Medical complications are uncommon following anterior cruciate ligament reconstruction (ACLR), yet their impact is not insignificant because of the frequency with which this procedure is performed. Further, these adverse events involve significant patient morbidity and represent a financial burden of over $1 billion annually, underscoring the importance of adequate management and prevention. This chapter will focus on the two medical complications most frequently encountered following ACLR: septic arthritis and venous thromboembolism. Although the relatively low incidence of these events precludes the establishment of robust clinical guidelines, the current literature allows for conclusions to be drawn regarding appropriate practices for diagnosis, treatment, and prevention.
The incidence of septic arthritis following ACLR is low, with accounts in the literature placing the incidence rate at 0.15% to 0.98%. Yet, septic arthritis has been reported as the third most common reason for reoperation after ACLR, accounting for approximately 15% of returns to the operating room. ACLR procedures involve several factors that may increase the risk of septic arthritis compared with simple arthroscopy. Drilling of bone tunnels and passage of a graft requires extensive soft tissue dissection, creating surgical sites that are easily contaminated. The incision for the tibial tunnel allows hematoma formation in the pretibial subcutaneous tissue, making it a particularly common site for infection. Although superficial infections are typically benign, newly formed bone tunnels may provide a conduit to the intraarticular space. , , , Furthermore, ACLR involves a high foreign body load. Graft tissue, suture material, and interference screws can all serve as a nidus for glycocalyx formation. , ,
Recent studies demonstrate lower rates than those reported historically, indicating that our ability to prevent postoperative infection has improved. Despite this, the implications of knee septic arthritis cannot be underemphasized. Bacterial toxins are destructive, and can destroy greater than half the glycosaminoglycan content of articular cartilage within one week. , , Consequences range from chondral thinning to full-thickness lesions, accelerating the development of osteoarthritis. , , , , Toxins may also contribute to arthrofibrosis, meniscal tears, and graft failure. If not addressed appropriately, the resulting joint dysfunction and instability can prolong recovery and lead to poor outcomes. , , ,
Diabetes mellitus appears to be the patient-specific factor that confers the highest relative risk for infection. Although the prevalence of diabetes is low among the population undergoing ACLR, these patients have almost twenty times the odds of infection relative to nondiabetics, and patients should be counseled appropriately. Tobacco use, known for its broad detrimental effects on tissue healing, was recently identified as an independent risk factor for infection after ACLR. Patients with a prior intraarticular surgery of the ipsilateral knee, as well as those requiring concomitant procedures at the time of ACLR, are also at increased risk of infection. This is likely influenced by a combination of factors, including prolonged operative and tourniquet time, additional or larger incisions, and increased foreign-body loads. , , , , , Interestingly, status as a professional athlete is associated with an increased risk of infection compared with amateurs. This may be explained by a greater exposure to infective organisms and a higher likelihood of undergoing more extensive procedures in this population.
Surgeon-Dependent Risk Factors
The primary surgeon-dependent factor associated with infection following ACLR is the choice of graft material for reconstruction. Bone-patellar tendon-bone (BPTB) autograft carries a low incidence of postoperative infection. , Hamstring tendon, the other primary source of autograft, has been shown by multiple authors to be associated with an increased risk of infection after ACLR compared with other graft types. , , , , , Although the reason for this disparity is unclear, several possible explanations exist. Soft tissue dissection during hamstring tendon harvest may increase the likelihood of hematoma formation, facilitating bacterial growth near the tibial tunnel. Further, additional time required for hamstring graft harvest and preparation may increase exposure to bacterial contamination. , Others have postulated that the lack of bone and presence of suture material makes hamstring grafts more susceptible to infection. , Although initial studies comparing autograft and allograft tissue showed conflicting results, a recent metaanalysis reported no difference in the incidence of infection between these two groups. Among allograft choices, there is also no difference between processed and nonprocessed grafts.
A higher incidence of infection has been reported in cases using flash sterilization compared with standard methods. Although the difference was not statistically significant, this topic warrants mention. Flash sterilization meets only minimum sterilization standards, and flash-sterilized instruments can easily be contaminated during transport because they are not enclosed in sterile packaging.
Many patients who develop an intraarticular infection following ACLR will present with acutely worsening pain, effusion, erythema, warmth, and fever. , , , , , , , However, the clinical picture is not always clear. The majority of patients with septic arthritis will present within 1 to 2 weeks postoperatively, , , , , , blurring the distinction between symptoms caused by an inflammatory process and those attributable to the normal postoperative course. , , , A smaller subset will present subacutely during the first two months. In these cases, the pathogen is often more indolent, producing symptoms that may be less obvious to the patient and examiner. , , , , Further, erythema and drainage occur with approximately equal frequency in septic arthritis as with extraarticular processes, such as superficial infection or hematoma formation. , , This overlap of symptoms can make the clinical diagnosis challenging. In fact, as many as 50% of cases may be missed at the initial visit. Thus it is imperative for the surgeon to be discerning in the postoperative evaluation, with a low threshold for suspicion of intraarticular infection. Several factors in particular may help distinguish septic arthritis from a more benign process. Fever, while present in less than 50% of cases of septic arthritis, should always alert the surgeon to the likelihood of a septic etiology. In additional, worsening pain and new-onset effusion should raise concern for a more serious underlying cause. , ,
Serum studies typically include a complete blood count with differential, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP). A white blood cell (WBC) count greater than 10,000 may indicate the presence of infection; however, this marker is abnormally elevated in only half of infected cases. , , , , , In addition, the WBC count may be elevated as part of the expected postoperative course, or in cases of superficial infection. An elevated ESR indicates an increase in fibrinogen levels, which occurs 1 to 2 days after the onset of infection. Although the ESR may be elevated in anywhere from 50% to 100% of infected cases, , , , , , increased values are expected as part of the normal postoperative course secondary to surgical trauma. , , , Further, the ESR failed to show significant association with septic arthritis in a general patient population. Thus it not surprising that many surgeons find this marker to be of little use. In contrast, CRP is an acute-phase protein, which serves as a rapid indicator of inflammation and tissue necrosis. Mediated by cytokines, CRP elevations occur within 6 to 8 hours of tissue insult. Although variable CRP elevations are expected in the postoperative period, , , , , this marker returns to normal values within 7 to 10 days postoperatively. Thus CRP elevations—especially those markedly above normal cutoff values—are considered more useful in the diagnosis of septic arthritis than elevations in ESR. , , , , , A CRP threshold of 41 mg/L has optimal sensitivity and specificity, reported at 94.1% and 97.6%, respectively. When interpreting postoperative CRP values, the practitioner should note that males, patients operated on by less experienced surgeons, and those undergoing concomitant chondral surgery may demonstrate higher values without an infectious source.
Synovial Fluid Studies
Joint aspiration and synovial fluid analysis are essential in the diagnostic evaluation of septic arthritis. Synovial fluid WBC count and differential is the most useful study in predicting septic arthritis. A WBC count greater than 50,000 with a differential showing greater than 75% polymorphonuclear (PMN) cells is commonly accepted as diagnostic. Successively higher WBC counts and PMN percentages are associated with increasing likelihood of septic arthritis.
Microbiology studies of synovial fluid provide more utility in guiding treatment than in establishing the diagnosis of septic arthritis. Gram stain of the synovial fluid yields minimal diagnostic information, with a sensitivity ranging from 29% to 50%. Although the sensitivity of culture results is higher (82% on average), rates of positive cultures in cases of confirmed septic arthritis can be as low as 14%. This may be attributed to the close temporal relation of culture attainment with preoperative antibiotic dosing, or to the reflex administration of therapeutic antibiotics upon suspicion of septic arthritis. In addition, coagulase-negative Staphylococcus species, a common cause of septic arthritis after ACLR, can be difficult to detect on culture media. , , , In light of this, it is advisable to send multiple cultures and communicate to the laboratory that sparse growth of a coagulase- negative Staphylococcus species should not be dismissed. ,
The organisms involved in cases of septic arthritis following ACLR are typically skin flora. Although reports in the literature vary, coagulase-negative Staphylococcus species appear to be the most common culprit. , , , , , , , , , Septic arthritis caused by these bacteria tends to follow a more indolent course. With less impressive signs and symptoms, these infections may be more difficult to diagnose in the immediate postoperative period. , , Fortunately, their potential for articular damage is also less significant. In animal models, these bacteria cause minimal destruction of cartilage unless the animal is immunocompromised. This stands in contrast to the second most common pathogen, Staphylococcus aureus , , , , , , , , , which can destroy more than one half of the glycosaminoglycan and collagen content of articular cartilage within 1 week if not treated appropriately. Patients suffering from septic arthritis caused by this species tend to have a higher graft removal rate, longer antibiotic course, and inferior clinical outcomes. , , Although infections caused by nonhemolytic Streptococcus, Peptostreptococcus, Proprionibacterium, Enterobacter , anaerobic, Gram-negative, fungal, and Mycobacterium species have all been reported, these occurrences are exceedingly rare. , , , , , , , , , , Suspicion for a polymicrobial infection should be heightened in cases of septic arthritis resistant to standard treatments. , , ,
Although the low incidence of septic arthritis following ACLR precludes the creation of validated clinical management guidelines, , sufficient evidence exists to establish overarching principles informing proper treatment of this condition. The ultimate goals of management for septic arthritis following ACLR are simple: first, protect the articular cartilage; second, protect the graft. Articular cartilage can lose greater than half of its glycosaminoglycan content if treatment is delayed for more than 1 week. , , The graft may be endangered within 24 hours of onset of a Staphylococcus aureus infection. Therefore expeditious initiation of appropriate treatment is vital. Doing so can lead to acceptable outcomes, even in cases caused by more virulent strains of Staphylococcus aureus. Treatment delays greater than 7 days lead to a longer duration of therapy, higher incidence of graft removal, and higher likelihood of restricted knee range of motion.
Once a strong suspicion for septic arthritis is established, and after joint aspiration has been performed and blood cultures have been drawn, intravenous antibiotics should be initiated promptly. , , , Delaying treatment in anticipation of synovial fluid culture results is not recommended because it is fairly common for cultures to show no growth. , , Initial therapy should consist of empiric agents providing broad coverage of Staphylococcus aureus and coagulase-negative Staphylococcus species, commonly accomplished with a third-generation cephalosporin or vancomycin. , , , , , , , , Some suggest using a beta-lactam antibiotic combined with an aminoglycoside, whereas others recommend a beta-lactamase–resistant agent. The antibiotic regimen should subsequently be narrowed with the assistance of an infectious disease specialist, using culture results as a guide. , , , There is no consensus to support a specific duration of intravenous antibiotics. , , , Conservative recommendations suggest a minimum of 4 to 6 weeks of therapy, , , , , , , whereas others posit that intravenous agents are needed only until the temperature and CRP normalize. , , ,
While arranging intravenous antibiotics, the patient should concurrently be scheduled for urgent arthroscopic irrigation and debridement (I&D). The current literature supports arthroscopic I&D as the mainstay of treatment. , , , , , , , , , , Arthroscopy is cost effective and allows for complete synovectomy while avoiding violation of the extensor mechanism and decreasing the risk of hemarthrosis. Multiple arthroscopic debridements may be required to eradicate the infection. , , In fact, some recommend repeat debridement even if clinical and laboratory metrics improve. Graft integrity must be assessed during the initial arthroscopic procedure, both directly using arthroscopic instruments and indirectly with a pivot shift test under anesthesia. The importance of extensive debridement should not be overlooked––inadequate synovectomy has been identified as a key factor in cases of persistent infection. , , , , If double fixation was used at the initial procedure, the tibial screw may be exchanged to decrease bacterial biofilm load. Although standard arthroscopic portals are sufficient, a posterior portal may improve examination and debridement of the posterior compartment. The amount of normal saline required for proper irrigation has not been established, but 6 to 9 L is generally considered to be adequate. Postoperatively, a drain system is often used to assist with wound drainage until minimal output is observed. , , Although the use of continual irrigation systems and antibiotic beads postoperatively have been reported, there is minimal support for their efficacy. , , ,
Many authors previously advocated for removal of the graft during initial debridement. As an avascular biological foreign body, the graft was thought to provide a medium for bacterial growth and biofilm formation. , , , , , Whereas acceptable outcomes can be attained with early graft removal, many now support graft retention as the initial treatment. , , , , , , , , , , , , , , This may be owing to improvements in early diagnosis and appropriate antibiotic therapy, as well as considerations of both the morbidity of an additional reconstruction and the instability of a knee left without a graft. Scenarios in which graft removal should be considered include signs of graft incompetence on initial arthroscopic I&D, , , , , delayed presentation of a Staphylococcus aureus infection, , , , and persistent infection despite appropriate initial treatment. , , , , , The ultimate decision should be made in concert with the patient, weighing the risk of persistent infection against the potential morbidity of graft removal and additional procedures. , If graft removal is undertaken, most surgeons recommend delaying revision for 6 to 9 months.
Once the patient has undergone arthroscopic I&D and completed an appropriate course of intravenous antibiotics, they should be transitioned to oral antibiotics. When available, culture results can guide the choice of antimicrobial agent. In cases of negative culture growth, cephalexin is recommended. Clindamycin can be substituted for patients with a beta-lactam allergy. Similar to the use of intravenous agents, there is no consensus regarding duration of treatment, with recommendations varying from 2 to 6 weeks. Some authors prefer to continue antibiotics until the CRP has normalized. , , , , However, this involves additional cost in the form of clinic visits and phlebotomy, and the efficacy of this practice has not been proven.
Of equal importance for long-term functional outcomes is the rehabilitation process. Although patients may be immobilized for a short period after arthroscopic I&D to allow for tissue rest, gentle range-of-motion exercises should begin within 24 to 48 hours. It is also imperative that aggressive therapy be initiated early, as soon as the infection is cleared clinically.
The potential for significant patient morbidity associated with septic arthritis mandates the inclusion of effective preventative measures in the surgical treatment plan. Antibiotic prophylaxis reduces the risk of infection 4-fold in clean surgical cases, and is therefore justified in ACLR procedures, where the consequences of infection can be devastating. Intravenous administration of a first-generation cephalosporin is the method of choice and has demonstrated superiority to clindamycin for preoperative prophylaxis. Further, this antibiotic class provides adequate coverage for outdoor athletes exposed to unique strains of bacteria. , , ,
Multiple reports have demonstrated that septic arthritis associated with ACLR can be successfully managed with minimal long-term detriment to the patient. Both graft retention and revision produce acceptable patient-reported outcomes, objective clinical measures, and return to activity. Less favorable outcomes are associated with delays in treatment, irrigation without debridement, , and graft removal without revision. Although posttreatment imaging may show evidence of accelerated osteoarthritic degeneration, , , these changes occur to a degree in all patients undergoing ACLR, and are influenced by other injury factors such as meniscal tears. , Despite largely positive results, it is not uncommon for patients to experience functional deficits, residual pain, and stiffness, even with prompt initiation of appropriate treatment. , , ,
Symptomatic deep venous thromboembolism (DVT) requiring treatment was recently reported as the most common complication following ACLR.
Reported rates of venous thromboembolism ( VTE) following ACLR vary widely in the literature, ranging from 0.05% to 14% for DVT and 0% to 0.2% for pulmonary embolism (PE). , , , This variance is explained by the fact that some studies analyzed large databases, where prophylactic practices and diagnostic methods are unclear, whereas others prospectively screened for VTE with imaging. Higher incidences were reported in the latter scenario because nearly three-quarters of patients diagnosed with VTE by imaging will be asymptomatic.
Despite the relatively low incidence of symptomatic VTE following ACLR, this complication is associated with significant morbidity, mortality, and resource expenditure.
Several risk factors for VTE have been identified within the ACLR population. Patients older than 35 years may be at increased risk of VTE after ACLR. , Concurrent use of oral contraceptive pills (OCPs) is another factor associated with increased risk. Although female gender has been reported as a risk factor for VTE following ACLR, OCP use within this study population was unknown. Others have found no significant relationship between gender and VTE. Personal history of DVT, chronic venous insufficiency, thrombophilia, or malignancy have all been associated with VTE after ACLR. Although family history of coagulation disorders is a known risk factor for knee arthroscopy, , , this relationship has not been established for ACLR specifically. Procedure-related risk factors for VTE following ACLR include tourniquet time in excess of 90 to 120 minutes , and increased surgical complexity (i.e., concomitant high tibial osteotomy or multiligamentous repair). , Modifiable lifestyle factors associated with increased risk include nicotine use , , and a body mass index over 30.
The presence of DVT may be suspected in patients with increased lower extremity pain, swelling, warmth, and erythema. However, this classic constellation of symptoms is found in fewer than half of cases. Furthermore, the signs and symptoms associated with DVT can be attributed to expected postoperative changes or to other disease etiologies such as infection. As a result, ultrasonography or venography are required to establish the diagnosis. Ultrasound has proven to be highly sensitive and accurate in detecting DVT postoperatively. It is also cheaper and less invasive than venography, making it an appropriate first choice in the diagnostic evaluation. , ,
The spectrum of presenting features can vary widely for PE as well. Patients with significant dyspnea, chest pain, or hemodynamic instability will likely be evaluated in an emergency department. However, the signs and symptoms of PE can also be subtle. Therefore all surgeons should be highly attuned to even slight changes in their patients’ cardiopulmonary status during the postoperative period. Workup for PE is outside of the scope of this chapter. Prompt referral for evaluation and treatment is warranted if PE is suspected.
Standard protocols for DVT management are sufficient in patients diagnosed with a postoperative DVT at the level of, or proximal to, the knee. Initial anticoagulation therapy for a duration of 10 days is begun immediately upon diagnosis and can be achieved using low molecular weight heparin (LMWH) or fondaparinux subcutaneously, factor Xa inhibitors orally, or unfractionated heparin. Inpatient admission is not necessary: several randomized trials and metaanalyses have demonstrated the safety of outpatient administration of LMWH in patients who are hemodynamically stable, at low risk of bleeding, with no renal insufficiency, and who have a support system. Once therapeutic anticoagulation has been confirmed, the patient can be transitioned to long-term therapy for a minimum of 3 months. , Factor Xa and thrombin inhibitors are the preferred agents, but warfarin, LMWH, and fondaparinux can also be used. Agent selection for both initial and long-term therapy should be based on clinician experience, patient factors, and cost. In the case of a distal DVT, patients may either forgo chemical anticoagulation and opt for serial screening to monitor progression, or proceed with treatment using a standard anticoagulation regimen. The ultimate decision should be determined on a case-by-case basis after discussion of the risks and benefits with the patient. Aside from the risk of bleeding, some believe that treatment of acute DVT may increase the risk of PE because of embolization of clot fragments. , However, others argue that this risk is low, and that thrombolysis can be safely carried out without the need for an inferior vena cava filter.
Although the incidence of symptomatic VTE following ACLR is relatively low, the potential for significant morbidity and mortality warrants prophylactic considerations. Preventative measures in the setting of ACLR are typically simple: the majority of sports medicine surgeons encourage early mobilization within 24 hours postoperatively, with smaller subsets also incorporating sequential compression devices and compression stockings. Routine screening for VTE is not recommended . Administration of chemical prophylaxis requires careful consideration because these medications carry risks of bleeding and drug reactions, and may involve an unacceptable cost or inconvenience for the patient. , , , Only one prospective randomized controlled trial has evaluated VTE prophylaxis in patients undergoing ACLR, comparing LMWH with a placebo. The authors reported a significantly lower incidence of DVT in patients who received LMWH preoperatively, with no significant difference in bleeding observed between the study groups. However, cases of DVT included both distal clots and asymptomatic proximal clots identified using magnetic resonance venography, which are generally considered to be clinically insignificant. ,
The Clinical Practice Guidelines released by the American Academy of Orthopaedic Surgeons provides recommendations regarding VTE prophylaxis after elective total knee arthroplasty, but not for arthroscopic procedures. Guidelines from the Eighth American College of Chest Physicians Consensus Conference on Antithrombotic Therapy recommended against routine pharmacological prophylaxis following knee arthroscopy, which was supported by a Cochrane review published the same year. The guidelines stated that prophylaxis with LMWH could be considered in patients at higher risk as a result of a complicated or prolonged procedure, or owing to patient-associated risk factors. Subsequent consensus from the Ninth American College of Chest Physicians Consensus Conference on Antithrombotic Therapy narrowed the indications for chemical prophylaxis after knee arthroscopy to patients with personal history of DVT; however, this was a grade 2B (weak) recommendation. , Despite this, approximately 50% of sports medicine surgeons routinely administer chemical prophylaxis, with the supermajority using aspirin. Ultimately, the decision to enact prophylactic anticoagulation in the setting of ACLR must be individualized to each case. The risks, advantages, and costs involved should be carefully weighed in conjunction with the patient and his or her preferences. ,