Risk Of Early Metastases Research Articles

Long-term survival of uveal melanoma patients with respect to clinical, hemodynamic and morphological characteristics of the tumor

Purpose. Evaluation of long-term results of uveal melanoma (UM) treatment in patients depending on clinical, hemodynamic and morphological parameters of the tumor. Material and methods. Long-term results of treatment were analyzed in 394 UM patients aged 17 to 82 (mean age 53.1 ± 13.2). The median follow-up was 56 months. Group A (n = 242, organ-preserving treatment) consisted of patients with a tumor height of 5.0 ± 1.8 mm and base diameter of 12.7 ± 2.9 mm. Group B (n = 152, enucleation) included patients with 8.3 ± 2.8 mm tumor height and 15.2 ± 4.1 mm base diameter. Overall survival and disease-dependent survival were determined. Results. The cumulative 3-, 5-, and 7-year survival rates of patients with choroidal melanoma were 91.6 ± 1.7, 84.0 ± 2.5, 74.0 ± 3.8 % , and those with ciliochoroidal melanoma were 84.2 ± 5.6, 79.9 ± 5.8, and 53.0 ± 14.7 %, respectively (p = 0.048). In group A, the 5-year survival rates of patients with mushroom-shaped and irregular forms were 80.6 ± 5.9 and 74.6 ± 10.9 %, respectively, and 96.0 ± 2.0 % for dome-shaped UM. The worst 5-year survival rates were shown by patients with tumor stabilization, and the best ones, by patients with complete tumor resorption (x2 = 6.880, p = 0.008). Patients with post-radiation optical neuropathy (n = 193) had better 5-year survival rates (90.3 ± 2.9 %) than patients (n = 49) without neuropathy (77.5 ± 8.0 %). In group B, cumulative 1-year survival was 95.2 ± 1.9 %, 2-year survival, 89.2 ± 2.8 % , 3-year survival, 84.9 ± 2.9 % , 4-year survival, 81.8 ± 3.6 %, 5-year survival, 77.5 ± 4.0 % , 6-year survival, 67.9 ± 4.8 % , and 7-year survival was 64.7 ± 5.2 %. In epithelioid-cell, mixed-cell, and spindle-cell types of UM, the 5-year survival rates were 46.3 ± 12.0, 70.7 ± 8.0 %, and 84.8 ± 5.4 %, respectively (x2 = 11.571, p = 0.004). The 5-year survival rate of patients with extrascleral growth was 56.0 ± 10.4 % and without extrascleral growth it was 81.7 ± 4.0 %. Conclusion. In organ-preserving treatment, the combination of prognostic factors allows us to stratify the patients based on the risk of early metastasis, which allows for personalized dynamic monitoring, taking into account the vital prognosis of a particular patient.

Abstract LB532: Bifidobacterium longum RAPO alleviates the risk of immune-related adverse events of anti-PD-1 immunotherapy in a mouse model of triple-negative breast cancer

Abstract Background: Triple negative breast cancer (TNBC) has higher risk of early metastasis and recurrence compared with other types of breast cancer. Although immune checkpoint inhibitors (ICIs) have emerged as a feasible strategy for the treatment of TNBC, its use causes various side effects known as immune-related adverse events (irAEs) including pneumonitis, myocarditis, colitis, hepatitis, and nephritis. Considering severe irAEs could lead to ICIs termination and fatal outcomes, studies into immunotherapy that increases anti-tumor efficacy while reducing side effects are needed. We have previously reported that combination therapy of Bifidobacterium longum RAPO and anti-PD-1 shows a better efficacy compared with monotherapy in mice with TNBC. In the present study, we further investigated whether this combination therapy with probiotics would affect the risk of irAEs induced by anti-PD-1 therapy against TNBC. Methods: BALB/c mice bearing the 4T1 murine metastatic breast cancer cells were randomly assigned to tumor control, Bifidobacterium longum RAPO (RAPO), Anti-PD-1, or Anti-PD-1+RAPO group. Tumor tissues were analyzed by Affymetrix Mouse Gene 2.0 ST Array, and functional analysis of microarray data was performed using Ingenuity Pathway Analysis (IPA®, QIAGEN). Lung, heart, colon, liver, and kidney tissues were analyzed by qRT-PCR to determine expression of pro- and anti-inflammatory cytokines. Colon and liver tissues were also analyzed by a Western blot and immunohistochemistry (IHC), respectively. Results: Functional analysis revealed that inflammation in lung and cardiac lesion was predicted to be reduced in Anti-PD-1+RAPO group compared with Anti-PD-1 group. Indeed, expression of pro-inflammatory cytokines including IL1β, IL6, and TNFα was significantly decreased, while anti-inflammatory cytokine IL10 expression was increased in the lung, liver, and kidney tissues of Anti-PD-1+RAPO group compared with Anti-PD-1 group. Similarly, in the heart tissue, TNFα expression was decreased, whereas IL10 was significantly increased in Anti-PD-1+RAPO group than monotherapy group. In the colon tissue, combination therapy of Anti-PD-1 with RAPO significantly enhanced expression of IL10 and tight junction proteins such as occludin and ZO-1 compared with Anti-PD-1 treatment alone. Colonic expression of pro-inflammatory cytokines tended to be reduced in combination therapy group than Anti-PD-1 group. Furthermore, Anti-PD-1+RAPO group was associated with lower level of myeloperoxidase, a pro-inflammatory enzyme released by activated neutrophils, in the liver than Anti-PD-1 group. Conclusions: Collectively, in addition to the effects on anti-tumor efficacy, combination therapy of anti-PD-1 with B. longum RAPO alleviates the risk of irAEs in mice with TNBC. This study provides evidence that B. longum RAPO supplementation might be a novel therapeutic approach in cancer immunotherapy. [This work was supported by BIFIDO Co. and NRF (2020R1C1C1007553 to S-EK).] Citation Format: Hyeyoon Kim, Rira Oh, Ji-Won Heo, Geun Eog Ji, Myeong Soo Park, Sung-Eun Kim. Bifidobacterium longum RAPO alleviates the risk of immune-related adverse events of anti-PD-1 immunotherapy in a mouse model of triple-negative breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr LB532.

Electrochemical neuron-specific enolase (NSE) immunosensor based on CoFe2O4@Ag nanocomposite and AuNPs@MoS2/rGO

Small cell lung cancer (SCLC) is highly associated with the risk of early metastasis. Neuron-specific enolase (NSE), a biomarker of SCLC, is directly related to tumor burden and early diagnosis. This biomarker exists in nerve tissue and neuroendocrine tissue. In this study, an electrochemical NSE immunosensor based on gold nanoparticles modified molybdenum disulfide and reduced graphene oxide (AuNPs@MoS2/rGO) as the electrode platform and CoFe2O4@Ag nanocomposite as the signal amplification was developed. The immobilization of anti-NSE capture antibody was successfully performed on AuNPs@MoS2/rGO modified electrode surface by amino-gold affinity and the conjugation of anti-NSE secondary antibody on CoFe2O4@Ag nanocomposite was successfully completed by the strong esterification reaction. The final immunosensor was designed by the specific interactions of electrode platform and signal amplification. The fabricated nanocomposites and electrochemical immunosensor were characterized by both physicochemical characterization techniques including transmission electron microscopy (TEM), scanning electron microscopy (SEM), x-ray diffraction (XRD), x-ray photoelectron spectroscopy (XPS), fourier transform infrared spectroscopy (FTIR), and electrochemical methods such as cyclic voltammetry (CV), square wave voltammetry (SWV), and electrochemical impedance spectroscopy (EIS). The quantification limit (LOQ) and the determination limit (LOD) were computed to be 0.01 pg mL−1 and 3.00 fg mL−1, respectively. In brief, it can be speculated that the constructed electrochemical NSE immunosensor can be successfully utilized in the early diagnosis for lung cancer.

Abstract PS6-15: Mammaprint risk stratification associated with presence of circulating tumor cells in breast cancer

Abstract Background: We previously reported the prognostic and predictive value of circulating tumor cells (CTCs) in early and late-stage breast cancer (BC). While the evaluation of CTCs via liquid biopsy in oncology practice guidelines is still evolving, other biomarkers are more commonly used, including multigene sequencing. The 21-gene Oncotype DX (ODX) assay predicts a risk of disease recurrence in patients with estrogen receptor (ER)-positive disease. The 70-gene MammaPrint (MP) panel, on the other hand, predicts a risk of early metastasis, and was developed using ER- and ER+ samples. We sought to understand whether these genomic assays predicted the presence of CTCs. Methods: A cohort of patients with BC and CTC data (i.e. positive or negative) was identified in this retrospective study using the National Cancer Database (NCDB) 2004-2016 registry. Clinicopathologic characteristics, including ODX and MP data, were described using frequencies/proportions. At the univariate level, chi-squared tests were used to evaluate for an association between the results of ODX (low-risk as recurrence score of <11, intermediate-risk as 11-25, and high-risk as >25), or MP (low or high risk), and the presence of CTCs. These variables were then assessed at the multivariate level using multiple logistic regression, controlling for other variables we previously showed predicted CTCs, including: race, receptor status, histology and AJCC clinical disease stage. Analyses were performed using SPSS version 26.0. Results: A total of n=4577 patients were evaluated for CTCs: n=940 (20.5%) of which underwent ODX testing, while n=282 (6.2%) underwent MP. Chi-squared analyses showed no association between ODX risk categories and the presence of CTCs (p=0.204), but a significant association between MP and CTCs (p=0.005). Through multivariate analysis, we found the association between a high-risk designation as per MP and the presence of CTCs remained significant (OR 2.65, 95% CI 1.23-5.70, p=0.013), even after controlling for race, receptor status, histology, and AJCC clinical disease stage. A similar multivariate model including ODX as a predictor of CTCs confirmed non-significance (p=0.664). Of patients with MP and CTC data, the majority were 50-70 years of age (n=164, 58.2%), White (n=244, 88.1%), with ductal carcinoma (n=211, 74.8%) on histological evaluation (Table 1). N=12 (4.6%) were diagnosed at stage 0, n=173 (66.0%) at stage I, n=63 (24.0%) at stage II, n=7 (2.7%) at stage III, and n=7 (2.7%) at stage IV. Most were hormone receptor-positive (HR+) and HER2-negative (HER2-) (n=234, 86.3%); n=11 (4.1%) were HR+/HER2+, n=5 (1.8%) were HR-/HER2, while n=21 (7.7%) were triple-negative breast cancers (TNBC). N=62 (22.0%) were assigned high-risk as per MP, while n=220 (78.0%) were assigned low-risk. Conclusion: High-risk of early metastasis, per MammaPrint, was significantly associated with the presence of CTCs. However, a significant association was not observed between Oncotype DX risk stratification and CTCs. This may suggest clinical utility in combining ODX metrics with liquid biopsy in order to maximize survival prognostication, when ODX is utilized as the biomarker. Further evaluation will be required to understand if the MP assay’s clinical value contribution is independent of CTC status. Table 1: Multiple logistic regression model predicting presence of CTCs in patients with breast cancer.VariableCirculating tumor cellsNegative (n=211)Positive (n=71)OR*95% CIp-valueMammaPrintLow-risk (ref)173 (61.3%)47 (16.7%)1--High-risk38 (13.5%)24 (8.5%)2.651.23-5.700.013**Age<5045 (16.0%)23 (8.2%)---50-70126 (44.7%)38 (13.5%)>7040 (14.2%)10 (3.5%)Race0.993White (ref)182 (65.7%)62 (22.4%)1--Black18 (6.5%)7 (2.5%)0.930.29-2.990.903Asian7 (2.5%)1 (0.4%)0.000.00-0.000.999AJCC clinical staging0.0910 (ref)10 (3.8%)2 (0.8%)1--I136 (51.9%)37 (14.1%)0.600.10-3.560.572II49 (18.7%)14 (5.3%)0.550.09-3.470.526III1 (0.4%)6 (2.3%)10.910.61-194.790.104IV3 (1.1%)4 (1.5%)3.670.273-49.520.327Estrogen receptor statusNegative (ref)21 (7.5%)6 (2.1%)1--Positive190 (67.6%)64 (22.8%)2.190.42-11.270.350Progesterone receptor statusNegative (ref)39 (13.9%)11 (3.9%)1--Positive172 (61.2%)59 (21.0%)1.550.49-4.890.457HER2 receptor statusNegative (ref)192 (69.6%)61 (22.1%)1--Positive16 (5.8%)7 (2.5%)1.580.53-4.700.412Histology0.503Ductal (ref)162 (57.4%)49 (17.4%)1--Lobular32 (11.3%)21 (7.4%)1.600.73-3.530.241Other17 (6.0%)1 (0.4%)0.00NE-NE0.998*OR predicting presence of CTCs. **Age not included in the multivariate model. Citation Format: Nadeem Bilani, Leah Elson, Hong Liang, Elizabeth Elimimian, Zeina Nahleh. Mammaprint risk stratification associated with presence of circulating tumor cells in breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS6-15.

Diffusion kurtosis imaging in patients with locally advanced rectal cancer: current status and future perspectives.

Morphological magnetic resonance imaging is currently the best imaging technique for local staging in patients with rectal cancer. However, morphological sequences have some limitations, especially after preoperative chemoradiotherapy (pCRT). Diffusion-weighted imaging has been applied to rectal cancer for detection of lesions, characterization of tissue, and evaluation of the response to therapy. In 2005, a non-Gaussian diffusion model called diffusion kurtosis imaging (DKI) was suggested. Several electronic databases were evaluated in the present review. The search included articles published from January 2000 to May 2018. The references of all articles were also evaluated. All titles and abstracts were assessed, and only the studies of DKI in patients with rectal cancer were retained. We identified 35 potentially relevant references through the electronic search. According to the inclusion and exclusion criteria, we retained five clinical studies that met the inclusion criteria. DKI is a useful tool for assessment of tumor aggressiveness, the nodal status, and the risk of early metastases as well as prediction of the response to pCRT. The results of DKI should be considered in treatment decision-making during the work-up of patients with rectal cancer.

Progress in clinical research of pancreatic cancer: from "resection" to "cure"

Because of the high malignancy of pancreatic ductal adenocarcinoma, the cancer-related mortality of pancreatic ductal adenocarcinoma is increasing year by year. Despite advance in surgical techniques, the 5-year survival rate of patients after resection is still less than 30%. Recent studies have found that pancreatic ductal adenocarcinoma is a systemic disease, which may not be cured completely by up-front resection, but requires perioperative multidisciplinary therapy. With the concept of "potentially curable pancreatic cancer" , clinicians need to evaluate the resectability of pancreatic ductal adenocarcinoma accurately before operation, and use the innovative multidisciplinary therapy including neoadjuvant chemoradiotherapy,surgery and adjuvant chemoradiotherapy to improve the R0 resection rate and reduce the risk of early metastasis. Therefore, the therapeutic goal of pancreatic ductal adenocarcinoma is no longer "simple resection" , but long survival through perioperative multidisciplinary treatment. In this article, we briefly introduce the progress of resectability assessment, surgical techniques and perioperative adjuvant therapy of "potentially curable pancreatic cancer" .

Association of Circulating Tumor Cells (CTCs) and Genomic Signatures in Prostate Cancer Patients

Intermediate and high-risk prostate cancer patients who are candidates for radiotherapy (RT) are very heterogeneous. Genomic classifiers add to clinical-pathologic risk factors and have the potential to become integral to risk stratification. Since circulating tumor cells (CTCs) are determinate of metastasis and genomic predictors of CTCs are not yet understood in this population, we investigated the associations of CTC counts with several genomic classifiers in patients enrolled in a randomized Phase II radiotherapy clinical trial (BLaStM, NCT02307058). CTCs were enumerated in blood samples from 31 patients using a microfilter system. The number of single CTCs, CTC clusters and their sum were recorded. Several gene signatures were investigated. Decipher® (metastatic risk), PORTOS (predictor of RT response) and PAM50 (predictor hormone response, which includes gene expression-based "intrinsic" subtypes: luminal A, luminal B, HER2-enriched, and basal-like), were quantified in tissue obtained by MRI-Ultrasound fused biopsies. A high-density gene expression microarray processed at a biosciences laboratory was used. A Poisson regression model with demographic and disease characteristics variables (PSA, Gleason Score (GS), T stage, %tumor in biopsies, age, race/ethnicity) were used in the univariable analysis (UVA). Multivariable analysis (MVA) included the statistically significant variables, where each biomarker was examined, along with chosen covariates for association with CTC counts in a one-at-a-time manner. CTCs were evident in over 70% of patients tested. The single cell and cluster median values in 8 mL of blood were 33 (range 0-343) and 4 (range 0-26). In MVA, several of the signatures were associated with total CTC number (single cells and clusters). When the biopsy core containing the highest Gleason score (GS) and %tumor were analyzed, the significant relationships included PORTOS (RR 6.00; p<0.001), PAM50_basal (RR 2.21; p<0.001), and Decipher (RR 1.16; p=0.048). Since multiple biopsies were analyzed for gene expression, when the cores with the highest signature value were selected, PORTOS surged (RR=47.10; p<0.0001) in the MVA. Despite the relatively small sample size, the results indicate a strong signal related to radiation sensitivity. NCCN risk categories were not significantly related to CTC counts or gene signatures. Determinants of prostate cancer lethality in intermediate to high-risk prostate cancer are poorly understood. Our results show an association between tumor gene expression alterations and CTC number, indicating that genes linked to radiation sensitivity have a role in early metastatic risk.

Uveal Melanomas with SF3B1 Mutations: A Distinct Subclass Associated with Late-Onset Metastases

To investigate the prevalence and prognostic value of SF3B1 and EIF1AX mutations in uveal melanoma (UM) patients. Case series. Cohort of 151 patients diagnosed with and treated for UM. SF3B1 and EIF1AX mutations in primary tumors were investigated using whole-exome sequencing (n= 25) and Sanger sequencing (n= 151). For the detection of BAP1 mutations, a previously reported cohort of 90 patients was extended using BAP1 sequencing or immunohistochemistry. The status of SF3B1, EIF1AX, and BAP1 in tumors of patients were correlated to clinical, histopathologic, and genetic parameters. Survival analyses were performed for patients whose tumors had SF3B1, EIF1AX, and BAP1 mutations. Patients with tumors harboring EIF1AX mutations rarely demonstrated metastases (2 of 28 patients) and overall had a longer disease-free survival (DFS; 190.1 vs. 100.2 months; P < 0.001). Within the patient group with disomy 3, UM patients with an SF3B1 mutation had an increased metastatic risk compared with those without an SF3B1 mutation (DFS, 132.8 vs. 174.4 months; P= 0.008). Patients with such a mutation were more prone to demonstrate late metastases (median, 8.2 years; range, 23-145 months). Patients with UM and loss of BAP1 expression had a significantly decreased survival (DFS, 69.0 vs. 147.9 months; P < 0.001). According to our data, patients with UM can be classified into 3 groups, of which EIF1AX-mutated tumors and tumors without BAP1, SF3B1, or EIF1AX mutations are associated with prolonged survival and low metastatic risk, SF3B1-mutated tumors are associated with late metastasis, and tumors with an aberrant BAP1 are associated with an early metastatic risk and rapid decline in patient DFS.

Biologic markers determine both the risk and the timing of recurrence in breast cancer

Breast cancer has a long natural history. Established and emerging biologic markers address overall risk but not necessarily timing of recurrence. 346 adjuvant naïve breast cancer cases from Guy's Hospital with 23years minimum follow-up and archival blocks were recut and reassessed for hormone-receptors (HR), HER2-receptor and grade. Disease-specific survival (DSS) was analyzed by recursive partitioning. To validate insights from this analysis, gene-signatures (proliferative and HR-negative) were evaluated for their ability to predict early versus late metastatic risk in 683 node-negative, adjuvant naïve breast cancers annotated with expression microarray data. Risk partitioning showed that adjuvant naïve node-negative outcome risk was primarily partitioned by tumor receptor status and grade but not tumor size. HR-positive and HER2-negative (HRpos) risk was partitioned by tumor grade; low grade cases have very low early risk but a 20% fall-off in DSS 10 or more years after diagnosis. Higher grade HRpos cases have risk over >20years. Triple-negative (Tneg) and HER2-positive (HER2pos) cases DSS events occurred primarily within the first 5years. Among node-positive cases, only low grade conferred late risk, suggesting that proliferative gene signatures that identify proliferation would be important for predicting early but not late recurrence. Using pooled data from four publicly available data sets for node-negative tumors annotated with gene expression and outcome data, we evaluated four prognostic gene signatures: two proliferation-based and two immune function-based. Tumor proliferative capacity predicted early but not late metastatic risk for HRpos cases. The immune function or HRneg specific signatures predicted only early metastatic risk in Tneg and HER2pos cases. Breast cancer prognostic signatures need to inform both risk and timing of metastatic events and may best be applied within subsets. Current signatures predict for outcome risk within 5years of diagnosis. Predictors of late risk for HR positive disease are needed.

The 70-gene prognosis profile predicts early metastases in postmenopausal breast cancer patients.

Abstract Abstract #1063 Introduction&amp;#x2028; By using gene expression profiling, Van 't Veer and colleagues developed a 70-gene prognosis profile to identify breast cancer patients who are at low risk of developing distant metastases and therefore might be safely spared chemotherapy. This 70-gene profile has been developed using tumors from a selected group of patients, i.e. women under the age of 55 years at diagnosis with stage I or II tumors. However, the majority of breast cancer patients are older, postmenopausal women who are increasingly being offered adjuvant chemotherapy despite the more favorable biological characteristics of their tumors, such as a lower proliferation rate and a high endocrine sensitivity. The aim of this study was to evaluate the prognostic value of the 70-gene prognosis profile in postmenopausal women with node-negative breast cancer.&amp;#x2028; Methods&amp;#x2028; Frozen tumor samples from 148 node-negative, breast cancer patients aged between 55 and 70 years at diagnosis were selected. Patients were treated with breast conserving therapy or mastectomy with axillary lymph node dissection between 1984 and 1996 at the Netherlands Cancer Institute. Eighteen percent received adjuvant tamoxifen, none of the patients received adjuvant chemotherapy. Samples were evaluated by gene expression profiling for the 70-gene profile and were classified as genomic high risk or genomic low risk. Clinical risk was assessed using Adjuvant! Online. The median follow-up was 11.6 years.&amp;#x2028; Results&amp;#x2028; Among the 148 patients, 91 (61%) were classified as genomic low risk, whereas 57 (39%) were classified as genomic high risk. The breast cancer specific survival (BCSS) at 5 and 10 years were 99% (SE 0.01) and 90% (SE 0.04), respectively, for the genomic low risk group versus 80% (SE 0.05) and 69% (SE 0.06), respectively, for the genomic high risk group (log rank p=0.036). The 70-gene prognosis profile was a predictor for BCSS at 5 and 10 years with hazard ratios (HR) of 19.1 (95% CI 2.5-148; p=0.005) and 3.9 (95% CI 1.7-9.0; p=0.002), respectively. In a multivariate model for BCSS including the genomic risk classification by the 70-gene prognosis profile and the clinical risk assessment by Adjuvant! online the profile outperformed the clinical risk assessment with a HR of 14.4 (95% CI 1.7-122.2; p=0.01) for BCSS at 5 years. The adjusted HR for the 70-gene profile for BCSS at 10 years was 2.2 (0.9-5.5; p=0.08).&amp;#x2028; Conclusion &amp;#x2028; Our results show that the 70-gene prognosis profile is also a strong predictor of BCSS in postmenopausal women. In addition, the profile can accurately select patients at high risk of early metastases within 5 years of diagnosis. Since the beneficial effect of chemotherapy in older women predominantly occurs in the first 5 years, the profile can be of clinical use in selecting postmenopausal women for adjuvant chemotherapy. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 1063.

Smoking and the Risk of Early Metastases from Uveal Melanoma

Smoking is suspected of altering host immunity and may therefore hasten the development of metastases among cancer patients. The authors followed 946 patients with melanomas of the choroid and/or ciliary body who had been treated with proton beam irradiation and who had provided a smoking history during their evaluation before treatment. After a mean follow-up of 33 months, 98 patients were diagnosed with metastatic disease. Metastasis-free survival rates 3 years after irradiation were similar in association with never, past, and current cigarette smoking (86%, 89%, and 90%, respectively; P = 0.90). Rates also were similar with increasing pack-years of exposure and when the product smoked was cigarettes versus pipes or cigars only. Estimates for smoking effects were unchanged after adjusting for established prognostic factors for metastases. Results suggest that smoking does not alter the risk of metastases during the first few years after irradiation of choroidal melanomas.