RESUMEN
Multiple myeloma management requires a balance between maximizing survival, minimizing adverse events to therapy, and monitoring disease progression. While previous work has proposed data-driven models for individual tasks, these approaches fail to provide a holistic view of a patient's disease state, limiting their utility to assist physician decision-making. To address this limitation, we developed a transformer-based machine learning model that jointly (1) predicts progression-free survival (PFS), overall survival (OS), and adverse events (AE), (2) forecasts key disease biomarkers, and (3) assesses the effect of different treatment strategies, e.g., ixazomib, lenalidomide, dexamethasone (IRd) vs lenalidomide, dexamethasone (Rd). Using TOURMALINE trial data, we trained and internally validated our model on newly diagnosed myeloma patients (N = 703) and externally validated it on relapsed and refractory myeloma patients (N = 720). Our model achieved superior performance to a risk model based on the multiple myeloma international staging system (ISS) (p < 0.001, Bonferroni corrected) and comparable performance to survival models trained separately on each task, but unable to forecast biomarkers. Our approach outperformed state-of-the-art deep learning models, tailored towards forecasting, on predicting key disease biomarkers (p < 0.001, Bonferroni corrected). Finally, leveraging our model's capacity to estimate individual-level treatment effects, we found that patients with IgA kappa myeloma appear to benefit the most from IRd. Our study suggests that a holistic assessment of a patient's myeloma course is possible, potentially serving as the foundation for a personalized decision support system.
RESUMEN
OBJECTIVES: Evaluating potential relationships between progression-free survival (PFS) and tumor gene expression patterns and mutational status was an exploratory objective of the phase 3 TOURMALINE-MM1 study (NCT01564537) of ixazomib-lenalidomide-dexamethasone (IRd) vs placebo-Rd in 722 patients with relapsed/refractory multiple myeloma (MM). METHODS: We utilized gene expression and mutation data from screening bone marrow aspirates to identify tumors with non-canonical nuclear factor-κB (NF-κB) signaling pathway activation. RESULTS: DNA/RNA sequencing data were available for 339 (47.0%)/399 (55.2%) patients; 49/339 (14.5%) patients had non-canonical NF-κB pathway gene mutations (tumor-necrosis-factor receptor-associated factor 2, 3 [TRAF2, TRAF3], baculoviral-inhibitor-of-apoptosis repeat-containing 2/3 [BIRC2/3]), and PFS was significantly longer with IRd vs placebo-Rd in these patients (hazard ratio [HR] 0.23). In patients with lower TRAF3 expression (median not reached vs 11 months, HR 0.47) and higher NF-κB-inducing kinase (NIK) expression (median not reached vs 14 months, HR 0.45), both associated with non-canonical NF-κB pathway activation, PFS was significantly longer with IRd vs placebo-Rd. TRAF3 expression was decreased in patients harboring t(4;14) and 1q21 amplification, suggesting increased non-canonical NF-κB pathway activation. CONCLUSIONS: Adding ixazomib to Rd provides clinical benefit in MM tumors with increased non-canonical NF-κB pathway activity. This is a potential mechanism for activity in 1q21 amplified high-risk tumors.
Asunto(s)
Protocolos de Quimioterapia Combinada Antineoplásica/uso terapéutico , Mieloma Múltiple/tratamiento farmacológico , Mieloma Múltiple/metabolismo , FN-kappa B/metabolismo , Transducción de Señal/efectos de los fármacos , Protocolos de Quimioterapia Combinada Antineoplásica/efectos adversos , Biomarcadores de Tumor , Compuestos de Boro/administración & dosificación , Aberraciones Cromosómicas , Dexametasona/administración & dosificación , Regulación Neoplásica de la Expresión Génica , Glicina/administración & dosificación , Glicina/análogos & derivados , Humanos , Lenalidomida/administración & dosificación , Mieloma Múltiple/mortalidad , Mieloma Múltiple/patología , Mutación , Pronóstico , Supervivencia sin Progresión , Transcriptoma , Resultado del TratamientoRESUMEN
OBJECTIVES: In the TOURMALINE-MM1 phase 3 trial in relapsed/refractory multiple myeloma, ixazomib-lenalidomide-dexamethasone (IRd) showed different magnitudes of progression-free survival (PFS) benefit vs placebo-Rd according to number and type of prior therapies, with greater benefit seen in patients with >1 prior line of therapy or 1 prior line of therapy without stem cell transplantation (SCT). METHODS: RNA sequencing data were used to investigate the basis of these differences. RESULTS: The PFS benefit of IRd vs placebo-Rd was greater in patients with tumors expressing high c-MYC levels (median not reached vs 11.3 months; hazard ratio [HR] 0.42; 95% CI, 0.26, 0.66; P < .001) compared with in those expressing low c-MYC levels (median 20.6 vs 16.6 months; HR 0.75; 95% CI, 0.42, 1.2). Expression of c-MYC in tumors varied based on the number and type of prior therapy received, with the lowest levels observed in tumors of patients who had received 1 prior line of therapy including SCT. These tumors also had higher expression levels of CD19 and CD81. CONCLUSIONS: PFS analyses suggest that lenalidomide and ixazomib target tumors with different levels of c-MYC, CD19, and CD81 expression, thus providing a potential rationale for the differential benefits observed in the TOURMALINE-MM1 study. This trial was registered at www.clinicaltrials.gov as: NCT01564537.
Asunto(s)
Expresión Génica , Mieloma Múltiple/genética , Mieloma Múltiple/patología , Proteínas Proto-Oncogénicas c-myc/genética , Anciano , Anciano de 80 o más Años , Protocolos de Quimioterapia Combinada Antineoplásica/efectos adversos , Protocolos de Quimioterapia Combinada Antineoplásica/uso terapéutico , Biomarcadores de Tumor , Compuestos de Boro , Dexametasona , Femenino , Perfilación de la Expresión Génica , Glicina/análogos & derivados , Humanos , Lenalidomida , Masculino , Persona de Mediana Edad , Mieloma Múltiple/tratamiento farmacológico , Mieloma Múltiple/mortalidad , Clasificación del Tumor , Estadificación de Neoplasias , Pronóstico , Supervivencia sin Progresión , Resultado del TratamientoRESUMEN
BACKGROUND: Alisertib (MLN8237) is an investigational, oral, selective Aurora A kinase inhibitor. Aurora A contains two functional single nucleotide polymorphisms (SNPs; codon 31 [F/I] and codon 57 [V/I]) that lead to functional changes. This study investigated the prognostic and predictive significance of these SNPs. METHODS: This study evaluated associations between Aurora A SNPs and overall survival (OS) in The Cancer Genome Atlas (TCGA) database. The Aurora A SNPs were also evaluated as predictive biomarkers for clinical outcomes to alisertib in two phase 2 studies (NCT01045421 and NCT01091428). Aurora A SNP genotyping was obtained from 85 patients with advanced solid tumors receiving single-agent alisertib and 122 patients with advanced recurrent ovarian cancer treated with alisertib plus weekly paclitaxel (n=62) or paclitaxel alone (n=60). Whole blood was collected prior to treatment and genotypes were analyzed by PCR. FINDINGS: TCGA data suggested prognostic significance for codon 57 SNP; solid tumor patients with VV and VI alleles had significantly reduced OS versus those with II alleles (HR 1.9 [VI] and 1.8 [VV]; p<0.0001). In NCT01045421, patients carrying the VV alleles at codon 57 (n=53, 62%) had significantly longer progression-free survival (PFS) than patients carrying IV or II alleles (n=32, 38%; HR 0.5; p=0.0195). In NCT01091428, patients with the VV alleles at codon 57 who received alisertib plus paclitaxel (n=47, 39%) had a trend towards improved PFS (7.5months) vs paclitaxel alone (n=32, 26%; 3.8months; HR 0.618; p=0.0593). In the paclitaxel alone arm, patients with the VV alleles had reduced PFS vs modified intent-to-treat (mITT) patients (3.8 vs 5.1months), consistent with the TCGA study identifying the VV alleles as a poor prognostic biomarker. No significant associations were identified for codon 31 SNP from the same data set. INTERPRETATION: These findings suggest that Aurora A SNP at codon 57 may predict disease outcome and response to alisertib in patients with solid tumors. Further investigation is warranted.
Asunto(s)
Aurora Quinasa A/genética , Azepinas/administración & dosificación , Biomarcadores de Tumor/genética , Neoplasias/tratamiento farmacológico , Pirimidinas/administración & dosificación , Adulto , Anciano , Alelos , Azepinas/efectos adversos , Supervivencia sin Enfermedad , Femenino , Humanos , Masculino , Persona de Mediana Edad , Estadificación de Neoplasias , Neoplasias/genética , Neoplasias/patología , Paclitaxel/administración & dosificación , Paclitaxel/efectos adversos , Polimorfismo de Nucleótido Simple , Inhibidores de Proteínas Quinasas/administración & dosificación , Pirimidinas/efectos adversos , Resultado del TratamientoRESUMEN
BACKGROUND: Patient weights are frequently used in surgery for essential calculations including preoperative surgical planning, dosing of medications, and qualification for insurance reimbursement. Often, there is a disparity between patient-reported and actual measured weights. This study examines differences between self-reported and measured weights and implications in breast reduction surgery. METHODS: A review was performed of 197 consecutive women who presented for breast surgery at an academic institution. Self-reported weights were recorded during the initial encounter, and the patients were subsequently weighed on the same day. A paired t-test was used to compare the self-reported and actual measured weights and stratified analysis performed based on body mass index (BMI). The Schnur sliding scale was used to estimate resection weights for reduction mammaplasty. RESULTS: The overall mean difference in reported and measured weight was 3.0 lbs standard deviation (SD) 8.9 (P < 0.0001) with a maximum overestimation of 25 lb and underestimation of 80 lb. Statistically significant differences were found when stratified analysis was performed based on BMI as mean differences in the overweight (BMI 25-30) and obese (BMI > 30) groups were 1.7 lb SD 5.5 (P < 0.026) and 4.9 lbs SD 11.8 (P < 0.0002), respectively. Significant differences in calculated breast reduction resection weights, based on the Schnur sliding scale, were also found when comparing self-reported and measured weights in all groups. CONCLUSIONS: Significant disparities between self-reported and measured weights were identified in patients presenting for breast surgery. These differences can influence important calculations of resection weights for breast reduction surgery. These differences may also affect dosing of medications and preoperative planning.
Asunto(s)
Peso Corporal , Mamoplastia , Autoinforme , Femenino , Humanos , Cuidados Preoperatorios , Estudios RetrospectivosRESUMEN
PURPOSE: Global gene expression analysis was performed on pre-treatment biopsies from patients with squamous cell carcinoma of the head and neck (SCCHN) to discover biomarkers that can predict outcome of radiation based therapy. METHODS: We initially evaluated RNA expression using cDNA microarray analysis of 38 patients that received radiotherapy (RT). The five strongest candidates (VEGF, BCL-2, CLAUDIN-4, YAP-1 and c-MET) were then analysed in pre-treatment biopsies in a second group of 86 patients who received radiation based treatment using immunohistochemical staining (IHC), prepared by tissue microarray. RESULTS: In the first population, 13 of 38 (34%) had no (NR) or partial response (PR) to RT. cDNA microarrays revealed 60 genes that were linked to response to therapy. In the second series, 12 of 86 patients (14%) experienced NR or PR to CRT. Cause specific survival (CSS) and recurrence free survival (RFS) at 2 years was 85% and 90% and at 3 years 81% and 84%, respectively. Biomarkers predictive for NR/PR were increased expression of vascular endothelial growth factor (VEGF) (p=0.02), Yes-associated protein (YAP-1) (p<0.01), CLAUDIN-4 (p<0.01), c-MET (p<0.01) and BCL-2 (p=0.02). Biomarkers predictive of poor RFS were YAP-1 (p=0.01) and BCL-2 (p<0.01). Biomarkers predictive of poor CSS were YAP-1 (p=0.04), VEGF (p=0.03) and CLAUDIN-4 (p=0.03). Furthermore, when YAP-1 and c-MET expression levels were combined the prediction of radio-resistance was increased. CONCLUSION: All five biomarkers were predictive of poor response to radiation based therapy. In particular, YAP-1 and c-MET have synergistic power and could be used to make treatment decisions.