Neutrophil-to-Lymphocyte Ratio as a Marker for Postoperative Stress in Robot-Assisted Total Knee Arthroplasty

Introduction

Total joint arthroplasty aims to relieve end-stage knee pain, restore function, and improve quality of life. While the success of the traditional manual total knee arthroplasty (mTKA) heavily depends on the surgeon’s skill, robot-assisted total knee arthroplasty (rTKA) has gained popularity for its potential to enhance outcomes, including reduced postoperative pain and improved prosthesis positioning [1]. rTKA offers precise alignment of the prosthesis, resulting in fewer alignment issues and better adherence to expected lower-extremity force lines [2]. This precision enhances postoperative comfort, reduces inflammation, and decreases medication needs [3].

Despite these advantages, rTKA carries also additional risks. rTKA is associated with longer surgical times, leading to extended wound exposure and potentially higher risks of wound infection [4,5]. Critics have also raised concerns about increased tourniquet durations, higher costs, and the potential for peri-prosthetic joint infections [2,4,6]. These challenges underscore the need for research on the inflammatory and stress responses associated with rTKA.

The neutrophil-to-lymphocyte ratio (NLR) has emerged as a valuable novel marker of systemic inflammation and stress. NLR reflects the balance between neutrophil-mediated innate immunity and lymphocyte-driven adaptive responses. Thus, it provides insights into the balance between innate and adaptive immune responses, serving as a valuable indicator of inflammation and stress [7]. Elevated NLR values are linked to stress conditions, such as sepsis, trauma, and major surgery, making it a reliable indicator of acute stress and systemic inflammation [8]. Preoperative NLR levels have also been identified as predictors of various complications, such as deep vein thrombosis following TKA [9].

Major surgery can cause lymphocytes to migrate from peripheral blood to lymphatic tissue, leading to lymphopenia [10]. Consequently, surgical stress has the potential to trigger a significant systemic inflammatory response syndrome (SIRS), leading to a level of lymphocyte depletion more profound than the levels seen in severe infections, such as severe sepsis and septic shock [7,11].

Postoperative elevated NLR values may indicate severe inflammation or trauma and predict poorer prognosis in morbidity and mortality [12,13], while declining levels may suggest clinical improvement.

Normal adult NLR values typically range between 1 and 2, with values exceeding 3 or falling below 0.7 considered pathological [13–19]. Aydin et al reported a range of 1.0 to 2.3, with a median of 1.9, in healthy Turkish adults [16]. In advanced knee osteoarthritis (OA), NLR levels are often elevated compared to mild OA and healthy controls. Ionitescu et al reported NLR values of 2.82 in advanced OA, compared to 1.99 in mild OA and 1.98 in controls [20]. Given that robot-assisted total knee arthroplasty (rTKA) is primarily performed in advanced OA patients, these findings provide context for interpreting systemic inflammation and stress responses in this population. NLR levels tend to increase in orthopedic surgery, polytrauma, and burn patients, with reported values ranging from 3.3 to 5.5 [7,21].

Therefore, in clinical practice, tracking the NLR values daily, focusing on the absolute values and the fluctuations, may help capturing the stress response and monitoring of the progression of the patient:

Given its affordability, accessibility, and sensitivity, NLR appears to be a user-friendly, practical tool for monitoring stress and inflammation. This study investigates the acute postoperative stress response following rTKA using NLR values, offering insights into inflammatory processes associated with rTKA and comparing these findings with outcomes from traditional mTKA documented in the literature.

Material and Methods

PRIMARY STUDY OUTCOME:

The primary outcome measure, based on retrospective evaluation results, was that the POD 1 NLR value is above 3 in more than 90% of patients who underwent robot-assisted total knee arthroplasty.

SECONDARY STUDY OUTCOME:

The secondary outcome of the retrospective evaluation was that the average NLR value obtained on POD 3 in patients who underwent robot-assisted total knee arthroplasty decreased to 50% of the average NLR value obtained on POD 1.

ADDITIONAL ANALYSIS:

Additional factors examined included age, perioperative hemorrhage, American Society of Anesthesiologists (ASA) classification, and comorbidities.

INCLUSION AND EXCLUSION CRITERIA:

Our study’s patient inclusion criteria are as follows:

We excluded patients based on the following criteria:

DATA COLLECTION:

NLR values were analyzed at 4 time points: preoperatively, on POD 1 (NLR1), POD 2 (NLR2), and POD 3 (NLR3). Due to the retrospective nature of the study, complete NLR values were unavailable for all patients. NLR1 values were available for all 50 patients, while preoperative NLR, NLR2, and NLR3 values were collected for 48 patients.

STATISTICAL ANALYSIS:

This retrospective cohort study included all eligible patients identified from hospital records, thereby minimizing selection bias. The sample was not based on a predetermined size; rather, all consecutive patients who met the inclusion criteria during the study period were included. This empirical approach ensured a comprehensive representation of the patient population.

Patient characteristics were summarized in terms of the count, the mean, or the ratio, as appropriate to the context. Descriptive statistics were used to summarize patient characteristics, including frequencies for categorical variables and means with standard deviations for continuous variables. Changes in NLR values were reported using tables, including mean values and standard deviations observed during follow-up assessments. Distributions of NLR values and count levels were assessed using histograms.

All statistical analyses and graph plotting were performed using Microsoft Excel (Version 16.95; Microsoft Corporation, Redmond, WA, USA). Results were reported with a focus on clinical relevance and practical applicability.

Results

PATIENT COHORT:

Our study included 50 patients, of whom 41 (82%) were female and 9 (18%) were male, as shown in Table 1. The mean age of the patients was 67.8 years, with an average age of 69.2 years for women and 66.3 years for men. One patient underwent combined epidural anesthesia while the others received general anesthesia for the operation. For postoperative pain management, 84% of patients were administered epidural analgesia with patient-controlled analgesia (PCA), while 12% received intravenous PCA. Among the patients, 28 had unilateral knee osteoarthritis, while 21 had bilateral knee osteoarthritis, for which they underwent rTKA surgery.

Table 2 details the age distribution of patients undergoing rTKA. The largest age group was 70–79 years (48%), followed by the 60–69 age group (32%). Table 3 presents the comorbidities of the study participants.

PRIMARY AND SECONDARY STUDY OUTCOMES:

The retrospective results showed that only 2 out of 50 patients had an NLR value below 3, indicating that the study achieved its primary outcome, with over 90% of patients having a POD 1 NLR value above 3. The secondary outcome measure was also achieved, with average NLR values of 9.48 on POD 1 versus 4.54 on POD 3, demonstrating that, on the average, NLR value dropped by POD 3 to under 50% of the POD 1 values.

Figure 1A displays the minimum, maximum, and mean NLR values for all 50 rTKA patients during the preoperative period and postoperative follow-up days. The mean preoperative NLR value was recorded as 2.37±1.34. On POD 1, 2, and 3, the corresponding mean values were 9.48±6.53, 5.04±2.76, and 4.54±4.11, respectively.

AGE GROUPS:

Preoperative NLR values did not show statistically significant variation across the groups aged <60, 60–69, and 70–79 years, with mean values of 2.2, 2.09, and 2.25, respectively (Figure 1B). In the group aged 80 year and older, the mean NLR was notably higher than in the other age groups, with the value of 4.71. However, given that this group comprised only 6% of all patients, further data collection is necessary to confirm whether elevated preoperative NLR values are consistently observed in this subgroup.

Focusing on the groups aged 60–69 (34%) and 70–79 (48%) years, which included most patients, a significant rise in NLR was observed in the immediate postoperative period. On POD 1, mean NLR values were 8.08 and 9.93, respectively. These values gradually declined over time, with POD 2 readings of 4.23 and 5.24, and POD 3 readings of 3.36 and 4.16, respectively (Figure 1C).

UNILATERAL VS BILATERAL PATIENTS:

Figure 2 presents mean NLR values in unilateral and bilateral rTKA patients. Figure 3 provides a detailed comparison of minimum, maximum, and mean NLR values during the preoperative period and postoperative follow-up days for these 2 groups. Among unilateral rTKA patients, preoperative NLR was 2.61±1.58, which rose to 8.43±4.60 on POD 1. Corresponding values on POD 2 and 3 were 4.81±2.78 and 4.73±5.22, respectively. In contrast, bilateral rTKA patients had a lower preoperative NLR mean of 2.07±0.88, which increased to 10.93±8.44 on POD 1. POD 2 and 3 values for this group were 5.37±2.76 and 4.30±2.04, respectively.

TEMPORAL PROGRESSION OF NLR LEVELS:

Figure 4 shows the distribution of patients according to their NLR levels across various perioperative time points. The distribution of patients was analyzed based on their NLR levels and corresponding stress levels. NLR levels were categorized as:

Figure 4A shows that during the preoperative period only 10 patients had elevated NLR levels indicative of moderate to severe inflammation, potentially meeting the criteria for diagnosis of non-infectious systemic inflammatory response syndrome (SIRS). Figure 4B shows that 48 patients had surgical systemic inflammatory response syndrome (surgical SIRS), as indicated by NLR values exceeding 3 (reflecting a decrease in lymphocyte count post-surgery, attributed to ‘surgical SIRS’). On POD 1 (NLR1), 29 patients had elevated NLR levels (>7), potentially indicating severe inflammation and stress levels. Given the dynamic nature of NLR, which can fluctuate significantly within a few hours in pathological conditions, it is not surprising that 48 patients showed elevated NLR levels on POD 1.

The anticipated trend is a decline in NLR values during the subsequent postoperative days. NLR levels below 7 are linked with enhanced clinical outcomes and a lowered risk of mortality. As previously noted, on POD 1 (NLR1), 29 patients had elevated NLR levels (>7). By the second and third days, these values decreased to 9 and 7, respectively, indicating improvements in most patients during the early postoperative period (Figure 4B–4D). These improvements were also evident in Figure 1A, presenting the minimum, maximum, and mean NLR values for all 50 rTKA patients during both the preoperative period and postoperative follow-up day. The mean preoperative NLR value was 2.37±1.34, whereas the values on POD 1, 2, and 3 were 9.48±6.53, 5.04±2.76, and 4.54±4.11, respectively.

Discussion

DISCUSSION OF THE PRIMARY STUDY OUTCOME – NLR ON POD 1:

As reported in Figure 4, in the preoperative period, only 10 patients had elevated NLR levels, indicating moderate to severe inflammation, potentially meeting the criteria for non-infectious systemic inflammatory response syndrome (SIRS). After surgery, 48 patients showed evidence of surgical SIRS, with NLR values exceeding 3, reflecting a decrease in lymphocyte count. On POD 1, 29 patients had NLR values >7, suggesting severe inflammation and stress. As expected, NLR levels declined over the next 2 days, with values dropping from POD 1 to POD 3, showing improvement in most patients. The average preoperative NLR was 2.37±1.34, while POD 1, 2, and 3 values were 9.48±6.53, 5.04±2.76, and 4.54±4.11, respectively, indicating a trend towards recovery.

To assess our primary outcome, we considered the percentage of patients undergoing rTKA who exhibited an NLR value exceeding 3 on POD 1. The distributions of NLR during both the pre- and postoperative periods are visually represented in Figure 5. Our findings revealed that 20.83% of patients exceeded this threshold in the preoperative period. Subsequently, we evaluated the increase in NLR (>3) percentage over consecutive postoperative days, including POD 1, 2, and 3, with results indicating improvements: 96%, 77%, and 62.50%, respectively. Based on these results, an NLR >3 on POD 1 was found in 96% of patients. This indicates that our primary outcome – over 90% of patients undergoing robot-assisted total knee arthroplasty (rTKA) having a neutrophil-to-lymphocyte ratio (NLR) exceeding 3 on POD 1 – was confirmed.

DISCUSSION OF THE SECONDARY STUDY OUTCOME – NLR ON POD 3:

The secondary objective of the study was to confirm a larger than 50% reduction in the mean NLR value on POD 3 relative to POD 1. As we see in Figure 5, a wide range of NLR values were observed on POD 1, with the highest mean NLR value recorded at 9.48. Following this, the distribution range narrows in the subsequent days, accompanied by a decrease in mean NLR values, reaching 5.04 on POD 2 and 4.54 on POD 3. Therefore, the secondary study outcome was also confirmed, with mean NLR values of 9.48 on POD 1 and 4.54 on POD 3, indicating a 52.1% reduction in NLR. When examining the 50% NLR reduction on a patient basis, our results showed that this decrease was observed in 23 out of 49 patients (46.09%) on POD 2, and in 25 out of 49 patients (51.02%) on POD 3.

IMPACT OF TKA ON INFLAMMATORY RESPONSE; COMPARISON OF RTKA VS MTKA:

Recent studies have linked rTKA to a reduction in the early postoperative local inflammatory response [25,26]. This improvement may be attributed to the advantages of rTKA, including reduced iatrogenic periarticular soft tissue injury, decreased femoral and tibial bone trauma, and improved accuracy in component positioning compared to mTKA [26]. For instance, Khlopas et al reported that rTKA, particularly when preserving the posterior cruciate ligament (PCL), reduced PCL injury and tibial subluxation relative to mTKA [27].

In a study by Xu et al, involving 65 patients (34 rTKA and 31 mTKA), preoperative NLR values were similar between groups (2.7±2.5 for rTKA vs 2.5±1.2 for mTKA). However, postoperative NLR differed significantly: on POD 1, NLR was 13.2±7.6 in the rTKA group versus 10.4±4.8 in the mTKA group; on POD 3, values were 5.0±2.8 (rTKA) and 5.8±4.0 (mTKA) [2]. In our study involving 50 rTKA patients, the preoperative NLR results closely resembled those reported in Xu’s study, with a value of 2.37±1.34. The increase in NLR observed in our study was not as pronounced as that reported in Xu’s study. Despite a significant rise in NLR on the first day compared to the preoperative value, the NLR values did not reach the levels observed in the rTKA group of Xu’s study, which had an average value of 9.48±6.53. This suggests a lower level of inflammation and stress response following the rTKA operation in our study, potentially influenced by factors such as variations in age, surgical incision, and the comorbidities present in the patient group. On POD 3, patients in Xu’s study exhibited an NLR value of 5.0±2.8, which closely mirrors the corresponding NLR values observed on our study (4.54±4.11).

Kayani et al also found that rTKA patients exhibited a transient reduction in early (day 7) postoperative inflammatory response compared to mTKA. However, no significant differences were noted during the immediate (<48 hours) or late (day 28) postoperative phases [26]. As we discuss below, in our study, we also observed a reduction in NLR during the early period of the postoperative inflammatory response, aligning with and corroborating the findings of the Kayani study. However, our study lacked a control group and did not cover the late response of NLR after rTKA.

COMPARISON OF THE EFFECTS OF UNILATERAL VS BILATERAL RTKA ON INFLAMMATORY RESPONSE:

Additionally, we investigated the variation in minimum, maximum, and mean NLR values among patients undergoing unilateral and bilateral robot-assisted total knee arthroplasty during both the preoperative period and postoperative follow-up days, as depicted in Figure 3. Initially, preoperative NLR values were 2.61±1.58 for unilateral rTKA patients and 2.07±0.88 for bilateral rTKA patients. These values rose to 8.43±4.60 and 10.93±8.44, respectively, on POD 1. NLR values on POD 2 were 4.81±2.78 and 5.37±2.76, and on POD 3 they were 4.73±5.22 and 4.3±2.04, respectively. Similarity of the results in Figures 2 and 3, which provided a detailed representation of the minimum, maximum, mean NLR values, and standard deviations during the follow-up days for patients who underwent unilateral and bilateral rTKA, indicates that there is no significant difference in the elevation of NLR after the postoperative period, regardless of whether the operation was unilateral or bilateral.

In selected patients, bilateral TKA may offer advantages, such as single administration of anesthesia, shorter hospital stay, reduced costs, and more efficient correction of bilateral bone deformities. Since our results indicate that bilateral rTKA may not increase the inflammation and the surgical stress when compared with unilateral rTKA, bilateral rTKA may be recommended for select patients. Additionally, considering our results, we believe that designing a prospective study to obtain clearer and more comprehensive information regarding the lack of difference in inflammation between unilateral and bilateral rTKA could provide valuable information.

ADDITIONAL FACTORS FOR RTKA:

In addition to the main objectives, we examined 4 additional factors within our dataset;

AGE AND NLR VARIATIONS:

Since age appears to be a factor influencing NLR, we examined the variations in NLR across different age groups [7,14]. As shown in Table 2, our analysis did not reveal statistically significant difference in preoperative NLR values among the groups aged <60, 60–69, and 70 years, with respective averages of 2.2, 2.09, and 2.25. However, the 80-year-old group exhibited a notably higher mean NLR of 4.71 compared to other age groups. Given that this group constituted only 6.12% of the total patient population, further data collection is warranted to confidently ascertain whether consistently elevated preoperative NLR values are characteristic of this demographic.

Patients in the group aged 60–69 years (34%) and the group aged 70–79 years (48%) are highly represented in our study. We did not observe statistically significant difference in preoperative NLR values between the groups aged 60–69 years and 70–79 years with the averages of 2.09 and 2.25, respectively. Notably, there was a significant increase in NLR observed in the immediate postoperative period, with values of 8.08 and 9.93 on POD 1, respectively. Subsequently, NLR values began to decline, registering readings of 4.23 and 5.24 on POD 2, and 3.36 and 4.16 on POD 3, respectively. These 2 age groups exhibited similar trends in NLR changes during the postoperative period spanning 3 days.

COMORBIDITIES AND NLR VARIATIONS:

In Figure 6A, we report the variation in NLR at different time points (pre-NLR, NLR1, NLR2, and NLR3) based on various conditions, such as diabetes (DM), hypertension (HT), heart disease, and thyroid disease. There was a similar progression of NLR across time for all comorbidities.

PERIOPERATIVE HEMORRHAGE AND NLR VARIATIONS:

Since perioperative bleeding is a significant stress factor in TKA operations, we further investigated its impact on postoperative NLR values to better understand the resulting stress and inflammatory response.

Figure 6B shows NLR variations at different time points (pre-NLR, NLR1, NLR2, NLR3) based on different amounts of perioperative bleeding. The results indicate that an inflammatory response is significant, especially on POD 1, when the bleeding amount exceeded 1000 ml compared to lesser amounts of bleeding. On POD 2 and 3, we did not observe any differences in NLR values among the groups. These results indicate that the amount of bleeding influences the stress and inflammatory response especially during POD 1.

ASA (AMERICAN SOCIETY OF ANESTHESIOLOGISTS) CLASSIFICATION AND MULTIPLE COMORBIDITIES:

Among our patients, 35 were classified in the (ASA <3) group, while 13 were classified in the (ASA=3) group. We compared NLR values between patients with higher ASA scores (ASA=3) and those with lower ASA scores (ASA<3). As depicted in Figure 7, NLR values varied across ASA classifications, with notably more pronounced elevations observed in patients with higher ASA scores on POD 1. Specifically, the pre-NLR values were 2.1±0.96 for the ASA<3 group and 3.05±1.93 for the ASA=3 group, resulting in elevations of 7.88±5.24 and 13.62±8.36, respectively, on POD 1. By POD 3, NLR levels decreased to 3.65±1.52 for the ASA<3 group and 6.71±7.22 for the ASA=3 group. These findings indicate that NLR levels remained higher in patients with higher ASA scores compared to those with lower ASA scores.

In our study group, 1 patient with multiple comorbidities (including diabetes mellitus, hypertension, COPD, and cardiac problems) had elevated NLR values both preoperatively and postoperatively (pre-NLR: 8.72, NLR1: 17.39). Another case, involving a patient with restrictive lung disease, who also showed a significant increase in NLR on POD 1, had a history of reintubation in the intensive care unit after operation (pre-NLR: 3.73, NLR1: 37). These results indicated that, although the progression of NLR did not significantly vary based on the underlying comorbidities, the values tended to be higher in patients classified as (ASA=3) and patients with multiple comorbidities.

LIMITATIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH:

A recent study suggested that in patients undergoing total knee arthroplasty, spinal anesthesia reduces the postoperative rise in CRP levels, while showing no significant difference in NLR between the spinal and general anesthesia groups [28]. Thus, an important limitation of our study is that it did not address the effects of anesthesia and postoperative pain management on stress and inflammatory response following rTKA, as most patients were given general anesthesia and were managed with epidural PCA.

Our retrospective study focused solely acute postoperative stress response by short-term NLR outcomes (the first 3 postoperative days), limiting our ability to assess long-term NLR normalization. Additionally, the retrospective design prevents follow-ups and assessment of longer-term quality of life. To address these limitations, future research should prioritize prospective studies with control groups larger sample sizes. Such studies will provide a comprehensive evaluation over time, improve quality-of-life assessments, and yield more valuable insights.

Conclusions

Our findings validate the utility of integrating NLR monitoring into clinical practice to assess the acute postoperative stress response following robot-assisted total knee arthroplasty (rTKA) in advanced knee osteoarthritis. NLR emerges as a cost-effective, user-friendly, swiftly responsive, and readily available marker of both stress and inflammation. Its ability to capture dynamic changes holds promise for clinicians in detecting the early stages of pathological processes. In particular, patients whose NLR values remain high may potentially benefit from additional postoperative care. For a more comprehensive understanding, further prospective studies are warranted, incorporating control groups and evaluating the normalization period of NLR levels.

Lastly, since our results indicate that bilateral rTKA does not significantly increase inflammation or surgical stress compared to unilateral rTKA, we believe bilateral rTKA can be recommended for select patients.