- Email: [email protected]
Wound Dehiscence or Failure to Heal following Venous Access Port Placement in Patients Receiving Bevacizumab Therapy
Wound Dehiscence or Failure to Heal following Venous Access Port Placement in Patients Receiving Bevacizumab Therapy Walter J. Zawacki, NP, T. Gregory Walker, MD, Emily DeVasher, RN, Elkan F. Halpern, PhD, Arthur C. Waltman, MD, Stephan T. Wicky, MD, David P. Ryan, MD, and Sanjeeva P. Kalva, MD
PURPOSE: To determine the incidence of wound dehiscence or failure to heal after port placement in patients receiving bevacizumab therapy. A hypothesis was tested that the mean interval between bevacizumab administration and port placement was shorter in patients who had dehiscence than in those who did not. MATERIALS AND METHODS: Medical records of all patients who had venous access ports placed from July 2006 through December 2007 were retrospectively reviewed. A total of 195 ports were placed in 189 patients (106 men) who were treated with bevacizumab within 120 days of port placement. The incidence of wound dehiscence and the significance of dose timing relative to port placement in these patients were calculated. RESULTS: Six of 195 ports (3.1%) were associated with wound dehiscence requiring port removal. The mean interval between bevacizumab dosing and port placement in patients without dehiscence (n ⴝ 189) was 16.9 days. The mean interval in patients with dehiscence (n ⴝ 6) was 10.8 days. A two-tailed Wilcoxon test was performed, which yielded a P value of .0150. A statistically significant difference in the mean interval between bevacizumab dosing and port placement exists between patients with dehiscence and those without. CONCLUSIONS: Wound dehiscence after port placement was related to timing of bevacizumab therapy. Patients receiving bevacizumab within 10 days of port placement had a higher incidence of wound dehiscence. J Vasc Interv Radiol 2009; 20:624 – 627 Abbreviation:
VEGF ⫽ vascular endothelial growth factor
A totally implanted venous access system to replace external catheters in cancer treatment was first tested and described in 1982 (1). Since that time, implantable venous access ports have gained wide acceptance, and are routinely used for the delivery of chemotherapy, intravenous fluids, blood prod-
From the Department of Radiology, Massachusetts General Hospital, Harvard Medical School, 55 Fruit Street, GRB 290, Boston, MA 02114. Received March 28, 2008; final revision received January 9, 2009; accepted January 18, 2009. Address correspondence to W.J.Z.; E-mail: [email protected] None of the authors have identified a conflict of interest. © SIR, 2009 DOI: 10.1016/j.jvir.2009.01.022
624
ucts, and total parenteral nutrition, and for concurrent blood sampling. Common complications associated with port placement include infection (1%– 6%), venous thrombosis (ⱕ5%), and catheter occlusion (1%– 6%) (2– 4). Port erosion through the skin is an infrequently reported complication (⬍1%), and has been associated with cachexia and suboptimal device selection (ie, high-profile ports in thin patients) (2,5,6). Bevacizumab (Avastin; Genentech, South San Francisco, California) is a recombinant humanized monoclonal immunoglobulin G1 antibody. It has been approved by the United States Food and Drug Administration for the treatment of metastatic colorectal cancer; unresectable, locally advanced, recurrent, or metastatic nonsquamous
non–small-cell lung cancer; and metastatic HER2-negative breast cancer. Bevacizumab binds to and inhibits the biologic activity of human vascular endothelial growth factor (VEGF), preventing the interaction of VEGF with its receptors (VEGF receptors 1 and 2, or Flt-1 and KDR, respectively) on the surface of endothelial cells (7). Normally, the interaction of VEGF with its receptors leads to endothelial cell proliferation and new blood vessel formation (ie, angiogenesis). When VEGF is bound to bevacizumab, it cannot stimulate endothelial cell proliferation, and angiogenesis is inhibited. Label warnings for bevacizumab administration include gastrointestinal perforations, wound healing complications, and hemorrhage. Administration guide-
Volume 20
Number 5
lines advise avoiding the initiation of bevacizumab therapy for at least 28 days after major surgery. The appropriate interval between termination of bevacizumab therapy and subsequent elective surgery required to avoid potential impaired wound healing or dehiscence has not been studied to our knowledge. Additionally, treatment guidelines for the timing of bevacizumab treatment with respect to minimally invasive procedures are lacking. This study seeks to determine the incidence of wound dehiscence or failure to heal after port placement in patients receiving bevacizumab therapy. It also seeks to establish any possible relationship between the time of administration of the agent and the development of wound complications related to port placement. A hypothesis was tested that the mean interval between bevacizumab administration and port placement was shorter in patients who had dehiscence than in those who did not.
MATERIALS AND METHODS In this institutional review board– approved, Health Insurance Portability and Accountability Act– compliant retrospective study, informed consent was waived, and research was limited to the use of health/medical records. A search of the electronic medical records at our institution, a large urban tertiary-care hospital, was conducted for the period from July 1, 2006, through December 31, 2007. Port placements and removals during the study period were identified by review of the departmental scheduling software (Hi-IQ; ConexSys, Albion, Rhode Island). The institutional electronic medical record system (CAS; Massachusetts General Hospital, Boston, Massachusetts) was subsequently reviewed for patient-specific data, which were then captured in Filemaker Pro software (version 9.0v3; Filemaker, Santa Clara, California) and subsequently exported to an Excel 2003 spreadsheet (Microsoft, Redmond, Washington). SAS software (version 9.1; SAS, Cary, North Carolina) was used for statistical analysis.
Zawacki et al
•
625
Patients
Port Placement
During the period of study, there were 1,110 ports placed in 1,082 patients. Of the 1,110 ports placed, we further identified 195 port placements in 189 patients who were treated with bevacizumab therapy within 120 days of placement (ie, the period encompassing 60 days before and 60 days after port placement). Indications for port placement in patients receiving bevacizumab therapy included the following malignancies: colorectal (n ⫽ 66), lung (n ⫽ 8), breast (n ⫽ 13), head and neck (n ⫽ 28), esophageal (n ⫽ 9), cholangiocarcinoma (n ⫽ 20), gastric (n ⫽ 4), duodenal (n ⫽ 1), pancreatic (n ⫽ 16), endometrial (n ⫽ 2), ovarian (n ⫽ 10), uterine (n ⫽ 2), hepatocellular (n ⫽ 10), renal (n ⫽ 2), and appendiceal (n ⫽ 4). There were 106 men and 89 women. Mean age at port placement was 56 years (range, 26 – 86 y; median, 58 y). The mean age among male patients was 57 years (range, 26 – 86; median, 60 y), and that among female patients was 54 years (range, 27– 83 y; median, 53 y). The date of port placement and device type were recorded: single-lumen port (n ⫽ 87), doublelumen port (n ⫽ 14), or single-lumen computed tomography power injection– capable port (n ⫽ 94; all from Bard, Murray Hill, New Jersey). Venous access for port placement was obtained by the right internal jugular vein (n ⫽ 170), left internal jugular vein (n ⫽ 19), right arm (n ⫽ 2), left arm (n ⫽ 2), or translumbar route (n ⫽ 2). Comorbid risk factors were recorded, including the presence of diabetes (n ⫽ 19) and nutritional status as indicated by serum albumin levels (mean, 3.8 g/dL; range, 1.7– 8.3 g/dL; median, 3.9 g/dL). Renal and hepatic function in the dehiscence group at the time of port placement were recorded as follows: serum creatinine (mean, 0.9 mg/dL; range, 0.6 –1.1 mg/dL; median, 0.9 mg/dL); alanine aminotransferase (mean, 36.8 U/L; range, 12– 67 U/L; median, 36 U/L); aspartate aminotransferase (mean, 25.0 U/L; range, 16 –39 U/L; median, 22 U/L); serum alkaline phosphatase (mean, 96.7 U/L; range, 61–188 U/L; median, 83 U/L); and total bilirubin (mean, 0.3 mg/dL; range, 0.2– 0.5 mg/dL; median, 0.3 mg/dL).
All ports were placed by the interventional radiology service according to a standard protocol. Ports were placed by attending physicians, interventional radiology fellows, experienced procedural nurse practitioners, and radiology residents under direct supervision. Patients were advised to take no food for 8 hours before port placement. Clear liquids were withheld for 2 hours before the procedure. Coagulation parameters were evaluated before each port placement, and coagulopathies (International Normalized Ratio ⱖ1.5 and platelet count ⱕ50,000/ mm3) were corrected as needed. Skin preparation included hair removal followed by antiseptic wash with combination 2% chlorhexidine gluconate and 70% isopropyl alcohol. Intravenous moderate sedation was provided per hospital protocol with fentanyl citrate and midazolam hydrochloride (Versed; Roche, Nutley, New Jersey). Port placement in irradiated or potentially irradiated tissue was avoided. Large or pendulous chest wall tissue was retracted caudally to facilitate tunneling and catheter positioning. Local anesthesia was achieved with 2% lidocaine with 8.4% sodium bicarbonate in a 9:1 proportion. Intravenous cefazolin 1 g (or clindamycin 600 mg) was administered for antibiotic prophylaxis. The port was sutured deep in the subcutaneous pocket and was secured with absorbable 3– 0 Vicryl sutures (Ethicon, Somerville, New Jersey). The skin incision was closed with interrupted 3– 0 Vicryl sutures, compound tincture of benzoin and alcohol, and Steri-Strips (3M, St. Paul, Minnesota). A transparent polyurethane film dressing over gauze was then applied. All patients received oral and written postprocedural instructions to avoid strenuous activity and heavy lifting for 7 days and to keep the dressing dry for 3 days after the procedure. Patients were permitted to shower after removal of the polyurethane film and gauze dressings. Bevacizumab Therapy Intravenous infusion protocols for bevacizumab therapy vary depending on the underlying malignancy. Metastatic colorectal carcinoma is treated with an intravenous infusion of 5 mg/
626
•
Wound Dehiscence after Port Placement during Bevacizumab Therapy
Table 1 Frequency Distribution, Minimum Interval between Bevacizumab Dosing and Port Placement in Patients without Dehiscence (n ⴝ 189) Minimum Interval (d) 0–7 8–14 15–21 22–28 29–35 36–42 43–49 50–56 57–60 Total
Incidence 44 (23.3) 63 (33.3) 41 (21.7) 13 (6.9) 9 (4.8) 5 (2.6) 9 (4.8) 4 (2.1) 1 (0.5) 189 (100)
Note.—The mean minimum interval between bevacizumab therapy and port placement was 16.9 days among patients without dehiscence.
kg or 10 mg/kg every 14 days; nonsquamous non–small-cell lung cancer with 15 mg/kg every 3 weeks; and metastatic breast cancer with 10 mg/kg every 14 days. The estimated half-life of bevacizumab is 20 days (range, 11–50 d). Clearance varies by body weight, sex, and tumor burden. No studies have been conducted to examine the pharmacokinetics of bevacizumab in patients with renal or hepatic impairments. There are no recommended dose reductions for the use of bevacizumab (7). The mean, minimum, maximum, and median cumulative bevacizumab dose administered to patients in the 120-day period was calculated for patients who had wound dehiscence (mean, 1,923 mg; range, 1,398 –2,546 mg; median, 1,863 mg) and those who did not (mean, 1,987 mg; range, 221– 6,133 mg; median, 1,709 mg). Follow-up Patients were followed up (mean, 307 days; range, 44 –581 d; median, 281 d) to determine if there were any wound healing complications and if ports were removed or remained in place. The indication for removal and days to removal from port placement were recorded. Statistical Analysis Statistical analysis was conducted to determine whether a significant dif-
JVIR
Table 2 Minimum Interval between Bevacizumab Dosing and Port Placement in Patients with Dehiscence (n ⴝ 6) Timing of Bevacizumab Dosing (d)
Cumulative Incidence (%) 23.3 58.2 78.3 85.2 89.9 92.6 97.4 99.5 100.0 —
May 2009
Pt. No.
Before Port Placement
After Port Placement
Minimum Interval, Bevacizumab to Port Placement
1 2 3 4 5 6
7 NA NA 7 20 44
7 3 3 7 1 NA
7 3 3 7 1 44
Note.—The mean minimum interval between bevacizumab therapy and IVAP placement was 10.8 days among patients with dehiscence. NA ⫽ not applicable.
ference existed in the timing of administration of bevacizumab therapy relative to port placement in patients who had dehiscence versus those who did not. The number of days elapsed between port placement and administration of bevacizumab was identified and recorded for the 60-day periods before and after port placement in both groups. A nonparametric twotailed Wilcoxon two-sample test was conducted and a P value was calculated. A P value less than .05 was considered significant.
RESULTS At the end of the follow-up period, of the 1,110 ports placed, 891 remained in place. Indications for port removal through the follow-up period were as follows: completion of treatment in 153 (13.78%), infection in 42 (3.78%), malfunction in eight (0.72%; secondary to kinked catheter in three [0.27%], extravasation in three [0.27%], and catheter tip migration/malposition in two [0.18%]), fibrin sheath in four (0.36%), erosion in one (0.09%), thrombus in two (0.18%), a flipped port in one (0.09%), exposed catheter at venotomy site in two (0.18%), and wound dehiscence in six (0.54%). The six patients who had wound dehiscence after port placement had all been treated with bevacizumab therapy. There were no cases of wound dehiscence in patients who did not receive bevacizumab therapy. A single patient who was not receiving bevacizumab therapy presented with thinning skin over the port/catheter connection and the device
was removed prophylactically. This patient, with pancreatic adenocarcinoma, had a body mass index of 19.2 kg/m2 at the time of port placement, which subsequently diminished to 18.9 kg/m2 at the time of port removal. Of the 195 ports placed in patients receiving bevacizumab, 26 (13.3%) required removal by the end of the follow-up period. Port removal was necessary in six patients (3.1%) because of wound dehiscence and in 20 patients (10.3%) for other reasons. Indications for removal other than dehiscence were as follows: completion of treatment in seven (3.59%), infection in eight (4.10%), malfunction in two (1.03%; secondary to a kinked catheter in one [0.51%] and extravasation in one [0.51%]), fibrin sheath in one (0.68%), thrombus in one (0.51%), and a flipped port in one (0.68%). Of the 189 patients receiving bevacizumab therapy who did not have wound dehiscence, 11 received bevacizumab therapy solely during the 60day period before port placement, 23 patients during both 60-day periods before and after port placement, and 155 patients during only the 60-day period after port placement. In the dehiscence group (n ⫽ 6), one patient received bevacizumab therapy solely during the 60-day period before port placement, three patients during the 60day periods before and after port placement, and two patients during only the 60-day period after port placement. A frequency distribution of the minimum interval in days between bevacizumab dosing and port placement in the no-dehiscence group is shown in
Volume 20
Number 5
Table 1. Table 2 provides the minimum interval between bevacizumab dosing and port placement in the 60-day periods before and after port placement in the dehiscence group. The mean interval between bevacizumab dosing and port placement in the 189 patients without dehiscence was 16.9 days (range, 2–58 d; median, 14 d). The mean interval in the six patients with dehiscence was 10.8 days (range, 1– 44 d; median, 5 d). A nonparametric two-tailed Wilcoxon two-sample test was performed to test the directional hypothesis, and the P value was .0150. This suggested a significant relationship between the timing of administration of bevacizumab therapy and wound dehiscence. All ports in the dehiscence group were subsequently removed after a mean period of 43.5 days (range, 21–75 d; median, 39.5 d). No patient in the dehiscence group had diabetes. Albumin levels in the dehiscence group were normal in all patients except for one whose albumin level was 3.0 g/dL (mean, 3.8 g/dL; range, 3.0 – 4.7 g/dL; median, 3.8 g/dL; institutional reference range, 3.3– 5.0 g/dL). Although the pharmacokinetics of bevacizumab in patients with renal or hepatic dysfunction has not been studied to our knowledge, laboratory results in our study suggest no significant renal or hepatic impairment in the patients who had dehiscence.
DISCUSSION Wound dehiscence after port placement in our study occurred only in patients receiving bevacizumab therapy. Of the 195 patients receiving bevacizumab therapy, all ports in the dehiscence group (n ⫽ 6) were single-lumen ports placed by experienced operators, with venous access via the right internal jugular vein. The intraprocedural technical success rate was 100%, and no violations of sterile technique or cases of excessive bleeding were noted. Wound dehiscence after placement of implantable ports necessitates removal of the device. In all the patients with wound dehiscence, we were unable to suture the incision after removal. The wounds were packed for healing by secondary intention, and home nursing services were arranged for wound care. Wound dehiscence as a complication of port placement causes patient distress, disruption of chemotherapeutic regi-
Zawacki et al men, risk of infection, and exposure to additional invasive procedures. As can be seen in Table 2, patient 6 represents a relative outlier in the dehiscence group, as the bevacizumab therapy was administered 44 days before port placement. This patient had a series of arterial thromboembolic cerebral vascular accidents after bevacizumab therapy. He was additionally found to have lupus antibody–positive coagulopathy. In all other patients with wound dehiscence after port placement, bevacizumab was given within 1 week of port placement. Berry et al (8), in a poster presentation at the 2006 Gastrointestinal Cancers Symposium, reported preliminary results of a study on the effects of bevacizumab therapy on wound healing after minor surgical procedures. The study was designed to involve as many as 2,000 patients in 40 countries. Their findings included the incidence of wound healing complications in patients starting bevacizumab therapy shortly after venous access device implantation, with a reported 1.1% incidence (n ⫽ 6) of wound healing complications in 534 patients treated with bevacizumab after port placement, and one (0.6%) wound healing complication in 182 patients who started bevacizumab therapy within 7 days after port implantation. The study did not consider patients who received bevacizumab therapy before port implantation. We found a higher incidence of wound dehiscence when bevacizumab was given within 1 week of port placement. In our study, five of 49 patients (10.2%) received bevacizumab within 7 days of port placement had wound dehiscence. Our findings are similar to a recent case series that reported three instances of implantable venous access device erosion through the skin in patients treated with anti-VEGF therapy (9). Our study is limited by the retrospective, single-center nature of its design and the relatively small number of patients who had wound dehiscence after port placement. Although we did not specifically address the effects of dose quantity and cumulative dose in our study, we noted the comparable mean cumulative dose administered in the dehiscence and no-dehis-
•
627
cence groups. The effects of renal and hepatic impairment on bevacizumab clearance have not been studied. Further study is warranted to address these issues. We conclude that wound dehiscence after port placement in our study was related to the timing of bevacizumab therapy. Patients receiving bevacizumab within 10 days of port placement had a higher incidence of wound dehiscence. References 1. Neiderhuber JE, Ensminger W, Gyves JW, Liepman M, Doan K, Cozzi E. Totally implanted venous and arterial access system to replace external catheters in cancer treatment. Surgery 1982; 92:706 –712. 2. Cil BE, Murat C, Bora P, et al. Subcutaneous venous port implantation in adult patients: a single center experience. Diagn Intervent Radiol 2006; 12:93–98. 3. Simpson KR, Hovsepian DM, Picus D. Interventional radiologic placement of chest wall ports: results and complications in 161 consecutive patients. J Vasc Interv Radiol 1997; 8:189 –195. 4. Shetty PC, Mody MK, Kastan DJ, et al. Outcome of 350 implanted chest ports placed by interventional radiologists. J Vasc Interv Radiol 1997; 8:991–995. 5. Morris SL, Jaques PF, Mauro MA. Radiology-assisted placement of implantable subcutaneous infusion ports for longterm venous access. Radiology 1992; 184: 149 –151. 6. Hills JR, Cardella JF, Cardella K, Waybill PN. Experience with 100 consecutive central venous access arm ports placed by interventional radiologists. J Vasc Interv Radiol 1997; 8:983–989. 7. Avastin (bevacizumab) [prescribing information]. San Francisco: Genentech, 2008. Available at http://www.gene.com/ gene/products/information/pdf/avastinprescribing.pdf. Accessed December 30, 2008. 8. Berry S, Kretzschmar MA, Cunningham D, et al. Lack of effect of starting bevacizumab shortly after venous access device implantation on wound healing/bleeding complications: preliminary results from first BEAT. Presented at the American Society of Clinical Oncology Gastrointestinal Cancers Symposium; January 26 –28, 2006; San Francisco, CA. 9. Almhanna K, Pelley R, Thomas Budd G, Davidson J, Moore HCF. Subcutaneous implantable venous access device erosion through the skin in patients treated with anti-vascular endothelial growth factor therapy: a case series. Anti-Cancer Drugs 2008; 19:217–219.
PURPOSE: To determine the incidence of wound dehiscence or failure to heal after port placement in patients receiving bevacizumab therapy. A hypothesis was tested that the mean interval between bevacizumab administration and port placement was shorter in patients who had dehiscence than in those who did not. MATERIALS AND METHODS: Medical records of all patients who had venous access ports placed from July 2006 through December 2007 were retrospectively reviewed. A total of 195 ports were placed in 189 patients (106 men) who were treated with bevacizumab within 120 days of port placement. The incidence of wound dehiscence and the significance of dose timing relative to port placement in these patients were calculated. RESULTS: Six of 195 ports (3.1%) were associated with wound dehiscence requiring port removal. The mean interval between bevacizumab dosing and port placement in patients without dehiscence (n ⴝ 189) was 16.9 days. The mean interval in patients with dehiscence (n ⴝ 6) was 10.8 days. A two-tailed Wilcoxon test was performed, which yielded a P value of .0150. A statistically significant difference in the mean interval between bevacizumab dosing and port placement exists between patients with dehiscence and those without. CONCLUSIONS: Wound dehiscence after port placement was related to timing of bevacizumab therapy. Patients receiving bevacizumab within 10 days of port placement had a higher incidence of wound dehiscence. J Vasc Interv Radiol 2009; 20:624 – 627 Abbreviation:
VEGF ⫽ vascular endothelial growth factor
A totally implanted venous access system to replace external catheters in cancer treatment was first tested and described in 1982 (1). Since that time, implantable venous access ports have gained wide acceptance, and are routinely used for the delivery of chemotherapy, intravenous fluids, blood prod-
From the Department of Radiology, Massachusetts General Hospital, Harvard Medical School, 55 Fruit Street, GRB 290, Boston, MA 02114. Received March 28, 2008; final revision received January 9, 2009; accepted January 18, 2009. Address correspondence to W.J.Z.; E-mail: [email protected] None of the authors have identified a conflict of interest. © SIR, 2009 DOI: 10.1016/j.jvir.2009.01.022
624
ucts, and total parenteral nutrition, and for concurrent blood sampling. Common complications associated with port placement include infection (1%– 6%), venous thrombosis (ⱕ5%), and catheter occlusion (1%– 6%) (2– 4). Port erosion through the skin is an infrequently reported complication (⬍1%), and has been associated with cachexia and suboptimal device selection (ie, high-profile ports in thin patients) (2,5,6). Bevacizumab (Avastin; Genentech, South San Francisco, California) is a recombinant humanized monoclonal immunoglobulin G1 antibody. It has been approved by the United States Food and Drug Administration for the treatment of metastatic colorectal cancer; unresectable, locally advanced, recurrent, or metastatic nonsquamous
non–small-cell lung cancer; and metastatic HER2-negative breast cancer. Bevacizumab binds to and inhibits the biologic activity of human vascular endothelial growth factor (VEGF), preventing the interaction of VEGF with its receptors (VEGF receptors 1 and 2, or Flt-1 and KDR, respectively) on the surface of endothelial cells (7). Normally, the interaction of VEGF with its receptors leads to endothelial cell proliferation and new blood vessel formation (ie, angiogenesis). When VEGF is bound to bevacizumab, it cannot stimulate endothelial cell proliferation, and angiogenesis is inhibited. Label warnings for bevacizumab administration include gastrointestinal perforations, wound healing complications, and hemorrhage. Administration guide-
Volume 20
Number 5
lines advise avoiding the initiation of bevacizumab therapy for at least 28 days after major surgery. The appropriate interval between termination of bevacizumab therapy and subsequent elective surgery required to avoid potential impaired wound healing or dehiscence has not been studied to our knowledge. Additionally, treatment guidelines for the timing of bevacizumab treatment with respect to minimally invasive procedures are lacking. This study seeks to determine the incidence of wound dehiscence or failure to heal after port placement in patients receiving bevacizumab therapy. It also seeks to establish any possible relationship between the time of administration of the agent and the development of wound complications related to port placement. A hypothesis was tested that the mean interval between bevacizumab administration and port placement was shorter in patients who had dehiscence than in those who did not.
MATERIALS AND METHODS In this institutional review board– approved, Health Insurance Portability and Accountability Act– compliant retrospective study, informed consent was waived, and research was limited to the use of health/medical records. A search of the electronic medical records at our institution, a large urban tertiary-care hospital, was conducted for the period from July 1, 2006, through December 31, 2007. Port placements and removals during the study period were identified by review of the departmental scheduling software (Hi-IQ; ConexSys, Albion, Rhode Island). The institutional electronic medical record system (CAS; Massachusetts General Hospital, Boston, Massachusetts) was subsequently reviewed for patient-specific data, which were then captured in Filemaker Pro software (version 9.0v3; Filemaker, Santa Clara, California) and subsequently exported to an Excel 2003 spreadsheet (Microsoft, Redmond, Washington). SAS software (version 9.1; SAS, Cary, North Carolina) was used for statistical analysis.
Zawacki et al
•
625
Patients
Port Placement
During the period of study, there were 1,110 ports placed in 1,082 patients. Of the 1,110 ports placed, we further identified 195 port placements in 189 patients who were treated with bevacizumab therapy within 120 days of placement (ie, the period encompassing 60 days before and 60 days after port placement). Indications for port placement in patients receiving bevacizumab therapy included the following malignancies: colorectal (n ⫽ 66), lung (n ⫽ 8), breast (n ⫽ 13), head and neck (n ⫽ 28), esophageal (n ⫽ 9), cholangiocarcinoma (n ⫽ 20), gastric (n ⫽ 4), duodenal (n ⫽ 1), pancreatic (n ⫽ 16), endometrial (n ⫽ 2), ovarian (n ⫽ 10), uterine (n ⫽ 2), hepatocellular (n ⫽ 10), renal (n ⫽ 2), and appendiceal (n ⫽ 4). There were 106 men and 89 women. Mean age at port placement was 56 years (range, 26 – 86 y; median, 58 y). The mean age among male patients was 57 years (range, 26 – 86; median, 60 y), and that among female patients was 54 years (range, 27– 83 y; median, 53 y). The date of port placement and device type were recorded: single-lumen port (n ⫽ 87), doublelumen port (n ⫽ 14), or single-lumen computed tomography power injection– capable port (n ⫽ 94; all from Bard, Murray Hill, New Jersey). Venous access for port placement was obtained by the right internal jugular vein (n ⫽ 170), left internal jugular vein (n ⫽ 19), right arm (n ⫽ 2), left arm (n ⫽ 2), or translumbar route (n ⫽ 2). Comorbid risk factors were recorded, including the presence of diabetes (n ⫽ 19) and nutritional status as indicated by serum albumin levels (mean, 3.8 g/dL; range, 1.7– 8.3 g/dL; median, 3.9 g/dL). Renal and hepatic function in the dehiscence group at the time of port placement were recorded as follows: serum creatinine (mean, 0.9 mg/dL; range, 0.6 –1.1 mg/dL; median, 0.9 mg/dL); alanine aminotransferase (mean, 36.8 U/L; range, 12– 67 U/L; median, 36 U/L); aspartate aminotransferase (mean, 25.0 U/L; range, 16 –39 U/L; median, 22 U/L); serum alkaline phosphatase (mean, 96.7 U/L; range, 61–188 U/L; median, 83 U/L); and total bilirubin (mean, 0.3 mg/dL; range, 0.2– 0.5 mg/dL; median, 0.3 mg/dL).
All ports were placed by the interventional radiology service according to a standard protocol. Ports were placed by attending physicians, interventional radiology fellows, experienced procedural nurse practitioners, and radiology residents under direct supervision. Patients were advised to take no food for 8 hours before port placement. Clear liquids were withheld for 2 hours before the procedure. Coagulation parameters were evaluated before each port placement, and coagulopathies (International Normalized Ratio ⱖ1.5 and platelet count ⱕ50,000/ mm3) were corrected as needed. Skin preparation included hair removal followed by antiseptic wash with combination 2% chlorhexidine gluconate and 70% isopropyl alcohol. Intravenous moderate sedation was provided per hospital protocol with fentanyl citrate and midazolam hydrochloride (Versed; Roche, Nutley, New Jersey). Port placement in irradiated or potentially irradiated tissue was avoided. Large or pendulous chest wall tissue was retracted caudally to facilitate tunneling and catheter positioning. Local anesthesia was achieved with 2% lidocaine with 8.4% sodium bicarbonate in a 9:1 proportion. Intravenous cefazolin 1 g (or clindamycin 600 mg) was administered for antibiotic prophylaxis. The port was sutured deep in the subcutaneous pocket and was secured with absorbable 3– 0 Vicryl sutures (Ethicon, Somerville, New Jersey). The skin incision was closed with interrupted 3– 0 Vicryl sutures, compound tincture of benzoin and alcohol, and Steri-Strips (3M, St. Paul, Minnesota). A transparent polyurethane film dressing over gauze was then applied. All patients received oral and written postprocedural instructions to avoid strenuous activity and heavy lifting for 7 days and to keep the dressing dry for 3 days after the procedure. Patients were permitted to shower after removal of the polyurethane film and gauze dressings. Bevacizumab Therapy Intravenous infusion protocols for bevacizumab therapy vary depending on the underlying malignancy. Metastatic colorectal carcinoma is treated with an intravenous infusion of 5 mg/
626
•
Wound Dehiscence after Port Placement during Bevacizumab Therapy
Table 1 Frequency Distribution, Minimum Interval between Bevacizumab Dosing and Port Placement in Patients without Dehiscence (n ⴝ 189) Minimum Interval (d) 0–7 8–14 15–21 22–28 29–35 36–42 43–49 50–56 57–60 Total
Incidence 44 (23.3) 63 (33.3) 41 (21.7) 13 (6.9) 9 (4.8) 5 (2.6) 9 (4.8) 4 (2.1) 1 (0.5) 189 (100)
Note.—The mean minimum interval between bevacizumab therapy and port placement was 16.9 days among patients without dehiscence.
kg or 10 mg/kg every 14 days; nonsquamous non–small-cell lung cancer with 15 mg/kg every 3 weeks; and metastatic breast cancer with 10 mg/kg every 14 days. The estimated half-life of bevacizumab is 20 days (range, 11–50 d). Clearance varies by body weight, sex, and tumor burden. No studies have been conducted to examine the pharmacokinetics of bevacizumab in patients with renal or hepatic impairments. There are no recommended dose reductions for the use of bevacizumab (7). The mean, minimum, maximum, and median cumulative bevacizumab dose administered to patients in the 120-day period was calculated for patients who had wound dehiscence (mean, 1,923 mg; range, 1,398 –2,546 mg; median, 1,863 mg) and those who did not (mean, 1,987 mg; range, 221– 6,133 mg; median, 1,709 mg). Follow-up Patients were followed up (mean, 307 days; range, 44 –581 d; median, 281 d) to determine if there were any wound healing complications and if ports were removed or remained in place. The indication for removal and days to removal from port placement were recorded. Statistical Analysis Statistical analysis was conducted to determine whether a significant dif-
JVIR
Table 2 Minimum Interval between Bevacizumab Dosing and Port Placement in Patients with Dehiscence (n ⴝ 6) Timing of Bevacizumab Dosing (d)
Cumulative Incidence (%) 23.3 58.2 78.3 85.2 89.9 92.6 97.4 99.5 100.0 —
May 2009
Pt. No.
Before Port Placement
After Port Placement
Minimum Interval, Bevacizumab to Port Placement
1 2 3 4 5 6
7 NA NA 7 20 44
7 3 3 7 1 NA
7 3 3 7 1 44
Note.—The mean minimum interval between bevacizumab therapy and IVAP placement was 10.8 days among patients with dehiscence. NA ⫽ not applicable.
ference existed in the timing of administration of bevacizumab therapy relative to port placement in patients who had dehiscence versus those who did not. The number of days elapsed between port placement and administration of bevacizumab was identified and recorded for the 60-day periods before and after port placement in both groups. A nonparametric twotailed Wilcoxon two-sample test was conducted and a P value was calculated. A P value less than .05 was considered significant.
RESULTS At the end of the follow-up period, of the 1,110 ports placed, 891 remained in place. Indications for port removal through the follow-up period were as follows: completion of treatment in 153 (13.78%), infection in 42 (3.78%), malfunction in eight (0.72%; secondary to kinked catheter in three [0.27%], extravasation in three [0.27%], and catheter tip migration/malposition in two [0.18%]), fibrin sheath in four (0.36%), erosion in one (0.09%), thrombus in two (0.18%), a flipped port in one (0.09%), exposed catheter at venotomy site in two (0.18%), and wound dehiscence in six (0.54%). The six patients who had wound dehiscence after port placement had all been treated with bevacizumab therapy. There were no cases of wound dehiscence in patients who did not receive bevacizumab therapy. A single patient who was not receiving bevacizumab therapy presented with thinning skin over the port/catheter connection and the device
was removed prophylactically. This patient, with pancreatic adenocarcinoma, had a body mass index of 19.2 kg/m2 at the time of port placement, which subsequently diminished to 18.9 kg/m2 at the time of port removal. Of the 195 ports placed in patients receiving bevacizumab, 26 (13.3%) required removal by the end of the follow-up period. Port removal was necessary in six patients (3.1%) because of wound dehiscence and in 20 patients (10.3%) for other reasons. Indications for removal other than dehiscence were as follows: completion of treatment in seven (3.59%), infection in eight (4.10%), malfunction in two (1.03%; secondary to a kinked catheter in one [0.51%] and extravasation in one [0.51%]), fibrin sheath in one (0.68%), thrombus in one (0.51%), and a flipped port in one (0.68%). Of the 189 patients receiving bevacizumab therapy who did not have wound dehiscence, 11 received bevacizumab therapy solely during the 60day period before port placement, 23 patients during both 60-day periods before and after port placement, and 155 patients during only the 60-day period after port placement. In the dehiscence group (n ⫽ 6), one patient received bevacizumab therapy solely during the 60-day period before port placement, three patients during the 60day periods before and after port placement, and two patients during only the 60-day period after port placement. A frequency distribution of the minimum interval in days between bevacizumab dosing and port placement in the no-dehiscence group is shown in
Volume 20
Number 5
Table 1. Table 2 provides the minimum interval between bevacizumab dosing and port placement in the 60-day periods before and after port placement in the dehiscence group. The mean interval between bevacizumab dosing and port placement in the 189 patients without dehiscence was 16.9 days (range, 2–58 d; median, 14 d). The mean interval in the six patients with dehiscence was 10.8 days (range, 1– 44 d; median, 5 d). A nonparametric two-tailed Wilcoxon two-sample test was performed to test the directional hypothesis, and the P value was .0150. This suggested a significant relationship between the timing of administration of bevacizumab therapy and wound dehiscence. All ports in the dehiscence group were subsequently removed after a mean period of 43.5 days (range, 21–75 d; median, 39.5 d). No patient in the dehiscence group had diabetes. Albumin levels in the dehiscence group were normal in all patients except for one whose albumin level was 3.0 g/dL (mean, 3.8 g/dL; range, 3.0 – 4.7 g/dL; median, 3.8 g/dL; institutional reference range, 3.3– 5.0 g/dL). Although the pharmacokinetics of bevacizumab in patients with renal or hepatic dysfunction has not been studied to our knowledge, laboratory results in our study suggest no significant renal or hepatic impairment in the patients who had dehiscence.
DISCUSSION Wound dehiscence after port placement in our study occurred only in patients receiving bevacizumab therapy. Of the 195 patients receiving bevacizumab therapy, all ports in the dehiscence group (n ⫽ 6) were single-lumen ports placed by experienced operators, with venous access via the right internal jugular vein. The intraprocedural technical success rate was 100%, and no violations of sterile technique or cases of excessive bleeding were noted. Wound dehiscence after placement of implantable ports necessitates removal of the device. In all the patients with wound dehiscence, we were unable to suture the incision after removal. The wounds were packed for healing by secondary intention, and home nursing services were arranged for wound care. Wound dehiscence as a complication of port placement causes patient distress, disruption of chemotherapeutic regi-
Zawacki et al men, risk of infection, and exposure to additional invasive procedures. As can be seen in Table 2, patient 6 represents a relative outlier in the dehiscence group, as the bevacizumab therapy was administered 44 days before port placement. This patient had a series of arterial thromboembolic cerebral vascular accidents after bevacizumab therapy. He was additionally found to have lupus antibody–positive coagulopathy. In all other patients with wound dehiscence after port placement, bevacizumab was given within 1 week of port placement. Berry et al (8), in a poster presentation at the 2006 Gastrointestinal Cancers Symposium, reported preliminary results of a study on the effects of bevacizumab therapy on wound healing after minor surgical procedures. The study was designed to involve as many as 2,000 patients in 40 countries. Their findings included the incidence of wound healing complications in patients starting bevacizumab therapy shortly after venous access device implantation, with a reported 1.1% incidence (n ⫽ 6) of wound healing complications in 534 patients treated with bevacizumab after port placement, and one (0.6%) wound healing complication in 182 patients who started bevacizumab therapy within 7 days after port implantation. The study did not consider patients who received bevacizumab therapy before port implantation. We found a higher incidence of wound dehiscence when bevacizumab was given within 1 week of port placement. In our study, five of 49 patients (10.2%) received bevacizumab within 7 days of port placement had wound dehiscence. Our findings are similar to a recent case series that reported three instances of implantable venous access device erosion through the skin in patients treated with anti-VEGF therapy (9). Our study is limited by the retrospective, single-center nature of its design and the relatively small number of patients who had wound dehiscence after port placement. Although we did not specifically address the effects of dose quantity and cumulative dose in our study, we noted the comparable mean cumulative dose administered in the dehiscence and no-dehis-
•
627
cence groups. The effects of renal and hepatic impairment on bevacizumab clearance have not been studied. Further study is warranted to address these issues. We conclude that wound dehiscence after port placement in our study was related to the timing of bevacizumab therapy. Patients receiving bevacizumab within 10 days of port placement had a higher incidence of wound dehiscence. References 1. Neiderhuber JE, Ensminger W, Gyves JW, Liepman M, Doan K, Cozzi E. Totally implanted venous and arterial access system to replace external catheters in cancer treatment. Surgery 1982; 92:706 –712. 2. Cil BE, Murat C, Bora P, et al. Subcutaneous venous port implantation in adult patients: a single center experience. Diagn Intervent Radiol 2006; 12:93–98. 3. Simpson KR, Hovsepian DM, Picus D. Interventional radiologic placement of chest wall ports: results and complications in 161 consecutive patients. J Vasc Interv Radiol 1997; 8:189 –195. 4. Shetty PC, Mody MK, Kastan DJ, et al. Outcome of 350 implanted chest ports placed by interventional radiologists. J Vasc Interv Radiol 1997; 8:991–995. 5. Morris SL, Jaques PF, Mauro MA. Radiology-assisted placement of implantable subcutaneous infusion ports for longterm venous access. Radiology 1992; 184: 149 –151. 6. Hills JR, Cardella JF, Cardella K, Waybill PN. Experience with 100 consecutive central venous access arm ports placed by interventional radiologists. J Vasc Interv Radiol 1997; 8:983–989. 7. Avastin (bevacizumab) [prescribing information]. San Francisco: Genentech, 2008. Available at http://www.gene.com/ gene/products/information/pdf/avastinprescribing.pdf. Accessed December 30, 2008. 8. Berry S, Kretzschmar MA, Cunningham D, et al. Lack of effect of starting bevacizumab shortly after venous access device implantation on wound healing/bleeding complications: preliminary results from first BEAT. Presented at the American Society of Clinical Oncology Gastrointestinal Cancers Symposium; January 26 –28, 2006; San Francisco, CA. 9. Almhanna K, Pelley R, Thomas Budd G, Davidson J, Moore HCF. Subcutaneous implantable venous access device erosion through the skin in patients treated with anti-vascular endothelial growth factor therapy: a case series. Anti-Cancer Drugs 2008; 19:217–219.