Total metabolic tumor volume on 18F-FDG PET/CT is a game-changer for patients with metastatic lung cancer treated with immunotherapy

PURPOSE

Because of atypical response imaging patterns in patients with metastatic non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICPIs), new biomarkers are needed for a better monitoring of treatment efficacy. The aim of this prospective study was to evaluate the prognostic value of volume-derived positron-emission tomography (PET) parameters on baseline and follow-up 18F-fluoro-deoxy-glucose PET (18F-FDG-PET) scans and compare it with the conventional PET Response Criteria in Solid Tumors (PERCIST).

METHODS

Patients with metastatic NSCLC were included in two different single-center prospective trials. 18F-FDG-PET studies were performed before the start of immunotherapy (PETbaseline), after 6-8 weeks (PETinterim1) and after 12-16 weeks (PETinterim2) of treatment, using PERCIST criteria for tumor response assessment. Different metabolic parameters were evaluated: absolute values of maximum standardized uptake value (SUVmax) of the most intense lesion, total metabolic tumor volume (TMTV), total lesion glycolysis (TLG), but also their percentage changes between PET studies (ΔSUVmax, ΔTMTV and ΔTLG). The median follow-up of patients was 31 (7.3-31.8) months. Prognostic values and optimal thresholds of PET parameters were estimated by ROC (Receiver Operating Characteristic) curve analysis of 12-month overall survival (12M-OS) and 6-month progression-free survival (6M-PFS). Tumor progression needed to be confirmed by a multidisciplinary tumor board, considering atypical response patterns on imaging.

RESULTS

110 patients were prospectively included. On PETbaseline, TMTV was predictive of 12M-OS [AUC (Area Under Curve) =0.64; 95% CI: 0.61 to 0.66] whereas SUVmax and TLG were not. On PETinterim1 and PETinterim2, all metabolic parameters were predictive for 12M-OS and 6M-PFS, the residual TMTV on PETinterim1 (TMTV1) being the strongest prognostic biomarker (AUC=0.83 and 0.82; 95% CI: 0.74 to 0.91, for 12M-OS and 6M-PFS, respectively). Using the optimal threshold by ROC curve to classify patients into three TMTV1 subgroups (0 cm3; 0-57 cm3; >57 cm3), TMTV1 prognostic stratification was independent of PERCIST criteria on both PFS and OS, and significantly outperformed them. Subgroup analysis demonstrated that TMTV1 remained a strong prognostic biomarker of 12M-OS for non-responding patients (p=0.0003) according to PERCIST criteria. In the specific group of patients with PERCIST progression on PETinterim1, low residual tumor volume (<57 cm3) was still associated with a very favorable patients' outcome (6M-PFS=73%; 24M-OS=55%).

CONCLUSION

The absolute value of residual metabolic tumor volume, assessed 6-8 weeks after the start of ICPI, is an optimal and independent prognostic measure, exceeding and complementing conventional PERCIST criteria. Oncologists should consider it in patients with first tumor progression according to PERCIST criteria, as it helps identify patients who benefit from continued treatment.

TRIAL REGISTRATION NUMBER

2018-A02116-49; NCT03584334.

[18F]-FDopa positron emission tomography imaging in corticobasal syndrome

PURPOSE

First, to investigate the patterns of [18F]-FDOPA positron emission tomography imaging in corticobasal syndrome using visual and semi-quantitative analysis and to compare them with patterns found in Parkinson's disease and progressive supranuclear palsy. Then, to search for correlations with clinical features and [18F]-FDG positron emission tomography imaging.

METHODS

27 corticobasal syndrome patients who underwent [18F]-FDOPA positron emission tomography imaging were retrospectively studied. They were compared to 27 matched Parkinson's disease patients, 12 progressive supranuclear palsy patients and 53 normal controls. Scans were visually assigned to one of the following patterns: normal; unilateral homogeneous striatal uptake reduction; putamen uptake reduction with putamen-caudate gradient. A semi-quantitative analysis of striatal regional uptake and asymmetry was performed and correlated to clinical features and [18F]-FDG positron emission tomography patterns.

RESULTS

[18F]-FDOPA positron emission tomography appeared visually abnormal in only 33.5% of corticobasal syndrome patients. However, semi-quantitative analysis found putaminal asymmetry in 63%. Striatal uptake was homogeneously reduced in both putamen and caudate nucleus in corticobasal syndrome patients unlike in Parkinson's disease and progressive supranuclear palsy. No correlation was found between [18F]-FDOPA positron emission tomography and clinical features. Half of corticobasal syndrome patients presented a corticobasal degeneration pattern on [18F]-FDG positron emission tomography.  CONCLUSION: [18F]-FDOPA positron emission tomography can often be normal in corticobasal syndrome patients. Semi-quantitative analysis is useful to unmask a significant asymmetry in many of them. Homogeneous striatal uptake reduction contralateral to the clinical signs is highly suggestive of corticobasal syndrome. This finding can be helpful to better characterize this syndrome with respect to Parkinson's disease and progressive supranuclear palsy.

Identification of Metabolomic Markers in Frozen or Formalin-Fixed and Paraffin-Embedded Samples of Diffuse Glioma from Adults

The aim of this study was to identify metabolomic signatures associated with the gliomagenesis pathway (IDH-mutant or IDH-wt) and tumor grade of diffuse gliomas (DGs) according to the 2021 WHO classification on frozen samples and to evaluate the diagnostic performances of these signatures in tumor samples that are formalin-fixed and paraffin-embedded (FFPE). An untargeted metabolomic study was performed using liquid chromatography/mass spectrometry on a cohort of 213 DG samples. Logistic regression with LASSO penalization was used on the frozen samples to build classification models in order to identify IDH-mutant vs. IDH-wildtype DG and high-grade vs low-grade DG samples. 2-Hydroxyglutarate (2HG) was a metabolite of interest to predict IDH mutational status and aminoadipic acid (AAA) and guanidinoacetic acid (GAA) were significantly associated with grade. The diagnostic performances of the models were 82.6% AUC, 70.6% sensitivity and 80.4% specificity for 2HG to predict IDH status and 84.7% AUC, 78.1% sensitivity and 73.4% specificity for AAA and GAA to predict grade from FFPE samples. Thus, this study showed that AAA and GAA are two novel metabolites of interest in DG and that metabolomic data can be useful in the classification of DG, both in frozen and FFPE samples.

Prognostic impact of post-treatment FDG PET/CT in anal canal cancer: A prospective study

BACKGROUND AND PURPOSE

The aim of our prospective study was to assess the prognostic value of 18F-FDG PET/CT performed two months post treatment for anal canal neoplasm.

POPULATION AND METHODS

Consecutive patients with histologically proved anal cancer, with 18F-FDG PET/CT pre and two months post treatment were included. Patients were not previously treated for this neoplasm and then received radiotherapy ± chemotherapy. Clinical and pathologic data were collected and for 18F-FDG PET/CT visual and quantitative analysis (standardized uptake value, metabolic volume) were performed; response was classified according to EORTC and PERCIST criteria. The results were assessed for disease free survival and local recurrence free survival using the log-Rank test RESULTS: From December 2014 to September 2019, 94 consecutive patients were screened and 78 were included in this study. Median follow-up was 51 months. Two months post treatment, 37 patients (47.4%) had a complete radiological response according to both EORTC and PERCIST criteria, 66 patients (84.6%) had a clinical complete response. For disease free survival, the prognostic value of complete response was statistically significant (p=0.02) with 18F-FDG PET/CT and with clinical examination (p<0.001). For local recurrence free survival, the prognostic value with 18F-FDG PET/CT was lower (p=0.04) than clinical examination (p < 0.007).

CONCLUSION

While clinical examination remains the gold standard for post treatment evaluation in anal cancer, 18F-FDG PET/CT has a statistically significant prognostic value. These two assessments could be combined to improve early evaluation.

Survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data

Purpose

Meta-analyses failed to accurately identify patients with non-metastatic breast cancer who are likely to benefit from chemotherapy, and metabolomics could provide new answers. In our previous published work, patients were clustered using five different unsupervised machine learning (ML) methods resulting in the identification of three clusters with distinct clinical and simulated survival data. The objective of this study was to evaluate the survival outcomes, with extended follow-up, using the same 5 different methods of unsupervised machine learning.

Experimental design

Forty-nine patients, diagnosed between 2013 and 2016, with non-metastatic BC were included retrospectively. Median follow-up was extended to 85.8 months. 449 metabolites were extracted from tumor resection samples by combined Liquid chromatography-mass spectrometry (LC-MS). Survival analyses were reported grouping together Cluster 1 and 2 versus cluster 3. Bootstrap optimization was applied.

Results

PCA k-means, K-sparse and Spectral clustering were the most effective methods to predict 2-year progression-free survival with bootstrap optimization (PFSb); as bootstrap example, with PCA k-means method, PFSb were 94% for cluster 1&2 versus 82% for cluster 3 (p = 0.01). PCA k-means method performed best, with higher reproducibility (mean HR=2 (95%CI [1.4-2.7]); probability of p ≤ 0.05 85%). Cancer-specific survival (CSS) and overall survival (OS) analyses highlighted a discrepancy between the 5 ML unsupervised methods.

Conclusion

Our study is a proof-of-principle that it is possible to use unsupervised ML methods on metabolomic data to predict PFS survival outcomes, with the best performance for PCA k-means. A larger population study is needed to draw conclusions from CSS and OS analyses.

Added value of [18F]FDOPA PET to the management of high-grade glioma patients after their initial treatment: a prospective multicentre study

BACKGROUND

Diagnostic value of 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine ([18F]FDOPA) PET in patients with suspected recurrent gliomas is recognised. We conducted a multicentre prospective study to assess its added value in the practical management of patients suspected of recurrence of high grade gliomas (HGG).

METHODS

Patients with a proven HGG (WHO grade III and IV) were referred to the multidisciplinary neuro-oncology board (MNOB) during their follow-up after initial standard of care treatment and when MRI findings were not fully conclusive. Each case was discussed in 2 steps. For step 1, a diagnosis and a management proposal were made only based on the clinical and the MRI data. For step 2, the same process was repeated taking the [18F]FDOPA PET results into consideration. A level of confidence for the decisions was assigned to each step. Changes in diagnosis and management induced by [18F]FDOPA PET information were measured. When unchanged, the difference in the confidence of the decisions were assessed. The diagnostic performances of each step were measured.

RESULTS

107 patients underwent a total of 138 MNOB assessments. The proposed diagnosis changed between step 1 and step 2 in 37 cases (26.8%) and the proposed management changed in 31 cases (22.5%). When the management did not change, the confidence in the MNOB final decision was increased in 87 cases (81.3%). Step 1 had a sensitivity, specificity and accuracy of 83%, 58% and 66% and step 2, 86%, 64% and 71% respectively.

CONCLUSION

[18F]FDOPA PET adds significant information for the follow-up of HGG patients in clinical practice. When MRI findings are not straightforward, it can change the management for more than 20% of the patients and increases the confidence level of the multidisciplinary board decisions.

Metabolomic Signatures of Scarff-Bloom-Richardson (SBR) Grade in Non-Metastatic Breast Cancer

PURPOSE

Identification of metabolomic biomarkers of high SBR grade in non-metastatic breast cancer.

METHODS

This retrospective bicentric metabolomic analysis included a training set (n = 51) and a validation set (n = 49) of breast cancer tumors, all classified as high-grade (grade III) or low-grade (grade I-II). Metabolomes of tissue samples were studied by liquid chromatography coupled with mass spectrometry.

RESULTS

A molecular signature of the top 12 metabolites was identified from a database of 602 frequently predicted metabolites. Partial least squares discriminant analyses showed that accuracies were 0.81 and 0.82, the R2 scores were 0.57 and 0.55, and the Q2 scores were 0.44431 and 0.40147 for the training set and validation set, respectively; areas under the curve for the Receiver Operating Characteristic Curve were 0.882 and 0.886. The most relevant metabolite was diacetylspermine. Metabolite set enrichment analyses and metabolic pathway analyses highlighted the tryptophan metabolism pathway, but the concentration of individual metabolites varied between tumor samples.

CONCLUSIONS

This study indicates that high-grade invasive tumors are related to diacetylspermine and tryptophan metabolism, both involved in the inhibition of the immune response. Targeting these pathways could restore anti-tumor immunity and have a synergistic effect with immunotherapy. Recent studies could not demonstrate the effectiveness of this strategy, but the use of theragnostic metabolomic signatures should allow better selection of patients.

Abstract P2-26-23: Transcriptomic modification induced by the first cycle of neoadjuvant chemotherapy impacts response to treatment in triple negative breast cancer (TNBC)

Background Despite adequate neoadjuvant chemotherapy, TNBC remains poor prognosis. Non-responding patient have 25-40% risk of relapse at 5 years. pCR is therefore currently considered as a major goal in TNBC and new tools of early prediction of residual disease should be identified. TransTep is a phase 2 monocentric clinical trial which aims to identify transcriptomic profile of triple-negative cancer cells associated with early tumor chemoresistance, as identified by FDG PET after the first course of neoadjuvant chemotherapy. Twenty patients were included between January 2015 and October 2017, with stage II or III of the UICC classification (except stage T4d). All patients received neoadjuvant chemotherapy with anthracyclines and taxanes sequentially, none of them had a dose-dense chemotherapy. Six patients obtained metabolic response (Delta SUV &lt; -30%) after one cycle of anthracycline. Methods Total RNA was extracted from biopsies (HES &gt; 30%) and used to prepare libraries thanks to TruSeq RNA library Prep kit (Illumina) and sequenced on NextSeq500 device. Results Transcriptomic expression analysis between tumors before chemotherapy and tumors after one course of chemotherapy showed that tumors showing a significant decrease of their Delta SUV (←30%) had a much higher gene expression variation than tumors with a Delta SUV &gt;-30%. These tumors presented, after one course of chemotherapy, a decrease of cell cycle, DNA replication and Fanconi anemia pathway related genes, while they harbored an increase of genes belonging to immunity related pathways such as natural killer cell mediated toxicity, TH17 cell differentiation, or chemokine signaling pathway. As patients had a surgery after neoadjuvant chemotherapy, we had Chevalier tumor response. As observed with FDG PET data, patients with a good response according to Chevalier (class 1) criteria presented a much higher transcriptomic modification after one course of chemotherapy. Patients with a little gene expression modification were classified as Chevalier 3. When we focused on pathways impacted by chemotherapy in tumors classified as Chevalier 1, we observed a significant decrease of cell cycle, DNA replication and DNA repair pathways related gene after one course of chemotherapy, whereas a high number of genes belonging to immunity related genes (NK cells, antigen processing and presentation, or chemokine signaling pathway) were increased. Conclusion Our results tend to indicate that transcriptomic tumor response after one course of chemotherapy could forecast final response to treatment. Moreover, it seems that one cycle of anthracycline-based chemotherapy is able to get hot some breast tumors, and that this tumor warming could be a marker of good response to the full treatment. In the following months, we will test whether transcriptomic could predict progression free survival and/or overall survival of patients. We will also look for the best early marker of response between PET FDG and tumor transcriptomic modifications. Based on these results, a new trial will be organized. A therapeutic adaptation depending on the transcriptomic or metabolic data will be proposed to increase pCR rate and prognosis of patients. Moreover, with years and therapeutic innovations, patient managment obviously evolves. This was recently the case with a new standard of care in triple negative breast cancer with combination of platinum salt and immunotherapy. Therefore, a new trial, similar to TRANSTEP, will be carried out with this new standard of care.

Citation Format: Isabelle Desmoulins, David Chardin, Corentin Richard, Laurent Arnould, ALEXANDRE COCHET, Olivier Humbert, Romain Boidot. Transcriptomic modification induced by the first cycle of neoadjuvant chemotherapy impacts response to treatment in triple negative breast cancer (TNBC) [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P2-26-23.

Learning a confidence score and the latent space of a new supervised autoencoder for diagnosis and prognosis in clinical metabolomic studies

Background

Presently, there is a wide variety of classification methods and deep neural network approaches in bioinformatics. Deep neural networks have proven their effectiveness for classification tasks, and have outperformed classical methods, but they suffer from a lack of interpretability. Therefore, these innovative methods are not appropriate for decision support systems in healthcare. Indeed, to allow clinicians to make informed and well thought out decisions, the algorithm should provide the main pieces of information used to compute the predicted diagnosis and/or prognosis, as well as a confidence score for this prediction.

Methods

Herein, we used a new supervised autoencoder (SAE) approach for classification of clinical metabolomic data. This new method has the advantage of providing a confidence score for each prediction thanks to a softmax classifier and a meaningful latent space visualization and to include a new efficient feature selection method, with a structured constraint, which allows for biologically interpretable results.

Results

Experimental results on three metabolomics datasets of clinical samples illustrate the effectiveness of our SAE and its confidence score. The supervised autoencoder provides an accurate localization of the patients in the latent space, and an efficient confidence score. Experiments show that the SAE outperforms classical methods (PLS-DA, Random Forests, SVM, and neural networks (NN)). Furthermore, the metabolites selected by the SAE were found to be biologically relevant.

Conclusion

In this paper, we describe a new efficient SAE method to support diagnostic or prognostic evaluation based on metabolomics analyses.

Development and validation of a radiomic model for the diagnosis of dopaminergic denervation on [18F]FDOPA PET/CT

Purpose

FDOPA PET shows good performance for the diagnosis of striatal dopaminergic denervation, making it a valuable tool for the differential diagnosis of Parkinsonism. Textural features are image biomarkers that could potentially improve the early diagnosis and monitoring of neurodegenerative parkinsonian syndromes. We explored the performances of textural features for binary classification of FDOPA scans.

Methods

We used two FDOPA PET datasets: 443 scans for feature selection, and 100 scans from a different PET/CT system for model testing. Scans were labelled according to expert interpretation (dopaminergic denervation versus no dopaminergic denervation). We built LASSO logistic regression models using 43 biomarkers including 32 textural features. Clinical data were also collected using a shortened UPDRS scale.

Results

The model built from the clinical data alone had a mean area under the receiver operating characteristics (AUROC) of 63.91. Conventional imaging features reached a maximum score of 93.47 but the addition of textural features significantly improved the AUROC to 95.73 (p < 0.001), and 96.10 (p < 0.001) when limiting the model to the top three features: GLCM_Correlation, Skewness and Compacity. Testing the model on the external dataset yielded an AUROC of 96.00, with 95% sensitivity and 97% specificity. GLCM_Correlation was one of the most independent features on correlation analysis, and systematically had the heaviest weight in the classification model.

Conclusion

A simple model with three radiomic features can identify pathologic FDOPA PET scans with excellent sensitivity and specificity. Textural features show promise for the diagnosis of parkinsonian syndromes.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00259-022-05816-7.

Prognostic value of immunotherapy-induced organ inflammation assessed on 18FDG PET in patients with metastatic non-small cell lung cancer

PURPOSE

We evaluated the prognostic value of immunotherapy-induced organ inflammation observed on ¹⁸FDG PET in patients with non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICPIs).

METHODS

Data from patients with IIIB/IV NSCLC included in two different prospective trials were analyzed. ¹⁸FDG PET/CT exams were performed at baseline (PET_Baseline) and repeated after 7-8 weeks (PET_Interim1) and 12-16 weeks (PET_Interim2) of treatment, using iPERCIST for tumor response evaluation. The occurrence of abnormal organ ¹⁸FDG uptake, deemed to be due to ICPI-related organ inflammation, was collected.

RESULTS

Exploratory cohort (Nice, France): PET_Interim1 and PET_Interim2 revealed the occurrence of at least one ICPI-induced organ inflammation in 72.8% of patients, including midgut/hindgut inflammation (33.7%), gastritis (21.7%), thyroiditis (18.5%), pneumonitis (17.4%), and other organ inflammations (9.8%). iPERCIST tumor response was associated with improved progression-free survival (p < 0.001). iPERCIST tumor response and immuno-induced gastritis assessed on PET were both associated with improved overall survival (OS) (p < 0.001 and p = 0.032). Combining these two independent variables, we built a model predicting patients' 2-year OS with a sensitivity of 80.3% and a specificity of 69.2% (AUC = 72.7). Validation cohort (Genova, Italy): Immuno-induced gastritis (19.6% of patients) was associated with improved OS (p = 0.04). The model built previously predicted 2-year OS with a sensitivity and specificity of 72.0% and 63.6% (AUC = 70.7) and 3-year OS with a sensitivity and specificity of 69.2% and 80.0% (AUC = 78.2).

CONCLUSION

Immuno-induced gastritis revealed by early interim ¹⁸FDG PET in around 20% of patients with NSCLC treated with ICPI is a novel and reproducible imaging biomarker of improved OS.

Primal-dual for classification with rejection (PD-CR): a novel method for classification and feature selection-an application in metabolomics studies

Background

Supervised classification methods have been used for many years for feature selection in metabolomics and other omics studies. We developed a novel primal-dual based classification method (PD-CR) that can perform classification with rejection and feature selection on high dimensional datasets. PD-CR projects data onto a low dimension space and performs classification by minimizing an appropriate quadratic cost. It simultaneously optimizes the selected features and the prediction accuracy with a new tailored, constrained primal-dual method. The primal-dual framework is general enough to encompass various robust losses and to allow for convergence analysis. Here, we compare PD-CR to three commonly used methods: partial least squares discriminant analysis (PLS-DA), random forests and support vector machines (SVM). We analyzed two metabolomics datasets: one urinary metabolomics dataset concerning lung cancer patients and healthy controls; and a metabolomics dataset obtained from frozen glial tumor samples with mutated isocitrate dehydrogenase (IDH) or wild-type IDH.

Results

PD-CR was more accurate than PLS-DA, Random Forests and SVM for classification using the 2 metabolomics datasets. It also selected biologically relevant metabolites. PD-CR has the advantage of providing a confidence score for each prediction, which can be used to perform classification with rejection. This substantially reduces the False Discovery Rate.

Conclusion

PD-CR is an accurate method for classification of metabolomics datasets which can outperform PLS-DA, Random Forests and SVM while selecting biologically relevant features. Furthermore the confidence score provided with PD-CR can be used to perform classification with rejection and reduce the false discovery rate.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-021-04478-w.

Performance of Multiparametric Functional Imaging to Assess Peritoneal Tumor Burden in Ovarian Cancer

PURPOSE

The aim of the study was to evaluate the clinical utility of pretreatment 18F-FDG PET/CT with quantitative evaluation of peritoneal metabolic cartography in relation to staging laparoscopy for ovarian carcinomatosis.

PATIENTS AND METHODS

A retrospective review of prospectively collected data from 84 patients with FIGO (International Federation of Gynecology and Obstetrics) stage IIIC to IV ovarian cancer was carried out. All patients had a double-blinded 18F-FDG PET/CT review. Discriminant capacity of metabolic parameters to identify peritoneal carcinomatosis in the 13 abdominal regions according to the peritoneal cancer index was estimated with area under the receiver operating characteristic curve (AUC).

RESULTS

The metabolic parameter showing the best trade-off between sensitivity and specificity to predict peritoneal extension compared with peritoneal cancer index score was the metabolic tumor volume (MTV), with a Spearman ρ equal to 0.380 (P < 0.001). The AUC of MTV to diagnose peritoneal involvement in the upper abdomen (regions 1, 2, and 3) ranged from 0.740 to 0.765. MTV AUC values were lower in the small bowel regions (9-12), ranging from 0.591 to 0.681, and decreased to 0.487 in the pelvic region 6. 18F-FDG PET/CT also improved the detection of extra-abdominal disease, upstaging 35 patients (41.6%) from stage IIIC to IV compared with CT alone and leading to treatment modification in more than one third of patients.

CONCLUSIONS

18F-FDG PET/CT metrics are highly accurate to reflect peritoneal tumor burden, with variable diagnostic value depending on the anatomic region. MTV is the most representative metabolic parameter to assess peritoneal tumor extension.

Dissociated Response in Metastatic Cancer: An Atypical Pattern Brought Into the Spotlight With Immunotherapy

When evaluating metastatic tumor response to systemic therapies, dissociated response is defined as the coexistence of responding and non-responding lesions within the same patient. Although commonly observed on interim whole-body imaging, the current response criteria in solid cancer do not consider this evolutive pattern, which is, by default, assimilated to progression. With targeted therapies and chemotherapies, dissociated response is observed with different frequencies, depending on the primary cancer type, treatment, and imaging modality. Because FDG PET/CT can easily assess response on a lesion-by-lesion basis, thus quickly revealing response heterogeneity, a PET/CT dissociated response has been described in up to 48% of women treated for a metastatic breast cancer. Although some studies have underlined a specific prognostic of dissociated response, it has always ended up being described as an unfavorable prognostic pattern and therefore assimilated to the “Progressive Disease” category of RECIST/PERCIST. This dichotomous imaging report (response vs. progression) provides a simple information for clinical decision-support, which probably explains the relatively low consideration for the dissociated response pattern to chemotherapies and targeted therapies until now. With immune checkpoint inhibitors, this paradigm is quickly changing. Dissociated response is observed in around 10% of advanced lung cancer patients and appears to be associated to treatment efficiency. Indeed, for this subset of patients, a clinical benefit of immunotherapy and favorable prognosis are usually observed. This specific pattern should therefore be considered in the future immunotherapy-adapted criteria for response evaluation using CT and PET/CT, and specific clinical managements should be evaluated for this response pattern.

Baseline metabolic tumor volume as a strong predictive and prognostic biomarker in patients with non-small cell lung cancer treated with PD1 inhibitors: a prospective study

BACKGROUND

Reliable predictive and prognostic markers are still lacking for patients treated with programmed death receptor 1 (PD1) inhibitors for non-small cell lung cancer (NSCLC). The purpose of this study was to investigate the prognostic and predictive values of different baseline metabolic parameters, including metabolic tumor volume (MTV), from 18F-fluorodeoxyglucose positron emission tomography-computed tomography (¹⁸F-FDG PET/CT) scans in patients with NSCLC treated with PD1 inhibitors.

METHODS

Maximum and peak standardized uptake values, MTV and total lesion glycolysis (TLG), as well as clinical and biological parameters, were recorded in 75 prospectively included patients with NSCLC treated with PD1 inhibitors. Associations between these parameters and overall survival (OS) were evaluated as well as their accuracy to predict early treatment discontinuation (ETD).

RESULTS

A high MTV and a high TLG were significantly associated with a lower OS (p<0.001). The median OS in patients with MTV above the median (36.5 cm³) was 10.5 months (95% CI: 6.2 to upper limit: unreached), while the median OS in patients with MTV below the median was not reached. Patients with no prior chemotherapy had a poorer OS than patients who had received prior systemic treatment (p=0.04). MTV and TLG could reliably predict ETD (area under the receiver operating characteristic curve=0.76, 95% CI: 0.65 to 0.87 and 0.72, 95% CI: 0.62 to 0.84, respectively).

CONCLUSION

MTV is a strong prognostic and predictive factor in patients with NSCLC treated with PD1 inhibitors and can be easily determined from routine ¹⁸F-FDG PET/CT scans. MTV, could help to personalize immunotherapy and be used to stratify patients in future clinical studies.

Comparison of unsupervised machine-learning methods to identify metabolomic signatures in patients with localized breast cancer

Genomics and transcriptomics have led to the widely-used molecular classification of breast cancer (BC). However, heterogeneous biological behaviors persist within breast cancer subtypes. Metabolomics is a rapidly-expanding field of study dedicated to cellular metabolisms affected by the environment. The aim of this study was to compare metabolomic signatures of BC obtained by 5 different unsupervised machine learning (ML) methods. Fifty-two consecutive patients with BC with an indication for adjuvant chemotherapy between 2013 and 2016 were retrospectively included. We performed metabolomic profiling of tumor resection samples using liquid chromatography-mass spectrometry. Here, four hundred and forty-nine identified metabolites were selected for further analysis. Clusters obtained using 5 unsupervised ML methods (PCA k-means, sparse k-means, spectral clustering, SIMLR and k-sparse) were compared in terms of clinical and biological characteristics. With an optimal partitioning parameter k = 3, the five methods identified three prognosis groups of patients (favorable, intermediate, unfavorable) with different clinical and biological profiles. SIMLR and K-sparse methods were the most effective techniques in terms of clustering. In-silico survival analysis revealed a significant difference for 5-year predicted OS between the 3 clusters. Further pathway analysis using the 449 selected metabolites showed significant differences in amino acid and glucose metabolism between BC histologic subtypes. Our results provide proof-of-concept for the use of unsupervised ML metabolomics enabling stratification and personalized management of BC patients. The design of novel computational methods incorporating ML and bioinformatics techniques should make available tools particularly suited to improving the outcome of cancer treatment and reducing cancer-related mortalities.

18FDG PET/CT in the early assessment of non-small cell lung cancer response to immunotherapy: frequency and clinical significance of atypical evolutive patterns

PURPOSE

This prospective study aimed (1) to assess the non-small cell lung cancer (NSCLC) evolutive patterns to immunotherapy using FDG-PET and (2) to describe their association with clinical outcome.

DESIGN

Fifty patients with metastatic NSCLC were included before pembrolizumab or nivolumab initiation. FDG-PET scan was performed at baseline and after 7 weeks of treatment (PET_interim1) and different criteria/parameters of tumor response were assessed, including PET response criteria in solid tumors (PERCIST). If a first PERCIST progressive disease (PD) without clinical worsening was observed, treatment was continued and a subsequent FDG-PET (PET_interim2) was performed at 3 months of treatment. Pseudo-progression (PsPD) was defined as a PERCIST response/stability on PET_interim2 after an initial PD. If a second PERCIST PD was assessed on PET_interim2, a homogeneous progression of lesions (termed immune homogeneous progressive-disease: iPD_homogeneous) was distinguished from a heterogeneous evolution (termed immune dissociated-response: iDR). A durable clinical benefit (DCB) of immunotherapy was defined as treatment continuation over a 6-month period. The association between PET evolutive profiles and DCB was assessed.

RESULTS

Using PERCIST on PET_interim1, 42% (21/50) of patients showed a response or stable disease, most of them (18/21) reached a DCB. In contrast, 58% (29/50) showed a PD, but more than one-third (11/29) were misclassified as they finally reached a DCB. No standard PET_interim1 criteria could accurately distinguished responding from non-responding patients. Treatment was continued in 19/29 of patients with a first PERCIST PD; the subsequent PET_interim2 demonstrated iPD_homogeneous, iDR and PsPD in 42% (8/19), 26% (5/19), and 32% (6/19), respectively. Whereas no patients with iPD_homogeneous experienced a DCB, all patients with iDR and PsPD reached a clinical benefit to immunotherapy.

CONCLUSION

In patients with a first PD on PERCIST and treatment continuation, a subsequent PET identifies more than half of them with iDR and PsPD, both patterns being strongly associated with a clinical benefit of immunotherapy.

Abstract 2449: Unsupervised machine learning methods reveal metabolomic based clusters in breast cancer patients

Background Transcriptomic have led to the now widely used sub-type based classification of breast cancer first described by Perou in 2000. Yet there persists heterogeneity in biological behaviors within breast cancer subtypes, underlining the need to refine the taxonomy of breast cancer. Metabolomics is a rapidly expanding field dedicated to the study of metabolism which integrates the impact of the environment on cell biology. The aim of this study was to identify new biological breast cancer clusters using different unsupervised machine learning (ML) methods based on metabolomic features. Those methods, in which no a priori class label information is given to guide the algorithm, seem suitable to address this type of problem.

Methods 52 patients with breast cancer and an indication for adjuvant chemotherapy between 2013 and 2016, were retrospectively included. Tumor resection specimens were analyzed. 1300 metabolomic were extracted by combined liquid chromatography-mass spectroscopy and processed using MZmine software and the “Human Metabolome” database. 5 unsupervised ML methods were used: PCA-Kmeans, Sparcl, SIMLR, Spectral clustering and K-sparse. Clinical differences between clusters and variations for every metabolite of interest were analyzed for each clustering method. Cluster separability and homogeneity was evaluated using the silhouettes method and t-sne visual evaluation.

Results Among the 5 clustering methods, with a partitioning optimum parameter k=3, only K-sparse and SIMLR methods generated 3 clusters with significant clinical differences, unmatched to traditional subtypes. These differences concerned: tumor stage, axillary lymph node invasion, histological grade, ki-67 proliferation index, and tumor phenotype. With a silhouette average of 0.84 and 0.85 for K-sparse and SIMLR methods respectively, those 2 methods gave the best score in terms of silhouette average and they showed a better gradient for tumor aggressiveness compared to the 3 other methods. Among them, 42 and 55 metabolites were selected for the construction of tumor metabolome profiles for K-sparse and SIMLR, respectively. Among selected metabolites we found a significant increase of L-methionine, L-phenylalanine, L-isoleucine and L-proline along with a significant decrease in glutathione (also characteristic of oxidative stress) and glutamate in the cluster associated with poorer histopronostic factors. This high concentration of proteinogenic amino-acid and low concentration of amino-acid precursors could be correlated to poorer prognosis.

Conclusion Unsupervised ML methods generate heterogeneous results when applied to metabolomics data extracted from breast cancer patients. K-sparse and SIMLR were able to identify three different groups based on tumor metabolome. Tumors with the worst histopronostic factors seemed to present higher concentrations of protienogenic amino-acids.

Note: This abstract was not presented at the meeting.

Citation Format: Jocelyn Gal, Caroline Bailleux, David Chardin, Thierry Pourcher, Lun Jing, Jean-Marie Guignonis, Jean-Marc Ferrero, Renaud Schiappa, Emmanuel Chamorey, Olivier Humbert. Unsupervised machine learning methods reveal metabolomic based clusters in breast cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2449.

Correction to: 18F-DOPA PET/CT in brain tumors: impact on multidisciplinary brain tumor board decisions

Jérôme Barriere was inadvertently missing in the original version of this article. He has participated to the study design, protocol writing and inclusion of a significant number of patients.

18F-DOPA PET/CT in brain tumors: impact on multidisciplinary brain tumor board decisions

PURPOSE

This study aimed to assess the therapeutic impact and diagnostic accuracy of ¹⁸F-DOPA PET/CT in patients with glioblastoma or brain metastases.

METHODS

Patients with histologically proven glioblastoma or brain metastases were prospectively included in this monocentric clinical trial (IMOTEP). Patients were included either due to a clinical suspicion of relapse or to assess residual tumor infiltration after treatment. Multimodality brain MRI and ¹⁸F-DOPA PET were performed. Patients' data were discussed during a Multidisciplinary Neuro-oncology Tumor Board (MNTB) meeting. The discussion was first based on clinical and MRI data, and an initial diagnosis and treatment plan were proposed. Secondly, a new discussion was conducted based on the overall imaging results, including ¹⁸F-DOPA PET. A second diagnosis and therapeutic plan were proposed. A retrospective and definitive diagnosis was obtained after a 3-month follow-up and considered as the reference standard.

RESULTS

One hundred six cases were prospectively investigated by the MNTB. All patients with brain metastases (N = 41) had a clinical suspicion of recurrence. The addition of ¹⁸F-DOPA PET data changed the diagnosis and treatment plan in 39.0% and 17.1% of patients' cases, respectively. Concerning patients with a suspicion of recurrent glioblastoma (N = 12), the implementation of ¹⁸F-DOPA PET changed the diagnosis and treatment plan in 33.3% of cases. In patients evaluated to assess residual glioblastoma infiltration after treatment (N = 53), ¹⁸F-DOPA PET data had a lower impact with only 5.7% (3/53) of diagnostic changes and 3.8% (2/53) of therapeutic plan changes. The definitive reference diagnosis was available in 98/106 patients. For patients with tumor recurrence suspicion, the adjunction of ¹⁸F-DOPA PET increased the Younden's index from 0.44 to 0.53 in brain metastases and from 0.2 to 1.0 in glioblastoma, reflecting an increase in diagnostic accuracy.

CONCLUSION

¹⁸F-DOPA PET has a significant impact on the management of patients with a suspicion of brain tumor recurrence, either glioblastoma or brain metastases, but a low impact when used to evaluate the residual glioblastoma infiltration after a first-line radio-chemotherapy or second-line bevacizumab.

SUV calculation in breast cancer: which normalization should be applied when using 18F-FDG PET?

BACKGROUND

When using 18F-FDG PET, glucose metabolism quantification is affected by various factors. We aimed to investigate the benefit of different standardized uptake value (SUV) normalizations to improve the accuracy of 18F-FDG uptake to predict breast cancer aggressiveness and response to treatment.

METHODS

Two hundred fifty-two women with locally advanced breast cancer treated with neoadjuvant chemotherapy (NAC) were included. Women underwent 18F-FDG PET before and after the first course of NAC. Glucose serum levels, patient heights and weights were recorded at the time of each PET exam. Four different procedures for SUV normalization of the primary tumor were used: by body weight (SUVBW) by blood glucose level (SUVG), by lean body mass (SUL) and then corrected for both lean body mass and blood glucose level (SULG).

RESULTS

At baseline, SUL was significantly lower than SUVBW (5.9±4.0 and 9.5±6.5, respectively; P<0.0001), whereas SUVG and SUVBW were not significantly different (9.7±6.4 and 9.5±6.5, respectively; P=0.67). Concerning SUV changes (ΔSUV), the different normalizations methods did not induce significant quantitative differences. The correlation coefficients were high between the four normalizations methods of SUV1, SUV2 and ΔSUVB (R>0.95; P<0.0001). High baseline SUVBW measures were positively correlated with the biological tumor characteristics of aggressiveness and proliferation (P<0.001): ductal carcinoma, high tumor grading, high mitotic activity, negative estrogen receptor status and the TNBC subtype. ΔSUVBW was highly predictive of pCR (AUC=0.76 on ROC curve analysis; P<0.0001). The different SUV normalizations yields identical statistical results and AUC to predict tumor biological aggressiveness and response to therapy.

CONCLUSIONS

In the present setting, SUVBW and SUL can be considered as robust measures and be used in future multicenter trials. The additional normalization of SUV by glycemia involves stringent methodologic procedures to avoid biased risk measurements and offers no statistical advantages.

False positive 18FDG PET-CT results due to exogenous lipoid pneumonia secondary to oily drug inhalation: A case report

RATIONALE

Exogenous lipoid pneumonia is a rare condition due to abnormal presence of oily substances in the lungs. It is a rarely known cause for false positive FDG PET-CT results and can sometimes lead to invasive investigations. Searching and finding the source of the oily substance is one of the keys to the diagnosis. Inhalation of oily drugs during snorting has rarely been described.

PATIENT CONCERNS

A patient with well controlled HIV infection was referred for an FDG PET-CT to assess extension of Kaposi's disease, recently removed from his right foot. The patient had no particular symptoms.

DIAGNOSES

Abnormal uptake of FDG was found in a suspicious lung nodule. An experienced radiologist thought the nodule was due to lipoid pneumonia.

INTERVENTIONS

Bronchoalveolar lavage fluid did not contain lipid-laden macrophages but bronchoscopy showed violet lesions resembling Kaposi's disease lesions. Lobectomy was performed after a multidisciplinary discussion.

OUTCOMES

Anatomopathological analysis revealed the nodule was due to lipoid pneumonia. The patient's quality of life did not diminish after the operation and he is still in good health. The source of the oily substance causing lipoid pneumonia was found after the surgery: the patient used to snort oily drugs.

LESSONS

The presence of a suspicious lung nodule possibly due to lipoid pneumonia in a patient with known Kaposi's disease was difficult to untangle and lead to invasive surgery. It is possible that if a source of exogenous lipoid pneumonia had been found beforehand, surgery could have been prevented.