Appropriateness of Empiric Gentamicin and Vancomycin Therapy for Bacteremias in Chronic Dialysis Outpatient Units in the Era of Antibiotic Resistance

Bacteremias in inpatient chronic HD units have been described, but there is little information on bacteremias in ambulatory HD units. To determine the frequency of bacteremia and pathogen distribution in ambulatory chronic HD units, we retrospectively reviewed our experience with 107 bacteremias in 5 chronic ambulatory HD units over a 3 year period. The object of the study was twofold. The first objective was to determine if bacteremias in ambulatory HD setting were substantially different in frequency or type than in the inpatient HD setting. Secondly, febrile patients suspected of having bacteremia in chronic HD patients are often empirically treated with vancomycin and gentamicin. Chronic HD patients require repeated and frequent venous access for HD. Bacteremias are common in chronic HD patients and may be primary or secondary and are often related to venous access site infections. The distributions of bacteremia pathogens in chronic HD patients are predominantly reflective of skin flora, i.e., staphylococci and to lesser extent aerobic Gram-negative bacilli. After S. aureus (MSRA/MSSA) and coagulase-negative staphylococcus (CoNS), enterococci are the next most important Gram-positive pathogens in bacteremic HD patients. Most strains of E. faecalis are sensitive to vancomycin and for practical purposes should be considered as vancomycin sensitive enterococci (VSE). In contrast, most strains of E. faecium are resistant to vancomycin and should be considered as vancomycin resistant enterococci (VRE). We retrospectively reviewed 107 patients on chronic ambulatory HD to determine the adequacy of empiric vancomycin and gentamicin prophylaxis. We found amikacin is preferred to gentamicin and that meropenem is an effective alternate substitution for gentamicin and vancomycin combination therapy.

P3-07-22: Combined Approach for Staging the Axilla in Breast Cancer Patients with Clinically (−)Nodes Versus Sentinel Node Biopsy Alone

Background

Prognostic information and local control for managing the majority of clinically node (−) breast cancer patients may be achieved by sentinel node biopsy (SNB) alone and/or limited axillary dissections (LAD). Currently, 20 - 30% of clinically node (−) patients have unnecessary surgery. This study compares the results from staging the axilla using SNB alone versus using the combined approach shown below.

Methods: Clinically node (−) patients (n=176) were subclassified according to their primary tumor histology, axillary ultrasound (US) data, and US guided Fine Needle Aspiration (US-FNA) results, as follows: Low Risk (LR) for axillary metastasis (n = 62); High Risk (HR) with normal axillary US (n = 17); HR with US suggesting minimal N1a disease (n = 23); HR with US suggesting N1a disease (n = 52); HR with US suggesting N2-3 disease (n = 22). All patients with (+) SNB or (+) US-FNA had Axillary Lymph Node Dissections (ALND). The number of (+) Sentinel Nodes (SN), Non Sentinel Nodes (NSN) and (+) LN after a (+) US-FNA from each patient category was correlated with corresponding preoperative data. HR patients were defined as having grade II tumors ≥ 1.5 cm and grade III tumors > 1.0 cm. US abnormalities in the axilla were interpreted as follows: minimal N1a disease equivalent to cortical defects < 5mm in 1–3 LN; N1a disease, cortical defects > 5mm in 1–3 LN and N2-3 disease, complete nodal replacement in ≥ 1 LN. LAD refers to level I dissections (1-5 LN).

Results: Three subgroups of patients were identified: Group A, patients not requiring ALND (128/176 = 72%); Group B, patients requiring ALND bypassing SNB (22/176 =13%) and Group C, patients requiring LAD (26/176 = 15%). Preoperatively these 3 groups were categorized as follows: Group A included patients at LR for axillary metastasis, HR patients with normal axillary US, HR patients with axillary US suggesting minimal N1a disease and HR patients with axillary US suggesting N1a disease with (−) US-FNA; Group B included HR patients with axillary US suggesting N2-3 disease and (+) US-FNA; Group C included HR patients with axillary US suggesting N1a disease and (+) US-FNA. The post ALND characteristics for these 3 groups are summarized as follows: all Group A patients had N1a disease represented by ≤ 2 (+) LN, 94% (17/18) were SN (+) only, 85% (15/18) with 1 (+) LN and 15% with 2 (+) LN; in Group B, 20 patients had N2-3 disease and 2 patients had N1a disease, all Group B patients had > 2 (+) LN; in group C, 20 patients had N1a disease and 2 patients had N2 disease, and 77% had single (+) node disease.

Conclusion: By following this approach a more patient oriented method for staging the axilla can be implemented as follows: 1. SNB alone for LR patients and for HR patients with axillary US findings suggesting no axillary disease, minimal N1a disease and/or N1a disease with (−) US-FNA; 2. ALND for HR patients with axillary findings suggesting N2-3 disease and a (+) US-FNA; 3. LAD for HR patients with US findings suggesting N1a disease and (+) US-FNA. This approach would result in a 38% (48/176) reduction in the number of SNB and a 30% (22/66) reduction in the number of ALND. This translates in to $200,000 (30-40%) in procedure-associated savings.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P3-07-22.