06) and 143 ml (-58;64 p = 0.9). The respective optimal TMTV0 threshold and area under curve (AUC) for prediction of progression free survival (PFS) were respectively: 313 ml and 0.70, 432 ml and 0.68, 450 ml and 0.68, 330 ml and 0.68. There was no significant difference between ROC curves. High TMTV0 value was predictive of poor PFS in all methodologies: 4-years PFS was 83% vs 42% (p = 0.006) for TMTV02.5, 83% vs 41% (p = 0.003) for TMTV041, 85% vs 40% (p<0.001) for TMTV0125 and 83% vs 42% (p = 0.004) for TMTV0140.
In newly diagnosed HL, baseline metabolic tumor volume values were significantly influenced by the choice of the method used for determination of volume. However, no significant differences were found in term of prognosis.
Summary Background An effective and well tolerated treatment is needed for patients with early HER2-positive breast cancer who do not achieve a pathological complete response after neoadjuvant ...therapy. The AVATAXHER trial aimed to predict pathological complete response early with the use of PET and to investigate whether the addition of bevacizumab could improve the proportion of patients achieving a pathological complete response in patients unlikely to respond to treatment. Methods AVATAXHER was a randomised, open-label, non-comparative, multicentre phase 2 study that enrolled women (≥18 years of age) with early-stage HER2-positive breast cancer from 26 oncology centres in France. Patients initially received two cycles of neoadjuvant docetaxel (100 mg/m2 intravenously every 3 weeks) plus trastuzumab (8 mg/kg intravenously every 3 weeks then 6 mg/kg intravenously every 3 weeks for the second course). Before the first and second cycles, 18 F-fluorodeoxyglucose (FDG) PET was done and the change in standardised uptake value was used to predict pathological complete response in each patient. Patients who were predicted to be responders on PET continued to receive standard therapy. Predicted non-responders were randomly assigned (2:1) to receive four cycles of docetaxel (100 mg/m2 intravenously every 3 weeks) and trastuzumab (6 mg/kg intravenously every 3 weeks) plus bevacizumab (15 mg/kg intravenously every 3 weeks; group A) or continue on docetaxel plus trastuzumab alone (group B). Randomisation was open label and was done by an adaptive minimisation method. Although investigators and patients were aware of group assignment, the anatomo-pathologist in charge of centralised review of surgical samples and lymph nodes was masked to treatment assignment. The primary endpoint was centrally assessed pathological complete response according to the Chevallier classification. Efficacy analyses were done in the intention-to-treat population. Safety analyses in this Article were done on all patients who received at least one dose of treatment starting from cycle 3. Survival outcomes are not yet mature. This study is registered with ClinicalTrials.gov ( NCT01142778 ) and EUDRACT (2009-013410-26). Findings Between May 19, 2010, and Oct 1, 2012, 152 patients were recruited for the study. Ten patients were subsequently excluded, leaving 142 patients in the intention-to-treat population. Of these 142 patients, 69 were predicted by 18 F-FDG PET to be treatment responders after two cycles of treatment. The 73 predicted non-responders were randomly assigned to group A (n=48) and group B (n=25). Pathological complete responses were noted in 37 (53·6%, 95% CI 41·2–65·7) of the PET responders, 21 (43·8%, 29·5–58·8) of those in group A, and six (24·0%, 9·4–45·1) of those in group B. Incidences of grade 3–4 adverse events were similar in all three groups. The most common grade 3–4 adverse events were neutropenia (four in PET responders, five in group A, and three in group B), febrile neutropenia (one, three, and one, respectively), and myalgia (four, none, and one, respectively). Overall, 24 serious adverse events were reported in 15 patients (PET responders: nine events in four 6% of 67 patients; group A: 14 events in ten 21% of 47 patients; group B: one event in one 4% of 25 patients). No deaths occurred during the study. Interpretation In patients with HER2-positive breast cancer, early PET assessment can help to identify non-responders to neoadjuvant docetaxel plus trastuzumab therapy. In these patients, the addition of bevacizumab can increase the proportion of patients achieving a pathological complete response. This potential new role for PET and the activity of bevacizumab in this setting need to be confirmed in larger phase 3 trials. Funding Roche France.
The prognostic value of interim positron emission tomography (PET) interpreted according to visual criteria is a matter of debate in diffuse large B-cell lymphoma (DLBCL). Maximal standardized uptake ...value reduction (ΔSUVmax) may better predict outcome. To compare the prognostic value of both methods, we analyzed PET done at baseline (PET0) and after 2 (PET2) and 4 (PET4) cycles in 85 patients with high-risk DLBCL enrolled on a prospective multicenter trial. All images were centrally reviewed and interpreted visually according to the International Harmonization Project criteria and by computing ΔSUVmax between PET0 and PET2 (ΔSUVmaxPET0-2) or PET4 (ΔSUVmaxPET0-4). Optimal cutoff to predict progression or death was 66% for ΔSUVmaxPET0-2 and 70% for ΔSUVmaxPET0-4. Outcomes did not differ significantly whether PET2 and PET4 were visually positive or negative. Inversely, ΔSUVmaxPET0-2 analysis (> 66% vs ≤ 66%) identified patients with significantly different 2-year progression-free survival (77% vs 57%; P = .0282) and overall survival (93% vs 60%; P < .0001). ΔSUVmaxPET0-4 analysis (> 70% vs ≤ 70%) seemed even more predictive for 2-year progression-free survival (83 vs 40%; P < .0001) and overall survival (94% vs 50%; P < .0001). ΔSUVmax analysis of sequential interim PET is feasible for high-risk DLBCL and better predicts outcome than visual analysis. The trial was registered at http://clinicaltrials.gov as NCT00498043.
Both total metabolic tumor volume (TMTV), computed on baseline positron emission tomography (PET), and end of induction (EOI) PET are imaging biomarkers showing promise for early risk stratification ...in patients with high-tumor-burden follicular lymphoma. A model was built incorporating these 2 factors in 159 patients from three prospective trials: 2 Lymphoma Study Association (LYSA) studies and 1 Fondazione Italiana Linfomi (FIL) trial. Median follow up was 64 months. High TMTV (>510 cm3) and positive EOI PET were independent, significant risk factors for progression. Their combination stratified the population into 3 risk groups: patients with no risk factors (n = 102; 64%) had a 5-year progression-free survival (PFS) of 67% vs 33% (hazard ratio HR, 2.9; 95% confidence interval CI, 1.8-4.9) for patients with 1 risk factor (n = 44; 27%) and only 23% (HR, 4.6; 95% CI, 2.3-9.2) for patients with both risk factors (n = 13; 8%). 2-year PFS was respectively 90% vs 61% (HR, 4.8; 95% CI, 2.2-10.4) and 46% (HR, 8.1; 95%CI, 3.1-21.3). This model enhances the prognostic value of PET staging and response assessment, identifying a subset of patients with a very high risk of progression and early treatment failure at 2 years.
•A model combining baseline metabolic tumor volume and EOI PET identify follicular lymphoma patients with a very high risk of progression.
Display omitted
(18)FFluorodeoxyglucose positron emission tomography (PET) is widely used for the staging and restaging of patients with aggressive lymphoma, but less is known about the utility of PET in patients ...with follicular lymphoma (FL). In a prospective study, we evaluated the prognostic value of PET performed during treatment and at the end of treatment in 121 patients with FL treated with first-line immunochemotherapy.
Patients with previously untreated high-tumor burden FL were treated with six cycles of R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) plus two cycles of rituximab, without rituximab maintenance. PET was performed before treatment, after four cycles of R-CHOP (interim PET), and at the end of treatment (final PET). PET scans were centrally reviewed.
The total number of patients included was 121. Median age was 57 years. After central review, interim PET (n = 111) was negative in 76% of patients, and final PET (n = 106) was negative in 78%. With a median follow-up of 23 months, 2-year progression-free survival rates were 86% for interim PET-negative versus 61% for interim PET-positive patients (P = .0046) and 87% for final PET-negative versus 51% for final PET-positive patients (P < .001), respectively. Two-year overall survival also significantly differed according to final PET results: 100% versus 88% (P = .0128).
PET performed either after four cycles of R-CHOP or at the end of therapy was strongly predictive of outcome in this prospective study. Therapeutic intervention based on PET results during or after inductive treatment should be evaluated.
Evaluate response and predict prognosis of patients with newly diagnosed metastatic breast cancer treated with first line systemic therapy using European Organization for Research and Treatment of ...Cancer (EORTC) criteria and PET Response Criteria in solid Tumours (PERCIST).
From December 2006 to August 2013, 57 women with newly diagnosed metastatic breast cancer were retrospectively evaluated. FDG-PET/CT was performed within one month before treatment and repeated after at least 3 cycles of treatment. Metabolic response evaluation was evaluated by two readers according to both EORTC criteria and PERCIST, classifying the patients into 4 response groups: complete metabolic response (CMR), partial metabolic response (PMR), stable metabolic disease (SMD), and progressive metabolic disease (PMD).
With EORTC criteria, 22 patients had CMR, 17 PMR, 6 SMD and 12 PMD. With PERCIST, 20 patients had CMR, 15 PMR, 10 SMD and 12 PMD. There was agreement between EORTC and PERCIST in 84% of the patients. By log-rank analysis, metabolic response evaluated with both EORTC criteria and PERCIST was able to predict overall survival (p = 0.028 and 0.002 respectively). CMR patient group had longer median OS than patients in the combined PMR+SMD+PMD group (60 vs 26 months both with EORTC and PERCIST; p = 0.009 and 0.006 respectively). By multivariate analysis, CMR either with EORTC or PERCIST remained an independent predictor of survival.
Metabolic response evaluation with EORTC criteria and PERCIST gave similar prognostic stratification for metastatic breast cancer treated with a first line of systemic therapy.
Outcomes for follicular lymphoma (FL) have greatly improved, but most patients will ultimately relapse. High total metabolic tumor volume (TMTV), computed from baseline ...18F-fluorodeoxyglucose–positron emission tomography (PET), is associated with shorter progression-free survival (PFS), but circulating tumor cells (CTCs) and cell-free DNA (cfDNA) may also reflect tumor burden and be of prognostic value. The aim of our study was to correlate CTCs and cfDNA with TMTV in FL at diagnosis and to determine their prognostic values. We retrospectively analyzed 133 patients (with previously untreated FL and a baseline PET) from 2 cohorts with either a baseline plasma sample (n = 61) or a bcl2-JH-informative peripheral blood (PB) sample (n = 68). Quantification of circulating bcl2-JH+ cells and cfDNA was performed by droplet digital polymerase chain reaction. A significant correlation was found between TMTV and both CTCs (P < .0001) and cfDNA (P < .0001). With a median 48-month follow-up, 4-year PFS was lower in patients with TMTV > 510 cm3 (P = .0004), CTCs >0.0018 PB cells (P = .03), or cfDNA >2550 equivalent-genome/mL (P = .04). In comparison with TMTV alone, no additional prognostic information was obtained by measuring CTCs. In contrast, Cox multivariate analysis, including cfDNA and TMTV, showed that both cfDNA and TMTV remained predictive of outcome. In conclusion, CTCs and cfDNA correlate with TMTV in FL, and all 3 influence patient outcome. PFS was shorter for patients with high cfDNA and TMTV, suggesting that these parameters provide relevant information for tumor-tailored therapy.
•Cell-free DNA and circulating tumor cells are related to TMTV in follicular lymphoma at diagnosis.•Cell-free DNA levels, in addition to TMTV, define high-risk follicular lymphoma patients.
Display omitted