COVID-19 in patients with cancer: can baseline radiologic severity and early evolution predict clinical outcomes?

OBJECTIVE: To determine whether the degree of parenchymal involvement on chest radiograph (CXR) at the time of COVID-19 diagnosis and its early radiologic evolution can predict adverse events including hospitalization, intubation, and death in patients with cancer. METHODS: Retrospective study of 627 COVID-19-positive patients between March and April 2020, of which 248 had baseline CXR within 72 h of diagnosis and 64 patients had follow-up wihtin72 h. CXRs were classified as abnormal (i.e., radiologic findings suggestive of COVID-19 infection were noted), normal, or indeterminate. Baseline and follow-up severity scores were calculated based on lung regions in abnormal CXRs. Statistical analysis was performed to determine associations between abnormal CXR or severity score with adverse events. RESULTS: Of 248 patients (median age = 65) with a baseline CXR, 172/248 (69%) had an abnormal baseline study, which was associated with hospitalization (p < 0.001), intubation (p = 0.001), and death (p = 0.005). For patients with solid neoplasms, when adjusted for stage, it was associated with hospitalization (p = 0.0002), intubation (p = 0.019), and death (p = 0.03). The median baseline severity score was 3 (range = 1-10); the greater the score, the higher the likelihood of adverse outcome (p < 0.003 for all). A baseline severity score > 9 predicted > 50% probability of intubation and a score of ≥ 10 predicted > 50% of probability of death. The baseline severity score was not correlated with cancer-related treatment. Early radiologic progression was not correlated with hospitalization, intubation, or death. CONCLUSION: The degree of parenchymal involvement on CXR within 72 h of COVID-19 diagnosis is associated with adverse outcomes in patients with cancer. KEY POINTS: • In patients with cancer, the presence and severity of radiologic manifestation of COVID-19 on chest radiographs within 72 h of COVID-19 diagnosis are associated with hospitalization, intubation, and death. • Early radiologic progression on chest radiographs is not correlated with adverse outcomes.

Imaging Characteristics of Prostate Cancer Patients Who Discontinued Active Surveillance on 3-T Multiparametric Prostate MRI.

OBJECTIVE: Early-stage prostate cancer may be followed with active surveillance to avoid overtreatment. Our institution's active surveillance regimen uses annual MRI in place of serial biopsies, and biopsies are performed only when clinically necessary. The objective of our study was to report the multiparametric MRI characteristics of prostate cancer patients who discontinued active surveillance at our institution after repeat imaging revealed possible evidence of tumor upgrading. MATERIALS AND METHODS: The Department of Urology at Winthrop University Hospital prospectively maintains a database of prostate cancer patients who are monitored with active surveillance. At the time of this study, there were 200 prostate cancer patients being monitored with active surveillance. Of those patients, 114 patients had an initial multiparametric MRI study that was performed before active surveillance started and at least one follow-up multiparametric MRI study that was performed after active surveillance began. The MRI findings were evaluated and correlated with pathology results, if available. RESULTS: Fourteen patients discontinued active surveillance because changes on follow-up MRI suggested progression of cancer. Follow-up MRI showed an enlarged or more prominent lesion compared with the appearance on a previous MRI in three (21.4%) patients, a new lesion or lesions suspicious for cancer in two (14.3%) patients, and findings suspicious for or confirming extracapsular extension in nine (64.3%) patients. Seven of the 14 (50.0%) patients had a biopsy after follow-up multiparametric MRI, and biopsy results led to tumor upgrading in six of the 14 (42.9%) patients. The duration of active surveillance ranged from 4 to 110 months. All patients received definitive treatment. CONCLUSION: The small number of patients with follow-up multiparametric MRI findings showing worsening disease supports the role of MRI in patients with early-stage prostate cancer. Multiparametric MRI is useful in monitoring patients on active surveillance and may identify patients with clinically significant cancer amenable to definitive treatment.

Clinicopathologic outcomes of preoperative targeted therapy in patients with clinical stage I to III non-small cell lung cancer.

OBJECTIVE: Targeted therapy improves outcomes in patients with advanced-stage non-small cell lung cancer (NSCLC) and in the adjuvant setting, but data on its use before surgery are limited. We sought to investigate the safety and feasibility of preoperative targeted therapy in patients with operable NSCLC. METHODS: We retrospectively reviewed 51 patients with clinical stage I to III NSCLC who received targeted therapy, alone or in combination with chemotherapy, before surgical resection with curative intent, treated from 2004 to 2021. The primary outcome was the safety and feasibility of preoperative targeted therapy; secondary outcomes included objective response rate, major pathologic response (defined as ≤10% viable tumor) rate, recurrence-free survival (RFS), and overall survival. RESULTS: Of the 51 patients included, 46 had an activating epidermal growth factor receptor gene alteration and 5 had an anaplastic lymphoma kinase fusion. Overall, 37 of 46 evaluable patients experienced at least 1 adverse event before surgery; however, only 3 patients experienced a grade 3 or 4 event. The objective response rate was 38% (17/45) for all evaluable patients and 44% (14/32) for patients with clinical stage II or III disease. The major pathologic response rate was 20% (9/44); 2 patients had a complete pathologic response. Median RFS was 3.8 years (95% CI, 2.8 to not reached). Targeted therapy alone was associated with better RFS than combination therapy (P = .009) in patients with clinical stage II or III disease. CONCLUSIONS: Preoperative targeted therapy was well tolerated and associated with good outcomes, with or without induction chemotherapy. In addition, radiographic response and pathologic response were strongly correlated.

Developing a Cancer Digital Twin: Supervised Metastases Detection From Consecutive Structured Radiology Reports.

The development of digital cancer twins relies on the capture of high-resolution representations of individual cancer patients throughout the course of their treatment. Our research aims to improve the detection of metastatic disease over time from structured radiology reports by exposing prediction models to historical information. We demonstrate that Natural language processing (NLP) can generate better weak labels for semi-supervised classification of computed tomography (CT) reports when it is exposed to consecutive reports through a patient's treatment history. Around 714,454 structured radiology reports from Memorial Sloan Kettering Cancer Center adhering to a standardized departmental structured template were used for model development with a subset of the reports included for validation. To develop the models, a subset of the reports was curated for ground-truth: 7,732 total reports in the lung metastases dataset from 867 individual patients; 2,777 reports in the liver metastases dataset from 315 patients; and 4,107 reports in the adrenal metastases dataset from 404 patients. We use NLP to extract and encode important features from the structured text reports, which are then used to develop, train, and validate models. Three models-a simple convolutional neural network (CNN), a CNN augmented with an attention layer, and a recurrent neural network (RNN)-were developed to classify the type of metastatic disease and validated against the ground truth labels. The models use features from consecutive structured text radiology reports of a patient to predict the presence of metastatic disease in the reports. A single-report model, previously developed to analyze one report instead of multiple past reports, is included and the results from all four models are compared based on accuracy, precision, recall, and F1-score. The best model is used to label all 714,454 reports to generate metastases maps. Our results suggest that NLP models can extract cancer progression patterns from multiple consecutive reports and predict the presence of metastatic disease in multiple organs with higher performance when compared with a single-report-based prediction. It demonstrates a promising automated approach to label large numbers of radiology reports without involving human experts in a time- and cost-effective manner and enables tracking of cancer progression over time.

3-T multiparametric MRI characteristics of prostate cancer patients suspicious for biochemical recurrence after primary focal cryosurgery (hemiablation).

INTRODUCTION: We aimed to report on multiparametric MRI (mpMRI) characteristics of post-primary focal cryosurgery (PFC) patients suspected of biochemical recurrence (BCR) by the Phoenix criteria. METHODS: We retrospectively reviewed all patients at our institution who had undergone PFC. Prostate-specific antigen nadir was determined using 2 or more post-PFC values. Suspicion of BCR was determined using the Phoenix criteria (nadir + 2 ng/ml). At the discretion of the physician, pre-and post-PFC 3-T mpMRIs were obtained and in a few cases biopsies were performed. RESULTS: Ninety (58.4%) of 154 consecutive patients who underwent PFC were included in our analysis and had a median (range) age and prostate volume of 66.5 (48-82) years and 40.5 (16-175) ml, respectively. Of those suspected of BCR (37/90, 41.1%), with a median time to BCR of 19.9 (7.0-38.5) months, 27 patients (73.0%) underwent a post-PFC mpMRI. Twenty-two (81.5%) of these mpMRIs were found with 24 suspicious lesions. A considerable number (9/24, 37.5%) of these lesions were located in the central gland of the prostate. Seven of 24 lesions exhibited adverse mpMRI characteristic; 4 (16.7%) had capsular contact, 2 (8.2%) showed frank extracapsular extension, and 1 (4.2%) showed seminal vesicle invasion. Five (45.5%) of 11 patients with positive post-PFC mpMRIs were positive on biopsy (4/5, 80% were clinically significant prostate cancer). CONCLUSION: Post-PFC mpMRI, at Phoenix suspicion of BCR, may help identify a significant number of patients failing post-PFC.

Clinical consequences of an indeterminate CT pulmonary angiogram in cancer patients.

Our aim was to evaluate clinical management and outcomes in cancer patients who had an indeterminate Computed Tomographic Pulmonary Angiogram (CTPA) for the assessment of pulmonary embolus. We reviewed 1000 CTPA studies and identified 251 limited (indeterminate) CTPA. We examined follow-up imaging and reviewed clinical management decisions and any positive diagnosis of venous thromboembolic disease (VTE) within the subsequent 90 days. 60 patients (23.9%) had a follow-up imaging study within five days. 8 had a positive study for VTE disease within 5 days. 3 patients (1.2%) were placed on anticoagulation therapy based on the limited CT result.