Stem cell-associated genes are extremely poor prognostic factors for soft-tissue sarcoma patients.

Cancer stem cells can play an important role in tumorigenesis and tumor progression. However, it is still difficult to detect and isolate cancer stem cells. An alternative approach is to analyse stem cell-associated gene expression. We investigated the coexpression of three stem cell-associated genes, Hiwi, hTERT and survivin, by quantitative real-time-PCR in 104 primary soft-tissue sarcomas (STS). Multivariate Cox's proportional hazards regression analyses allowed correlating gene expression with overall survival for STS patients. Coexpression of all three stem cell-associated genes resulted in a significantly increased risk of tumor-related death. Importantly, tumors of patients with the poorest prognosis were of all four tumor stages, suggesting that their risk is based upon coexpression of stem cell-associated genes rather than on tumor stage.

Ceramide-induced G2 arrest in rhabdomyosarcoma (RMS) cells requires p21Cip1/Waf1 induction and is prevented by MDM2 overexpression.

The sphingoplipid ceramide is responsible for a diverse range of biochemical and cellular responses including a putative role in modulating cell cycle progression. Herein, we describe that an accumulation of ceramide, achieved through the exogenous application of C(6)-ceramide or exposure to sphingomyelinase, induces a G(2) arrest in Rhabdomyosarcoma (RMS) cell lines. Utilizing the RMS cell line RD, we show that this G(2) arrest required the rapid induction of p21(Cip1/Waf1) independent of DNA damage. This was followed at later time points (48 h) by the commitment to apoptosis. Apoptosis was prevented by Bcl-2 overexpression, but permitted the maintenance of elevated p21(Cip1/Waf1) protein expression and the stabilization of the G(2) arrest response. Inhibition of p21(Cip1/Waf1) protein synthesis with cyclohexamide (CHX) or silencing of p21(Cip1/Waf1) with siRNA, prevented ceramide-mediated G(2) arrest and the late induction of apoptosis. Further, adopting the recent discovery that murine double minute 2 (MDM2) controls p21(Cip1/Waf1) expression by presenting this CDK inhibitor to the proteasome for degradation, RD cells overexpressing MDM2 abrogated ceramide-mediated p21(Cip1/Waf1) induction, G(2) arrest and the late ensuing apoptosis. Collectively, these data further support the notion that ceramide accumulation can modulate cell cycle progression. Additionally, these observations highlight MDM2 expression and proteasomal activity as key determinants of the cellular response to ceramide accumulation.

A retrospective review of an epidural blood patch database: the incidence of epidural blood patch associated with obstetric neuraxial anesthetic techniques and the effect of blood volume on efficacy.

BACKGROUND: The optimal volume of blood required to treat post-dural puncture headache remains in question. In our institution a target volume of 30mL is used for an epidural blood patch unless the patient experiences pain during injection. METHODS: The institutional database was retrospectively reviewed for epidural blood patch and delivery statistics over a 15-year period to determine if the volume of blood administered during the procedure directly correlated with the number of epidural blood patches administered. The primary endpoint was defined as the need for a repeat epidural blood patch. RESULTS: There were 466 epidural blood patches performed on 394 patients, associated with 84 804 obstetric neuraxial procedures. Thirty-two percent (95% CI 28.3 to 34.9%) of patients who had an inadvertent dural puncture with an epidural needle received an epidural blood patch versus 0.19% (0.16% to 0.22%) of patients who received neuraxial anesthesia with no documented dural puncture with an epidural needle. All patients experienced relief of post-dural puncture headache, although 17% required two and 1.5% required three epidural blood patches. The mean±SD volume of blood administered was 20.5±5.4mL and only 35 patients (8.9%) received 30mL. CONCLUSION: Increasing blood volumes up to 30mL did not reduce the need for repeat epidural blood patch. Although the optimal volume of blood to administer during epidural blood patch placement remains unknown, our institution will continue to administer up to 30mL or until the patient experiences pain during epidural injection.

MDM2 splice variants and their therapeutic implications.

MDM2 splice variants have now been identified in many different tumor types, and their expression has been associated with advanced disease. However, published data concerning their function is contradictory, and therefore their role in tumorigenesis and their potential as a therapeutic target are unclear. Expression of a specific splice variant, MDM2-B, in a transgenic mouse model results in tumor development; and expression of several splice variants has been shown to enhance tumor formation in Emu-myc transgenic mice. However, expression of similar variants in vitro results in growth inhibition, an observation inconsistent with a transformed phenotype. The observed growth inhibition is p53-dependent, resulting from the binding of splice variants with an intact C-terminal RING finger domain to full-length MDM2 protein. In doing so, p53 can no longer bind MDM2, and p53 activity is elevated. Subsequent inactivation of p53 or p53-mediated apoptosis could contribute to the MDM2 splice variant-mediated tumorigenesis observed in vivo. However, MDM2 splice variants, like full-length MDM2, probably display p53-independent activities. Therefore, the potential for MDM2 splice variants as therapeutic targets will be dependent upon their phenotype within specific tumor types.

A comparison of human O6-methylguanine-DNA methyltransferase promoter activity in Mer+ and Mer- cells.

The activity of the human O6-methylguanine-DNA methyltransferase (MGMT) gene promoter was determined in eight human cell lines by measuring chloramphenicol acetyltransferase activity in a reporter gene system. MGMT promoter activities in cells that do not express MGMT (Mer-) fell within the range of activities seen in cells that do express MGMT (Mer+). The promoter region contains 11 potential binding sites for the transcription factor Sp1, but no correlation was seen between cellular Sp1 protein and MGMT promoter chloramphenicol acetyltransferase activity. Because Mer- cells are not deficient in the factors needed for transcription of MGMT, we suggest that at least two mechanisms regulate MGMT expression. One suppresses MGMT mRNA and protein in Mer- cells, and another regulates the levels of constitutive expression in Mer+ cells. Sp1 is not a limiting factor in MGMT expression.

Wild-type p53 suppresses transcription of the human O6-methylguanine-DNA methyltransferase gene.

High-level expression of the DNA repair protein O6-methylguanine-DNA methyltransferase (MGMT) correlates with cellular resistance to the chloroethylnitrosourea (CENU) class of alkylating agents. Consequently, tumors expressing low levels of MGMT are sensitive to CENU chemotherapy, and any mechanism that can be used to reduce MGMT levels could sensitize resistant tumors. We have demonstrated that transient transfection of wild-type, but not mutant, p53 protein into a p53-null cell line, Saos-2, suppresses MGMT promoter activity in a reporter gene system. In addition, following a 24-h transduction of IMR90 fibroblasts with a wild-type p53-adenoviral vector, endogenous MGMT protein is down-regulated and is no longer detectable 5 days following infection. Because p53 is inducible by ionizing radiation, we propose that existing cancer therapy regimens that combine radiotherapy with CENU chemotherapy may be improved by altering scheduling and allowing enough time between the two therapies for the relatively stable MGMT protein to degrade.

Wild-type p53 induction mediated by replication-deficient adenoviral vectors.

Replication-deficient E1-/E3-deleted adenoviral vectors are commonly used to introduce transgenes into cells in vitro and have been used for certain kinds of gene therapy protocols in vivo. We have demonstrated that transduction of cells using these vectors can induce p53 expression in cells containing a wild-type p53 gene. This response is different from p53 induction observed after DNA damage because the time course of induction is slower and because it occurs in cells with an attenuated DNA damage response. However, this vector-induced p53 is transcriptionally active and, therefore, p53 function is not inactivated by viral proteins. The mechanism of induction appears to be an increased rate of protein translation because immunoprecipitation analyses demonstrated increased levels of 35S-labeled p53 protein, even after a short 15-min labeling time. Induction of p53 by adenoviral vectors may have various effects on transduced cells, including apoptosis and altered chemotherapy chemosensitivity. Therefore, the influence of the vector might confound the impact of any particular gene used in a gene therapy application.

Sensitization of human tumor cells to CPT-11 via adenoviral-mediated delivery of a rabbit liver carboxylesterase.

Irinotecan, 7-ethyl-10-[4-(1-piperidino)-1-piperidino]carbonyloxycamptothecin (CPT-11) is activated by carboxylesterases (CE) to yield the potent topoisomerase I inhibitor, SN-38. We have demonstrated previously that a rabbit liver CE is approximately 100-1000-fold more efficient at drug activation than a highly homologous human CE. In an attempt to use rabbit CE expression in combination with CPT-11 for gene therapy approaches for the treatment of cancer, we have developed an adenoviral vector expressing this intracellular CE. After transduction, this virus produces very high levels of CE activity in a panel of human tumor cell lines and results in marked sensitization to CPT-11 of all of the transduced cells. Reductions in IC(50) values for this drug ranged from 11-127-fold. Additionally, comparison with an adenovirus expressing a secreted form of the rabbit CE indicated that a collateral effect could be achieved with reductions in the IC(50) values ranging from 4-19-fold. These data suggest that the described reagents may be suitable for use in vivo in a viral-directed enzyme prodrug therapy approach using CPT-11.

Ribozyme-mediated modulation of human O6-methylguanine-DNA methyltransferase expression.

A synthetic oligonucleotide containing ribozyme sequences targeted to the 5' region of the human O6-methylguanine-DNA methyltransferase (MGMT) mRNA has been constructed. This ribozyme demonstrates cleavage activity in vitro in the presence of Mg2+. To determine whether this ribozyme can function in vivo, HeLa CCL2 cells were transfected with a mammalian expression vector containing the ribozyme sequence. Following selection and expansion of individual transfectants, a stable clone was isolated that lacks both MGMT mRNA and protein. Molecular analysis of this cell line indicates that in vivo cleavage of MGMT mRNA is responsible for the lack of MGMT activity.

Use of a modified ornithine decarboxylase promoter to achieve efficient c-MYC- or N-MYC-regulated protein expression.

One of the advantages of viral-directed enzyme prodrug therapy (VDEPT) is its potential for tumor-specific cytotoxicity. However, the viruses used to deliver cDNAs encoding prodrug-activating enzymes transduce normal cells as well as tumor cells, and several approaches to achieve tumor-specific expression of the delivered cDNAs are being investigated. One such approach is to regulate transcription of the prodrug-activating enzyme with a promoter that is preferentially activated by tumor cells. Published data suggest that the most promising transcription factor/promoter/enhancer combinations are those activated by a tumor-specific transcription factor to retain tumor cell specificity but that are equal in strength to nonspecific viral promoters in their ability to up-regulate target cDNAs. This report identifies MYC-responsive, modified ornithine decarboxylase (ODC) promoter/enhancer sequences that up-regulate target protein expression in tumor cells overexpressing either N-MYC or c-MYC protein. The most efficient of the four constructs assessed contained six additional CACGTG MYC binding sites 5' to the endogenous ODC promoter (R6ODC). Reporter assays with this chimeric promoter/enhancer regulating expression of chloramphenicol acetyltransferase demonstrated 50-250-fold more activity in MYC-expressing cells compared with similar assays with promoterless plasmids. The R6ODC regulatory sequence was approximately equivalent to the CMV promoter in inducing expression of the neomycin resistance gene in c-MYC-expressing SW480 and HT-29 colon carcinoma cells and in N-MYC-expressing NB-1691 neuroblastoma cells. The modified ODC promoter may, therefore, be useful in achieving tissue-specific expression of target proteins in tumor cells that overexpress c- or N-MYC.

A virus-directed enzyme prodrug therapy approach to purging neuroblastoma cells from hematopoietic cells using adenovirus encoding rabbit carboxylesterase and CPT-11.

Tumor cells that contaminate hematopoietic cell preparations contribute to the relapse of neuroblastoma patients who receive autologous stem cell rescue as a component of therapy. Therefore, effective purging methods are needed. This study details in vitro experiments to develop a viral-directed enzyme prodrug purging method that specifically targets neuroblastoma cells. The approach uses an adenovirus to deliver the cDNA encoding a rabbit liver carboxylesterase that efficiently activates the prodrug irinotecan,7-ethyl-10-[4-(1-piperidino)-1-piperidino]carbonyloxycamptothecin (CPT-11). The data show that an adenoviral multiplicity of infection of 50 transduces 100% of cultured neuroblastoma cells and primary tumor cells, irrespective of the level of tumor cell line contamination. Exposure of neuroblastoma cell lines or of mixtures of these cell lines with CD34(+) cells at a ratio of 10:90 to replication-deficient AdRSVrCE for 24 h and subsequent exposure of cells to 1-5 microM CPT-11 for 4 h increased the toxicity of CPT-11 to three neuroblastoma cell lines (SJNB-1, NB-1691, and SK-N-SH) from approximately 20-50-fold and eradicated their clonogenic potential. Also, after "purging," RNA for neuroblastoma cell markers (tyrosine hydroxylase, synaptophysin, and N-MYC) was undetectable by reverse transcription-PCR. In contrast, the purging protocol did not affect the number or type of colonies formed by CD34(+) cells in an in vitro progenitor cell assay. No bystander effect on CD34(+) cells was observed. The method described is being investigated for its potential clinical utility, particularly its efficacy for use with patients having relatively high tumor burdens, because no published methods have been shown to be efficacious when the tumor burden exceeds 1%.

In vitro methylation of the human O6-methylguanine-DNA methyltransferase promoter reduces transcription.

Approx. 20% of human tumor cell lines (termed Mer-) are deficient in the DNA repair enzyme O6-methylguanine-DNA methyltransferase (MGMT; E.C.2.1.1.63). Such cells possess the MGMT gene and promoter sequences but have virtually no mRNA or protein. Cytosine methylation of gene sequences has been proposed as a mechanism by which MGMT could be suppressed in Mer- cells; however, the experimental evidence does not uniformly support this idea. We therefore investigated the effect of in vitro methylation of the MGMT promoter in a reporter gene construct transfected into cultured human cells. DNA methylation by HpaII or HhaI methylases suppressed the activity of the promoter, although the effect was not absolute. The occurrence of partial intracellular demethylation of promoter sequences may account for the incomplete inhibition of transcription. A model that attempts to reconcile the opposing views on the role of cytosine methylation in MGMT gene expression is presented.

Differences in epitope accessibility of p53 monoclonal antibodies suggest at least three conformations or states of protein binding of p53 protein in human tumor cell lines.

The p53 tumor suppressor gene is deleted or mutated in over 50% of human tumors. Mutations frequently extend the half-life of the p53 protein; and a high level of nuclear p53 expression, detected by immunohistochemistry, has been used to predict the p53 status of tumors. We compared the sensitivity and reactivity of five frequently used, commercially available monoclonal antibodies (1801, DO1, DO7, BP53.12 and 421) in immunoblot and immunofluorescence assays, and found that results differed among the antibodies. Comparison of immunoblot analysis of denatured nuclear and cytoplasmic p53 protein were consistent with antibodies DO1, DO7 and BP53.12, each of which generated a strong specific signal in both cell fractions. However, in situ analysis demonstrated that although all antibodies recognized nuclear p53, only BP53.12 and 421 recognized p53 protein in the cytoplasm. In addition, 1801 produced a signal in p53-negative tumor cell lines. Differences in situ among the antibodies were probably due to the accessibility of their respective epitopes and suggested that nuclear and cytoplasmic p53 either have different three-dimensional conformations or are bound to different proteins. A third p53 protein conformation was also suggested by the observation that only two of the five antibodies (BP53.12 and DO7) detected induced levels of p53 in situ following exposure to ionizing radiation. In summary, except for the fact that DO7 does not recognize cytoplasmic p53 in situ, we found it to be the most specific, versatile, and reliable antibody. We conclude that the p53 antibody of choice depends upon the specific goal of a study and the method used to detect this protein.

Retroviral transfer of a bacterial alkyltransferase gene (ada) into human bone marrow cells protects against O6-benzylguanine plus 1, 3-bis(2-chloroethyl)-1-nitrosourea cytotoxicity.

The antitumor activity of 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) is limited by the O6-alkylguanine-DNA alkyltransferase (ATase) in tumor cells and by delayed myelosuppression. Inactivation of neoplastic ATase by O6-benzylguanine (BG) improves the therapeutic index for BCNU. We have demonstrated previously that BG + BCNU-induced myelosuppression in mice is reduced by expression of the BG-resistant ATase ada in murine bone marrow. We have now generated an amphotropic retrovirus containing the ada gene and tested the effectiveness of ada expression in preventing BG + BCNU cytotoxicity in human hematopoietic progenitor cells. A retroviral producer clone with a biological titer of 6.5 x 10(4) colony-forming units/ml and 4.4 pmol ATase/mg protein was used for transduction of bone marrow. Cocultivation of these ada producer cells with progenitor cells from six normal individuals resulted in 1.9-3. 9-fold protection against BG + BCNU-induced cytotoxicity in committed progenitor cell assays. Furthermore, this cytoprotective effect was associated with a high transduction efficiency (40%) and a 2-fold increase of ATase activity in the surviving committed progenitor cell colonies. These data provide a basis for testing the clinical effectiveness of retroviral ada gene transfer into hematopoietic cells to increase the therapeutic index of BG + BCNU.

Retroviral transfer of a bacterial alkyltransferase gene into murine bone marrow protects against chloroethylnitrosourea cytotoxicity.

The chloroethylnitrosoureas (CENUs) are important antineoplastic drugs for which clinical utility has been restricted by the development of severe delayed myelosuppression in most patients. To investigate the potential of DNA repair proteins to reduce bone marrow sensitivity to the CENUs, we transferred the Escherichia coli ada gene, which encodes a Mr 39,000 O6-alkylguanine-DNA alkyltransferase (ATase), into murine bone marrow cells by the use of a high-titer ecotropic retrovirus. The ada-encoded ATase is resistant to O6-benzylguanine (O6-BG), a potent inhibitor of the mammalian ATases, thus affording the bone marrow an additional level of protection against CENUs. In methylcellulose cultures, ada-infected hematopoietic progenitor cells were twice as resistant as uninfected cells to the toxic effects of 1, 3-bis(2-chloroethyl)-1-nitrosourea (BCNU) following treatment with O6-BG. Although showing no obvious protective effects against leukopenia, overexpression of the bacterial ATase activity reduced the severity of anemia and thrombocytopenia in mice treated with O6-BG and BCNU. These effects, which were maximal at a BCNU dose of 12.5 mg/kg, were associated with improved survival when BCNU was given at this dose. At lower BCNU doses cytotoxicity was limited in both transduced and control mice, and at higher doses the protective effect was saturated due to cytotoxicity. These results suggest that ada gene therapy may be a feasible approach to amelioration of delayed myelosuppression following O6-BG plus CENU combination chemotherapy.

Wild-type p53 can induce p21 and apoptosis in neuroblastoma cells but the DNA damage-induced G1 checkpoint function is attenuated.

p53 is a tumor suppressor protein important in the regulation of apoptosis. Because p53 functions as a transcription factor, cellular responses depend upon activity of p53 localized in the nucleus. Cytoplasmic sequestration of p53 has been proposed as a mechanism by which the function of this protein can be suppressed, particularly in tumor types such as neuroblastoma in which the frequency of mutations of p53 is low. Data presented here demonstrate that nuclear p53 protein is expressed in a panel of neuroblastoma cell lines, and after exposure to DNA damage, transcriptionally active p53 expression can be induced. After exposure to both equitoxic IC80 and 10-Gy doses of ionizing radiation, both p53 and p21 were induced, but G1 cell cycle arrest was attenuated. To investigate whether the DNA damage signaling pathway was incapable of inducing sufficient p53 in these cells, we expressed additional wild-type p53 after adenoviral vector transduction. This exogenous p53 expression also resulted in p21 induction but was unable to enhance the G1 arrest, suggesting that the pathway downstream from p53 is nonfunctional. Although p53-mediated G1 arrest is attenuated in neuroblastoma cells, the ability of p53 to induce apoptosis appears functional, consistent with its chemosensitive phenotype. This work demonstrates that p53 is expressed in the nucleus of neuroblastoma cells and can mediate induction of p21. However, this cell type appears to have an attenuated ability to mediate a DNA damage-induced G1 cell cycle arrest.

Ocular pharmacology of bicyclic hexahydroaporphines.

This paper explores the ocular hypotensive actions of bicyclic analogs of hexahydroaporphine (HHA), specifically nor-HHA, in an attempt to shed light on the mechanism(s) by which they lower intraocular pressure (IOP). Studies involving the measurement of IOP and aqueous humor production were conducted in ocular normotensive albino rabbits, while those involving smooth muscle contractility utilized isolated bovine iris. The ability of nor-HHA to produce a sustained drop in IOP is linked to both a functioning adrenergic nervous system and the availability of the products of cyclooxygenase metabolism. Although aqueous flow is not impacted by the bicyclic structures, the significant enhancement of outflow facility points to a probable mechanism of IOP-lowering action. Nor-HHA had no direct contractile or relaxant action on bovine irides, but does cause a concentration-dependent inhibition of carbachol-evoked contractions. This inhibition was reversed by inhibitors of phospholipase A(2) and cyclooxygenase, but not by inhibitors of lipoxygenase, again indicating a role for prostaglandins in the ocular pharmacological action of bicyclic HHAs. Pretreatment with a nitric oxide (NO) scavenger also reversed the ability of nor-HHA to inhibit carbachol-induced contractions, implying a role for NO in the postjunctional actions of HHAs.

Contribution of social and developmental factors to lead exposure during the first year of life.

The social and developmental correlates of early lead exposure were explored in an interim analysis of data from an ongoing longitudinal investigation in Cincinnati. Regardless of the apparent net availability of lead in the infant's physical environment, parental behavior was still significantly associated with infant blood lead levels. However, this was only the case after infants in the study reached 6 months of age and beyond when prewalking progression and early walking made parental management all the more critical. Future lead screening and abatement programs should include supports for the caretaker-child relationship.

A comparison of three pulmonary artery oximetry catheters in intensive care unit patients.

OBJECTIVE: To compare the clinical performance of three pulmonary artery oximetry catheters (Oximetrix 3, SAT-2, and HEMOPRO2) in intensive care unit (ICU) patients. DESIGN: Unblinded comparison of performance over 24 h using an IL-282 CO-oximeter as a criterion standard. SETTING: Multispecialty adult ICU at a university teaching hospital. PATIENTS: Thirty critically ill patients selected from those requiring hemodynamic monitoring for medical management. MEASUREMENTS AND MAIN RESULTS: By all measures, performance of the Oximetrix 3 and SAT-2 systems were comparable; bias +/- precision were -1.98 +/- 3.07 and +1.80 +/- 3.49, respectively, vs -2.28 +/- 5.24 for the HEMOPRO2. The Oximetrix 3 and SAT-2 systems demonstrated consistent performance over the range of saturations tested, though Oximetrix 3 tended to underestimate and SAT-2 tended to overestimate the CO-oximeter value. The HEMOPRO2 underestimated the CO-oximetry-derived saturation, although this was not constant across the range of values tested. The 95 percent confidence limits based on intrasubject variability were similar (+/- 4.59, +/- 5.66, and +/- 6.56 for the Oximetrix 3, SAT-2, and HEMOPRO2, respectively); however, the 95 percent confidence limits based on total variability, while similar for Oximetrix 3 (+/- 6.03) and SAT-2 (+/- 6.86), were larger for the HEMOPRO2 (+/- 10.30). The expected SD was similar for the three systems (2.03, 2.50, and 2.90 for the Oximetrix 3, SAT-2, and HEMOPRO2 systems, respectively). None of the systems equaled or exceeded (p greater than 0.05) the manufacturers' published specifications, which, in all cases, are listed as +/- 2 percent (saturation; 1 SD) when compared with bench oximetry. CONCLUSIONS: Although each system measures mixed venous oxygen saturation, the Oximetrix 3 and SAT-2 systems demonstrate closer agreement with CO-oximetry. However, none of these catheters provided statistically significant evidence that they would perform within +/- 2 percent of CO-oximetry. As a continuous monitor used to detect changes or trends, any of the three may be acceptable.

The risk of peripheral vein phlebitis associated with chlorhexidine-coated catheters: a randomized, double-blind trial.

OBJECTIVE: To evaluate the risk of phlebitis associated with chlorhexidine-coated polyurethane catheters in peripheral veins. DESIGN: A randomized, double-blinded trial comparing chlorhexidine-coated polyurethane catheters with uncoated polyurethane catheters. SETTING: A university hospital. PATIENTS: Adult medicine and surgery patients. INTERVENTIONS: Certified registered nurse anesthetists or an infusion team consisting of nurses and physicians inserted the catheters. Catheter insertion sites were scored twice daily for evidence of phlebitis. At the time catheters were removed, a quantitative blood culture was performed, and catheters were sonicated for quantitative culture. RESULTS: Of 221 evaluable catheters, phlebitis developed in 18 (17%) of 105 coated catheters, compared to 27 (23%) of 116 uncoated catheters (relative risk [RR], 0.74; 95% confidence interval [CI95], 0.43-1.26; P = .32). By survival analysis, chlorhexidine-coated catheters had a lower risk of phlebitis during the first 3 days (P = .06), but not when all catheters were considered in both patient groups (P = .31). In the absence of catheter colonization, the incidence of phlebitis was 21% (16/76) and 24% (20/86) for coated and uncoated catheters, respectively (P = .85), whereas in the presence of catheter colonization, the incidence of phlebitis was 14% (1/7) and 80% (4/5) for coated and uncoated catheters, respectively (RR, 0.18; CI95, 0.03-1.15; P = .07). CONCLUSION: The risk of phlebitis in the presence of catheter colonization was 82% lower for chlorhexidine-coated polyurethane catheters compared to otherwise identical uncoated catheters.