The Effect of Normal Saline and Plasmalyte on Acid-Base Status in Patients Undergoing Head-and-Neck Surgery with Free Flap Reconstruction: A Prospective, Observational Cohort Study.

Background: Intraoperative fluid strategy may affect the graft viability in head-and-neck surgeries with free flap reconstruction (HNS-FFR). Studies to guide regarding association of intraoperative fluid with metabolic parameters during such surgeries are infrequent. Aim: This study aimed to compare plasmalyte (PL) and normal saline (NS) (0.9%) in terms of acid-base balance and electrolytes in the peri-operative period along with graft viability during above-mentioned surgeries. Settings and Design: Prospective, observational cohort study was conducted in patients, 18-65 years, undergoing HNS-FFR at a tertiary care center. Materials and
Methods: The cohort was categorized into two groups based on the intraoperative fluid used, i.e., PL (Group A) and NS (Group B) group. The primary objective was to compare arterial blood gas parameters at seven time points till the 3rd postoperative day. We studied the effect on graft viability and length of hospital stay. Statistical Analysis Used: The independent t-tests, Chi-square, or Fisher's exact test were used to evaluate the categorical variables with a repeated measures analysis of variance for inter-group comparison with P < 0.05 as significant.
Results: Seventy-one (36 in Group A and 35 in Group B) patients were included in the study with comparable baseline characteristics. Group A had a better acid-base status, especially after the conclusion of vascular anastomosis (pH 7.37 ± 0.06 vs. 7.33 ± 0.04, P = 0.014) and in the postoperative period (pH 7.35 ± 0.07 vs. 7.31 ± 0.05, P = 0.013). No statistically significant difference was observed in outcome parameters between the groups. Conclusions: PL may be preferred over NS due to better metabolic milieu during HNS-FFR surgery. Copyright:

INTRODUCTION

Anesthetic management of head-and-neck surgeries with free flap reconstruction (HNS-FFR) is challenging, especially with regard to intraoperative fluid management.[12345] Although proper fluid management is essential, the debate over correct perioperative fluid therapy remains unresolved.

The choice of intravenous fluids during surgery is guided by proper understanding of the fluid's properties. Normal saline (NS), an isotonic solution, is a readily available, cheap, and frequently used crystalloid.[678] Multiple studies have supported the use of plasmalyte (PL), a balanced isotonic solution, in critically ill and hemorrhagic shock patients where massive crystalloid infusion is required for resuscitation. Still, there is a paucity of data to validate use of any one crystalloid over the other in prolonged surgeries like HNS-FFR to change practice.

In this study, we hypothesized that NS and PL may affect the acid–base status in HNS-FFR. Thus, we have conducted this prospective observational study to compare the effects of two types of crystalloids (i.e., PL and NS) administered during HNS-FFR on metabolic profile during and after surgery. Further, our secondary objective was to compare outcomes in terms of blood loss, urine output, and graft viability assessed by Doppler flow, length of intensive care unit (ICU), and hospital stay between the two groups.

MATERIALS AND METHODS

This prospective, single-center, observational, cohort study was conducted in a tertiary care institute from January 01, 2016, to April 30, 2017, after approval from Institutional Ethics Committee (approval number-MICR-551/2015). We included all patients, aged 18–65 years undergoing head-and-neck surgery with free flap reconstruction after getting written informed consent for study. The exclusion criteria were presence of any preexisting acid–base disorders, chronic renal failure (glomerular filtration rate - <60%), liver diseases, severe cardiac diseases (ejection fraction – <40%), and need for intraoperative tourniquet application. We also excluded any patient where group allocation was not possible as per the study design. All data including demographics (age, gender), comorbidities, and clinical and laboratory findings were recorded.

Preoperative fluid deficit was replaced by 10–20 mL.kg−1 of crystalloid, followed by 4–8 mL.kg−1.h−1 as maintenance. In case of bleeding (<750 mL) or mean arterial pressure decrease (>20% of baseline), 250 mL boluses of 6% hydroxyethyl starch were given. If bleeding was profuse with hemoglobin (Hb) <7 g.dL−1, packed red blood cells were transfused to maintain Hb >8 g.dL−1. Since this was an observational study, the type of intravenous crystalloid was decided on the discretion of the consultant anesthesiologist. However, depending on the majority of the type of crystalloid given during surgery, the patients were allocated by the study investigator into two groups. In case more than 75% fluid administered was PL, the patient was assigned to Group A, whereas when more than 75% fluid was NS, the patient was assigned to Group B.

As a protocol, even prior to this study, anesthetic management was standardized in the HNS-FFR. In the operation theater, standard monitors (electrocardiogram [ECG], noninvasive blood pressure monitoring [NIBP], and peripheral capillary oxygen saturation [SpO2]) were attached and baseline parameters were noted. A wide bore intravenous access was attained after local anesthetic infiltration.

Standard intravenous induction (midazolam 0.01 mg.kg−1, fentanyl 2 μg.kg−1, propofol 2 mg.kg−1, atracurium 0.5 mg.kg−1) was done in all patients followed by tracheal intubation with a suitable sized nasotracheal tube. Postinduction, invasive arterial monitoring, and arterial blood gas analysis were done as a routine in such long-duration surgeries. Anesthesia was maintained on isoflurane (minimum alveolar concentration of 0.7–1.0) in oxygen and air mixture with low flow ventilation adjusted to maintain end-tidal carbon dioxide between 30 and 40 mmHg throughout the procedure.

Samples for arterial blood gas monitoring were drawn at the following time points (T): T1 – immediate postinduction; T2 – 4 h postinduction; T3 – anastomosis complete (blood flow is established); T4 – surgery completed (last suture); T5 – 12 h after the end of surgery; T6 – 36 h after the end of surgery, and T7 – 60 h after the end of surgery.

We noted changes in pH, base excess (BE), serum sodium (Na+), serum potassium (K+), serum bicarbonate (HCO3−), serum chloride (Cl−), and lactates at these predefined time points.

During surgery, blood loss, blood products transfused (if any), volume of intravascular fluids, and urine output were noted. The total duration of surgery was noted. Postoperative elective ventilation was done in intensive care unit for 1 day followed by spontaneous breathing trial and extubation. Follow-up visits were continued till 3rd postoperative day with assessment of vitals, arterial blood gas analysis, urine output, and graft vascularity in first 3 postoperative days. Assessment of graft vascularity was done by Doppler by the radiologist and surgical team. Graft vascularity was documented as good or compromised as a surrogate marker of graft viability. The total length of intensive care unit and hospital stay (in days) was noted in all patients.

Statistical analysis

Data were entered, cleaned, and coded into Microsoft Excel 2013. Data analysis was done using IBM Statistical Package for the Social Sciences (SPSS) version 24.0, Armonk, New York. Quantitative data were expressed in terms of means and standard deviation. Qualitative (categorical) data were expressed as absolute numbers and proportions. Patient characteristics were analyzed using independent t-tests and categorical variables were evaluated using Chi-square or Fisher's exact test. Variables between the groups and within each group were compared by using a repeated measures analysis of variance. P < 0.05 was considered significant.

RESULTS

During the study period, a total of 82 patients underwent HNS-FFR. However, only 75 met the inclusion and exclusion criteria of the study. Further, four patients were excluded as group allocation was not possible due to equal use of Saline and PL. A total of 71 patients were included in the final data analysis, with 36 patients in Group A and 35 patients in Group B. The two groups were comparable in age (52.5 ± 12.1 years in Group A vs. 46.5 ± 13.7 years in Group B, P = 0.105) and gender (male: female: 32:4 in Group A, vs. 32:3 in Group B, P = 0.690) [Table 1]. The study population was comparable in terms of baseline vitals (blood pressure and heart rate) and comorbidities (diabetes mellitus, hypertension, coronary heart disease, and thyroid diseases) [Table 1]. All the 71 patients enrolled successfully completed the study. Based on the arterial blood gas analysis done at predefined points, we observed the following parameters.

Table 1

Baseline characteristics of patients included in the study

Parameter	Overall	Group A (PL)	Group B (NS)	P
Age (years)*	50±11.64	51.5±12.1	46.5±13.7	0.105
Gender (male/female)	64/7	32/4	32/3	0.690
Comorbidities
Diabetes (%)	31.4	25.7	31.4	0.303
Hypertension (%)	70	77.1	62.9	0.192
Thyroid disorder (%)	10	8.6	11.4	0.690
CAD (%)	7.1	11.4	2.9	0.164

*Mean±SD. P<0.05 is considered significant. CAD=Coronary artery disease, PL=Plasmalyte, NS=Normal saline, SD=Standard deviation

Metabolic parameters

The metabolic profile was more favorable in Group A. The pH was statistically lower in NS group as compared to PL group at completion of anastomosis (7.37 in Group A vs. 7.33 in Group B, P = 0.014) and at the completion of the surgery (7.35 in Group A vs. 7.31 in Group B, P = 0.013) [Table 2]. However, the pH values were within the normal physiological range in both groups [Figure 1].

Table 2

Descriptive statistics of patients for metabolic parameters, i.e., pH, base excess, and serum bicarbonate in Group A (plasmalyte) and Group B (normal saline)

	Group A-pH†	Group-pH†	P-value of pH	Group A BE (mEq/L)†	Group B BE (mEq/L)†	P-value BE	Group A HCO3-(mEq/L)†	Group B HCO3-(mEq/L)†	P-value HCO3-
T1	7.38±0.05	7.37±0.04	0.213	−0.94±2.44	−1.61±2.15	0.230	23.63±2.26	22.99±1.99	0.216
T2	7.39±0.05	7.36±0.04	0.056	−1.40±2.32	−2.27±2.11	0.103	23.07±2.28	22.20±2.09	0.099
T3	7.37±0.06	7.33±0.04	0.014*	−2.57±2.52	−3.80±2.52	0.045*	21.88±2.52	20.96±2.33	0.116
T4	7.35±0.07	7.31±0.05	0.013*	−3.13±3.20	−5.45±2.93	0.002*	21.38±3.12	19.51±2.76	0.010*
T5	7.36±0.06	7.35±0.06	0.265	−2.67±2.83	−3.99±2.84	0.056	21.87±2.84	20.67±2.79	0.079
T6	7.37±0.06	7.36±0.06	0.459	−2.11±2.41	−2.60±2.99	0.446	22.54±2.21	21.86±2.86	0.270
T7	7.38±0.05	7.37±0.05	0.184	−1.23±1.98	−1.79±2.07	0.249	22.67±4.32	22.63±1.89	0.966

*P<0.05 is considered significant, †Mean±SD. BE=Base excess, SD=Standard deviation, HCO3 −=Serum bicarbonate

Figure 1

Comparison of pH between the study groups

The BE values were statistically insignificant between the two groups immediately after induction (BE of −0.94 in Group A vs. −1.61 in Group B, P = 0.230) and at 4 h postinduction (BE of −1.40 in Group A vs. −2.27 in Group B, P = 0.103). However, a significant difference was observed in the BE values at the end of anastomosis completion (BE of −2.57 in Group A vs. −3.80 in Group B, P = 0.045) and at the completion of the surgery (BE of −3.13 in Group A vs. −5.45 in Group B, P = 0.002) [Figure 2]. The BE values were comparable on postoperative days 1, 2, and 3 [Table 2].

Figure 2

Comparison of base excess between the study groups

There was a significant difference between the two groups in bicarbonate values at the completion of the surgery (P = 0.010), being higher in PL group (21.38 ± 3.12) as compared to NS group (19.51 ± 2.76) [Table 2].

The lactate levels were comparable in the study, except at T6 (postoperative day 2) when lactate level was significantly high (1.11 ± 0.66 in Group A vs. 0.72 ± 0.38 in Group B, P = 0.042) in PL group.

Electrolytes

Both the study groups showed an increasing trend in serum sodium levels throughout the surgery. Although sodium levels were raised comparatively more in PL group as compared to NS group, it was statistically insignificant throughout the study [Table 3].

Table 3

Descriptive statistics of patients for electrolytes, i.e., serum sodium, serum potassium, and serum chloride in Group A (plasmalyte) and Group B (normal saline)

	Na+ (mEq/L)			K+ (mEq/L)			Cl- (mEq/L)
	Group A†	Group B†	P	Group A†	Group B†	P	Group A†	Group B†	P
T1	134.47±4.66	132.6±4.92	0.107	4.03±0.43	3.95±0.49	0.452	105.20±6.23	104.14±5.36	0.450
T2	133.68±3.67	132.94±4.13	0.321	4.16±0.64	4.05±0.50	0.419	106.26±5.60	105.69±5.17	0.659
T3	134.74±3.62	133.81±4.90	0.308	4.11±0.52	4.20±0.58	0.482	106.77±4.74	107.69±4.64	0.418
T4	135.63±3.57	134.81±4.90	0.426	4.19±0.58	4.30±0.65	0.476	106.97±4.71	108.60±5.52	0.189
T5	135.74±3.18	136.65±3.35	0.250	4.15±0.66	4.10±0.60	0.737	106.83±4.85	108.00±4.75	0.311
T6	136.98±3.46	137.17±3.45	0.812	4.08±0.52	3.88±0.52	0.103	107.0±3.96	107.25±4.82	0.814
T7	137.52±3.21	137.40±3.69	0.885	3.94±0.57	3.96±0.48	0.843	106.29±5.05	105.20±5.36	0.386

P<0.05 is considered significant, †Mean±SD. SD=Standard deviation, Na+=Serum sodium, K+=Serum potassium, HCO3 −=Serum bicarbonate, Cl−=Serum chloride

The baseline serum potassium (K+) levels were comparable in both the groups. Surprisingly, Group B had shown a slight rise in K+ levels (insignificant) intraoperatively, which normalized after surgery [Table 3].

The baseline mean serum chloride (Cl−) levels were similar in both the study groups. Patients receiving NS showed increasing levels of Cl−, which, however, was insignificant from the PL group [Table 3].

Respiratory parameters

There was no significant difference in pO2 and pCO2 values at any timepoint in both the study groups.

Other parameters

The total volume of crystalloid administered during surgery was comparable in both the study groups. The urine output at the end of surgery, on postoperative days 1, 2, and 3, and amount of blood lost during surgery in both the study groups were also comparable [Table 4].

Table 4

Descriptive statistics of patients for total crystalloid used intraoperatively, urine output intra- and postoperatively, intraoperative phenylephrine used, intensive care unit stay, and hospital stay in Group A (plasmalyte) and Group B (normal saline)

	Group	Mean±SD	P*
Intraoperative total crystalloid used (mL)	A-PL	3445.71±588.79	0.322
	B-NS	3565.71±632.89
Intraoperative urine output (mL)	A-PL	952.86±263.45	0.194
	B-NS	875.43±303.88
Urine output POD 1 (mL)	A-PL	1858.97±581.73	0.765
	B-NS	1824.85±338.83
Urine output POD 2 (mL)	A-PL	1299.57±371.74	0.316
	B-NS	1391.60±389.68
Urine output POD 3 (mL)	A-PL	1026.29±243.28	0.250
	B-NS	1093.43±241.29
Intraoperative total phenylephrine used (mg)	A-PL	0.950±0.398	0.012*
	B-NS	1.139±1.185
ICU stay (days)	A-PL	2.09±0.288	0.326
	B-NS	2.06±0.236
Hospital stay (days)	A-PL	7.09±0.288	0.942
	B-NS	7.09±0.284

*P<0.05 is considered significant. POD=Postoperative day, SD=Standard deviation, ICU=Intensive care unit, PL=Plasmalyte, NS=Normal saline

Cumulative amount of phenylephrine required intraoperatively was higher in NS group as compared to PL group (1.139 ± 1.18 mg vs. 0.95 ± 0.39 mg, P = 0.012) [Table 4].

Outcome parameters

The graft viability, ICU stay postsurgery (2.09 ± 0.288 days in Group A vs. 2.06 ± 0.236 days in Group B, P = 0.326) and the total hospital stay (7.09 ± 0.288 days in Group A vs. 7.09 ± 0.284 days in Group B, P = 0.942) were comparable in both the study groups [Table 4].

DISCUSSION

This prospective, single-center, observational cohort study was conducted to compare the effects of NS and PL on the metabolic status of patients who underwent HNS-FFR, in our tertiary care institute. It was found that infusion of PL was associated with a more favorable metabolic profile at the time of completion of anastomosis and at the end of surgery as compared to NS. However, outcomes in terms of graft viability, urine output, and blood loss were similar.

Hypovolemia, improper crystalloid use, acidosis, and alkalosis can affect flap functioning.[9101112] Varying crystalloids have different chemical compositions and can cause differing metabolic consequences. Hence, the choice of intravenous replacement fluid is important. NS and PL are the two crystalloids used commonly during surgery. NS is an isotonic crystalloid solution containing Na+ and Cl− with an in vitro osmolality of 287 mOsmol.kg−1.[7] However, the nonphysiological ion content and the lack of buffering capacity may lead to complications of metabolic acidosis.[8] On the other hand, PL is a balanced solution, with a composition akin to that of plasma with an osmolality of 271 mOsmol.kg−1, making it an ideal intraoperative crystalloid solution.[8913] However, there is a paucity of literature to validate this assumption in real-time studies on patients undergoing long-duration surgeries like HNS-FFR, where graft viability may be affected by the metabolic milieu of the body.

The most important and determining factor for the success of graft is the anesthetic technique used due to its role in hemodynamic stability and regional blood flow.[101112] As a protocol in our tertiary care institute, the anesthetic regimen was standardized in this group of patients in a protocolized manner even prior to initiation of this study to ensure that anesthetic factors had minimal effect, if any, on the graft viability. Thus, the anesthetic agents used and the ventilation parameters maintained during surgery were consistent in all patients.

In NS group, a declining trend was seen in pH values from time of induction (T1) to completion of surgery (T4), with the development of acidosis at completion of anastomoses and surgery. However, this acidosis resolved in the postoperative period and a rising trend was seen in the first 3 postoperative days. In contrast to this, no variation was observed in the pH throughout the study period in the PL group. This is consistent with the study done by Williams et al. in which they concluded that the use of NS in healthy volunteers leads to a decrease in pH levels attributed to metabolic acidosis.[14] Kim et al. also demonstrated a significantly lower value of pH during the postreperfusion period with the use of NS during living donor kidney transplantation.[15] Bruegger et al. also demonstrated the development of intraoperative acidosis, which resolved spontaneously within 24 h, thus emphasizing the transient nature and short longevity of acidosis.[16] The mechanism for this can be explained using the Stewart hypothesis through the strong anion effects of chloride.[1718] BE declined (commonly presenting as base deficit) during surgery in both the study groups, with a greater decline in saline as compared to PL group. This difference was statistically significant at the completion of anastomosis and surgery. It was at this point of time, most of the fluid volume had been infused and cumulative acidity of intravenous fluids was then manifested. Furthermore, in this phase of surgery, mean arterial perfusion pressure was on the lower side which may have led to metabolic acidosis. The acidosis resolved in the first 3 days after surgery, possibly due to hemostasis, normalization of fluid status, emergence from anesthesia, and good graft function. These findings are consistent with the study done by Kim et al. which showed significantly lower values of BE during postreperfusion period in NS group during living donor kidney transplantation.[15]

Changes in ventilation strategy may contribute to changes in bicarbonate level. However, in our study, as per the protocolized anesthesia management, ventilation was tightly controlled throughout the surgery, maintaining end-tidal carbon dioxide (ETCO2) levels within a narrow range. Hence, the presence of respiratory compensation for metabolic changes was ruled out. A similar decrease in bicarbonate levels has been reported in patients given saline during major hepatobiliary or pancreatic surgery by McFarlane and Lee.[19] Contribution to bicarbonate levels in PL group may also have been due to the acetate (27 mmol.L−1 present in PL). Acetate has been shown to be metabolized effectively, regardless of liver failure or severe shock. Acetate can be metabolized in several extrahepatic tissues, including the muscles, brain, myocardium, and renal cortex because they all have the required enzymes.[202122]

A rise in lactate values was seen intraoperatively which showed a similar trend in both the study groups. Lactate, a marker of anaerobic metabolism, is increased in conditions of regional hypoperfusion and hypoxia.[23] Hence, this intraoperative rise in lactate values can be explained by intraoperative blood loss and low blood pressures.

Both the groups showed an increasing trend in mean serum sodium levels throughout the study, which, though more in Group B (NS), was statistically insignificant. However, significant hyperkalemia was noted in patients receiving saline in spite of comparable initial baseline levels. The mechanism is presumably through an extracellular shift of potassium caused by acute changes in blood hydrogen ion concentration, which occurs in association with hyperchloremic metabolic acidosis.[1920]

Hyperchloremia was observed in NS group which is not surprising, as saline with its higher chloride content as compared to PL (154 mEq.L−1 vs. 98 mEq.L−1) will result in hyperchloremia.[2425] The hyperchloremia was statistically insignificant in both the groups in our study. Young et al. observed significantly higher and persistent hyperchloremia even till 24 h after surgery after infusion of NS in comparison to PL A for resuscitation in trauma patients.[26] The volume of fluid infused was higher in this study of trauma resuscitation as compared to our study, which may explain this difference.

The cumulative amount of phenylephrine utilized intraoperatively was significantly more (P = 0.012) in NS group as compared to PL group, indicating more requirement of vasopressors, possibly due to metabolic acidosis in saline group. This is consistent with the finding of Potura et al., who observed higher catecholamine requirements in saline group as compared to acetate-buffered balanced crystalloid solution in patients with end-stage renal disease undergoing cadaveric renal transplantation.[27]

However, though the infusion of PL resulted in more favorable metabolic parameters, especially after the conclusion of vascular anastomosis and in the postoperative period, there was no effect on outcome in terms of graft viability, renal function as assessed by urine output, and postoperative morbidity with similar ICU and hospital stay. This is consistent with the study done by Kim et al., who observed that though NS infusion results in significant metabolic changes as compared to PL in renal transplant recipients, more specifically during postperfusion period, early postoperative graft functions in terms of serum creatinine, urine output, and graft failure requiring dialysis were not significantly different between the groups.[15]

Strength of the study

Although previous studies have highlighted the metabolic parameters after use of PL and NS in patients with critical illness or hemorrhagic shock, no study has compared the metabolic outcomes in terms of acid–base status in patients undergoing HNS-FFR. This novel observational study can be used as a pilot for future interventional studies.

Limitation of the study

This study was a single center study with a small sample size. However, even with this small sample size, we were able to get a perspective on the changes in the metabolic parameters perioperatively in the two study groups

Multiple factors may be associated with metabolic derangements. However, our study has focussed on intraoperative fluid used. Confounding factors cannot possibly be removed in absence of a randomized control trial

Group allocation was done based on the majority (>75%) of intravenous fluid infused as the study design was observational. We strongly recommend a randomized controlled trial to validate the finding of our study so that intravenous fluid can be appropriately and judiciously used in these types of surgeries.

CONCLUSIONS

In this prospective, observational, cohort study, PL was associated with a better acid–base status, especially at conclusion of vascular anastomosis and in the postoperative period in patients undergoing HNS-FFR. NS administration worsened the acid–base status with a significant reduction in pH, BE and bicarbonate levels after anastomosis and reperfusion of the graft with a need for higher phenylephrine used intraoperatively. However, there was no difference in outcome-based parameters, i.e., graft viability, length of intensive care, and hospital stay. It is suggested that NS should be used cautiously in HNS-FFR, with close monitoring of acid–base balance.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.