Chinese Medical Journal 2014;127(4):675-679:10.3760/cma.j.issn.0366-6999.20131586
Symptomatic pulmonary lipiodol embolism after transarterial chemoembolization for hepatic malignant tumor: clinical presentation and chest imaging findings

Xu Haifeng, Yang Renjie, Wang Xiaodong, Zhu Xu and Chen Hui

lipiodol; pulmonary embolism; hepatic tumor; complications; transarterial chemoembolization
Background Pulmonary lipiodol embolism after transarterial chemoembolization (TACE) was rare and life-threatening, occasionally reported in previous literatures. We aimed to review the records of 11 patients with pulmonary oily embolism and analyze their characteristics of radiographic findings and risk factors.
Methods Records of 478 consecutive patients who underwent 1 026 percutaneous TACE procedures were retrospectively analyzed. Eleven cases with respiratory symptoms were identified as having symptomatic pulmonary lipiodol embolism after TACE. Data of these patients, including clinical presentation, techniques of TACE, imaging features of tumor and chest imaging findings, were assessed.
Results Eleven (2.3%) of 478 consecutive patients who underwent percutaneous TACE procedures had a pulmonary oily embolism after procedures. The mean size of target tumors embolized was (13.6±2.0) cm. All were hyper-vascular. The mean volume of lipiodol was (21.8±8.2) ml. Pulmonary oily embolisms were revealed within 12–48 hours after TACE. The most severe respiratory symptoms and imaging abnormalities of the eight patients who survived presented between 2 and 5 days after TACE, becoming normal between 12 and 35 days after TACE. Three patients died. Chest CT revealed retention of radiopaque lipiodol in lungs.
Conclusions Pulmonary lipiodol embolism occurs easily in patients who have large hyper-vascular hepatic malignant tumor. The high-density lipiodol deposition in the lung field can be used as diagnostic feature.
Transarterial chemoembolization (TACE) is a widely applied effective therapeutic modality for advanced unresectable hepatocellular carcinoma (HCC) and some hepatic metastases,1,2taking advantage of the fact that most hepatic malignancies derive their primary blood supply from the hepatic artery rather than the portal vein. Although a variety of microspheres, especially drug-loading beads, were developed and approved as effective materials for arterial embolization, there were no evidence-based data to support their superiority over traditional TACE with lipiodol. Chemoembolization with lipiodol remains the mainstay in the world and the standard of care in our hospital for its relative excellent visibility, effectiveness, lower cost, and safety. According to SIR guidelines,3patients with complications need a greater level of postprocedure care; postembolization syndrome (fever, pain, increased white blood cell count) after TACE is not considered a complication but an expected outcome. Complications included liver failure, hepatic abscess, postembolization syndrome requiring extended stay or readmission, surgical cholecystitis, biloma requiring percutaneous drainage, pulmonary arterial oil embolus, and gastrointestinal hemorrhage/ulceration.4-13Pulmonary oily embolism following TACE with lipiodol was occasionally reported in previous literatures; some of the complications were serious and life-threatening.
It was well known that hepatic arteriovenous (AVF) fistula was the main cause of pulmonary mis-embolization during hepatic arterial embolization.14,15However, in our experience, some patients with postprocedure pulmonary embolization did not show any AVF on pre-embolization angiography; rather, they had other features in common. Besides, the chest imaging characteristics of symptomatic pulmonary oily embolism which are different from thrombotic pulmonary embolism were not well addressed previously.4-7
The purpose of this study was to review the clinical, laboratory, and imaging findings of 11 patients with pulmonary oily embolism following TACE procedures in the past 2 years.
Study population
The study protocol was approved by the Ethics Committee of Peking University Cancer Hospital & Institute (Beijing, China). The medical records between January 2010 and March 2012 from the Hospital Information system (HIS) at our hospital were reviewed. A total of 478 consecutive patients underwent 1026 percutaneous TACE procedures during this period. Postprocedure pulmonary embolism was used as the search term to identify patients of interest from the above patient population using the HIS. Twenty-three patients who had pulmonary embolism records were found. Nine patients were excluded from pulmonary embolism. Three other patients were excluded from this study, because of preexisting (or concurrent) thrombotic pulmonary embolism. Eleven cases (all men whose ages ranged between 37 and 79 years) with respiratory symptoms were identified as having a major complication of TACE procedure.
Transarterial embolization
HCC and metastatic liver cancer chosen for TACE treatment were all unresectable. Among them 265 cases were HCC and 213 cases were metastatic liver cancer. Indications for TACE in the 11 patients are hepatocellular carcinoma(8 patients), hepatic metastases of gastrointestinal stromal tumors (2 patients) andhepatic metastases of lung cancer (1 patient). The diagnosis of HCC was based on biopsy, or clinical feature(s) including elevated levels of serum a-fetoprotein, the presence of chronic liver disease associated with hepatitis B or C virus, and classic imaging (CT, MRI, US) features. The diagnosis of liver metastasis was based on the history of primary tumor, classic imaging (CT, MRI) features of the tumor, or biopsy. The technique for TACE has been previously described.3,5,14Briefly, hepatic angiography was performed from a common femoral approach with a 5 F angiographic catheter to depict the hepatic artery anatomy, location of the tumors, tumor feeding artery, and portal vein patency. A 2.7 F microcatheter (Progreat, TERUMO, Tokyo, Japan) was superselectively inserted into tumor feeding arteries with the tip of the catheter in lobar, segmental, or subsegmental artery depending on the tumor size, the number of tumor nodules, and feeding arteries. TACE was performed with a suspension of 5–40 ml of iodized oil and epirubicin hydrochloride (40–80 mg, Pfizer, New York, NY, USA) followed by gelatin sponge particles 1–2 mm in diameter and/or polyvinyl alcohol (PVA, 90–500 µm, COOK, Bloomington, IN, USA) depending on the size and vascularity of the tumor.
Diagnosis of pulmonary oily embolism
In case the respiratory symptoms such ascough, dyspnea, or hemoptysis occurred after TACE procedure, a regular chest X-ray was performed to compare with the pre-procedure chest X-ray. For cases with positive chest X-ray suggestive of pulmonary embolism, chest CT scans were then usually performed. High-density lipiodol accumulation in chest imaging with or without the surrounding pulmonary interstitial exudation in patients with acute respiratory symptoms is regarded as the imaging diagnostic characteristic for pulmonary lipiodol embolism.
Treatments of pulmonary oily embolism
According to the medical records of cases with clinically diagnosed pulmonary oily embolism, close monitoring of vital signs, oxygen inhalation, anticoagulation medication, bronchodilation, and high-dose corticosteroids and antibiotics were applied. Assisted mechanical ventilator support therapies were given in cases of acute respiratory distress syndrome (ARDS).
Eleven (2.3%) of 478 consecutive patients who underwent percutaneous TACE procedures had a pulmonary oily embolism after procedures. Table 1 shows the diagnosis, tumor size, AVF, the embolization technique, oil consumption, and embolization materials. Table 2 shows the clinical symptoms, signs, pulmonary embolism occurrence time after the procedure, laboratory examination, treatment, and efficacy. The mean size of target tumors embolized was (13.6±2.0) cm, (range 10.0–16.6 cm); tumors larger than 10cm in diameter were found in all 11 patients. All the tumors were markedly hyper-vascular (Figures 1 and 2). Hepatic arteriovenous shunts were seen in only four patients (three HCCs and one metastasis) retrospectively but not identified in the procedure.
TACE was performed via right hepatic artery in nine patients, left hepatic artery in two, and right phrenic artery in two of the nine cases (Figure 3).
The mean volume of lipiodol was (21.8±8.2) ml. More than 30 ml of iodized oil was infused in two patients, 20–29 ml in six patients, 15 ml in two patients, and 10 ml in one patient. Gelatin sponge particles were used additionally in six patients, and gelatin sponge particles and PVA in four patients according to the feeding vessels of the tumors.
Pulmonary oily embolisms were revealed within 12 hours in five cases, 24 hours in three cases, and 48 hours in three cases after the TACE procedures. Common respiratory symptoms were found, such as cough in all cases, hemoptysis in four cases, tachypnea in nine cases, and dyspnea in eight cases.
PaO2decreased to between 35 and 80 mmHg and oxygen saturation to between 55% and 95%. A temperature higher than 38°C (maximum temperature 38.9°C) was found in 10 patients. White blood cell counts and the neuter granulocyte percentage increased in eight patients. Serum D-dimer increased in seven patients. Fine crackles were heard in both lungs of the symptomatic patients, especially prominent in the lower lungs.
In nine of all patients, chest radiographs showed diffuse and bilateral pulmonary parenchymal infiltrate, ranging from fine reticulonodular interstitial areas of opacity (2/11, Figure 4) to flocculent alveolar areas of opacity (7/11) (Figure 5); six cases can be seen with high-density scattered lipiodol deposition (Figure 6). In two other patients, chest radiographs did not show any pulmonary parenchymal infiltrate (2/11). Chest CTs were obtained in nine patients and revealed retention of radiopaque iodized oil in lower lobe of both lungs (9/9), atelectasis (7/9), interspersed patchy ground-glass opacity (6/9), and predominantly in bilateral lower lungs and pleural effusion (8/9) (Figure 7).
The most severe respiratory symptoms and imaging abnormalities of the eight patients presented between 2 and 5 days after TACE, even under active treatment with medicine and oxygen inhalation, then gradually resolved, becoming normal between 12 and 35 days after TACE, except for only one patient who had few fiber stripes left in the lower lungs for a long time. Three patients with irreversible severe respiratory distress syndrome eventually died: one of them with simultaneous coma caused by cerebral lipiodol embolism diagnosed with CT died 10 days later, one with severe ARDS within 3 hours after the procedure who refused ventilator treatment died 3 days later, one with ARDS during the procedure died of respiratory failure 14 days later.
TACE is an effective and safe method in the treatment of inoperable hepatic malignant tumors.1,2There are few reports of severe complications of TACE.4-7,14,16Pulmonary lipiodol embolism, first reported by Samejima et al in 1990,17is one of the most severe complications because it may be a potentially life-threatening emergency.4,6,14,16The incidence of symptomatic pulmonary oily embolism after TACE ranged between 0.05 and 3%.3-5,16The symptomatic pulmonary lipiodol embolism rate of 2.3% revealed in this study is comparable to prior studies.
It was well known that hepatic AVF was the main cause of pulmonary mis-embolization during hepatic arterial embolization.14,15Kan et al18reported the pathway of oil flowing into the pulmonary circulation after TACE is through hepatic AVF and normal hepatic arteriovenous anastomosis. Most of the previously reported pulmonary oily embolism was revealed in patients with AVF,4,6,14,15but was only shown in 4/11 cases in this study. At the same time, tumor larger than 10 cm was present in all 11 patients, and lipiodol consumption more than 20 ml in 8/11 patients. Although we did not have a matched control group for comparison, our data suggest that tumor larger than 10 cm and lipiodol consumption more than 20 ml may be other potential risk factors for pulmonary oily embolism, a possible mechanism of which is that larger tumors have relatively larger size of tumor feeding vessels and over-consumption of lipiodol increases the pressure on vessel wall causing more lipiodol passing through it.
The pathophysiology of lipiodol pulmonary embolism is different from that of thrombotic pulmonary embolism. The mechanism of symptomatic pulmonary injury is most likely a high concentration of free fatty acids resulting from breakdown of oil microemboli, which might lead to pulmonary capillary leakage and non-cardiogenic pulmonary edema,4,5,7There are also reports of acute-onset pulmonary edema after the administration of chemotherapeutic drugs, such as doxorubicin.19Alveolar capillary bed oily embolism caused decline in alveolar microcirculation function and gas exchange capacity. Iodized oil emboli catabolized and cleared by phagocyte caused non-specific inflammatory reaction followed by increase in alveolar wall vascular permeability and resulted in a large number of exudates filling the alveoli, affecting ventilation.10,15
Clinical manifestations, imaging findings, and the prognosis of these patients were similar to those in patients from previous researches with oil embolism after TACE for HCC4,6,7,14and similar to those in patients with oil embolism after lymphangiography,20and hysterosalpingography21performed with administration of iodized oil. But the imaging findings of pulmonary oily embolism have not been previously addressed systemically. We analyzed the patients’ chest X-ray and CT findings and found that the image findings of pulmonary oily embolism are different from thrombotic pulmonary embolism, which are summarized as follows: (1) high-density lipiodol deposition in the lung field was the characteristic finding of the pulmonary oily embolism and (2) scattered ground-glass opacity, atelectasis, and pleural effusion were non-characteristic. These imaging findings differ from those of thrombotic pulmonary embolism.
Regarding the treatment and outcome, all patients were given oxygen, anticoagulation, anti-infection, high-dose corticosteroids, and supportive treatment. Assisted mechanical ventilator support therapy was given in cases of ARDS (3/11). These methods of treatment were not specific and were similar to those reported previously,4-6,14We believed that the patients’ clinical symptoms and imaging manifestations varied depending on the amount of lipiodol in the pulmonary circulation.
How should we prevent iodine oil pulmonary embolism? Based on our experience, we suggest paying attention to the following points: (1) First, we should be aware of the high risk of pulmonary oily embolism when treating large liver tumor (>10 cm) with rich blood supply (including benign liver tumor). (2) During angiography, we should observe carefully whether hepatic AVF exists. If it does, we should reduce the degree of AV shunting by temporarily occluding the hepatic venous branch involved with AV shunting using balloon, the most successful method so far,22or by embolizing the AV shunts with polyvinyl alcohol, gelatin sponges, or coils as the alternative strategy.8,17,23TACE is contraindicated if AV shunting is refractory and severe after the above-mentioned strategies. Remember in our study, hepatic arteriovenous shunts were seen only in four patients retrospectively; we did not even see AV shunting in other symptomatic patients. Our experience from this study suggests that if the tumor is huge and hyper-vascular, “sandwich” embolization (PVA–lipiodol–PVA and/or gelatin sponges) can be applied, even when AV shunt is not seen. (3) Single lipiodol consumption cannot exceed 20 ml. (4) We should monitor the patient’s respiratory symptoms closely both during and after the procedure.
Pulmonary lipiodol embolism is a serious and rare complication after TACE. We believe that if the tumor is large, with rich blood supply, especially with hepatic AVF, pulmonary oily embolism risk increases greatly after TACE. Diagnostic features in our study were respiratory symptoms, signs, and chest imaging findings. The chest imaging characteristics of symptomatic pulmonary oily embolism are different from thrombotic pulmonary embolism. Especially, chest CT can show lipiodol deposition directly and be helpful for diagnosis of the complication. Therefore, for patients with suspected pulmonary oily embolism, chest CT examination can be used. The “sandwich” embolization technique can be applied in patients with huge and hyper-vascular hepatic tumor, even if AV shunt is not seen.
  • 1.

    Lo CM, Ngan H, Tso WK, Liu CL, Lam CM, Poon RT et al. Randomized controlled trial of transarterial lipiodol chemoembolization for unresectable hepatocellular carcinoma. Hepatology 2002; 35: 1164-1171.

  • 2.

    Gupta S, Yao JC, Ahrar K, MJ Wallace, FA Morello, DC Madoff, et al. Hepatic artery embolization and chemoembolization for treatment of patients with metastatic carcinoid tumors: the M.D. Anderson experience. Cancer J 2003; 9: 261-267.

  • 3.

    Brown DB, Nikolic B, Covey AM, Nutting CW, Saad WEA, Salem R, et al. Quality improvement guidelines for transhepatic arterial chemoembolization, embolization, and chemotherapeutic infusion for hepatic malignancy. J Vasc Interv Radiol 2012; 23: 287-294.

  • 4.

    Chung JW, Park JH, Im JG, Han JK, Han MC. Pulmonary oil embolism after transcatheter oily chemoembolization of hepatocellular carcinoma. Radiology 1993; 187: 689-693.

  • 5.

    Chung JW, Park JH, Han JK, Choi BI, Han MC, Lee HS, et al. Hepatic tumors: predisposing factors for complications of transcatheter oily chemoembolization. Radiology 1996; 198: 33-40.

  • 6.

    Wu JJ, Chao M, Zhang GQ, Li B, Dong F. Pulmonary and cerebral lipiodol embolism after transcatheter arterial chemoembolization in hepatocellular carcinoma. World J Gastroenterol 2009; 15: 633-635.

  • 7.

    Xia J, Ren Z, Ye S, Sharma D, Lin Z, Gan Y, et al. Study of severe and rare complications of transarterial chemoembolization (TACE) for liver cancer. Eur J Radiol 2006; 59: 407-412.

  • 8.

    Brown KT, Nevins AB, Getrajdman GI, Brody LA, Kurtz RC, Fong Y, et al. Particle embolization for hepatocellular carcinoma. J Vasc Interv Radiol 1998; 9: 822-828.

  • 9.

    Geschwind JF, Kaushik S, Ramsey DE, Choti MA, Fishman EK, Kobeiter H. Influence of a new prophylactic antibiotic therapy on the incidence of liver abscesses after chemoembolization treatment of liver tumors. J Vasc Interv Radiol 2002; 13: 1153-1156.

  • 10.

    Gates J, Hartnell GG, Stuart KE, Clouse ME. Chemo-embolization of hepatic neoplasms: safety, complications, and when to worry. Radiographics1999; 19: 399-414.

  • 11.

    Kim W, Clark TWI, Baum RA, Soulen MC. Risk factors for liver abscess formation after hepatic chemoembolization. J Vasc Interv Radiol 2001; 12: 965-968.

  • 12.

    Carter S, Martin RC II. Drug-eluting bead therapy in primary and metastatic disease of the liver. HPB (Oxford) 2009; 11: 541-550.

  • 13.

    Sakamoto I, Aso N, Nagaoki K, Matsuoka Y, Uetani M, Ashizawa K, et al. Complications associated with transcatheter arterial embolization for hepatic tumors. Radiographics 1998; 18: 605-619.

  • 14.

    Tajima T, Honda H, Kuroiwa T, Yabuuchi H, Okafuji T, Yosimitsu K, et al. Pulmonary complications after hepatic artery chemoembolization or infusion via the inferior phrenic artery for primary liver cancer. J Vasc Interv Radiol 2002; 13: 893-900.

  • 15.

    Lee JH, Won JH, Park SI, Won JY, Lee do Y, Kang BC. Transcatheter arterial chemoembolization of hepatocellular carcinoma with hepatic arteriovenous shunt after temporary balloon occlusion of hepatic vein. J Vasc Interv Radiol 2007; 18: 377-382.

  • 16.

    Shiah HS, Liu TW, Chen LT, Chang JY, Liu JM, Chuang TR, et al. Pulmonary embolism after transcatheter arterial chemoembolization. Eur J Cancer Care 2005:14: 440-442.

  • 17.

    Samejima M, Tamura S, Kodama T, Yuuki Y, Takasaki J, Sekiva R, et al. Pulmonary complication following intra-arterial infusion of lipiodol-adriamycin emulsion for hepatocellular carcinoma, report of a case. Nippon Igaku Hoshasen Gakkai Zasshi 1990; 50: 24-28.

  • 18.

    Kan Z, Ivancev K, Hagerstrad I, Chuang VP, Lunderquist A. In vivo microscopy of the liver after injection of Lipiodol into the hepatic artery and portal vein in the rat. Acta Radiol 1989; 30: 417-425.

  • 19.

    Moon TY, Kim BH, Koo BS. Experimental study on the toxicity of Lipiodol injected into proper hepatic artery of rabbits. J Korean Radiol Soc 1990; 26: 449-461.

  • 20.

    Clouse ME, Hallgrimsson J, Wenlund DE. Complications following lymphography with particular reference to pulmonary oil embolization. Am J Roentqenol Radium Ther Nucl Med 1966; 96: 972-978.

  • 21.

    Shapiro JH, Rubinstein B, Jacobson HG, Poppel MH. Pulmonary oil embolism: a complication of hysterosalpingography. Am J Roentqenol Radium Ther Nucl Med 1957; 77: 1055-1105.

  • 22.

    Murata S, Tajima H, Nakazawa K, Onozawa S, Kumita S, Nomura K. Initial experience of transcatheter arterial chemoembolization during portal vein occlusion for unresectable hepatocellular carcinoma with marked arterioportal shunts. Eur Radiol 2009; 19: 2016-2023.

  • 23.

    Kim YJ, Lee HG, Park JM. Polyvinyl alcohol embolization adjuvant to oily chemoembolization in advanced hepatocellular carcinoma with arterioportal shunts. Korean J Radiol 2007; 8: 311-319.

(Received June 19, 2013)
Edited by Cui Yi
Table 1.General conditions of the patients and TACE method
Patient No./sex/age (year) Diagnosis Tumor size (cm) AVF IO (ml) CD/dose (mg) GSP PVA
01/M/49 HCC 14.0×12.0 Not seen 20 EPI/40 mg A NA
02/M/37 HCC 15.0×11.0 Not seen 25 EPI/40 mg A NA
03/M/75 HCC 11.0×10.6 Seen 10 EPI/20 mg A NA
04/M/79 HCC 12.0×11.5 Not seen 30 EPI/60 mg A NA
05/M/78 HCC 14.5×12.0 Not seen 40 EPI/60 mg A NA
06/M/63 HCC 16.0×14.6 Seen 20 EPI/40 mg A A
07/M/38 HCC 13.0×12.6 Seen 15 EPI/40 mg A A
08/M/65 HCC 10.0×9.5 Not seen 25 EPI/60 mg A NA
09/M/63 Meta 13.6×12.0 Not seen 20 NA A A
10/M/73 Meta 14.5×11.0 Not seen 20 EPI/40 mg NA NA
11/M/55 Meta 16.6×14.0 Seen 15 EPI/60 mg A A
A: applicable; AVF: arteriovenous fistula; CD:chemotherapeuticdrug; EPI:epirubicin hydrochloride; GSP: gelatin sponge particle; HCC: Hepatocellular carcinoma; IO: iodized oil;M: male; Meta: Metastases; NA: not applicable. PVA: polyvinyl alcohol.
Table 2.Clinical symptoms, signs, and results of laboratory examination from 11 patients with symptomatic pulmonary lipiodol embolism after TACE
Patient No./
sex/age (y)
Respiratory symptoms
Maximum hypoxemia
Maximum body temperature (°C)
WBC counts (×109)
Symptom Onset time (h or d) BOS (%) PaO2(mmHg) Day
01/M/49 C.T.D 1 d 86 49 5 38.2 13.14 918 Recovered14 d after TACE
02/M/37 C 2 h 95 80 NM 38.5 8.98 NA Recovered 12 d after TACE
03/M/75 C.H 2 d 95 NA NM 38.4 11.02 224 Recovered 13 d after TACE
04/M/79 C.H.T.D 2 d 90 71 5 38.9 11.49 489 Recovered 14 d after TACE
05/M/78 C.T.D 1 d 70 56 2 38.2 15.20 367 Recovered25 d after TACE
06/M/63 C.H.T.D 2 d 90 NA 2 38.3 7.56 NA Recovered 14 d after TACE
07/M/38 C.T.D 1 d 93 75 2 38.0 9.40 593 Recovered 16 d after TACE
08/M/65 C.T.D 5 h 89 58 3 38.1 10.89 NA Dead
09/M/63 C.T.D 1 h 86 69 3 38.7 28.14 1390 Dead
10/M/73 C.T.D 0 h 55 35 2 38.2 17.65 783 Dead
11/M/55 C.H.T 0 h 94 NA 2 37.8 11.23 NA Recovered 14 d after TACE
A:applicable; BOS: Blood oxygen saturation; C:cough; D: dyspnea; H:hemoptysis; M: male; NA:not applicable; NM:not measured; PaO2: partial pressure of arterial oxygen; T:tachypnea; y: years; h: hours; d: day.

view in a new window
Figure 1. Abdominal enhanced arterial phase CT image shows the liver tumor located at the right lobe with rich blood supply.
Figure 2. Hepatic artery angiography shows a lot of tumor vessels.
Figure 3. The right phrenic arteriography shows visible phrenic artery participation in the tumor blood supply.

view in a new window
Figure 4.Patient 3: the day after TACE, chest radiographs show reticulonodular change.
Figure 5.Patient 6: the chest radiograph showed normal before TACE (5A), then with bilateral scattered patchy exudation and left pleural effusion the next day after TACE (5B), exudation absorption 8 days after TACE (5C).
Figure 6.Patient 7: chest X-ray showed bilateral pulmonary scattered lipiodol deposition as soon as TACE was finished.

view in a new window
Figure 7. Patient 7: interspersed patchy ground-glass distributed in bilateral lung (A) and (B); high-density lipiodol deposition, atelectasis, and pleural effusion (C) and (D).
  1. grants from National Natural Science Foundation of China (No. 81172120) and China International Medical Foundation (No. 2008-17).