Early periprosthetic infections (generally defined as those occurring within the first 3 months after surgery) are due to organisms that are inoculated during surgery or in the postoperative window prior to skin and soft tissue healing. Prolonged drainage, hematoma, and seroma may contribute to this risk by impairing skin closure and facilitating the introduction of organisms from the cutaneous microbiome. Although the likelihood of such infection is reduced by surgical antibiotic prophylaxis and by optimal cutaneous antisepsis, some organisms survive and are not effectively cleared by surgical antisepsis.¹⁰ Further, organisms may be inoculated in the perioperative window after surgical antimicrobial prophylaxis has completed. Most early periprosthetic infections are caused by virulent organisms, as these organisms replicate more rapidly and induce a more robust inflammatory response, leading to their manifesting earlier. For example, S. aureus comprises over 50% of early periprosthetic knee infections³; other organisms include streptococci, Gram-negative bacilli, and enterococci. Notably, the cutaneous microbiome varies between different regions of the body.¹¹ As early postoperative infections are primarily due to organisms that reside at the surgical site, the microbiology of early infections differs between hip and knee arthroplasty infections. Infections due to organisms that comprise enteric flora are more common on perineal skin than about the knee; these organisms (including Gram-negative bacilli and enterococci) are therefore seen less commonly in the knee than the hip in early infection.¹ Notably, the type of antibiotic prophylaxis may influence the types of organisms seen in early PJIs. In one series, patients who received vancomycin for surgical antimicrobial prophylaxis were more likely to suffer infection due to Gram-negative organisms than those who received cefazolin; however, they were less likely to develop infection due to MRSA.¹²

Delayed periprosthetic infections may also be caused by organisms introduced in the perioperative window but present in a more subacute to chronic fashion. These infections are caused by organisms that grow more slowly and are of lower virulence and are therefore less likely to be identified in the early perioperative window. Patients
with infections caused by these organisms often achieve an early clinical response after arthroplasty, before presenting with increasing pain 3 to 6 months after surgery. While virulent organisms such as S. aureus and Staphylococcus lugdunensis can present this way, in many studies nonvirulent organisms including coagulase-negative staphylococci are more common. Gram-negative organisms are less common causes of delayed arthroplasty infections.³ One way that organisms causing chronic periprosthetic infection are able to survive the host immune response is by producing biofilm. Biofilms are comprised of clusters of microorganisms that adhere to the surface of the device and elaborate a polymer matrix.¹³ Organisms within biofilms often replicate more slowly and are protected from the host immune response and antimicrobial therapies by mechanical and electrostatic barriers. Due both to their growth characteristics and the impenetrability of the biofilm matrix, sessile organisms within biofilms may be 10 to >500 times less susceptible to antimicrobials than their planktonic counterparts (those organisms that have left the biofilm).¹⁴

Late infections are primarily due to hematogenous spread of organisms, i.e., due to virulent organisms that gain access to the bloodstream. Organisms can translocate into the bloodstream through weaknesses or loss of integrity of epidermal and mucosal surfaces, including breaks in the skin, oropharyngeal and nasopharyngeal mucosa, bladder endothelium, and gut mucosal surfaces. The types of organisms causing hematogenous infection may be inferred based on the portal of entry. Organisms that enter through breaks in the skin are those that reside on the skin surface, including S. aureus, S. lugdunensis, Streptococcus agalactiae (Group B streptococcus) and pyogenic streptococci such as Streptococcus pyogenes (Group A), and Streptococcus dysgalactiae (Groups C/G). These may be more common in the setting of cutaneous diseases like eczema, psoriasis, and ulcers caused by venous and/or arterial insufficiency. Overall, compared with early and delayed infections, streptococcal species are more common causes of hematogenous PJI, in some series causing up to 37% of late hematogenous PJI.¹⁵ Similarly, coagulase-negative staphylococci are less common causes of late hematogenous infection.⁵,¹⁶

Antecedent dental infection and gingivitis may contribute to infection with viridans streptococci and other mouth flora, including organisms collectively grouped by the HACEK acronym (Haemophilus aphrophilus, Actinobacillus actinomycetomcomitans, Cardiobacterium hominis, Eikenella corrodens, and Kingella kingae). Urinary tract infection and bacteriuria are a risk for later PJI.¹⁷ Although still less common than Gram-positive organisms as a cause of PJI, Gram-negative infections are more commonly seen in patients with a history of bacteriuria.¹⁸

Injection drug use is unfortunately an increasing source of hematogenous infection, though still not as common among patients who have undergone arthroplasty. Bacteria may be introduced because of inadequate skin preparation (e.g., skin organisms such as S. aureus and S. pyogenes) because of the practice of needle-licking (oral streptococci and oral anaerobes such as E. corrodens) or because of contamination of the drug itself. Other organisms seen more commonly in the setting of injection drug use include Pseudomonas aeruginosa (because of the use of tap water for preparation) and Candida species.¹⁹

Rarely, the organisms causing late periprosthetic infections may be introduced directly into the joint, primarily in the setting of trauma. In addition to the organisms that reside on the host’s skin, organisms associated with traumatic inoculation include environmental organisms: Gram-negative bacilli, mycobacteria, fungal species, and environmental anaerobes such as Clostridia. Organisms that reside within water sources (e.g., P. aeruginosa and other nonenteric Gram-negative bacilli, nontuberculous mycobacteria) may be seen after marine trauma.

With advances in microbiologic diagnostic techniques, the spectrum of organisms capable of causing periprosthetic infection is expanding. While still rare, fungal and mycobacterial infections are increasingly being reported.²⁰,²¹,²²,²³ These organisms should be considered in the setting of prior antibiotic exposure (Candida), injection
drug use (Candida), and environmental trauma. As diagnostic techniques improve, many additional organisms that do not grow well in culture and were not previously thought to be pathogens in the setting of arthroplasty (e.g., Mycoplasma and Lyme) are being identified.²⁴,²⁵,²⁶ We anticipate that the spectrum of organisms capable of causing PJI will continue to expand as our diagnostic modalities improve.

Despite advancements in microbiologic diagnostics, the organisms causing some periprosthetic knee infections remain elusive. Depending on the study, between 7% and 25% of infections are culture-negative.¹,⁴,⁷,⁸ Culture negativity may be more common in chronic biofilm-associated infections, as the organisms may be encapsulated within the biofilm matrix and less adapted to growth in culture media.¹³ Prior antibiotic exposure (defined as antibiotic treatment within 3 months) increases the likelihood that an infection may be culture-negative.²⁷,²⁸ Fastidious organisms or those that require special growth media (e.g., nutritionally deficient streptococci, HACEK organisms) are often implicated as causes of culture negativity. Mycoplasma, fungi, and mycobacteria do not always grow in routine bacterial cultures and therefore suspicion of their presence is important to ensure that the correct culture media are utilized. Newer non-culture-based diagnostic modalities (discussed later in the chapter) may ultimately reduce the proportion of infections for which the causative pathogen is unknown.

With the expanding use of antimicrobial therapies worldwide, more infections are being caused by antibiotic-resistant organisms, including MRSA and methicillin-resistant coagulase-negative staphylococci, ampicillin- and vancomycin-resistant enterococci, and Gram-negative organisms with extended-spectrum beta-lactamases (ESBL). Some studies have also identified an increasing prevalence of resistant organisms causing PJI over time.¹,⁸ Predicting the likelihood of resistance is important in empiric antimicrobial management and also in considering different surgical strategies, as resistant organisms have been more commonly associated with treatment failure.²⁹,³⁰ Patients with prior antimicrobial exposure and those with comorbidities may be more likely to harbor resistant organisms.³¹ The prevalence of certain resistant organisms differs geographically and institutionally as well: MRSA is more commonly found as a cause of PJI in the United States than at European Centers.⁷ As clinicians, understanding the local microbiology is paramount.