Primary tumor standardized uptake value (SUVmax) measured on fluorodeoxyglucose positron emission tomography (FDG-PET) is of prognostic value for survival in non-small cell lung cancer (NSCLC): a systematic review and meta-analysis (MA) by the European Lung Cancer Working Party for the IASLC Lung Cancer Staging Project

Assessment of PD-L1 Expression in Non-Small Cell Lung Cancers Using [68Ga]Ga-DOTA-WL12 PET/CT

Assessing programmed death ligand-1 (PD-L1) expression in non-small cell lung cancer (NSCLC), particularly in metastatic cases, remains challenging. In this study, surface plasmon resonance (SPR) analysis and [68Ga]Ga-DOTA-WL12 micro-PET/CT imaging are performed. [68Ga]Ga-DOTA-WL12 PET/CT and [18F]FDG PET/CT are performed on a cohort of 20 patients with NSCLC. Semi-quantitative assessments include SUVmax, metabolic tumor volume (MTV), total lesion glycolysis (TLG), and target-to-background ratio (TBR). DOTA-WL12 exhibits robust PD-L1 binding with a KD value of 0.2 nM. Subsequent human studies reveal significant correlations between PD-L1 expression and the [68Ga]Ga-DOTA-WL12 SUVmax in primary and metastatic lesions, surpassing the [18F]FDG results (r = 0.8889, p <0.0001 vs r = 0.0469, p = 0.8127). Notably, [68Ga]Ga-DOTA-WL12 imaging discerned SUVmax and TBR differences between PD-L1 TPS ≤1% and PD-L1 TPS > 1% groups (p all <0.001). In an NSCLC patient with brain metastases, [68Ga]Ga-DOTA-WL12 shows a SUVmean of 0.04 in the brain background, with TBR values of 17 and 23, underscoring its potential for detecting brain metastases. The study provides initial evidence for the clinical utility of [68Ga]Ga-DOTA-WL12 PET/CT for lesion detection, immunotherapy selection, and therapeutic efficacy evaluation in PD-L1-expressing NSCLC, demonstrating its potential as a valuable tool in NSCLC research and management.

Imaging at the nexus: how state of the art imaging techniques can enhance our understanding of cancer and fibrosis

Both cancer and fibrosis are diseases involving dysregulation of cell signaling pathways resulting in an altered cellular microenvironment which ultimately leads to progression of the condition. The two disease entities share common molecular pathophysiology and recent research has illuminated the how each promotes the other. Multiple imaging techniques have been developed to aid in the early and accurate diagnosis of each disease, and given the commonalities between the pathophysiology of the conditions, advances in imaging one disease have opened new avenues to study the other. Here, we detail the most up-to-date advances in imaging techniques for each disease and how they have crossed over to improve detection and monitoring of the other. We explore techniques in positron emission tomography (PET), magnetic resonance imaging (MRI), second generation harmonic Imaging (SGHI), ultrasound (US), radiomics, and artificial intelligence (AI). A new diagnostic imaging tool in PET/computed tomography (CT) is the use of radiolabeled fibroblast activation protein inhibitor (FAPI). SGHI uses high-frequency sound waves to penetrate deeper into the tissue, providing a more detailed view of the tumor microenvironment. Artificial intelligence with the aid of advanced deep learning (DL) algorithms has been highly effective in training computer systems to diagnose and classify neoplastic lesions in multiple organs. Ultimately, advancing imaging techniques in cancer and fibrosis can lead to significantly more timely and accurate diagnoses of both diseases resulting in better patient outcomes.

Quantitative evaluation of lesion response heterogeneity for superior prognostication of clinical outcome

PURPOSE

Standardized reporting of treatment response in oncology patients has traditionally relied on methods like RECIST, PERCIST and Deauville score. These endpoints assess only a few lesions, potentially overlooking the response heterogeneity of all disease. This study hypothesizes that comprehensive spatial-temporal evaluation of all individual lesions is necessary for superior prognostication of clinical outcome.

METHODS

[18F]FDG PET/CT scans from 241 patients (127 diffuse large B-cell lymphoma (DLBCL) and 114 non-small cell lung cancer (NSCLC)) were retrospectively obtained at baseline and either during chemotherapy or post-chemoradiotherapy. An automated TRAQinform IQ software (AIQ Solutions) analyzed the images, performing quantification of change in regions of interest suspicious of cancer (lesion-ROI). Multivariable Cox proportional hazards (CoxPH) models were trained to predict overall survival (OS) with varied sets of quantitative features and lesion-ROI, compared by bootstrapping with C-index and t-tests. The best-fit model was compared to automated versions of previously established methods like RECIST, PERCIST and Deauville score.

RESULTS

Multivariable CoxPH models demonstrated superior prognostic power when trained with features quantifying response heterogeneity in all individual lesion-ROI in DLBCL (C-index = 0.84, p < 0.001) and NSCLC (C-index = 0.71, p < 0.001). Prognostic power significantly deteriorated (p < 0.001) when using subsets of lesion-ROI (C-index = 0.78 and 0.67 for DLBCL and NSCLC, respectively) or excluding response heterogeneity (C-index = 0.67 and 0.70). RECIST, PERCIST, and Deauville score could not significantly associate with OS (C-index < 0.65 and p > 0.1), performing significantly worse than the multivariable models (p < 0.001).

CONCLUSIONS

Quantitative evaluation of response heterogeneity of all individual lesions is necessary for the superior prognostication of clinical outcome.

Beneficial implications of adjuvant chemotherapy for stage IB lung adenocarcinoma exhibiting elevated SUVmax in FDG-PET/CT: a retrospective study from a single center

Background

Controversy surrounds the efficacy of adjuvant chemotherapy (ACT) in the treatment of stage I lung adenocarcinoma (LUAD). The objective of this study was to examine the impact of the maximum standardized uptake value (SUVmax) as measured by 18F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) on the efficacy of ACT in patients diagnosed with stage I LUAD.

Methods

We scrutinized the medical records of 928 consecutive patients who underwent complete surgical resection for pathological stage I LUAD at our institution. The ideal cut-off value for primary tumor SUVmax in terms of disease-free survival (DFS) and overall survival (OS) was determined using the X-tile software. The Kaplan-Meier method and Cox regression analysis were used for survival analysis.

Results

Based on the SUVmax algorithm, the ideal cutoff values were determined to be 4.9 for DFS and 5.0 for OS. We selected 5.0 as the threshold because OS is the more widely accepted predictive endpoint. In a multivariate Cox regression analysis, SUVmax ≥ 5.0, problematic IB stage, and sublobectomy were identified as independent risk factors for poor DFS and OS. It is noteworthy that patients who were administered ACT had significantly longer DFS and OS than what was observed in the subgroup of patients with pathological stage IB LUAD and SUVmax ≥ 5.0 (p < 0.035 and p ≤ 0.046, respectively). However, there was no observed survival advantage for patients in stages IA or IB who had an SUVmax < 5.0.

Conclusion

The preoperative SUVmax of tumors served as an indicator of the impact of ACT in the context of completely resected pathological stage I LUAD. Notably, patients within the Stage IB category exhibiting elevated SUVmax levels emerged as a subgroup experiencing substantial benefits from postoperative ACT.

Impact of preoperative [18F]FDG PET/CT vs. contrast-enhanced CT in the staging and survival of patients with clinical stage I and II non-small cell lung cancer: a 10-year follow-up study

OBJECTIVES

To elucidate the impact of [18F]FDG positron emission tomography/computed tomography (PET/CT) vs. CT workup on staging and prognostic evaluation of clinical stage (c) I-II NSCLC.

METHODS

We retrospectively identified 659 cI-II NSCLC who underwent CT (267 patients) or preoperative CT followed by PET/CT (392 patients), followed by curative-intended complete resection in our hospital from January 2008 to December 2013. Differences were assessed between preoperative and postoperative stage. Five-year disease-free survival (DFS) and overall survival (OS) rates were calculated using the Kaplan-Meier approach and compared with log-rank test. Impact of preoperative PET/CT on survival was assessed by Cox regression analysis.

RESULTS

The study included 659 patients [mean age, 59.5 years ± 10.8 (standard deviation); 379 men]. The PET/CT group was superior over CT group in DFS [12.6 vs. 6.9 years, HR 0.67 (95% CI 0.53-0.84), p < 0.001] and OS [13.9 vs. 10.5 years, HR 0.64 (95% CI 0.50-0.81), p < 0.001]. In CT group, more patients thought to have cN0 migrated to pN1/2 disease as compared with PET/CT group [26.4% (66/250) vs. 19.2% (67/349), p < 0.001], resulting in more stage cI cases being upstaged to pII-IV [24.7% (49/198) vs. 16.1% (47/292), p = 0.02], yet this was not found in cII NSCLC [27.5% (19/69) vs. 27.0% (27/100), p = 0.94]. Cox regression analysis identified preoperative PET/CT as an independent prognostic factor of OS and DFS (p = 0.002, HR = 0.69, 95% CI 0.54-0.88; p = 0.004, HR = 0.72, 95% CI 0.58-0.90).

CONCLUSION

Addition of preoperative [18F]FDG PET/CT was associated with superior DFS and OS in resectable cI-II NSCLC, which may result from accurate staging and stage-appropriate therapy.

PET/CT-based deep learning grading signature to optimize surgical decisions for clinical stage I invasive lung adenocarcinoma and biologic basis under its prediction: a multicenter study

PURPOSE

No consensus on a grading system for invasive lung adenocarcinoma had been built over a long period of time. Until October 2020, a novel grading system was proposed to quantify the whole landscape of histologic subtypes and proportions of pulmonary adenocarcinomas. This study aims to develop a deep learning grading signature (DLGS) based on positron emission tomography/computed tomography (PET/CT) to personalize surgical treatments for clinical stage I invasive lung adenocarcinoma and explore the biologic basis under its prediction.

METHODS

A total of 2638 patients with clinical stage I invasive lung adenocarcinoma from 4 medical centers were retrospectively included to construct and validate the DLGS. The predictive performance of the DLGS was evaluated by the area under the receiver operating characteristic curve (AUC), its potential to optimize surgical treatments was investigated via survival analyses in risk groups defined by the DLGS, and its biological basis was explored by comparing histologic patterns, genotypic alternations, genetic pathways, and infiltration of immune cells in microenvironments between risk groups.

RESULTS

The DLGS to predict grade 3 achieved AUCs of 0.862, 0.844, and 0.851 in the validation set (n = 497), external cohort (n = 382), and prospective cohort (n = 600), respectively, which were significantly better than 0.814, 0.810, and 0.806 of the PET model, 0.813, 0.795, and 0.824 of the CT model, and 0.762, 0.734, and 0.751 of the clinical model. Additionally, for DLGS-defined high-risk population, lobectomy yielded an improved prognosis compared to sublobectomy p = 0.085 for overall survival [OS] and p = 0.038 for recurrence-free survival [RFS]) and systematic nodal dissection conferred a superior prognosis to limited nodal dissection (p = 0.001 for OS and p = 0.041 for RFS).

CONCLUSION

The DLGS harbors the potential to predict the histologic grade and personalize the surgical treatments for clinical stage I invasive lung adenocarcinoma. Its applicability to other territories should be further validated by a larger international study.

Pretreatment and Posttreatment Tumor Metabolic Activity Assessed by FDG-PET/CT as Predictors of Tumor Recurrence and Survival Outcomes in Early-Stage Non-Small Cell Lung Cancer Treated With Stereotactic Body Radiation Therapy

Purpose

Stereotactic body radiation therapy (SBRT) is considered the standard of care for medically inoperable early-stage non-small cell lung cancer. There is mixed evidence on the prognostic significance of tumor metabolic activity assessed by positron emission tomography combined with computed tomography (PET/CT) using F-18 fluorodeoxyglucose (FDG). The objectives of this study were to evaluate the maximum standardized uptake value (SUVmax) pretreatment and at 3 and 6 months after SBRT for prediction of tumor control and survival outcomes.

Methods and Materials

Consecutive patients from a single institution with T12N0M0 non-small cell lung cancer receiving primary treatment with SBRT with pretreatment FDG-PET/CT (n = 163) and follow-up FDG-PET/CT at 3 or 6 months (n = 71) were included. Receiver operator characteristic analysis was performed to dichotomize variables for Kaplan-Meier survival analysis. Multivariate analysis was performed with Cox proportional hazards regression.

Results

Median follow-up was 19 months. For the whole cohort, 1-year and 2-year local control, progression-free survival (PFS), and overall survival (OS) were 95.0% and 80.3%, 87.1% and 75.4%, and 67.0% and 49.6% respectively. The following pre-SBRT SUVmax cutoffs were significant: SUV > 4.0 for distant failure-free survival (adjusted hazard ratio [aHR], 3.33, P = .006), >12.3 for PFS (aHR, 2.80, P = .011), and >12.6 for OS (aHR, 3.00, P = .003). SUVmax decreases of at least 45% at 3 months (aHR, 0.15, P = .018), and 53% at 6 months (aHR, 0.12, P = .046) were associated with improved local failure-free survival.

Conclusions

Pre-SBRT SUVmax cutoffs can predict distant failure, PFS, and OS. At both 3 and 6 months after SBRT, cutoffs for percentage change in SUVmax can potentially stratify risk of local recurrence.

Methodological evaluation of original articles on radiomics and machine learning for outcome prediction based on positron emission tomography (PET)

AIM

Despite a vast number of articles on radiomics and machine learning in positron emission tomography (PET) imaging, clinical applicability remains limited, partly owing to poor methodological quality. We therefore systematically investigated the methodology described in publications on radiomics and machine learning for PET-based outcome prediction.

METHODS

A systematic search for original articles was run on PubMed. All articles were rated according to 17 criteria proposed by the authors. Criteria with >2 rating categories were binarized into "adequate" or "inadequate". The association between the number of "adequate" criteria per article and the date of publication was examined.

RESULTS

One hundred articles were identified (published between 07/2017 and 09/2023). The median proportion of articles per criterion that were rated "adequate" was 65% (range: 23-98%). Nineteen articles (19%) mentioned neither a test cohort nor cross-validation to separate training from testing. The median number of criteria with an "adequate" rating per article was 12.5 out of 17 (range, 4-17), and this did not increase with later dates of publication (Spearman's rho, 0.094; p = 0.35). In 22 articles (22%), less than half of the items were rated "adequate". Only 8% of articles published the source code, and 10% made the dataset openly available.

CONCLUSION

Among the articles investigated, methodological weaknesses have been identified, and the degree of compliance with recommendations on methodological quality and reporting shows potential for improvement. Better adherence to established guidelines could increase the clinical significance of radiomics and machine learning for PET-based outcome prediction and finally lead to the widespread use in routine clinical practice.

CircANKRD17 promotes glycolysis by inhibiting miR-143 in breast cancer cells

Glucose metabolic reprogramming, known as the Warburg effect, is one of the metabolic hallmarks of tumor cells. Cancer cells preferentially metabolize glucose by glycolysis rather than mitochondrial oxidative phosphorylation regardless of oxygen availability, but the regulatory mechanism underlying this switch has been incompletely understood. Here, we report that the circular RNA circ ankyrin repeat domain 17 (ANKRD17) functions as a key regulator for glycolysis to promote cell growth, migration, invasion, and cell-cycle progression in breast cancer (BC) cells. We further show that circANKRD17 acts to accelerate glycolysis in BC cells by acting as a sponge for miR-143 and in turn overrides the repressive effect of miR-143, a well-documented glycolytic repressor, on hexokinase 2 in BC cells, thus resulting in enhanced glycolysis in BC cells. These data suggest the circANKRD17-miR-143 cascade as a novel mechanism in controlling glucose metabolic reprogramming in BC cells and suggest circANKRD17 as a promising therapeutic target to interrupt cancerous glycolysis.

PET/CT based cross-modal deep learning signature to predict occult nodal metastasis in lung cancer

Occult nodal metastasis (ONM) plays a significant role in comprehensive treatments of non-small cell lung cancer (NSCLC). This study aims to develop a deep learning signature based on positron emission tomography/computed tomography to predict ONM of clinical stage N0 NSCLC. An internal cohort (n = 1911) is included to construct the deep learning nodal metastasis signature (DLNMS). Subsequently, an external cohort (n = 355) and a prospective cohort (n = 999) are utilized to fully validate the predictive performances of the DLNMS. Here, we show areas under the receiver operating characteristic curve of the DLNMS for occult N1 prediction are 0.958, 0.879 and 0.914 in the validation set, external cohort and prospective cohort, respectively, and for occult N2 prediction are 0.942, 0.875 and 0.919, respectively, which are significantly better than the single-modal deep learning models, clinical model and physicians. This study demonstrates that the DLNMS harbors the potential to predict ONM of clinical stage N0 NSCLC.

A fingerprint of 2-[18F]FDG radiometabolites - How tissue-specific metabolism beyond 2-[18F]FDG-6-P could affect tracer accumulation

Studies indicate that the radiotracer 2-[18F]fluoro-2-deoxy-D-glucose (2-[18F]FDG) can be metabolized beyond 2-[18F]FDG-6-phosphate (2-[18F]FDG-6-P), but its metabolism is incompletely understood. Most importantly, it remains unclear whether downstream metabolism affects tracer accumulation in vivo. Here we present a fingerprint of 2-[18F]FDG radiometabolites over time in cancer cells, corresponding tumor xenografts and murine organs. Strikingly, radiometabolites representing glycogen metabolism or the oxPPP correlated inversely with tracer accumulation across all examined tissues. Recent studies suggest that not only hexokinase, but also hexose-6-phosphate dehydrogenase (H6PD), an enzyme of the oxidative pentose phosphate pathway (oxPPP), determines 2-[18F]FDG accumulation. However, little is known about the corresponding enzyme glucose-6-phosphate dehydrogenase (G6PD). Our mechanistic in vitro experiments on the role of the oxPPP propose that 2-[18F]FDG can be metabolized via both G6PD and H6PD, but data from separate enzyme knockdown suggest diverging roles in downstream tracer metabolism. Overall, we propose that tissue-specific metabolism beyond 2-[18F]FDG-6-P could matter for imaging.

Association between the Preoperative Standard Uptake Value (SUV) and Survival Outcomes after Robotic-Assisted Segmentectomy for Resectable Non-Small Cell Lung Cancer (NSCLC)

BACKGROUND

Lung-sparing procedures, specifically segmentectomies and wedge resections, have increased over the years to treat early-stage non-small cell lung cancer (NSCLC). We investigate here the perioperative and long-term outcomes of patients who underwent robotic-assisted segmentectomy (RAS) at an NCI-designated cancer center and aim to show associations between the preoperative standard update value (SUV) to tumor stage, recurrence patterns, and overall survival.

METHODS

A retrospective analysis was performed on 166 consecutive patients who underwent RAS at a single institution from 2010 to 2021. Of this number, 121 robotic-assisted segmentectomies were performed for primary NSCLC, and a total of 101 patients were evaluated with a PET-CT scan. The SUV from the primary tumor was determined from the PET-CT. The clinical, surgical, and pathologic profiles and perioperative outcomes were summarized via descriptive statistics. Numerical variables were described as the median and interquartile range because all numerical variables were not normally distributed as assessed by the Shapiro-Wilk test of normality. Categorical variables were described as the count and proportion. Chi-square or Fisher's exact test was used for association. The main outcomes were overall survival (OS) and recurrence-free survival (RFS). Kaplan-Meier (KM) curves were constructed to visualize the OS and RFS, which were also stratified according to tumor histology, the pathologic stage, and standard uptake value. A log-rank test for the equality of survival curves was performed to determine significant differences between groups.

RESULTS

The most common postoperative complications were atrial fibrillation (8.8%, 9/102), persistent air leak (7.84%, 8/102), and pneumonia (4.9%, 5/102). The median operative duration was 168.5 min (IQR 59), while the median estimated blood loss was 50 mL (IQR 125). The conversion rate to thoracotomy in this cohort was 3.9% (4/102). Intraoperative complications occurred in 2.9% (3/102). The median hospital length of stay was 3 days (IQR 3). The median chest tube duration was 3 days (IQR 2), but 4.9% (5/102) of patients were sent home with a chest tube. The recurrence for this cohort was 28.4% (29/102). The time to recurrence was 353 days (IQR 504), while the time to mortality was 505 days (IQR 761). The NSCLC patients were divided into the following two groups: low SUV (<5, n = 55) and high SUV (≥5, n = 47). Statistically significant associations were noted between SUV and the tumor histology (p = 0.019), tumor grade (p = 0.002), lymph-vascular invasion (p = 0.029), viscera-pleural invasion (p = 0.008), recurrence (p < 0.001) and the site of recurrence (p = 0.047). KM survival analysis showed significant differences in the curves for OS (log-rank p-value 0.0204) and RFS (log-rank p-value 0.0034) between the SUV groups.

CONCLUSION

Robotic-assisted segmentectomy for NSCLC has reasonable perioperative and oncologic outcomes. Furthermore, we demonstrate here the prognostic implication of preoperative SUV to pathologic outcomes, recurrence-free survival, and overall survival.

Exploring the correlation between HER2 alterations and 18F-FDG PET/CT metabolic parameters and their prognostic value in EGFR-negative non-small-cell lung cancer patients

PURPOSE

Our study intended to explore the correlation between HER2 alterations and 18F-FDG PET/CT metabolic parameters and their prognostic value in EGFR-negative non-small-cell lung cancer (NSCLC) patients detected by next-generation sequencing (NGS).

METHODS

NGS assay was performed in 1737 NSCLC patients, a total of 88 HER2 alterations and 176 negative HER2 with EGFR-negative patients were randomly selected for this study.

RESULTS

When the HER2 status with EGFR-negative group was analyzed, multivariate analysis showed that smoking status, primary tumor SUVmax (pSUVmax) < 13.03 and stage were the independent deterministic factors of HER2 alterations. Multivariate cox regression analysis revealed that HER2 status, age, smoking status and stage were independent risk factors for overall survival (OS) in EGFR-negative NSCLC patients with different HER2 status. When the HER2 alterations group was separately analyzed, multivariate analysis demonstrated that low pSUVmax < 15.32 and histology were the independent deterministic factors of HER2 mutation. Multivariate cox regression analysis revealed that pSUVmax, smoking status, nodal involvement and treatment methods were independent risk factors for OS in EGFR-negative NSCLC patients with HER2 alterations.

CONCLUSION

The study revealed that low pSUVmax was associated with HER2 alterations in EGFR-negative NSCLC patients, moreover HER2 mutation and HER2 amplification exhibited distinct 18F-FDG metabolic and clinical characteristics. Furthermore, it explored the prognostic value of HER2 alterations and 18F-FDG PET/CT metabolic parameters of pSUVmax in EGFR-negative NSCLC patients.

A machine learning-based PET/CT model for automatic diagnosis of early-stage lung cancer

Objective

The aim of this study was to develop a machine learning-based automatic analysis method for the diagnosis of early-stage lung cancer based on positron emission tomography/computed tomography (PET/CT) data.

Methods

A retrospective cohort study was conducted using PET/CT data from 187 cases of non-small cell lung cancer (NSCLC) and 190 benign pulmonary nodules. Twelve PET and CT features were used to train a diagnosis model. The performance of the machine learning-based PET/CT model was tested and validated in two separate cohorts comprising 462 and 229 cases, respectively.

Results

The standardized uptake value (SUV) was identified as an important biochemical factor for the early stage of lung cancer in this model. The PET/CT diagnosis model had a sensitivity and area under the curve (AUC) of 86.5% and 0.89, respectively. The testing group comprising 462 cases showed a sensitivity and AUC of 85.7% and 0.87, respectively, while the validation group comprising 229 cases showed a sensitivity and AUC of 88.4% and 0.91, respectively. Additionally, the proposed model improved the clinical discrimination ability for solid pulmonary nodules (SPNs) in the early stage significantly.

Conclusion

The feature data collected from PET/CT scans can be analyzed automatically using machine learning techniques. The results of this study demonstrated that the proposed model can significantly improve the accuracy and positive predictive value (PPV) of SPNs at the early stage. Furthermore, this algorithm can be optimized into a robotic and less biased PET/CT automatic diagnosis system.

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.

Relationship between the expression of PD-L1 and 18F-FDG uptake in pancreatic ductal adenocarcinoma

BACKGROUND

PD-L1 promotes glycolysis in tumour cells. We observed a correlation between high PD-L1 expression and high 18F-FDG uptake in patients with pancreatic ductal adenocarcinoma (PDAC) in a previous study. This study aims to determine the usefulness of 18F-FDG PET/CT for evaluating the PD-L1 status in PDAC and to elucidate its rationality by integrated analyses.

METHODS

For bioinformatics analysis, WGCNA, GSEA and TIMER were applied to analyse the pathways and hub genes associated with PD-L1 and glucose uptake. 18F-FDG uptake assay was used to determine the glucose uptake rate of PDAC cells in vitro. Related genes expression were verified by RT-PCR and western blot. A retrospective analysis was performed on 47 patients with PDAC who had undergone 18F-FDG PET/CT. Maximum standardised uptake values (SUVmax) were determined. The usefulness of SUVmax for evaluating PD-L1 status was determined by receiver operating characteristic (ROC) curve analysis.

RESULTS

Bioinformatics analysis showed that several signalling pathways are associated with both PD-L1 expression and tumour glucose uptake, among which JAK-STAT may be an important one. By in vitro experiments, the regulatory role of PD-L1 on glucose uptake was demonstrated, and its dependency on the JAK-STAT pathway was also verified by the rescue study. The SUVmax of PD-L1-positive patients was significantly higher than PD-L1-negative in tumour cells (TCs) (6.1 ± 2.3 vs. 11.1 ± 4.2; P < 0.001), and in tumour-infiltrating immune cells (TIICs) (6.4 ± 3.2 vs. 8.4 ± 3.5; P < 0.001). In a multivariate analysis, SUVmax was significantly associated with PD-L1 expression in TCs and TIICs (P < 0.001 and P = 0.018, respectively). Using SUVmax cut-off values of 8.15 and 7.75, PD-L1 status in TCs and TIICs could be predicted with accuracies of 91.5% and 74.5%, respectively.

CONCLUSION

Higher 18F-FDG uptake by PDAC is associated with elevated PD-L1 expression. JAK-STAT is an important pathway that mediates PD-L1 to promote glucose uptake in PDAC.

Metabolic kinetic modeling of [11C]methionine based on total-body PET in multiple myeloma

PURPOSE

Multiple myeloma (MM) is a malignant disease characterized by the secretion of monoclonal immunoglobulins and has a high demand for amino acids. [11C]methionine total-body PET is capable of noninvasive dynamic monitoring of radiotracer in vivo, thus providing a way to reveal the dynamic changes of myeloma metabolism. This study aims to analyze the metabolic process of [11C]methionine based on kinetic modeling, and to preliminary reveal its application value in MM.

METHODS

Dynamic total-body [11C]methionine PET/CT was conducted with uEXPLORER in 12 subjects (9 MM patients and 3 controls). The tissue time activity curves (TACs) of organs and bone marrows were extracted. Model fitting of TACs was operated using PMOD Kinetic Modeling. After validation by Goodness of fit (GOF), the reversible two-tissue compartment model (2T4k) was used to further analysis. R software was used to analyze the correlation between kinetic parameters and clinical indicators.

RESULTS

The 2T4k has passed the criterion of GOF and was used to fit the data of 0-20 minutes. The [11C]methionine net uptake rate (Ki) was significantly higher in the MM lesions than in the non-myeloma controls (control: 0.040±0.007 mL/g/min, MM: 0.171±0.108 mL/g/min, p=0.009). The Ki values were found to be correlated with M protein levels in MM patients. MM patients with t(4;14) translocations had an elevated k4 value compared with t(4;14) negative patients.

CONCLUSION

MM lesions have a propensity for uptake of [11C]methionine. The serum levels of M protein are correlated with [11C]methionine uptake rate in myeloma. Metabolic classification based on the k4 value may be a promising strategy for risk stratification in MM.

Predictive Value of Total Metabolic Tumor Burden Prior to Treatment in NSCLC Patients Treated with Immune Checkpoint Inhibition

OBJECTIVES

We aimed to assess the predictive value of the total metabolic tumor burden prior to treatment in patients with advanced non-small-cell lung cancer (NSCLC) receiving immune checkpoint inhibitors (ICIs).

METHODS

Pre-treatment 2-deoxy-2-[¹⁸F]fluoro-D-glucose positron emission tomography/computed tomography (PET/CT) scans performed in two consecutive years for staging in adult patients with confirmed NSCLC were considered. Volume, maximum/mean standardized uptake value (SUVmax/SUVmean), metabolic tumor volume (MTV) and total lesion glycolysis (TLG) were assessed per delineated malignant lesion (including primary tumor, regional lymph nodes and distant metastases) in addition to the morphology of the primary tumor and clinical data. Total metabolic tumor burden was captured by _totalMTV and _totalTLG. Overall survival (OS), progression-free survival (PFS) and clinical benefit (CB) were used as endpoints for response to treatment.

RESULTS

A total of 125 NSCLC patients were included. Osseous metastases were the most frequent distant metastases (n = 17), followed by thoracal distant metastases (pulmonal = 14 and pleural = 13). Total metabolic tumor burden prior to treatment was significantly higher in patients treated with ICIs (mean _totalMTV ± standard deviation (SD) 72.2 ± 78.7; mean _totalTLG ± SD 462.2 ± 538.9) compared to those without ICI treatment (mean _totalMTV ± SD 58.1 ± 233.8; mean _totalTLG ± SD 290.0 ± 784.2). Among the patients who received ICIs, a solid morphology of the primary tumor on imaging prior to treatment was the strongest outcome predictor for OS (Hazard ratio HR 28.04, p < 0.01), PFS (HR 30.89, p < 0.01) and CB (parameter estimation PE 3.46, p < 0.01), followed by the metabolic features of the primary tumor. Interestingly, total metabolic tumor burden prior to immunotherapy showed a negligible impact on OS (p = 0.04) and PFS (p = 0.01) after treatment given the hazard ratios of 1.00, but also on CB (p = 0.01) given the PE < 0.01. Overall, biomarkers on pre-treatment PET/CT scans showed greater predictive power in patients receiving ICIs, compared to patients without ICI treatment.

CONCLUSIONS

Morphological and metabolic properties of the primary tumors prior to treatment in advanced NSCLC patients treated with ICI showed great outcome prediction performances, as opposed to the pre-treatment total metabolic tumor burdens, captured by _totalMTV and _totalTLG, both with negligible impact on OS, PFS and CB. However, the outcome prediction performance of the total metabolic tumor burden might be influenced by the value itself (e.g., poorer prediction performance at very high or very low values of total metabolic tumor burden). Further studies including subgroup analysis with regards to different values of total metabolic tumor burden and their respective outcome prediction performances might be needed.

Four-dimensional quantitative analysis using FDG-PET in clinical oncology

Positron emission tomography (PET) with F-18 fluorodeoxyglucose (FDG) has been commonly used in many oncological areas. High-resolution PET permits a three-dimensional analysis of FDG distributions on various lesions in vivo, which can be applied for tissue characterization, risk analysis, and treatment monitoring after chemoradiotherapy and immunotherapy. Metabolic changes can be assessed using the tumor absolute FDG uptake as standardized uptake value (SUV) and metabolic tumor volume (MTV). In addition, tumor heterogeneity assessment can potentially estimate tumor aggressiveness and resistance to chemoradiotherapy. Attempts have been made to quantify intratumoral heterogeneity using radiomics. Recent reports have indicated the clinical feasibility of a dynamic FDG PET-computed tomography (CT) in pilot cohort studies of oncological cases. Dynamic imaging permits the assessment of temporal changes in FDG uptake after administration, which is particularly useful for differentiating pathological from physiological uptakes with high diagnostic accuracy. In addition, several new parameters have been introduced for the in vivo quantitative analysis of FDG metabolic processes. Thus, a four-dimensional FDG PET-CT is available for precise tissue characterization of various lesions. This review introduces various new techniques for the quantitative analysis of FDG distribution and glucose metabolism using a four-dimensional FDG analysis with PET-CT. This elegant study reveals the important role of tissue characterization and treatment strategies in oncology.

Evaluation of Data-Driven Respiration Gating in Continuous Bed Motion in Lung Lesions

Respiration gating is used in PET to prevent image quality degradation due to respiratory effects. In this study, we evaluated a type of data-driven respiration gating for continuous bed motion, OncoFreeze AI, which was implemented to improve image quality and the accuracy of semiquantitative uptake values affected by respiratory motion. Methods: ¹⁸F-FDG PET/CT was performed on 32 patients with lung lesions. Two types of respiration-gated images (OncoFreeze AI with data-driven respiration gating, device-based amplitude-based OncoFreeze with elastic motion compensation) and ungated images (static) were reconstructed. For each image, we calculated SUV and metabolic tumor volume (MTV). The improvement rate (IR) from respiration gating and the contrast-to-noise ratio (CNR), which indicates the improvement in image noise, were also calculated for these indices. IR was also calculated for the upper and lower lobes of the lung. As OncoFreeze AI assumes the presence of respiratory motion, we examined quantitative accuracy in regions where respiratory motion was not present using a ⁶⁸Ge cylinder phantom with known quantitative accuracy. Results: OncoFreeze and OncoFreeze AI showed similar values, with a significant increase in SUV and decrease in MTV compared with static reconstruction. OncoFreeze and OncoFreeze AI also showed similar values for IR and CNR. OncoFreeze AI increased SUV_max by an average of 18% and decreased MTV by an average of 25% compared with static reconstruction. From the IR results, both OncoFreeze and OncoFreeze AI showed a greater IR from static reconstruction in the lower lobe than in the upper lobe. OncoFreeze and OncoFreeze AI increased CNR by 17.9% and 18.0%, respectively, compared with static reconstruction. The quantitative accuracy of the ⁶⁸Ge phantom, assuming a region of no respiratory motion, was almost equal for the static reconstruction and OncoFreeze AI. Conclusion: OncoFreeze AI improved the influence of respiratory motion in the assessment of lung lesion uptake to a level comparable to that of the previously launched OncoFreeze. OncoFreeze AI provides more accurate imaging with significantly larger SUVs and smaller MTVs than static reconstruction.

Metabolic activity grows in human cancers pushed by phenotypic variability

Different evolutionary processes push cancers to increasingly aggressive behaviors, energetically sustained by metabolic reprogramming. The collective signature emerging from this transition is macroscopically displayed by positron emission tomography (PET). In fact, the most readily PET measure, the maximum standardized uptake value (SUV_max), has been found to have prognostic value in different cancers. However, few works have linked the properties of this metabolic hotspot to cancer evolutionary dynamics. Here, by analyzing diagnostic PET images from 512 patients with cancer, we found that SUV_max scales superlinearly with the mean metabolic activity (SUV_mean), reflecting a dynamic preferential accumulation of activity on the hotspot. Additionally, SUV_max increased with metabolic tumor volume (MTV) following a power law. The behavior from the patients data was accurately captured by a mechanistic evolutionary dynamics model of tumor growth accounting for phenotypic transitions. This suggests that non-genetic changes may suffice to fuel the observed sustained increases in tumor metabolic activity.

Prognostic Significance of the Maximum Standardized Uptake Value on the Prognosis of Clinical Stage IA Lung Adenocarcinoma Based on the 8th Edition TNM Classification

BACKGROUND

There are few reports on the utility of the maximum standardized uptake value (SUVmax) for predicting the prognosis of early-stage lung adenocarcinoma based on the latest tumor-node-metastasis (TNM) classification. This study aimed to determine whether clinicopathologic factors, including the SUVmax, affect prognosis in these patients.

PATIENTS AND METHODS

We enrolled 527 patients with c-stage IA lung adenocarcinoma who underwent lobectomy or greater resection between 2011 and 2017. Recurrence-free survival (RFS) and overall survival (OS) were analyzed using Kaplan-Meier curves and compared using the log-rank test. Factors associated with RFS and OS were determined using the Cox proportional hazards model.

RESULTS

RFS was significantly different based on tumor stage. In contrast, there was no significant difference in OS between patients with stage IA2 and IA3 disease (p = 0.794), although there were significant differences in OS between patients with stage IA1 and IA2 disease (p = 0.024) and between patients with stage IA1 and IA3 disease (p = 0.012). Multivariate analysis demonstrated that SUVmax was independently associated with both RFS and OS among patients with c-stage IA lung adenocarcinoma (RFS, p = 0.017; OS, p = 0.047). Further, even though there was no significant difference in OS between patients with stage IA2 and IA3 disease (n = 410), SUVmax was able to stratify patients with high and low RFS and OS among these patients (RFS, p < 0.001; OS, p < 0.001).

CONCLUSION

SUVmax was an important preoperative factor to evaluate prognosis among patients with c-stage IA lung adenocarcinoma as well as the current TNM classification.

Prognostic Role of Biologically Active Volume of Disease in Patients With Metastatic Lung Adenocarcinoma

BACKGROUND

Number of metastatic sites can identify patient populations with non-small cell lung cancer (NSCLC) that benefit from aggressive therapy. Total volume of disease is also relevant. We evaluated the prognostic impact of biologically active volume of disease (BaVD) on patients with metastatic lung adenocarcinoma.

MATERIALS AND METHODS

Positron emission tomography/computerized tomography (PET/CT) scans from patients with newly diagnosed lung adenocarcinoma prior to starting any therapy were identified. SUV thresholds of 3 and 4 were used to auto-contour all FDG avid areas. Kaplan-Meier analysis and Cox regression were performed to examine influence on OS.

RESULTS

One hundred forty-eight patients were included in the analysis. The median BaVD when using an SUV threshold of 3 was 122.8 mL. The median BaVD when using an SUV threshold of 4 was 46.2 mL When stratified by median BaVD using an SUV of 3, median OS was higher for patients with <=122.8 mL (2.12 years) compared to patients with >122.8 mL (1.46 years) (log-rank P = .001). Similarly, when stratified by median BaVD using an SUV of 4, median OS was higher for patients with <=46.2 mL (1.91 years; 95% CI: 1.65-3.22 years) compared to patients with >46.2 mL (1.48 years; 95% CI: 1.07-1.80 years) (log-rank P = .007). On multivariable analysis, BaVD was significantly associated with OS when using an SUV threshold of 3 (HR: 20.169, P < .001) and 4 (HR: 4.117, P < .001).

CONCLUSION

BaVD is an important prognostic factor in metastatic lung adenocarcinoma and may aid identification of patients with limited disease who may be candidates for more aggressive therapies.

The role of artificial intelligence based on PET/CT radiomics in NSCLC: Disease management, opportunities, and challenges

Objectives

Lung cancer has been widely characterized through radiomics and artificial intelligence (AI). This review aims to summarize the published studies of AI based on positron emission tomography/computed tomography (PET/CT) radiomics in non-small-cell lung cancer (NSCLC).

Materials and methods

A comprehensive search of literature published between 2012 and 2022 was conducted on the PubMed database. There were no language or publication status restrictions on the search. About 127 articles in the search results were screened and gradually excluded according to the exclusion criteria. Finally, this review included 39 articles for analysis.

Results

Classification is conducted according to purposes and several studies were identified at each stage of disease:1) Cancer detection (n=8), 2) histology and stage of cancer (n=11), 3) metastases (n=6), 4) genotype (n=6), 5) treatment outcome and survival (n=8). There is a wide range of heterogeneity among studies due to differences in patient sources, evaluation criteria and workflow of radiomics. On the whole, most models show diagnostic performance comparable to or even better than experts, and the common problems are repeatability and clinical transformability.

Conclusion

AI-based PET/CT Radiomics play potential roles in NSCLC clinical management. However, there is still a long way to go before being translated into clinical application. Large-scale, multi-center, prospective research is the direction of future efforts, while we need to face the risk of repeatability of radiomics features and the limitation of access to large databases.

Predictive Value of Baseline FDG-PET/CT for the Durable Response to Immune Checkpoint Inhibition in NSCLC Patients Using the Morphological and Metabolic Features of Primary Tumors

Objectives: We aimed to investigate the predictive value of baseline 2-deoxy-2-[18F]fluoro-D-glucose positron emission tomography/computed tomography (FDG-PET/CT) for durable responses to immune checkpoint inhibitors (ICIs) by linking the morphological and metabolic features of primary tumors (PTs) in nonsmall cell lung cancer (NSCLC) patients. Methods: For the purpose of this single-center study, the imaging data of the patients with a first diagnosis of NSCLC and an available baseline FDG-PET/CT between 2020 and 2021 were retrospectively assessed. The baseline characteristics were collected based on clinical reports and interdisciplinary tumor board documentation. The metabolic (such as standardized uptake value SUV maximum and mean (SUVmax, SUV mean), metabolic tumor volume (MTV), total lesion glycolysis (TLG)) and morphological (such as volume, morphology, margin, and presence of lymphangiosis through imaging) features of all the PTs were retrospectively assessed using FDG-PET/CT. Overall survival (OS), progression-free survival (PFS), clinical benefit (CB) and mortality rate were used as endpoints to define the long-term response to therapy. A backward, stepwise logistic regression analysis was performed in order to define the best model for predicting lasting responses to treatment. Statistical significance was assumed at p < 0.05. Results: A total of 125 patients (median age ± standard deviation (SD) 72.0 ± 9.5 years) were enrolled: 64 men (51.2%) and 61 women (48.8%). Adenocarcinoma was by far the most common histological subtype of NSCLC (47.2%). At the initial diagnosis, the vast majority of all the included patients showed either locally advanced disease (34.4%) or metastatic disease (36.8%). Fifty patients were treated with ICIs either as a first-line (20%) or second-line (20%) therapy, while 75 patients did not receive ICIs. The median values ± SD of PT SUVmax, mean, MTV, and TLG were respectively 10.1 ± 6.0, 6.1 ± 3.5, 13.5 ± 30.7, and 71.4 ± 247.7. The median volume of PT ± SD was 13.7 ± 30.7 cm3. The PTs were most frequently solid (86.4%) with irregular margins (76.8%). Furthermore, in one out of five cases, the morphological evidence of lymphangiosis was seen through imaging (n = 25). The median follow-up ± SD was 18.93 ± 6.98 months. The median values ± SD of OS and PFS were, respectively, 14.80 ± 8.68 months and 14.03 ± 9.02 months. Age, PT volume, SUVmax, TLG, the presence of lymphangiosis features through imaging, and clinical stage IV were very strong long-term outcome predictors of patients treated with ICIs, while no significant outcome predictors could be found for the cohort with no ICI treatment. The optimal cut-off values were determined for PT volume (26.94 cm3) and SUVmax (15.05). Finally, 58% of NSCLC patients treated with ICIs had a CB vs. 78.7% of patients in the cohort with no ICI treatment. However, almost all patients treated with ICIs and with disease progression over time died (mortality in the case of disease progression 95% vs. 62.5% in the cohort without ICIs). Conclusion: Baseline FDG-PET/CT could be used to predict a durable response to ICIs in NSCLC patients. Age, clinical stage IV, lymphangiosis features through imaging, PT volume (thus PT MTV due to a previously demonstrated linear correlation), PT SUVmax, and TLG were very strong long-term outcome predictors. Our results highlight the importance of linking clinical data, as much as morphological features, to the metabolic parameters of primary tumors in a multivariate outcome-predicting model using baseline FDG-PET/CT.

Correlation of Transcriptomics and FDG-PET SUVmax Indicates Reciprocal Expression of Stemness-Related Transcription Factor and Neuropeptide Signaling Pathways in Glucose Metabolism of Ewing Sarcoma

BACKGROUND

In Ewing sarcoma (EwS), long-term treatment effects and poor survival rates for relapsed or metastatic cases require individualization of therapy and the discovery of new treatment methods. Tumor glucose metabolic activity varies significantly between patients, and FDG-PET signals have been proposed as prognostic factors. However, the biological basis for the generally elevated but variable glucose metabolism in EwS is not well understood.

METHODS

We retrospectively included 19 EwS samples (17 patients). Affymetrix gene expression was correlated with maximal standardized uptake value (SUVmax) using machine learning, linear regression modelling, and gene set enrichment analyses for functional annotation.

RESULTS

Expression of five genes correlated (MYBL2, ELOVL2, NETO2) or anticorrelated (FAXDC2, PLSCR4) significantly with SUVmax (adjusted p-value ≤ 0.05). Additionally, we identified 23 genes with large SUVmax effect size, which were significantly enriched for "neuropeptide Y receptor activity (GO:0004983)" (adjusted p-value = 0.0007). The expression of the members of this signaling pathway (NPY, NPY1R, NPY5R) anticorrelated with SUVmax. In contrast, three transcription factors associated with maintaining stemness displayed enrichment of their target genes with higher SUVmax: RNF2, E2F family, and TCF3.

CONCLUSION

Our large-scale analysis examined comprehensively the correlations between transcriptomics and tumor glucose utilization. Based on our findings, we hypothesize that stemness may be associated with increased glucose uptake, whereas neuroectodermal differentiation may anticorrelate with glucose uptake.

The prognostic relationship of 18F-FDG PET/CT metabolic and volumetric parameters in metastatic ALK + NSCLC

OBJECTIVE

The aim of this study is to determine the role of metabolic and volumetric parameters obtained from 18Fluorine-Fluorodeoxyglucose PET/computed tomography (18F-FDG PET/CT) imaging on progression-free survival (PFS) and overall survival (OS) in patients with advanced nonsquamous cell lung carcinoma (NSCLC) with anaplastic lymphoma kinase (ALK) rearrangement.

METHODS

Pre and post-treatment PET/CT images of the ALK + NSCLC patients between January 2015 and July 2020 were evaluated. The highest standardized uptake value (SUVmax), metabolic tumor volume (MTV) and total lesion glycolysis (TLG) values were obtained from pre-tyrosine kinase inhibitor (TKI) basal PET/CT (PETpre) and post-TKI PET/CT (PETpost) images. Total MTV (tMTV) and total TLG (tTLG) values were calculated by summing MTV and TLG values in all tumor foci. The change (Δ) in pSUVmax, pMTV, pTLG, tMTV and tTLG before and after treatment was calculated.The relationship of these parameters with OS and PFS was analyzed.

RESULTS

tTLGpre, tMTVpre, pTLGpre, pMTVpre, ∆SUVmax, ∆tMTV and ∆tTLG values were found to be associated with OS; ∆tMTV, ∆tTLG, tTLGpre, tMTVpre, pTLGpre and pMTVpre were associated with PFS. The cutoff values in both predicting OS and PFS were calculated as -31.6 and 391.1 for ∆tMTV and tTLGpre, respectively. In Cox regression analysis, ∆tMTV and stage for OS and ∆tMTV and tTLGpre for PFS were obtained as prognostic factors.

CONCLUSIONS

Metabolic and volumetric parameters, especially TLG values in the whole body before treatment and change in whole body MTV value, obtained from PET/CT may be useful in predicting prognosis and determining treatment strategies for patients with advanced ALK + NSCLC.

Consistency and prognostic value of preoperative staging and postoperative pathological staging using 18F-FDG PET/MRI in patients with non-small cell lung cancer

OBJECTIVE

In recent years, positron emission tomography/magnetic resonance imaging (PET/MRI) has been clinically used as a method to diagnose non-small cell lung cancer (NSCLC). This study aimed to evaluate the concordance of staging and prognostic ability of NSCLC patients using thin-slice computed tomography (CT) and ¹⁸F-fluorodeoxyglucose (FDG) PET/MRI.

METHODS

This retrospective study was performed on consecutive NSCLC patients who underwent both diagnostic CT and ¹⁸F-FDG PET/MRI before surgery between November 2015 and May 2019. The cTNM staging yielded from PET/MRI was compared with CT and pathological staging, and concordance was investigated, defining pathological findings as reference. To assess the prognostic value of disease-free survival (DFS) and overall survival (OS), we dichotomized the typical prognostic factors and TNM classification staging (Stage I vs. Stage II or higher). Kaplan-Meier curves derived by the log-rank test were generated, and univariate and multivariate analyses were performed to identify the factors associated with DFS and OS.

RESULTS

A total of 82 subjects were included; PET/MRI staging was more consistent (59 of 82) with pathological staging than with CT staging. There was a total of 21 cases of CT and 11 cases of PET/MRI that were judged as cStage I, but were actually pStage II or pStage III. CT tended to judge pN1 or pN2 as cN0 compared to PET/MRI. There was a significant difference between NSCLC patients with Stage I and Stage II or higher by PET/MRI staging as well as prognosis prediction of DFS by pathological staging (P < 0.001). In univariate analysis, PET/MRI, CT, and pathological staging (Stage I or lower vs. Stage II or higher) all showed significant differences as prognostic factors of recurrence or metastases. In multivariate analysis, pathological staging was the only independent factor for recurrence (P = 0.009), and preoperative PET/MRI staging was a predictor of patient survival (P = 0.013).

CONCLUSIONS

In NSCLC, pathologic staging was better at predicting recurrence, and preoperative PET/MRI staging was better at predicting survival. Preoperative staging by PET/MRI was superior to CT in diagnosing hilar and mediastinal lymph-node metastases, which contributed to the high concordance with pathologic staging.

Principal component analysis of texture features derived from FDG PET images of melanoma lesions

Background

The clinical utility of radiomics is hampered by a high correlation between the large number of features analysed which may result in the “bouncing beta” phenomenon which could in part explain why in a similar patient population texture features identified and/or cut-off values of prognostic significance differ from one study to another. Principal component analysis (PCA) is a technique for reducing the dimensionality of large datasets containing highly correlated variables, such as texture feature datasets derived from FDG PET images, increasing data interpretability whilst at the same time minimizing information loss by creating new uncorrelated variables that successively maximize variance. Here, we report on PCA of a texture feature dataset derived from 123 malignant melanoma lesions with a significant range in lesion size using the freely available LIFEx software.

Results

Thirty-eight features were derived from all lesions. All features were standardized. The statistical assumptions for carrying out PCA analysis were met. Seven principal components with an eigenvalue > 1 were identified. Based on the “elbow sign” of the Scree plot, only the first five were retained. The contribution to the total variance of these components derived using Varimax rotation was, respectively, 30.6%, 23.6%, 16.1%, 7.4% and 4.1%. The components provided summarized information on the locoregional FDG distribution with an emphasis on high FDG uptake regions, contrast in FDG uptake values (steepness), tumour volume, locoregional FDG distribution with an emphasis on low FDG uptake regions and on the rapidity of changes in SUV intensity between different regions.

Conclusions

PCA allowed to reduce the dataset of 38 features to a set of 5 uncorrelated new variables explaining approximately 82% of the total variance contained within the dataset. These principal components may prove more useful for multiple regression analysis considering the relatively low numbers of patients usually included in clinical trials on FDG PET texture analysis. Studies assessing the superior differential diagnostic, predictive or prognostic value of principal components derived using PCA as opposed to the initial texture features in clinical relevant settings are warranted.

Convolutional neural networks for automatic image quality control and EARL compliance of PET images

Background

Machine learning studies require a large number of images often obtained on different PET scanners. When merging these images, the use of harmonized images following EARL-standards is essential. However, when including retrospective images, EARL accreditation might not have been in place. The aim of this study was to develop a convolutional neural network (CNN) that can identify retrospectively if an image is EARL compliant and if it is meeting older or newer EARL-standards.

Materials and methods

96 PET images acquired on three PET/CT systems were included in the study. All images were reconstructed with the locally clinically preferred, EARL1, and EARL2 compliant reconstruction protocols. After image pre-processing, one CNN was trained to separate clinical and EARL compliant reconstructions. A second CNN was optimized to identify EARL1 and EARL2 compliant images. The accuracy of both CNNs was assessed using fivefold cross-validation. The CNNs were validated on 24 images acquired on a PET scanner not included in the training data. To assess the impact of image noise on the CNN decision, the 24 images were reconstructed with different scan durations.

Results

In the cross-validation, the first CNN classified all images correctly. When identifying EARL1 and EARL2 compliant images, the second CNN identified 100% EARL1 compliant and 85% EARL2 compliant images correctly. The accuracy in the independent dataset was comparable to the cross-validation accuracy. The scan duration had almost no impact on the results.

Conclusion

The two CNNs trained in this study can be used to retrospectively include images in a multi-center setting by, e.g., adding additional smoothing. This method is especially important for machine learning studies where the harmonization of images from different PET systems is essential.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40658-022-00468-w.

Correlation between apparent diffusion coefficient and tumor-stroma ratio in hybrid 18F-FDG PET/MRI: preliminary results of a rectal cancer cohort study

Background

To explore possible correlations between the tumor-stroma ratio (TSR) and different imaging features of fluorine-18-fluorodeoxyglucose positron emission tomography/magnetic resonance imaging (¹⁸F-FDG PET/MRI) in untreated rectal cancer patients.

Methods

A patients with rectal cancer were included in this study. All participants were examined preoperatively with whole-body ¹⁸F-FDG PET/MRI. Two pathologists evaluated the TSR of tumors together. Apparent diffusion coefficient (ADC) values and PET-related parameters of the primary lesions were measured and compared between the stroma-high and stroma-low groups. Pearson's correlation or Spearman's rank correlation were used to evaluate the correlation between the ADC values, PET-related parameters, and pathological indices.

Results

Our results showed that in the untreated rectal cancer patients, the ADC mean values correlated with the TSR (r=0.327; P=0.007), and stroma-high (low TSR) rectal cancer corresponded to relatively lower ADC mean values (813.54±88.68 vs. 879.92±133.18; P=0.018). The ADC mean and ADC minimum (ADCmin) values were found to be negatively correlated with the pathological T stages (r=-0.384, P=0.001; r=-0.416, P=0.001, respectively) as well as the largest tumor diameters (r=-0.340, P=0.005; r=-0.314, P=0.010, respectively) of rectal cancer. In addition, the pathological T stages correlated with all PET-related metabolic parameters, including mean standard uptake value (SUV), maximum SUV (SUVmax), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) (r=0.338, P=0.006; r=0.350, P=0.004; r=0.326, P=0.007; and r=0.472, P<0.001, respectively). Our results also identified associations between the ADCmin values and SUVmean, SUVmax, and TLG (r=-0.335, P=0.006; r=-0.343, P=0.005; and r=-0.343, P=0.005, respectively). However, there were no statistical correlations between the PET/MRI parameters and the immunohistochemical (IHC) results.

Conclusions

This study indicated that the intratumoral heterogeneity measured by PET/MRI may reflect characteristics of the tumor microenvironment. Hence, PET/MRI parameters might be helpful in predicting tumor aggressiveness and prognosis.

Prognostic impact of an integrative analysis of [18F]FDG PET parameters and infiltrating immune cell scores in lung adenocarcinoma

Background

High levels of ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) tumor uptake are associated with worse prognosis in patients with non-small cell lung cancer (NSCLC). Meanwhile, high levels of immune cell infiltration in primary tumor have been linked to better prognosis in NSCLC. We conducted this study for precisely stratified prognosis of the lung adenocarcinoma patients using the integration of ¹⁸F-FDG positron emission tomography (PET) parameters and infiltrating immune cell scores as assessed by a genomic analysis.

Results

Using an RNA sequencing dataset, the patients were divided into three subtype groups. Additionally, 24 different immune cell scores and cytolytic scores (CYT) were obtained. In ¹⁸F-FDG PET scans, PET parameters of the primary tumors were obtained. An ANOVA test, a Chi-square test and a correlation analysis were also conducted. A Kaplan–Meier survival analysis with the log-rank test and multivariable Cox regression test was performed to evaluate prognostic values of the parameters. The terminal respiratory unit (TRU) group demonstrated lower ¹⁸F-FDG PET parameters, more females, and lower stages than the other groups. Meanwhile, the proximal inflammatory (PI) group showed a significantly higher CYT score compared to the other groups (P = .001). Also, CYT showed a positive correlation with tumor-to-liver maximum standardized uptake value ratio (TLR) in the PI group (P = .027). A high TLR (P = .01) score of ¹⁸F-FDG PET parameters and a high T follicular helper cell (TFH) score (P = .005) of immune cell scores were associated with prognosis with opposite tendencies. Furthermore, TLR and TFH were predictive of overall survival even after adjusting for clinicopathologic features and others (P = .024 and .047).

Conclusions

A high TLR score was found to be associated with worse prognosis, while high CD8 T cell and TFH scores predicted better prognosis in lung adenocarcinoma. Furthermore, TLR and TFH can be used to predict prognosis independently in patients with lung adenocarcinoma.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13550-022-00908-9.

Healthy Organs Uptake on Baseline 18F-FDG PET/CT as an Alternative to Total Metabolic Tumor Volume to Predict Event-Free Survival in Classical Hodgkin's Lymphoma

Purpose

Healthy organs uptake, including cerebellar and liver SUVs have been reported to be inversely correlated to total metabolic tumor volume (TMTV), a controversial predictor of event-free survival (EFS) in classical Hodgkin's Lymphoma (cHL). The objective of this study was to estimate TMTV by using healthy organs SUV measurements and assess the performance of this new index (UF, Uptake Formula) to predict EFS in cHL.

Methods

Patients with cHL were retrospectively included. SUV values and TMTV derived from baseline 18F-FDG PET/CT were harmonized using ComBat algorithm across PET/CT systems. UF was estimated using ANOVA analysis. Optimal thresholds of TMTV and UF were calculated and tested using Cox models.

Results

163 patients were included. Optimal UF model of TMTV included age, lymphoma maximum SUVmax, hepatic SUVmean and cerebellar SUVmax (R2 14.0% - p &lt; 0.001). UF &gt; 236.8 was a significant predictor of EFS (HR: 2.458 [1.201–5.030], p = 0.01) and was not significantly different from TMTV &gt; 271.0 (HR: 2.761 [1.183–5.140], p = 0.001). UF &gt; 236.8 remained significant in a bivariate model including IPS score (p = 0.02) and determined two populations with different EFS (63.7 vs. 84.9%, p = 0.01).

Conclusion

The Uptake Formula, a new index including healthy organ SUV values, shows similar performance to TMTV in predicting EFS in Hodgkin's Lymphoma. Validation cohorts will be needed to confirm this new prognostic parameter.

Respiratory gated multistatic PET reconstructions to delineate radiotherapy target volume in patients with mobile lung tumors

BACKGROUND

PET-CT with ¹⁸F-FDG or other radiopharmaceuticals is a recommended tool to help the delineation of lung cancers candidate to radiotherapy. The motion artifacts caused by respiratory movements are reduced by 4D acquisitions. We introduced an extended reconstruction algorithm (multiple reconstruct register and average [multi-RRA]) which requires much shorter acquisition times than standard 4D PET-CT. Our aim was to evaluate the interest on multi-RRA images as an alternative of 3D and 4D PET-CT for the delineation of lung lesion.

METHODS

PET acquisitions synchronized to the respiratory signal were obtained in 18 patients with mobile lung tumors. We compared the tumor volumes delineated on Multi-RRA images to 3D and 4D PET-CT, considering the 4D CT as a reference. The tumor volumes were delineated and compared with topologic similarity indexes (Dice, Jaccard and overlap).

RESULTS

Twenty tumors were delineated. The volumes delineated with multi-RRA and 4D PET were not significantly different (mean difference of 0.2±0.7 mL). Comparison by pairs (Tukey-Kramer test) showed that 3D-PET volumes were significantly smaller than 4D-PET and multi-RRA volumes (P<0.001). Topologic similarity indexes with 4D-PET were slightly statistically higher with multi-RRA than with 3D-PET (Dice and Jaccard) or 4D-CT (Dice, Jaccard and Overlap).

CONCLUSIONS

The tumor volumes delineated on multi-RRA are similar to the volumes obtained with 4D PET, with shorter acquisition time.

The value of the Standardized Uptake Value (SUV) and Metabolic Tumor Volume (MTV) in lung cancer

The diagnosis, staging and therapeutic monitoring of lung cancer were amongst the first applications for which the utility of FDG PET was documented and FDG PET/CT is now a routine diagnostic tool for clinical decision-making. As well as having high sensitivity for detection of disease sites, which provides critical information about stage, the intensity of uptake provides deeper biological characterization, while the burden of disease also has potential clinical significance. These disease characteristics can easily be quantified on delayed whole-body imaging as the maximum standardized uptake value (SUV_max) and metabolic tumor volume (MTV), respectively. There have been significant efforts to harmonize the measurement of these features, particularly within the context of clinical trials. Nevertheless, however calculated, in general, a high SUV_max and large MTV have been shown to have an adverse prognostic significance. Nevertheless, the use of these parameters in the interpretation and reporting of clinical scans remains inconsistent and somewhat controversial. This review details the current status of semi-quantitative FDG PET/CT in the evaluation of lung cancer.

T Stage and Pretreatment SUV Predict for Tumor Recurrence with 5 Fraction SAbR in Early-Stage Non-Small Cell Lung Cancer

Purpose

Five-fraction stereotactic ablative radiotherapy (SABR) regimens are frequently used to treat centrally located early-stage non-small cell lung cancer or disease in the proximity of the chest wall as a means of optimizing tumor control and reducing treatment toxicity. However, increasing these SABR regimens to 5 fractions may reduce tumor control outcomes. We sought to identify the clinical parameters predictive of treatment failures with these 5-fraction courses.

Methods

Ninety patients with T1-2 non-small cell lung cancer were treated with 50 or 60 Gy in 5 fractions. Failure over time was modeled using cumulative incidences of local, regional, or distant failure, with death as a competing risk. Cox proportional hazards analysis for incidences of failure was performed to control for patient variables.

Results

Of 90 patients, 24 of 53 patients with T1 tumors and 19 of 37 patients with T2 tumors received 50 Gy SABR, and the other 47 patients received 60 Gy. Two-year overall survival and progression-free survival for the whole cohort were 75.8% and 59.3%, respectively. Total SABR dose (50 vs 60 Gy) did not influence survival nor failure rates at 2 and 5 years. Within 2 years of treatment, 7.8% of all patients developed local failure. For all patient and tumor characteristics evaluated, only T stage and pretreatment positron emission tomography standardized uptake values served as predictors of local, regional, and distant failure at 2 and 5 years posttreatment on univariate and multivariable analysis.

Conclusions

Five-fraction SABR provides excellent in-field control. T2 and high fluorodeoxyglucose uptake tumors have increased failure rates, suggesting the potential need for adjuvant therapies, which are being assessed in randomized phase 3 trials.

A Review of the Correlation Between Epidermal Growth Factor Receptor Mutation Status and 18F-FDG Metabolic Activity in Non-Small Cell Lung Cancer

PET/CT with 18F-2-fluoro-2-deoxyglucose (18F-FDG) has been proposed as a promising modality for diagnosing and monitoring treatment response and evaluating prognosis for patients with non-small cell lung cancer (NSCLC). The status of epidermal growth factor receptor (EGFR) mutation is a critical signal for the treatment strategies of patients with NSCLC. Higher response rates and prolonged progression-free survival could be obtained in patients with NSCLC harboring EGFR mutations treated with tyrosine kinase inhibitors (TKIs) when compared with traditional cytotoxic chemotherapy. However, patients with EGFR mutation treated with TKIs inevitably develop drug resistance, so predicting the duration of resistance is of great importance for selecting individual treatment strategies. Several semiquantitative metabolic parameters, e.g., maximum standard uptake value (SUVmax), metabolic tumor volume (MTV), and total lesion glycolysis (TLG), measured by PET/CT to reflect 18F-FDG metabolic activity, have been demonstrated to be powerful in predicting the status of EGFR mutation, monitoring treatment response of TKIs, and assessing the outcome of patients with NSCLC. In this review, we summarize the biological and clinical correlations between EGFR mutation status and 18F-FDG metabolic activity in NSCLC. The metabolic activity of 18F-FDG, as an extrinsic manifestation of NSCLC, could reflect the mutation status of intrinsic factor EGFR. Both of them play a critical role in guiding the implementation of treatment modalities and evaluating therapy efficacy and outcome for patients with NSCLC.

Prognostic value of 18F-FDG PET/CT and MRI features in patients with high-risk and very-high-risk cutaneous squamous cell carcinoma

Objective:

This study aimed to evaluate the prognostic value of 18F-fludeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) and MRI features in patients with high-risk and very-high-risk cutaneous squamous cell carcinoma (cSCC).

Methods:

This study included 54 consecutive patients with surgically resected primary high-risk and very-high-risk cSCC who underwent pre-operative FDG-PET/CT and/or MRI. Among them, 14 patients (26%) had recurrences. We retrospectively reviewed the FDG-PET/CT (n = 34) and MRI (n = 48) and investigated the clinical significance and prognostic value of imaging features in cSCC.

Results:

On FDG-PET/CT, the maximum standardized uptake value (SUVmax) of the primary tumor (13.0 ± 6.4 vs. 6.9 ± 5.3, p < 0.05) was higher in cSCC with recurrence than in cSCC without recurrence. On MRI, the maximum diameter of the lesion (46.8 ± 24.1 mm vs 30.4 ± 17.0 mm, p < 0.05) and the frequency of muscle/tendon/bone invasion (42% vs 11%, p < 0.05) were significantly greater in cSCC with recurrence than in cSCC without recurrence. In the univariate analysis, prognostic factors for recurrence were SUVmax of the primary tumor (p < 0.01), the maximum diameter of the lesion (p < 0.05), and depth of invasion (p < 0.05). The areas under the receiver operating characteristic curves of the SUVmax (0.78) were superior to those of the maximum diameter (0.71) and depth of invasion (0.60).

Conclusion:

SUVmax, maximum diameter, and depth of invasion were useful parameters for prognostic factors predicting recurrence in patients with high-risk and very-high-risk cSCC.

Advances in knowledge:

SUVmax represents a prognostic factor.

Correlation of preoperative PET/computer tomography 18F-fluorodeoxyglucose uptake (maximum standardized uptake value) with prognosis in patients with operated lung cancer

PURPOSE

The aim of this study was to investigate the correlation of preoperative 18F-fluorodeoxyglucose PET/computed tomography maximum standardized uptake value (SUVmax) in operated non-small cell lung cancer (NSCLC) cases with other prognostic parameters and survival.

PATIENTS AND METHODS

NSCLC patients treated by surgical resection were imaged with PET within 60 days before surgery.

RESULTS

Overall, 525 cases consecutive patients were retrospectively reviewed. The median value of SUVmax in a total of 525 cases was 12.1, and the mean was 13.3 ± 7.13. Logistic regression analysis performed to identify the variables that have an impact on SUVmax revealed that histology [hazard ratio (HR: 1.893; 95% CI; P = 0.001) and T status (HR: 8.991; 95% CI; P = 0.000) are correlated with SUVmax. Kaplan-Meier analysis revealed a mean survival of 73.7 ± 1.95 months and a median survival of 85.6 ± 6.03 months. In the group with an SUVmax value of less than 10, the mean survival was 81.9 ± 3.02 months (76.0-87.8), and in the group with SUVmax greater than 10.1, the mean survival was 68.6 ± 2.4 months (63.9-73.3) (P = 0.000). In the multivariate analysis, SUVmax, age, tumor histology, lymph node metastasis, comorbid diseases and complete/incomplete status of the resection were identified as the factors predictive of prognosis.

CONCLUSION

It is seen that preoperative SUVmax is a parameter with prognostic significance at least as much as histopathology, age, complete/incomplete status of resection and lymph node involvement.

SUVmax to tumor perimeter distance: a robust radiomics prognostic biomarker in resectable non-small cell lung cancer patients

OBJECTIVE

The purpose of this study was to evaluate the prognostic value of novel geometric variables obtained from pre-treatment [¹⁸F]FDG PET/CT with respect to classical ones in patients with non-small cell lung cancer (NSCLC).

METHODS

Retrospective study including stage I-III NSCLC patients with baseline [¹⁸F]FDG PET/CT. Clinical, histopathologic, and metabolic parameters were obtained. After tumor segmentation, SUV and volume-based variables, global texture, sphericity, and two novel parameters, normalized SUVpeak to centroid distance (nSCD) and normalized SUVmax to perimeter distance (nSPD), were obtained. Early recurrence (ER) and short-term mortality (STM) were used as end points. Univariate logistic regression and multivariate logistic regression with respect to ER and STM were performed.

RESULTS

A cohort of 173 patients was selected. ER was detected in 49/104 of patients with recurrent disease. Additionally, 100 patients died and 53 had STM. Age, pathologic lymphovascular invasion, lymph nodal infiltration, TNM stage, nSCD, and nSPD were associated with ER, although only age (aOR = 1.06, p = 0.002), pathologic lymphovascular invasion (aOR = 3.40, p = 0.022), and nSPD (aOR = 0.02, p = 0.018) were significant independent predictors of ER in multivariate analysis. Age, lymph nodal infiltration, TNM stage, nSCD, and nSPD were predictors of STM. Age (aOR = 1.05, p = 0.006), lymph nodal infiltration (aOR = 2.72, p = 0.005), and nSPD (aOR = 0.03, p = 0.022) were significantly associated with STM in multivariate analysis. Coefficient of variation (COV) and SUVmean/SUVmax ratio did not show significant predictive value with respect to ER or STM.

CONCLUSION

The geometric variables, nSCD and nSPD, are robust biomarkers of the poorest outcome prediction of patients with NSCLC with respect to classical PET variables.

KEY POINTS

• In NSCLC patients, it is crucial to find prognostic parameters since TNM system alone cannot explain the variation in lung cancer survival. • Age, lymphovascular invasion, lymph nodal infiltration, and metabolic geometrical parameters were useful as prognostic parameters. • The displacement grade of the highest point of metabolic activity towards the periphery assessed by geometric variables obtained from [¹⁸F]FDG PET/CT was a robust biomarker of the poorest outcome prediction of patients with NSCLC.

Immunohistochemical expression of VEGFR1 in non small cell lung carcinomas: Lower VEGFR1 expression is asociated with squamous cell carcinoma subtype and high SUV max values in 18F-FDG PET

BACKGROUND

To study the possible relation between immunohistochemical expression of vascular endothelial growth factor receptor 1 (VEGFR1) and the maximum standardised uptake value (SUV max) of ¹⁸F-FDG PET in patients with non small cell lung cancer (NSCLC).

MATERIAL AND METHODS

The study included 39 patients with NSCLC (24 squamous cell carcinomas and 15 adenocarcinomas). According to the clinical stage, the patients were distributed as follows: 8 stage I, 7 stage II, 15 stage III and 9 stage IV. Immunohistochemical expression of VEGFR1 was studied through the technique of tissue-matrix using Tissue Arrayer Device (Beecher Instruments, Sun Prairie, WI), using the polyclonal antibody against VEGFR1 (Santa Cruz Biotechnology, California, USA).

RESULTS

Positive VEGFR1 immunohistochemical expression was noted in 23 cases (59%). The number of positive tumours was not related with clinical stage but there was a different statistically significant association (p:0,0009) between VEGFR1 positivity and histological type, corresponding the greater percentages of positive results to adenocarcinomas (93,3%) versus in squamous cell carcinomas (37,5%). Likewise, SUV max values were higher (p: 0,039) in negative VEGFR1 carcinomas than in positive VEGFR1 tumors (r: 4-32,1; 16,4+/-6,4 (median 16,1) vs r: 3-47; 14,5+/-8,6 (12,8)).

CONCLUSIONS

Our results led us to consider that in NSCLC, the negative VEGFR1 immunohistochemical expression is associated significantly with squamous cell carcinomas subtype and with higher SUV max values in ¹⁸F-FDG-PET.

Heterogeneous radiological response to chemotherapy is associated with poor prognosis in advanced non-small-cell lung cancer

Background

A heterogeneous radiological response is frequently observed in cancer patients and could reflect tumor heterogeneity. We investigated the prognostic impact of heterogeneous radiological responses in patients with advanced non‐small‐cell lung cancer (NSCLC) who received platinum‐based chemotherapy.

Methods

The treatment response according to Response Evaluation Criteria in Solid Tumors (RECIST) 1.1 criteria was evaluated in 212 patients with advanced NSCLC who received platinum‐based chemotherapy. Patients with partial response (PR) or stable disease (SD) were classified into “PR homo,” “PR hetero,” “SD homo,” and “SD hetero” by the presence of a heterogeneous radiological response, and survival was compared between groups. We also compared survival based on the presence of metabolic responses in lesions with heterogeneous radiological responses.

Results

Fifty‐two patients (24.5%) were classified as PR, 112 patients (52.8%) as SD, and 48 patients (22.7%) as progressive disease (PD). There was no significant difference in progression‐free survival (PFS) and overall survival (OS) between the PR homo and PR hetero groups. The SD homo group had a longer PFS and OS than the SD hetero group. In the SD hetero group, patients with increased maximum standardized uptake value (SUVmax) in lesions with heterogeneous radiological responses had a shorter PFS than those with a stable SUVmax.

Conclusions

The presence of lesions with radiological heterogeneity was associated with disease progression and poor prognosis in the SD group. Patients with heterogeneous radiological responses require careful monitoring.

Weakly supervised segmentation of tumor lesions in PET-CT hybrid imaging

Purpose: We introduce and evaluate deep learning methods for weakly supervised segmentation of tumor lesions in whole-body fluorodeoxyglucose-positron emission tomography (FDG-PET) based solely on binary global labels (“tumor” versus “no tumor”).

Approach: We propose a three-step approach based on (i) a deep learning framework for image classification, (ii) subsequent generation of class activation maps (CAMs) using different CAM methods (CAM, GradCAM, GradCAM++, ScoreCAM), and (iii) final tumor segmentation based on the aforementioned CAMs. A VGG-based classification neural network was trained to distinguish between PET image slices with and without FDG-avid tumor lesions. Subsequently, the CAMs of this network were used to identify the tumor regions within images. This proposed framework was applied to FDG-PET/CT data of 453 oncological patients with available manually generated ground-truth segmentations. Quantitative segmentation performance was assessed for the different CAM approaches and compared with the manual ground truth segmentation and with supervised segmentation methods. In addition, further biomarkers (MTV and TLG) were extracted from the segmentation masks.

Results: A weakly supervised segmentation of tumor lesions was feasible with satisfactory performance [best median Dice score 0.47, interquartile range (IQR) 0.35] compared with a fully supervised U-Net model (median Dice score 0.72, IQR 0.36) and a simple threshold based segmentation (Dice score 0.29, IQR 0.28). CAM, GradCAM++, and ScoreCAM yielded similar results. However, GradCAM led to inferior results (median Dice score: 0.12, IQR 0.21) and was likely to ignore multiple instances within a given slice. CAM, GradCAM++, and ScoreCAM yielded accurate estimates of metabolic tumor volume (MTV) and tumor lesion glycolysis. Again, worse results were observed for GradCAM.

Conclusions: This work demonstrated the feasibility of weakly supervised segmentation of tumor lesions and accurate estimation of derived metrics such as MTV and tumor lesion glycolysis.

Clinical correlates of circulating cell-free DNA tumor fraction

BACKGROUND

Oncology applications of cell-free DNA analysis are often limited by the amount of circulating tumor DNA and the fraction of cell-free DNA derived from tumor cells in a blood sample. This circulating tumor fraction varies widely between individuals and cancer types. Clinical factors that influence tumor fraction have not been completely elucidated.

METHODS AND FINDINGS

Circulating tumor fraction was determined for breast, lung, and colorectal cancer participant samples in the first substudy of the Circulating Cell-free Genome Atlas study (CCGA; NCT02889978; multi-cancer early detection test development) and was related to tumor and patient characteristics. Linear models were created to determine the influence of tumor size combined with mitotic or metabolic activity (as tumor mitotic volume or excessive lesion glycolysis, respectively), histologic type, histologic grade, and lymph node status on tumor fraction. For breast and lung cancer, tumor mitotic volume and excessive lesion glycolysis (primary lesion volume scaled by percentage positive for Ki-67 or PET standardized uptake value minus 1.0, respectively) were the only statistically significant covariates. For colorectal cancer, the surface area of tumors invading beyond the subserosa was the only significant covariate. The models were validated with cases from the second CCGA substudy and show that these clinical correlates of circulating tumor fraction can predict and explain the performance of a multi-cancer early detection test.

CONCLUSIONS

Prognostic clinical variables, including mitotic or metabolic activity and depth of invasion, were identified as correlates of circulating tumor DNA by linear models that relate clinical covariates to tumor fraction. The identified correlates indicate that faster growing tumors have higher tumor fractions. Early cancer detection from assays that analyze cell-free DNA is determined by circulating tumor fraction. Results support that early detection is particularly sensitive for faster growing, aggressive tumors with high mortality, many of which have no available screening today.

Maximum standardized uptake value of the primary tumor does not improve candidate selection for sublobar resection

OBJECTIVE

This retrospective study examined whether adding the maximum standardized uptake value of a primary tumor to the consolidation-to-tumor ratio from a high-resolution computed tomography scan can improve the predictive accuracy for pathological noninvasive lung cancer and lead to better patient selection for sublobar resection.

METHODS

We included 926 patients with clinical stage IA non-small cell lung cancer. Pathological noninvasive cancer (n = 515) was defined as any case without lymphatic invasion, vascular invasion, or lymph node metastasis. The prediction accuracies of maximum standardized uptake value and consolidation-to-tumor ratio were evaluated using receiver operating characteristic curves and area under the curve.

RESULTS

For consolidation-to-tumor ratio or maximum standardized uptake value alone, the area under the curves were 0.733 (95% confidence interval, 0.708-0.758) and 0.842 (95% confidence interval, 0.816-0.866), respectively. When the consolidation-to-tumor ratio and maximum standardized uptake value were combined, the area under the curve was 0.854 (95% confidence interval, 0.829-0.876). However, to obtain a predictive specificity of 97%, sensitivity needed to be 42.5% for the consolidation-to-tumor ratio, 38.3% for the maximum standardized uptake value, and 45.0% for these 2 in combination.

CONCLUSIONS

Our results suggest that despite the high area under the curve for maximum standardized uptake value, caution is needed when using maximum standardized uptake value to select candidates for sublobar resection. We found that a low maximum standardized uptake value did not mean the tumor was a pathological noninvasive lung cancer. Therefore, using consolidation-to-tumor ratios from high-resolution computed tomography to decide whether sublobar resection is appropriate for patients with clinical stage IA non-small cell lung cancer is better than using maximum standardized uptake value when setting specificity to a conservative 97% for predicting pathological noninvasive lung cancer.

Imaging biomarkers for evaluating tumor response: RECIST and beyond

Response Evaluation Criteria in Solid Tumors (RECIST) is the gold standard for assessment of treatment response in solid tumors. Morphologic change of tumor size evaluated by RECIST is often correlated with survival length and has been considered as a surrogate endpoint of therapeutic efficacy. However, the detection of morphologic change alone may not be sufficient for assessing response to new anti-cancer medication in all solid tumors. During the past fifteen years, several molecular-targeted therapies and immunotherapies have emerged in cancer treatment which work by disrupting signaling pathways and inhibited cell growth. Tumor necrosis or lack of tumor progression is associated with a good therapeutic response even in the absence of tumor shrinkage. Therefore, the use of unmodified RECIST criteria to estimate morphological changes of tumor alone may not be sufficient to estimate tumor response for these new anti-cancer drugs. Several studies have reported the low reliability of RECIST in evaluating treatment response in different tumors such as hepatocellular carcinoma, lung cancer, prostate cancer, brain glioma, bone metastasis, and lymphoma. There is an increased need for new medical imaging biomarkers, considering the changes in tumor viability, metabolic activity, and attenuation, which are related to early tumor response. Promising imaging techniques, beyond RECIST, include dynamic contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI), diffusion-weight imaging (DWI), magnetic resonance spectroscopy (MRS), and ¹⁸ F-fluorodeoxyglucose (FDG) positron emission tomography (PET). This review outlines the current RECIST with their limitations and the new emerging concepts of imaging biomarkers in oncology.

Clinical Characteristics and Outcome of Pathologic N0 Non-small Cell Lung Cancer Patients With False Positive Mediastinal Lymph Node Metastasis on FDG PET-CT

BACKGROUND/AIM

Preoperative fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET-CT) is a non-invasive and useful diagnostic tool to evaluate mediastinal lymph node (LN) metastasis in lung cancer. However, there are often false-positive LN cases in FDG PET-CT. This study aimed to explore the clinical characteristics and outcome of pathologic N0 non-small cell lung cancer patients with false-positive mediastinal LN on FDG PET-CT.

PATIENTS AND METHODS

We enrolled 147 patients who underwent preoperative FDG PET-CT scan and mediastinal LN dissection. These patients were re-evaluated for post-operative pathologic nodal metastasis and divided into a false-positive group and a group of others.

RESULTS

Among 40 patients diagnosed with clinical N1-3 on FDG PET-CT, 19 (47.5%) patients were pathologic N0, meaning false-positive LN by PET-CT. Preoperative absolute platelet count and platelet-lymphocyte ratio were significantly higher in patients with pathologic N0. The presence of lymphatic invasion was significantly lower in patients with pathologic N0 than in the group of others. Recurrence-free survival was significantly shorter in patients with false positive LN than in patients with true positive LN or true negative LN at the same pathologic stage.

CONCLUSION

Higher absolute platelet count and PLR, lower proportion of lymphatic invasion and shorter recurrence-free survival were associated with false positive mediastinal LN on preoperative FDG PET-CT.