Rapid Communication
Copyright ©2006 Baishideng Publishing Group Co., Limited. All rights reserved.
World J Gastroenterol. Apr 21, 2006; 12(15): 2427-2431
Published online Apr 21, 2006. doi: 10.3748/wjg.v12.i15.2427
Effect of oral garlic on arterial oxygen pressure in children with hepatopulmonary syndrome
Mehri Najafi Sani, Hamid Reza Kianifar, Abdolrazagh Kianee, Gholamreza Khatami
Mehri Najafi Sani, Gholamreza Khatami, Division of pediatric Gastroenterology, Department of Pediatrics, Children’s Medical Center Hospital of Tehran University , Tehran, Iran
Hamid Reza Kianifar, Division of Pediatric Gastroenterology, Department of Pediatrics, Ghaem Medical Center of Mashhad University of Medical Sciences, Mashhad, Iran
Abdolrazagh Kianee, Division of Pediatric Cardiology, Department of Pediatrics, Children’s Medical Center Hospital of Tehran University, Tehran, Iran
Correspondence to: Hamid Reza Kianifar, Division of Pediatric Gastroenterology, Department of Pediatrics, Ghaem Medical Center of Mashhad University of Medical Sciences, Ahmadabad Ave, Mashhad, Iran. hr-kianifar@Mums.Ac.ir
Telephone: +98-511-7640815 Fax: +98-511-8417451
Received: October 9, 2005
Revised: November 2, 2005
Accepted: November 10, 2005
Published online: April 21, 2006


AIM: To study the effect of oral garlic on arterial oxygen pressure in children with hepatopulmonary syndrome.

METHODS: Garlic powder in a capsule form was given to 15 children with hepatopulmonary syndrome (confirmed by contrast echocardiography) at the dosage of 1 g/1.73 m2 per day. Patients were evaluated clinically and by arterial blood gas every four weeks.

RESULTS: The garlic capsule was administered to 15 patients with hepatopulmonary syndrome. There were 10 boys and 5 girls with a mean age of 9.4 ± 3.9 years. The underlying problems were biliary tract atresia (4 patients), autoimmune hepatitis (4 patients), cryptogenic cirrhosis (4 patients) and presinusoidal portal hypertension (3 patients). Eight patients (53.3%) showed an increase of 10 mmHg in their mean arterial oxygen pressure. The baseline PaO2 was 65.6 ± 12.1 mmHg in the responder group and 47.1 ± 11.2 mmHg in non-responder group. At the end of treatment the mean PaO2 in responders and non-responders was 92.2 ± 7.75 mmHg and 47.5 ± 11.87 mmHg, respectively (P < 0.01).

CONCLUSION: Garlic may increase oxygenation and improve dyspnea in children with hepatopulmonary syndrome.

Key Words: Hepatopulmonary syndrome, Garlic, Arterial oxygen pressure, Pediatric


In 1995, Lange et al [1] introduced the term “hepatopulmonary syndrome” for those patients with pulmonary vasodilatation associated with chronic liver disease and cyanosis. Hepatopulmonary syndrome is defined by an arterial oxygen pressure < 70 mmHg in room air. In adults, the prevalence is 5%-29% and the overall mortality is 41% in hospitalized patients. In children, the prevalence ranges from 0.5% in those with portal vein obstruction to 20% in children with biliary atresia and polysplenia syndrome. In other causes of childhood cirrhosis, the prevalence is 2% -4%. Hepatopulmonary syndrome has been described in children as young as 6 months. Generally, hepatopulmonary syndrome is seen in chronic liver disease, but it has also been described in acute liver failure and extrahepatic portal venous obstruction [2].

A right to left pulmonary shunt due to intrapulmonary vascular dilation characterizes this syndrome. Blood flowing through the dilated capillaries is less exposed to oxygen contained in the alveoli, resulting in a ventilation-perfusion mismatch [3-6]. In animal models, elevated levels of nitric oxide have been detected in the lung homo-genates of animals with the clinical features of hepatopulmonary syndrome. This finding appears to be supported by reports documenting increased exhaled nitric oxide concentration in patients with hepatopulmonary syndrome [7,8].

The best method to evaluate the shunt in hepatopulmonary syndrome is the contrast-enhanced echocardiogram [1,2,9,10]. Technetium 99m-labeled macroaggregated albumin scanning is a second method of detecting intrapulmonary vascular dilatations. Pulmonary arteriography can suggest two angiographic patterns [1]. Schneck et al [13] have reported the beneficial effect of methylene blue on this syndrome [13]. The transjugular intrahepatic portosystemic shunt (TIPS) [14], cavoplasty [15], octreotide [16], indomethacin [17,18] and embolization [19] in the management of this syndrome have been assessed.

In 1992, clinical improvement in a case report was achieved following the use of garlic [20]. In a study by Abrams et al [21] the use of garlic in adults in treatment of this syndrome was evaluated and garlic was found to be an effective therapy for this syndrome especially in young groups. Ku et al [22] revealed that active garlic metabolites are capable of pulmonary vasodilatation in rats. In 1997 Battaglia et al [23] reported an improvement in right to left shunt and arterio-alveolar gradient in such patients after liver transplantation. Further studies have confirmed the beneficial effect of liver transplantation on this syndrome [24,25]. In our country, liver transplantation has a long waiting list. If the oral garlic can improve the arterial oxygenation and dyspnea, it may represent a palliative therapy for patients with hepatopulmonary syndrome awaiting liver transplantation.


Fifteen patients (aged 6 months - 14 years) suffering from chronic liver disease, portal hypertension or cirrhosis and intrapulmonary shunt which were confirmed by saline contrast echocardiography were enrolled in this pilot study. Patients with congenital heart diseases, acute and chronic lung diseases and symptoms of severe gastritis, were excluded from the study. This study was performed at Children’s Medical Center Hospital of Tehran University from 2002-2003.


Arterial blood gases were obtained from the radial artery in a sitting position at the same time and in the same room. Echocardiography with saline contrast was performed as previously described [13]. The presence of air bubbles in the left heart between 3 and 6 cardiac contractions indicated intrapulmonary shunt. Garlic was administered to the patients at the dose of 0.5-2 g/1.73 m2 per day. We obtained a special kind of dried garlic (Hamedan city garlic), which was prepared in the form of 250 mg capsules by the hospital pharmacy.

The parents were asked to add the capsule or its content to their food once or twice a day at the recommended dose. We evaluated the patients every month. Clinical symptoms and signs, probable drug complication, and patient’s compliance to the drug were asked in each visit. Arterial blood gas (ABG) was taken in the same condition as mentioned above. During the study we provided some facilities such as free contact if needed.

Response to garlic was defined as a 10 mmHg increment in the PaO2 or when PaO2 was more than 70 mmHg. Baseline and monthly data between responders and non-responders were analyzed statistically.

The Ethics Committee of Tehran University approved the protocol. The parents of the children gave their informed consent.

Stastistical analysis

The results were expressed as mean ± SD. Differences between two groups were evaluated by Fisher’s exact test, Mann Whitney U and Wilcoxson rank-sum tests. P < 0.05 was considered statistically significant.


The baseline data of patients are summarized in Table 1. Fifteen patients with hepatopulmonary syndrome were treated with garlic. One child expired after one month of treatment.

Table 1 Baseline characteristics of children with hepatopulmonary syndrome.
Age (yr)GenderType of diseaseDuration (yr)Child scoreDyspnea
Non-responders14MCryptogenic cirrhosis10C+
6MBiliary atresia6C+
3FBiliary atresia3C+
9MBiliary atresia9C+
12MBiliary atresia11C+
14MCryptogenic cirrhosis9B+
11MCryptogenic cirrhosis5B-

At the end of the first month arterial O2 pressure was increased significantly (P = 0.004). The mean of arterial O2 pressure increment at the end of the second month was higher than 10 mmHg and over 70 mmHg at the end of the third month. At the end of the fifth month, the mean increment in the arterial O2 pressure was 15 mmHg (Table 2).

Table 2 PaO2 values of responders and non-responders in relation to dyspnea after treatment.
RespondersBaseline1st mo2nd mo3rd mo4th mo5th moDypnea after Rx
1343332354040No change
5374240394547No change
6454951545052No change

Eight patients (53.3%) were considered as respon-ders. The baseline PaO2 was 65.6 ± 12.1 mmHg in the responder group and 47.1 ± 11.2 mmHg in non-responder group. There was a statistically significant difference in the initial O2 pressure between the responders and non-responders (P = 0.009). At the end of treatment the mean PaO2 in responders and non-responders was 92.2 ± 7.75 mmHg and 47.5 ± 11.87 mmHg, respectively (Table 3).

Table 3 PaO2 values after oral garlic administration (mean±SD, n =15).
Baseline1st mo2nd mo3rd mo4th mo5th mo
Non-responders47.1 ± 11.248.8 ± 13.247.5 ± 14.249 ± 13.147.3 ± 11.847.5 ± 11.8
Respondersb65.6 ± 12.182 ± 6.885.8 ± 6.189.5 ± 5.190.7 ± 792.2 ± 7.7d
Total57 ± 14.766.5 ± 19.769.4 ± 2272.1 ± 22.672.1 ± 2473 ± 24.7 d

Dyspnea (including shortness of breath at rest or exercise, orthopnea and platypnea) occurred in 5 responders and 6 non-responders. Four out of these 11 patients had their dyspnea improved after treatment (Table 2).

According to the Child score, 5 patients (33.3%) were in group A, 4 (26.6%) in group B, and 6 (40%) in group C. The mean of arterial oxygen pressure in groups A, B and C, at the beginning of therapy was 56.4 ± 15.6 mmHg, 68.7 ± 11.2 mmHg and 49.6 ± 12.8 mmHg, respectively. At the end of therapy the mean PaO2 in A, B and C groups was 79.4 ± 19.6 mmHg, 88.2 ± 13.3 mmHg and 53.6 ± 26 mmHg, respectively. No statistical difference in the mean PaO2 was noted between groups A, B, and C.


The role of garlic in the management of hepatopulmonary syndrome is controversial. In 1992, clinical improvement was reported following the use of garlic [20]. In 1998 Abrams et al [21] studied the effects of garlic on hepatopulmonary syndrome in adults [21]. Chan et al [26] showed that garlic has no effect on the improvement of PaO2 and hypoxemia as well as clinical status, suggesting that deteriorating oxygenation in patients with chronic liver disease may be an indication for liver transplantation. Abrams et al [21] showed that 40% of patients have at least an increase of 10 mmHg in their arterial oxygen pressure after treatment with garlic. They also studied the dyspnea index in these patients and reported that all patients responding to therapy have an improvement in dyspnea [21]. In the pre-sent study, 53.3% of patients following the use of garlic had an arterial oxygen pressure > 70 mmHg or an increment > 10 mmHg. Four out of the 8 patients responding to the therapy had an improvement in dyspnea, but non-responders did not show any improvement in their respiratory symptoms.

It was reported that the clinical response is better in younger patients [21]. The results of the present study are consistent with those of Abrams et al [21]. We showed an improvement rate of 53.4% in children and Abrames et al showed 40% in adults. Nevertheless, no statistically significant differences were noted between responders and non-responders with respect to age.

It is interesting to note that responders had a higher baseline PaO2, suggesting that non-responders have a severe and profound disease and that medical management can be effective when baseline PaO2 is higher. Since TC-99 macro-aggregated albumin scan was not performed in the present study, there was a possibility of selection bias, because anatomic shunt failing to respond to medical therapy was not excluded.

The exact mechanism of garlic therapy for hepatopulmonary syndrome is not known. Garlic appears to cause pulmonary vasodilatation [22] and increases the rate of NO synthesis [25,27,28]. Thus one can conclude that garlic wor-sens hepatopulmonary syndrome. Garlic causes reduced NO synthesis in macrophages, resulting in reduced concentration in hypoxic tissue [29].

Intrapulmonary vasodilatation occurs primarily in the bases of the lungs, resulting in significant V/Q mismatch in this region. Abrams et al [21] speculated that if garlic results in uniform vasodilatation throughout the lung, then a redistribution of pulmonary blood flow to apical and mid lung fields could significantly improve V/Q ratios in these regions. Therefore, garlic may improve V/Q and reduce NO synthesis in the lung bases. Also garlic is known as a hepatoprotective agent and protects liver against tissue and chemical injures. If this is the case, it is probable that garlic might improve liver function and hepatopulmonary syndrome. Garlic might also reduce portal hypertension, indirectly resulting in an improvement in hepatopulmonary syndrome. Additionally, Allicin, Ajoene, and diallyl sulfur may be beneficial for hepatopulmonary syndrome treatment. These ingredients are present in dried garlic in sufficient amount [22]. However, the amounts of these ingredients vary depending on different types of garlic [29].

In children with chronic cholestasis, repeated transcutaneous bedside measurements are a rapid and reliable noninvasive test for characterizing the severity of abnormal oxygenation, and may prove useful also in post liver transplantation monitoring [30]. Although the existence of hypoxemia and hepatopulmonary syndrome has a prognostic value in patients with chronic liver disease [31], controlling the oxygen pressure, is not recommended as a screening tool [9,10]. Therefore disease in diagnosis and control echocardiography with contrast should be more exact.

In conclusion, garlic can reduce the severity of hepatopulmonary syndrome. Arterial oxygen pressure is increased after treatment with garlic. Further studies are needed to evaluate the garlic therapy for hepatopulmonary syndrome in children.


The authors thank nurses of Gastroenterology and Cardiology Wards in Children’s Medical Center Hospital and also Miss. Khaef and Miss. Javadzadeh for their kind assistance in this study.


S- Editor Wang J L- Editor Wang XL E- Editor Zhang Y

1.  Lange PA, Stoller JK. The hepatopulmonary syndrome. Ann Intern Med. 1995;122:521-529.  [PubMed]  [DOI]
2.  Mc Diarmid SV. Treatment of end stage liver disease. Pediatric Gastrointestinal Disease. 3rd ed. Ontario: Bc Decker 2000; 1264-1265.  [PubMed]  [DOI]
3.  Davis HH 2nd, Schwartz DJ, Lefrak SS, Susman N, Schainker BA. Alveolar-capillary oxygen disequilibrium in hepatic cirrhosis. Chest. 1978;73:507-511.  [PubMed]  [DOI]
4.  Hedenstierna G, Söderman C, Eriksson LS, Wahren J. Ventilation-perfusion inequality in patients with non-alcoholic liver cirrhosis. Eur Respir J. 1991;4:711-717.  [PubMed]  [DOI]
5.  Hopkins WE, Waggoner AD, Barzilai B. Frequency and significance of intrapulmonary right-to-left shunting in end-stage hepatic disease. Am J Cardiol. 1992;70:516-519.  [PubMed]  [DOI]
6.  Whyte MK, Hughes JM, Peters AM, Ussov W, Patel S, Burroughs AK. Analysis of intrapulmonary right to left shunt in the hepatopulmonary syndrome. J Hepatol. 1998;29:85-93.  [PubMed]  [DOI]
7.  Fallon MB, Abrams GA, Luo B, Hou Z, Dai J, Ku DD. The role of endothelial nitric oxide synthase in the pathogenesis of a rat model of hepatopulmonary syndrome. Gastroenterology. 1997;113:606-614.  [PubMed]  [DOI]
8.  Cremona G, Higenbottam TW, Mayoral V, Alexander G, Demoncheaux E, Borland C, Roe P, Jones GJ. Elevated exhaled nitric oxide in patients with hepatopulmonary syndrome. Eur Respir J. 1995;8:1883-1885.  [PubMed]  [DOI]
9.  Abrams GA, Jaffe CC, Hoffer PB, Binder HJ, Fallon MB. Diagnostic utility of contrast echocardiography and lung perfusion scan in patients with hepatopulmonary syndrome. Gastroenterology. 1995;109:1283-1288.  [PubMed]  [DOI]
10.  Krowka MJ, Cortese DA. Hepatopulmonary syndrome. Current concepts in diagnostic and therapeutic considerations. Chest. 1994;105:1528-1537.  [PubMed]  [DOI]
11.  Krowka MJ, Wiseman GA, Burnett OL, Spivey JR, Therneau T, Porayko MK, Wiesner RH. Hepatopulmonary syndrome: a prospective study of relationships between severity of liver disease, PaO(2) response to 100% oxygen, and brain uptake after (99m)Tc MAA lung scanning. Chest. 2000;118:615-624.  [PubMed]  [DOI]
12.  Vachiéry F, Moreau R, Hadengue A, Gadano A, Soupison T, Valla D, Lebrec D. Hypoxemia in patients with cirrhosis: relationship with liver failure and hemodynamic alterations. J Hepatol. 1997;27:492-495.  [PubMed]  [DOI]
13.  Schenk P, Madl C, Rezaie-Majd S, Lehr S, Müller C. Methylene blue improves the hepatopulmonary syndrome. Ann Intern Med. 2000;133:701-706.  [PubMed]  [DOI]
14.  Riegler JL, Lang KA, Johnson SP, Westerman JH. Transjugular intrahepatic portosystemic shunt improves oxygenation in hepatopulmonary syndrome. Gastroenterology. 1995;109:978-983.  [PubMed]  [DOI]
15.  De BK, Sen S, Biswas PK, Sanyal R, Majumdar D, Biswas J. Hepatopulmonary syndrome in inferior vena cava obstruction responding to cavoplasty. Gastroenterology. 2000;118:192-196.  [PubMed]  [DOI]
16.  Söderman C, Juhlin-Dannfelt A, Lagerstrand L, Eriksson LS. Ventilation-perfusion relationships and central haemodynamics in patients with cirrhosis. Effects of a somatostatin analogue. J Hepatol. 1994;21:52-57.  [PubMed]  [DOI]
17.  Andrivet P, Cadranel J, Housset B, Herigault R, Harf A, Adnot S. Mechanisms of impaired arterial oxygenation in patients with liver cirrhosis and severe respiratory insufficiency. Effects of indomethacin. Chest. 1993;103:500-507.  [PubMed]  [DOI]
18.  Hamilton G, Phing RC, Hutton RA, Dandona P, Hobbs KE. The relationship between prostacyclin activity and pressure in the portal vein. Hepatology. 1982;2:236-242.  [PubMed]  [DOI]
19.  Felt RW, Kozak BE, Rosch J, Duell BP, Barker AF. Hepatogenic pulmonary angiodysplasia treated with coil-spring embolization. Chest. 1987;91:920-922.  [PubMed]  [DOI]
20.  Caldwell SH, Jeffers LJ, Narula OS, Lang EA, Reddy KR, Schiff ER. Ancient remedies revisited: does Allium sativum (garlic) palliate the hepatopulmonary syndrome. J Clin Gastroenterol. 1992;15:248-250.  [PubMed]  [DOI]
21.  Abrams GA, Fallon MB. Treatment of hepatopulmonary syndrome with Allium sativum L. (garlic): a pilot trial. J Clin Gastroenterol. 1998;27:232-235.  [PubMed]  [DOI]
22.  Ku DD, Abdel-Razek TT, Dai J, Kim-Park S, Fallon MB, Abrams GA. Garlic and its active metabolite allicin produce endothelium- and nitric oxide-dependent relaxation in rat pulmonary arteries. Clin Exp Pharmacol Physiol. 2002;29:84-91.  [PubMed]  [DOI]
23.  Battaglia SE, Pretto JJ, Irving LB, Jones RM, Angus PW. Resolution of gas exchange abnormalities and intrapulmonary shunting following liver transplantation. Hepatology. 1997;25:1228-1232.  [PubMed]  [DOI]
24.  Collisson EA, Nourmand H, Fraiman MH, Cooper CB, Bellamy PE, Farmer DG, Vierling JM, Ghobrial RM, Busuttil RW. Retrospective analysis of the results of liver transplantation for adults with severe hepatopulmonary syndrome. Liver Transpl. 2002;8:925-931.  [PubMed]  [DOI]
25.  Stoller JK, Moodie D, Schiavone WA, Vogt D, Broughan T, Winkelman E, Rehm PK, Carey WD. Reduction of intrapulmonary shunt and resolution of digital clubbing associated with primary biliary cirrhosis after liver transplantation. Hepatology. 1990;11:54-58.  [PubMed]  [DOI]
26.  Chan CC, Wu HC, Wu CH, Hsu CY. Hepatopulmonary syndrome in liver cirrhosis: report of a case. J Formos Med Assoc. 1995;94:185-188.  [PubMed]  [DOI]
27.  Egen-Schwind C, Eckard R, Kemper FH. Metabolism of garlic constituents in the isolated perfused rat liver. Planta Med. 1992;58:301-305.  [PubMed]  [DOI]
28.  Ip C, Lisk DJ. Bioavailability of selenium from selenium-enriched garlic. Nutr Cancer. 1993;20:129-137.  [PubMed]  [DOI]
29.  Nagae S, Ushijima M, Hatono S, Imai J, Kasuga S, Matsuura H, Itakura Y, Higashi Y. Pharmacokinetics of the garlic compound S-allylcysteine. Planta Med. 1994;60:214-217.  [PubMed]  [DOI]
30.  Santamaria F, Sarnelli P, Celentano L, Farina V, Vegnente A, Mansi A, Montella S, Vajro P. Noninvasive investigation of hepatopulmonary syndrome in children and adolescents with chronic cholestasis. Pediatr Pulmonol. 2002;33:374-379.  [PubMed]  [DOI]
31.  Schwarzenberg SJ, Freese DK, Regelmann WE, Gores PF, Boudreau RJ, Payne WD. Resolution of severe intrapulmonary shunting after liver transplantation. Chest. 1993;103:1271-1273.  [PubMed]  [DOI]