Carrying out streamlined routine data analyses with reports for observational studies: introduction to a series of generic SAS ® macros.

For a typical medical research project based on observational data, sequential routine analyses are often essential to comprehend the data on hand and to draw valid conclusions.  However, generating reports in SAS ® for routine analyses can be a time-consuming and tedious process, especially when dealing with large databases with a massive number of variables in an iterative and collaborative research environment. In this work, we present a general workflow of research based on an observational database and a series of SAS ® macros that fits this framework, which covers a streamlined data analyses and produces journal-quality summary tables. The system is generic enough to fit a variety of research projects and enables researchers to build a highly organized and concise coding for quick updates as research evolves. The result reports promote communication in collaborations and will escort the research with ease and efficiency.

Relative validation of a short questionnaire to assess the dietary habits of pregnant American Indian women.

The objective of this study was to compare a short dietary screener developed to assess diet quality with interviewer-administered telephone 24-hour dietary recalls in a population of pregnant Northern Plains (NP) American Indian women. Participants were recruited from NP clinical sites of the Prenatal Alcohol and SIDS and Stillbirth (PASS) Network, as part of a large, prospective, multidisciplinary study. Prenatal PASS participants who enrolled prior to 24 weeks gestation were eligible to participate. Repeated 24-hour dietary recalls were collected using the Nutrition Data System for Research (NDSR) software and a short dietary screener was administered intended to capture usual dietary intake during pregnancy. The available recalls were averaged across days for analysis. Items were grouped from the recalls to match the food group data estimates for the screener (e.g., total vegetables, total fruit, total dairy, total and whole grains). Deattenuated Pearson correlation coefficients were calculated between the two data sources after correcting for the within-person variation in the 24-hour recall data. A total of 164 eligible women completed the screener and at least two 24-hour dietary recalls and were included in the analyses. Pearson deattenuated correlation coefficients between the diet screener and the dietary recalls for the majority of food groups were 0.40 or higher. This short diet screener to assess usual diet appears to be a valid instrument for use in evaluating diet quality among pregnant American Indian women.

Socioeconomic Factors Are Associated With Readmission After Lobectomy for Early Stage Lung Cancer.

BACKGROUND: Data regarding risk factors for readmissions after surgical resection for lung cancer are limited and largely focus on postoperative outcomes, including complications and hospital length of stay. The current study aims to identify preoperative risk factors for postoperative readmission in early stage lung cancer patients. METHODS: The National Cancer Data Base was queried for all early stage lung cancer patients with clinical stage T2N0M0 or less who underwent lobectomy in 2010 and 2011. Patients with unplanned readmission within 30 days of hospital discharge were identified. Univariate analysis was utilized to identify preoperative differences between readmitted and not readmitted cohorts; multivariable logistic regression was used to identify risk factors resulting in readmission. RESULTS: In all, 840 of 19,711 patients (4.3%) were readmitted postoperatively. Male patients were more likely to be readmitted than female patients (4.9% versus 3.8%, p < 0.001), as were patients who received surgery at a nonacademic rather than an academic facility (4.6% versus 3.6%; p = 0.001) and had underlying medical comorbidities (Charlson/Deyo score 1+ versus 0; 4.8% versus 3.7%; p < 0.001). Readmitted patients had a longer median hospital length of stay (6 days versus 5; p < 0.001) and were more likely to have undergone a minimally invasive approach (5.1% video-assisted thoracic surgery versus 3.9% open; p < 0.001). In addition to those variables, multivariable logistic regression analysis identified that median household income level, insurance status (government versus private), and geographic residence (metropolitan versus urban versus rural) had significant influence on readmission. CONCLUSIONS: The socioeconomic factors identified significantly influence hospital readmission and should be considered during preoperative and postoperative discharge planning for patients with early stage lung cancer.

Cholangiocarcinoma size on magnetic resonance imaging versus pathologic specimen: Implications for radiation treatment planning.

PURPOSE: The accuracy of abdominal magnetic resonance imaging (MRI) in measuring gross tumor volume in patients with resectable cholangiocarcinoma (CC) is unknown. CC is a highly difficult tumor to visualize and treatment with dose-escalated radiation therapy requires clear tumor delineation. We aim to investigate the concordance between imaging and pathologic size in patients with resected CC to determine the usefulness of MRI for image guided treatment modalities. METHODS AND MATERIALS: The records of 51 patients with resected CC who underwent preoperative MRI were evaluated. Each preoperative MRI was individually reviewed by a diagnostic radiologist (P.M.), who was blinded to pathologic measurements. A combination of dynamic multiphase contrast-enhanced T1- and T2-weighted images, original imaging reports, and pathologic reports were reviewed for greatest tumor dimensions. A general linear regression model was used to examine the outcome MRI minus pathology using MRI report, T1-weighted measurement, or T2-weighted measurement. A multivariable regression model was fit to assess the association of other factors with pathologic underestimation. RESULTS: The median age was 69 years. Eleven tumors were categorized distal/extrahepatic, 17 hilar, and 23 intrahepatic CC. The median tumor size on pathology report was 3.00 cm (range, 0.3-19). The median tumor size from the MRI report was 3 cm (range, 0.80-16.20) and median tumor size on independent radiological review was 3 cm (range, 0.90-17) on the T1-weighted and 3 cm (range, 0.90-17) on the T2-weighted MRI sequences. When compared with pathologic tumor size, the MRI report dimension was found to underestimate tumor size by 4.1 mm (P = .04). On multivariable analysis, pathologic size underestimation was influenced by increasing tumor size (slope, -0.20; P < .001); however, underestimation was not affected by tumor location or MRI sequence. CONCLUSIONS: MRI underestimates tumor size, which was more pronounced with larger tumors, but not influenced by tumor location. The potential for gross tumor volume underestimation should be considered when planning highly conformal radiation therapy treatment of CC.

Nodal Upstaging Is More Common with Thoracotomy than with VATS During Lobectomy for Early-Stage Lung Cancer: An Analysis from the National Cancer Data Base.

INTRODUCTION: Questions remain regarding differences in nodal evaluation and upstaging between thoracotomy (open) and video-assisted thoracic surgery (VATS) approaches to lobectomy for early-stage lung cancer. Potential differences in nodal staging based on operative approach remain the final significant barrier to widespread adoption of VATS lobectomy. The current study examines differences in nodal staging between open and VATS lobectomy. METHODS: The National Cancer Data Base was queried for patients with clinical stage T2N0M0 or lower lung cancer who underwent lobectomy in 2010-2011. Propensity score matching was performed to compare the rate of nodal upstaging in VATS with that in open approaches. Additional subgroup analysis was performed to assess whether rates of upstaging differed by specific clinical setting. RESULTS: A total of 16,983 lobectomies were analyzed; 4935 (29.1%) were performed using VATS. Nodal upstaging was more frequent in the open group (12.8% versus 10.3%; p < 0.001). In 4437 matched pairs, nodal upstaging remained more common for open approaches. For a subgroup of patients who had seven lymph or more nodes examined, propensity matching revealed that nodal upstaging remained more common after an open approach than after VATS (14.0% versus 12.1%; p = 0.03). For patients who were treated in an academic/research facility, however, the difference in nodal upstaging between an open and VATS approach was no longer significant (12.2% versus 10.5%, p = 0.08). CONCLUSIONS: For early-stage lung cancer, nodal upstaging was observed more frequently with thoracotomy than with VATS. However, nodal upstaging appears to be affected by facility type, which may be a surrogate for expertise in minimally invasive surgical procedures.

Novel risk stratification score for predicting early distant brain failure and salvage whole-brain radiotherapy after stereotactic radiosurgery for brain metastases.

BACKGROUND: The purpose of this study was to evaluate predictors of early distant brain failure (DBF) and salvage whole-brain radiotherapy (WBRT) after treatment with stereotactic radiosurgery (SRS) for brain metastases and create a clinically relevant risk score to stratify patients' risk for these events. METHODS: The records of 270 patients with brain metastases who were treated with SRS between 2003 and 2012 were reviewed. Pretreatment patient and tumor characteristics were analyzed with univariate and multivariate analyses. The cumulative incidences of first DBF and salvage WBRT were calculated. Significant factors were used to create a score for stratifying early (6-month) DBF risk. RESULTS: No prior WBRT, a total lesion volume < 1.3 cm(3), primary breast cancer or malignant melanoma histology, and multiple metastases (≥2) were found to be significant predictors of early DBF. Each factor was ascribed 1 point because of similar hazard ratios. Scores of 0 to 1, 2, and 3 to 4 were considered to indicate low, intermediate, and high risk, respectively. This correlated with 6-month cumulative incidences of DBF of 16.6%, 28.8%, and 54.4%, respectively (P < .001). For patients without prior WBRT, the 6-month cumulative incidence of salvage WBRT was 2%, 17.7%, and 25.7%, respectively (P < .001). CONCLUSIONS: Early DBF after SRS requiring salvage WBRT remains a significant clinical problem. Patient stratification for early DBF can better inform the decision for the initial treatment strategy for brain metastases. The provided risk score may help to predict early DBF and subsequent salvage WBRT if SRS is initially used. External validation is needed before clinical implementation.

Nonclinical Factors Associated with 30-Day Mortality after Lung Cancer Resection: An Analysis of 215,000 Patients Using the National Cancer Data Base.

BACKGROUND: Clinical variables associated with 30-day mortality after lung cancer surgery are well known. However, the effects of nonclinical factors, including insurance coverage, household income, education, type of treatment center, and area of residence, on short-term survival are less appreciated. We studied the National Cancer Data Base, a joint endeavor of the Commission on Cancer of the American College of Surgeons and the American Cancer Society, to identify disparities in 30-day mortality after lung cancer resection based on these nonclinical factors. STUDY DESIGN: We performed a retrospective cohort analysis of patients undergoing lung cancer resection from 2003 to 2011 using the National Cancer Data Base. Data were analyzed using a multivariable logistic regression model to identify risk factors for 30-day mortality. RESULTS: During our study period, 215,645 patients underwent lung cancer resection. We found that clinical variables, such as age, sex, comorbidity, cancer stage, preoperative radiation, extent of resection, positive surgical margins, and tumor size were associated with 30-day mortality after resection. Nonclinical factors, including living in lower-income neighborhoods with a lesser proportion of high school graduates, and receiving cancer care at a nonacademic medical center were also independently associated with increased 30-day postoperative mortality. CONCLUSIONS: This study represents the largest analysis of 30-day mortality for lung cancer resection to date from a generalizable national cohort. Our results demonstrate that, in addition to known clinical risk factors, several nonclinical factors are associated with increased 30-day mortality after lung cancer resection. These disparities require additional investigation to improve lung cancer patient outcomes.

Locoregional prostate cancer treatment pattern variation in independent cancer centers: policy effect, patient preference, or physician incentive?

Surveillance, Epidemiologic, and End Results (SEER) registry data abstracted from a priority 2 or higher reporting source from 2006 to 2008 were used to compare treatment patterns in 45-64-year old men diagnosed with locoregional prostate cancer (LRPC) across states with or without radiation therapy-directed certificate of need (CON) laws and across independent cancer centers (ICCs) compared to large multi-specialty groups (LMSGs). Adjusted treatment percentages for the five most common LRPC treatments (surgery, external beam radiation therapy (EBRT), combination brachytherapy with EBRT, brachytherapy, and observation) were compared using cross-sectional logistic regression between CON-unregulated and -regulated states and between LMSGs and ICCs. LRPC EBRT rates were no different across CON regions, but are increased in ICCs compared to LMSGs (37.00% vs. 13.23%, P < 0.001). Variation in LRPC treatment patterns by reporting source merits further scrutiny under the Affordable Care Act of 2010, considering the intent of incentivized accountable care organizations (ACOs) established by the Patient Protection and Affordable Care Act of 2010 (PPACA) and the implications of early descriptions of these new healthcare provider organizations on prostate cancer treatment patterns.

18F-FDG-PET/CT parameters as imaging biomarkers in oral cavity squamous cell carcinoma, is visual analysis of PET and contrast enhanced CT better than the numbers?

PURPOSE: This study was designed to seek associations between positron emission tomography/computed tomography (PET/CT) parameters, contrast enhanced neck computed tomography (CECT) and pathological findings, and to determine the potential prognostic value of PET/CT and CECT parameters in oral cavity squamous cell carcinoma (OCSCC). MATERIALS AND METHOD: 36 OCSCC patients underwent staging PET/CT and 30/36 of patients had CECT. PET/CT parameters were measured for the primary tumor and the hottest involved node, including maximum, mean, and peak standardized uptake values (SUV max, SUV mean, and SUV peak), metabolic tumor volume (MTV), total lesion glycolysis (TLG), standardized added metabolic activity (SAM), and normalized standardized added metabolic activity (N SAM). Qualitative assessment of PET/CT and CECT were also performed. Pathological outcomes included: perineural invasion, lymphovascular invasion, nodal extracapsular spread, grade, pathologic T and N stages. Multivariable logistic regression models were fit for each parameter and outcome adjusting for potentially confounding variables. Multivariable Cox proportional hazards models were used for progression free survival (PFS), locoregional recurrence free survival (LRFS), overall survival (OS) and distant metastasis free survival (DMFS). RESULTS: In multivariable analysis, patients with high (≥ median) tumor SUV max (OR 6.3), SUV mean (OR 6.3), MTV (OR 19.0), TLG (OR 19.0), SAM (OR 11.7) and N SAM (OR 19.0) had high pathological T-stage (T3/T4) (p<0.05). Ring/heterogeneous pattern on CECT qualitative assessment was associated with worse DMFS and OS. CONCLUSION: High PET/CT parameters were associated with pathologically advanced T stage (T3/T4). Qualitative assessment of CECT has prognostic value. PET/CT parameters did not predict clinical outcome.

Do 18F-FDG PET/CT parameters in oropharyngeal and oral cavity squamous cell carcinomas indicate HPV status?

OBJECTIVE: The aim of this study was to explore the relationship of PET/CT parameters with human papillomavirus (HPV) status of oropharyngeal (OP) and oral cavity (OC) squamous cell carcinomas (SCCs). PATIENTS AND METHODS: We retrospectively reviewed 39 patients with OC and OP-SCC who underwent staging 18F-FDG PET/CT. PET/CT parameters were measured for the primary tumor and the hottest involved node, including SUV max, SUV mean, SUV peak, metabolic tumor volume, total lesion glycolysis, standardized added metabolic activity (SAM), and normalized SAM. Patient characteristics were compared between HPV positive (HPV+) and negative (HPV-) groups. Receiver operating characteristic analysis was used to dichotomize PET/CT parameters into high and low. Logistic regression models predicting HPV status were fit for each PET/CT parameter. RESULTS: The HPV+ group was composed of 18 patients all with OP-SCC; the HPV- group consisted of 21 patients, 4 OP cancer patients and 17 OC cancer patients. The HPV+ group had a higher proportion of N2 stage (94% vs 43%; P < 0.001). Nodal PET/CT parameters were higher in the HPV+ group (P < 0.01); this difference was not present for the primary lesion. After adjusting for sex and age, the association of higher nodal SUV max (odds ratio [OR], 9.67), SUV mean (OR, 10.48), SUV peak (OR 9.67), metabolic tumor volume (OR, 14.52), total lesion glycolysis (OR, 11.84), and SAM, normalized SAM (OR, 16.21) with HPV+ status remained statistically significant (P < 0.05). CONCLUSIONS: Nodal PET/CT parameters predict HPV status. High nodal FDG uptake should raise suspicion for positive HPV status in the evaluation of the primary lesion.

Postoperative radiotherapy is associated with better survival in non-small cell lung cancer with involved N2 lymph nodes: results of an analysis of the National Cancer Data Base.

INTRODUCTION: Use of postoperative radiotherapy (PORT) in non-small-cell lung cancer remains controversial. Limited data indicate that PORT may benefit patients with involved N2 nodes. This study evaluates this hypothesis in a large retrospective cohort treated with chemotherapy and contemporary radiation techniques. METHODS: The National Cancer Data Base was queried for patients diagnosed 2004-2006 with resected non-small-cell lung cancer and pathologically involved N2 (pN2) nodes also treated with chemotherapy. Multivariable Cox proportional hazards model was used to assess factors associated with overall survival (OS). Inverse probability of treatment weighting (IPTW) using the propensity score was used to reduce selection bias. OS was compared between patients treated with versus without PORT using the adjusted Kaplan-Meier estimator and weighted log-rank test based on IPTW. RESULTS: Two thousand and one hundred and fifteen patients were eligible for analysis. 918 (43.4%) received PORT, 1197 (56.6%) did not. PORT was associated with better OS (median survival time 42 months with PORT versus 38 months without, p = 0.048). This effect was significant in multivariable and IPTW Cox models (hazard ratio: 0.87, 95% confidence interval: 0.78-0.98, p = 0.026, and hazard ratio: 0.89, 95% confidence interval: 0.79-1.00, p = 0.046, respectively). No interaction was seen between the effects of PORT and number of involved lymph nodes (p = 0.615). CONCLUSIONS: PORT was associated with better survival for patients with pN2 nodes also treated with chemotherapy. No interaction was seen between benefit of PORT and number of involved nodes. These findings reinforce the benefit of PORT for N2 disease in modern practice using the largest, most recent cohort of chemotherapy-treated pN2 patients to date.

Socioeconomic risk factors for long-term mortality after pulmonary resection for lung cancer: an analysis of more than 90,000 patients from the National Cancer Data Base.

BACKGROUND: Several clinical variables, such as tumor stage and age, are well established factors associated with long-term survival after surgical resection of lung cancer. Our aim was to examine the impact of other clinical and demographic variables, controlling for known predictors of long-term survival, in order to investigate how outcomes varied according to important nonclinical factors. STUDY DESIGN: The National Cancer Data Base, jointly supported by the Commission on Cancer of the American College of Surgeons and the American Cancer Society, was used to identify patients undergoing pulmonary resection for lung cancer and perform a retrospective cohort study. The cohort consisted of patients diagnosed with nonsmall cell lung cancer from 2003 to 2006, who underwent resection; overall survival data are available only for patients diagnosed through 2006. A Cox proportional hazards survival model was used to examine factors associated with risk of mortality. RESULTS: A total of 92,929 patients were identified as diagnosed during the study period and undergoing surgical resection for lung cancer. On multivariable analysis, several socioeconomic factors such as lack of insurance, lower income, less education, and treatment at community centers vs academic or research programs predicted worse overall survival after controlling for disease characteristics known to be predictors of worse survival, such as tumor stage, histology, age, and extent of resection. CONCLUSIONS: Diminished long-term survival after pulmonary resection was associated with a number of socioeconomic factors. To date, this represents the largest database analysis of long-term mortality in patients undergoing surgical resection for lung cancer. The disparities in survival outcomes reported here require further detailed investigation.

Neighborhood influences on girls' obesity risk across the transition to adolescence.

BACKGROUND AND OBJECTIVES: The neighborhoods in which children live, play, and eat provide an environmental context that may influence obesity risk and ameliorate or exacerbate health disparities. The current study examines whether neighborhood characteristics predict obesity in a prospective cohort of girls. METHODS: Participants were 174 girls (aged 8-10 years at baseline), a subset from the Cohort Study of Young Girls' Nutrition, Environment, and Transitions. Trained observers completed street audits within a 0.25-mile radius around each girl's residence. Four scales (food and service retail, recreation, walkability, and physical disorder) were created from 40 observed neighborhood features. BMI was calculated from clinically measured height and weight. Obesity was defined as BMI-for-age ≥ 95%. Logistic regression models using generalized estimating equations were used to examine neighborhood influences on obesity risk over 4 years of follow-up, controlling for race/ethnicity, pubertal status, and baseline BMI. Fully adjusted models also controlled for household income, parent education, and a census tract measure of neighborhood socioeconomic status. RESULTS: A 1-SD increase on the food and service retail scale was associated with a 2.27 (95% confidence interval, 1.42 to 3.61; P < .001) increased odds of being obese. A 1-SD increase in physical disorder was associated with a 2.41 (95% confidence interval, 1.31 to 4.44; P = .005) increased odds of being obese. Other neighborhood scales were not associated with risk for obesity. CONCLUSIONS: Neighborhood food and retail environment and physical disorder around a girl's home predict risk for obesity across the transition from late childhood to adolescence.

Necrotizing sialometaplasia-like change of the esophageal submucosal glands is associated with Barrett's esophagus.

The esophageal submucosal glands (SMG) protect the squamous epithelium from insults such as gastroesophageal reflux disease by secreting mucins and bicarbonate. We have observed metaplastic changes within the SMG acini that we have termed oncocytic glandular metaplasia (OGM), and necrotizing sialometaplasia-like change (NSMLC). The aim of this study is to evaluate the associated clinicopathological parameters of, and to phenotypically characterize the SMG metaplasias. Esophagectomy specimens were retrospectively assessed on hematoxylin and eosin sections and assigned to either a Barrett's esophagus (BE) or non-BE control group. Clinicopathologic data was collected, and univariate analysis and multivariate logistic regression models were performed to assess the adjusted associations with NSMLC and OGM. Selected cases of SMG metaplasia were characterized. SMG were present in 82 esophagi that met inclusion criteria. On univariate analysis, NSMLC was associated with BE (p = 0.002). There was no relationship between NSMLC and patient age, sex, tumor size, or treatment history. OGM was associated with BE (p = 0.031). No relationship was found between OGM and patient age, sex, or tumor size. On multivariate analysis, BE was independently associated with NSMLC (odds ratio [OR] 4.95, p = 0.003). Treatment history was also independently associated with OGM (p = 0.029), but not NSMLC. Both NSMLC and OGM were non-mucinous ductal type epithelia retaining a p63-smooth muscle actin co-positive myoepithelial cell layer. NSMLC and OGM were present in endoscopic mucosal resection specimens. Our study suggests that SMG metaplasia is primarily a reflux-induced pathology. NSMLC may pose diagnostic dilemmas in resection specimens or when only partially represented in mucosal biopsies or endoscopic resection specimens.

The association between C-reactive protein (CRP) level and biochemical failure-free survival in patients after radiation therapy for nonmetastatic adenocarcinoma of the prostate.

BACKGROUND: C-reactive protein (CRP) has been associated with outcomes in patients with metastatic adenocarcinoma of the prostate. Associations between prostate adenocarcinoma-specific endpoints and CRP in patients who are treated for localized disease remain unknown. METHODS: In total, 206 patients who received radiation therapy for adenocarcinoma of the prostate had at least 1 CRP measured in follow-up and were analyzed. The primary outcome was biochemical failure-free survival. In addition, associations were examined between CRP and prostate-specific antigen (PSA). RESULTS: On univariate analysis, higher CRP levels were associated significantly with shorter biochemical failure-free survival for patients who received radiation therapy after undergoing radical prostatectomy. For patients who were managed with definitive radiation therapy alone, higher CRP levels also were associated significantly with shorter biochemical failure-free survival on univariate and multivariable analyses (hazard ratio, 2.03; 95% confidence interval, 1.19-3.47; P = .009). In addition, CRP levels were associated significantly with PSA after radical prostatectomy for patients who had Gleason scores ≥ 8 (P = .037), for high-risk patients (P = .008), and for those with pretreatment PSA levels > 20 ng/mL (P = .05). In patients who received definitive radiation therapy, CRP levels also were associated with PSA both for those with pretreatment PSA levels > 20 ng/mL (P < .001), and for the intermediate-risk (P = .029) and high-risk (P = .009) subgroups. CONCLUSIONS: A higher CRP level was associated with shorter biochemical failure-free survival on univariate and multivariable analyses in patients who received definitive radiation therapy. CRP was also associated with PSA in exploratory subgroups. These findings warrant further exploration in a prospectively enrolled patient cohort.

CHFR protein expression predicts outcomes to taxane-based first line therapy in metastatic NSCLC.

PURPOSE: Currently, there is no clinically validated test for the prediction of response to tubulin-targeting agents in non-small cell lung cancer (NSCLC). Here, we investigated the significance of nuclear expression of the mitotic checkpoint gene checkpoint with forkhead and ringfinger domains (CHFR) as predictor of response and overall survival with taxane-based first-line chemotherapy in advanced stage NSCLC. METHODS: We studied a cohort of 41 patients (median age 63 years) with advanced NSCLC treated at the Atlanta VAMC between 1999 and 2010. CHFR expression by immunohistochemistry (score 0-4) was correlated with clinical outcome using chi-square test and Cox proportional models. A cutoff score of "3" was determined by receiver operator characteristics analysis for "low" CHFR expression. Results were validated in an additional 20 patients who received taxane-based chemotherapy at Emory University Hospital and the Atlanta VAMC. RESULTS: High expression (score = 4) of CHFR is strongly associated with adverse outcomes: the risk for progressive disease after first-line chemotherapy with carboplatin-paclitaxel was 52% in patients with CHFR-high versus only 19% in those with CHFR-low tumors (P = 0.033). Median overall survival was strongly correlated with CHFR expression status (CHFR low: 9.9 months; CHFR high: 6.2 months; P = 0.002). After multivariate adjustment, reduced CHFR expression remained a powerful predictor of improved overall survival (HR = 0.24; 95% CI, 0.1-0.58%; P = 0.002). In the validation set, low CHFR expression was associated with higher likelihood of clinical benefit (P = 0.03) and improved overall survival (P = 0.038). CONCLUSIONS: CHFR expression is a novel predictive marker of response and overall survival in NSCLC patients treated with taxane-containing chemotherapy.