Mackenzie Daly

Prognostic microRNA signatures derived from The Cancer Genome Atlas for head and neck squamous cell carcinomas

Identification of novel prognostic biomarkers typically requires a large dataset which provides sufficient statistical power for discovery research. To this end, we took advantage of the high‐throughput data from The Cancer Genome Atlas (TCGA) to identify a set of prognostic biomarkers in head and neck squamous cell carcinomas (HNSCC) including oropharyngeal squamous cell carcinoma (OPSCC) and other subtypes. In this study, we analyzed miRNA‐seq data obtained from TCGA patients to identify prognostic biomarkers for OPSCC. The identified miRNAs were further tested with an independent cohort. miRNA‐seq data from TCGA was also analyzed to identify prognostic miRNAs in oral cavity squamous cell carcinoma (OSCC) and laryngeal squamous cell carcinoma (LSCC). Our study identified that miR‐193b‐3p and miR‐455‐5p were positively associated with survival, and miR‐92a‐3p and miR‐497‐5p were negatively associated with survival in OPSCC. A combined expression signature of these four miRNAs was prognostic of overall survival in OPSCC, and more importantly, this signature was validated in an independent OPSCC cohort. Furthermore, we identified four miRNAs each in OSCC and LSCC that were prognostic of survival, and combined signatures were specific for subtypes of HNSCC. A robust 4‐miRNA prognostic signature in OPSCC, as well as prognostic signatures in other subtypes of HNSCC, was developed using sequencing data from TCGA as the primary source. This demonstrates the power of using TCGA as a potential resource to develop prognostic tools for improving individualized patient care.

Comparison of unilateral versus bilateral intensity‐modulated radiotherapy for surgically treated squamous cell carcinoma of the palatine tonsil

The authors hypothesized that unilateral intensity‐modulated radiotherapy (IMRT) would decrease toxicity compared with bilateral IMRT for patients with lateralized palatine tonsillar cancer and a neck classification of N0 to N2b, with similar oncological outcomes.

nab-Paclitaxel, cisplatin, and 5-fluorouracil followed by concurrent cisplatin and radiation for head and neck squamous cell carcinoma

OBJECTIVES We previously reported the efficacy of nab-paclitaxel added to cisplatin, 5-FU, and cetuximab (APF-C) followed by concurrent high dose bolus cisplatin and radiation therapy (CRT) in patients with locally advanced head and neck squamous cell carcinoma (HNSCC). In this phase II trial, we determined the efficacy of APF (without cetuximab) followed by CRT in similar patients. MATERIALS AND METHODS Eligible patients had stage III-IV oropharynx (OP), larynx, or hypopharynx SCC and adequate organ function and performance status. T1 tumors were excluded. Patients were treated with three cycles of APF followed by CRT. Efficacy endpoints included two-year disease-specific survival (DSS), progression-free survival (PFS), overall survival (OS), and relapse rate. RESULTS Thirty patients were enrolled. Most patients were smokers (77%) with bulky T3/4 (73%) and N2/3 (83%) tumors. Analyses were stratified for human papilloma virus (HPV) status: HPV-related OPSCC (n=17; 57%) and HPV-unrelated HNSCC (n=13; 43%). With a minimum follow-up of 21months, relapse occurred in 1 (3%) patient. Two-year DSS was 94% in HPV-related OPSCC and 100% in HPV-unrelated HNSCC. Two-year PFS was 94% in HPV-related OPSCC and 100% in HPV-unrelated HNSCC. Two-year OS was 94% in HPV-related OPSCC and 92% in HPV-unrelated HNSCC. Causes of death were relapse (1), treatment-related mortality (1), and co-morbidity (1). Two patients with HPV-unrelated HNSCC treated with APF declined CRT and remained free of relapse at 36 and 28months of follow-up. CONCLUSION This phase II trial demonstrated favorable two-year DSS, PFS, and OS and a low relapse rate in HPV-unrelated HNSCC and HPV-related OPSCC treated with APF followed by CRT.

nab-Paclitaxel-based induction chemotherapy with or without cetuximab for locally advanced head and neck squamous cell carcinoma

OBJECTIVES To explore the effect of incorporating cetuximab into induction chemotherapy in locally advanced head and neck squamous cell carcinoma (HNSCC). MATERIALS AND METHODS Retrospective comparative analysis of two consecutive prospective phase II trials was performed: trial 1 with nab-paclitaxel/cisplatin/5-FU and cetuximab (APF-C; n=30) and trial 2 with APF (n=30). Patients were scheduled to receive chemoradiation therapy (CRT) with cisplatin. T2-4 classification oropharynx (OP)/larynx/hypopharynx SCC were included. Cumulative incidence of death of disease (CIDD), overall survival (OS), and cumulative incidence of relapse were compared between APF-C and APF. RESULTS No significant differences in patient or tumor characteristics were noted between the groups. Median follow-up of surviving patients was 52 (25-95) months. Relapse occurred in 5 (17%) patients treated with APF-C and in 2 (7%) treated with APF (p=0.37). In human papillomavirus (HPV)-related OPSCC (n=34), the CIDD at 52months was 3.4% with APF-C and 2.6% with APF and the two-year OSs were 94%. In HPV-unrelated HNSCC (n=25), the CIDD at 52months was 4.4% with APF-C and 3.3% with APF and two-year OSs were 83% and 92%, respectively. CIDD or OS did not differ when stratified by treatment group and HPV status (CIDD: p=0.80; OS: p=0.30). CONCLUSION This exploratory retrospective comparative analysis demonstrated no significant difference in CIDD, OS, or cumulative incidence of relapse between patients treated with APF-C or APF.

Correlation of Ki-67 Proliferative Antigen Expression and Tumor Response to Induction Chemotherapy Containing Cell Cycle-Specific Agents in Head and Neck Squamous Cell Carcinoma.

Determine if highly proliferative head and neck squamous cell carcinomas, assessed by pretreatment Ki-67 expression, respond more robustly to induction chemotherapy (IC) that is selectively toxic to cycling cells. Retrospective analysis of 59 patients treated with IC and chemoradiation. IC included either nab-paclitaxel, cisplatin, 5-FU and cetuximab (APF-C, n = 27) or docetaxel, cisplatin, 5-FU +/− cetuximab (TPF+/−C, n = 32). Ki-67 expression was assessed by immunohistochemistry. Tumor response (complete/partial/stable/progressive) at the primary site after two IC cycles was evaluated by visual examination in all patients. In the APF-C sub-group, tumor response (primary site and neck nodes) after two IC cycles was evaluated by computed tomography (CT) and fluorodeoxyglucose-positron emission tomography (FDG-PET)/CT. Ki-67 expression (median 66%, range: 16–97) did not differ across the tumor response categories assessed by visual examination (p = 0.95), CT (p = 0.30), or FDG-PET/CT (p = 0.65). Median decrease in summed SUVmax of measured lesions was 71.6% (range: 8.3–100%). The Pearson correlation coefficient between Ki-67 expression and the percent decrease in summed SUVmax was 0.48 (p = 0.02). Ki-67 expression was not different between those with or without a relapse (median: 60 and 71%, p = 0.10). In multivariate regression analysis (MVA) controlling for p16 positive oropharyngeal SCC status and smoking status, Ki-67 expression was not significantly associated with tumor response by visual examination (coefficient estimate −0.002, standard error 0.010, p = 0.84), CT (coefficient estimate −0.007, standard error 0.011, p = 0.54), FDG-PET/CT (coefficient estimate 0.006, standard error 0.008, p = 0.51), the percent decrease in summed SUVmax (coefficient estimate 0.389, standard error 0.222, p = 0.09), or relapse events (OR = 1.02(95%CI:0.99–1.05), p = 0.28). No significant relationships were found in MVA between pretreatment Ki-67 expression and tumor response to IC or to relapse.

Post-operative radiation effects on lymphopenia, neutrophil to lymphocyte ratio, and clinical outcomes in palatine tonsil cancers

OBJECTIVE To evaluate radiation-induced lymphopenia associated with unilateral vs. bilateral neck radiation and to test post-treatment neutrophil to lymphocyte ratio (NLR) as a prognostic clinical biomarker. METHODS This was a single academic center retrospective review of palatine tonsil squamous cell cancer patients treated with post-operative intensity modulated radiation therapy (IMRT) from 1997 to 2013. Absolute lymphocyte count (ALC) and NLR were evaluated during and after radiation for up to a year. Correlations of lab values with loco-regional control (LRC), freedom from distant metastases (FFDM), and overall survival (OS) were assessed. RESULTS Ninety-nine patients with median follow up 5.8 years had ALC recorded at least at baseline and within one year of starting RT. Acute grade 3-4 lymphopenia (<10 weeks from RT start) occurred in 79% of bilateral neck RT patients (n = 70) and 58% of unilateral neck RT patients (n = 29), p = 0.03. There was no significant difference in late grade 3-4 (p = 0.12) lymphopenia. In a multivariable Cox regression model, acute NLR > 11.875 correlated with worse OS (HR = 4.4, 95% CI 1.2-16). Late NLR > 6.875 independently correlated with significantly worse FFDM (HR = 16, 95% CI 1.9-137) and OS (HR = 12, 95% CI 3.0-48). CONCLUSIONS Unilateral neck radiation may prevent acute iatrogenic immunosuppression. In exploratory analyses, elevated post-treatment NLR was associated with risk for distant metastases and death.

Predictors of acute throat or esophageal patient reported pain during radiation therapy for head and neck cancer

Background and purpose Acute pain during weekly radiotherapy (RT) to the head and neck is not well characterized. We studied dose-volume metrics and clinical variables that are plausibly associated with throat or esophageal pain as measured with a weekly questionnaire during RT. Materials and methods We prospectively collected weekly patient-reported outcomes from 122 head and neck cancer patients during RT. The pain score for each question consisted of a four-level scale: none (0), mild (1), moderate (2), and severe (3). Univariate and multivariate ordinal logistic regression analyses were performed to investigate associations between both esophageal and throat pain and clinical as well as dosimetric variables. Results In multivariate analysis, age was significantly associated with both types of pain, leading to odds ratio (OR) = 0.95 (p = 0.008) and OR = 0.95 (p = 0.007) for esophageal and throat pain, respectively. For throat pain, sex (OR = 4.12; p = 0.010), with females at higher risk, and fractional organ at risk (OAR) mean dose (OR = 3.30; p = 0.014) were significantly associated with throat pain. Conclusions A fractional OAR mean dose of 1.1 Gy seems a reasonable cutoff for separating no or mild pain from moderate to severe esophageal and throat pain. Younger patients who received RT experienced more esophageal and throat pain. Females experienced more throat pain, but not esophageal pain.

Induction chemotherapy in the treatment of nasopharyngeal carcinoma: Clinical outcomes and patterns of care

The role of induction chemotherapy in nasopharyngeal carcinoma (NPC) remains controversial. The primary aim of this study was to use the National Cancer Database to evaluate the patterns of care of induction chemotherapy in NPC and its impact on overall survival (OS). Patients with NPC from 2004 to 2014 were obtained from the NCDB. Patients were considered to have received induction chemotherapy if it was started ≥43 days before the start of RT and concurrent CRT if chemotherapy started within 21 days after the start of RT. Propensity score matching was used to control for selection bias. Cox proportional hazards model was used to determine significant predictors of OS. Logistic regression model was used to determine predictors of the use of induction chemotherapy. Significance was defined as a P value <.05. A total of 4857 patients were identified: 4041 patients (87.2%) received concurrent CRT and 816 patients (16.8%) received induction chemotherapy. The use of induction therapy remained stable between 2004 and 2014. Younger patients and those with higher T‐ and N‐stage had a higher likelihood of being treated with induction chemotherapy. The 5‐year OS in patients treated with induction chemotherapy and CRT was 66.3% vs 69.1%, respectively (P = .25). There was no difference in OS when these two groups were analyzed after propensity score matching. No differences in OS existed between these treatment groups in patients with T3‐T4N1 or TanyN2‐3 disease (P = .76). Propensity score matching also did not reveal any difference in OS in patients with T3‐T4N1 or TanyN2‐3 disease. The use of induction chemotherapy has remained stable in the last decade. In this study of patients with NPC, induction chemotherapy was not associated with improved OS compared to CRT alone.

Palliative radiation therapy (RT) for prostate cancer patients with bone metastases at diagnosis: A hospital-based analysis of patterns of care, RT fractionation scheme, and overall survival

Prostate cancer (PCa) is one of the most common malignancies associated with bone metastases, and palliative radiation therapy (RT) is an effective treatment option. A total of 2641 patients were identified with PCa and bone metastases at diagnosis from 2010 to 2014 in the NCDB. Fractionation scheme was designated as short course ([SC‐RT]: 8 Gy in 1 fraction and 20 Gy in 5 fractions) vs long course ([LC‐RT]: 30 Gy in 10 fractions and 37.5 Gy in 15 fractions). Patient characteristics were correlated with fractionation scheme using logistic regression. Overall survival was analyzed using the Kaplan‐Meier method, log‐rank test, Cox proportional hazards models, and propensity score‐matched analyses. A total of 2255 (85.4%) patients were included in the LC‐RT group and 386 (14.6%) patients in the SC‐RT group. SC‐RT was more common in patients over 75 years age (odds ratio [OR]: 1.70, 95% confidence interval [CI] 1.32‐2.20), treatment at an academic center (OR: 1.76, 1.20‐2.57), living greater than 15 miles distance to treatment facility (OR: 1.38, 1.05‐1.83), treatment to the rib (OR: 2.99, 1.36‐6.60), and in 2014 (OR: 1.73, 1.19‐2.51). RT to the spine was more commonly long course (P < .0001). In the propensity‐matched cohort, LC‐RT was associated with improved OS (P < .0001), but no OS difference was observed between 37.5 Gy and either 8 Gy in one fraction or 20 Gy in 5 fractions (P > .5). LC‐RT remains the most common treatment fractionation scheme for palliative bone metastases in PCa patients. Use of palliative SC‐RT is increasing, particularly in more recent years, for older patients, treatment at academic centers, and with increasing distance from a treatment center.

Determining the Role of Gene Expression Assays in Local Therapy for Breast Cancer: First Steps with a Positive Impression

Determining when post-mastectomy radiation therapy (PMRT) should be utilized in breast cancer patients with T1 or T2 tumors and 1–3 positive lymph nodes is an area of uncertainty. In this patient population, balancing the benefits of improved local control and overall survival with the long-term complications of therapeutic radiation relies heavily on the ‘art of medicine’. The paper submitted by Dr. Jegadeesh and colleagues published in this issue of Annals of Surgical Oncology suggests that the 21-gene recurrence score assay may be a tool to optimize local therapy for breast cancer patients. Furthermore, the authors offer a novel perspective on utilizing the recurrence score for clinical decision making. For the reader, it is helpful to appreciate the precarious position that PMRT has held in treatment paradigms. Early trials, such as the DBCG 82 b&c trials, included patients with one or more positive axillary nodes and showed improved local control as well as improved survival. Questions regarding the adequacy of surgery and older systemic agents used in these trials make many clinicians wary about applying historic results to modern-day patients. By the late 1990s a consensus report was released stating that insufficient evidence exists to make evidencebased recommendations regarding PMRT in patients with small tumors and 1–3 positive axillary lymph nodes. More recently, the treatment pendulum has been swinging back with the results of the Early Breast Cancer Trialist’s Collaborative Group meta-analysis, reporting reduction of locoregional recurrence (LRR) and improvement in breast cancer mortality for this patient population. In the absence of clear, randomized data, oncologists are forced to weigh clinicopathologic risk factors, ultimately relying on their clinical judgment to recommend or not recommend PMRT in patients with small primary tumors (T1–T2) and 1–3 positive lymph nodes. While tallying up an individual patient’s risk factors for local recurrence, lymphovascular invasion, tumor size, nodal involvement, extracapsular extension, age, etc., I cannot help but reflect on my medical oncology colleagues. They too were once weighing clinical and pathologic risk factors down to the millimeter, using actuarial science to treat individual patients with biologically heterogeneous tumors. That is until the 21-gene recurrence score became widely available, allowing medical oncologists to tailor systemic therapy recommendations for estrogen receptor (ER)positive, HER2-negative breast cancer patients by integrating their individual molecular profiles. Although systemic management of early-stage breast cancer patients has moved forward into a genomic era, local therapies such as radiation and surgery continue to treat cancer empirically. In the current issue of Annals of Surgical Oncology, the article entitled ‘‘21-Gene Recurrence Score and Locoregional Recurrence in Breast Cancer Patients’’ attempts to clarify if there is a relationship between the 21-gene recurrence score and local recurrence. The authors retrospectively reviewed records of stages I and II, ER-positive breast cancer patients who were treated with breast-conserving therapy (BCT) or mastectomy alone. Patients received systemic therapy based upon their 21-gene recurrence score. The study was limited by rather marked differences in patient and tumor characteristics for each local control group. Most notably, there was a statistically significant difference in patient age, stage, and lymph node status between the BCT and mastectomy-alone groups. Additionally, there was a very low event rate in the Society of Surgical Oncology 2014