High-volume, Multilevel Local Anesthetics-Epinephrine Infiltration in Kyphoscoliosis Surgery: Blood Conservation.

INTRODUCTION: Scoliosis surgery is usually associated with severe bleeding. Various systemic strategies for blood conservation were applied, while the local techniques get less attention. The preemptive use of sufficient volume for proper tissue infiltration at two levels was applied. The local epinephrine may control bleeding without reliance upon deliberate hypotension, permitting a higher tissue perfusion.
MATERIALS AND METHODS: This prospective study included 46 patients scheduled for posterior spinal fusion for scoliosis correction. Patients were randomized into two groups. group I received a cocktail of high volume (100 ml/each 10 cm of wound length) of local anesthetics and epinephrine tumescent infiltration at the subcutaneous (SC) followed by intramuscular level bilaterally. Group C received saline infiltration in the same technique. Statistically, data were analyzed according to its distribution using the t-test, Mann-Whitney, and Chi-square tests as appropriate.
RESULTS: There was a significant reduction in blood loss (38%), reduced blood and fluid transfusion (36% and 23%), and reduced operative time (23%), with higher surgeon satisfaction. The surgical field visibility (Fromme's scale) was much better during SC and muscular dissection in Group I, while it was fair during the bony work. The satisfactory field in spite of higher mean blood pressure in Group I greatly omitted the reliance upon deliberate hypotension.
CONCLUSION: The high-volume multilevel infiltration of epinephrine cocktail can provide a significant blood and operative time conservation in kyphoscoliosis surgery. Copyright:

INTRODUCTION

Since the first spinal fusion by Hibbs in 1911,[1] anesthesia for correction of scoliosis is challenging due to frequently associated comorbidities, extensive nature of the surgery, and liability for several complications, such as blood loss and respiratory and gastrointestinal problems.[2]

Among the major concerns for anesthesiologists are the pain and bleeding. Scoliosis correction accounts for massive blood loss that may exceed more than half of blood volume.[34] In addition, both dilutional and consumptive coagulopathy may augment this problem.[5] The long incision, wide exposure with muscle stripping off the bone, long surgical time, and wide decortications of an osteoporotic vascular bone are important factors correlating with blood loss.[34]

There are many strategies for blood conservation; however, sometimes it may be unsuitable as the autologous donation in small children and acute normovolemic hemodilution because of possible exacerbation of spinal cord ischemia by anemia.[6] Cell savers and erythropoietin for costs and availability.[7] Aprotinin may carry a risk of hypersensitivity reaction, particularly with re-exposure.[8] Furthermore, antifibrinolytics may possess a risk of thromboembolism, seizures, and central retinal vascular thrombosis,[9] and the induced hypotension regards concerns about spinal cord ischemia.[10]

For the purpose of decreasing bleeding and pain in spine surgery, a standard practice is the use of epinephrine and local anesthetics (LA).[11] In spite of wide application, the use of vasoconstrictor infiltration in spine surgery has scarce literature and little evidence.[12] Usually, the infiltration volume is less (20-30 ml).[11] To the authors’ knowledge, this is the first trial of applying the local tumescent technique in kyphoscoliosis surgery.

Our proposal is that proper use of preemptive high-volume multilevel epinephrine cocktail infiltration may offer bleeding control without the reliance upon deliberate hypotension. This technique may benefit all spine surgery patients.

MATERIALS AND METHODS

This prospective randomized double-blind study was executed from November 2016 to November 2017, including patients scheduled for posterior spinal fusion for kyphoscoliosis correction; patients aged from 8 to 18 years, and patients of American Society of Anesthesiologists Status I and II, after approval by the local institutional review board (R/17.02/85) and clinical trial registration (NCT03319563). Written consents were obtained.

A priori G-power analysis was used to estimate the study sample size (G*power version 3.1: Faul, Erdfelder, Lang, and Buchner, Germany). A preliminary study was performed on ten patients. the results were; mean estimated blood loss (EBL) (1620 vs. 970 ml), standard deviation (940 vs. 410), and an effect size 0.89. A power of 80%, with Type I error of 0.05, yields a total sample size of 42 patients. A drop out of 10% of cases was expected. Therefore, a total number of 46 cases was enrolled (23/group).

The exclusion criteria were parent's refusal, surgical site infection, implants’ removal, hypersensitivity to the involved drugs, coagulopathy, blood diseases, severe hepatic, cardiac, respiratory, renal, and cognitive impairment, or intensive care admission.

All patients were evaluated clinically for history, examination, and standard investigations. Pulmonary function test, arterial blood gases analysis, and echocardiography were requested as indicated. The degree of spinal curvature was assessed by the Cobb angle. It is the angle between the perpendiculars to the lines parallel to the surface of the upper and lowermost vertebrae of the curve. Scoliosis surgery is required with 40°–50° or more. The flexibility index is calculated from the equation: (Cobb angle − bending Cobb angle)/Cobb angle. This index of <50% denotes rigidity.[13] The visual analog score (VAS) for pain assessment and the possible wake up test were explained. Lactated Ringer's 10 ml.kg−1 started the morning of the operation.

After establishing standard monitoring (electrocardiography, noninvasive blood pressure, pulse oximetry, urine output, and end-tidal carbon dioxide), anesthesia was induced by propofol 2 mg/kg and fentanyl 1 microgram (μg).kg−1. Endotracheal intubation was facilitated by atracurium (0.5 ml.kg−1). The peak airway pressure was recorded after establishing a prone position. Anesthesia was maintained by isoflurane in a mixture of air and 40% oxygen. Tranexamic acid 15 ml.kg−1 intravenous (Kapron® Amoun Pharmaceutical Co., Egypt) was given to all patients.

The infiltration technique

After measuring the marked wound length in centimeters, an anesthetist not involved in the study prepared the infiltration cocktail. The cocktail contained 2 ml.kg−1 plain bupivacaine 0.5%, and lidocaine 5 ml.kg−1 for analgesia, plus epinephrine 5 mcg/ml in a total volume of 100 ml/10 cm of the wound length. The assessors of intra- and post-operative data were blinded for the cocktail. That was replaced by normal saline in the same volume in the control group.

The infiltration was done at two levels:

Subcutaneous (SC) level – the skin infiltration was done at the mid-line by the surgeon 10 min before skin incision. The allowed volume is 30–40 ml/10 cm of the wound length

Muscular level – the paramedian intramuscular infiltration was done bilaterally before the incision of thoracolumbar fascia [Figure 1]. The infiltration extended from superficial to deep level on the vertebral lamina. Then, muscle dissection off the vertebrae was done [Figure 2]. The infiltration volume is 20–30 ml/10 cm/side of the same cocktail.

Figure 1

Intramuscular infiltration under the thoracolumbar fascia

Figure 2

Muscle dissection

The patients were randomly allocated using opaque-sealed envelopes into two groups:

Control Group (C) – n = 23: Patients received general anesthesia and then preemptive infiltration with the blindly prepared solution of normal saline in the same manner as described above

Infiltration Group (I) – n = 23: Patients received general anesthesia, and then, the infiltration was done using bupivacaine, lidocaine, and epinephrine as described above.

The operative field visibility was judged by 0–5 Fromme's scale as follows:[14]

5: Massive uncontrollable bleeding requires a constant suctioning so surgery is impossible

4: Heavy but controllable bleeding requires prompt suctioning and interferes with dissection

3: Moderate bleeding requires frequent suctioning and moderately compromises dissection

2: Moderate bleeding requires occasional suctioning and no interference with accurate dissection. The surgical field is not threatened

1: Mild bleeding requires no suctioning and no surgical nuisance

0: No bleeding, virtually bloodless field.

The Fromme's scale was assessed by the surgeon at four occasions: (1) after SC dissection, (2) after bilateral muscle dissection, (3) after screws’ application, and (4) after osteotomies and decortication.

Hypertension and tachycardia defined as 25% higher values of mean blood arterial pressure (MAP) and heart rate than the basal values for 5 min. Hypertension was managed by increasing isoflurane concentration up to 2%, fentanyl 0.5 μg.kg−1 increments. Nitroglycerine 0.5–10 Nitroglycerine 0.5–10 μg.kg−1.min−1 (Nitronal® Merck Serono Ltd., UK., 1 mg.ml−1) was used if the Fromme's operative visibility scale ≥3 provided that the minimum MAP is not <55 mmHg. Resistant tachycardia and hypertension were managed additionally by diluted propranolol 1 mg. Hypotension (MAP <55 mmHg) was managed by discontinuation of nitroglycerine, decreasing isoflurane concentration, proper preload compensation. and ephedrine 6 mg increments if not responding. Ringer's acetate was applied according to the following equation:

Infusion volume (mL) = EBV × (Ho − H1)/H1 × 100/80

where EBV is the estimated blood volume, Ho is the basal hematocrit (Hct), and H1 is the target Hct of 25%.[15] The blood loss is compensated as triple its volume by Ringer's acetate to a maximum 30% of the EBV, and then, blood was transfused or started after decortication. Furosemide was used after a blood transfusion.

Side effects as arrhythmia and hypertension during infiltration were recorded and managed as appropriate.

Data were statistically analyzed using SPSS software version 17 (SPSS, Inc., Chicago, IL, USA). All variables were tested for normality using the Shapiro–Wilk test. Parametric data were compared by t-test. The Mann–Whitney and Chi-square tests were applied for nonparametric and ordinal variables. The significance level of P value was ≤0.05.

The primary outcome was the intraoperative EBL calculated from the suction after extraction of the irrigation fluids and towel's weighting.[3] The other secondary outcomes were the number of transfused blood units, the intraoperative utilized dose of nitroglycerine, the hemodynamics (MAP), and the transfused fluid volume. Surgeon satisfaction regards the surgical field at the end of the operation using a visual analog scale (VAS) of 0–10, 10 being the best. The incidence of wound complications within 2 weeks was infection, dehiscence, seroma, hematoma, and bleeding.

RESULTS

This study included 49 patients. Three patients were excluded due to intensive care admission [Figure 3]. The demographic and hematological data were not different between the groups [Table 1]. Furthermore, the surgical data including the types of scoliosis, degree of rigidity, the extent of fusion, and osteotomies were not significantly different [Table 2].

Figure 3

The study flowchart

Table 1

The demographic and hematological

	Group C	Group I	P
Age (years)	13.2±4.2	15±2.2	0.097
Female/male sex (%)	15/8 (65/35)	16/7 (70/30)	0.862
BMI (kg.m−2)	27.1±5.6	24.9±3.8	0.127
Body weight (kg)	58.5±15	48.5±10	0.055
Preoperative
Hemoglobin (g.dl−1)	12.8±1.6*	11.1±1.7	0.029
Hct	39.4±7	34.2±3.5	0.168
Platelet count	255±59	311±58	0.073
White blood cell count	7.67±1.4	7.5±2.5	0.880
Postoperative
Hemoglobin (g.dl−1)	10.4±2.2	10.8±1.0	0.695
Hct	30.3±6.7	32.2±2.3	0.741
Platelet count	150±44	176±74	0.574
White blood cell count	12.3±1.9	15.5±3.0	0.166

*Significant difference between the groups P<0.05. Data are in mean±SD or n (%), n=23. Hct=Hematocrit, BMI=Body mass index, SD=Standard deviation

Table 2

The surgical and operative data

	Group C	Group I	P
Type of scoliosis (%)
Idiopathic	19 (82)	17 (74)	0.764
Congenital	2 (9)	2 (9)
Neuromuscular	0 (0)	1 (4)
Neurofibromatosis	2 (9)	3 (13)
Cobb angle (°)	59.3±9.8	63±12.4	0.604
Flexibility index (%)	40.3±7.1	39.8±8.6	0.839
Fused levels (n)	12.3±1.4	12.8±1.7	0.350
Osteotomies (n)	3±2.8	3±3.2	0.898
Infiltration volume (ml)	420±40	402±32	0.103
Nitroglycerine consumption (ml)	5.7±3.2*	0.4±1.0	0.000
Operative duration (h)	5.2±1.2*	4.0±1.1	0.002
Surgeon satisfaction score (0-10)	6.05±0.7*	7.95±0.8	0.002
Peak airway pressure (cm H2O)	19.8±3.5	21.8±6.3	0.335

*Significant difference between the groups P<0.05. Data are in n (%) or the mean±SD n=23. SD=Standard deviation

This study showed a significantly lower EBL in Group I than Group C (38% difference in mean values). In addition, there was 36% reduction in blood transfusion, and 22% less fluid administration in Group I [Table 3]. The Fromme's scales were lower in Group I denoting better field visualization especially during SC and muscular dissection than during bony work and instrumentation [Table 3]. The blood loss was about 100–200 ml after SC and bilateral muscular dissection. The operative time was 23% shorter in Group I, and surgeon satisfaction was higher [Table 2]. The intraoperative MAP was higher in Group I [Figure 4]. There was no arrhythmia or hypertension during infiltration. Two patients – one in each group – presented by wound seroma that was drained. One patient in Group C showed a superficial wound infection that was treated with antibiotics.

Table 3

Intraoperative blood conservation data

	Group C	Range	Group I	Range	Difference (%)	P
Fromme’s field visibility score (0-5)
Subcutaneous dissection	2.2±0.6*	1-3	0.3±0.5	0-1	86	0.003
Muscle dissection	3.5±0.6*	2-4	1.4±0.6	1-3	58	0.011
Screws fixation	3.6±0.6*	2-5	3.0±0.6	2-4	17	0.012
Osteotomy	3.7±0.7*	3-5	3.1±0.6	2-4	16	0.042
Estimated blood loss (ml)	1346±691*	700-4200	830±461	300-2000	38	0.005
Blood transfusion (unit)	2.8±1.4*	1-8	1.8±1.1	0-4	36	0.010
Fluid transfusion (l)	3.8±1.2*	2.5-7	2.9±4.3	2-3.5	23	0.003
Urine output (ml)	762±471	300-2000	772±334	300-1500	-	0.938

*Significant difference between the two groups P≤0.05. Data in mean±SD, Range=Minimum-maximum, Difference=The percent difference between the means of two groups. SD=Standard deviation

Figure 4

Intraoperative mean blood pressure changes in the studied groups. *Significant difference between the groups, P < 0.05

DISCUSSION

This study showed that the high-volume multilevel infiltration technique produced a significant reduction of intraoperative blood loss (38%). Generally, scoliosis surgery is associated with severe bleeding.[16] Möller and Hedlund found that the average blood loss during spinal fusion is 1.5 liter but can be severe as much as a 7 liters loss.[17] The intraoperative mean blood loss in the control group in this study was 1346 ml. Kolban et al. reported a loss about 1500 ml,[18] Guay et al. reported 2000 ml loss,[19] while it was 1250 ml in patients who received an aminocaproic acid infusion.[20]

In this study, the blood transfusion was 36% lower in Group I. The median value of blood transfusion reduced from 3 to 2 units. Usually, the blood transfusion is around 3–6 units in a similar category of patients.[21] The predictors of blood transfusion may include operative time, number of vertebral fused levels, male sex, MAP, platelet abnormalities, and coagulopathy.[22] Actually, there is no perfect preoperative screening method to identify patients with a higher risk of bleeding.[23] There is an association between the excessive bleeding during spine surgery and increased complications such as wound infection, implant failure, neurological problems, renal failure, and longer hospital stay.[1624] In addition, there are the risks of transfusion reactions, transmission of infection, and immunomodulation.[25]

The properly dry field permitted higher MAP in our patients. This beneficial effect agrees with the trend of less emphasis on hypotension for blood conservation in scoliosis surgery, while the reliance is toward cell salvage and antifibrinolytics.[6] According to this study, we can add the local infiltration technique as an effective method for blood conservation.

Deliberate hypotension can be associated with reduced blood loss by 33%.[26] However, MAP <60 mmHg during correction showed significant changes in somatosensory-evoked potential with a potential risk of cord ischemia.[27] Hence, an MAP >65 mmHg during dissection and a higher pressure goal of 75 mmHg are recommended during correction.[28] In Group I, there was almost no demand for nitroglycerine. The favorable surgical field allowed higher MAP that permits better tissues’ perfusion.

Another benefit of infiltration in this study is the reduced operative time by 23%. The operative duration is a determinant of blood loss and short-term morbidity and mortality.[22]

The secret of success in blood conservation in this study may originate from using preemptive, high-volume infiltration that achieved a good distribution along the long incision at multiple levels. In addition, the administration of diluted LA with epinephrine in a large volume tensing the injected area is the principle of tumescent anesthesia that has a different pharmacokinetic profile. The blood loss is reduced by both the vasoconstrictor effect of epinephrine and the hydrostatic action of tumescence.[29] The tumescent mechanical expansion allows the use of higher doses with reduced systemic toxicity and prolonged local drug effects.[30] Consequently, epinephrine altered the LA pharmacokinetics. The peak plasma level of lidocaine is shifted from 2 to 18 h.[31] Another benefit of epinephrine is the antagonism of the vasodilator action of LA, thus reducing blood loss.[32] A low concentration of epinephrine (0.6–1 mg.l−1) is effective for hours with the low incidence of side effects.[33] However, Guay et al. found no apparent relation between the infiltration dose and blood loss.[19] For safety, the maximum recommended dose is limited to 0.07 mg.kg−1.[34] Again, the tumescence showed a peak effect of epinephrine at 3 h after SC infiltration, with nearly four times the physiological level, but without signs of toxicity.[35] Guay et al. could not demonstrate a clear benefit of infiltration that epinephrine 1:1,000,000 in a volume of 20–250 ml did not correlate with the blood loss, but their infiltration was limited to the SC level.[19]

The surgical field was better during SC and muscular dissection so that blinding using epinephrine-free saline was easy detectable like a red and white watermelon. Most of the blood loss occurred during the bony work. At this level, the sources of bleeding are the vertebral and epidural vasculature and the cancellous bone that is not amenable to the action of epinephrine. We used the antifibrinolytic tranexamic acid in all patients to reduce blood loss. In addition, topical hemostatic agents containing collagen, thrombin, and fibrin could be helpful for control of blood losses from bony sites.[36]

The fluid administration was not markedly decreased as much as the blood loss (22% vs. 38%); using sufficient fluids, there was low urine output that may be related to excess antidiuretic hormone rather than hypovolemia.[37] Furosemide was frequently used in our patients late in the operation to avoid the complications of large crystalloid volume utilization.

In this study, the peak airway pressure was not different in both groups. Avoiding the abdominal compression in the prone position reduces the degree of bleeding.[38]

The infiltration of the epinephrine cocktail was safely applied in this study. However, the lake of determination of the optimal concentration, the duration of action, and the pharmacokinetics of epinephrine may represent a limitation. Furthermore, we did not assess the effect of this technique on postoperative blood loss.

CONCLUSION

The high-volume multilevel infiltration of epinephrine cocktail can provide a significant blood and operative time conservation in kyphoscoliosis surgery.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.