10 Graft Infections


Robert Tisherman, Itamar Neto, Orr Limpisvasti, and Carola F. van Eck

Ligament reconstructions are one of the most frequently performed orthopaedic procedures. Postoperative graft infection, although a rare complication, is one of the most serious complications of ligament reconstructions. Graft infections represent a uniquely challenging situation with the goal of maintaining joint stability while eradicating the infectious process. Intra-articular infections occur following 0.05 to 1.9% of anterior cruciate ligament (ACL) reconstructions and 0.5% of posterior cruciate ligament (PCL) reconstructions, and gram-positive bacteria are typically responsible for infection. Patients typically present with signs and symptoms of septic arthritis during the acute (<2 weeks) period postoperatively, but graft infections have been reported for up to 15 months after ACL reconstructions. Risks for infection following ACL reconstruction include hamstring autograft usage, prior knee surgery, and hemarthrosis. Graft infections typically require multiple surgical debridements and prolonged antibiotic management, adding to the overall healthcare cost. Non-operative and operative measures that preserve the graft tissue have been successful, but removal of the graft and subsequent reimplantation are sometimes necessary. Additionally, the situation of intraoperative graft contamination during ACL reconstruction is discussed and whether a contaminated graft can be safely implanted. This chapter reviews multiple aspects of graft infections including demographics, risk factors, diagnosis, management, complications, and prevention.

10 Graft Infections

Practical Tips

  • Knee aspiration can help to differentiate between superficial postoperative infection and septic arthritis in the early postoperative period.

  • Surgical debridement for septic arthritis should include all prior arthroscopic, meniscal repair, and graft harvest sites, as these can represent niduses for ongoing infection.

  • Arthroscopic debridement for septic arthritis after knee ligament reconstruction has a high rate of successful graft retention.

  • Intraoperative graft contamination can be managed with soaking the graft in 4% chlorhexidine gluconate or polymyxin B–bacitracin solution for 3 minutes without discarding the harvested graft in many cases.

10.1 Introduction

Graft infections represent an uncommon, but potentially devastating, complication of ligament reconstruction surgery. Due to its low incidence, the quality of literature relating to graft infections varies widely. The majority of studies looking at graft infections focus on anterior cruciate ligament (ACL) reconstructions. As the incidence of anterior cruciate ligament (ACL) reconstruction continues to rise, 1 and other ligamentous reconstruction procedures becomes more common, it is becoming more important for surgeons and other clinicians involved in the care of surgical patients to understand the etiology, diagnosis, management, and outcomes of graft infections.

Septic arthritis represents one of the most serious, but well known, complications of ligament reconstruction. Studies looking at patient risk factors, optimal management strategies, and outcomes often focus on the preferred method of the authors and are therefore should be scrutinized for their widespread applicability. Many studies on this subject are also inconclusive due to the low number of patients presenting with postoperative septic arthritis and graft infection.

Septic arthritis in the postoperative period can lead to numerous complications, including permanent cartilage damage, increased risk for graft failure, need for hardware removal, and even death. 2 , 3 , 4 Patients who present with graft infections will typically undergo multiple debridement procedures and require lengthy antibiotic therapy. Overall, this constitutes a large burden on healthcare providers, the healthcare system, and most importantly, the patient.

The objective of this chapter is to review the epidemiology, diagnosis, management, and complications of orthopaedic graft infection.

10.2 Risk Factors for Graft Infection

Various studies looking at the rate of graft infection following ACL surgery have found a rate of postoperative intra-articular infection between 0.23 and 1.9% (▶ Table 10.1). 2 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 The risk factors for generalized orthopaedic infections including age, diabetes, renal disease, immune disorders, and rheumatologic conditions are not routinely seen in the ligament reconstruction patient. The prototypical patient undergoing ligament reconstruction are often younger, more active, and present with fewer medical comorbidities, but several patient and operative factors have been associated in increased risk for graft failure. Patients with morbid obesity, diabetes, and other generalized risk factors for orthopaedic infection do present with low-energy dislocations requiring multi-ligament reconstruction and should be considered high risk for graft infection postoperatively. 27

Table 10.1 Summary of published articles relating to the incidence and patient characteristics of patients presenting with septic arthritis after knee ligament reconstruction. Values presented as median (range)


Number of patients

Septic arthritis – n (incidence)

Graft type (n)

Average age (yrs)

Average time to presentation (days)

Number of surgical procedures (range)

Follow–up (months)

Anterior cruciate ligament reconstruction

Williams et al (1997) 37


7 (0.28%)

Hamstring (3), BPTP (4)

31.3 (17–50)

21.8 (3–79)

1.6 (1–2)

29 (7–71)

McAllister et al (1999) 16


4 (0.48%)

Hamstring (1), BPTB (3)

26 (20–34)

11 (8–18)

1.5 (1–2)

36 (28–42)

Viola et al (2000) 20


14 (0.78%)

BPTB (14)

21 (17–29)

7.7 (2–20)

0.4 (0–1)

14.4 (5–43)

Indelli et al (2002) 21


6 (0.14%)

BPTB (4), Achilles allograft (2)

32.5 (20–51)

20 (9–34)

1.3 (1–3)

36 (24–96)

Schollin–Borg et al (2003) 22


10 (1.7%)

Hamstring (4), BPTB (6)

28.3 (19–39)

15 (4–40)

1.3 (1–3)

35.8 (19–56)

Fong and Tan (2004) 23


7 (1.5%)

Hamstring (7)

23 (19–30)

24 (7–56)

1.4 (1–3)

11.7 (5–26)

Judd et al (2006) 24


11 (0.68%)

Hamstring (11)

28 (22–35)

14 (6–45)

2.4 (2–4)

22 (10–48)

Van Tongel et al (2007) 25


9 (0.52%)

Hamstring (9)

33 (17–50)

10.9 (3–455)

1.9 (1–4)

58 (9–99)

Binnet and Basarir (2007) 26


6 (0.49%)

Hamstring (2), BPTB (4)


22 (14–35)

2.7 (1–5)

102 (30–196)

Wang et al (2009) 6


21 (0.52%)

Hamstring (20), Allograft (1)

28.6 (16–58)

16 (5–32)



Sajovic et al (2009) 7


3 (0.23%)

Hamstring (3)

33 (23–48)

8 (2–14)


33 (4–61)

Monaco et al (2010) 8


12 (1.0%)

Hamstring (12)

24 (16–43)

16 (10–20)

0.3 (0–1)

38 (6–54)

Barker et al (2010) 9


18 (0.58%)

Hamstring (5), BPTB (7), Allograft (6)

34.1 (16–48)

28 (5–122)

1.6 (1–3)


Sonnery–Cottet et al (2011) 2



Hamstring (4), BPTB (7), QT (1)

29.2 (18–49)

16 (2–37)

1.3 (1–2)


Torres–Claramunt et al (2013) 10


15 (1.9%)

Hamstring (13), BPTB (2)




39 (n/a)

Maletis et al (2013) 11


34 (0.32%)

Hamstring (24), Allograft (17), BPTB (10)


20 (12–30)


88% >12 mo

Calvo et al (2014) 12



Hamstring (7)

27.8 (14–51)


n/a (1–4)


Abdel–Aziz et al (2014) 13


24 (0.93%)

Hamstring (24)

26 (19–35)

12 (5–45)

3 (1–6)

59 (18–96)

Bostrom Windhamre et al (2014) 14


43 (0.98%)

Hamstring (27/27)

27 (16–43)

8 (1–22)

3.7 (1–11)

60 (13–108)

Schuster et al (2015) 15


36 (0.51%)

Hamstring (36)

33 (15–55)

17 (4–37)

2.25 (1–6)

56 (8–134)

Murphy et al (2016) 17


55 (0.46%)

Hamstring (36), BPTB (7), Allograft (12)





Bohu et al (2019) 18


5 (0.31%)

Hamstring (5)

36 (20–62)

18 (12–21)

1.6 (1–2)

34 (18–58)

Posterior cruciate ligament reconstruction

Schuster et al (2018) 19


4 (0.46%)

Hamstring (3), Allograft (1)

34.5 (18–47)

18 (7–35)

1.3 (1–2)

16.5 (12–24)

Abbreviation: BPTB, bone patellar-tendon bone.

10.2.1 The Effect of Graft Type on Rates of Postoperative Infection

For any ligament reconstruction, the choice of graft is dependent upon donor site morbidity, suitability of the graft for the desired reconstruction, number of ligaments in need of reconstruction, intraversus extra-articular nature of the ligament and prior surgical history. Donor site morbidity associated with hamstring autograft includes quadriceps weakness, donor site pain, and ecchymosis. Bone-patellar tendon-bone (BPTB) autograft is associated with increased anterior knee pain but provides the benefit of bony incorporation at both ends of the graft. Allograft is available in a wider range of sizes and does not have donor site morbidity. Allograft is widely available in the United States but increases the cost of the operation significantly and has limited availability outside of the United States.

Therefore, autograft is routinely chosen for ACL reconstruction because it is native tissue and does not have the potential disease transmission or rejection associated with allograft, and offers a cost-effective alternative to allograft. 28 Autograft has also been shown to incorporate in ACL reconstruction in less time than allograft. 29 Currently, the most widely used autografts are hamstring, BPTB, and quadriceps tendon. 27

The choice of donor graft is important when considering the risk of potential infection and is one of the strongest factors found across all studies. Hamstring autografts have consistently exhibited increased risk for postoperative infection compared to BPTB autograft and allograft, 7 , 9 , 17 , 30 as the relative risk of infection when using hamstring autograft for ACL reconstruction is 3.3 to 4.3 compared to BPTB. 9 , 30 Hamstring autograft has been shown to have a high risk for infection prior to transplantation, as 16 to 22% of hamstring autografts are culture-positive at the time of harvest, indicating that graft harvest and preparation are the likely source for introduction of inoculating bacterium. 31 It has been proposed that hamstring autograft has a higher rate of postoperative infection due to the tissue dissection proximity to the tibial tunnel and the possibility of hematoma formation that can extend intra-articularly. Multiple studies have shown no difference in deep infection rate when comparing BPTB to allograft (BPTB or Achilles). 9 , 30

Allograft processing and contamination are discussed later in this chapter, but there is no increased risk of graft infection with allograft despite its nonsterile harvest, avascular nature, and longer ligamentization time when compared with autograft. A combined prospective and retrospective multicenter cohort study of 1,298 ACL reconstruction patients with 74.3% allografts demonstrated no cases of septic arthritis, a superficial infection rate of 2.3%, and no increased risk of clinical infection with the use of allografts. 32 Several widely publicized cases of bacterial contamination and death following ligament reconstruction have occurred due to graft infections from allograft tissue in Minnesota, 33 Florida, and Louisiana 3 due to ineffective terminal sterilization. Allografts were culture positive 7.9% of the time for bacterial contamination, with no association seen between culture-positive allografts and clinical infections. 34 With regard to bacterial infections in allografts, Clostridium spp. (37.5% Clostridium sordelli 7 ) were liable for roughly 50% of cases. 35 Unlike autograft infections, gram-negative bacilli such as Pseudomonas aeruginosa, Serratia liquefaciens, and Escherichia coli as well as fungal infections (Candida sp.) have been seen at higher rates with allograft transplants. 35

Additional risk factors include previous knee surgery, 18 hospital admission following surgery, 36 and development of hemarthrosis in the immediate postoperative period. 18 In a review of the Multicenter Orthopaedic Outcomes Network (MOON) study database, which contained 17 patients (0.8% of the entire cohort) who presented with septic arthritis following ACL reconstruction, it was found that diabetes was a significant risk factor for graft infection. 27

There is conflicting evidence that ACL reconstruction in professional athletes has a higher rate for septic arthritis. One study found that rates of postoperative infection in professional athletes may be as high as 5.7%, 2 while other studies have shown no difference between nonprofessional athletes and professional athletes. 18

10.3 Clinical Presentation and Management

10.3.1 Clinical Presentation

In the early postoperative period, patients present with signs and symptoms of septic arthritis at an average of 18 days postoperatively (▶ Table 10.1). The most common presenting symptoms are fevers, pain, and effusion. 9 , 22 , 37 Other common symptoms include erythema, local knee joint drainage at the site of surgical incisions, and progressive knee pain (▶ Fig. 10.1). 18 , 23 While most patients present within the acute (<2 week) or subacute phase, there have been reports of graft infection occurring more than 1 year postoperatively, and the development of symptoms long time after the surgery should not be used to rule out septic arthritis. 25

Fig. 10.1 A 43-year-old female patient 6 months postoperatively from anterior cruciate ligament, posterolateral corner, and medial collateral ligament reconstruction using allograft who presents with new drainage from the tibial incision and surrounding erythema.

10.3.2 Laboratory Evaluation

Clinical examination of postoperative knee pain, swelling, and low-grade fevers can have a wide differential diagnosis, including hemarthrosis, tissue response to surgery, and superficial infection. Septic arthritis is often missed at the initial visit where signs and symptoms may be present, due to the overlap between postoperative healing response and acute intra-articular infection. 22 Confirming the presence of graft infection in the postoperative period is crucial for expedient management and prevention of long-term complications. Laboratory evaluation of joints with suspected graft infection should include serum inflammatory marker levels, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP), as well as serum white blood cell counts. Blood cultures are negative in the majority of cases of confirmed septic arthritis. 20

ESR and CRP are acute inflammatory markers that can help to differentiate postoperative septic arthritis from normal postoperative healing response. CRP is elevated within 12 to 24 hours of the onset of infection and ESR shows elevation 24 to 48 hours after the onset of infection. 38 A retrospective review of ESR and CRP in noninfected and septic knees following ACL reconstruction found that the optimal ESR and CRP cutoffs for septic arthritis were 32 mm/hour and 41 mg/L, respectively. 39 These ESR and CRP cutoffs provided a sensitivity of 91.2 and 94.1% and specificity of 80.5 and 97.6%, respectively. 39

Knee aspiration should be performed as part of the standard septic arthritis workup through superolateral aspiration under aseptic technique to avoid contamination at the portal holes. Knee aspiration revealed average leukocyte count in excess of 50,000 cells/mL in most cases, 9 , 21 , 25 but a high index of suspicion should exist for septic arthritis with aspirate cells counts over 20,000 cells/mL and a polymorphonuclear cell percentage >75%. 40 Some experts have recommended a lower threshold of >10,000 cells/mL from postoperative knee aspirations in patients presenting with signs and symptoms consistent with septic arthritis. 24

Laboratory evaluation can help to confirm when graft infections do occur and prevent unnecessary antibiotics and exploratory surgery.

Only gold members can continue reading. Log In or Register to continue

Jun 5, 2021 | Posted by in ORTHOPEDIC | Comments Off on 10 Graft Infections
Premium Wordpress Themes by UFO Themes