[Show abstract][Hide abstract] ABSTRACT: Background:
Head and neck cancer is morbid with a poor prognosis that has not significantly improved in the past several decades. The purpose of this study was to identify biological pathways underlying progressive head and neck cancer to inform prognostic and adjuvant strategies. We identified 235 head and neck cancer patients in The Cancer Genome Atlas (TCGA) with sufficient clinical annotation regarding therapeutic treatment and disease progression to identify progressors and non-progressors. We compared primary tumor gene expression and mutational status between these two groups.
105 genes were differentially expressed between progressors and nonprogressors (FDR < 0.05). Pathway analyses revealed deregulation (FDR < 0.05) of multiple pathways related to integrin signaling as well as IL-10 signaling. A number of genes were uniquely mutated in the progressor cohort including increased frequency of truncating mutations in CTCF (P = 0.007). An 11-gene signature derived from a combination of unique mutations and differential expression was identified (PAGE4, SMTNL1, VTN, CA5A, C1orf43, KRTAP19-1, LEP, HRH4, PAGE5, SEZ6L, CREB3). This signature was associated with decreased overall survival (Logrank Test; P = 0.03443). Cox modeling of both key clinical features and the signature was significant (P = 0.032) with the greatest prognostic improvement seen in the model based on nodal extracapsular spread and alcohol use alone (P = 0.004).
Molecular analyses of head and neck cancer tumors that progressed despite treatment, identified IL-10 and integrin pathways to be strongly associated with cancer progression. In addition, we identified an 11-gene signature with implications for patient prognostication. Mutational analysis highlighted a potential role for CTCF, a crucial regulator of long-range chromatin interactions, in head and neck cancer progression.
[Show abstract][Hide abstract] ABSTRACT: Prostate cancer patients undergoing localized external beam radiation therapy (EBRT) can experience a progressive increase in fatigue, which can affect physical functioning and quality of life. The purpose of this study was to develop a mouse EBRT prostate cancer treatment model with which to determine the role of pro-inflammatory cytokines in the genesis of EBRT-related fatigue. We assessed voluntary wheel-running activity (VWRA) as a proxy for fatigue, food intake and body weight in male C57BL/6 mice undergoing EBRT to the pelvis. In the first experiment, anesthetized male C57BL/6 mice underwent fractionated EBRT to the pelvis for a total dose of 68.2 Gy, thereby mimicking a clinically relevant therapeutic dose and frequency. The day after the last treatment, levels of IL-1β and TNF-α in plasma along with mRNA levels in liver, colon and whole brain were measured. EBRT-induced fatigue resulted in reduced body weight, diminished food intake, and increased plasma and tissue levels of IL-1β and TNF-α. In a follow-up experiment, we used TNF-α-deficient mice to further delineate the role of TNF-α signaling in EBRT-induced sickness behavior. EBRT-induced changes in fatigue, food intake and body weight were no different between TNF-α deficient mice and their wild-type counterparts. Taken together our data demonstrate that a clinically relevant localized irradiation of the pelvis induces a systemic IL-1β and TNF-α response and sickness behavior in mice, but the TNF-α signaling pathway alone does not independently mediate these effects.
No preview · Article · Dec 2015 · Radiation Research
[Show abstract][Hide abstract] ABSTRACT: Background:
Anal cancer treatment has evolved from abdominoperineal resection to chemoradiotherapy, which allows for sphincter preservation.
The aim of this study was to develop an accurate model and nomogram to predict overall survival and the probability of salvage abdominoperineal resection for anal cancer patients.
This is a retrospective cohort study.
Data were gathered from National Cancer Database entries from 1998 to 2010.
Patients with de novo anal cancer were selected from the National Cancer Database in the years 1998 through 2010; 1778 patients were included, and their data were analyzed.
Main outcome measures:
Variables included time to death, censoring indicator, age, race, sex, tumor size, year of diagnosis, surgery status, nodal status, TNM stage, and chemoradiation therapy. A stratified Cox proportional hazards model for overall survival and a logistic regression model for salvage abdominoperineal resection were developed. Our final models were internally validated for discrimination and validation.
Statistically significant variables in the salvage surgery model were tumor size and nodal status (p ≤ 0.001). For overall survival model, statistically significant variables (all with p ≤ 0.005), fitted across the strata of TNM clinical stage included age, sex, tumor size, nodal status, chemoradiotherapy treatment, and combination salvage surgery and chemoradiotherapy. Nomograms that predict events are based on our final models.
Limitations included clerical database errors and nonmeasured variables, such as HIV status.
A nomogram can predict overall survival and salvage surgery for an individual with anal cancer. Such tools may be used as decision support aids to guide therapy and predict whether or not patients may need salvage surgery.
No preview · Article · Dec 2015 · Diseases of the Colon & Rectum
[Show abstract][Hide abstract] ABSTRACT: Introduction:
Esophageal cancer is the eighth most common cancer worldwide and the sixth leading cause of cancer-related deaths. As a significant cause of morbidity and mortality, its burden on society has yet to be fully characterized. The aim of this study is to examine its global burden through estimation of the disability-adjusted life years (DALYs) attributable to it.
Global incidence and mortality estimates for esophageal cancer were obtained from the International Agency for Research on Cancer GLOBOCAN 2008 database. DALYs were calculated, using methodology established by the World Health Organization.
In 2008, 3,955,919 DALYs were attributed to esophageal cancer, at a global rate of 0.58 DALYs per 1000 people annually. Years of life lost (YLL) accounted for 96.8 % of DALYs, while years lived with disability (YLD) accounted for 3.2 %. 83.8 % of the global DALYs occurred in less-developed countries, with most accrued in Eastern Asia, comprising 50.9 % of the total. The highest rate of DALY accrual was in Southern Africa, at 1.62 DALYs per 1000 people annually.
A substantial number of years of life were lost or affected by esophageal cancer worldwide in 2008, with the burden resting disproportionately on less-developed countries. Geographically, the greatest burden is in Eastern Asia. The vast majority of DALYs were due to YLL, rather than YLD, indicating the need to focus resources on disease prevention and early detection. Our findings provide an additional basis upon which to formulate global priorities for interventions that affect DALY reduction in esophageal cancer.
Full-text · Article · Dec 2015 · World Journal of Surgery
[Show abstract][Hide abstract] ABSTRACT: Background and purpose:
Mechanisms of fatigue reported during radiotherapy are poorly defined but may include inflammatory cytokines and/or sleep disturbances. This prospective, longitudinal, phase II study assessed fatigue, sleep, and serum cytokine levels during radiotherapy for early-stage prostate cancer (PCa).
Material and methods:
Twenty-eight men undergoing radiotherapy for early-stage PCa wore an Actiwatch Score to record fatigue level, sleep time, onset latency, efficiency and wake after sleep onset. Serum levels of IL-1α, IL-1β, TNF-α, IL-6, IL-8, IL-10 and VEGF were measured weekly during radiotherapy. Patient reported quality of life (QOL) metrics were collected before and after treatment. Linear mixed effects models examined trajectories across treatment weeks.
Fatigue increased across treatment weeks (P<.01), and fatigue was associated with decreased patient-reported QOL. Sleep efficiency increased across treatment weeks (rate of change over time=.29, P=.03), and sleep onset latency decreased (rate of change over time=.86, P=.06). IL-6 tended to increase during treatment (P=0.09), but none of the cytokine levels or sleep variables were significantly related to fatigue trajectories.
Despite increased sleep efficiency across treatment weeks, fatigue significantly increased. Although IL-6 increased during the course of radiotherapy, cytokines levels were not associated with fatigue scores or sleep disturbance. Further studies are needed to define the mechanisms for fatigue during radiotherapy.
Full-text · Article · Dec 2015 · Radiotherapy and Oncology
[Show abstract][Hide abstract] ABSTRACT: Objectives
We sought to quantitatively determine the interobserver variability of expert radiotherapy target-volume delineation for thymic cancers, as part of a larger effort to develop an expert-consensus contouring atlas.
A pilot dataset was created consisting of a standardized case presentation with pre- and post-operative DICOM CT image sets from a single patient with Masaoka-Koga Stage III thymoma. Expert thoracic radiation oncologists delineated tumor targets on the pre- and post-operative scans as they would for a definitive and adjuvant case, respectively. Respondents completed a survey including recommended dose prescription and target volume margins for definitive and post-operative scenarios. Interobserver variability was analyzed quantitatively with Warfield’s simultaneous truth, performance level estimation (STAPLE) algorithm and Dice similarity coefficient (DSC).
Seven users completed contouring for definitive and adjuvant cases; of these, five completed online surveys. Segmentation performance was assessed, with high mean ± SD STAPLE-estimated segmentation sensitivity for definitive case GTV and CTV at 0.77 and 0.80, respectively, and post-operative CTV sensitivity of 0.55; all volumes had specificity of ≥0.99. Interobserver agreement was markedly higher for the definitive target volumes, with mean ± SD DSC of 0.88 ± 0.03 and 0.89 ± 0.04 for GTV and CTV, respectively, compared to post-op CTV DSC of 0.69 ± 0.06 (Kruskal-Wallis p < 0.01.
Expert agreement for definitive case volumes was exceptionally high, though significantly lower agreement was noted post-operatively. Technique and dose prescription between experts was substantively consistent, and these preliminary results will be utilized to create an expert-consensus contouring atlas to aid the nonexpert radiation oncologist in the planning of these challenging, rare tumors.
[Show abstract][Hide abstract] ABSTRACT: To obtain a favorable tradeoff between treatment benefits and morbidity ("therapeutic ratio"), radiotherapy (RT) dose is prescribed according to the tumor volume, with the goal of controlling the disease while respecting normal tissue tolerance levels. We propose a new paradigm for tumor dose prescription in stereotactic ablative radiotherapy (SABR) based on organ-at-risk (OAR) tolerance levels called isotoxic dose prescription (IDP), which is derived from experiences and limitations of conventionally fractionated radiotherapy. With IDP, the radiation dose is prescribed based on the predefined level of normal tissue complication probability of a nearby dose-limiting OAR at a prespecified dose-volume constraint. Simultaneously, the prescribed total tumor dose (TTD) is maximized to the technically highest achievable level in order to increase the local tumor control probability (TCP). IDP is especially relevant for tumors located at eloquent locations or for large tumors in which severe toxicity has been described. IDP will result in a lower RT dose or a treatment scheduled with more fractions if the OAR tolerance level is exceeded, and potential dose escalation occurs when the OAR tolerance level allows it and when it is expected to be beneficial (if TCP < 90%). For patients with small tumors at noneloquent sites, the current SABR dose prescription already results in high rates of local control at low toxicity rates. In this review, the concept of IDP is described in the context of SABR.
Full-text · Article · Oct 2015 · Journal of the National Cancer Institute
[Show abstract][Hide abstract] ABSTRACT: Background:
Local excision is an organ-preserving treatment alternative to transabdominal resection for patients with stage I rectal cancer. However, local excision alone is associated with a high risk of local recurrence and inferior survival compared with transabdominal rectal resection. We investigated the oncological and functional outcomes of neoadjuvant chemoradiotherapy and local excision for patients with stage T2N0 rectal cancer.
We did a multi-institutional, single-arm, open-label, non-randomised, phase 2 trial of patients with clinically staged T2N0 distal rectal cancer treated with neoadjuvant chemoradiotherapy at 26 American College of Surgeons Oncology Group institutions. Patients with clinical T2N0 rectal adenocarcinoma staged by endorectal ultrasound or endorectal coil MRI, measuring less than 4 cm in greatest diameter, involving less than 40% of the circumference of the rectum, located within 8 cm of the anal verge, and with an Eastern Cooperative Oncology Group performance status of at least 2 were included in the study. Neoadjuvant chemoradiotherapy consisted of capecitabine (original dose 825 mg/m(2) twice daily on days 1-14 and 22-35), oxaliplatin (50 mg/m(2) on weeks 1, 2, 4, and 5), and radiation (5 days a week at 1·8 Gy per day for 5 weeks to a dose of 45 Gy, followed by a boost of 9 Gy, for a total dose of 54 Gy) followed by local excision. Because of adverse events during chemoradiotherapy, the dose of capecitabine was reduced to 725 mg/m(2) twice-daily, 5 days per week, for 5 weeks, and the boost of radiation was reduced to 5·4 Gy, for a total dose of 50·4 Gy. The primary endpoint was 3-year disease-free survival for all eligible patients (intention-to-treat population) and for patients who completed chemotherapy and radiation, and had ypT0, ypT1, or ypT2 tumours, and negative resection margins (per-protocol group). This study is registered with ClinicalTrials.gov, number NCT00114231.
Between May 25, 2006, and Oct 22, 2009, 79 eligible patients were recruited to the trial and started neoadjuvant chemoradiotherapy. Two patients had no surgery and one had a total mesorectal excision. Four additional patients completed protocol treatment, but one had a positive margin and three had ypT3 tumours. Thus, the per-protocol population consisted of 72 patients. Median follow-up was 56 months (IQR 46-63) for all patients. The estimated 3-year disease-free survival for the intention-to-treat group was 88·2% (95% CI 81·3-95·8), and for the per-protocol group was 86·9% (79·3-95·3). Of 79 eligible patients, 23 (29%) had grade 3 gastrointestinal adverse events, 12 (15%) had grade 3-4 pain, and 12 (15%) had grade 3-4 haematological adverse events during chemoradiation. Of the 77 patients who had surgery, six (8%) had grade 3 pain, three (4%) had grade 3-4 haemorrhage, and three (4%) had gastrointestinal adverse events.
Although the observed 3-year disease free survival was not as high as anticipated, our data suggest that neoadjuvant chemoradiotherapy followed by local excision might be considered as an organ-preserving alternative in carefully selected patients with clinically staged T2N0 tumours who refuse, or are not candidates for, transabdominal resection.
National Cancer Institute and Sanofi-Aventis.
No preview · Article · Oct 2015 · The Lancet Oncology
[Show abstract][Hide abstract] ABSTRACT: Purpose:
Computer-based, patient-reported symptom survey tools have been described for patients undergoing chemotherapy. We hypothesized that patients undergoing radiotherapy might also benefit, so we developed a computer application to acquire symptom ratings from patients and generate summaries for use at point of care office visits and conducted a randomized, controlled pilot trial to test its feasibility.
Subjects were randomized prior to beginning radiotherapy. Both control and intervention group subjects completed the computerized symptom assessment, but only for the intervention group were printed symptom summaries made available before each weekly office visit. Metrics compared included the Global Distress Index (GDI), concordance of patient-reported symptoms and symptoms discussed by the physician and numbers of new and/or adjusted symptom management medications prescribed.
One hundred twelve patients completed the study: 54 in the control and 58 in the intervention arms. There were no differences in GDI over time between the control and intervention groups. In the intervention group, more patient-reported symptoms were actually discussed in radiotherapy office visits: 46/202 vs. 19/230. A sensitivity analysis to account for within-subjects correlation yielded 23.2 vs. 10.3 % (p = 0.03). Medications were started or adjusted at 15.4 % (43/280) of control visits compared to 20.4 % (65/319) of intervention visits (p = 0.07).
This computer application is easy to use and makes extensive patient-reported outcome data available at the point of care. Although no differences were seen in symptom trajectory, patients who had printed symptom summaries had improved communication during office visits and a trend towards a more active symptom management during radiotherapy.
No preview · Article · Oct 2015 · Supportive Care in Cancer
[Show abstract][Hide abstract] ABSTRACT: Purpose:
Radiation oncology curriculum development is challenging because of limited numbers of trainees at any single institution. The goal of this project is to implement and evaluate a standardized medical student clerkship curriculum following the multi-institutional cooperative group research model.
During the 2013 academic year, a standardized curriculum was implemented at 11 academic medical centers consisting of three 1-hour lectures and a hands-on radiation treatment planning workshop. After the curriculum, students completed anonymous evaluations using Likert-type scales (1 = "not at all" to 5 = "extremely") and free responses. Evaluations asked students to rate their comfort, before and after the curriculum, with radiation oncology as a specialty, knowledge of radiotherapy planning methods, and ability to function as a radiation oncology resident. Nonparametric statistical tests were used in the analysis.
Eighty-eight students at 11 academic medical centers completed the curriculum de novo, with a 72.7% (64 of 88) survey response rate. Fifty-seven students (89.1%) reported intent to pursue radiation oncology as their specialty. Median (interquartile range) student ratings of the importance of curricular content were as follows: overview, 4 (4-5); radiation biology/physics, 5 (4-5); practical aspects/emergencies, 5 (4-5); and planning workshop, 4 (4-5). Students reported that the curriculum helped them better understand radiation oncology as a specialty (5 [4-5]), increased specialty decision comfort (4 [3-5]), and would help the transition to radiation oncology residency (4 [4-5]). Students rated their specialty decision comfort significantly higher after completing the curriculum (4 [4-5] versus 5 [5-5]; P < .001).
A national standardized curriculum was successfully implemented at 11 academic medical centers, providing proof of principle that curriculum development can follow the multi-institutional cooperative group research model.
No preview · Article · Sep 2015 · Journal of the American College of Radiology: JACR
[Show abstract][Hide abstract] ABSTRACT: Low anterior resection or abdominoperineal resection are considered standard treatments for early rectal cancer but may be associated with morbidity in selected patients who are candidates for early distal lesions amenable to local excision (LE). The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed every 3 years by a multidisciplinary expert panel. The guideline development and review include an extensive analysis of current medical literature from peer reviewed journals and the application of a well-established consensus methodology (modified Delphi) to rate the appropriateness of imaging and treatment procedures by the panel. In those instances where evidence is lacking or not definitive, expert opinion may be used to recommend imaging or treatment. The panel recognizes the importance of accurate staging to identify patients who may be candidates for a LE approach. Patients who may be candidates for LE alone include those with small, low-lying T1 tumors, without adverse pathologic features. Several surgical approaches can be utilized for LE however none include lymph node evaluation. Adjuvant radiation±chemotherapy may be warranted depending on the risk of nodal metastases. Patients with high-risk T1 tumors, T2 tumors not amenable to radical surgery may also benefit from adjuvant treatment; however, patients with positive margins or T3 lesions should be offered abdominoperineal resection or low anterior resection. Neoadjuvant radiation±chemotherapy followed by LE in higher risk patients results in excellent local control, but it is not clear if this approach reduces recurrence rates over surgery alone.
No preview · Article · Sep 2015 · American journal of clinical oncology
[Show abstract][Hide abstract] ABSTRACT: Comparison of imaging measurement devices in the absence of a gold-standard comparator remains a vexing problem; especially in scenarios where multiple, non-paired, replicated measurements occur, as in image-guided radiotherapy (IGRT). As the number of commercially available IGRT presents a challenge to determine whether different IGRT methods may be used interchangeably, an unmet need conceptually parsimonious and statistically robust method to evaluate the agreement between two methods with replicated observations. Consequently, we sought to determine, using an previously reported head and neck positional verification dataset, the feasibility and utility of a Comparison of Measurement Methods with the Mixed Effects Procedure Accounting for Replicated Evaluations (COM3PARE), a unified conceptual schema and analytic algorithm based upon Roy's linear mixed effects (LME) model with Kronecker product covariance structure in a doubly multivariate set-up, for IGRT method comparison.
An anonymized dataset consisting of 100 paired coordinate (X/ measurements from a sequential series of head and neck cancer patients imaged near-simultaneously with cone beam CT (CBCT) and kilovoltage X-ray (KVX) imaging was used for model implementation. Software-suggested CBCT and KVX shifts for the lateral (X), vertical (Y) and longitudinal (Z) dimensions were evaluated for bias, inter-method (between-subject variation), intra-method (within-subject variation), and overall agreement using with a script implementing COM3PARE with the MIXED procedure of the statistical software package SAS (SAS Institute, Cary, NC, USA).
COM3PARE showed statistically significant bias agreement and difference in inter-method between CBCT and KVX was observed in the Z-axis (both p - value<0.01). Intra-method and overall agreement differences were noted as statistically significant for both the X- and Z-axes (all p - value<0.01). Using pre-specified criteria, based on intra-method agreement, CBCT was deemed preferable for X-axis positional verification, with KVX preferred for superoinferior alignment.
The COM3PARE methodology was validated as feasible and useful in this pilot head and neck cancer positional verification dataset. COM3PARE represents a flexible and robust standardized analytic methodology for IGRT comparison. The implemented SAS script is included to encourage other groups to implement COM3PARE in other anatomic sites or IGRT platforms.
Preview · Article · Aug 2015 · BMC Medical Imaging
[Show abstract][Hide abstract] ABSTRACT: For resectable gastric cancer, perioperative chemotherapy or adjuvant chemoradiation with chemotherapy are standards of care. The decision making for adjuvant therapeutic management can depend on the stage of the cancer, lymph node positivity, and extent of surgical resection. After gastric cancer resection, postoperative chemotherapy combined with chemoradiation should be incorporated in cases of D0 lymph node dissection, positive regional lymph nodes, poor clinical response to induction chemotherapy, or positive margins. In the setting of a D2 lymph node dissection, especially those with negative regional lymph nodes, adjuvant chemotherapy alone could be considered. The American College of Radiology (ACR) Appropriateness Criteria® are evidence-based guidelines for specific clinical conditions that are reviewed every 3 years by a multidisciplinary expert panel. The guideline development and review includes an extensive analysis of current medical literature from peer-reviewed journals and the application of a well-established consensus methodology (modified Delphi) to rate the appropriateness of imaging and treatment procedures by the panel. In those instances where evidence is lacking or not definitive, expert opinion may be used to recommend imaging or treatment.
No preview · Article · Aug 2015 · Oncology (Williston Park, N.Y.)
[Show abstract][Hide abstract] ABSTRACT: Purpose: To evaluate stressors among radiation oncology residency program directors (PDs) and determine the prevalence and indicators of burnout. Methods and Materials: An anonymous, online, cross-sectional survey was offered to PDs of US radiation oncology programs in the fall of 2014. Survey content examined individual and program demographics, perceptions surrounding the role of PD, and commonly encountered stressors. Burnout was assessed using the validated Maslach Burnout Inventory-Human Services Survey. Results: In total, 47 of 88 PDs (53%) responded to the survey. Although 78% of respondents reported feeling "satisfied" or "highly satisfied" with their current role, 85% planned to remain as PD for <5 years. The most commonly cited stressors were satisfying Accreditation Council for Graduate Medical Education/Residency Review Committee requirements (47%), administrative duties (30%) and resident morale (28%). Three-quarters of respondents were satisfied that they became PDs. Overall, 11% of respondents met criteria for low burnout, 83% for moderate burnout, and 6% for high burnout. Not having served as a PD at a prior institution correlated with high depersonalization (OR 6.75, P = .04) and overall burnout (odds ratio [OR], 15.6; P =. 04). Having more years on faculty prior to becoming PD correlated with less emotional exhaustion (OR, 0.44, P = .05) and depersonalization (OR, 0.20, P = .04). Finally, having dedicated time for PD duties correlated with less emotional exhaustion (OR, 0.27, P = . 04). Conclusions: Moderate levels of burnout are common in U.S. radiation oncology PDs with regulatory stressors being common. Despite this, many PDs are fulfilled with their role. Longitudinal studies assessing dynamic external factors and their influence on PD burnout would be beneficial.
No preview · Article · Aug 2015 · International journal of radiation oncology, biology, physics
[Show abstract][Hide abstract] ABSTRACT: The role of adjuvant radiation for gallbladder carcinoma (GBC) is uncertain. We combine the experience of six National Cancer Institute-designated cancer centers to explore the impact of adjuvant radiation following oncologic resection of GBC.
Patients who underwent extended surgery for GBC at Johns Hopkins, Mayo Clinic, Duke University, Oregon Health & Science University, University of Michigan, and University of Texas MD Anderson between 1985 and 2008 were reviewed. Patients with metastatic disease at surgery, gross residual disease, or missing pathologic information were excluded.
Of the 112 patients identified, 61 % received adjuvant radiation, 93 % of whom received concurrent chemotherapy. Median follow-up of surviving patients was 47.3 (range 2.2-167.7) months. Patients who received adjuvant radiation had a higher rate of advanced T-stage (57 vs. 16 %, p < 0.01), lymph node involvement (63 vs. 18 %, p < 0.01), and positive microscopic margins (37 vs. 9 %, p < 0.01) compared with patients managed with surgery alone, but overall survival (OS) was comparable between the two cohorts (5-year OS: 49.7 vs. 52.5 %, p = 0.20). Lymph node involvement had the strongest association with poor OS (p < 0.01). Adjuvant radiation was associated with decreased isolated local failure (hazard ratio 0.17, 95 % confidence interval 0.05-0.63, p = 0.01). However, 71 % of recurrences included distant failure.
Following oncologic resection for GBC, adjuvant radiation may offer improved local control compared with observation. The benefit of adjuvant radiation beyond chemotherapy alone should therefore be explored. Certainly, the high rate of distant failure highlights the need for more effective systemic therapy.
No preview · Article · Jul 2015 · Annals of Surgical Oncology