Do oral amino acid supplements facilitate the healing of rat lung injuries?

Abstract

OBJECTIVES

Prolonged air leaks following lung injury cause extended hospital stays. This study investigated the effect of nutritional supplements containing arginine, glutamine and β-hydroxy β-methyl butyrate, which were theoretically proven to accelerate wound healing, on air leak and wound healing parameters in a rat lung injury model.

METHODS

Twenty-eight female rats were randomly divided into 4 groups. Experimental groups were given glutamine (Resource Glutamine^®) or a mixture of arginine, glutamine and β-hydroxy β-methyl butyrate (Abound^®) as a dietary supplement at isonitrogenous and isocaloric doses. On day 3, standard sized lung injuries were created in all rats except the sham group. The rats were sacrificed on day 6, and the lungs were removed for air-leak threshold pressure measurement and histopathological and biochemical analyses.

RESULTS

Loss of body mass was greater in the glutamine group than in the other groups (P = 0.004). Rats that received the amino acid mixture had better results for mature collagen fibre density (P = 0.002) and inflammation suppression (P = 0.003). The sham group had higher values for air-leak threshold pressure and all other histochemical parameters compared to the other groups. Hydroxyproline level did not differ significantly in any of the groups.

CONCLUSIONS

Our results indicated that an oral amino acid mixture was effective in the healing of lung injuries. Isolated glutamine supplementation had an adverse impact on body mass. Randomized clinical studies including larger series are needed. Hydroxyproline does not seem to be a suitable marker for this purpose.

Arginine, Glutamine, Lung injury, Methyl butyrate, Nutritional support

INTRODUCTION

Prolonged air leak following lung injury results in decreased quality of life and extended hospital stay. Therefore, the search continues for different treatments to resolve air leaks as early as possible.

As lung injuries heal via collagen synthesis and inflammation just like other wound healing processes in the body, approaches that accelerate wound healing should in theory accelerate the healing of lung injury and help resolve air leaks [1–6].

Many studies have been conducted in an effort to increase collagen synthesis in wounds. These studies revealed a close relationship between patients’ nutritional status and collagen deposition during wound healing. More recently, it was shown that pharmacological formulations of certain nutritional elements administered alone or in combination can increase collagen deposition in wounds [1]. Glutamine, arginine and β-hydroxy β-methyl butyrate (HMB) in particular were shown to be highly effective in the promotion of wound healing [1–6].

However, we were unable to find any studies demonstrating the effect of the oral administration of these nutritional elements on lung injury. In the present study, we aimed to evaluate whether these nutritional elements have an impact on the healing of lung injuries.

MATERIALS AND METHODS

Ethics

This study was approved by the Dokuz Eylül University Animal Experiments Local Ethics Committee (date: 24 July 2018; ID number: 05/14/2018). This committee evaluates all procedures based on the principles of 4R (replacement, reduction, refinement and reproduction).

Sample size

When determining sample size, calculations were made based on air-leak threshold pressure values of the control and experiment groups in the study published in 2006 by Sanlı et al. [6]. To determine inter-group differences at 0.9 partial eta-squared, 95% power and 5% type I error, we decided to include 7 animals per group, resulting in a total of 28 animals.

Animals

The female albino Wistar rats weighing between 200 and 270 g were used. The rats were housed at 21–23°C and controlled humidity with a 12-h light/dark cycle.

Nutritional supplements

In this study, l-glutamine (Resource Glutamine^®, Nestle, Osthofen, Germany) or a combination of l-arginine, l-glutamine and HMB (Abound^®, Abbott, CA, USA) was used as nutritional supplement.

The 2 preparations were equalized to isonitrogenous doses by dissolving 24 g Abound (1 packet) and 20 g Resource Glutamine (4 packets) in 300 ml of tap water each to yield solutions containing 3.8 g of nitrogen. In terms of energy content, 24 g of Abound contains 89 kcal and 20 g of Resource Glutamine contains 80 kcal.

The dose of l-glutamine used in our study was 1.5 g/kg [6]. Accordingly, rats in the experimental groups were given 4.5–6 ml of Resource Glutamine or Abound solution daily.

Rats eat a mean of 15–25 g per day [7], and 1 g of rat food (Nükleon Bil-Yem^®, Ankara, Turkey) provides 2.5 kcal of energy. Based on this, the daily energy requirement of rats can be calculated as 37.5–62.5 kcal. The energy provided by 4.5–6 ml of either solution differs from this value by only 0.14–0.18 kcal, which cannot cause an animal to exceed its daily energy requirement.

Therefore, both nutritional supplements given were both isocaloric and isonitrogenous.

Groups and randomization

The rats were randomly assigned numbers and divided into 4 groups [control, sham, Resource Glutamine supplement group (GLU) and Abound supplement group (AGH)] by simple randomization using statistical software (PASS 11^®; NCSS, LLC, Kaysville, UT, USA).

In addition to ad libitum access to food during the 6-day experiment, the rats were also given either Abound, Resource Glutamine or the vehicle by oral gavage each day at 9:00 am. Rats in the control and sham groups were given a volume of vehicle (tap water) equal to that of the amino acid solutions (4.5–6 ml/day). The experimental period started on the day the animals were given their first dose of nutritional supplement/vehicle.

Thoracotomy and lung injury

On day 3 of the experiment, general anaesthesia was provided with ether and the rats were placed in left lateral position for anterolateral thoracotomy through a 3-cm incision in the fifth intercostal space. To ensure the incisions were of standard size, the thoracotomy site was measured using a ruler and marked with a marker pen (Fig. 1).

Figure 1:

Standardization of thoracotomy dimensions.

For the rats in the control, GLU and AGH groups, the tip of the scalpel was inserted perpendicularly into the lung parenchyma in the upper and lower lobes of the right lung to create 2 injuries 1-mm long and 2-mm deep. Standard lung injury dimensions were achieved by placing metal clips on the scalpel blades 2 mm from the tip (Fig. 2). The presence of pulmonary air leak secondary to injury was qualitatively observed in all rats (except sham group). Rats in the sham group were subjected to thoracotomy only, with no lung injury. All thoracotomy incisions were closed without pleural drainage [6, 8].

Figure 2:

Ensuring standardization of the experimental injury.

Animal sacrifice, median sternotomy and air-leak threshold pressure measurement

The rats were sacrificed on day 6 of the experiment by exsanguination through the caudal vena cava under ether anaesthesia. Median sternotomy was performed, and both lungs were removed together with the trachea (Fig. 3). The left main bronchus was ligated, and the trachea intubated using a cannula. A three-way valve was placed at the other end of the cannula; 1 outlet was connected to the pressure monitor (KMA 800^®, PETAS, Istanbul, Turkey) while the other was connected to an injector. The lungs were immersed in 0.9% sodium chloride solution, and the injector was used to apply positive airway pressure (Fig. 4). The pressure level shown on the monitor when air leak was detected from the injury sites was recorded as the air-leak threshold pressure [6].

Figure 3:

Median sternotomy and removal of lung tissue.

Figure 4:

Air-leak threshold pressure measurement.

Transfer of tissues to pathology and biochemistry laboratories and blinding procedures

The medical pathology and medical biochemistry specialists in our study were blinded to which group of rats the tissue and blood samples were obtained from. Only the number of the corresponding rat was written on the containers and tubes.

Immediately after measuring air-leak threshold pressure, the upper and lower lobes of the right lung were removed anatomically from the specimen. The upper lobes were placed in numbered containers containing 10% buffered neutral formalin and sent to the medical pathology laboratory the next day.

The lower lobes were placed in empty, numbered sterile containers. Blood collected during animal sacrifice was collected in biochemistry tubes. These tubes and containers were stored in a refrigerator at 4°C. At the end of the same day, the samples were removed from the refrigerator, put on ice packs in an insulated bag and transferred to the biochemistry laboratory within 15 min.

Criteria for exclusion or termination of the study

Rats were excluded from the study if they showed signs of illness, impaired oral nutrition or died during the experiment, if additional lung damage beyond the planned injury occurred during thoracotomy, or if the lung tissue collected after sacrifice sustained additional damage. We planned to terminate the study if 3 or fewer rats remained in any group.

Histopathological analysis

The tissue samples were fixed in 10% buffered neutral formalin solution for 24–48 h. After fixation, the samples were processed with 70% alcohol, 95% alcohol and 100% alcohol in the dehydration stage, xylene in the clearing stage and paraffin in the infiltration stage. Then, the samples were embedded in paraffin blocks. Tissue sections 5 µm in thickness were made from the paraffin blocks using a microtome and stained with haematoxylin–eosin and Masson’s trichrome stains in the automated staining device.

The haematoxylin–eosin-stained sections were analysed under the light microscope using 10× and 40× objectives, and histopathological parameters of inflammation and healing were semi-quantitatively scored [4–6]. The following histopathological parameters were evaluated while scoring:

• Inflammation score (scored separately for neutrophil, eosinophil, lymphoplasmocyte and histiocyte density): 0 = none, 1 = mild, 2 = moderate, 3 = severe;

• Granulation tissue maturation score: 0 = no, 1 = partial, 2 = complete;

• Fibroblastic proliferation score: 0 = none, 1 = mild, 2 = moderate, 3 = severe;

• Angiogenesis score (new vessels/1 high-power field): 0 = 0, 1 = <5, 2 = 5–10, 3 = >10;

• Necrosis score: 0 = no necrosis, 1 = necrosis

• Mature collagen score (using Masson trichrome): 0 = no collagen fibres, 1 = few, loose fibres, 2 = moderate collagen fibre accumulation, 3 = thick, dense collagen bundles

Biochemical analysis

Lung tissues were stored at −80°C until the day of analysis. Lung tissues thawed on the day of analysis were homogenized in 50 mM phosphate buffer solution (pH 7) and centrifuged at 5000 rpm for 10 min at 4°C. The supernatant was collected.

Tissue protein in the supernatants was measured spectrophotometrically using a urine protein kit (Abbott) in automated analyser (Abbott c8000^®) and protein concentrations were determined (mg/dl). Hydroxyproline levels were analysed by enzyme-linked immunosorbent assay (ELISA) method using a rat ELISA (Cusabio, Wuhan, Hubei, China) kit in an automated ELISA plate washer (BioTek ELx50^®, Winooski, Vermont, USA) and ELISA reader (BioTek). Concentrations were calculated (in ng/ml) based on a curve drawn from calibration standards using Gen5^® Microplate Reader and Imager Software (BioTek). Tissue hydroxyproline concentrations were expressed per milligram of tissue protein.

After delivery, blood samples were centrifuged at 1000 rpm for 10 min at 4°C. The separated serum was transferred to Eppendorf tubes and stored at −80°C until the day of analysis. On the day of biochemical analysis, all serum samples were thawed and serum hydroxyproline concentrations were measured as described above.

Statistical analysis

Data were analysed using IBM SPSS Statistics Standard Concurrent User version 25 (IBM Corp., Armonk, NY, USA) software. Descriptive statistics were given as number of units (n), mean, 95% confidence interval, median (M) and 25th and 75th percentile values. Normal distribution of numerical data was evaluated using Shapiro–Wilk normality test and Q–Q plots [9]. Homogeneity of variances was evaluated using Levene’s test. Between-group comparisons were performed using one-way analysis of variance for normally distributed variables. If a difference was detected by one-way analysis of variance, Tukey’s HSD test was used as a multiple comparison test. Kruskal–Wallis analysis was used for between-group comparisons of non-normally distributed variables. If a difference was detected by Kruskal–Wallis analysis, multiple comparisons were performed using Dunn–Bonferroni test. Given the narrow distribution range of histopathological scores, mean rank score values were used to interpret the variance among groups. A P-value of <0.05 was considered statistically significant.

RESULTS

None of the criteria for exclusion or study termination occurred, and the study was completed with 28 rats. According to post hoc power calculation, type 1 error was 5%, partial eta-squared was 0.991 and statistical power was 98.9%. Air-leak threshold pressure was similar in the AGH, GLU and control groups and was higher in the sham group (P < 0.001). It was noted that air-leak threshold pressure was higher in the AGH group than in the GLU and control groups (Table 1). Pairwise comparisons also showed same results (P < 0.001).

Table 1:

Comparisons of body weight and threshold air-leakage pressure values among the groups

	Groups				Partial eta- squared	P-value
	AGH (n = 7)	GLU (n = 7)	Control (n = 7)	Sham (n = 7)
	Mean (95% CI)	Mean (95% CI)	Mean (95% CI)	Mean (95% CI)
Pre-thoracotomy weight (g)	252.57 (244.15–260.99)	236.86 (228.43–245.28)	255.43 (247.01–263.85)	252.57 (244.15–260.99)	0.35	0.092
Pre-sacrifice weight (g)	244.57a (233.51–255.63)	220.57b (209.51–231.63)	248.29a (237.22–259.35)	246.71a (235.65–257.78)	0.43	0.004
Weight difference (g)	8.00 (2.26–13.74)	16.29 (10.55–22.03)	7.14 (1.40–12.88)	5.86 (0.12–11.60)	0.27	0.071
Air-leak threshold pressure (mmHg)	7.14a (5.61–8.68)	5.42a (3.90–6.96)	6.00a (4.47–7.53)	51.00b (49.47–52.53)	0.99	<0.001

^a and ^b show significant differences between groups. Groups with the same letter designations were statistically equivalent.

AGH: Abound supplement group; CI: confidence interval; GLU: Resource Glutamine supplement group.

There was no significant difference between the groups in terms of the animals’ body weight before the experiment (P = 0.092). However, comparison of body weights measured before sacrifice showed that rats in the GLU group were significantly lighter (P = 0.004) (Table 1). Pairwise comparisons also showed same results.

There were no differences among the groups in the biochemical analysis results (Table 2). The group data presented in this table were not normally distributed. The pairwise comparisons were similar again.

Table 2:

Results of biochemical analysis

	Groups				P-value
	AGH (n = 7)	GLU (n = 7)	Control (n = 7)	Sham (n = 7)
	M (25th–75th percentile)	M (25th–75th percentile)	M (25th–75th percentile)	M (25th–75th percentile)
Hydroxyproline level (ng/ml)	2049.34 (1892.95–3240.16)	3026.13 (2206.11–4376.72)	3491.23 (3223.88–5193.66)	2363.59 (1375.55–2685.75)	0.12
Tissue protein value (mg/dl)	468.10 (398.40–516.30)	491.10 (374.50–588.90)	587.50 (202.20–656.20)	562.80 (389.10–721.90)	0.86
Tissue protein value (mg/ml)	4.68 (3.98–5.16)	4.91 (3.75–5.89)	5.88 (2.02–6.56)	5.63 (3.89–7.22)	0.86
Hydroxyproline per milligram of tissue protein (ng/mg protein)	511.52 (312.03–848.21)	656.57 (409.22–808.05)	890.85 (778.65–1675.41)	410.40 (236.51–1130.77)	0.19
Serum hydroxyproline level (ng/ml)	609.79 (110.99–851.96)	416.49 (95.81–893.55)	754.64 (698.42–959.34)	441.19 (65.00–754.64)	0.26

AGH: Abound supplement group; GLU: Resource Glutamine supplement group; M: median.

Histopathological analyses revealed significant differences in inflammation and healing criteria in the sham group compared to the other groups (Table 3). According to Masson’s trichrome staining, the AGH group had statistically higher mature collagen fibre density values than the sham and control groups (Table 3). In pairwise comparisons, only comparisons with the sham group were different.

Table 3:

Histopathological scores

	Groups				P-value
	AGH (n = 7)	GLU (n = 7)	Control (n = 7)	Sham (n = 7)
	M (25th–75th percentile) mrs	M (25th–75th percentile) mrs	M (25th–75th percentile) mrs	M (25th–75th percentile) mrs
Neutrophil	1 (1–2)	2 (1–3)	2 (2–3)	1 (0–1)	0.026
	14.29a,b	16.71a	19.57a	7.43b
Eosinophil	1 (1–1)	1 (1–1)	2 (1–2)	0 (0–1)	0.005
	14.86a	16.00a	20.50a	6.64b
Lymphoplasmocyte	1 (1–2)	2 (2–2)	2 (2–2)	1 (1–1)	0.010
	12.79a,b	17.14a	20.21a	7.86b
Histiocyte	2 (1–3)	2 (1–3)	2 (2–3)	1 (0–1)	0.007
	16.71a	16.57a	19.00a	5.71b
Granulation tissue	1 (0–2)	1 (1–2)	2 (1–2)	0 (0)	0.007
	15.57a	17.00a	19.43a	6.00b
Necrosis	0 (0)	0 (0)	0 (0)	0 (0)	1.00
	14.50	14.50	14.50	14.50
Angiogenesis	2 (1–2)	1 (0–2)	1 (1–1)	0 (0–0)	0.003
	21.29a	14.64a	16.07a	6.00b
Fibroblastic proliferation	2 (1–3)	1 (1–2)	1 (0–2)	0 (0)	0.001
	22.00a	15.93a	14.57a	5.50b
Mature collagen fibre density (Masson’s trichrome staining)	2 (2–3)	2 (1–2)	1 (1–2)	0 (0–1)	0.002
	21.64a	17.07a,b	13.36b	5.93b

^a and ^b indicate differences between groups. Groups with the same letter designations were statistically equivalent.

AGH: Abound supplement group; GLU: Resource Glutamine supplement group; M: median; mrs: mean rank score.

The sum of neutrophil, eosinophil, lymphoplasmocyte and histiocyte scores was accepted as the inflammation criterion. In addition, the sum of neutrophil and eosinophil scores was used as the acute inflammation criterion and the sum of lymphoplasmocyte and histiocyte scores was defined as the chronic inflammation criterion. In all comparisons, the sham group differed from the GLU and control groups and was similar to the AGH group (Table 4).

Table 4:

Comparison of inflammation, acute inflammation and chronic inflammation criteria among the groups

	Groups				P-value
	AGH (n = 7)	GLU (n = 7)	Control (n = 7)	Sham (n = 7)
	M (25th–75th percentile) osp	M (25th–75th percentile) osp	M (25th–75th percentile) osp	M (25th–75th percentile) ARS
Inflammation criterion	5 (4–9)a,b	7 (5–9)a	8 (7–9)a	2 (2–3)b	0.003
	15.21	17.00	20.57	5.21
Acute inflammation criterion	2 (2–3)a,b	3 (2–4)a	4 (3–5)a	1 (0–1)b	0.005
	14.57	16.86	20.71	5.86
Chronic inflammation criterion	3 (3–5)a,b	4 (3–5)a	4 (4–5)a	2 (1–2)b	0.003
	15.29	17.14	20.36	5.21

^a and ^b indicate differences between groups. Groups with the same letter designations were statistically equivalent.

AGH: Abound supplement group; ARS: average rank score; GLU: Resource Glutamine supplement group; M: median.

DISCUSSION

The results of the present study indicate that oral administration of an amino acid mixture had a significant effect on the healing of lung injury in rats. However, these results should be supported by clinical studies including larger samples.

The fact that the animals in our study were the same species and sex, similar in weight, housed under the same conditions, and subjected to the same procedures throughout the experiment rules out the effect of these potential confounders. Furthermore, other than the amino acid solution received, there were no additional differences in nutrient intake, feeding patterns or stress factors between the animals. The groups were determined in a completely objective manner using simple randomization.

In previous experimental studies evaluating the effects of several different nutritional supplements, it was stated that the nutritional supplements given should be isonitrogenous and isocaloric [10, 11]. This was the basis of our use.

In the clinical setting, nutritional supplementation is provided for longer period. However, the patients in these practices are human. Humans and rats are very different in terms of vital functions such as metabolic rate, heart rate and protein turnover times. It has been shown that the protein turnover time is 9.6 times, the metabolic rate is 6.4 times and the lifespan is 26.7 times faster in rats [12]. Therefore, 2 rat days of protein turnover correspond to 19 human days and the total 6-day experimental duration corresponds to 58 human days. We believe that these times are suitable for the simulation of clinical setting.

The literature includes numerous studies related to the association between loss of body weight and deterioration of general health and performance status [13–15]. All animals in our study lost some amount of weight between the start of the experiment and sacrifice due to thoracotomy. However, the rats in the GLU group lost significantly more weight than those in the other groups, which suggests that Resource Glutamine may have had an adverse effect on the animals. This finding is consistent with previous studies. Laviano et al. [16] reported that glutamine supplementation was effective in weight reduction in obese female patients. In their 2019 study, Abboud et al. [17] also observed the same phenomenon in Wistar rats. In contrast, there were no studies demonstrating that HMB caused weight loss. Moreover, we found studies and reviews recommending the use of HMB in conditions such as cancer cachexia and showing that it improved fat-free body mass in healthy older adults [18–20].

Regarding the surgical procedure, closing the thoracotomy without placing a chest drain may initially suggest that the animals were unable to ventilate the injured lungs postoperatively due to lung collapse, but this is not the case. It must be remembered that there are anatomical differences between rats and humans. For instance, the mediastinal pleura is not fully developed in rats [8]. For this reason, the air spreads to mediastinum, contralateral hemithorax, abdomen and subcutaneous spaces, while the lung remains expanded. Therefore, despite the lung injury, thoracotomy closure without drainage in rats does not cause tension pneumothorax and is compatible with life. This indicates that our model provided a valid clinical simulation.

Air-leak threshold pressure has been used as an indicator of lung tissue healing in many studies [6, 21–23]. We also measured air-leak threshold pressure in our study along with histopathological and biochemical parameters. As lung injury was not made in the rats in the sham group, higher air-leak threshold pressure in healthy lung compared to the lungs of the other animals is an expected result. However, this finding shows that our results are independent of thoracotomy or sternotomy.

After the sham group, the AGH group had the highest air-leak threshold pressure values. We believe that this difference did not reach statistical significance due to the small number of animals used in the study. Clinical studies with larger series could yield results supporting the potential positive effect of Abound on air-leak threshold pressure.

The amount of collagen produced following an injury is an indicator of wound healing in the tissue. The predominant amino acid in collagen synthesis is glycine, at a ratio of 33%. This is followed by proline, hydroxyproline, lysine and hydroxylysine. In the present study, we measured tissue hydroxyproline concentration based on the possibility that it may be an indicator of collagen formation. However, we observed no difference between the groups in terms of hydroxyproline levels. Similar to our study, Bozkırlı et al. [5] reported that nutritional supplementation in healthy rats did not improve secondary wound healing. In another study from the same group, rats were given nutritional supplementation to treat ischaemic wound healing, and it was biochemically determined that hydroxyproline was an indicator of tissue regeneration [4]. Consistent with our findings, there was no difference in hydroxyproline values in the control and treatment groups in their study. In light of these studies and our results, we believe that hydroxyproline level is not a good marker for wound healing. In future studies on this issue, it may be more appropriate to study another component as a marker.

In our study, hydroxyproline results were non-significant not only in rat lung tissue but also in serum. This showed that the nutritional supplements given to the animals had no effect on wound healing or on systemic hydroxyproline levels.

The protein levels were higher in the sham and control groups (Table 2), which did not receive nutritional supplementation. This may be explained by the fact that proteins are made up of various amino acids, and for this reason, supplementation of certain isolated amino acids could create an imbalance of amino acid distribution in the body, which may result in reduced protein synthesis. Therefore, we believe that, like hydroxyproline, total protein may not be a good marker when used alone. Further studies are needed to test this hypothesis.

In complete contrast to this phenomenon, serum and tissue hydroxyproline levels were highest in the control group and lowest in the sham group, both of which were fed the same diet. We have no explanation for this finding and thus believed that hydroxyproline is also a poor marker.

The sham group had the best results in terms of inflammation and healing criteria. As with the air-leak threshold pressure measurements, these findings demonstrated that lungs without parenchymal damage had better outcomes than those with lung injury and that the thoracotomy and sternotomy had no effect on the results of our experiment.

Mature collagen fibre density values were higher in the AGH group than in the sham and control groups. This means that the animals given Abound exhibited a statistically significant benefit compared to those that did not receive it. Though it could not be confirmed directly with threshold air-leak pressure measurements, this was an indirect finding supporting our hypothesis that Abound would have a favourable effect on lung healing.

In addition, mature collagen fibre density values and the mean rank score for fibroblast proliferation were higher in the AGH group than in the control and GLU groups. This offered further evidence that Abound may have a positive effect on lung injury healing and should be corroborated in larger clinical studies.

Inflammation criteria were similar in the sham and AGH groups and different in the GLU and control groups. The statistical similarity between the AGH group and the uninjured sham group is a strong indication that inflammatory processes were reduced in the AGH group. In other words, Abound also suppressed inflammation at the wound site. Inflammation is a necessary part of the initial stage of wound healing, but a prolonged inflammatory phase can delay wound healing due to fibroplasia and inhibition of collagen and proteoglycan synthesis. Lower inflammation scores and higher fibroblastic proliferation scores in the AGH group suggest that AGH conferred a greater benefit in terms of wound healing.

The favourable effects of intraperitoneal glutamine on lung injury in rats were demonstrated previously [6]. However, we observed the opposite with oral glutamine. Studies comparing the efficacy of intravenous and oral glutamine are needed.

Limitations

Although power calculations performed both before and after the study indicated that an adequate number of animals were included, we believe that similar studies including larger sample sizes will provide clearer results. Furthermore, although the biochemists and pathologists who analysed the samples in our study were blinded, it may have been beneficial to blind the physicians who performed the surgical procedures as well to reduce potential bias. Another limitation is that these effects were analysed after only 2 preoperative days and a total of 6 days of nutritional supplementation. Although the evidence indicates that this period corresponds to 58 human days, this could still be considered a limitation. Furthermore, the results of this experimental study cannot be generalized to lung injury in humans without corroboration by randomized clinical trials. We recommend using a component other than hydroxyproline as a biochemical marker in future studies.

CONCLUSION

Our results indicated that an oral amino acid mixture was effective in the healing of lung injuries. Isolated glutamine supplementation had an adverse impact on body mass. Randomized clinical studies including larger series are needed. Hydroxyproline does not seem to be a suitable marker for this purpose.

Funding

This work was supported by the Scientific and Technological Research Council of Turkey (TÜBİTAK), under the scope of the national support project ‘1002-Quick Support’ (218S860).

Conflict of interest: none declared.

Author contributions

Hasan Ersöz: Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Validation; Writing—original draft; Writing—review & editing. İsmail Ağababaoğlu: Data curation; Investigation; Methodology; Validation; Writing—review & editing. İbrahim Taylan: Investigation; Methodology; Writing—review & editing. Ebru Çakır: Investigation; Methodology; Writing—review & editing. Saliha Aksun: Investigation; Methodology; Writing—review & editing. Ensari Güneli: Conceptualization; Data curation; Resources; Supervision; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Luca Bertolaccini, Emmanouil Ioannis, Michael John Shackcloth and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

Abbreviations

 
• AGH

  Abound supplement group

 
• ELISA

  Enzyme-linked immunosorbent assay

 
• GLU

  Resource Glutamine supplement group

 
• HMB

  β-Hydroxy β-methyl butyrate