Dataset Information

Comparison of Different Methods for Defining Hyperprogressive Disease in NSCLC.

ABSTRACT:

Introduction

Hyperprogressive disease (HPD) as a consequence of immune checkpoint inhibitors in NSCLC has been reported in multiple studies. However, inconsistent results in incidence and survival outcomes within studies, together with different assessment methods, have led to increasing controversy regarding the concept of HPD.

Methods

Consecutive patients treated with nivolumab (N = 42) or docetaxel (N = 37) were evaluated. HPD was quantified by applying three different methods (tumor growth rate [TGR], tumor growth kinetics [TGK], and Response Evaluation Criteria in Solid Tumors version 1.1 [RECIST 1.1]). HPD rates were compared between and within both cohorts using the different methods.

Results

Using TGR, TGK, and RECIST 1.1, we identified seven (16.7%), seven (16.7%), and six (14.3%) patients with HPD in the nivolumab cohort and three (8.1%), four (10.8%), and five (13.6%) in the docetaxel cohort, respectively. We observed a higher concordance between TGR and TGK (90.1%) compared with RECIST 1.1 (31.3% and 37.5% with TGR and TGK, respectively). We found no significant differences in the overall survival between patients with progressive disease and HPD in either cohort.

Conclusions

TGR and TGK revealed high concordance rates for identifying patients with HPD in NSCLC. The incidence of HPD was numerically higher in patients treated with immune checkpoint inhibitors. Standardization of methods for measuring HPD and its exploration in larger studies are needed to establish its clinical meaning in NSCLC.

SUBMITTER: Rocha P

PROVIDER: S-EPMC8474364 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BACKGROUND:Several neurodegeneration (N) metrics using structural MRI are used for the purpose of Alzheimer's disease (AD)-related staging, including hippocampal volume, global atrophy, and an "AD signature" composite consisting of thickness or volumetric estimates derived from regions impacted early in AD. This study sought to determine if less user-intensive estimates of global atrophy and hippocampal volume were equivalent to a thickness-based AD signature from FreeSurfer for defining N across the AD continuum (i.e., individuals who are amyloid-positive (A+)). METHODS:Cognitively unimpaired (CU) late middle-aged and older adults, as well as A+ mild cognitive impairment (MCI) and A+ AD dementia individuals, with available CSF and structural MRI scan <1.5 years apart, were selected for the study (n = 325, mean age = 62). First, in a subsample of A+ AD dementia and matched biomarker-negative (i.e., A- and tau tangle pathology (T)-) CU controls (n = 40), we examined ROC characteristics and identified N cut-offs using Youden's J for neurofilament light chain protein (NfL) and each of three MRI-based measures: a thickness-based AD signature from FreeSurfer, hippocampal volume (using FIRST), and a simple estimate of global atrophy (the ratio of intracranial CSF segmented volume to brain tissue volume, using SPM12). Based on the results from the ROC analyses, we then examined the concordance between NfL N positivity and N positivity for each MRI-based metric using Cohen's Kappa in the remaining subsample of 285 individuals. Finally, in the full sample (n = 325), we examined the relationship between the four measures of N and group membership across the AD continuum using Kruskal-Wallis tests and Cliff's deltas. RESULTS:The three MRI-based metrics and CSF NfL similarly discriminated between the A-T- CU (n = 20) and A+ AD (n = 20) groups (AUCs ≥0.885; ps < 0.001). Using the cut-off values derived from the ROCs to define N positivity, there was weak concordance between NfL and all three MRI-derived metrics of N in the subsample of 285 individuals (Cohen's Kappas ≤0.429). Finally, the three MRI-based measures of N and CSF NfL showed similar associations with AD continuum group (i.e., Kruskal-Wallis ps < 0.001), with relatively larger effect sizes noted when comparing the A-T- CU to the A+ MCI (Cliff's deltas ≥0.741) and A+ AD groups (Cliff's deltas ≥0.810) than to the A+T- CU (Cliff's deltas = 0.112-0.298) and A + T+ CU groups (Cliff's deltas = 0.212-0.731). CONCLUSIONS:These findings suggest that the three MRI-based morphometric estimates and CSF NfL similarly differentiate individuals across the AD continuum on N status. In many applications, a simple estimate of global atrophy may be preferred as an MRI marker of N across the AD continuum given its methodological robustness and ease of calculation when compared to hippocampal volume or a cortical thickness AD signature.

Project description:For in vitro culture of plant and animal cells, one of the critical steps is to adjust the initial cell density. A typical example of this is isolated microspore culture, where specific cell densities have been determined for different species. Out of these ranges, microspore growth is not induced, or is severely reduced. A similar situation occurs in many other plant and animal cell culture systems. Traditionally, researchers have used counting chambers (hemacytometers) to calculate cell densities, but little is still known about their technical advantages. In addition, much less information is available about other, alternative methods. In this work, using isolated eggplant microspore cultures and fluorescent beads (fluorospheres) as experimental systems, we performed a comprehensive comparison of six methods to calculate cell densities: (1) a Neubauer improved hemacytometer, (2) an automated cell counter, (3) a manual-counting method, and three flow cytometry methods based on (4) autofluorescence, (5) propidium iodide staining, and (6) side scattered light (SSC).Our results show that from a technical perspective, hemacytometers are the most reasonable option for cell counting, which may explain their widely spread use. Automated cell counters represent a good compromise between precision and affordability, although with limited accuracy. Finally, the methods based on flow cytometry were, by far, the best in terms of reproducibility and agreement between them, but they showed deficient accuracy and precision.Together, our results show a thorough technical evaluation of each counting method, provide unambiguous arguments to decide which one is the most convenient for the particular case of each laboratory, and in general, shed light into the best way to determine cell densities for in vitro cell cultures. They may have an impact in such a practice not only in the context of microspore culture, but also in any other plant cell culture procedure, or in any process involving particle counting.

Project description:BackgroundAutomated ICD coding on medical texts via machine learning has been a hot topic. Related studies from medical field heavily relies on conventional bag-of-words (BoW) as the feature extraction method, and do not commonly use more complicated methods, such as word2vec (W2V) and large pretrained models like BERT. This study aimed at uncovering the most effective feature extraction methods for coding models by comparing BoW, W2V and BERT variants.MethodsWe experimented with a Chinese dataset from Fuwai Hospital, which contains 6947 records and 1532 unique ICD codes, and a public Spanish dataset, which contains 1000 records and 2557 unique ICD codes. We designed coding tasks with different code frequency thresholds (denoted as [Formula: see text]), with a lower threshold indicating a more complex task. Using traditional classifiers, we compared BoW, W2V and BERT variants on accomplishing these coding tasks.ResultsWhen [Formula: see text] was equal to or greater than 140 for Fuwai dataset, and 60 for the Spanish dataset, the BERT variants with the whole network fine-tuned was the best method, leading to a Micro-F1 of 93.9% for Fuwai data when [Formula: see text], and a Micro-F1 of 85.41% for the Spanish dataset when [Formula: see text]. When [Formula: see text] fell below 140 for Fuwai dataset, and 60 for the Spanish dataset, BoW turned out to be the best, leading to a Micro-F1 of 83% for Fuwai dataset when [Formula: see text], and a Micro-F1 of 39.1% for the Spanish dataset when [Formula: see text]. Our experiments also showed that both the BERT variants and BoW possessed good interpretability, which is important for medical applications of coding models.ConclusionsThis study shed light on building promising machine learning models for automated ICD coding by revealing the most effective feature extraction methods. Concretely, our results indicated that fine-tuning the whole network of the BERT variants was the optimal method for tasks covering only frequent codes, especially codes that represented unspecified diseases, while BoW was the best for tasks involving both frequent and infrequent codes. The frequency threshold where the best-performing method varied differed between different datasets due to factors like language and codeset.