Surgical resection of breast cancers: Molecular analysis of cancer stem cells in residual disease.

BACKGROUND: Approximately 70% of breast cancer patients have residual disease after neoadjuvant chemotherapy. This study was designed to determine whether breast cancer cells with stemlike properties are present in residual disease after neoadjuvant chemotherapy and whether they exhibit oncogenic mutations. The presence of breast cancer cells with stemlike properties with specific mutations may help explain the poor prognosis associated with residual disease. METHODS: A total of 68 breast cancer specimens were collected at the time of mastectomy or lumpectomy. A total of 44 were chemotherapy naïve and 24 were collected as residual disease after neoadjuvant chemotherapy. Tumor cells were collected by fluorescence-activated cell sorting, with breast cancer cells with stemlike properties specifically identified using breast stem cell associated antibodies. Whole tumor specimens and fluorescence-activated cell sorting breast cancer cells with stemlike properties were analyzed for genetic mutations, including PIK3CA. RESULTS: Breast cancer cells with stemlike properties, demonstrating EpCAM-positive, CD44-positive, CD49f±, CD24± expression were present in chemotherapy-naïve tumors and residual disease. In both chemotherapy-naïve and residual disease specimens the highest frequency of PIK3CA mutations were detected in CD49f-CD24+ BCSCs (39% and 33%, respectively). PIK3CA mutations were detected in all stages of breast cancer (35%), in both chemotherapy naïve (39%) and residual disease (29%) and in both estrogen receptor positive (41%) and negative tumors (14%) (P = ns). Various PIK3CA mutations were identified in chemotherapy-naïve specimens versus residual disease specimens in both patient-paired and unpaired breast cancers. CONCLUSION: Breast cancer cells with stemlike properties with mutations in PIK3CA were present in chemotherapy-naïve breast cancers and residual disease after neoadjuvant chemotherapy. These results demonstrate that neoadjuvant chemotherapy does not completely eradicate PIK3CA-defective breast cancer cells with stemlike properties. Although these findings may help explain the poor clinical outcomes in patients with residual disease, they also identify breast cancer cells with stemlike-property targets for therapies.

Estrogen receptor, progesterone receptor, interleukin-6 and interleukin-8 are variable in breast cancer and benign stem/progenitor cell populations.

BACKGROUND: Estrogen receptor positive breast cancers have high recurrence rates despite tamoxifen therapy. Breast cancer stem/progenitor cells (BCSCs) initiate tumors, but expression of estrogen (ER) or progesterone receptors (PR) and response to tamoxifen is unknown. Interleukin-6 (IL-6) and interleukin-8 (IL-8) may influence tumor response to therapy but expression in BCSCs is also unknown. METHODS: BCSCs were isolated from breast cancer and benign surgical specimens based on CD49f/CD24 markers. CD44 was measured. Gene and protein expression of ER alpha, ER beta, PR, IL-6 and IL-8 were measured by proximity ligation assay and qRT-PCR. RESULTS: Gene expression was highly variable between patients. On average, BCSCs expressed 10-106 fold less ERα mRNA and 10-103 fold more ERß than tumors or benign stem/progenitor cells (SC). BCSC lin-CD49f-CD24-cells were the exception and expressed higher ERα mRNA. PR mRNA in BCSCs averaged 10-104 fold less than in tumors or benign tissue, but was similar to benign SCs. ERα and PR protein detection in BCSCs was lower than ER positive and similar to ER negative tumors. IL-8 mRNA was 10-104 higher than tumor and 102 fold higher than benign tissue. IL-6 mRNA levels were equivalent to benign and only higher than tumor in lin-CD49f-CD24-cells. IL-6 and IL-8 proteins showed overlapping levels of expressions among various tissues and cell populations. CONCLUSIONS: BCSCs and SCs demonstrate patient-specific variability of gene/protein expression. BCSC gene/protein expression may vary from that of other tumor cells, suggesting a mechanism by which hormone refractory disease may occur.

Red blood cells accelerate the onset of clot formation in polytrauma and hemorrhagic shock.

BACKGROUND: Hemorrhage and coagulopathy are major contributors to death after trauma. The contribution of red blood cells (RBCs) in correcting coagulopathy is poorly understood. Current methods of measuring coagulopathy may fail to accurately characterize in vivo clotting. We aimed to determine the effect of RBCs on clotting parameters by comparing resuscitation regimens containing RBCs and plasma with those containing plasma alone. METHODS: Thirty-two Yorkshire swine were anesthetized, subjected to a complex model of polytrauma and hemorrhagic shock, and resuscitated with either fresh frozen plasma, lyophilized plasma (LP), or 1:1 ratios of fresh frozen plasma:packed RBC (PRBC) or LP:PRBC. Activated clotting time, prothrombin time, partial thromboplastin time, and thrombelastography (TEG) were performed at 1 hour, 2 hours, 3 hours, and 4 hours after resuscitation. RESULTS: Animals treated with 1:1 LP:PRBC had less blood loss than the other groups (p < 0.05). The activated clotting time was shorter in the 1:1 groups when compared with the pure plasma groups at all time points (p < 0.05). The 1:1 groups had shorter TEG R times (time to onset of clotting) at 1 hour, 3 hours, and 4 hours compared with pure plasma groups (p < 0.05). Other TEG parameters did not differ between groups. Partial thromboplastin time was shorter in the pure plasma groups than the 1:1 groups at all time points (p < 0.05). CONCLUSIONS: Whole blood assays reveal that RBCs accelerate the onset of clot formation. Coagulation assays using spun plasma underestimate the effect of RBCs on clotting and do not completely characterize clot formation.

Reproducibility of an animal model simulating complex combat-related injury in a multiple-institution format.

We developed a complex combat-relevant model of abdominal and extremity trauma, hemorrhagic shock, hypothermia, and acidosis. We then simulated injury, preoperative, and operative phases. We hypothesized that this model is reproducible and useful for randomized multicenter preclinical trials. Yorkshire swine were anesthetized, intubated, and instrumented. They then underwent femur fracture, 60% total blood volume hemorrhage, a 30-min shock period, induced hypothermia to 33 degrees C, and hemorrhage volume replacement with 3:1 isotonic sodium chloride solution (NS) at each of three centers. Hemodynamic parameters were measured continuously. Thromboelastography, arterial blood gas, and laboratory values were collected at baseline, after the shock period, and after NS replacement. Thirty-seven animals were used for model development. Eight (21%) died before completion of the study period. Twenty-nine survivors were included in the analysis. MAP (+/-SEM) after the shock period was 32 +/- 2 mmHg and was similar between centers (P = 0.4). Mean pH, base deficit, and lactate levels were 7.29 +/- 0.02, 8.20 +/- 0.65 mmol/L, and 5.29 +/- 0.44 mmol/L, respectively, after NS replacement. These were similar between centers (P > 0.05). Prothrombin time values increased significantly over time at all centers, reflecting a progressive coagulopathy (P < 0.02). Thromboelastography maximum amplitude values were similar among centers (P > 0.05) and demonstrated progressively weakened platelet interaction over time (P < 0.03). Hematocrit was similar after controlled hemorrhage (P = 0.15) and dilution (P = 0.9). The pH, lactate, base deficit, and coagulation tests reflect a severely injured state. A complex porcine model of polytrauma and shock can be used for multi-institutional study with excellent reproducibility. A consistent severe injury profile was achieved, after which experimental interventions can be applied. This is the first report of a reproducible multicenter trauma and resuscitation-related animal model.

A novel highly porous silica and chitosan-based hemostatic dressing is superior to HemCon and gauze sponges.

BACKGROUND: Hemostatic dressings have become increasingly popular as the optimal initial treatment for severe hemorrhage. The purpose of this study was to compare the hemostatic properties of a novel highly porous silica and chitosan-based dressing (TraumaStat) to HemCon, and gauze dressing in a severe groin injury model in swine. METHODS: Thirty swine were blindly randomized to receive TraumaStat, HemCon, or standard gauze dressing for hemostatic control. A complex groin injury involving complete transaction of the femoral artery and vein was made. After 30 seconds of uncontrolled hemorrhage, the randomized dressing was applied and pressure was held for 5 minutes. Fluid resuscitation was initiated to achieve and maintain the baseline mean arterial pressure and the wound was inspected for bleeding. Failure of hemostasis was defined as pooling of blood outside of the wound. Animals were then monitored for 120 minutes and surviving animals were euthanized. RESULTS: Blood loss before treatment was similar between groups (p > 0.1). TraumaStat had one failure, compared with five for gauze, and eight for HemCon (p = 0.005, TraumaStat vs. HemCon). TraumaStat significantly reduced median blood loss when compared with both HemCon and gauze (117 vs. 774 and 268 mL respectively, p < 0.05). At study conclusion, TraumaStat animals had a greater median hematocrit than both HemCon (24 vs. 19, p = 0.033), and gauze (24 vs. 19, p = 0.049) animals. Median volume of fluid resuscitation and mortality were not different between groups (p > 0.1). CONCLUSIONS: TraumaStat was superior to HemCon and gauze dressings in controlling bleeding from a severe groin injury. TraumaStat may be a better hemostatic dressing for control of active hemorrhage than current standards of care.

Ketamine-based total intravenous anesthesia versus isoflurane anesthesia in a swine model of hemorrhagic shock.

BACKGROUND: Inhalational anesthetics can cause profound hemodynamic effects including decreases in systemic vascular resistance and cardiac inotropy. Although widely used in uncontrolled hemorrhagic shock (UHS), their consequences compared with other anesthetic regimens are not well-studied. Ketamine-based total intravenous anesthesia (TIVA) may produce less profound cardiovascular depression, and has been used during elective surgery but rarely during traumatic shock. The purpose of this study was to compare the effects of isoflurane (ISO) and TIVA regimens in a swine grade V liver injury model. We hypothesized that TIVA would result in less hypotension and dysfunctional inflammation than ISO. METHODS: Twenty swine were randomized blindly to receive either 1% to 3% ISO, or intravenous ketamine, midazolam, and buprenorphine for maintenance anesthesia. Six animals acted as controls. After sedation and intubation, randomized anesthesia was initiated and monitored by an independent animal technician. Invasive lines were placed followed by celiotomy and splenectomy. Baseline mean arterial pressure (MAP) was documented and a grade V liver injury created. After 30 minutes of UHS, animals were resuscitated with 8 mL of Ringer's lactate per milliliter blood loss at 165 mL/min. MAP and tissue oxygen saturation (StO2) were continuously recorded. The animals were sacrificed 120 minutes after injury and lung tissue was harvested. Serum cytokines (interleukin-6 [IL-6], IL-8, and tumor necrosis factor-alpha [TNF-alpha]) were quantified with enzyme-linked immunosorbent assay. Lung cytokine mRNA levels were quantified with real time reverse transcriptase polymerase chain reaction. RESULTS: Animal weight, liver injury pattern, and blood loss were similar (p > 0.1). The ISO group had a lower MAP at baseline (p = 0.02), at injury (p = 0.004), and study completion (p = 0.001). After resuscitation, MAP decreased in the ISO group but remained stable in the TIVA group. StO2 was significantly higher in the TIVA group immediately after injury (p = 0.004), but similar between groups throughout the remainder of the study. Animals who received TIVA trended toward higher levels of lactate and lower pH throughout the study, reaching significance at 30 minutes postinjury (p = 0.037 and 0.043). Inflammatory cytokine (IL-6, IL-8, and TNF-alpha) production did not differ between groups, however TNF-alpha mRNA production was significantly lower in the TIVA group (p = 0.04). CONCLUSION: Although a TIVA regimen produced less pronounced hypotension in a swine model of UHS than did ISO, end-organ perfusion with TIVA appeared to be equivalent or inferior to ISO. In circumstances of limited resources, such as those experienced by forward Army surgical teams, a ketamine-based TIVA regimen may be an option for use in UHS.

Lactated Ringer's is superior to normal saline in the resuscitation of uncontrolled hemorrhagic shock.

BACKGROUND: Normal saline (NS) and lactated Ringer's solution (LR) continue to be used interchangeably for the resuscitation of hemorrhagic shock in some institutions. We hypothesized that, aside from hyperchloremic acidosis, NS resuscitation would be similar to that of LR in a swine model of uncontrolled hemorrhage. METHODS: Twenty swine weighing a mean of 37 kg underwent invasive line placement, midline celiotomy, and splenectomy. After a 15-minute stabilization period, we recorded a baseline mean arterial pressure (MAP) and created a grade V liver injury. The animals bled freely for 30 minutes after which we measured blood loss. We blindly randomized the swine to receive NS (10 animals) versus LR (10 animals) to achieve and maintain the baseline MAP for 90 minutes postinjury. Laboratory values were obtained at baseline and upon completion of the 2-hour study period. RESULTS: Initial blood loss was 25 mL/kg in the NS group and 22 mL/kg in the LR group (p = 0.54). Animals required 256.3 +/- 145.4 mL/kg of fluid in the NS group as compared with 125.7 +/- 67.3 mL/kg in the LR group (p = 0.04). The urine output was higher in the NS group (46.6 +/- 39.5 mL/kg versus 18.9 +/- 12.9 mL/kg, p = 0.04). Upon study completion, the NS group had a significant hyperchloremia (119 +/- 1.9 mEq/L versus 105 +/- 2.9 mEq/L, p < 0.01) with acidosis (7.28 +/- 0.12 versus 7.45 +/- 0.06, p < 0.01) in comparison to the LR group. In addition, resuscitation with NS resulted in significantly lower fibrinogen levels (99 +/- 21 mg/dL versus 123 +/- 20 mg/dL, p = 0.02). The serum lactate was 4.7 +/- 2.2 in the LR group and 1.7 +/- 1.7 in the NS swine (p < 0.01) at the end of the study. CONCLUSIONS: Resuscitation of uncontrolled hemorrhagic shock with NS requires significantly greater volume and is associated with greater urine output, hyperchloremic acidosis, and dilutional coagulopathy as compared with LR. Resuscitation with LR results in an elevation of the lactate level that is not associated with acidosis. Lactated Ringer's solution is superior to NS for the resuscitation of uncontrolled hemorrhagic shock in swine.

Fluid resuscitation increases inflammatory gene transcription after traumatic injury.

BACKGROUND: The debate continues over type and quantity of fluid to administer for resuscitation after traumatic injury. This study aimed to examine effects of resuscitation with lactated Ringer's (LR) and Hextend (HEX) on the inflammatory response after uncontrolled hemorrhagic shock (UHS). METHODS: There were 38 swine randomized. Control swine were anesthetized and killed. Sham swine underwent laparotomy, splenectomy, and 2 hours of anesthesia. UHS swine received a grade V liver injury after laparotomy and splenectomy and were randomized to no fluid (NF) resuscitation or to blinded resuscitation 30 minutes after injury with LR or HEX. Fluids were infused as needed to maintain baseline blood pressure for 90 minutes. Lung tissue mRNA levels of interleukin-6 (IL-6), granulocyte colony stimulating factor (G-CSF), and tumor necrosis factor alpha (TNF-alpha) were determined. Lung sections were examined for neutrophils (PMNs) sequestered within alveolar walls. RESULTS: All UHS animals survived and initial blood loss was similar between groups. Mean arterial pressures (MAPs) were similar for all UHS animals until resuscitation was initiated. MAPs of resuscitated animals remained similar and were significantly higher than MAPs of the NF animals. Sequestered PMNs were equally elevated in all UHS animals. Cytokine analysis showed increased IL-6, G-CSF, and TNF-alpha gene transcription in resuscitated swine compared with NF swine. LR and HEX resuscitated swine tissue mRNA levels showed no differences. CONCLUSIONS: Fluid resuscitation after solid organ injury and uncontrolled hemorrhage results in greater proinflammatory gene transcription than no resuscitation. LR and HEX resuscitation have equivalent effects on indices of inflammation in the lungs.

Resuscitation with normal saline (NS) vs. lactated ringers (LR) modulates hypercoagulability and leads to increased blood loss in an uncontrolled hemorrhagic shock swine model.

BACKGROUND: Lactated ringers (LR) and normal saline (NS) are used interchangeably in many trauma centers. The purpose of this study was to compare the effects of LR and NS on coagulation in an uncontrolled hemorrhagic swine model. We hypothesized resuscitation with LR would produce hypercoagulability. METHODS: There were 20 anesthetized swine (35 +/- 3 kg) that underwent central venous and arterial catheterization, celiotomy, and splenectomy. After splenectomy blinded study fluid equal to 3 mL per gram of splenic weight was administered. A grade V liver injury was made and animals bled without resuscitation for 30 minutes. Animals were resuscitated with the respective study fluid to, and maintained, at the preinjury MAP until study end. Prothrombin Time (PT), Partial Thromboplastin Time (PTT), and fibrinogen were collected at baseline (0') and study end (120'). Thrombelastography was performed at 0'and postinjury at 30', 60', 90', and 120'. RESULTS: There were no significant baseline group differences in R value, PT, PTT, and fibrinogen. There was no significant difference between baseline and 30 minutes R value with NS (p = 0.17). There was a significant R value reduction from baseline to 30 minutes with LR (p = 0.02). At 60 minutes, R value (p = 0.002) was shorter while alpha angle, maximum amplitude, and clotting index were higher (p < 0.05) in the LR versus the NS group. R value, PT, and PTT were significantly decreased at study end in the LR group compared with the NS group (p < 0.05). Overall blood loss was significantly higher in the NS versus LR group (p = 0.009). CONCLUSIONS: This data indicates that resuscitation with LR leads to greater hypercoagulability and less blood loss than resuscitation with NS in uncontrolled hemorrhagic shock.

A single bolus of 3% hypertonic saline with 6% dextran provides optimal initial resuscitation after uncontrolled hemorrhagic shock.

BACKGROUND: The optimal fluid for early resuscitation of hemorrhagic shock would restore perfusion without increasing blood loss, hypothermia, acidosis, or coagulopathy. This study examined effects of a single bolus of hypertonic saline (HTS) with or without (+/-) dextran (D) after uncontrolled hemorrhage (UH) and determined optimal fluid composition. METHODS: Fifty swine were anesthetized and underwent invasive line placement, celiotomy, splenectomy, suprapubic catheterization, and grade V liver injury. After 30 minutes of UH, blinded fluid resuscitation was initiated with a 250-mL bolus. Animals were randomized to five groups: normal saline (NS), 3% HTS (3%), 3% HTS/6% D (3% D), 7.5% HTS (7.5%), or 7.5% HTS/6% D (7.5% D). Mean arterial pressure (MAP) and tissue oxygen saturation (StO2) were recorded. Laboratory and thrombelastography (TEG) data were collected every 30 minutes. Animals were sacrificed 120 minutes after injury. Analysis of variance was used to compare groups. Significance was defined as p < 0.05. RESULTS: Baseline characteristics and laboratory values were similar in all groups. All groups achieved a similar degree of shock. Two NS and two 3% animals did not survive to 120 minutes. Fluids containing dextran produced a significantly greater increase in MAP (p < 0.02). Animals receiving 3% D maintained a higher MAP 90 minutes after fluid bolus. Also, 7.5% +/- D produced a significantly greater initial increase in StO2 (p < 0.05). This effect declined after fluid bolus while 3% D continued to improve tissue oxygenation. Significant differences developed between groups in TEG values, hematocrit, fibrinogen, urine sodium, serum sodium, serum chloride, and urine output. CONCLUSIONS: A single bolus of 3% D after uncontrolled hemorrhagic shock produces an adequate and sustained rise in MAP and StO2 and attenuates hypercoagulability. Resuscitation with 7.5% +/- D produces significantly increased urine output accompanied by a decline in MAP and StO2 over time. A single bolus of 7.5% D results in significant dilutional anemia and relative hypofibrinogenemia.

Hextend attenuates hypercoagulability after severe liver injury in swine.

BACKGROUND: Hypercoagulability is a major source of morbidity and mortality after injury. A resuscitation regimen that modulates this coagulopathy may prove beneficial. We sought to evaluate the effects of lactated Ringer's (LR) solution and Hextend on the resuscitation of uncontrolled hemorrhagic shock. METHODS: Twenty swine underwent invasive line placement, midline celiotomy, and splenectomy. After a 15-minute stabilization period, we recorded a baseline mean arterial pressure and created a grade V liver injury. The animals bled freely for 30 minutes, after which we measured the initial blood loss (that after injury). We blindly randomized the swine to receive LR solution or Hextend to achieve and maintain the baseline mean arterial pressure for 90 minutes postinjury. Laboratory values were obtained at baseline and on completion of the 2-hour study period. RESULTS: The initial blood loss (before resuscitation) was 22 mL/kg in both treatment groups (p = 0.97). Animals required 119 +/- 78 mL/kg of fluid in the LR group and 40 +/- 21 mL/kg in the Hextend group (p = 0.01). After resuscitation, the secondary blood loss was 3.7 +/- 1.7 mL/kg in the LR group and 4.7 +/- 1.1 mL/kg in the Hextend group (p = 0.1). Thrombelastography revealed a hypercoagulable state in all animals after injury. This was less pronounced in those animals resuscitated with Hextend. Routine tests of coagulation did not reveal a hypercoagulable state. CONCLUSION: Modulation and restoration of normal coagulation is critical in the management of trauma patients. The patient's coagulation profile might determine the type of fluid to be used at various times during their course. Thrombelastography is superior to routine coagulation assays for the detection of a hypercoagulable state. Resuscitation with Hextend results in a decreased fluid requirement and attenuation of hypercoagulability after injury without increased blood loss.

Resuscitation with lactated ringer's does not increase inflammatory response in a Swine model of uncontrolled hemorrhagic shock.

Lactated Ringer's (LR) and normal saline (NS) are widely and interchangeably used for resuscitation of trauma victims. Studies show LR to be superior to NS in the physiologic response to resuscitation. Recent in vitro studies demonstrate equivalent effects of LR and NS on leukocytes. We aimed to determine whether LR resuscitation would produce an equivalent inflammatory response compared with normal saline (NS) resuscitation in a clinically relevant swine model of uncontrolled hemorrhagic shock. Thirty-two swine were randomized. Control animals (n = 6) were sacrificed following induction of anesthesia for baseline data. Sham animals (n = 6) underwent laparotomy and 2 h of anesthesia. Uncontrolled hemorrhagic shock animals (n = 10/group) underwent laparotomy, grade V liver injury, and blinded resuscitation with LR or NS to maintain baseline blood pressure for 1.5 h before sacrifice. Lung was harvested, and tissue mRNA levels of interleukin-6 (IL-6), granulocyte colony-stimulating factor (G-CSF), and tumor necrosis factor-alpha (TNF-alpha) were determined using quantitative reverse transcriptase polymerase chain reaction (Q-RT-PCR). Sections of lung were processed and examined for neutrophils sequestered within the alveolar walls. Cytokine analysis showed no difference in IL-6 gene transcription in any group (P = 0.99). Resuscitated swine had elevated G-CSF and TNF-alpha gene transcription, but LR and NS groups were not different from each other (P= 0.96 and 0.10, respectively). Both resuscitation groups had significantly more alveolar neutrophils present than controls (P < 0.01) and shams (P < 0.05) but were not different from one another (P= 0.83). LR and NS resuscitation have equivalent effects on indices of inflammation in the lungs in our model of uncontrolled hemorrhagic shock.