Comparing the Efficacy of a Compound Topical Anesthetic Versus Benzocaine: A Pilot Study

Abstract

To compare the effectiveness of a combination of 10% lidocaine, 10% prilocaine, and 4% tetracaine versus 20% benzocaine for use as a topical anesthetic agent prior to dental injections. A double-blind randomized prospective clinical trial was conducted with 26 participants receiving a topical anesthetic of 20% benzocaine (control) and 26 participants receiving a compound topical anesthetic mixture of 10% lidocaine, 10% prilocaine, and 4% tetracaine (experimental) prior to a maxillary infiltration injection. The procedure was conducted by 1 operator with the Wand^® injection system. Pain was assessed directly with visual analog scale (VAS) scores and indirectly by measuring changes in heart rate at 4 different time points. Complications associated with the application of the topical anesthetics were also assessed. The experimental group had a significantly higher mean VAS score of 19.5 ± 19.7 mm versus 14.2 ± 14.6 mm for the control group (p < .001). No significant differences in heart rate at any of the 4 measured time points compared with baseline were noted for either group. The experimental group had a significantly higher incidence of complications, including tissue sloughing, when compared with the control group (p < .001). Participants in the control group reported significantly lower VAS scores than those in the experimental group. Both types of topical anesthetic showed similar impacts on alterations to heart rate. No benefits were seen with the use of 10% lidocaine, 10% prilocaine, and 4% tetracaine as a topical anesthetic prior to a maxillary infiltration of local anesthetic when compared with 20% benzocaine.

Anxiety and fear of the dentist and dental work are prevalent throughout the general population. Surveys have shown that 22–79% of people have at least mild dental anxiety.¹ Fear of the injection is a specifically anxiety-provoking aspect of dental treatment. The direct experience of pain during the injection procedure is a primary contributor to this fear.^1–3

Topical anesthetics can reduce the pain felt during needle penetration in the administration of local anesthesia.^4,5 Specifically, topical anesthetic helps to relieve anxiety and pain through both psychological and pharmacologic effects.^3,5–7 Psychologically, subjects who are informed they are going to receive topical anesthetic may experience less anticipated pain and may experience decreased apprehension.^6–8 Physiologically, reduction in pain is achieved by blocking nerve impulse conduction in the free nerve endings located within the superficial tissues through the temporary decrease of sodium ion permeability in the nerve cell membrane.^4,9

The most commonly used intraoral topical anesthetic is 20% benzocaine; however, other alternatives have also been used.^3,9 Topical anesthetics can be altered by many factors, including altering the pharmaceutical components, the pH, and the additional additives.⁹ Compound topical anesthetics (CTAs), defined as topical anesthetics containing more than 1 active pharmaceutical component, are gaining popularity. However, it is uncertain whether CTAs are more effective than 20% benzocaine.

The majority of research in the efficacy of CTAs in the dental field is mostly limited to its use for laser gingival re-contouring and orthodontic mini-implant placement.^10–12 Therefore, more research of the use of CTA as an adjunct to intraoral local anesthetic injections is warranted. The aim of this study was to evaluate the efficacy of a CTA using 10% lidocaine, 10% prilocaine, and 4% tetracaine, which has been trademarked as Profound^®, in comparison with 20% benzocaine.¹³

METHODS

The hypothesis for this pilot study was that use of a CTA comprising 10% lidocaine, 10% prilocaine, and 4% tetracaine would be more effective than a topical anesthetic with 20% benzocaine for providing relief of pain during local anesthetic injection in adults.

Study Participants

This pilot study was a randomized, prospective, double-blinded clinical trial, in which 2 different topical anesthetic groups were compared. This study population included volunteers over the age of 21 at the Herman Ostrow School of Dentistry at the University of Southern California (USC).

Fifty-two participants were recruited by mass email in USC School of Dentistry's electronic list serve and by word of mouth and were offered a $10 gift card to compensate them for their time. Informed consent, with an explanation of potential risk and benefits, were discussed with each of the participants. The USC Health Science Campus institutional review board approved this study.

The inclusion criteria were as follows: ages 21 years and older, who were American Society of Anesthesiology (ASA) Type I, as defined by the ASA Physical Status Classification System.¹⁴ Exclusion criteria included the following: history of any known drug or food allergies, active oral lesions, open sores, or any recent surgery in the oral cavity. Pregnant women were also excluded from the study.

Power Analysis

An a priori power analysis was performed to calculate the sample size needed for the primary dependent variable, the evaluation of pain using visual analog scale (VAS) scores. A calculated effect size of 12.5 mm was established based initially on 2 prior studies^10,11 and further refined per clinical experience and judgment. Based upon this established effect size, an 80% power to reject the null hypothesis of equal means with a 2-sided alpha of 0.05 produced a sample size of 25 subjects per group with 50 total subjects. The PASS software, version 14 (Kaysville, UT) was used to calculate sample size.

Randomization and Blinding

Participants were randomly assigned to either the 20% benzocaine control group or the 10% lidocaine, 10% prilocaine, and 4% tetracaine experimental group using a random number table generated using the runiform command in Stata (StataCorp, College Station, TX). The table was managed by a faculty member who was not involved in the study and who dispensed the study medication. The 2 topical anesthetics tested were compounded by a compounding pharmacy (in Woodland Hills, CA) and mixed to resemble the same color, consistency, and flavor. All participants and the investigators were blinded as to which anesthetic they received.

Treatment Protocol

A digital pocket scale was used to weigh 0.25 g of the topical anesthetic. The oral mucosa adjacent to the maxillary right canine was dried with 2 × 2 gauze. The preweighed, topical anesthetic (control or experimental) was applied in the buccal vestibule with a cotton swab for 60 seconds. The Wand^®, a computer-controlled, local anesthesia delivery system, was used to regulate the flow rate. A 27-gauge needle was advanced approximately 2–5 mm into the buccal vestibule. Approximately a quarter of a cartridge, or 0.425 mL, of 2% lidocaine with 1:100,000 epinephrine was injected for a total dose of 8.5 mg lidocaine and .00425 mg epinephrine. The principal investigator (LP) administered the injection of the local anesthetic solution to all the participants under the supervision of a USC faculty dentist.

Outcome Measurements

After the procedure, immediately postinjection, the subjects were asked to rate their pain perception using the VAS. The VAS utilized was a visual aid of a 100-mm ruler with 0 mm on the left identified as “no pain” and 100 mm on the right representing “the worst pain imaginable” with 10-mm increments in between.¹⁵ Each participant was asked to identify his or her pain by pointing on the VAS to the measurement that most accurately represented his or her perceived pain during the most painful portion of the procedure.

Pain was also indirectly measured by assessing changes in the participant's heart rate¹ via pulse oximeter (Criticare^®) at 4 different time points: (a) baseline (B), (b) during administration of the topical gel (preinjection, PI), (c) during needle injection (I), and (d) 1 to 2 minutes after needle injection (postinjection, P).¹

All participants received a phone call the evening after the procedure to assess if there were any adverse events (AEs) from the injection. If any occurred, the participant reported what they were during the phone conversation with the investigator. These AEs were documented and followed until it had resolved.

Statistical Analysis

The primary outcome of interest was pain as assessed using the scores from the VAS (0–100 mm). The data were analyzed using a generalized linear model with a Poisson family and log link specified. Since the study was randomized, confounders that required adjustment were not expected. However, based on the distribution of participant characteristics across the 2 study groups (Table 1), gender, race/ethnicity, and baseline heart rate were all evaluated as possible predictors or confounders. If any of these variables were found to alter the unadjusted relative risk (incident rate ratio [IRR]) by more than 15%, they were considered a confounder and included in the final model. They were also included if they were found to significantly predict VAS scores. Gender, ethnicity, and baseline heart rate were all included in the final model as these 3 factors were found to be significant predictors of VAS scores, though none were confounders of the association between treatment group and VAS scores. Model fit was evaluated via q-normal plots of the residuals, and the likelihood ratio test was used to test differences between full and reduced candidate models. The Hosmer-Lemeshow test was used to assess for goodness of fit.

Table 1. .

Demographics and Clinical Characteristics of the Study Population

Variable*		Experimental, n = 26	Control, n = 26	P Value
Age, y		28.1 ± 4.0	29.8 ± 5.6	.31
Gender	Male	6 (23.1%)	9 (34.6%)	.54
	Female	20 (76.9)	17 (65.4%)
ASA class I		26 (100%)	26 (100%)
Baseline heart rate, bpm		71.9 ± 8.4	78.2 ± 12.0	.05
Ethnicity	Hispanic	5 (19.2%)	6 (23.1%)	.50
	Not Hispanic	21 (80.8%)	20 (76.9%)
Race†		n = 21	n = 20	.41
	White	6 (28.6)	8 (40.0)
	Black	0 (0)	1 (5.0)
	Asian	15 (71.4)	11 (55.0)

^*Continuous variables are presented as mean ± SD, and categorical variables are presented as count (%). P values were obtained by the Wilcoxon rank sum test for continuous variables and Fisher exact test for categorical variables. The missing data for race were excluded from the analysis.

^†Race was not collected on all the participants.

A secondary outcome, heart rate, was measured at 4 time points: (a) baseline, (b) preinjection, (c) injection, and (d) postinjection. Generalized linear mixed models were used to evaluate the effect of treatment on heart rate across time. To assess the effect of treatment on each time point separately, a linear regression adjusting for baseline heart rate was used, as it was a significant predictor of heart rate at all time points.

Results for the primary outcome were presented as relative risks (IRRs) and 95% CIs, and results for the secondary objective were presented as betas and (change in the outcome per unit of the predictor) CIs. For this study, a p value of <.05 was considered statistically significant. All statistical analyses were conducted using Stata 13.0 (StataCorp).

RESULTS

Sample Group

Age was similar between the 2 study groups (x_exp = 28.1 ± 4 vs x_con = 29.8 ± 5.6) as was the proportion of Hispanics (p_exp = 19.2% vs p_con = 23.1%; Table 1). All subjects in both groups were ASA 1, as specified in the inclusion criteria. There were slight differences between groups with regards to race, with the proportion of white subjects in the experimental group being 26.6 versus 40.0% in the control group. There were more males in the control group (34.6%) than in the experimental group (23.1%). Last, baseline heart rate was lower in the experimental group than in the control group (x_exp = 71.9 ± 8.4 vs x_con = 78.2 ± 1.0). There were no statistically significant differences between the 2 groups with regards to age, ethnicity, race, gender, ASA classification, or baseline heart rate.

Visual Analog Scale

The mean VAS score was 14.2 ± 14.6 mm for the control group and 19.5 ± 19.7 mm in the experimental group. When adjusting for gender, Hispanic ethnicity, and baseline heart rate, VAS scores were statistically significantly higher in the experimental group compared with the control group (IRR = 1.37, 95% CI: 1.19–1.58, p < .001). Other factors also found to be associated with VAS scores were male sex (IRR = 0.60, 95% CI: 0.50–0.71, p < .001), Hispanic ethnicity (IRR = 0.66, 95% CI: 0.55–0.80, p < .001), and baseline heart rate (IRR = 1.02, 95% CI: 1.00–1.02, p < .001). Specifically, males and Hispanics had a lower VAS scores, while higher baseline heart rates were associated with increased VAS scores (Table 2).

Table 2. .

Effect on VAS by Study Group Assignment (n = 52)

Variable	VAS IRR	95% CI	P Value*
Control	Referent	—	—
Experimental	1.37	1.19–1.58	< .001
Gender
Female	Referent	—	—
Male	0.60	0.50–0.71	< .001
Ethnicity
Non-Hispanic	Referent	—	—
Hispanic	0.66	0.55–0.80	< .001
Baseline heart rate	1.01	1.00–1.02	< .001

^*P values were obtained using a generalized linear model with a Poisson family and log link. All variables in the table are mutually adjusted for one another.

Heart Rate

The Figure shows the median heart rate in beats per minute for each group at each time point as measured by the pulse oximeter. There were no statistically significant differences between the 2 groups in heart rate at the preinjection, injection, and postinjection time points when compared with their group's respective baseline (Figure).

Box plot of mean heart rates at different time points. **Differences in heart rate at different time points compared with baseline were not statistically significant.

Adverse Events

AEs were reported in 6 of the 26 participants (23%) in the control group, whereas 19 of the 26 (73%) participants in the experimental group reported an AE (Table 3). AEs reported were tissue sloughing, redness, and ulceration. Tissue sloughing was the most common, occurring in 73.1% of the experimental compared with only 19.2% of the control group (p < .001).

Table 3. .

AEs Among Treatment Groups

Variable	Control, n = 26	Experimental, n = 26	P Value*
Sloughing	5 (19.2%)	19 (73.1%)	<.001
Redness	4 (15.4%)	7 (26.9%)	.50
Ulceration	2 (7.7%)	4 (15.4%)	.69

^*P value obtained by Fisher exact test.

DISCUSSION

In pediatric and general dentistry, providers are constantly searching for newer and better alternatives to deliver a more comfortable pain-free injection. Based on the results of this pilot study, the use of a CTA with 10% lidocaine, 10% prilocaine, and 4% tetracaine topical anesthesia was not found to be superior to 20% benzocaine for reducing local anesthetic injection pain, as assessed using both VAS measurements and alterations in heart rate. In fact, VAS scores were significantly higher in the experimental group and more postoperative issues, such as redness and tissue sloughing, were reported compared with the control group. Changes in heart rate between the 2 groups were not statistically or clinically significant. However, the lack of a significant difference for heart rate may be related to the small sample size enrolled in this pilot study.

Given the favorable results in the literature of similar CTAs, such as Eutecitic Mixture of Local Anesthetics (EMLA) and lidocaine 20%, tetracaine 4%, and phenylephrine 2%, it was initially predicted that the experimental group would have superior pain reduction effects compared with the 20% benzocaine control group. The reported effects of CTAs in the literature have been inconsistent. Several studies have shown favorable results of CTAs.^16–18 Many providers are using CTAs for periodontal and orthodontic procedures in lieu of traditional local anesthetic injections.^11–13,18 In a study by Milani et al¹⁶ and a study by Abu et al,¹⁷ EMLA (2.5% lidocaine and 2.5% prilocaine) was found to be significantly more effective in reducing the pain of maxillary infiltration injections than 20% benzocaine. Reznik et al¹⁸ found that a CTA containing lidocaine 20%, tetracaine 4%, and phenylephrine 2% was more effective than 20% benzocaine in pain reduction during placement of orthodontic temporary anchorage devices. In a study by Primosch and Rolland,¹⁹ EMLA and 20% benzocaine were found to be equal in pain reduction for palatal injections in children. Meanwhile, Tulga and Mutlu²⁰ observed 20% benzocaine to be more effective than EMLA for pain reduction of dental injections in their study of pediatric subjects.

Increased AEs, like tissue sloughing, have not been reported in similar studies using lidocaine and prilocaine CTAs.^16,17 The increase in tissue sloughing may be due to the tetracaine component of the CTA used in this study. Tetracaine is one of the most potent anesthetics of the ester-type anesthetic group, which is widely used in ophthalmology as a topical anesthetic agent. Additionally, it has been cited as being the most irritating of the topical anesthetic agents and most likely to cause contact irritation of the skin or mucosa.²¹ There have been reports of corneal ulceration after abuse of tetracaine as a topical ocular anesthesia.^22,23 Tetracaine at high doses was also found to cause gastric ulcerations in animal studies.²⁴ An important thing to note is that safety of CTA should be taken into consideration as compound agents are not directly regulated by the Food and Drug Administration (FDA).^13,25,26 In this study, the topical gel was not rinsed off with water immediately after its application. Instead the participants rinsed their mouth after the injection was completed. The prolonged contact time of the gel with the oral soft tissues may have contributed to some of the tissue sloughing.

Despite being a common belief, the addition of more component local anesthetic agents fails to equate to increased CTA efficacy. Many alternative topical anesthetics are available on the market; however, benzocaine still remains the most widely used.^3,26 Additionally, topical benzocaine is often used as the standard for comparing or measuring the efficacy of newer, alternative topical anesthetics.^{18–20,27,28} In many studies, benzocaine has been shown to be more effective than placebo,^27,29 while also being equal or superior to many of the newer topical anesthetics.^19,20,27,30 Benzocaine also has many desirable qualities that make it easy to apply in the oral cavity such as solubility, consistency, taste, and low systemic absorption.¹⁹

However, the use of benzocaine is not without risks. Methemoglobinemia is a concern whenever topical anesthetics are used, especially benzocaine. The FDA has also been evaluating the risk of methemoglobinemia associated with the use of benzocaine and lidocaine topical anesthetics. In the 2013 FDA's Adverse Events Reporting system, it identified a total of 375 cases of methemoglobinemia associated with benzocaine.³¹ Methemoglobinemia results from oxidation of iron, causing its inability to effectively bind and carry oxygen, which may result in anemia and hypoxia.¹⁷ Taleb³² stated that benzocaine accounted for 66% of the total cases of methemoglobinemia compared with lidocaine (5%) and prilocaine (28%) in adults undergoing transesophageal echocardiography. Methemoglobinemia associated with topical benzocaine is rather rare, certainly in comparison with the benzocaine spray; however, the spray may have the highest incidence simply because it is the most commonly used formulation in preparation for ENT procedures, which is not so in dentistry.³² The spray may be more commonly associated with methemoglobinemia due to the difficulty in regulating dose/volume during administration, so it may be quite easy to overdose. However, in this study there was not a single participant who experienced methemoglobinemia.

The time of onset of topical anesthetics should also be discussed. In this study, the compounded 20% benzocaine had a time of onset of 1 minute compared with 3 minutes for 10% lidocaine, 10% prilocaine, and 4% tetracaine gel. Both topical anesthetics remained in contact with the gingival mucosa for 60 seconds. This may have affected the study outcomes since the time of onset of 10% lidocaine, 10% prilocaine, and 4% tetracaine gel is slightly delayed by approximately 2 minutes. However, leaving the 10% prilocaine, and 4% tetracaine gel longer may have resulted in increased tissue sloughing and irritation.

Due to restrictions imposed by the associated institutional review board, this study was unable to include a third placebo group to serve as a true control group. Further studies should be done to test 10% lidocaine, 10% prilocaine, and 4% tetracaine and 20% benzocaine against a placebo group. The homogenous and small subject sample size may have limited the ability to detect small effects such as with changes in heart rate. The vast majority of the volunteers were dental students and auxiliaries, which led to a homogenous sample size that caused a lack of generalizability in this study. However, there are also strengths within this study. A randomized clinical trial prevented selection bias and confounding factors. Blinding of both the patient and the clinician limited the bias and placebo effects. There were high retention rates for the participants, all outcomes were measured, and no missing data on outcomes were reported.

Conclusion

In summary, a CTA comprised of 10% lidocaine, 10% prilocaine, and 4% tetracaine was less effective than 20% benzocaine at reducing pain associated with the injection of local anesthesia. The CTA mixture showed higher VAS pain scores and caused a significant increase in tissue sloughing at the injection site compared with 20% benzocaine when applied in oral soft tissues after 60 seconds. In this study, no benefits were seen using a CTA of 10% lidocaine, 10% prilocaine, and 4% tetracaine versus 20% benzocaine prior to maxillary infiltration injections.