**Hypothermic Cardiac Arrest to Remove Right Atrial Thrombi Due to Abdominal Malignancies**

Pier Luigi Stefano, Stefano Romagnoli, Donata Villari and Giacomo Batignani *Università degli Studi di Firenze, Italy* 

#### **1. Introduction**

146 Front Lines of Thoracic Surgery

Wippermann, J., Albes, JM., Hartrumpf, M., Kaluza, M., Vollandt, R., Bruhin, R. & Wahlers,

127-32.

T. (2005). Comparison of minimally invasive closed circuit extracorporeal circulation with conventional cardiopulmonary bypass and with off-pump technique in CABG patients: selected parameters of coagulation and inflammatory system. European Journal of Cardiothoracic Surgery. Vol. 28, No 1 (July 2005) pp.

> The first use of hypothermia in cardiac surgery is attributed to Dr. John Lewis who performed an atrial septal defect closure on September 2, 1952 at the University of Minnesota (Gott, 2005).

> Cardiopulmonary by-pass (CPB) and hypothermic cardio-circulatory arrest (HCCA) has been introduced for the first time in clinical practice for aortic arc substitution (Pierangeli et al., 1974). More recently the same technique has been proposed to remove atrial thrombi originated from renal carcinomas (Marshall et al., 1984), adrenal carcinomas (Shahinian et al., 1989), hepatocellular carcinoma (Hamazaki et al., 1995) and other abdominal malignancies such as caval leiomyosarcoma, uterine endometrial sarcoma or intravenous leiomyomatosis, ovarian or testicular tumors (Hassan et al.,, 2010; Vargas-Barron et al., 1990; Ariza et al., 1982; Kanda et al., 1991).

> There is still a debate pro and cons the use of hypothermic arrest with some favours the normothermic (Lubahn et al., 2006; Stewart et al., 1991) and some others the hypothermic, because of the operative field appears better exposed and almost completely bloodless so the blood loss resulted much lower. Furthermore they claim a better visualization of some critical areas with an easier removal of the tumoural thrombus, that may invade the hepatic veins or the coronaric sinus, may either remain attached to the tricuspid valve or may have embolized into the pulmonary artery (Chiappini et al., 2002; Kalkat et al., 2008; Leo et al., 2010; Topcouglu et al., 2004). In all these occasions HCCA offers the possibility of a better tumoural cleaning with the possibility of R0 resection. The incomplete removal of this tumoural thrombi in fact is correlated to an early recurrence and a worst postoperative survival (Skinner et al., 1989). Furthermore the hepatic, renal and splacnic damage from warm ischemia due to the Pringle's manoeuvre and/or aortic/mesenteric cross-clamping (usually necessary when CPB is used without cardio-circulatory arrest) is reduced when the HCCA is instead used (Chiappini et al., 2002, Davlouros et al., 2005). This permits its use even in the setting of mild hepatic damage (Leo et al., 2010).

#### **2. Kidney**

Tumour thrombus extension in the inferior vena cava (IVC) occurs in 4 % to 10% of patients with renal cell carcinoma (RCC) (Marshall et al., 1988). The cephalic extension of tumour

Hypothermic Cardiac Arrest to Remove Right Atrial Thrombi Due to Abdominal Malignancies 149

benefit (Hatcher et al., 1991). Grafting or reconstitution sometimes is required only in those patients with not complete occlusion of IVC and collateral blood flow. In 1913, Berg first described nephrectomy and vena caval thrombectomy as primary surgical treatment of these patients (Berg, 1913). Usually, in level I and II disease the tumour thrombus can safely be excised by means of proximal and distal control of the IVC. In level III and IV disease the exposure and isolation of IVC requires liver mobilization with or without the use of CBP. Several techniques have been used during the years to remove large caval thrombi when distal caval control cannot be achieved below the liver such as crossclamping of the supraceliac aorta (Cummings et al., 1979), occlusion of the intra pericardial vena cava with simultaneous occlusion of the "porta hepatis" and superior mesenteric artery (Skinner et al., 1989), use of an intraoperative venous (cava-atrial) shunt (Foster et al.,1988), and cardiopulmonary by pass alone (Novick & Cosgrove, 1980). The use of CBP alone is limited to the period during which the liver and the kidneys undergo warm ischemia. Skinner and co-workers reported an average warm ischemia of 14 minutes ranging from 8 to 20 minutes. Post operative complications occurred in 41% to 60% of the patients and included transient hyperbilirubinemia, renal dysfunction and respiratory failure (Skinner et al., 1989). When the thrombus extends above the diaphragm the use of CPB with deep HCCA have been used. Marshall et al., 1984 and Krane et al., 1984 first demonstrated the feasibility of this approach. The most common post operative complications related to this procedure are haemorrhage, coagulopathy, increased transient and/or permanent neurologic risk. Novick et al., 1990 in their experience reported up to 60 minutes of safe ischemia with direct visual inspection of the entire vena caval lumen in a bloodless field reducing risk of sudden massive intra-operative haemorrhage or distal pulmonary tumour Thrombo-embolization. Operative mortality rate reported was 4.7% (Novick et al., 1990). Heart surgery have been used more frequently to treat intra-cardiac extension of infra-diaphragmatic tumours and the safety of this procedure has increased. Chiappini and co-workers in their series of 13 patients reported no in-hospital deaths while 2 patients experienced postoperative complications such as respiratory failure and bleeding (Chiappini et al., 2002). On the other side, Ruel and co-workers in 2001 reported their experience in cavo-atrial thrombus excision in 6 patients affected by RCC without circulatory arrest (Ruel et al., 2001). During the last 10 years the traditional approach using CBP and HCCA was revisited. Deep hypothermic circulatory arrest was therefore not considered imperative and some studies reported that its avoidance may decrease morbidity and mortality. Shinghal et al., in 2003 reported in a single patient a novel technique by placing an aortic occlusive balloon in the abdominal aorta at the level of the diaphragm, limiting flow in the IVC and maintaining both cerebral and spinal cord perfusion during cardiopulmonary bypass (Shingai et al., 2003). In 2005 Ciancio & Soloway described a technique that can be safely used to approach RCC with a tumour thrombus extending into the supra-diaphragmatic IVC and right atrium through a trans-abdominal approach without sternotomy and/or cardiopulmonary bypass utilizing a complete liver mobilization and minimizing the vena caval cross–clamp time. The trans-abdominal incision exposes the intra-pericardial IVC and right atrium transdiaphragmatically. (Ciancio & Soloway, 2005; Ciancio G, Shirodkar S et al.,2009). Step by step description of this technique based on orthotopic liver transplantation was reported in 2011 by the same author regarding a cohort of 56 patients with retro-hepatic extension of the thrombus and 12 patients with a supra-hepatic/intra atrial localizations. Mean operative time was 5h and 32 min. Five patients required CPB (7.3%). Three patients died

thrombus in RCC was first defined in 1987 by Neves and Zincke (Neves & Zincke, 1987) and classified in four levels (level I : thrombus at the level of the renal vein, level II : thrombus extending into the infra-hepatic IVC, level III : thrombus extending into the retro-hepatic IVC, level IV : intra-atrial thrombus (suprahepatic IVC involvement or right atrium thrombus). Libertino classification system (Moinzadeh & Libertino, 2004) distinguishes: renal vein only involvement, level I caval involvement (IVC below the diaphragm), level II caval involvement (IVC above the diaphragm) and level III involvement (when the thrombus reaches the atrium).

Recently an updated staging system by the American Joint Committee on Cancer (AJCC) and the "Union International Contre le Cancer" (UICC) was utilized and the current staging, according to the 2009 UICC/AJCC TNM staging classification system, designates renal vein involvement as pT3a Level III, when the thrombus involves the intra-hepatic portion of the IVC but below the diaphragm is designated as p T3b, while level IV when the thrombus extends above the diaphragm, as T3c. The thrombus may extend as far as the right atrium (IV level) in 1% of patients with RCC. Accurate information about the presence and complete extent of IVC involvement is essential for surgical planning. Routine CT scanning and abdominal ultrasonography are less reliable in delineating the cephalic extent of the thrombus. Contrast inferior venacavography is an accurate, but invasive, diagnostic tool and a single anterograde study may be insufficient in complete caval obstruction. Occasionally has been employed trans oesophageal echocardiography and trans-abdominal colour flow Doppler ultrasonography. MRI has become the preferred, non invasive, diagnostic study though its frequent inability to differentiate actual invasion of the vena cava wall which is generally associate with a poor prognosis. Contrast inferior venography is reserved for patients in whom MRI finings are equivocal or for whom MRI is contraindicated. Another option in this group of patients is multiplanar CT.

Renal arteriography is useful in the preoperative study because it may highlighted a distinct vascularisation of the thrombus (observed in 35% to 40% of the cases) allowing to perform a preoperative embolization of the kidney, causing the shrinkage of the thrombus and its more easily intra-operative removal.

The prognostic significance of venous involvement and tumour thrombus level remains highly controversial. According to the literature in the absence of metastasis, when adjusting for clinical and pathologic features, radical nephrectomy with total tumour thrombus excision in patients with RCC and level I to III vena caval involvement is associated with a cancer specific 5 –year survival rate between 47% and 68% (Staheler & Brkovic, 2000; Belis et al., 2000).

Some authors generally accepted that neoplastic extension in the IVC was not a prognostic determining factor (Libertino et al., 1987; Skinner et al., 1989; Ciancio et al., 2010). Some others, more recently, have shown that tumour thrombus level is an independent predictor of survival (Mart´nez-Salamanca et al., 2011).

RCC with vascular invasion into the IVC, direct invasion of the vein's wall appears to be an important prognostic factor and should be noted during tumour staging. With no perinephric fat or lymph nodal involvement, patients who undergo tumour excision with radical nephrectomy and IVC thrombectomy have an overall and cancer specific 5–year survival of 30% to 72% with an operative mortality of 2.7% to 13.5 %. When the thrombus locally invades the wall of the vena cava, aggressive resection of the vascular wall with reconstruction along with the attainment of negative surgical margins provides a survival

thrombus in RCC was first defined in 1987 by Neves and Zincke (Neves & Zincke, 1987) and classified in four levels (level I : thrombus at the level of the renal vein, level II : thrombus extending into the infra-hepatic IVC, level III : thrombus extending into the retro-hepatic IVC, level IV : intra-atrial thrombus (suprahepatic IVC involvement or right atrium thrombus). Libertino classification system (Moinzadeh & Libertino, 2004) distinguishes: renal vein only involvement, level I caval involvement (IVC below the diaphragm), level II caval involvement (IVC above the diaphragm) and level III involvement (when the

Recently an updated staging system by the American Joint Committee on Cancer (AJCC) and the "Union International Contre le Cancer" (UICC) was utilized and the current staging, according to the 2009 UICC/AJCC TNM staging classification system, designates renal vein involvement as pT3a Level III, when the thrombus involves the intra-hepatic portion of the IVC but below the diaphragm is designated as p T3b, while level IV when the thrombus extends above the diaphragm, as T3c. The thrombus may extend as far as the right atrium (IV level) in 1% of patients with RCC. Accurate information about the presence and complete extent of IVC involvement is essential for surgical planning. Routine CT scanning and abdominal ultrasonography are less reliable in delineating the cephalic extent of the thrombus. Contrast inferior venacavography is an accurate, but invasive, diagnostic tool and a single anterograde study may be insufficient in complete caval obstruction. Occasionally has been employed trans oesophageal echocardiography and trans-abdominal colour flow Doppler ultrasonography. MRI has become the preferred, non invasive, diagnostic study though its frequent inability to differentiate actual invasion of the vena cava wall which is generally associate with a poor prognosis. Contrast inferior venography is reserved for patients in whom MRI finings are equivocal or for whom MRI is contraindicated. Another

Renal arteriography is useful in the preoperative study because it may highlighted a distinct vascularisation of the thrombus (observed in 35% to 40% of the cases) allowing to perform a preoperative embolization of the kidney, causing the shrinkage of the thrombus and its

The prognostic significance of venous involvement and tumour thrombus level remains highly controversial. According to the literature in the absence of metastasis, when adjusting for clinical and pathologic features, radical nephrectomy with total tumour thrombus excision in patients with RCC and level I to III vena caval involvement is associated with a cancer specific 5 –year survival rate between 47% and 68% (Staheler & Brkovic, 2000; Belis et

Some authors generally accepted that neoplastic extension in the IVC was not a prognostic determining factor (Libertino et al., 1987; Skinner et al., 1989; Ciancio et al., 2010). Some others, more recently, have shown that tumour thrombus level is an independent predictor

RCC with vascular invasion into the IVC, direct invasion of the vein's wall appears to be an important prognostic factor and should be noted during tumour staging. With no perinephric fat or lymph nodal involvement, patients who undergo tumour excision with radical nephrectomy and IVC thrombectomy have an overall and cancer specific 5–year survival of 30% to 72% with an operative mortality of 2.7% to 13.5 %. When the thrombus locally invades the wall of the vena cava, aggressive resection of the vascular wall with reconstruction along with the attainment of negative surgical margins provides a survival

thrombus reaches the atrium).

option in this group of patients is multiplanar CT.

of survival (Mart´nez-Salamanca et al., 2011).

more easily intra-operative removal.

al., 2000).

benefit (Hatcher et al., 1991). Grafting or reconstitution sometimes is required only in those patients with not complete occlusion of IVC and collateral blood flow. In 1913, Berg first described nephrectomy and vena caval thrombectomy as primary surgical treatment of these patients (Berg, 1913). Usually, in level I and II disease the tumour thrombus can safely be excised by means of proximal and distal control of the IVC. In level III and IV disease the exposure and isolation of IVC requires liver mobilization with or without the use of CBP. Several techniques have been used during the years to remove large caval thrombi when distal caval control cannot be achieved below the liver such as crossclamping of the supraceliac aorta (Cummings et al., 1979), occlusion of the intra pericardial vena cava with simultaneous occlusion of the "porta hepatis" and superior mesenteric artery (Skinner et al., 1989), use of an intraoperative venous (cava-atrial) shunt (Foster et al.,1988), and cardiopulmonary by pass alone (Novick & Cosgrove, 1980). The use of CBP alone is limited to the period during which the liver and the kidneys undergo warm ischemia. Skinner and co-workers reported an average warm ischemia of 14 minutes ranging from 8 to 20 minutes. Post operative complications occurred in 41% to 60% of the patients and included transient hyperbilirubinemia, renal dysfunction and respiratory failure (Skinner et al., 1989). When the thrombus extends above the diaphragm the use of CPB with deep HCCA have been used. Marshall et al., 1984 and Krane et al., 1984 first demonstrated the feasibility of this approach. The most common post operative complications related to this procedure are haemorrhage, coagulopathy, increased transient and/or permanent neurologic risk. Novick et al., 1990 in their experience reported up to 60 minutes of safe ischemia with direct visual inspection of the entire vena caval lumen in a bloodless field reducing risk of sudden massive intra-operative haemorrhage or distal pulmonary tumour Thrombo-embolization. Operative mortality rate reported was 4.7% (Novick et al., 1990). Heart surgery have been used more frequently to treat intra-cardiac extension of infra-diaphragmatic tumours and the safety of this procedure has increased. Chiappini and co-workers in their series of 13 patients reported no in-hospital deaths while 2 patients experienced postoperative complications such as respiratory failure and bleeding (Chiappini et al., 2002). On the other side, Ruel and co-workers in 2001 reported their experience in cavo-atrial thrombus excision in 6 patients affected by RCC without circulatory arrest (Ruel et al., 2001). During the last 10 years the traditional approach using CBP and HCCA was revisited. Deep hypothermic circulatory arrest was therefore not considered imperative and some studies reported that its avoidance may decrease morbidity and mortality. Shinghal et al., in 2003 reported in a single patient a novel technique by placing an aortic occlusive balloon in the abdominal aorta at the level of the diaphragm, limiting flow in the IVC and maintaining both cerebral and spinal cord perfusion during cardiopulmonary bypass (Shingai et al., 2003). In 2005 Ciancio & Soloway described a technique that can be safely used to approach RCC with a tumour thrombus extending into the supra-diaphragmatic IVC and right atrium through a trans-abdominal approach without sternotomy and/or cardiopulmonary bypass utilizing a complete liver mobilization and minimizing the vena caval cross–clamp time. The trans-abdominal incision exposes the intra-pericardial IVC and right atrium transdiaphragmatically. (Ciancio & Soloway, 2005; Ciancio G, Shirodkar S et al.,2009). Step by step description of this technique based on orthotopic liver transplantation was reported in 2011 by the same author regarding a cohort of 56 patients with retro-hepatic extension of the thrombus and 12 patients with a supra-hepatic/intra atrial localizations. Mean operative time was 5h and 32 min. Five patients required CPB (7.3%). Three patients died

Hypothermic Cardiac Arrest to Remove Right Atrial Thrombi Due to Abdominal Malignancies 151

with percentage of incidence reported of 2.9% by imaging techniques, 0.7% at operation, and

Therapeutics options at this stage are few and debatable because of the high risk for lung metastasis (Masaki et al., 1995) but, on the other hand, the risk of developing a "ball valve syndrome"(Hahne & Climie, 1962) or a pulmonary embolism is very high (Fujisaki et al.,1991). The overall survival time of these patients with cavo-atrial thrombi and who do not undergo operative treatment is reported to be very low: from 3 days to 2 months (Lin et

Liver resection and cavo-atrial thrombectomy remains, in this situation, the only effective therapeutic option with reported survival time ranging from 5 to 56 months (Miyazawa et

The firsts authors describing the removal of a tumoural thrombus in the RA from HCC on cirrhosis using CBP were Goto et al. in 1986 who removed only the thrombus. Fujisaki et al. in 1991 first removed the thrombus along with the primary tumour using CPB hepatic vascular exclusion while Hamazaki et al. in 1995 added the use of HCCA in a similar situation. Indeed, HCCA was first applied in liver surgery, in a young child to remove atrial thrombi from hepatoblastoma, which usually occurs on a normal liver parenchyma, in 1981

The majority of patients with HCC has a diseased liver because of the presence of hepatitis viruses infection and /or previous alcohol abuse. The presence of hepatic vein, vena cava and atrial thrombosis contraindicate liver transplantation and, liver resection with thrombectomy, is the only real therapeutic option in this setting (Wu et al., 2000). As previously stated, acute pulmonary embolism or congestive heart failure may complicate liver tumours involving hepatic vein with tumours thrombi in the inferior vena cava/right atrium. This latter situation represent an impending life-threatening condition that should

Atrial thrombus removal may be achieved under normothermic CPB and total liver vascular exclusion. Tolerance of the cirrhotic liver to normothermic ischemia is reported as safe up to 33 minutes (Huguet et al., 1994) which is quite similar to the safety time reported to avoid

Extracorporeal circulation with HCCA has been increasingly used to remove atrial thrombi due to retroperitoneal malignancies (Ogata et al., 2006) but its use in cirrhotic patients has been limited, for many years, because of the fear of post-operative liver failure and poor outcome. However, in more recent time, we have assisted to an increase number of reports where HCCA has been used to remove cavo-atrial thrombi due to HCC with no mortality and low morbidity, probably, lower than expected thus, we can assume that, a short lasting HCCA, has to be considered safe in a well compensated liver cirrhosis (Novick et al., 1990; Yogitaet al., 2000; Ohwada et al., 2008; Florman et al., 2009; Leo et al.,

Since the vast majority of patients with HCC and atrial tumour thrombi has large type of HCC, a major liver resection is usually required and the possibility of postoperative liver failure raises accordingly, liver function tests and indocyanin green retention test, should be carried out in these patients in order to demonstrate the presence of a good functional reserve (Kawasaki et al., 1995). Furthermore we think that adding a sequential arterial and portal embolization of a hemi-liver could lessen the possibilities of post-operative liver

be surgically treated as soon as possible (Novick et al., 1990).

failure when a major resection is planned (Ogata et al., 2006).

neurological complications when using HCCA (McCullough et al., 1999).

18.2% at autopsy in Japan (Yogita et al., 2000).

al., 2007)

al., 2005).

(Ein et al., 1981).

2010).

in the immediate post operative period (4.4%) (Ciancio G, Gonzalez J,et al., 2011). In 2007 Chowdhury et al., reported their experience in a cohort of 6 patients with RCC and intra atrial tumour thrombus using mild hypothermic CPB and intermittent cross- clamping of the supra-celiac abdominal aorta (Chowdhury et al., 2007)

RCC with vascular invasion into the IVC remains a crucial challenge to the urologic oncologist. Complete excision of the tumour together with the thrombus and negative margins offers the best chance for patients survival. Surgical resection of these complex tumours requires multidisciplinary approach.

## **2.1 Adrenal**

The use of CPB to remove the tumour thrombus from adrenocortical carcinoma (ACC) was first reported in 1976 (Scully et al., 1976) while the first procedure using HCCA was in 1989 (Shahian et al., 1989) but because of the poor prognosis of this tumour at this stage and the technical difficulties that has to be faced such as the invasion of surrounding organs and vascular control/reconstruction, radical surgical resection remains controversial. A recent review from France (Chiche et al., 2006) collected the outcome of 32 patients reported in literature with cavo-atrial invasion in whom a CPB was used in 26 including 6 under HCCA. Reported peri-operative mortality was 3/26 (11.5%) and took place in patients operated using CPB alone.

In their own series of ACC with tumour thrombi extending into the vena cava or in the right atrium, they reported the outcome of 15 patients treated by a combination of venous control ranging from cross clamping to hepatic vascular exclusion or normo/ipothermic cardiopulmonary by-pass. The extension grade C (above the liver) was in 7 patients, including 4 with atrial thrombi and represents the 7% of the overall series. Of these patients 3 had HCCA and 1 (33.3%) died p.o. from multi-organ failure but the patient was cirrhotic with a portal thrombosis.

They concluded that CPB should always be used for tumours extending to the cavo-atrial junction or into the right atrium and additional HCCA provides a bloodless field that allow a precise dissection of the thrombus while reducing the blood loss.

#### **2.2 Liver**

Macroscopic vascular invasion (either portal or supra-hepatic) is a major complication of hepatocellular carcinoma (HCC). Major vascular invasion is the evolution of microscopic vascular invasion over the time so the liver tumour is of the large type at diagnosis and usually require major liver resection.

The natural history of HCC complicated with macroscopic vascular invasion of the portal vein, shows a median survival time of only 9 to 10 weeks (Okada et al., 1992).

Treatments with systemic chemotherapy, intra-arterial chemotherapy or radiofrequency ablation result in a dismal survival at 1 year ranging from 7% to 18% (Raul et al., 1994; Akashi et al., 1991; Poon et al., 2003). Hepatic resection remains the only option for these patients and a recent multicentric study showed that, resected patients lived longer, compared to those not treated surgically, with a five-year survival rate of 10% (Pawlik et al., 2005; Lin et al., 2007).

Liver tumour with extension to the hepatic veins represents instead a controversial issue and its extension in the right atrium (RA) via the inferior vena cava is more uncommon,

in the immediate post operative period (4.4%) (Ciancio G, Gonzalez J,et al., 2011). In 2007 Chowdhury et al., reported their experience in a cohort of 6 patients with RCC and intra atrial tumour thrombus using mild hypothermic CPB and intermittent cross- clamping of

RCC with vascular invasion into the IVC remains a crucial challenge to the urologic oncologist. Complete excision of the tumour together with the thrombus and negative margins offers the best chance for patients survival. Surgical resection of these complex

The use of CPB to remove the tumour thrombus from adrenocortical carcinoma (ACC) was first reported in 1976 (Scully et al., 1976) while the first procedure using HCCA was in 1989 (Shahian et al., 1989) but because of the poor prognosis of this tumour at this stage and the technical difficulties that has to be faced such as the invasion of surrounding organs and vascular control/reconstruction, radical surgical resection remains controversial. A recent review from France (Chiche et al., 2006) collected the outcome of 32 patients reported in literature with cavo-atrial invasion in whom a CPB was used in 26 including 6 under HCCA. Reported peri-operative mortality was 3/26 (11.5%) and took place in patients operated

In their own series of ACC with tumour thrombi extending into the vena cava or in the right atrium, they reported the outcome of 15 patients treated by a combination of venous control ranging from cross clamping to hepatic vascular exclusion or normo/ipothermic cardiopulmonary by-pass. The extension grade C (above the liver) was in 7 patients, including 4 with atrial thrombi and represents the 7% of the overall series. Of these patients 3 had HCCA and 1 (33.3%) died p.o. from multi-organ failure but the patient was cirrhotic with a

They concluded that CPB should always be used for tumours extending to the cavo-atrial junction or into the right atrium and additional HCCA provides a bloodless field that allow

Macroscopic vascular invasion (either portal or supra-hepatic) is a major complication of hepatocellular carcinoma (HCC). Major vascular invasion is the evolution of microscopic vascular invasion over the time so the liver tumour is of the large type at diagnosis and

The natural history of HCC complicated with macroscopic vascular invasion of the portal

Treatments with systemic chemotherapy, intra-arterial chemotherapy or radiofrequency ablation result in a dismal survival at 1 year ranging from 7% to 18% (Raul et al., 1994; Akashi et al., 1991; Poon et al., 2003). Hepatic resection remains the only option for these patients and a recent multicentric study showed that, resected patients lived longer, compared to those not treated surgically, with a five-year survival rate of 10% (Pawlik et al.,

Liver tumour with extension to the hepatic veins represents instead a controversial issue and its extension in the right atrium (RA) via the inferior vena cava is more uncommon,

vein, shows a median survival time of only 9 to 10 weeks (Okada et al., 1992).

a precise dissection of the thrombus while reducing the blood loss.

the supra-celiac abdominal aorta (Chowdhury et al., 2007)

tumours requires multidisciplinary approach.

**2.1 Adrenal** 

using CPB alone.

portal thrombosis.

2005; Lin et al., 2007).

usually require major liver resection.

**2.2 Liver** 

with percentage of incidence reported of 2.9% by imaging techniques, 0.7% at operation, and 18.2% at autopsy in Japan (Yogita et al., 2000).

Therapeutics options at this stage are few and debatable because of the high risk for lung metastasis (Masaki et al., 1995) but, on the other hand, the risk of developing a "ball valve syndrome"(Hahne & Climie, 1962) or a pulmonary embolism is very high (Fujisaki et al.,1991). The overall survival time of these patients with cavo-atrial thrombi and who do not undergo operative treatment is reported to be very low: from 3 days to 2 months (Lin et al., 2007)

Liver resection and cavo-atrial thrombectomy remains, in this situation, the only effective therapeutic option with reported survival time ranging from 5 to 56 months (Miyazawa et al., 2005).

The firsts authors describing the removal of a tumoural thrombus in the RA from HCC on cirrhosis using CBP were Goto et al. in 1986 who removed only the thrombus. Fujisaki et al. in 1991 first removed the thrombus along with the primary tumour using CPB hepatic vascular exclusion while Hamazaki et al. in 1995 added the use of HCCA in a similar situation. Indeed, HCCA was first applied in liver surgery, in a young child to remove atrial thrombi from hepatoblastoma, which usually occurs on a normal liver parenchyma, in 1981 (Ein et al., 1981).

The majority of patients with HCC has a diseased liver because of the presence of hepatitis viruses infection and /or previous alcohol abuse. The presence of hepatic vein, vena cava and atrial thrombosis contraindicate liver transplantation and, liver resection with thrombectomy, is the only real therapeutic option in this setting (Wu et al., 2000). As previously stated, acute pulmonary embolism or congestive heart failure may complicate liver tumours involving hepatic vein with tumours thrombi in the inferior vena cava/right atrium. This latter situation represent an impending life-threatening condition that should be surgically treated as soon as possible (Novick et al., 1990).

Atrial thrombus removal may be achieved under normothermic CPB and total liver vascular exclusion. Tolerance of the cirrhotic liver to normothermic ischemia is reported as safe up to 33 minutes (Huguet et al., 1994) which is quite similar to the safety time reported to avoid neurological complications when using HCCA (McCullough et al., 1999).

Extracorporeal circulation with HCCA has been increasingly used to remove atrial thrombi due to retroperitoneal malignancies (Ogata et al., 2006) but its use in cirrhotic patients has been limited, for many years, because of the fear of post-operative liver failure and poor outcome. However, in more recent time, we have assisted to an increase number of reports where HCCA has been used to remove cavo-atrial thrombi due to HCC with no mortality and low morbidity, probably, lower than expected thus, we can assume that, a short lasting HCCA, has to be considered safe in a well compensated liver cirrhosis (Novick et al., 1990; Yogitaet al., 2000; Ohwada et al., 2008; Florman et al., 2009; Leo et al., 2010).

Since the vast majority of patients with HCC and atrial tumour thrombi has large type of HCC, a major liver resection is usually required and the possibility of postoperative liver failure raises accordingly, liver function tests and indocyanin green retention test, should be carried out in these patients in order to demonstrate the presence of a good functional reserve (Kawasaki et al., 1995). Furthermore we think that adding a sequential arterial and portal embolization of a hemi-liver could lessen the possibilities of post-operative liver failure when a major resection is planned (Ogata et al., 2006).

Hypothermic Cardiac Arrest to Remove Right Atrial Thrombi Due to Abdominal Malignancies 153

metastatic fibro-leiomyoma because pulmonary nodules (proliferating leiomycites) are often present. Although this tumour is usually confined to the pelvis, it sometime extends to the right cardiac cavities through the iliac vein and the vena cava (Wakiama et al., 2000) by means of intra-vascular proliferation and growth of fibro-myocites which results in an extension of the fibro-leiomyoma rather than in a thrombus formation conferring a typical consistency and resistance to traction to the endo-vascular proliferation (Galajda et al., 2010). IVL with cardiac extension is an exceedingly rare disease first reported in 1907 in an autopsy series (Durk, 1907). Since then a further 34 cases have been reported involving the right heart (Wakiyama et al., 2000). Patients often have a history of hysterectomy or leiomyoma resection and various lengths of intravascular proliferation up to 29 cm with a diagnosis

Complete surgical removal of the uterus and IVL is the therapy of choice (Topcouglu et al., 2004) and it has been carried out for the first time in 1974 using CPB and HCCA

We came across a similar case (Fig. 2) in whom we performed a total abdominal hysterectomy and IVL removal through a cavotomy and atriotomy under CPB and HCCA (Fig. 3). Due to the typical consistency of IVL and the usual absence of adhesions the complete heart and retrohepatic vena cava cleaning was achieved by simple traction from the infra-renal cavotomy site

Fig. 2. Preoperative ct scan of intra-vascular leiomyomatosis arising from uterine leiomyoma (white arrow) extending to the right iliac vein (blue arrow), infra and retro-hepatic vena cava (green arrow), right atrium (red arrow) and to the rignt ventricle (yellow arrow).

(Fig. 3A) while the iliac portion was stacked and required separate iliac vein incision.

performed up to 18 yrs after the primary surgery ( Galajda et al., 2010).

(Mandelbaum et al., 1974).

Fig. 1. A: Ct scan of a cavo-atrial tumour thrombus (yellow arrow). B and C: Extraction of the thrombus, after a left hemi-hepatectomy, from the sectioned left and middle hepatic vein ostium (white arrow) using a rings clamps to pull and an index finger, inserted from the atriotomy site, to push the thrombus downward. D: Cleaned ostium (red arrow), closed atriotomy (black arrow) and distal middle hepatic vein stump (green arrow) to be reconstructed. Please note the completely bloodless field of the procedure under hypothermic cardiocirculatory arrest.

#### **2.3 Uterus**

The uterus is seldom the site of origin of a cavo-atrial thrombi owing to two distinct tumours: the endometrial stromal sarcoma (ESS) and the "benign" intravenous leiomyomatosis (IVL). This latter situation differs from the other types of neoplastic thrombi because the endovascular material is represented mainly by myofibroblastic cells, growing into the vessel, rather than by a neoplastic thrombotic apposition.

#### **2.3.1 Endometrial stromal sarcoma**

ESS represents only the 0.2% of all uterine malignancies and it has been described to invade the great vessels rarely. In a recent review only 9 patients had a tumour thrombus reaching the right atrium, ventricle or the pulmonary artery that underwent radical resection by means of total abdominal hysterectomy and thrombectomy. Normothermic CPB was reported in 7 out the 9 patients with only 1 (14.3%) complication due to renal acute tubular necrosis. No perioperative deaths were reported (Renzulli et al., 2008).

#### **2.3.2 Intravenous leyomiomatosis**

IVL is an uncommon "benign" tumour of middle age parous women arising from either the uterine venous wall or uterine leiomyoma (Nam et al., 2002). It is known also as benign

Fig. 1. A: Ct scan of a cavo-atrial tumour thrombus (yellow arrow). B and C: Extraction of the thrombus, after a left hemi-hepatectomy, from the sectioned left and middle hepatic vein ostium (white arrow) using a rings clamps to pull and an index finger, inserted from the atriotomy site, to push the thrombus downward. D: Cleaned ostium (red arrow), closed atriotomy (black arrow) and distal middle hepatic vein stump (green arrow) to be reconstructed. Please note the completely bloodless field of the procedure under

The uterus is seldom the site of origin of a cavo-atrial thrombi owing to two distinct tumours: the endometrial stromal sarcoma (ESS) and the "benign" intravenous leiomyomatosis (IVL). This latter situation differs from the other types of neoplastic thrombi because the endovascular material is represented mainly by myofibroblastic cells, growing

ESS represents only the 0.2% of all uterine malignancies and it has been described to invade the great vessels rarely. In a recent review only 9 patients had a tumour thrombus reaching the right atrium, ventricle or the pulmonary artery that underwent radical resection by means of total abdominal hysterectomy and thrombectomy. Normothermic CPB was reported in 7 out the 9 patients with only 1 (14.3%) complication due to renal acute tubular

IVL is an uncommon "benign" tumour of middle age parous women arising from either the uterine venous wall or uterine leiomyoma (Nam et al., 2002). It is known also as benign

into the vessel, rather than by a neoplastic thrombotic apposition.

necrosis. No perioperative deaths were reported (Renzulli et al., 2008).

hypothermic cardiocirculatory arrest.

**2.3.1 Endometrial stromal sarcoma** 

**2.3.2 Intravenous leyomiomatosis** 

**2.3 Uterus** 

metastatic fibro-leiomyoma because pulmonary nodules (proliferating leiomycites) are often present. Although this tumour is usually confined to the pelvis, it sometime extends to the right cardiac cavities through the iliac vein and the vena cava (Wakiama et al., 2000) by means of intra-vascular proliferation and growth of fibro-myocites which results in an extension of the fibro-leiomyoma rather than in a thrombus formation conferring a typical consistency and resistance to traction to the endo-vascular proliferation (Galajda et al., 2010). IVL with cardiac extension is an exceedingly rare disease first reported in 1907 in an autopsy series (Durk, 1907). Since then a further 34 cases have been reported involving the right heart (Wakiyama et al., 2000). Patients often have a history of hysterectomy or leiomyoma resection and various lengths of intravascular proliferation up to 29 cm with a diagnosis performed up to 18 yrs after the primary surgery ( Galajda et al., 2010).

Complete surgical removal of the uterus and IVL is the therapy of choice (Topcouglu et al., 2004) and it has been carried out for the first time in 1974 using CPB and HCCA (Mandelbaum et al., 1974).

We came across a similar case (Fig. 2) in whom we performed a total abdominal hysterectomy and IVL removal through a cavotomy and atriotomy under CPB and HCCA (Fig. 3). Due to the typical consistency of IVL and the usual absence of adhesions the complete heart and retrohepatic vena cava cleaning was achieved by simple traction from the infra-renal cavotomy site (Fig. 3A) while the iliac portion was stacked and required separate iliac vein incision.

Fig. 2. Preoperative ct scan of intra-vascular leiomyomatosis arising from uterine leiomyoma (white arrow) extending to the right iliac vein (blue arrow), infra and retro-hepatic vena cava (green arrow), right atrium (red arrow) and to the rignt ventricle (yellow arrow).

Hypothermic Cardiac Arrest to Remove Right Atrial Thrombi Due to Abdominal Malignancies 155

The problems when dealing with these situations are almost the same of those reported above and at the moment it is impossible to draw any conclusion due the scarcity of the

In the following sections, an overview of the main technical features for the management of anaesthesia, extracorporeal circulation and hypothermic circulatory arrest for cavo-atrial

Any operative procedure involving large vessels entail the risk of developing hemodynamic instability, mainly due to massive bleeding. Moreover, a number of specific conditions have to be taken into consideration, including coexisting multisystemic diseases, nonphysiological conditions associated with CPB and deep HCCA. In light of these concerns, patients undergoing cavo-atrial thrombectomy require extensive monitoring to provide

An intraoperative electrocardiogram, using the five-leads technique, is the standard monitoring procedure always recommended during surgery and anesthesia. In this setting, it is useful for the diagnosis of both myocardial ischemia (DII – V5 identify 90% of ischemic

Intravascular pressure measurements represent the standard technique during this type of surgery. Arterial pressure is usually measured by placing a catheter in a peripheral (radial, ulnar, brachial) or central (femoral) artery. Direct arterial pressure measurement allows monitoring during the pre-CPB stage and nonpulsatile ECC. Moreover, during the post-CPB/HCCA period, patients are usually hemodynamically unstable, and close surveillance of arterial blood pressure, other than with blood-gas analysis, is of primary

A central venous line (internal jugular, subclavian) for pressure monitoring (central venous pressure [CVP]) is influenced by circulating blood volume, venous tone and capacitance, and right ventricular function; therefore, a number of pieces of information can be obtained from the CVP. A central venous catheter (3 lumens, 7-8.5 Fr) can be used for both pressure measurement and inotropic-vasoactive drug administration. Central venous lines can also be used for pulmonary artery catheter (PAC) positioning if the patient's comorbidities suggest the monitoring of pulmonary artery pressures, the measurement of cardiac output and mixed venous oxygen saturation. Central venous cannulation should be performed under ultrasound guidance because up to 30% of patients have some abnormalities of the

Hemodynamic monitoring is completed with transesophageal echocardiography, which has gained widespread use in cardiac operating rooms (where this intervention for cavo-atrial invasion is usually performed) because it provides a great deal of information about the heart's global performance during systole and diastole, the valve function and cardiac volume loads, as well as morphologic details on thrombotic cardiac invasion from the

jugular vein anatomy (Carid et al., 1988; Bevilacqua et al., 2005).

early warning of conditions that may lead to potentially life-threatening states.

cases.

**3. Anesthetic management** 

thrombectomy are summarized.

**3.1 Monitoring and anesthesia** 

**3.1.1 Monitoring** 

importance.

inferior vena cava.

episodes) and dysrhythmias.

Fig. 3. Same Patient of figure 2. Operative view of the removal, through a cavotomy (A) and atriotomy (B), of intravascular leiomyomatosis. Operative specimen removed is shown in C: 1 and 2 are the right atrial extensions, 3 the right ventricular extension, 4 the caval iliac extensions and 5 right ovarian vein involvement.

#### **2.4 Vena cava**

Leiomysarcomas of the inferior vena cava (IVC) are rare malignant tumours originating from the smooth muscle cells of the media layer that typically show three growth patterns: extra-luminal, intra-luminal and both (Mingoli et al., 1996). They may be classified anatomically according to Chiappini et al., 2002, as for renal tumours, into four types: Type I infra-hepatic, type II retro-hepatic, type III supra-hepatic IVC and type IV the thrombus extends in the right atrium.

Patient with type IV was treated and reported by Hassan et al., 2010, using normothermic CPB and Pringle's manoeuvre to remove the atrial thrombus reporting mild hyperbilirubinemia and normal renal function. They discussed the possible advantages of the technique used, compared to HCCA which they thought to be associated with an extended CPB time, increased postoperative bleeding and coagulopathy, and increased neurological risk. On the other hand HCCA provides a bloodless surgical field with reduced risk of spreading/embolization, and fatal haemorrhage. Furthermore the advantages include reduced liver and kidney warm ischemia, reduced risk of incomplete excision, optimal visualization of the IVC/right atrial lumen minimizing the need for a too extensive retro-peritoneal dissection (Chiappini et al., 2002).

#### **2.5 Others**

Scattered case reports of ovarian haemangioma (Tamburino et al., 1992), testicular tumour (Kanda et al., 1991), embrional carcinoma of the testis (Paule et al., 1991), testicular teratoma (Moon et al., 1992) and even pancreatic cancer (Ozben et al., 2007) have been reported so far.

The problems when dealing with these situations are almost the same of those reported above and at the moment it is impossible to draw any conclusion due the scarcity of the cases.
