Penile Prosthesis Infection: Approaches to Prevention and Treatment

Epidemiologic studies have estimated that more than 50% of men ages 40 to 70 have some form of erectile dysfunction. Penile prosthesis implantation remains a mainstay for treatment of erectile dysfunction unresponsive to other less-invasive methods. Improvements in penile prosthesis design have extended the long-term survival of implants. As the improved design of prostheses has led to their increased mechanical survival, other complications, such as infection, have emerged as the leading causes of implant failure. This article focuses on approaches to prevention and treatment of penile prosthesis infection.

Epidemiologic studies have estimated that more than 50% of men ages 40 to 70 have some form of erectile dysfunction, with nearly half of men ages 60 to 70 having moderate to severe erectile dysfunction. Among 40- to 69-year-old white men, it is estimated that nearly 620,000 will develop erectile dysfunction each year in the United States. Penile prosthesis implantation remains a mainstay for treatment of erectile dysfunction unresponsive to other less-invasive methods. An estimate of the number of penile prostheses implanted each year in the United States approaches 15,000 devices. Since the introduction of the inflatable penile prosthesis (IPP) by Scott and colleagues in 1973, inflatable prostheses have become the most commonly placed implant, and improvements in penile prosthesis design have extended the long-term survival of implants, with 5-year mechanical survival estimates ranging from 86% to 96% and longer-term studies suggesting 10-year mechanical survival of 67% to 88%. As the improved design of prostheses has led to their increased mechanical survival, other complications, such as infection, have emerged as the leading causes of implant failure. This article focuses on approaches to prevention and treatment of penile prosthesis infection.

Pathogens and risk factors for infection

Infection rates for virgin IPPs have typically been approximately 1% to 3%, but published rates have been significantly higher in revision surgery or when reconstructive procedures are involved. Compared with a 1.8% risk of infection in virgin penile prosthesis implantations, Jarow noted a 13.3% risk of infection after revision surgery for an uninfected malfunctioning penile prosthesis and a 21.7% risk if a penile reconstructive procedure was involved. The only significant difference between the three groups was mean operating time (98, 110, and 255 minutes for the primary, revision, and reconstructive groups, respectively), so the length of the operation was thought to have played a role in infection risk. Other studies have shown the risk for infection in revision surgery to range from 6.7% to 18.8%. The higher risk of infection during revision surgery is thought related to factors that impair host resistance to infection, including poor antibiotic penetration due to capsule formation around the previous implant and poor wound healing.

The source for infection is usually skin flora introduced at the time of the operation. The most common offending organism is Staphylococcus epidermidis in both primary and revision surgery, although gram-negative fecal pathogens can cause infection. Other organisms implicated include Pseudomonas aeruginosa , Escherichia coli , Serratia marcescens , Enterococcus species, Proteus mirabilis , and methicillin-resistant Staphylococcus aureus (MRSA). Several studies have suggested that the higher rate of infection during revision surgery may be explained by the fact that even clinically uninfected prostheses are often colonized by bacteria. This was suggested by Licht and colleagues, who noted that 40% of clinically uninfected penile prostheses were culture positive at the time of revision. Although three of the patients with positive cultures went on to develop infection requiring prosthesis removal, none of the patients with negative cultures developed infection. Henry and colleagues published a multicenter study in which 54 of 77 (70%) clinically uninfected patients had a culture-positive penile prosthesis at the time of revision surgery. Furthermore, of the 48 patients who had time to revision data culture results and underwent revision due to mechanical failure, those who had positive cultures had significantly shorter revision-free times: 6.3 years for positive cultures versus 8.9 years for negative cultures. The reason for this finding is unclear, but it suggests colonization may play a role in mechanical failure.

Persistent bacterial adherence to implanted medical devices is mediated by bacterial production of biofilm, a survival mechanism by which bacteria produce a slimy matrix that reduces phagocytosis, prevents antibiotic diffusion, and traps nutrients. The National Institutes of Health estimates that 80% of all infections are caused by biofilms. Henry and colleagues noted the visible presence of biofilm on some of the clinically uninfected prostheses removed in their multicenter study, and, using laser microscopy, Silverstein’s group noted biofilm on 8 of 10 clinically uninfected prosthesis removed for mechanical failure. They also suggested that it is likely that essentially all implants have some biofilm matrix formation. A patient begins to experience symptoms of infection when the bacteria in the biofilm propagate and become free-floating or planktonic, and currently, the only mechanism to eradicate biofilms is by removing the prosthetic. Antibiotics are able to suppress these planktonic bacteria, but the bacteria imbedded in the biofilm remain untouched. It is postulated that revision surgery could be a stimulus that activates the bacteria in these biofilms to become planktonic, and this theory has been the basis for the salvage protocols developed by surgeons. Future directions in preventing and treating prosthetic infections may involve disrupting the communication of bacteria in biofilms and administration of antibiotics concomitantly with drugs that can disperse biofilms into a planktonic state.

Other documented risk factors for infection include diabetes mellitus, spinal cord injury, urinary tract infection, immunosuppression, and distant sites of infection. The literature on the risk of infection in diabetic patients is mixed. Fallon and Ghanem had a 3-fold higher incidence of infection in their diabetic population. Bishop and colleagues suggested that elevated hemoglobin A1c (HgbA1c) levels in diabetics may predict a higher incidence of infection in this population. In a series of 90 patients in which 5 of 32 diabetics had postoperative infections, diabetics with an HgbA1c of greater than 11.5% had a 31% risk of infection compared with 5% in patients with an HgbA1c less than 11.5%. The investigators suggested better diabetes control preoperatively might help lower the risk of infection. In response to the Bishop and colleagues’ findings, one group noted in their series of 389 implants that diabetics had an 8.7% infection rate compared with 4.0% in nondiabetics, but there was no difference in the mean or median HgbA1c levels in the infected versus uninfected diabetic. Although the study suggested diabetics had higher rates of infection, they concluded that neither glycosylated hemoglobin levels, fasting sugar levels, nor insulin dependence increased the risk of infection in diabetics undergoing prosthesis implantation. Conversely, in Jarow’s series, diabetics had no increased incidence of infection in either the primary or revision implant group. In a series of 1251 operations, Wilson and Delk noted no significant increase in infections in diabetics receiving an initial implant. Other studies have shown that usage of antibiotic-coated penile prostheses can reduce the risk of infection in diabetics to 0% to 1%.

Spinal cord injury has been documented in numerous studies to increase the risk of IPP infection, although others have shown no increased incidence of infection in the spinal cord injury population. The spinal cord–injured population has unique risk factors, including a higher rate of urinary tract infections and decreased sensation that predisposes to implant erosion. Literature on studies of immunosuppressed patients has been mixed as well. In Wilson and Delk’s large series, 5 of 10 patients on chronic steroids developed infections; however, in a series of 13 diabetic patients on immunosuppression after organ transplantation who received a penile prosthesis for impotence, there was no increased incidence of infection. Cuellar and Sklar found no increased risk of infection in their series of organ transplant recipients who had penile implants placed. Barry suggested transplant recipients meet certain criteria before penile prosthesis implantation, including stable graft function for greater than 6 months, avoiding an intra-abdominal reservoir, and low-dose immunosuppression. Postorgan transplant patients with erectile dysfunction should appropriately be offered a penile prosthesis without undue risk for infection.

Prolonged hospitalization is a risk factor for infection due to skin flora changes to more virulent organisms; thus, same-day admission is advantageous. Others have shown that late hematogenous spread is possible, and one case of a Salmonella infection of an IPP in a renal transplant recipient has been reported. Although the risk of late spread should be considered rare, ruling out distant infection should be part of the workup before performing surgery.

Most infections of penile prostheses are caused by skin flora introduced at the time of surgery. Spinal cord injury, poorly controlled diabetes, distant active infections, prolonged hospitalization, and prior implantation can increase the risk of developing infection. Although these factors may increase the risk of infection in certain subsets of patients, appropriate perioperative planning and operative technique can minimize these risks.

Surgical planning

Having discussed preoperative risk factors associated with infection, other perioperative considerations are important for reducing the risk of infection. Patients should be assessed for genital skin lesions preoperatively to rule out active infections, and remote infections should be eliminated. Sterile urine cultures are an imperative measure to reduce the risk of infection from urine spillage, and placing a urethral catheter is advisable in patients with neurogenic bladder. Same-day admission is warranted to reduce the risk of skin flora changes to more virulent and resistant organisms. Increased risk of infection from performing simultaneous procedures at the time as IPP placement should be considered. Some investigators have suggested simultaneous circumcision and IPP increases risk of infection, whereas other investigators have suggested that closing the incisions related to the implant before performing concomitant minor genital procedures does not increase risk. Carson and Noh documented an increased risk of infection when additional foreign bodies, such as Gore-Tex or Dacron, were implanted at the same time as a penile prosthesis, whereas Jarow found a 21.7% incidence of infection when reconstructive procedures were involved. Simultaneous placement of an IPP and artificial urethral sphincter has been suggested to decrease infection risk due to the decrease in operative time and avoidance of reoperation in a surgical field that may have an asymptomatic colonized prosthetic. Thus, if multiple procedures are being considered, the benefit of a single course of anesthesia must be weighed against the possible increased risk of infection in some settings.

Preoperative bathing with an antibacterial shower has been suggested by several investigators to reduce risk of infection. Bathing with chlorhexidine has been shown to decrease colony counts on skin but has not been shown to reduce surgical site infection rates. At the authors’ institution, any preoperative antiseptic bath is not used for penile prosthesis patients. On the contrary, skin preparation with antimicrobial agents has been shown to reduce the incidence of surgical site infections. A recent prospective, randomized study showed that chlorhexidine-alcohol–based preparations were superior to povidone-iodine–based preparations in preventing surgical site infections in clean-contaminated surgery, and a meta-analysis by Noorani and colleagues noted similar findings. Consequently, the authors’ institution has mandated the use of chlorhexidine-alcohol–based solutions for preoperative surgical skin preparation. An no-touch technique devised by Siegrist and colleagues involves placing a drape over the incision that limits the contact between the skin and surgical field. The rate of infection in the no-touch group was 0.73% compared with 2.23% in the standard technique group.

Various studies have shown that clipping hair in the operating room significantly reduces the risk of infection compared with shaving the night before, and current recommendations are to perform clipping, if needed, immediately before the operation. The Sexual Medicine Society of North America (SMSNA) maintains that surgeons should be given the option of shaving with a razor as opposed to the more ubiquitous clippers mandated by most hospitals, because there is less tissue trauma with no increased risk of infection. No matter the method chosen, it should be performed immediately preceding the operation.

With the increasing rates of MRSA isolates in hospitals, preoperative measures to reduce MRSA colonization have been suggested. Prospective, randomized, controlled trials have shown no benefit to preoperative decolonization with mupirocin in unselected patients. Although current recommendations do not advise decolonization of unselected patients with mupirocin nasal ointment, recent studies in the cardiothoracic and orthopedic literature have shown a reduction in the rate of MRSA surgical site infections when mupirocin ointment applied to the nares was part of a regimen designed to decrease the risk of these infections. The generalizability of these results is unclear, and further study of this topic is needed before definitive recommendations can be made. Other mechanisms to minimize the risk of postoperative infection include limiting operating traffic, appropriate hemostasis of bleeding tissues, avoiding flashing of instruments, maintaining postoperative glucose levels below 200, and preventing intraoperative hypothermia.

Preoperative administration of prophylactic antibiotics is recommended by the American Urological Association for open procedures involving the placement of prosthetic implants. Antimicrobials of choice include an aminoglycoside plus a first-generation or second-generation cephalosporin or vancomycin administered within 1 to 2 hours before incision and given for a duration of 24 hours or less. This regimen targets the commonly implicated gram-positive skin flora and urinary tract pathogens most likely to cause infection. Although there are no recommendations on the practice, most urologists discharge postoperative IPP patients with oral prophylactic antibiotics. One abstract looking at practice patterns among urologists who perform penile prosthesis surgery noted that 94% of SMSNA urologists prescribe postoperative home oral antibiotics compared with 88% of non-SMSNA members. The same survey noted that 100% of SMSNA members and 92% of non-SMSNA urologists irrigate intraoperatively with antibiotic solution. Although it may seem prudent to irrigate with antibiotic solution during primary implantation, there are currently no recommendations on the practice to guide urologists. For revision surgery due to noninfectious reasons, a group of urologists at three institutions recommended using an antibiotic washout procedure because they found a rate of infection of 2.86% for the washout group versus 11.6% for the nonwashout group. The same group also noted that the rate of positive tissue capsule cultures performed during revision surgery decreased from 43% to 25% after the washout. Although this group used antibiotic irrigation, they postulated that the most important part of the washout may be the mechanical débridement of the biofilm, and some of the agents used in the washout (eg, hydrogen peroxide) may actually increase the risk of infection. A study of washout procedures comparing antibiotic irrigation with saline irrigation may shed some light on the issue and potentially provide urologists with guidelines on whether or not to irrigate with antibiotics. Caution should be advised when implanting a prosthesis coated with antibiotics because copious irrigation may cause elution of the antibiotics before insertion.

Although meticulous sterile technique is critical to reducing the risk of prosthetic infection, recent advances in prosthetic technology have further decreased the incidence of implant infection. American Medical Systems (AMS) (Minnetonka, MN, USA) and Coloplast Corporation (Minneapolis, MN, USA) each manufacture three-component prostheses that allow incorporation of antibiotics into the implant. The AMS 700 Series of IPPs is impregnated with InhibiZone, a coating on the external surface of the device that elutes rifampin and minocycline to create a zone of inhibition for bacterial growth. The amount of drug is less than a single oral dose, with a large drop-off in drug levels occurring at approximately 7 days. Coloplast manufactures the Titan IPP, which has a hydrophilic polyvinylpyrrolidone coating that can be soaked in antibiotics intraoperatively. Surgeons have the advantage of choosing the antibiotics in which to soak the Titan, and these then diffuse into the tissue around the implant to neutralize bacteria.

Many studies have documented the decreased incidence of infection associated with these implants. In 2004, Carson published retrospective data on 2261 men who received InhibiZone prostheses compared with 1944 men who received untreated prostheses; the InhibiZone group had an 82.4% reduction in infection at 60 days and a 57.8% reduction in infection at 180 days. Wilson and colleagues published data on InhibiZone-coated prostheses, showing a reduction in the rate of infection in virgin nondiabetic, virgin diabetic, and revision implants with washout compared with their historical rates. Furthermore, of the 223 nondiabetic patients who received virgin implants, none developed infection. In a similar study, a second group also noted no infections in 58 patients (0%) who had InhibiZone-coated prostheses placed compared with 3.2% (3/94) of patients who had uncoated prostheses. A retrospective study of patient information forms filed with the AMS covering 7 years and more than 40,000 IPPs (36,659 InhibiZone prostheses and 3456 noncoated prostheses) showed an infection rate of 1.77% for InhibiZone-coated prostheses and 3.09% for noncoated prostheses at 84 months. A study of 2357 implants of the Coloplast Titan and 482 Alpha-1 IPPs (Coloplast Corporation, Minneapolis, MN, USA) showed an infection rate of 1.06% in the antibiotic-coated Titan compared with 2.07% in the noncoated Alpha-1. An even larger study of 17,900 Titan implants showed an infection rate of 1.6% compared with 8825 Alpha-1 implants with an infection rate of 4%. Dhabuwala and colleagues recently compared the antibiotic-soaked Coloplast Titan to the InhibiZone-coated AMS prosthesis. The rates of infection for the vancomycin/gentamicin-coated Titan, the rifampin/gentamicin-coated Titan, and InhibiZone-coated AMS were 4.4%, 0%, and 1.3%, respectively. The investigators suggested that the infection rate between the rifampin/gentamicin-coated Titan and the AMS InhibiZone models were comparable and that all Titan prostheses should be coated with rifampin/gentamicin solution. Wilson recently presented an abstract in which he prefers coating the Titan with Bactrim due to its low cost, broad-spectrum activity, and ease of handling. Revision implants have shown similar trends. In one study of 55 patients who underwent revision surgery for mechanical failure and had an InhibiZone-coated prosthesis implanted, only one (1.8%) developed infection compared with the historical standard of 10%.

A final consideration to prevent complications is whether or not to place a drain, which some investigators have argued increases the risk of infection. Sadeghi-Najed and colleagues retrospectively reviewed data on 425 consecutive nonantibiotic-coated implants placed at three separate institutions. Each patient had a closed-suction drain placed for 12 to 24 hours postoperatively and the overall infection rate was 3.3%, which was similar to published historical rates. Garber used a closed-suction drain in 50 patients for 24 hours postoperatively and had one periprosthetic infection. Using a closed-suction drain does not seem to increase the risk of infection, and advocates recommend placing the drain through a separate incision and not the surgical incision to avoid bacterial migration into the dissection site.