Most electrical synapses in the mammalian central nervous system are composed of Connexin 36 (Cx36). Electrical synapses are functionally plastic, changing their degree of coupling based on the activity of the cell or connected cells, or based on activation of neurotransmitter or neurohormone receptors. Plasticity can reach an extreme in which electrical synapses become functionally silent, which is a normal operational condition for some circuits. Cx36 coupling is regulated by phosphorylation, which opens the channels. In retinal circuits, Cx36 is often maintained in a poorly phosphorylated, poorly coupled state. We reasoned that phosphomimetic mutants of Cx36 could remain constitutively open and maintain circuits in a well-coupled state that will interrogate the need for plasticity. We developed a constitutively open Cx36 mutant by systematically replacing phosphorylatable residues that regulate coupling with acidic residues. Single mutants of serine 315 significantly modified functional regulation of coupling in HeLa cells, but mutation of four residues was required to produce a mutant that was constitutively open. This mutant, Cx36-S110D, T111E, S293D, S315D, called Cx36-DEDD, displayed high coupling in control conditions and only modest changes under phosphorylating and dephosphorylating conditions. We developed a conditional knockin mouse that expresses Cx36-DEDD and cytoplasmic tdTomato in cells that expressed Cre recombinase. When crossed with Six3-Cre mice, Cx36-DEDD expressed widely in the retina including in photoreceptors, bipolar, amacrine and ganglion cells. Rod-cone electrical coupling displayed the maximum of its physiological dynamic range, and photopic visual acuity and contrast sensitivity were significantly reduced in Cx36-DEDD homozygous animals. We conclude that reduction of coupling in some retinal circuits is required for optimal daylight vision. Significance Statement:Two types of synapses, chemical and electrical, work together throughout the central nervous system to perform neurological functions. While it is widely understood for chemical synapses that plasticity, changing the strength of synaptic connections, plays critical roles in many processes, this is far less understood for electrical synapses. By developing an electrical synapse protein mutant that locks channels in an open state, we have investigated retinal circuits that retain functional electrical synapses but lack their latitude for plasticity. This perturbation significantly compromises visual acuity and contrast sensitivity in the daylight, revealing that electrical synapse plasticity is necessary to tune retinal functions for optimal performance. Thus, electrical synapse plasticity along with chemical synapse plasticity is required for neural function.

Calmodulin binding is a nearly universal property of gap junction proteins, imparting a calcium-dependent uncoupling behavior that can serve in an emergency to decouple a stressed cell from its neighbors. However, gap junctions that function as electrical synapses within networks of neurons routinely encounter large fluctuations in local cytoplasmic calcium concentration; frequent uncoupling would be impractical and counterproductive. We have studied the properties and functional consequences of calmodulin binding to the electrical synapse protein Connexin 35 (Cx35 or gjd2b), homologous to mammalian Connexin 36 (Cx36 or gjd2). We find that specializations in Cx35 calmodulin binding sites make it relatively impervious to moderately high levels of cytoplasmic calcium. Calmodulin binding to a site in the C-terminus causes uncoupling when calcium reaches low micromolar concentrations, a behavior prevented by mutations that eliminate calmodulin binding. However, milder stimuli promote calcium/calmodulin-dependent protein kinase II activity that potentiates coupling without interference from calmodulin binding. A second calmodulin binding site in the end of the Cx35 cytoplasmic loop, homologous to a calmodulin binding site present in many connexins, binds calmodulin with very low affinity and stoichiometry. Together, the calmodulin binding sites cause Cx35 to uncouple only at extreme levels of intracellular calcium.

A variety of electrical synapses are capable of activity-dependent plasticity, including both activity-dependent potentiation and activity-dependent depression. In several types of neurons, activity-dependent electrical synapse plasticity depends on changes in the local Ca2+ environment. To enable study of local Ca2+ signaling that regulates plasticity, we developed a GCaMP Ca2+ biosensor fused to the electrical synapse protein Connexin 36 (Cx36). Cx36-GCaMP transfected into mammalian cell cultures formed gap junctions at cell-cell boundaries and supported Neurobiotin tracer coupling that was regulated by protein kinase A signaling in the same way as Cx36. Cx36-GCaMP gap junctions robustly reported local Ca2+ increases in response to addition of a Ca2+ ionophore with increases in fluorescence that recovered during washout. Recovery was strongly dependent on Na+-Ca2+ exchange activity. In cells transfected with NMDA receptor subunits, Cx36-GCaMP revealed transient and concentration-dependent increases in local Ca2+ on brief application of glutamate. In HeLa cells, glutamate application increased Cx36-GCaMP tracer coupling through a mechanism that depended in part on Ca2+, calmodulin-dependent protein kinase II (CaMKII) activity. This potentiation of coupling did not require exogenous expression of glutamate receptors, but could be accomplished by endogenously expressed glutamate receptors with pharmacological characteristics reminiscent of NMDA and kainate receptors. Analysis of RNA Sequencing data from HeLa cells confirmed expression of NMDA receptor subunits NR1, NR2C, and NR3B. In summary, Cx36-GCaMP is an effective tool to measure changes in the Ca2+ microenvironment around Cx36 gap junctions. Furthermore, HeLa cells can serve as a model system to study glutamate receptor-driven potentiation of electrical synapses.

Calcium signals act as a ubiquitous secondary messenger in regulating many body functions. The detection of calcium microdomain signals is greatly facilitated by the existence of biomarker-targeted fluorescent probes. In this study, SRRF (super-resolution radial fluctuations) algorithm were used to compare the loci and the intensity of fluorescent probes before and after SRRF analysis. The implementation of SRRF algorithm was aimed for automatically resolving delicate and small calcium signals (to avoid the overlapped loci) on original images. For assessing the spatial accuracy of image intensity, immunofluorescence staining of retina cryostat slice for connexin 36 (Cx36) was microscopically imaged with or without the successive SRRF reconstruction. For characterizing the temporal association between SRRF and non-SRRF images, the changes of Cx36-GCaMP calcium indicator were recorded from transfected HeLa cells in response to the transient puffing of ionomycin. Image processing and analyses were conducted with Image J and Matlab. Through this study, SRRF reconstruction was found to confer an accurate measure for the identification of subcellular molecules, such as gap junctions. Compared with the conventional imaging, SRRF reconstruction generated better image resolution for the precise registration of individual signals. Temporally, the ratios of change in fluorescence intensity between SRRF and non-SRRF images were significantly correlated in the presence or absence of the subtraction of high background intensity. Quantitatively, the ratios of change in fluorescence intensity between SRRF and non-SRRF images with or without background subtraction were also significantly correlated. The merit of SRRF application on calcium live imaging was validated with the reporter gene system we worked on.

ABSTRACTIn the mammalian central nervous system, a remarkably small number of connexins is used in electrical synapses, with the majority formed from Cx36. A larger number has been detected in teleosts, with some seeming to serve restricted roles. Here, we report the discovery of a new connexin expressed in the zebrafish lens and a limited set of neurons. Zebrafish cx79.8 (gja8a), previously annotated incorrectly as cx50.5 based on a partial cDNA sequence, is a homologue of mammalian Cx50 (Gja8). We examined its expression through transgenic promoter‐reporter constructs, in situ hybridization, and immunolabeling, and examined regulation of coupling in transfected HeLa cells. cx79.8 was expressed most strongly in the lens, but expression was also found in several groups of neurons in the cerebellum and related areas at the midbrain‐hindbrain boundary, in cone photoreceptors, and in neurons in the retinal inner nuclear and ganglion cell layers. Labeling in the retina with antibodies against two C‐terminal epitopes revealed numerous small punctate spots in the inner plexiform layer and along the somata of cones. Abundant gap junctions were labeled in the outer 1/3 of the lens, but were absent from the center, suggesting that the epitopes or the entire protein was absent from the center. Cx79.8 tracer coupling was strongly regulated by phosphorylation, and was extremely low in control conditions in HeLa cells due to protein phosphatase 2A activity. These properties allow coupling to be strongly restricted in situ, a frequently observed property for electrical synapses. © 2016 Wiley Periodicals, Inc. Develop Neurobiol 77: 548–561, 2017

Gap junctions formed of connexin 36 (Cx36, also known as Gjd2) show tremendous functional plasticity on several time scales. Changes in connexin phosphorylation modify coupling in minutes through an order of magnitude, but recent studies also imply involvement of connexin turnover in regulating cell-cell communication. We utilized Cx36 with an internal HaloTag to study Cx36 turnover and trafficking in cultured cells. Irreversible, covalent pulse-chase labeling with fluorescent HaloTag ligands allowed clear discrimination of newly formed and pre-existing Cx36. Cx36 in junctional plaques turned over with a half-life of 3.1 h, and the turnover rate was unchanged by manipulations of protein kinase A (PKA) activity. In contrast, changes in PKA activity altered coupling within 20 min. New Cx36 in cargo vesicles was added directly to existing gap junctions and newly made Cx36 was not confined to points of addition, but diffused throughout existing gap junctions. Existing connexins also diffused into photobleached areas with a half-time of less than 2 s. In conclusion, studies of Cx36-HaloTag revealed novel features of connexin trafficking and demonstrated that phosphorylation-based changes in coupling occur on a different time scale than turnover.

Purpose/Objective(s)Limited data exists on the implications of lymphovascular space invasion (LVSI), extensive intraductal component (EIC), and multifocality (MF) in appropriately selecting and risk-stratifying patients undergoing accelerated partial breast irradiation (APBI). We present five-year outcomes in patients undergoing APBI based on these factors.Materials/MethodsBetween 1993 and 2011, 692 patients with early-stage breast cancer were treated with APBI at a single institution, receiving either interstitial brachytherapy (n = 195), balloon-based brachytherapy (n = 292), or 3D-CRT (n = 205) as part of breast conserving therapy. Patients were stratified into patients with (n = 64) and without (n = 419) LVSI, patients with (n = 24) and without (n = 289) EIC, and patients with (n = 8) and without (n = 668) MF. Clinical outcomes analyzed include ipsilateral breast tumor recurrence (IBTR), regional recurrence (RR), distant metastases (DM), disease-free survival (DFS), cause-specific survival (CSS), and overall survival (OS).ResultsPatients with and without LVSI were well balanced with respect to treatment type, age, and, histology; however patients with LVSI had larger tumors (15.6 vs. 11.4mm, p < 0.001), higher grade (44% vs. 14% grade 3, p < 0.001), higher rates of ER negative disease (19% vs.10%, p = 0.02), higher rates of HER-2 positive disease (19% vs. 7%, p = 0.006), more node positive disease (20% vs. 7%, p = 0.001), and were more likely to receive chemotherapy (31% vs. 14%, p < 0.001). No differences were noted in the rates of IBTR (1.8% vs. 3.7%, p = 0.10), CSS (98% vs. 95%, p = 0.29), or OS (90% vs. 87%, p = 0.45) while statistically significant differences were noted in RR (0% vs. 4.4%, p < 0.001), DM (1.9% vs. 8.7%, p = 0.005), and DFS (96% vs. 90%, p = 0.02) in LVSI negative and LVSI positive patients, respectively. Patients with and without EIC were well balanced with respect to age, tumor size, treatment type, grade, receptor status, LVSI, margin and nodal status; however, patients with EIC were more likely to have received chemotherapy (38% vs. 15%, p = 0.008). In patients with and without MF, characteristics were well balanced with the exception of increased tumor size in patients with MF (15.9 vs. 10.9mm, p = 0.03). There were no IBTRs in patients with and without EIC; however, presence of EIC was associated with a higher rate of RR (5.3% vs. 0.6%, p = 0.008) with a trend for increased DM (10.8% vs. 2.7%, p = 0.10). No events occurred in patients with and without MF limiting comparisons between the groups.ConclusionsLVSI and EIC were associated with modest increases in regional and distant failure. The limited number of events within our patient population precludes definitive conclusions regarding multifocality. Ongoing prospective study of risk factors for recurrence after APBI including LVSI and EIC is warranted. Purpose/Objective(s)Limited data exists on the implications of lymphovascular space invasion (LVSI), extensive intraductal component (EIC), and multifocality (MF) in appropriately selecting and risk-stratifying patients undergoing accelerated partial breast irradiation (APBI). We present five-year outcomes in patients undergoing APBI based on these factors. Limited data exists on the implications of lymphovascular space invasion (LVSI), extensive intraductal component (EIC), and multifocality (MF) in appropriately selecting and risk-stratifying patients undergoing accelerated partial breast irradiation (APBI). We present five-year outcomes in patients undergoing APBI based on these factors. Materials/MethodsBetween 1993 and 2011, 692 patients with early-stage breast cancer were treated with APBI at a single institution, receiving either interstitial brachytherapy (n = 195), balloon-based brachytherapy (n = 292), or 3D-CRT (n = 205) as part of breast conserving therapy. Patients were stratified into patients with (n = 64) and without (n = 419) LVSI, patients with (n = 24) and without (n = 289) EIC, and patients with (n = 8) and without (n = 668) MF. Clinical outcomes analyzed include ipsilateral breast tumor recurrence (IBTR), regional recurrence (RR), distant metastases (DM), disease-free survival (DFS), cause-specific survival (CSS), and overall survival (OS). Between 1993 and 2011, 692 patients with early-stage breast cancer were treated with APBI at a single institution, receiving either interstitial brachytherapy (n = 195), balloon-based brachytherapy (n = 292), or 3D-CRT (n = 205) as part of breast conserving therapy. Patients were stratified into patients with (n = 64) and without (n = 419) LVSI, patients with (n = 24) and without (n = 289) EIC, and patients with (n = 8) and without (n = 668) MF. Clinical outcomes analyzed include ipsilateral breast tumor recurrence (IBTR), regional recurrence (RR), distant metastases (DM), disease-free survival (DFS), cause-specific survival (CSS), and overall survival (OS). ResultsPatients with and without LVSI were well balanced with respect to treatment type, age, and, histology; however patients with LVSI had larger tumors (15.6 vs. 11.4mm, p < 0.001), higher grade (44% vs. 14% grade 3, p < 0.001), higher rates of ER negative disease (19% vs.10%, p = 0.02), higher rates of HER-2 positive disease (19% vs. 7%, p = 0.006), more node positive disease (20% vs. 7%, p = 0.001), and were more likely to receive chemotherapy (31% vs. 14%, p < 0.001). No differences were noted in the rates of IBTR (1.8% vs. 3.7%, p = 0.10), CSS (98% vs. 95%, p = 0.29), or OS (90% vs. 87%, p = 0.45) while statistically significant differences were noted in RR (0% vs. 4.4%, p < 0.001), DM (1.9% vs. 8.7%, p = 0.005), and DFS (96% vs. 90%, p = 0.02) in LVSI negative and LVSI positive patients, respectively. Patients with and without EIC were well balanced with respect to age, tumor size, treatment type, grade, receptor status, LVSI, margin and nodal status; however, patients with EIC were more likely to have received chemotherapy (38% vs. 15%, p = 0.008). In patients with and without MF, characteristics were well balanced with the exception of increased tumor size in patients with MF (15.9 vs. 10.9mm, p = 0.03). There were no IBTRs in patients with and without EIC; however, presence of EIC was associated with a higher rate of RR (5.3% vs. 0.6%, p = 0.008) with a trend for increased DM (10.8% vs. 2.7%, p = 0.10). No events occurred in patients with and without MF limiting comparisons between the groups. Patients with and without LVSI were well balanced with respect to treatment type, age, and, histology; however patients with LVSI had larger tumors (15.6 vs. 11.4mm, p < 0.001), higher grade (44% vs. 14% grade 3, p < 0.001), higher rates of ER negative disease (19% vs.10%, p = 0.02), higher rates of HER-2 positive disease (19% vs. 7%, p = 0.006), more node positive disease (20% vs. 7%, p = 0.001), and were more likely to receive chemotherapy (31% vs. 14%, p < 0.001). No differences were noted in the rates of IBTR (1.8% vs. 3.7%, p = 0.10), CSS (98% vs. 95%, p = 0.29), or OS (90% vs. 87%, p = 0.45) while statistically significant differences were noted in RR (0% vs. 4.4%, p < 0.001), DM (1.9% vs. 8.7%, p = 0.005), and DFS (96% vs. 90%, p = 0.02) in LVSI negative and LVSI positive patients, respectively. Patients with and without EIC were well balanced with respect to age, tumor size, treatment type, grade, receptor status, LVSI, margin and nodal status; however, patients with EIC were more likely to have received chemotherapy (38% vs. 15%, p = 0.008). In patients with and without MF, characteristics were well balanced with the exception of increased tumor size in patients with MF (15.9 vs. 10.9mm, p = 0.03). There were no IBTRs in patients with and without EIC; however, presence of EIC was associated with a higher rate of RR (5.3% vs. 0.6%, p = 0.008) with a trend for increased DM (10.8% vs. 2.7%, p = 0.10). No events occurred in patients with and without MF limiting comparisons between the groups. ConclusionsLVSI and EIC were associated with modest increases in regional and distant failure. The limited number of events within our patient population precludes definitive conclusions regarding multifocality. Ongoing prospective study of risk factors for recurrence after APBI including LVSI and EIC is warranted. LVSI and EIC were associated with modest increases in regional and distant failure. The limited number of events within our patient population precludes definitive conclusions regarding multifocality. Ongoing prospective study of risk factors for recurrence after APBI including LVSI and EIC is warranted.

To analyze the impact of change in glycolytic activity and FDG-PET volume over time on rate and pattern of local recurrence after stereotactic body radiation therapy (SBRT) for early-stage non-small cell lung cancer (NSCLC). One hundred sixty-four cases of NSCLC treated with SBRT at a single institution with follow-up ≥1 year were analyzed for factors related to local recurrence (LR vs. no-LR). In general, RT dose was 12 Gy x 4 for tumors ≤3 cm and 12 Gy x 5 for tumors >3 cm, heterogeneity-corrected. In-field LR was defined as recurrent tumor primarily within the 90% isodose line (IDL) versus marginal miss if the majority of a tumor recurred beyond (but still involving) the 90% IDL. Volumetric PET data using software with a PET-edge gradient tool was collected for all LR tumors. For comparison, volumetric PET data was also measured for a matched population of no-LR tumors (matched for tumor size, histology, and initial max SUV). Glycolytic activity (GA: PET volume*meanSUV) and PET volume (PV) were then recorded for all LR and matched no-LR tumors at baseline, and 6 weeks, 3-, 6-, and 12-months post treatment. Median follow-up was 1.7 yrs (range: 1.0-6.1 yrs), tumor size was 2.6 cm (range: 0.6-6.4 cm), mean GTV was 15.9 cc (range: 0.7-127.4 cc), and mean max-SUV was 8.6 (range: 1.0-32.1) for all patients. At baseline, mean overall GA was 100.5 SUVbw*cc (range: 1.57-861.6 SUVbw*cc) and PV was 12.4 cc (range: 0.65-92.82 cc). In total, 15 LRs (8.6%) were identified, 13 of which were in-field recurrences (86%). The GTV was larger for LR versus no-LR tumors (25.1 vs. 14.9 cc, p = 0.81) as was max pre-tx SUV (12.5 vs. 8.2, p = 0.02). Twenty-six no-LR tumors were suitably matched to the LR cohort for comparison of GA and PV over time; thus, volumetric assessment of FDG-PET avidity at five potential time points was performed for 41 total tumors. Following SBRT, GA and PV changed at different rates for LR versus no-LR tumors. At 3 months post-tx, mean GA was 168.7 SUV*cc lower than baseline for no-LRs versus 36.7 SUV*cc higher for tumors that would eventually recur (p = 0.056). Similarly, PV decreased more in no-LR versus LR-tumors at 3 months (-14.9 cc vs. +14.2 cc, p = 0.037). Volume change at other time points was not statistically different. Although a similar proportion of LR and no LR patients received 48 and 60 Gy (p = 0.70) a trend towards decreased total dose per centimeter of maximum tumor dimension (20.1 Gy/cm vs. 24.2 Gy/cm, p = 0.07) and total dose per volume of tumor (7.3 Gy/cc vs. 10.2 Gy/cc, p = 0.15) was seen. Local recurrence after SBRT is rare. Change in glycolytic activity and PET volume from baseline are statistically different at three months post-treatment for locally recurrent versus non-recurrent lung tumors. These metrics, as well as dose per tumor dimension, deserve additional study regarding their role in the management of early-stage NSCLC.

To analyze factors related to local recurrence (LR) following stereotactic body radiation therapy (SBRT) for early-stage non-small cell lung cancer (NSCLC). One hundred sixty-four cases of NSCLC were treated with SBRT between 9/2005 and 12/2010 at a single institution. Per institutional policy, tumors < 3cm received 48 Gy in 4 fractions vs. 60 Gy in 5 fractions for tumors > 3cm, heterogeneity-corrected. A clinical/internal target volume (CTV-ITV) was created using 4DCT simulation to account for tumor motion due to respiration as well as microscopic extension. SBRT fractions were delivered with a minimum interfraction time of 40 hours and prescription dose covering 95% of the PTV. Patients with > 1 year follow up were analyzed for evidence of LR, which was defined as recurrent tumor within the same lobe of the lung. Patient demographics, tumor characteristics, and treatment parameters were analyzed for differences between LR and no-LR cases for all patients. Median follow-up was 1.9 years (range: 1.0-6.1 yrs) for all patients and 2.1 years (range: 1.2-5.7 yrs) for the patients with a LR. Median age was 75.6 years (range: 42-93), median tumor size was 2.6cm (range: 0.6-6.4cm), and mean GTV was 15.9cc (range: 0.7-127.4cc). Mean maximum SUV was 8.6 (range: 1.0-32.1) for all patients. Forty percent of patients were treated as part of an IRB-approved in-house SBRT protocol and most were not candidates for surgery (88.4%). 60% of tumors were located within the right lung and most tumors were located in the upper lobes (53%). We identified 15 tumors out of 164 cases treated meeting criteria for LR (8.6%). A positive biopsy was obtained in 6 of 15 cases (40%) and LR identified on imaging alone in the remaining 9 cases. Gross tumor volume (GTV) was larger for LR vs. no-LR tumors (25.1cc vs. 14.9cc, p=0.081) as was mean baseline SUV max (12.5 vs. 8.2, p=0.021). Patients with LR were also more likely to have squamous cell carcinoma histology (71% vs. 32%, p=0.004) and receive chemotherapy (40% vs. 17%, p=0.034) versus those that did not have a recurrence. There was no statistical difference between prescription dose for LR vs. no-LR groups (48 Gy: 65%, 60Gy: 35%, both groups, p=0.698). Other dosimetric parameters such as percent of prescription dose to 80 and 90 percent of PTV were not different for LR vs. no-LR tumors (p=0.543/p=0.623). Time from first evidence of tumor on CT until simulation, however, trended towards being longer for patients with LR (5.3 vs. 3.8 months, p=0.116). Our data show that SBRT for early-stage NSCLC is an effective treatment with a local failure rate less than 9% using 48-60Gy in 4-5 fractions. Recurrent tumors were more likely to be squamous cell histology and have a higher SUV prior to treatment with SBRT. Delays in start of treatment may contribute to local recurrence after SBRT for early-stage lung cancer.

Abstract Purpose: To determine the impact of estrogen receptor (ER) negativity on clinical outcomes for patients treated with Accelerated Partial Breast Irradiation (APBI). Materials/Methods(s): We evaluated 506 consecutive patients treated with interstitial brachytherapy (n= 199), balloon-based brachytherapy (n=203), and 3D-CRT (n=104). ER negative (ERN) status was assigned using the traditional definition of an ER nuclear IHC stain < 10%, which corresponds to an Allred/NSABP staining score of < 2. 63 patients (12.5%) were ER negative and 443 (87.5%) were ER positive (ERP). Patient demographics and clinical outcomes (IBTR, RNF, DM, DFS, CSS, OS) were analyzed for each group. Results: The two groups had similar patient characteristics. Tumor sizes were slightly larger for the ERN group at 11.9mm vs. 10.7mm, although this was not statistically significant (p=0.14). No differences were seen in median age (63 vs. 64 years, p=0.36), rate of HER-2/neu overexpression (83% vs. 91%, p=0.11), or lymph node positivity (6% vs. 9%, p=0.55) between the ERN vs. ERP groups, respectively. There were an equal distribution of invasive ductal carcinoma (ERN n=55, 87%; ERP n=387, 87%) and DCIS (ERN n=8, 13%; ERP n=56, 13%) patients within each group. The use of chemotherapy (55% vs. 15%, p<0.001) and nuclear grade (71% vs. 12%, p<0.001) were higher in the ERN vs. ERP cohort. With a mean follow up of 6.1 years, the 5-year actuarial rates of ipsilateral breast tumor recurrence (IBTR), regional nodal failure (RNF), and distant metastasis (DM) for the entire cohort were 1.8%, 0.6%, and 3.2%. Although this was not statistically significant, ERN patients appear to have an increased rate of local failure than patients with ERP histology (4.0% vs. 1.5%, p=0.13). Rates of RNF and DM were, however, significantly higher for the ERN group (RNF: 4.9% ERN vs. 0% ERP, p<0.001; DM: 12.1% ERN vs. 2.0% ERP, p<0.001). Although there was no difference in overall survival at six years (86% vs. 90%, p=0.67), we observed a shorter disease-free survival (86.4% vs. 96.5%, p= 0.01) and cause-specific survival (90% vs. 98%, p=0.01) for the ERN vs. ERP groups. Conclusion: The ER negative phenotype of early-stage breast cancer may have a decreased rate of locoregional control. We observed a higher rate of DM with reduced disease-free and cause-specific survival in ER negative cases, emphasizing the importance of systemic therapy and careful, long-term follow up for these patients. Prospective study of this histologic subtype with a larger cohort of patients is needed to substantiate these findings. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P3-13-09.