Association of Obesity With Inflammation and Pain After Total Hip Arthroplasty

Abstract

Background

The prevalence of obesity is increasing, and obesity often leads to degenerative joint disease requiring total hip arthroplasty (THA). Obesity is a proinflammatory state associated with an increase in chronic, low-grade inflammatory response. As such, it may augment the postoperative inflammatory response, which has been associated with postoperative pain and complications.

Questions/purposes

We determined whether severity of obesity was associated with (1) severity of inflammatory response, as measured by the in vivo circulating levels of cytokines and ex vivo functional reactivity of mononuclear blood cells, and (2) severity of pain, as measured by verbal pain scores and analgesic consumption, in the first 24 hours after THA.

Methods

We studied 60 patients (20 normal weight, 20 overweight, 20 obese) undergoing elective primary unilateral THA in this prospective cross-sectional study. Blood samples were collected for C-reactive protein and cytokine levels, including IL-1β, IL-2, IL-6, IL-8, and tumor necrosis factor α (TNF-α), from patients before and 24 hours after surgery. Cytokine response of whole blood was evaluated ex vivo with or without two standard activators, phorbol-12-myristate-13-acetate and lipopolysaccharide, using standardized blood sample from patients at 24 hours. These standard immune activators are implicated in the inflammatory response to gram-negative infection, translocation of microbial products, pathophysiology of septic shock syndrome in human, and tumor promotion. Pain response was gauged using verbal pain scores (on a 0- to 10-point scale, where 0 = no pain and 10 = worst pain) at rest and with activity at 24 hours after surgery and analgesic consumption of volume of epidural analgesic solution for the first 24 hours after surgery.

Results

No correlation was found between BMI and postoperative spontaneous circulating cytokine levels. However, after activation of blood leukocytes with lipopolysaccharide, there was a significant positive correlation between the BMI and IL-1β, IL-6, and TNF-α levels (r = 0.26–0.32; p = 0.03, p = 0.03, and p = 0.01, respectively), suggesting priming of the innate immune system in obesity and potential for excessive postoperative inflammatory response. Obesity was not associated with increased pain or analgesic consumption in the first 24 hours after surgery.

Conclusions

Obesity is associated with a proinflammatory state after THA as demonstrated by enhanced cytokine reactivity. Larger studies exploring the specific impact of obesity and inflammation on surgical outcomes, including pain, are warranted.

Level of Evidence

Level II, therapeutic study. See the Instructions for Authors for a complete description of levels of evidence.

Introduction

The prevalence of obesity is increasing, reaching epidemic proportions in parts of the world [30]. In the most recent National Health and Nutrition Examination Survey, it was reported that 36% of the US adult population is obese [8]. Increased BMI is associated with several chronic comorbid conditions, including degenerative joint disease [13, 31], often requiring surgical intervention such as total hip arthroplasty (THA). An analysis of the National Hospital Discharge Survey from the CDC identified 2,288,579 patients having undergone THA from 1990 to 2004, with greater than 100% increase in the incidence of obesity among patients during that time period [20]. Indeed, the rising rate of obesity is a key factor in current projections for the United States that by 2030 the number of THAs will increase by 174% to nearly 600,000 procedures per year [17]. Pharmacokinetics, respiratory and cardiac physiology, monitoring, and postoperative care are all profoundly affected by increased body mass; thus, the increasing number of obese patients undergoing THA have increased risk for postoperative complications [18, 25].

The postoperative period is characterized by a prolonged systemic inflammatory response and depressed cell-mediated immunity, which have been shown to play a major role in multiple organ failure, delayed healing, and risk of infection during the postoperative period. For example, systemic levels of several cytokines, including IL-6, increase after THA and are associated with negative outcomes [16, 28]. Additionally, the extent of postsurgical inflammatory reaction correlates with the severity of pain experienced by patients after total joint arthroplasty [15]. Obesity has recently been found to be an independent risk factor for severity of postoperative pain after musculoskeletal surgery and for risk of surgical site infection after THA [21, 23]. Although the exact underlying mechanisms remain obscure, some of the consequences and complications of obesity likely are caused by the condition. Obesity is considered a chronic systemic inflammatory state at baseline [9, 29]. Recent studies have shown that fat tissue functions as an endocrine and immunoactive organ and secretes a number of inflammatory proteins, including adiponectin, tumor necrosis factor α (TNF-α), and IL-6. This potential for obesity to play a causative or synergistic role in postoperative inflammation and pain has not, to our knowledge, been evaluated carefully to date. Only one report has suggested that IL-6 level is exaggerated in severely obese patients after surgery [10]. Therefore, further characterization of the inflammatory response and functional capacity of leukocytes in obese patients should be valuable in predicting postoperative course and outcomes after THA.

We therefore determined whether severity of obesity is associated with (1) severity of inflammatory response, as measured by both circulating levels of cytokines and the functional capacity of leukocytes activated to produce inflammatory mediators, and (2) severity of pain, as measured by verbal pain scores and analgesic consumption, in the first 24 hours after THA.

Patients and Methods

Study Cohort

This study was an observational prospective cross-sectional study of male and female adults between the ages of 18 and 85 years scheduled for elective primary unilateral THA at the Hospital for Special Surgery (HSS) (New York, NY, USA). The study was approved by the institutional review board, and written informed consent was obtained. Patients were selected from those scheduled to start between 0800 and 1400 hours with the standard HSS surgical, anesthesia, and postoperative analgesia technique. Patients were excluded if affected with severe comorbid conditions (eg, American Society of Anesthesiologists physical classification > 3). Between December 22, 2010, and October 31, 2011, we approached patients who met the above-mentioned inclusion criteria when research personnel were available until we had achieved the desired number of participants (n = 60). A total of 63 patients were enrolled, which included 29 men and 34 women, with a mean age of 64 ± 11 years and a mean BMI of 27 ± 6 kg/m². The patients were divided into three groups according to BMI: normal weight (BMI < 25 kg/m²), overweight (25 kg/m² ≤ BMI < 30 kg/m²), and obese (BMI ≥ 30 kg/m²). Three patients were eliminated from final analysis: one patient with normal weight was later found to have chronic renal insufficiency and two overweight patients were lost to followup. Of the remaining 60 patients, 20 had normal weight (mean BMI, 21.7 ± 1.9 kg/m²), 20 were overweight (mean BMI, 27.1 ± 1.6 kg/m²), and 20 were obese (mean BMI, 35.0 ± 3.8 kg/m²).

Surgical, Anesthesia, and Postoperative Analgesia Technique

All patients received our standard anesthetic that was combined spinal epidural anesthesia with intravenous sedation with an arterial line for blood pressure monitoring and for blood draw. The standard postoperative analgesic was epidural patient-controlled analgesia (PCA) with a combination of bupivacaine and hydromorphone that continued for the first 24 hours after surgery. Patients received meloxicam 7.5 to 15 mg and dexamethasone 6 mg orally in the holding area. The meloxicam was continued through Postoperative Day (POD) 2. For pharmacologic deep venous thrombosis prophylaxis, aspirin 325 mg twice daily was used.

Outcomes

The primary outcome was the relationship between BMI and postoperative inflammatory response, specifically, induced secretion of cytokines. Blood was obtained before surgery in the operating room and on POD 1 for markers of inflammation, including C-reactive protein (CRP), TNF-α, IL-1β, IL-2, IL-6, and IL-8. Secondary clinical outcomes were the relationship between BMI and pain scores at 24 hours, as measured by a 0- to 10-point verbal rating scale (0 = no pain to 10 = worst pain) at rest and with activity, and the relationship between BMI and analgesic consumption, assessed using the epidural PCA pump at 24 hours. Anthropometrical measures, including age, sex, height, body weight, and BMI, were collected before surgery.

Laboratory Assays

The serum concentrations of IL-1β, IL-6, IL-8, and TNF-α were determined in the General Core Laboratory of the Clinical and Translational Science Center in Weill Cornell Medical College (New York, NY, USA) using a quantitative sandwich enzyme immunoassay kit from Meso Scale Discovery (Gaithersburg, MD, USA) following the manufacturer’s instruction. The assays had 100% reactivity with all of the measured cytokines. The intra- and interassay coefficients of variation for these assays were 15% or less and 18% or less, respectively. The sensitivities of the IL-2, IL-1β, IL-6, IL-8, and TNF-α assays were 1.3 pg/mL, 0.18 pg/mL, 0.26 pg/mL, 0.10 pg/mL, and 0.37 pg/mL, respectively.

The VITROS^® 5.1 FS Chemistry System methods (Ortho-Clinical Diagnostics, Rochester, NY, USA) were performed at the HSS Laboratory to determine the concentrations of CRP.

Cytokine Induction Assay

We measured the ex vivo functional reactivity of diluted whole-blood samples by stimulating them with two standard activators, bacterial lipopolysaccharide (LPS) and phorbol-12-myristate-13-acetate (PMA) with a calcium ionophore, ionomycin. These activators evaluate dynamic response of the immune system and provide mechanistic insight into pathways leading to the release of cytokines and inflammatory response [27]. LPS is used to assess innate immune response to bacterial infection. LPS activates Toll-like receptor 4 (TLR4) and the nuclear factor-kappa B signaling pathway that leads to cytokine production and release in monocytes, while PMA is a protein kinase C activator that bypasses the T cell receptors and promotes T cell and monocyte terminal differentiation and activation [4, 14].

Peripheral blood was collected into Vacutainer^® tubes (BD Biosciences, San Jose, CA, USA) containing sodium heparin from fasting patients the day after surgery and transported at room temperature to the Weill Cornell Cellular Immunology Laboratory. A standardized aliquot of each blood sample was used for the cytokine induction assays. Blood was mixed by inversion, diluted 1:1 in RPMI 1640 with HEPES and GlutaMAX™ (Invitrogen™; Life Technologies Corp, Grand Island, NY, USA) containing penicillin at 40 IU/mL and streptomycin at 40 μg/mL (Gemini Bio-Products, West Sacramento, CA, USA), and aliquoted at 1 mL into sterile tubes (Falcon™ 14-mL tubes; BD Biosciences). Activators were added as follows to duplicate tubes and one set was used as a control. PMA (Sigma-Aldrich, St Louis, MO, USA) was prepared at 0.1 mg/mL in dimethyl sulfoxide (Sigma), frozen until day of use, and then diluted to 1 μg/mL with sterile phosphate-buffered saline (PBS) (cellgro^®; Mediatech, Inc, Manassas, VA, USA). PMA was added to duplicate tubes at 10 μL of diluted stock to give a final concentration of 10 ng/mL of cell suspension. Ionomycin (Sigma) was reconstituted in EtOH at 5.0 mg/mL, stored frozen in aliquots until use, and then diluted 1:10 in sterile PBS before adding to aliquots of blood with PMA at a final concentration of 1 μg/ml. LPS (Sigma) was reconstituted in PBS at 0.5 mg/mL and frozen until use. LPS diluted 1:10 in sterile PBS was added to cell cultures at 20 μL to give a final concentration of 1 μg/mL. Cultures were gently vortexed and incubated at 37° C in a humidified incubator with 5% CO₂ for 5 hours. After incubation, cultures were centrifuged at 400g for 10 minutes, and supernatants were removed and frozen at −80° C in microcentrifuge tubes (SafeSeal™; Sorenson BioScience, Salt Lake City, UT, USA) until assayed for cytokine level measurement, as described above. Amounts of cytokines were measured and reported before adding activators (in the medium) and after adding each activator (PMA and LPS) separately.

Statistical Analysis

Patient demographics were summarized by means and SDs for continuous variables and frequencies and percentages for categorical or discrete variables. Cytokine levels and pain scores were summarized by means and SDs. Pre- and postoperative cytokine levels were compared using the Wilcoxon signed-rank test. The strength of the association between BMI and pre- and postoperative cytokine levels was assessed using the Spearman’s rank correlation coefficient. All statistical analyses were performed using SAS^® Version 9.1.3 (SAS Institute, Inc, Cary, NC, USA). We considered p values of less than 0.05 significant.

Sample Size Calculation

Response of IL-6 to LPS induction on POD 1 was considered our primary outcome, as Naito et al. [22] previously examined inflammatory markers after THA with epidural anesthesia and noted peak increases by POD 1 with return to baseline by POD 2. No data were available for an a priori power calculation. Using our observed data with sample sizes of 21, 21, and 17 for the normal, overweight, and obese groups, respectively, with corresponding mean LPS-induced IL-6 values of 7603 ± 2633, 8063 ± 2747, and 8932 ± 1508 pg/mL, a sample size of 32 per group would be needed for a power of 0.8 and an alpha of 0.05 for the observed 17% difference.

Results

Correlation Between Obesity and Inflammatory Response

No correlation was found between BMI and pre- or postoperative in vivo circulating levels or ex vivo spontaneous secretions without activators for any of the cytokines, including IL-1β, IL-2, IL-6, and IL-8 (Table 1). However, there was a positive correlation between BMI and IL-1β, IL-6, and TNF-α levels after leukocyte activation with LPS (but not with PMA), but no correlation between BMI and IL-8 levels with the numbers available (Table 1). CRP levels were not correlated with BMI either before (r = 0.21; p < 0.1) or after (r = 0.23; p = 0.06) surgery (Table 1).

Table 1.

Correlations of BMI with cytokines

Cytokine/chemokine	Preoperative circulating mediator in vivo		Postoperative circulating mediator in vivo		Activator-induced cytokine secretion in vitro
	r value	p value	r value	p value	Spontaneous (no activator)		LPS		PMA
					r value	p value	r value	p value	r value	p value
CRP	0.21	0.1	0.23	0.06
IL-1β	−0.12	0.41	−0.23	0.1	0.18	0.1	0.27	0.03	0.11	0.39
IL-2	ND		ND		−0.06	0.60	0.22	0.1	0.12	0.38
IL-6	0.23	0.1	0.05	0.70	−0.11	0.38	0.28	0.03	−0.02	0.85
IL-8	0.05	0.73	0.006	0.96	−0.02	0.85	0.25	0.05	0.1	0.41
TNF-α	0.07	0.6	−0.13	0.36	0.07	0.59	0.32	0.01	0.13	0.32

LPS = lipopolysaccharide; PMA = phorbol-12-myristate-13-acetate; r = correlation using Spearman test; CRP = C-reactive protein; TNF-α = tumor necrosis factor α; ND = not detected.

Baseline circulating levels of CRP were uniformly low in patients (< 1 mg/dL) preoperatively and increased up to 10-fold (average 4.7) postoperatively (Table 2). Levels of IL-1β and IL-2 were negligible before and after surgery, while low and moderate levels of TNF-α, IL-6, and IL-8 increased significantly from baseline in all patients (p < 0.001) (Table 2).

Table 2.

Levels of cytokines before and after THA for all patients

Cytokine/chemokine	Preoperative circulating mediator in vivo*	Postoperative circulating mediator in vivo*	p value (Wilcoxon signed-rank test)	Activator-induced postoperative cytokine secretion in vitro
				Spontaneous (no activator)	LPS	PMA
CRP (ng/mL)	0.8 ± 0.42	4.6 ± 5.2	< 0.001
IL-1β (pg/mL)	1.2 ± 1.2	1.4 ± 1.3	0.005	52	1468	734
IL-2 (pg/mL)	< 1	< 1	0.179	< 1	< 1	11,909
IL-6 (pg/mL)	3.9 ± 3.2	63.7 ± 51.9	< 0.001	737	8150	616
IL-8 (pg/mL)	11.2 ± 4.7	23.7 ± 21.9	< 0.001	2438	7974	8304
TNF-α (pg/mL)	10.3 ± 6	10.99 ± 5.9	0.005	525	6582	4328

* Values are expressed as mean ± SD; LPS = lipopolysaccharide; PMA = phorbol-12-myristate-13-acetate; CRP = C-reactive protein; TNF-α = tumor necrosis factor α.

Correlation Between Obesity and Pain

With the numbers available, there was no relationship between obesity and pain scores. Mean preoperative pain scores were similar regardless of weight: normal weight, 3.4 ± 3.3; overweight, 2.9 ± 2.9; and obese, 2.9 ± 3.2. Mean postoperative pain scores at rest were similar regardless of weight: normal weight, 2.1 ± 2.2; overweight, 2.1 ± 2.3; and obese: 1.7 ± 1.6. Mean postoperative pain scores with activity were also similar regardless of weight: normal weight, 3.3 ± 2.3; overweight, 3.8 ± 2.2; and obese, 4.6 ± 2.5. Analgesic consumption via the epidural PCA pump was similar regardless of weight: normal weight, 194 ± 79 mL; overweight, 164 ± 78 mL; and obese, 179 ± 82 mL. There were no deaths, major complications, or surgical site infections in any of the patients.

Discussion

Cytokines are mediators released from a variety of immune cells that promote hemodynamic, metabolic, and immunologic changes after surgery. Certain cytokines augment and others attenuate immune system activity, and the balance between the proinflammatory and antiinflammatory mediators is important in appropriate healing and prevention of infection after tissue injury [24]. While cytokines are important in recovery from injury and infections, exaggerated systemic inflammatory response can lead to increased pain, hemodynamic instability, and end organ damage [3, 11]. Obesity is now recognized to be a proinflammatory state. Some of the inflammatory cytokines released in obesity are secreted from adipose tissue and some from macrophages in the adipose tissue stroma [7]. Due to the increasing prevalence of obesity and obesity-related musculoskeletal morbidities, documenting the association of postoperative immunologic response in obese patients has potential clinical value. We therefore determined whether severity of obesity is associated with (1) severity of inflammatory response, as measured by both circulating levels of cytokines and the functional capacity of leukocytes activated to produce inflammatory mediators, and (2) severity of pain, as measured by verbal pain scores and analgesic consumption, in the first 24 hours after THA.

Our study has inherent limitations common to exploratory pilot observational studies. Correlations were significant but modest and need to be reexamined in future larger studies with perhaps more invasive orthopaedic procedures. As a pilot study, our sample size was modest and not based on a formal power analysis due to lack of preexisting data. Our study now allows a post hoc power analysis for outcomes of interest such as pain scores for activity on POD 1 (power of 0.31 to detect a 20% difference with an alpha of 0.05). Our modest sample size and elective surgical population may not have allowed us to differentiate effects of obesity and enhanced inflammatory state on pain. This exploratory pilot study was primarily intended to identify all potential inflammatory markers for future study from our panel of measured mediators and was not intended to be a definitive study. Due to this exploratory nature, multiple statistical comparisons without adjustment were performed to increase ability to detect potential biomarker candidates. Future focused studies with more rigorous statistical methodology will be needed to definitively confirm our unadjusted findings or identify spurious associations from our pilot data. Finally, the observational nature of our study would not have allowed formal assignment of causation between obesity, inflammation, and pain.

Our results support our hypothesis that severity of obesity is related to severity of postoperative inflammatory response as measured by induced cytokine secretion from activated leukocytes. This observation suggests that obese patients may be primed for an exaggerated inflammatory response to perioperative stress. However, we did not observe clinical correlation with pain, and there were no differences in spontaneous levels of cytokines or activated IL-2 and IL-8. If our findings are confirmed in future definitive studies, obesity may be a potentially modifiable risk for increased acute and chronic pain and complications such as systemic inflammatory response syndrome.

In this study, we considered the ex vivo induced cytokine release to be our primary outcome for several reasons. THA is an elective surgery with limited tissue trauma and we did not necessarily expect to observe high levels of spontaneous circulating cytokines in our patients. Furthermore, cytokines have often a very short half-life in circulation [19], and circulating levels may not reflect the dynamic nature of immune response and function of monocytes. Therefore, in addition to measuring circulating levels of cytokines, we assessed the expression of cytokines after incubation of blood samples with LPS and PMA. These standard immune activators assess different aspects of immune response to injury (TLR4 for LPS and protein kinase C for PMA). LPS is implicated in the inflammatory response to gram-negative infection, translocation of microbial products, and pathophysiology of septic shock syndrome in humans [26]. To our knowledge, this is the first study to observe inflammatory priming with this methodology in patient undergoing THA; thus, clinical correlation will need to be established. However, the general belief is that modest stimulation of TLR4 facilitates the elimination of invading microorganisms, while potent TLR4 stimulation leads to severe reactions in the host, often leading to multiple organ failure, septic shock, and death. Overall, we observed that spontaneous circulating levels of IL-1β and IL-2 were negligible before and after surgery, confirming that the procedures were modestly traumatic. The circulatory levels of CRP rose modestly but significantly after surgery compared to baseline, reflecting the normal response to perioperative events. In contrast, IL-6, TNF-α, and IL-8 increased significantly in all patients (p < 0.001) (Table 2). TNF-α is an early potent mediator of cytokine release. Substantial elevations in IL-6 (eg, 334–4000 pg/mL) and IL-8 (eg, 75 pg/mL) have been associated with hepatic and pulmonary dysfunction, sepsis, and death after high-risk major abdominal, aortic, and cardiac surgery [3, 19].

We did not observe a correlation between inflammatory response and pain outcomes. The most likely reason was a limitation in ability to detect differences due to our small number of patients in this pilot study. Several factors in our routine clinical care may have further minimized potential effects of obesity per se on postoperative pain scores. All patients received perioperative dexamethasone and NSAIDs as part of multimodal analgesia and postoperative nausea-vomiting prophylaxis. These are commonly used and recommended agents [1, 5, 6] but may have diminished the magnitude of inflammatory and pain responses in all patients due to their antiinflammatory nature [12]. All patients received aspirin as standard deep venous thrombosis prophylaxis, which again may have diminished inflammatory and pain responses in all patients. All patients received regional anesthesia and analgesia (combined spinal epidural). Previous studies have reported marked suppression of inflammatory markers, including TNF-α and IL-6, with use of central neuraxial versus general anesthesia for THA [22]. The use of epidural analgesia for postoperative pain control provides excellent analgesia [32] that is superior to systemic analgesia including intravenous PCA and may have minimized potential differences in pain scores and analgesic consumption in obese versus nonobese patients. Indeed, a prior study that claimed obesity as a risk factor for increased postoperative pain deliberately excluded patients who had received regional anesthesia and analgesia [2].

Overall, our findings are aligned with the hypothesis that obesity is a proinflammatory condition with the potential to augment the postoperative immune response. More detailed examination is needed to confirm our findings, establish definitive clinical implications for THA and other more invasive procedures such as major spine surgery or knee arthroplasty, and identify the cellular mechanism involved in this observation.