Continuation of fluoropyrimidine treatment with S-1 after cardiotoxicity on capecitabine- or 5-fluorouracil-based therapy in patients with solid tumours: a multicentre retrospective observational cohort study.

BACKGROUND: Capecitabine- or 5-fluorouracil (5-FU)-based chemotherapy is widely used in many solid tumours, but is associated with cardiotoxicity. S-1 is a fluoropyrimidine with low rates of cardiotoxicity, but evidence regarding the safety of switching to S-1 after 5-FU- or capecitabine-associated cardiotoxicity is scarce. PATIENTS AND METHODS: This retrospective study (NCT04260269) was conducted at 13 centres in 6 countries. The primary endpoint was recurrence of cardiotoxicity after switch to S-1-based treatment due to 5-FU- or capecitabine-related cardiotoxicity: clinically meaningful if the upper boundary of the 95% confidence interval (CI; by competing risk) is not including 15%. Secondary endpoints included cardiac risk factors, diagnostic work-up, treatments, outcomes, and timelines of cardiotoxicity. RESULTS: Per protocol, 200 patients, treated between 2011 and 2020 [median age 66 years (range 19-86); 118 (59%) males], were included. Treatment intent was curative in 145 (73%). Initial cardiotoxicity was due to capecitabine (n = 170), continuous infusion 5-FU (n = 22), or bolus 5-FU (n = 8), which was administered in combination with other chemotherapy, targeted agents, or radiotherapy in 133 patients. Previous cardiovascular comorbidities were present in 99 (50%) patients. Cardiotoxic events (n = 228/200) included chest pain (n = 125), coronary syndrome/infarction (n = 69), arrhythmia (n = 22), heart failure/cardiomyopathy (n = 7), cardiac arrest (n = 4), and malignant hypertension (n = 1). Cardiotoxicity was severe or life-threatening in 112 (56%) patients and led to permanent capecitabine/5-FU discontinuation in 192 (96%). After switch to S-1, recurrent cardiotoxicity was observed in eight (4%) patients (95% CI 2.02-7.89, primary endpoint met). Events were limited to grade 1-2 and occurred at a median of 16 days (interquartile range 7-67) from therapy switch. Baseline ischemic heart disease was a risk factor for recurrent cardiotoxicity (odds ratio 6.18, 95% CI 1.36-28.11). CONCLUSION: Switching to S-1-based therapy is safe and feasible after development of cardiotoxicity on 5-FU- or capecitabine-based therapy and allows patients to continue their pivotal fluoropyrimidine-based treatment.

Treatment response of colorectal cancer liver metastases to neoadjuvant or conversion therapy: a prospective multicentre follow-up study using MRI, diffusion-weighted imaging and 1H-MR spectroscopy compared with histology (subgroup in the RAXO trial).

BACKGROUND: Colorectal cancer liver metastases respond to chemotherapy and targeted agents not only by shrinking, but also by morphologic and metabolic changes. The aim of this study was to evaluate the value of advanced magnetic resonance imaging (MRI) methods in predicting treatment response and survival. PATIENTS AND METHODS: We investigated contrast-enhanced MRI, apparent diffusion coefficient (ADC) in diffusion-weighted imaging and 1H-magnetic resonance spectroscopy (1H-MRS) in detecting early morphologic and metabolic changes in borderline or resectable liver metastases, as a response to first-line neoadjuvant or conversion therapy in a prospective substudy of the RAXO trial (NCT01531621, EudraCT2011-003158-24). MRI findings were compared with histology of resected liver metastases and Kaplan-Meier estimates of overall survival (OS). RESULTS: In 2012-2018, 52 patients at four Finnish university hospitals were recruited. Forty-seven patients received neoadjuvant or conversion chemotherapy and 40 liver resections were carried out. Low ADC values (below median) of the representative liver metastases, at baseline and after systemic therapy, were associated with partial response according to RECIST criteria, but not with morphologic MRI changes or histology. Decreasing ADC values following systemic therapy were associated with improved OS compared to unchanged or increasing ADC, both in the liver resected subgroup (5-year OS rate 100% and 34%, respectively, P = 0.022) and systemic therapy subgroup (5-year OS rate 62% and 23%, P = 0.049). 1H-MRS revealed steatohepatosis induced by systemic therapy. CONCLUSIONS: Low ADC values at baseline or during systemic therapy were associated with treatment response by RECIST but not with histology, morphologic or detectable metabolic changes. A decreasing ADC during systemic therapy is associated with improved OS both in all patients receiving systemic therapy and in the resected subgroup.

Impact of deep learning-determined smoking status on mortality of cancer patients: never too late to quit.

BACKGROUND: Persistent smoking after cancer diagnosis is associated with increased overall mortality (OM) and cancer mortality (CM). According to the 2020 Surgeon General's report, smoking cessation may reduce CM but supporting evidence is not wide. Use of deep learning-based modeling that enables universal natural language processing of medical narratives to acquire population-based real-life smoking data may help overcome the challenge. We assessed the effect of smoking status and within-1-year smoking cessation on CM by an in-house adapted freely available language processing algorithm. MATERIALS AND METHODS: This cross-sectional real-world study included 29 823 patients diagnosed with cancer in 2009-2018 in Southwest Finland. The medical narrative, International Classification of Diseases-10th edition codes, histology, cancer treatment records, and death certificates were combined. Over 162 000 sentences describing tobacco smoking behavior were analyzed with ULMFiT and BERT algorithms. RESULTS: The language model classified the smoking status of 23 031 patients. Recent quitters had reduced CM [hazard ratio (HR) 0.80 (0.74-0.87)] and OM [HR 0.78 (0.72-0.84)] compared to persistent smokers. Compared to never smokers, persistent smokers had increased CM in head and neck, gastro-esophageal, pancreatic, lung, prostate, and breast cancer and Hodgkin's lymphoma, irrespective of age, comorbidities, performance status, or presence of metastatic disease. Increased CM was also observed in smokers with colorectal cancer, men with melanoma or bladder cancer, and lymphoid and myeloid leukemia, but no longer independently of the abovementioned covariates. Specificity and sensitivity were 96%/96%, 98%/68%, and 88%/99% for never, former, and current smokers, respectively, being essentially the same with both models. CONCLUSIONS: Deep learning can be used to classify large amounts of smoking data from the medical narrative with good accuracy. The results highlight the detrimental effects of persistent smoking in oncologic patients and emphasize that smoking cessation should always be an essential element of patient counseling.

Centralized repeated resectability assessment of patients with colorectal liver metastases during first-line treatment: prospective study.

BACKGROUND: Metastasectomy is probably underused in metastatic colorectal cancer. The aim of this study was to investigate the effect of centralized repeated assessment on resectability rate of liver metastases. METHODS: The prospective RAXO study was a nationwide study in Finland. Patients with treatable metastatic colorectal cancer at any site were eligible. This planned substudy included patients with baseline liver metastases between 2012 and 2018. Resectability was reassessed by the multidisciplinary team at Helsinki tertiary referral centre upfront and twice during first-line systemic therapy. Outcomes were resectability rates, management changes, and survival. RESULTS: Of 812 patients included, 301 (37.1 per cent) had liver-only metastases. Of these, tumours were categorized as upfront resectable in 161 (53.5 per cent), and became amenable to surgery during systemic treatment in 63 (20.9 per cent). Some 207 patients (68.7 per cent) eventually underwent liver resection or ablation. At baseline, a discrepancy in resectability between central and local judgement was noted for 102 patients (33.9 per cent). Median disease-free survival (DFS) after first resection was 20 months and overall survival (OS) 79 months. Median OS after diagnosis of metastatic colorectal cancer was 80, 32, and 21 months in R0-1 resection, R2/ablation, and non-resected groups, and 5-year OS rates were 68, 37, and 9 per cent, respectively. Liver and extrahepatic metastases were present in 511 patients. Of these, tumours in 72 patients (14.1 per cent) were categorized as upfront resectable, and 53 patients (10.4 per cent) became eligible for surgery. Eventually 110 patients (21.5 per cent) underwent liver resection or ablation. At baseline, a discrepancy between local and central resectability was noted for 116 patients (22.7 per cent). Median DFS from first resection was 7 months and median OS 55 months. Median OS after diagnosis of metastatic colorectal cancer was 79, 42, and 17 months in R0-1 resection, R2/ablation, and non-resected groups, with 5-year OS rates of 65, 39, and 2 per cent, respectively. CONCLUSION: Repeated centralized resectability assessment in patients with colorectal liver metastases improved resection and survival rates.

EGFR gene copy number assessment from areas with highest EGFR expression predicts response to anti-EGFR therapy in colorectal cancer.

BACKGROUND: Only 40-70% of metastatic colorectal cancers (mCRCs) with wild-type (WT) KRAS oncogene respond to anti-epidermal growth factor receptor (anti-EGFR) antibody treatment. EGFR amplification has been suggested as an additional marker to predict the response. However, improved methods for bringing the EGFR analysis into routine laboratory are needed. METHODS: The material consisted of 80 patients with mCRC, 54 of them receiving anti-EGFR therapy. EGFR gene copy number (GCN) was analysed by automated silver in situ hybridisation (SISH). Immunohistochemical EGFR protein analysis was used to guide SISH assessment. RESULTS: Clinical benefit was seen in 73% of high (≥ 4.0) EGFR GCN patients, in comparison with 59% of KRAS WT patients. Only 20% of low EGFR GCN patients responded to therapy. A high EGFR GCN number associated with longer progression-free survival (P<0.0001) and overall survival (P=0.004). Together with KRAS analysis, EGFR GCN identified the responsive patients to anti-EGFR therapy more accurately than either test alone. The clinical benefit rate of KRAS WT/high EGFR GCN tumours was 82%. CONCLUSION: Our results show that automated EGFR SISH, in combination with KRAS mutation analysis, can be a useful and easily applicable technique in routine diagnostic practise for selecting patients for anti-EGFR therapy.

Expression of the cell-cell adhesion molecule beta-catenin in colorectal carcinomas and their metastases.

To study the dynamic events leading to impaired cell-cell adhesion upon transition to the invasive phenotype of colorectal cancer (CRC), we examined three distinct beta-catenin expression patterns (membranous, cytoplasmic, and nuclear) in the paired samples of the primary tumours (P) and their metastatic lesions (M). beta-catenin expression was detected by immunohistochemistry (IHC) in 33 pairs of the primary CRC and their metastases. In a pair-wise (P-M) comparison, the membranous index (MI) was significantly different between P and M (p=0.036, Wilcoxon Signed-Ranks test), while cytoplasmic index (CI) and nuclear index (NI) values did not significantly deviate between P and M. MI in primary tumours was inversely related to the patient's age (p=0.04) and tumour grade (p=0.03), while patients with low MI in M had a high rate of metastasis at diagnosis (p=0.06). CI in P was lower in patients with LN involvement (p=0.02) and in advanced tumour stage (p=0.002). Tumours of the ascending colon had the highest CI in their M (p=0.04). Interestingly, high MI of the M lesions was a significant predictor of favourable overall survival (OS) in univariate (Kaplan-Meier) survival analysis (p=0.035). In conclusion, significant aberrations in beta-catenin expression probably take place in CRC cells during the development of metastatic phenotype, but a change from membrane expression to cytoplamic and/or nuclear expression is not a prerequisite for metastasis in all cases.

Comparison of CD44 expression in primary tumours and metastases of colorectal cancer.

A better understanding on the development of a metastatic phenotype in colorectal cancer (CRC) is essential to help identify patients at high risk for metastasis. Therefore, we have studied the role of the CD44 family of trans-membrane glycoprotein in the process of CRC metastasis, by examining the expression of CD44s and CD44v6 in primary tumours and their metastatic lesions in 46 patients using immunohistochemistry. The expression of both CD44s and CD44v6 was significantly higher (moderate/strong) in primary tumours as compared to their metastases (p=0.008, p=0.0001, respectively). CD44s expression in metastases increased with the degree of the histological grade (p=0.009) and invasiveness of the primary tumour (p=0.002). Disease-free survival (DFS) was shorter in patients who had metastases with a strong/moderate expression of CD44s as compared to those with negative/weak expression (8.3 months vs 16.9 months p=0.221, respectively). Our finding that CD44s expression in metastatic lesions may reflect the aggressiveness of the primary tumour from which it has originated implicates an important link between the two lesions. CD44 expression may also provide valuable biological information as suggested by the observation that up-regulated CD44s expression in metastases is associated with a shorter DFS.

DNA image cytometry is a useful adjunct tool in the prediction of disease outcome in patients with stage II and stage III colorectal cancer.

BACKGROUND: We assessed the prognostic value of the nuclear DNA content measured in the primary tumours of 123 patients with stage II or stage III colorectal cancer (CRC). METHODS: Isolated nuclei from paraffin sections were stained with the Feulgen reaction, and DNA was measured using a computer-assisted image analysis cytometry system. We applied 4 different approaches in the analysis of DNA histograms: the ABCDE approach, histogram range, peak evaluation and DNA cut-off values. RESULTS: Using the histogram range, the narrow range was rare (3.7%) in patients who died of disease (n = 28) as compared with 16.4% among those alive (n = 74; p = 0.017). Modal peak evaluation was a significant predictor of disease-free survival (DFS; Kaplan-Meier log-rank p = 0.0235). In the range evaluation, the 1st set (low-start gates) was a significant predictor of DFS (log-rank p = 0.0121), where disease recurrence was closely associated with the widest range (1.8->10c; c = haploid DNA content) gates. Recurrence-free survival was 3 times better in narrow-gate histograms than wide-range histograms (p < 0.03). The 1st set also proved to be a significant predictor of disease-specific survival (DSS; log-rank p = 0.0045), which was markedly better (77.8-90.0%) among the patients with the narrow-gate histograms. Grading of the histogram range into 2 categories (with 6.0c as cut-off), was a powerful predictor of both DSS (log-rank p = 0.0092) and 5-year DFS (p = 0.0106) in the whole series, and separately in stage III (but not stage II) disease, with p = 0.0131 and p = 0.0201, respectively. CONCLUSION: The DNA image cytometry with careful analysis of the histograms may provide valuable prognostic information in CRC, with potential clinical implications in patient management, particularly in predicting the patients at high risk for recurrence who should be considered as candidates for adjuvant therapy.

Response to chemotherapy (irinotecan plus 5-fluorouracil) in colorectal carcinoma can be predicted by tumour DNA content.

OBJECTIVE: The aim of this study was to identify markers that might predict response to chemotherapy. Postoperative chemotherapy improves the outcome in stage III colon cancer and is widely accepted as a standard therapy, but there are currently no reliable predictors to identify and select patients that will benefit. METHODS: Using DNA image cytometry, the DNA content was determined from the isolated nuclei of 56 primary colorectal carcinomas of patients who received chemotherapy (either irinotecan or irinotecan plus 5-fluorouracil and folinic acid) for advanced disease. Response to chemotherapy could be reliably evaluated in 53 patients. RESULTS: The modal DNA content (ploidy status) of the tumour correlated with the observed response to chemotherapy (p = 0.01). An objective response was observed in 56% of patients whose tumour histograms displayed tetraploid, peri-tetraploid or multiploid patterns of peaks, compared with 19% in patients with diploid, peri-diploid or aneuploid peaks. Notably, 86% (6/7) of patients whose tumours displayed a multiploid peak pattern showed an objective response and 1 patient had stable disease. CONCLUSIONS: This study suggests that modal DNA content can be used to predict a patient's response to chemotherapy in advanced colorectal carcinoma. This may help in identifying patients who will benefit most from therapy for advanced colorectal cancer.