Does hydrogen peroxide help mitigate the incidence of Cutibacterium acnes in cervical spine surgeries?

Abstract

Background

Surgical site infection (SSI) is a common yet serious complication of cervical spine surgery. While initially thought to be clinically insignificant, Cutibacterium acnes (C. acnes) is an important cause of infection. The purpose of this study was to investigate the ability of a hydrogen peroxide (H₂O₂) application during standard presurgical skin preparation to reduce the burden of C. acnes in patients undergoing cervical spine surgery.

Methods

This was a retrospective review of prospectively collected data. Subjects were randomly assigned to either standard surgical preparation plus H₂O₂ (experimental) or without H₂O₂ (control). Prescrub, postscrub, and dermal cultures were obtained to assess the C. acnes burden after cultures on an aerobic and anaerobic growth medium were held for 21 days. Multivariate analysis was conducted to determine factors associated with presence of C. acnes. Outcome measures included the results of intraoperative cultures and the development of a SSI within 90 days postoperatively.

Results

Patients (n=86) undergoing elective 2- or 3-level fusion via anterior approach were included. Prior to application of the antiseptic solution, 65% (28/43) of the experimental cohort and 77% (33/43) of the control cohort had positive C. acnes cultures (p=.34). Following application of antiseptic solution, there were no differences in positive C. acnes culture rates between the experimental and control cohorts in the epidermal (30% vs. 28%, p=1.00) or dermal (40% vs. 42%, p=1.00) cultures. No differences in the rates of C. acnes eradication from preantiseptic to postantiseptic application occurred for epidermal (p=1.00) or dermal (p=1.00) skin layers. None of the factors were associated with positive C. acnes epidermal cultures on multivariable logistic regression analysis (p>.05).

Conclusions

While there is potential for H₂O₂ to reduce the positive culture rate of C. acnes in cervical spine patients, no difference was seen when compared to standard surgical skin preparation.

Introduction

As cervical spine surgery continues to see an increasing trend in its utilization, these procedures are no more immune in their risk to postoperative complications than the other most commonly performed orthopedic surgeries [[1], [2], [3], [4], [5], [6]]. Secondary to a higher prevalence of cervical spine operations conducted in our aging population, the associated presence of concomitant medical comorbidities further exacerbates the potential for such complications [[6], [7], [8], [9]]. Specifically, a surgical site infection (SSI) following cervical spine surgery is among the most dreaded postoperative complications and is associated with an increased rate of morbidity and mortality [[10], [11], [12], [13], [14]]. With incidence rates ranging as high as 18%, such variability in the literature can be attributed to diagnostic methods improvements, differing criteria for physicians defining a SSI, as well as operative factors regarding instrumentation, case complexity, and the surgical approach used [10,11,[15], [16], [17], [18], [19], [20]]. While Staphylococcus aureus is historically associated with postoperative infections, 1 bacteria, named Cutibacterium acnes (C. acnes), has warranted particular concern in orthopedic spine and shoulder surgery as the etiology for SSI, as it was previously discredited for pathological insignificance [10,12,13,16].

C. acnes is a nonspore forming, gram positive, anaerobic bacilli regarded as an essential part of the normal microbiota where it resides on the skin, hair follicles, and sebaceous glands primarily concentrated in the back, neck, axilla, and chest wall [12,21,22]. Though its role in postoperative deep infections continues to be defined, C. acnes has been associated with up to 10%–20% of all infections following orthopedic surgery and is reported to be the primary source of postoperative infection in shoulder surgery [16,21,[23], [24], [25], [26]]. C. acnes has been identified as an important pathogen in spine surgery, and specifically cervical spine surgery [[27], [28], [29]]. C. acnes was shown to cause as many as 21.6% of all disk infections and 37% of all spinal implant associated infections [29,30].

C. acnes has also been reported as an underlying etiology for patients with nonpyogenic intervertebral infection causing sciatica, Modic changes, and nonspecific lower back pain despite no index procedure predisposing one to bacterial seeding of the deep tissue layers [12,18,[31], [32], [33], [34], [35], [36], [37]]. Though controversial as some studies have failed to reproduce these findings, it is believed that damage to the disc renders it susceptible to bacterial seeding with ensuing inflammatory changes and damage to the encompassing vertebrae [12,18,32,[38], [39], [40]]. As more studies are needed to evaluate the role of C. acnes in SSIs and intervertebral disc degeneration, eradication of the organism from the surgical site is of particular interest to both spine and shoulder surgery.

Surgical disruption of follicles harboring C. acnes provides a means for direct inoculation of the surgical site, deeper tissue layers, instrumentation, and implants allowing for biofilm formation [10,16,17,41,42]. As C. acnes is a slow-growing commensal bacteria, SSI can often be difficult to diagnose due to its indolent nature and atypical presentation often lacking common signs or markers of an underlying infective process [12,20,21,26,42]. Despite advances in detection and management, no study has established a definitive means for eradication of C. acnes during surgical site preparation [17,21].

Colonizing as many as 10⁵ C. acnes organisms per follicular pore, typical surgical site preparation methods using isopropyl alcohol, chlorhexidine gluconate (CHG), or Betadine (Purdue Pharma LP, Stam- ford, CT, USA), as well as intravenously administered preoperative antimicrobial prophylaxis have failed to show successful elimination from the surgical site [21,[43], [44], [45], [46], [47], [48]]. Using a topical solution of 3% hydrogen peroxide (H₂O₂) has demonstrated to be an effective bactericidal agent against C. acnes both in vitro and clinically in the setting of shoulder surgery [21,[49], [50], [51]]. The purpose of this study was to investigate the ability of H₂O₂ application during standard sterile presurgical skin preparation to reduce the burden of C. acnes in patients undergoing cervical spine surgery. We sought to investigate what factors were predictive of a positive C. acnes epidermal culture. We hypothesize that H₂O₂ application will reduce the burden of C. acnes at the surgical site and male sex will show a greater propensity for a positive culture, and that the addition of H₂O₂ to skin preparation will lower the rate of infections, especially infections with C. acnes [16,26,49,52].

Methods

Ethics

Approval was granted for this study by the Nassau University Medical Center Institutional Review Board. All subjects provided informed consent prior to enrollment.

Study Design

This study was a retrospective review of prospectively collected data performed at a level 1 trauma center conducted between August 2020 and July 2022 in patients undergoing surgery by 3 fellowship trained orthopedic spine surgeons. Subjects were included in the study if they met the inclusion criteria of: (1) elected to undergo primary cervical spine surgery via anterior approach, (2) 18 years of age or older at time of surgery, and (3) no mental handicap precluding one's ability to provide informed consent. Subjects were excluded from our study if they failed to meet the aforementioned inclusion criteria or if they had a: (1) previous incision in the area of surgery or were undergoing a revision surgery, (2) traumatic etiology necessitating surgical treatment, (3) history of ongoing infection, (4) known use of antibiotics within 6 weeks prior to surgery, (5) known allergy to H₂O₂, benzoyl peroxide, or any of the other materials of standard sterile skin preparation, or (6) were undergoing primary or concomitant cervical spine surgery via a posterior approach. Inclusion criteria were similar to those seen in previous studies [53].

Sample acquisition

All patients were given weight-based Cefazolin IV 1 hour prior to surgery or Clindamycin if they were allergic to Cefazolin. If necessary, patients with hair at the anticipated area for surgical approach had hair removed using a battery powered hair clipper before any skin preparation was begun. The first culture, labeled “prescrub culture” was obtained prior to any preparation to determine the incidence of C. acnes in the normal skin flora of the respective patient's cervical region. In accordance with experimental protocol, all culture swabs were taken using a sterile cotton swab. The swab was then placed on the skin/dermis and rotated for 5 seconds, ensuring that all sides of the swab made contact with the skin/dermis. The swab was then placed in a sterile container. All participating surgeons were required to follow this swabbing protocol.

All surgeons participating in this study were required to follow the same protocol for both the experimental and control cohorts. In both treatment groups, 70% ethyl alcohol was applied to the appropriate area to clean the skin and remove any gross debris. Patients then underwent skin preparation with either Duraprep and 3% H₂O₂ (experimental) or Duraprep alone (control). Skin preparation included the application of four 1,010 drapes on the surgical site and then cleaning with either Duraprep or 3% H₂O₂ followed by Duraprep. A second culture stick was then used to swab over a 1-cm linear area near the incision site and labeled “skin culture.” Formal draping was subsequently applied and Ioban (St. Paul, MN) draping was used to cover the incision site. The surgeon then made an incision with a third swab, labeled “dermal culture”, taken in the dermal layer. A fourth culture was taken as a control and waved in the air during the start of the case and labeled “air culture.”

All 4 cultures were taken to the microbiology laboratory and grown in the appropriate media with incubation for 21 days to evaluate for C. acnes. Though the surgeon knew what skin preparation was used, this is a single blinded study, as neither the patient nor the microbiology lab was aware of the perioperative preparation.

Data collection

Chart review was conducted independently. Factors evaluated were based on their relevance to the literature pertaining to SSI in spine surgery in general, as well as C. acnes specific infections in patients undergoing spine surgery. These factors included: sex, age, body mass index (BMI), medical comorbidities such as diabetes, hypertension, hyperlipidemia, smoking status (never vs current smoker), Hispanic vs. non-Hispanic heritage, and invasiveness of procedure relative to size of the incision (2 vs. 3 vertebral level surgery) [14,20,38,48,[50], [51], [52]]. Similarly, culture results for each patient were obtained by chart review upon final growth determination after a 21-day incubation period. Finally, patient charts were reviewed for the development of SSI, defined as return to the operating room for irrigation and debridement within 90-days of the index procedure.

Statistical analysis

All statistical analysis was performed via Jupyter Notebook Version 6.4.8 (Open Source) using Python programming language (Willmington, DE) [52,54]. A priori power analysis was performed with G*Power Version 3.1.9.7 (Dusseldorf, Germany) using an effect size of 0.3 in positive culture rates as previously described in the literature as the threshold for clinical significance [53,55]. With the goal of achieving a minimum of 80% power with alpha level of 0.05, it was determined that 36 patients were needed per each cohort. Chi-square or the Fisher's exact test were used for analysis of categorical variables and continuous variables were analyzed using the 2-sample t-test.

Multivariable logistic regression analysis was performed to evaluate factors predictive of a positive C. acnes culture while controlling for covariates (age and BMI). The primary outcome evaluated was presence of Duraprep and 3% H₂O₂ scrub vs. Duraprep. Possible confounding variables assessed included sex (male vs. female), comorbidities (diabetes, hypertension, and hyperlipidemia), smoking status (never vs. ever smoker), Hispanic heritage, and invasiveness of procedure (2-level vs. 3-level fusion). For all analyses, a p-value ≤.05 was determined to be statistically significant. All analyses were 2-tailed.

Results

A total of 86 patients were enrolled in this study upon application of the inclusion and exclusion criteria. Subjects were divided into an experimental and a control cohort each consisting of 43 patients, making our study powered to determine significant differences as defined by our power analysis. Cohorts were similar with respect to age (43.7 ± 11.6 vs. 42.3 ± 13.4, p=.61), male sex (37% (16/43) vs. 27% (12/43), p=.49), BMI (29.6 ± 6.3 vs. 29.7 ± 6.1, p=.61) when comparing the experimental and control cohorts, respectively. A complete comparison of patient demographics between cohorts is available in Table 1.

Table 1.

Patient demographics.

		Overall, n=86	Traditional, n=43	Hydrogen peroxide, n=43	p-value
Demographics	Age	43.0 (12.5)	42.3 (13.4)	43.7 (11.6)	.61
	Male	28 (32.6)	12 (27.9)	16 (37.2)	.49
	BMI	29.6 (6.1)	29.7 (6.1)	29.6 (6.3)	.99
	African	22 (25.6)	11 (25.6)	11 (25.6)	.976
	Hispanic	34 (39.5)	18 (41.9)	16 (37.2)
	White	15 (17.4)	7 (16.3)	8 (18.6)
	Other	15 (17.4)	7 (16.3)	8 (18.6)
Levels fused	II	69 (80.2)	37 (86.0)	32 (74.4)	.28
	III	17 (19.8)	6 (14.0)	11 (25.6)
Other	Diabetes	11 (12.8)	7 (16.3)	4 (9.3)	.52
	Hypertension	15 (17.4)	7 (16.3)	8 (18.6)	1.00
	Hyperlipidemia	5 (5.8)	3 (7.0)	2 (4.7)	1.000
	Smoking	15 (17.4)	8 (18.6)	7 (16.3)	1.000
	Revision	0 (0)	0 (0)	0 (0)	1.000

BMI, body mass index.

Prior to application of the antiseptic solution, 65% (28/43) of patients in the experimental cohort and 77% (33/43) of patients in the control cohort had positive C. acnes cultures (p=.34). Following application of antiseptic solution, there were no differences in positive C. acnes culture rates in the epidermal (30% (13/43) vs. 28% (12/43), p=.81) or dermal (40% (17/43) vs. 42% (18/43), p=.83) cultures for the experimental vs. control cohorts, respectively. There were no differences in the rates of C. acnes eradication from preantiseptic to postantiseptic application for epidermal (61% (17/43) vs. 70% (23/43), p=1.00) or dermal (61% (17/43) vs. 70% (23/43), p=1.00) skin layers for the experimental vs. control cohorts, respectively. A complete comparison of positive C. acnes cultures rates can be seen in Table 2. After controlling for covariates (age and BMI), none of the factors evaluated demonstrated a significant association with positive C. acnes epidermal cultures on multivariable logistic regression analysis (p>.05). A comprehensive list of factors evaluated are seen in Table 3.

Table 2.

Positive Cutibacterium acnes cultures.

	Prescrub	Postscrub	Dermal	Eradicated
Traditional, n=43	33 (76.7)	12 (27.9)	18 (41.9)	23 (69.7)
Hydrogen peroxide, n=43	28 (65.1)	13 (30.2)	17 (39.5)	17 (60.7)
p-value	.24	.81	.83	.46

Positive culture rates at baseline, following presurgical skin preparation, and upon dermal incision, as well as count of positive bacteria cultures which were eradicated between pre and postscrub.

Table 3.

Multivariable logistic regression analysis for factors predictive of positive Cutibacterium acnes epidermal culture.

Factors	OR	95% Confidence Interval	p-value
Male vs. female	0.65	0.23–1.83	.42
Diabetes	1.92	0.33–11.08	.46
Hypertension	0.49	0.13–1.86	.29
Hyperlipidemia	1.73	0.15–19.23	.66
Never vs. ever smoker	0.81	0.23–2.84	.75
Hispanic vs. vs. non-Hispanic	1.27	0.44–3.63	.66
Hydrogen peroxide vs. control	0.60	0.22–1.60	.31
II vs. III Levels	0.81	0.22–2.86	.75

Odds ratios and 95% confidence intervals for patient factors assessed with multivariate analysis, controlling for age and body mass index.

With regard to the development of surgical site infection within 90 days postoperatively, infection rates in the total cohort were low (1/86 = 1.16%). The 1 infection was noted in the control cohort. Chi-square analysis demonstrated no significance between the 2 cohorts (Duraprep and 3% H₂O₂ 0% vs. Duraprep 2.33%, p=.31) (Table 4).

Table 4.

Results of chi-square for surgical site infection.

	No SSI	SSI	p
Traditional, n=43	43 (100.0%)	0 (0%)	.31
Hydrogen Peroxide, n=43	42 (97.7%)	1 (2.3%)

Table 4. Results of chi-square analysis comparing the development of SSI within 90 days postoperatively between the traditional skin preparation cohort and the hydrogen peroxide treated cohort.

SSI, surgical site infection.

Discussion

This is the first study to investigate the application of H₂O₂ preoperatively during standard surgical site preparation in patients undergoing elective primary cervical spine surgery via an anterior approach. When comparing the incidence of C. acnes at baseline (p=.34), postscrub (p=.81), or upon dermal incision (p=.82) between our study cohorts, no difference was found. Furthermore, no difference between the experimental and control cohort was noted when comparing eradication of C. acnes following presurgical skin preparation on the epidermal (p=1.00) or dermal (p=1.00) cultures. When examining factors associated with a positive presurgical culture, neither sex, medical comorbidities (diabetes, hypertension, hyperlipidemia), ethnicity, or invasiveness of procedure were predictive. Further, there was no statistically significant difference in development of SSI between the treatment and control groups. Age, positive preoperative cultures, and male sex were analyzed as these factors have been shown to be highly associated with C. acnes colonization [49,50,53]. Medical comorbidities and ethnicity were analyzed as these have been shown to predict SSI [8,14].

Acting by free radical generation to induce oxidative stress to disrupt the cellular processes of bacteria, H₂O₂ is an inexpensive antimicrobial agent having shown effectiveness at a 3% concentration in reducing C. acnes from the surgical site [13,21,41,[49], [50], [51],53,56]. While potential risk for contact dermatitis, skin blanching, or blistering has been described, no adverse reactions were observed in our sample or in related literature [21,53,57,58]. Moreover, the potential clinical application outweighs these rare and typically transient symptoms, as standard skin preparation methods of ChloraPrep, DuraPrep, and Povidone-Iodine scrub have proven inadequate for eradicating C. acnes from the surgical site [13,53,59]. As the investigation of concomitant application of H₂O₂ presurgically has primarily been the focus of shoulder surgery, our study is the first to investigate its impact in the setting of the cervical spine.

Though our study showed no difference between our control and experimental cohorts, several studies have reported on the potential for H₂O₂ application presurgically. In vitro studies by Hernandez et al. [51] and Ohlin et al. [56] showed a bactericidal effect of H₂O₂ against C. acnes after 5 minutes of application and its potential use in eradicating bacteria from seeding the surgical site by pretreating implants with H₂O₂, respectively. Notably, these studies used lab isolates of C. acnes and did not examine clinical variants of the pathogen. This may account for some of the difference noted between our study and theirs. Topical use of benzoyl peroxide or its active product H₂O₂ prior to shoulder surgery has supported these findings by showing up to 50% reduction in positive C. acnes skin cultures with an even greater effect on deep cultures when combined with topical clindamycin [21,53,57,60]. These studies likely have positive results due to differing methodologies. Kolakowski et al., compared C. acnes derived their eradication rates based on the difference between contralateral sides of the same patient, 1 side treated with benzoyl peroxide and the other chlorhexidine gluconate [21]. This limits the generalizability of the study as it limits comparison across clinical variants of C. acnes. In another example, Stull et al., used punch biopsy as opposed culture swabs. [53] Both Sabetta et al. [60] and Dizay et al. [57] allowed for more extended periods of application and penetration of benzoyl peroxide, thus allowing for greater efficacy. Thus, many of these observed differences were due to methodology.

Furthermore, our study found that overall rates of infection were low. Our total cohort SSI incidence of 1.16% matches that described in the literature. [61] While no significant difference was noted between the H₂O₂-treated group and the control group, interestingly the singular infection found among the total sample was only in the control group. More research is needed to further confirm this pattern.

Though determining the optimal presurgical treatment to mitigate the risk of a SSI is paramount, identifying what factors increase the risk of a positive C. acnes culture is prudent as well. In general, obesity, smoking, alcohol abuse, steroid use, malnutrition, medical comorbidities, extremes of age, and surgical approach all play a role in SSI development [10,11,14,62] Specific to C. acnes, men have been found to have up to a 5-fold greater incidence of C. acnes, while geography and ethnicity have shown variations in the skin microbiome composition [26,[47], [48], [49],52,[63], [64], [65]].

While no factors evaluated in our study were associated with a positive C. acnes culture at baseline when controlling for confounding factors, understanding what factors predispose to a greater biologic burden and facilitate early intraoperative colonization are of the utmost importance in patients electing to undergo spinal surgery. Thus, it is important to also discuss the limitations of our study and possible further areas of study. While our study sought to evaluate the impact of H₂O₂ on reducing C. acnes burden in patients undergoing cervical spine surgery, each subject in our sample underwent surgery via an anterior approach. This limits the generalizability of our findings as different surgical approaches and locations are inherently subject to their own respective risk of SSI, as well as C. acnes burden [10,11,14,21,53,62]. Despite this restraint on external validity, we sought to provide the foundation for understanding the implication behind the use of H₂O₂ in patients undergoing cervical spine surgery. Similar to the findings in the shoulder literature, future studies may wish to investigate the effect of application of H₂O₂ with or without topical antibiotics as this may yield a difference in culture rates [21,53,57,60].

Another possible limitation of our study was that implant design and composition data was not collected. A study by Garcia et al. found polyether ether ketone implants to show the greatest biologic burden for rapid biofilm formation for C. acnes in the first 24 hours from exposure [66]. This constitutes a potential future direction and further research should address this topic. The inclusion of 3 surgeons in this study is another possible limitation, as there could be some variance in preparation methods. However, this was mitigated by necessitating that all participating surgeons follow the above outlined protocols. This study was also limited by its limited follow-up. Finally, this study was limited by its inclusion of only a single-center, limiting the generalizability of the results as rates of SSI and C. acnes colonization vary between locations.

Conclusion

While there is potential for H₂O₂ to reduce the positive culture rate of C. acnes in cervical spine patients, no difference was seen when compared to standard surgical skin preparation. Future studies are necessary to determine if there is an optimal duration of application or different combination of skin preparations, or whether the addition of antibiotics to skin preparation might help to reduce C. acnes burden.

Funding

No funding was received for the conduct of this research. The authors have no other relevant conflicts of interest to disclose.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. MCM: Nothing to disclose. NAO: Nothing to disclose. HG: Nothing to disclose. JF: Nothing to disclose. JT: Nothing to disclose. AA:Nothing to disclose.