Retrospective Review of Complications and Outcomes in COVID-19–Positive Patients With Comorbidities Undergoing Limb Salvage Procedures in a Tertiary Care Wound Center

As of February 2022, more than 78 million cases of SARS-CoV-2, the virus that causes COVID-19, have been documented in the United States alone.¹ As the virus has spread globally, considerable evidence has emerged linking COVID-19 with a hypercoagulable, hyperinflammatory state that leads to venous and arterial thromboembolism and dysregulation of wound healing pathways.^2-4 Upregulation of proinflammatory cytokines and subsequent dysregulation of coagulation in COVID-19 infection can lead to PE and deep vein thrombosis,^2,5-7 as well as extrapulmonary thromboses such as acute limb ischemia, large vessel arteriosclerosis obliterans, myocardial infarction, and stroke.^8-12 This same hyperinflammatory state has been associated with poor wound healing and locally elevated proinflammatory cytokines.¹³ While COVID-19 is an independent risk factor for surgical mortality, the attributable risk of perioperative death due to infection remains difficult to isolate.¹⁴

At the authors’ limb salvage center, patients with complex LE wounds often require microsurgical reconstruction. Patients typically undergo extensive surgical debridement before, during, and after LE reconstruction and amputation procedures. When patients continue progressing with standard treatment but are not candidates for FTT, the need for eventual amputation is discussed. Patients with chronic LE wounds typically have multiple coexisting medical comorbidities—such as diabetes, hypertension, and PVD—that negatively affect their microvasculature and wound healing. Superimposing the robust inflammation of COVID-19 infection in those with baseline vascular dysfunction poses theoretical additional perioperative risks and alterations in wound healing amongst patients undergoing LE reconstruction surgery. Although surgical guidelines have been proposed for managing patients during the pandemic,^15-18 there is currently a paucity of literature specifically focused on LE reconstructive procedures to guide surgeons in making evidence-based operative decisions. The authors describe their recent experiences with 17 patients who underwent LE limb salvage procedures—including FTT, amputation, and debridement—within 30 days of their COVID-19 positive diagnosis.

All operative procedures for chronic LE wounds performed by the senior authors (K.K.E. and C.E.A.) at a tertiary wound care center from February 2020 to March 2022 were retrospectively reviewed. Of 1462 procedures performed during this period, 597 patients were screened for a COVID-19 diagnosis, and of those, 67 (11.2%) patients who were diagnosed with COVID-19 were identified. The surgical procedure that fell within 30 days of COVID-19 diagnosis was considered the “index” procedure. All patients who were COVID-negative or were diagnosed more than 30 days before or after their date of surgery were excluded. Ultimately, 17 patients were identified who had been diagnosed with COVID-19 within 30 days before or after their index procedure and were included in the study. Patient data on demographics, comorbidities, wound and operative details, postoperative complications, and long-term outcomes were collected from patient charts. Primary outcomes included rates of 1) perioperative mortality, 2) SICU admission, and 3) thrombosis. Secondary outcomes collected included the development of or treatment for COVID-related symptoms and ambulation status. COVID-related symptoms were defined as the onset of fever, chills, cough, shortness of breath, dyspnea, fatigue, myalgias, headache, sore throat, nausea, or diarrhea in the setting of a positive nasopharyngeal reverse transcriptase-polymerase chain reaction SARS-CoV-2 test. Decision to administer antiviral treatment with remdesivir was made on a case-by-case basis based on severity of clinical presentation.

Patient characteristics

Seventeen patients received a diagnosis of COVID-19 within 30 days before or following LE reconstructive surgery. The patients were predominantly male (n = 12, 70.5%), had an average age of 56 ± 17.3 years, and a mean BMI of 28.7 ± 6.3 kg/m². The median Charlson Comorbidity Index score was 5 (IQR: 2, 8.5). Most patients were Black (n = 10, 58.8%), 4 were former smokers (23.5%), and 2 were active smokers (11.8%). Table 1 summarizes patient demographics.

While all 17 included patients tested positive for COVID-19 within 30 days either before or after their LE reconstructive procedure, 10 (58.8%) patients were diagnosed preoperatively and 7 (41.2%) postoperatively relative to the index procedure. The mean time between COVID-19 diagnosis and index procedure was 8.5 ± 15.2 days. Of the 5 patients (29.4%) who had been vaccinated before COVID-19 infection, 2 had completed the 3-dose series, with the remaining 3 patients only having received an initial dose. Infections varied in clinical presentation. COVID-19 infections were asymptomatic in 10 cases (58.8%) and symptomatic in 7 (41.2%), with 1 patient requiring treatment with remdesivir. Additional COVID-19 characteristics are detailed in Table 2.

Operative details and postoperative outcomes

Across the 17 included patients, index procedures performed included: incision and drainage or debridement (n = 10, 58.8%), amputation (n = 6, 35.3%), and FTT to the LE (n = 3, 17.6%). Two patients underwent LE amputation followed by FTT. Amputation levels consisted of AKA (n = 2, 11.8%), BKA (n = 3, 17.6%), and metatarsophalangeal joint (n = 1, 5.9%). The 3 FTT procedures utilized 2 ALT flaps and 1 latissimus dorsi flap. Osteomyelitis was the most common indication for operation, occurring in 13.3% (n = 6) of procedures.

Postoperatively, all 17 patients received anticoagulation therapy, including subcutaneous heparin (n = 8, 47.1%), enoxaparin sodium injection (n = 7, 41.2%), clopidogrel bisulfate (n = 1, 5.9%,), and continuous heparin drip (n = 1, 5.9%). At least 1 postoperative complication occurred in 29.4% (n = 5) of patients, including necrosis/gangrene (n = 2, 11.8%), partial flap necrosis (n = 1, 5.9%), hypotension with delirium (n = 1, 5.9%), and a nonhealing wound (n = 1, 5.9%).

Two patients were admitted to the SICU following index procedure, 1 for hypotension necessitating pressor treatment and 1 after the patient developed a PE intraoperatively. One patient was taken back to the operating room on postoperative day 1 following an ALT flap to the lateral foot for concern of poor flap perfusion. On reoperation, venous congestion and intraluminal thrombi were noted. The patient was subsequently switched from subcutaneous heparin to heparin infusion following the procedure and was eventually discharged on postoperative day 11 with adequate flap healing. Table 3 details operative and postoperative outcomes.

Long-term outcomes

The average follow-up duration was 4.3 ± 3.2 months, at which point, 6 (35.3%) of the 17 patients were fully healed. Less than half of the patients (n = 6, 35.3%) were ambulatory at the most recent follow-up visit, though 5 generally (29.4%) required ambulatory assistance, with only 1 patient being independently ambulatory. The mortality rate at the most recent follow-up visit was 29.4% of patients (n = 5) (Table 4). Of the 5 patients that died, the mean duration between index procedure and date deceased was 61 ± 37.7 days. The most common causes of death were PE (n = 2, 40.0%) and pneumonia (n = 2, 40.0%), of which 1 was suspected to be aspiration pneumonia and the other a sequela of COVID-19 infection. The patients who developed PEs died 40 and 86 days postoperatively, and those with pneumonia died 26 and 39 days postoperatively. The patient who developed aspiration pneumonia originally underwent AKA as revision of a prior total knee arthroplasty secondary to infection, which likely contributed to their death. The fifth patient died 115 days postoperatively following an acute gastrointestinal bleed in the setting of suspected chronic alcohol misuse.

The onset of the COVID-19 pandemic affected almost every realm of medical care, with surgical procedures being no exception. In March 2020, several surgical specialty societies, including the American Society of Plastic Surgeons, strongly recommended minimizing, postponing, or canceling elective procedures. Consequently, there was a 48% decrease in total surgical volume compared with 2019.¹⁹ Patients with chronic LE wounds were therefore advised to delay surgical intervention for as long as possible. Patients who underwent LE reconstruction during this time were consequently some of the sickest patients at the authors’ institutions, whose cases could not afford to be postponed. Consequently, to the authors’ knowledge, this study presents the first review assessing the effects of COVID-19 infection, specifically, on limb salvage operative outcomes in a highly comorbid patient population. Although this is a heterogenous group of patients in a single institution, the study highlights an increased perioperative mortality and morbidity in the study population during the COVID-19 pandemic.

While the pathophysiology remains incompletely understood, the COVID-19 pandemic has caused delayed patient presentations and worsened outcomes in highly comorbid patients undergoing limb salvage procedures, such as those with poorly controlled diabetes.²⁰ One presumed theory documented across multiple specialties is that the induced hypercoagulable and hyperinflammatory state may pathologically affect wound healing and operative outcomes.^21–26

The exact duration of COVID-associated coagulopathy or hyperinflammatory state has yet to be established. An analysis of 531 patients found that thrombotic events in the setting of COVID-19 infection occurred at a median of 11 days (range, 5–20 days) after the first symptoms of infection.²¹ However, thrombosis can also occur in cases of asymptomatic COVID-19 infections.²² In free flap surgery, most vascular complications manifest within the first 12 to 24 hours^,27,28 yet multiple cases of delayed flap thrombosis in COVID-19–positive patients have been reported.^23,24,29,30 Only 1 of the 3 patients in the current study who received a free flap within 30 days of a COVID-19 diagnosis developed postoperative flap thrombosis, occurring on postoperative day 1. Comparatively, 2 patients (11.8%) died from PE following BKA procedures. A previous study investigating venous thromboembolism rates following LE amputation in 49 patients without COVID-19 reported an overall PE rate of 12.2%, of which 4% were fatal.³¹ The differences in fatality rate secondary to PE between the 2 studies can likely be explained by the presence of COVID-19 infection, which was associated with a thromboembolic risk ratio of 33.05 during days 1 through 30 following infection.³² However, the presence of multiple underlying comorbidities such as diabetes and PVD make it difficult to attribute the increase in PE fatality rate strictly to COVID-19 infection. Future studies are therefore warranted to quantify the duration of coagulopathy and inflammatory states in patients who test positive for COVID-19 to better understand the pathology underlying increased mortality risk.

Ultimately, it remains difficult to determine whether COVID-19 infection is the sole cause of death in postoperative patients due to the multitude of confounding comorbidities of the study population. However, COVID-19 was identified as an independent risk factor for surgical mortality.^.14 The current study corroborates these findings by reporting a 29.4% mortality rate in patients undergoing limb salvage with COVID-19 positivity, a drastically higher rate in comparison to previous studies from the authors’ institution.^33–35 Although additional larger, multi-institutional studies are needed, COVID-19 infections are beginning to show dangerous implications on a patient’s perioperative course. It is critical for both surgeons and patients to have an understanding of the increased risk of mortality and consider the necessary steps to delay surgery until outside this window of associated risk.

COVID-19 infection is also associated with perioperative delayed wound healing.²⁴ Inouye highlighted near-total skin graft loss in the donor sites of COVID-19–positive patients who underwent mandibular reconstructions.²⁴ Da Silveira et al described sternal dehiscence following a coronary artery bypass graft.³⁶ In this study, 4 out of the 5 patients developed postoperative complications that were wound-related—including necrosis/gangrene (n = 2, 11.8%), partial flap necrosis (n = 1, 5.9%), and a nonhealing wound (n = 1, 5.9%)—which prolonged their hospital stay. These findings may be related to the unbalanced proinflammatory cytokine response that occurs during COVID-19 infection. This proinflammatory imbalance has the potential to impact wound healing and the patient’s overall hospital course.^37,38 Previous studies assessing cytokine levels of chronic healing and nonhealing leg ulcers have identified persistently elevated levels of pro-inflammatory markers such as interleukin-1, interleukin-6, and tumor necrosis factor-α.^13,39 Inflammatory stimuli trigger enzymatic and reactive oxygen species-mediated endothelial glycocalyx degradation, expose cell adhesion molecules that promote leukocyte and platelet adhesion, and ultimately induce a pro-thrombotic state.^40–42 Mitigating cytokine storms with low molecular weight heparin, particularly in patients with severe disease, may underlie the mortality benefit.^43,44 Furthermore, these biomarkers have been found in higher concentrations from fluid drainage of chronic nonhealing wounds compared to healing wounds, suggesting their nocuous impact on the healing process.¹³ Additional studies are warranted to assess the local changes in the inflammatory response in COVID-19–positive patients with chronic wounds and guide the medical management of these high-risk patients.

This study has several limitations, including its small sample size and retrospective nature, which relies on the quality of electronic medical records. Furthermore, COVID-19 documentation may be underreported, as the identification of patients is limited to those who underwent testing within the authors’ hospital system medical records. It is not possible to determine the degree of impact SARS-CoV-2 had on wound healing, as most of the included patients had multiple medical comorbidities that may have also played a role in delayed wound healing or hypercoagulability. The shift in the health care landscape, knowledge of disease, and hospital management during the 2-year study period may also influence perioperative outcomes and decisions to operate. Nonetheless, these results highlight the multitude of risks present when performing surgical procedures in COVID-19–positive patients with multiple comorbidities.

This study highlights the possible implications a COVID-19 infection may have on the postoperative course among patients with multiple or severe comorbidities undergoing limb salvage procedures. Such patients likely require additional perioperative risk stratification or close follow-up to optimize outcomes.