- © 2006 by American Society of Clinical Oncology
Docetaxel, Cisplatin, and Trastuzumab As Primary Systemic Therapy for Human Epidermal Growth Factor Receptor 2–Positive Locally Advanced Breast Cancer
- Judith Hurley,
- Philomena Doliny,
- Isildinha Reis,
- Orlando Silva,
- Carmen Gomez-Fernandez,
- Pedro Velez,
- Giovanni Pauletti,
- Jodeen E. Powell,
- Mark D. Pegram and
- Dennis J. Slamon
- From the Department of Surgery; Sylvester Cancer Center, Miller School of Medicine, University of Miami; Taylor Breast Center, Jackson Memorial Hospital, Miami, FL; and University of California at Los Angeles, Los Angeles, CA
- Address reprint requests to Judith Hurley, MD, Sylvester Cancer Center/University of Miami, 1475 NW 12th Ave, D8-4, Miami, FL 33136; e-mail: jhurley{at}miami.edu
Abstract
Purpose To evaluate the efficacy and safety of docetaxel, cisplatin, and trastuzumab as primary systemic therapy for human epidermal growth factor receptor 2 (HER2) –positive, locally advanced breast cancer (LABC).
Patients and Methods Forty-eight patients with immunohistochemistry-confirmed HER2-positive LABC or inflammatory breast cancer received 12 weeks of docetaxel, cisplatin, and trastuzumab with filgrastim, followed by surgery, adjuvant doxorubicin and cyclophosphamide, and locoregional radiotherapy with or without tamoxifen. The primary end point was pathologic complete response (pCR) in breast.
Results Baseline mean tumor size was 9.2 cm (range, 4 to 32 cm). pCR occurred in breast in 11 patients (23%; 95% CI, 12% to 37%) and breast and axilla in eight patients (17%; 95% CI, 8% to 30%). pCR rates in breast (HER2 positive, seven of 30 patients, 23% v HER2 negative, four of 18 patients, 22%; P > .05) and breast and axilla (four of 30 patients, 13% v four of 18 patients, 22%, respectively; P > .05) were similar regardless of HER2 status by fluorescence in situ hybridization (FISH). At a median follow-up time of 43 months, 4-year progression-free survival (PFS) rate was 81% (95% CI, 64% to 90%); overall survival (OS) rate was 86% (95% CI, 71% to 94%). In patients with pCR in breast and axilla, PFS and OS rates were 100% (95% CI, inestimable). In patients without pCR, PFS rate was 76% (95% CI, 57% to 88%; P = .15, log-rank test), and OS rate was 83% (95% CI, 66% to 92%; P = .21). Survival rates were similar regardless of FISH status. There were only two grade 4 adverse events.
Conclusion Twelve weeks of docetaxel, cisplatin, and trastuzumab is clinically active and leads to excellent survival in patients with large, HER2-positive tumors.
INTRODUCTION
Locally advanced breast cancer (LABC) is the most common presentation for breast cancer outside the United States, with incidences ranging from 33% in Peru1 to 73% in Nigeria.2 Only 6% of women present with LABC in the United States,3 but 23% of women present with LABC at Jackson Memorial Hospital in Miami (10-year, tumor registrar statistics).
Primary systemic therapy (PST) has become the standard of care for LABC, but the optimal regimen is unknown.4 Regimens are often derived from metastatic breast cancer regimens and are usually anthracycline based, with recent regimens including taxanes. Pathologic complete response (pCR) rates in the breast range from 6% to 26% in randomized studies of women with small tumors.5-11 Little has been published regarding large tumors, although the National Surgical Adjuvant Breast and Bowel Project reported a 4% pCR rate in tumors more than 5 cm with doxorubicin and cyclophosphamide.
We began studying platinum-based regimens in 1990 at the University of Miami. Our first regimen, methotrexate, vinblastine, doxorubicin, and cisplatin, was highly active, with a pCR rate in breast and axilla of 27%.12 We then added docetaxel to cisplatin because of in vitro synergism13-15 and clinical activity against metastatic breast cancer16,17 and reported pCR rates for LABC of 26% in breast and 20% in breast and axilla.18
We hypothesized that adding trastuzumab to platinum-docetaxel PST would be active in women with human epidermal growth factor receptor 2 (HER2) –positive LABC based on its activity, both as a single agent19 and combined with other agents,20 against advanced breast cancer. The greatest synergy in vitro occurs when trastuzumab is added to docetaxel and cisplatin21,22; this combination is active against advanced breast cancer.22 Furthermore, adding a platinum agent to a taxane and trastuzumab improved the response rate and prolonged time to progression in a randomized study of women with advanced breast cancer.23 We conducted a phase II study of docetaxel, cisplatin, and trastuzumab as PST in patients with HER2-positive LABC and inflammatory breast cancer. The primary end point was pCR in breast; secondary end points were pCR in breast and axilla, progression-free survival (PFS) and overall survival (OS), and safety.
PATIENTS AND METHODS
Eligibility
Study patients had untreated HER2-positive LABC (T3, T4, N2, or N3)24 or inflammatory breast cancer (T4d) diagnosed by needle biopsy. Eligibility criteria were bidimensionally measurable disease without metastases. Metastatic assessment included hepatic function enzymes, chest x-ray, and bone scan. If normal, no further work-up was performed. Computed tomography and other extensive studies were not routinely performed.
Tumors had to be HER2 positive by immunohistochemistry (IHC; HerceptTest, DAKO 2+ or 3+; DAKO, Copenhagen, Denmark) or by fluorescence in situ hybridization (FISH). Adequate hematologic (absolute neutrophil count > 1.5 × 109/L, platelet count > 100 × 109/L, and hemoglobin > 10 g/dL), renal (blood urea nitrogen < 25 mg/dL and serum creatinine < 1.5 mg/dL), and hepatic function (serum bilirubin < 1 mg/dL and transaminases and alkaline phosphatase < 1.5× upper limit of normal); normal ejection fraction (> 49% by multigated acquisition [MUGA] scan) or normal echocardiogram; and negative pregnancy test were required. Fertile women had to use contraceptives. Exclusion criteria were previous malignancy, active infection, grade 2 of greater peripheral neuropathy, lactation, or other comorbid conditions that would affect drug tolerance or limit compliance.
The University of Miami’s Institutional Review Board approved the protocol. All patients provided written informed consent.
Treatment
PST comprised four cycles over 12 weeks. Trastuzumab 4 mg/kg was administered intravenously over 90 minutes on day 1, and then 2 mg/kg was administered over 30 minutes weekly, for a total of 12 doses. Docetaxel 70 mg/m2 was administered intravenously over 30 minutes, followed by cisplatin 70 mg/m2 intravenously over 2 hours, on days 1, 22, 43, and 64. Dose adjustments were planned for severe toxicities. Premedication was dexamethasone 8 mg every 12 hours for three doses starting 12 hours before docetaxel and a serotonin 5-hydroxytryptamine-3 antagonist. Filgrastim 5 μg/kg was administered daily starting on days 2, 23, 44, and 65 until the WBC count was more than 10 × 109/L. When hematocrit decreased to less than 35%, epoetin alfa 40,000 U was administered weekly for up to 12 doses.
Breast surgery with level I or II axillary dissection was performed 2 to 6 weeks after the last chemotherapy dose. Surgery type was at the surgeon’s discretion. Conservation was allowed.
Adjuvant chemotherapy began 3 to 6 weeks postoperatively. Cyclophosphamide 600 mg/m2 and doxorubicin 60 mg/m2 were administered intravenously every 21 days for four cycles. Filgrastim and epoetin alfa were administered the same as for PST. Within 4 weeks after completion, patients received standard locoregional radiotherapy to the chest or ipsilateral breast. For hormone receptor–positive tumors, tamoxifen 20 mg orally daily was initiated with radiotherapy and continued for 5 years. For hormone receptor–positive tumors and age less than 40 years, goserelin acetate 3.6 mg/m2 was added subcutaneously every 28 days, for a total of 24 doses.
Assessment
FISH was performed retrospectively on all tumors at the University of California, Los Angeles. HER2/neu gene amplification was analyzed using FISH on formalin-fixed, paraffin-embedded tissue.25,26 A Vysis SpectrumOrange-labeled HER2/neu probe (Abbott Molecular, Downers Grove, IL) was used to determine HER2 gene copy number. To account for increased HER2 gene copy number as a result of chromosome 17 polysomy, specimens were cohybridized with a chromosome 17 centromeric (chr 17cen) probe labeled with Vysis SpectrumGreen fluorophore (Abbott Molecular). Data were expressed as HER2 genes per chr 17cen, with two HER2 genes per chr 17cen as the cutoff for HER2 amplification.
Clinical response was assessed by physical examination on the final day of PST and defined as complete response, partial response, and progressive disease (PD) according to Response Evaluation Criteria in Solid Tumors.27 Pathologic response was assessed postoperatively. pCR in breast, which was defined as disappearance of invasive disease in breast by pathologic examination, included specimens containing only residual ductal carcinoma-in-situ; pCR in axilla was absence of positive lymph nodes by hematoxylin and eosin staining.
PFS was defined as time from starting study drug to PD or cancer-related death. OS was defined as time from starting PST to death from any cause. In the absence of death or PD, follow-up was censored at last contact.
Adverse events were assessed weekly during PST and graded by National Cancer Institute Common Toxicity Criteria (version 2). cardiotoxicity was monitored by MUGA scans at baseline, postoperatively, and after adjuvant radiotherapy.
Statistics
The two-stage study was designed to detect a pCR rate of 25% v 10% in breast. Significance was set at 5%, and power was set at 80%. Eighteen patients were enrolled onto the first stage and, if three or more pCRs occurred in breast, 30 patients were enrolled onto the second stage.28
Baseline patient characteristics and adverse events were presented as percentages or mean values with standard deviations, medians, and ranges. Response rates with 95% CIs were calculated. Survival was estimated using the Kaplan-Meier product-limit method.29 The log-rank test was used to determine whether survival rates corresponded to pCR in breast, pCR in breast and axilla, disease stage, or HER2 status.
RESULTS
Patients
Forty-eight patients were enrolled between February 2000 and April 2003 and observed for a median of 43 months (range, 11 to 63 months). Forty-five patients completed PST without dose reduction. Cisplatin was changed to carboplatin after three doses in one patient with tinnitus. Trastuzumab was stopped after one dose in one patient with bradycardia; she completed PST with docetaxel and cisplatin. The third patient developed catheter-related Staphylococcus aureus sepsis after two courses; therapy was halted, and surgery was cancelled.
Patients were generally young (mean age, 51 years), white (83%), and Hispanic (69%; Table 1). Twenty-one patients (44%) were premenopausal.
Breast tumors were large (mean size, 9.2 cm; Table 2) and HER2 positive by IHC. Thirty patients (62%) were FISH positive. In 34 patients with adequate preoperative and postoperative specimens, FISH status became negative after PST in 10 (43%) of 23 women whose tumors were FISH positive at baseline.
Response Rates
In the intent-to-treat evaluation of clinical response, there were 22 complete responses (46%; 95% CI, 31% to 61%) and 26 partial responses (54%; 95% CI, 39% to 69%), for a total of 48 objective responses (100%; 95% CI, 94% to 100%).
The intent-to-treat evaluation of pCR included the patient whose surgery was cancelled (ie, no pCR). At surgery, pCR in breast was found in 11 patients (23%; 95% CI, 12% to 37%; Table 3). Seven patients (15%) had only microscopic residual disease, 19 patients (40%) had residual tumors ≤ 2 cm, and 10 patients (21%) had residual tumors more than 2 cm. Eight pCRs occurred in both breast and axilla (17%; 95% CI, 8% to 30%). pCR rates in breast (FISH positive, seven of 30 patients, 23% v FISH negative, four of 18 patients, 22%) and breast and axilla (FISH positive, four of 30 patients, 13% v FISH negative, four of 18 patients, 22%) were similar regardless of FISH status.
At surgery, 22 patients (46%; 95% CI, 32% to 61%) were lymph node negative. Node-negative rates were similar for FISH-positive tumors (15 of 30 patients, 50%) and FISH-negative tumors (seven of 18 patients, 39%). Estrogen receptor (ER) –negative tumors were more likely than ER-positive tumors to be node negative in all patients (64% v 31%, respectively; P = .04) and in the FISH-positive subset (67% v 25%, respectively; P = .06). Otherwise, there were no differences between pCR rates by patient or disease characteristics including FISH and IHC status.
PFS and OS
Eight patients experienced relapses, including one local relapse at 8 months, four distant relapses at 8 to 40 months, and three local and distant relapses at 13 to 18 months. Relapsed tumors were generally aggressive and refractory to therapy; five involved the CNS. Survival time after distant relapse ranged from only 2.5 to 10 months. One patient was alive with distant disease at 6 months. The patient with local relapse only was in remission 20 months later.
Seven deaths occurred a median of 21 months after initiating chemotherapy (range, 11 to 50 months). Six were attributable to FISH-positive (n = 4) or FISH-negative (n = 2) breast cancer. The remaining cause of death is unknown; this 65-year-old patient did not undergo surgery, was relapse free 32 months after initiating PST, had diabetes and hypertension, and died suddenly at an outlying hospital.
The 4-year PFS rate was 81% (95% CI, 64% to 90%), and the OS rate was 86% (95% CI, 71% to 94%). In eight patients with pCR in breast and axilla, PFS and OS rates were 100% (95% CI, inestimable;Fig 1). In 40 patients without pCR in breast and axilla, PFS rate was 76% (95% CI, 57% to 88%), and OS rate was 83% (95% CI, 66% to 92%). Although these data suggest pCR in breast and axilla decreased risk of progression and death, differences were insignificant (P = .15). Similar trends were seen for pCR in breast alone (P = .36) and axilla (P = .14).
FISH status did not affect PFS rate (FISH negative, 87% v FISH positive, 77%; log-rank test, P = .34) or OS rate (FISH negative, 87% v FISH positive, 85%; log-rank test, P = .85; Fig 2). Survival rates were similarly unaffected by ER status (PFS rate: ER positive, 83% v ER negative, 76%; P = .30; OS rate: ER positive, 91% v ER negative, 80%; P = .50), disease stage (PFS rate: IIB and IIIA, 86% v IIIB, 71%; P = .21; OS rate: IIB and IIIA, 90% v IIIB, 79%; P = .47), and menopausal status (PFS rate: premenopausal, 87% v postmenopausal, 75%; P = .54; OS rate: premenopausal, 91% v postmenopausal, 82%; P = .27).
Safety
Only two grade 4 adverse events occurred (infection and leukopenia; Table 4). Twenty-one grade 3 adverse events occurred; the most common was corticosteroid-induced hyperglycemia (23%).
Seven patients were hospitalized during therapy. Four had catheter-related infections, but only one was neutropenic. One patient had grade 4 S aureus septicemia with multiple septic emboli; PST was stopped. The other three patients had grade 3 infections treated with antibiotics and catheter removal. Subsequent patients received prophylactic oral cephalexin, avoiding further catheter-related infections. Two chemotherapy hospitalizations were extended overnight to treat dental abscess (n = 1) or biliary colic (n = 1). The seventh patient was hospitalized for overnight observation for filgrastim-induced bone pain.
Of 47 patients who had post-PST MUGA scans, cardiotoxicity was grade 1 in 10 patients (21%), grade 2 in 10 patients (21%), and grade 3 in one patient (2%). The woman with grade 3 cardiotoxicity had congestive heart failure on preoperative chest x-ray and acute diastolic dysfunction on echocardiogram. She was asymptomatic and was treated with oral diuretics. Chest x-ray normalized, and ejection fraction decreased from 75% to 61%; she refused follow-up cardiac testing. One patient experienced asymptomatic sinus bradycardia (40 beats/min) within 1 hour after beginning trastuzumab. She completed PST with docetaxel and cisplatin; bradycardia resolved over 4 weeks. cardiotoxicity was unrelated to age, race, ethnicity, diabetes, or hypertension.
Eighteen patients refused postadjuvant chemotherapy MUGA scans because they were asymptomatic. Of the remaining 30 patients, 14 patients’ ejection fractions remained normal, whereas 10 patients (33%) had grade 1 cardiotoxicity, and six patients (20%) had grade 2 cardiotoxicity.
DISCUSSION
pCR has become the preferred end point for assessing PST since being shown to correlate with survival12,30,31; however, its definition is controversial. When our study was conceptualized in 1999, we chose pCR in breast for the stopping rule because the National Surgical Adjuvant Breast and Bowel Project used it.30 We have since shifted to pCR in breast and axilla because it may be a better predictor.31 We changed our eligibility criteria after trastuzumab’s benefit was shown to be limited to IHC 3+ or FISH-positive tumors32; however, 13 patients with IHC 2+ tumors had already been enrolled.
Docetaxel, cisplatin, and trastuzumab seemed active as PST in our patients with HER2-positive LABC, with pCR rates of 23% in breast and 17% in breast and axilla. The 4-year PFS rate of 81% and OS rate of 86% are notable because our staging work-up was limited and some patients may have had metastases at baseline. The survival rate was higher in patients with pCR in breast and axilla (100%) compared with patients without pCR in breast and axilla (83%; P = .21); but the patients with residual tumor after PST also had good survival rates. Importantly, our patients had high PFS and OS rates regardless of FISH status. In contrast, survival curves diverged early after adjuvant chemotherapy for HER2-positive, node-positive tumors not treated with trastuzumab.33,34 These subset analyses are exploratory because the study was not powered to detect between-subgroup differences.
Our current pCR rates approximate those from our previous study18 in which 57 women with LABC received identical doses of docetaxel and cisplatin. pCR rates were 26% in breast and 20% in breast and axilla. Our previous pCR rates were lower for patients with HER2-positive (IHC 2+ or 3+) versus HER2-negative tumors (19% v 28%, respectively). Unlike the current study, the 4-year disease-free survival rate was only 65%, and the OS rate was 73%,18 providing further evidence that adding trastuzumab to docetaxel and cisplatin may have benefited women with HER2-positive tumors.
Our pCR rates generally approximate those from other studies of trastuzumab as PST, except for the randomized study by Buzdar et al.35 Adding trastuzumab to sequential PST with paclitaxel followed by fluorouracil, epirubicin, and cyclophosphamide in 42 women with HER2-positive, operable cancer improved the pCR rate in breast and axilla from 26% to 65% (P = .02).35 Small initial tumor size, sequential PST, and prolonged, 24-week trastuzumab exposure may explain these remarkable results. In a study similar to ours, Burstein et al36 evaluated PST with paclitaxel and trastuzumab followed by adjuvant doxorubicin and cyclophosphamide in 40 women with HER2-positive LABC. Paclitaxel and trastuzumab produced a pCR rate of 18% in breast and axilla and a 2-year disease-free survival rate of 83%. Harris et al37 reported that vinorelbine and trastuzumab produced a pCR rate of 19% in 42 women. Docetaxel and trastuzumab produced pCR rates of 11% in 27 patients when the taxane was administered weekly38 and 25% in 16 patients when it was administered every 3 weeks.39 It is unclear whether response rates in the last three studies37-39 reflect pCR in breast or in breast and axilla. Finally, Van Pelt et al40 evaluated PST with 3 weeks of trastuzumab, followed by trastuzumab and docetaxel. The patients were a mixed group with LABC and metastatic breast cancer. pCR rates were not reported; complete response rate was 41% of 22 patients.40 The last group41 recently reported partial response after only 3 weeks of trastuzumab in 23% of 35 women with LABC. Only Burstein et al36 reported survival data.
Although studies having different end points and populations should not be compared, we emphasize our conservative definition of pCR and high risk of relapse from large tumors (mean size, 9 cm). Others rarely indicated tumor size, except for the groups from Brazil (median diameter, 8 cm)38 and from Houston and New Haven (median volume, 10 × 10 cm2).41 Thirty-one percent of our patients had T4 lesions, including 13% with inflammatory carcinoma. Similarly, 15% of the patients in the study by Burstein et al36 had inflammatory disease; however, fewer had N2 or N3 disease (5% of his patients v 33% of ours). Only 4% of the patients in the study by Buzdar et al35 had N2 disease, and 4% had T4 lesions.
PST provides an opportunity to study biologic mechanisms because specimens are available before and after drug exposure. An unexpected finding in our study was that 43% of FISH-positive tumors became FISH negative after PST, which is unlikely to be an artifact of sampling or testing. High copy number chromosomal polysomy occurred in eight of 48 patients, including six of 10 patients with tumors that became FISH negative. This high copy number polysomy rate, presumably a result of large tumor size and ensuing genomic instability, might have affected interpretation of FISH values near the cutoff level. Alternatively, the change could reflect the intrinsic heterogeneity of HER2 expression and trastuzumab destruction of FISH-positive cells. Regardless of the mechanism, a 12-week course of trastuzumab plus chemotherapy seemed sufficient to obliterate the FISH-positive clone in many tumors.
Safety is important because the goal of PST is long-term survival. With routine filgrastim and frequent epoetin alfa support, hematologic adverse events were usually grade 1 or 2. In contrast, Buzdar et al35 reported neutropenic fever in 35% of 23 patients who received paclitaxel; fluorouracil, epirubicin, and cyclophosphamide; and trastuzumab without routine growth factor support. Similarly, Van Pelt et al40 reported grade 3 or 4 hematologic toxicity in 32% of 22 patients; filgrastim was not used until neutropenic sepsis caused two deaths (9%). We monitored cardiotoxicity with serial MUGA scans, but compliance was poor. Left ventricular ejection fraction decreased in 45% of patients after PST and in 53% of patients after adjuvant therapy. Burstein et al36 also reported asymptomatic cardiotoxicity after both PST with trastuzumab and adjuvant chemotherapy.
In conclusion, the combination of docetaxel, cisplatin, and trastuzumab seems active and feasible as PST for HER2-positive LABC. This regimen produced a respectable pCR rate, using a conservative definition, and excellent survival, considering the prognosis of HER2-positive large tumors. Growth factors prevented most hematologic toxicities. cardiotoxicity was common but asymptomatic. Although more studies are needed to understand trastuzumab’s cellular and molecular effects and the chemotherapy-trastuzumab interaction with HER2 amplification, our findings suggest it is reasonable to add trastuzumab to taxane-containing PST for HER2-positive breast cancer.
Authors’ Disclosures of Potential Conflicts of Interest
Although all authors completed the disclosure declaration, the following authors or thier immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information about ASCO’s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
| Authors | Employment | Leadership | Consultant | Stock | Honoraria | Research Funds | Testimony | Other |
|---|---|---|---|---|---|---|---|---|
| Judith Hurley | Aventis (A) | Aventis (B) | ||||||
| Giovanni Pauletti | Vysis Inc for FDA hearing (N/R) | |||||||
| Mark D. Pegram | Genentech (B); Aventis (A) | Genentech (B); Aventis (B) | ||||||
| Dennis J. Slamon | Genentech (B); Aventis (B) | Aventis (B); Genentech (B) |
Dollar Amount Codes (A) < $10,000 (B) $10,000–99,999 (C) ≥ $100,000 (N/R) Not Required
Author Contributions
Conception and design: Judith Hurley
Administrative support: Judith Hurley
Provision of study materials or patients: Philomena Doliny, Orlando Silva, Pedro Velez, Giovanni Pauletti, Jodeen E. Powell, Dennis J. Slamon
Collection and assembly of data: Judith Hurley, Philomena Doliny, Carmen Gomez-Fernandez, Giovanni Pauletti
Data analysis and interpretation: Judith Hurley, Isildinha Reis
Manuscript writing: Judith Hurley, Isildinha Reis, Mark D. Pegram
Progression-free survival and overall survival by pathologic complete response (pCR) in breast and axilla.
Progression-free survival and overall survival by fluorescence in situ hybridization (FISH).
Baseline Characteristics in 48 patients
Baseline Disease Characteristics in 48 Patients
pCR by FISH, ER, and menopause Status in 48 Patients
Adverse Events in 48 Patients
Acknowledgments
We thank Cindy W. Hamilton for assistance with manuscript preparation.
Footnotes
-
Supported by grants from Aventis Oncology and Amgen.
Presented in part at the 24th Annual San Antonio Breast Cancer Symposium, San Antonio, TX, December 10-13, 2001; 26th Annual San Antonio Breast Cancer Symposium, San Antonio, TX, December 3-6, 2003; 37th Annual Meeting of the American Society of Clinical Oncology, San Francisco, CA, May 12-15, 2001; 38th Annual Meeting of the American Society of Clinical Oncology, Orlando, FL, May 18-21, 2002; and 39th Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, May 31-June 3, 2003.
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
- Received May 31, 2005.
- Accepted February 3, 2006.
