Long-term survival after radical surgery for renal cell carcinoma with tumour thrombus extension into the right atrium


Pavel Zacek, Department of Cardiac Surgery, Charles University in Prague, Faculty of Medicine and Faculty Hospital Hradec Kralove, 50005 Hradec Kralove, Czech Republic. e-mail: zacek@fnhk.cz


What's known on the subject? and What does the study add?

Surgical treatment of renal cell carcinoma (RCC) with tumour thrombus extending into the right atrium remains, despite its complexity and specific technical aspects, the only radical therapeutic option.

This single-centre study, unique in size for this rare condition, reports early and late results over a period of 18 years. All patients were operated on using a standardised protocol with use of cardiopulmonary bypass and deep hypothermic circulatory arrest. Overall and cancer-specific cumulative survival was better than in other reports.


  • • To evaluate the long-term results of radical surgical management of renal cell carcinoma (RCC) with tumour thrombus extension (TTE) level IV into the right atrium (RCC/TTE IV) in a large single-institution series.


  • • Radical complex urological and cardio-surgical procedure was performed over a period of 18 years (1993–2010) on 21 patients with RCC/TTE IV. A radical nephrectomy was performed followed by sternotomy, institution of cardiopulmonary bypass and extraction of the intracardiac tumour thrombus under direct visual control during deep hypothermic circulatory arrest (DHCA).
  • • Perioperative and postoperative variables, and long-term overall and cancer-specific survival using the Kaplan–Meier method were analysed.


  • • In all patients, precise removal of tumour thrombus was accomplished in a bloodless field during DHCA.
  • • The mean (sd) duration of circulatory arrest was 16 (6) min at a mean hypothermia of 20 (3) °C. In-hospital mortality was 9.5% (two patients).
  • • The median survival (including in-hospital mortality) was 25 months.
  • • In Kaplan–Meier analysis, 2- and 5-year overall cumulative survival rate was 57 (95% confidence interval, CI 36–78)% and 37 (95% CI 15–58)%, respectively.
  • • Cancer-specific cumulative survival was 68 (95% CI 49–89)% at 2 years and 51 (95% CI 28–74)% at 5 years.


  • • Late outcome after radical surgical treatment in patients with RCC and TTE reaching up to the right atrium justifies this extensive procedure.
  • • Cardiopulmonary bypass with DHCA allows safe and precise extirpation of all intracaval and intracardiac tumour mass.

cardiopulmonary bypass


deep hypothermic circulatory arrest


RCC with tumour thrombus extension level IV


tumour thrombus extension.


Radical surgery remains a mainstay of curative treatment for RCC, in the Czech population displaying very high incidence of this disease [1], but may become a great challenge in the presence of advanced i.v. progression of a tumour thrombus. An i.v. tumour thrombus is diagnosed in 4–15% of patients operated for RCC. Clinically, four levels of thrombus extension have been defined by Nesbitt et al.[2]. In the most advanced level, IV, the thrombus extends above the diaphragm into the right atrium.

Urological surgery is performed routinely in patients with level I and II thrombus (renal vein or infra-hepatic extension); in the case of intra- or supra-hepatic thrombus (level III) the surgical procedure becomes highly demanding.

The most challenging circumstance is in patients with RCC and tumour thrombus extension (TEE) level IV (RCC/TTE IV) where surgery has to be performed in collaboration with a urologist and a cardiac surgeon. In these cases, strong controversy still exists concerning the indication, choice of the safest surgical strategy and the overall benefit of surgery considering early and late results. Due to the scarcity of these operations, usually only case reports or small series with short follow-up have been published to date. The present cohort of 21 patients, operated with cardiopulmonary bypass (CPB) and deep hypothermic circulatory arrest (DHCA) over the period of 18 years is, to our best knowledge, the second largest single institutional experience, with the longest follow-up. The accumulated information may, therefore, serve for re-evaluation of the chosen strategy and the outcome of treatment.


Over the period of 18 years (January1993 to December 2010), 1650 patients were operated for RCC at our institution. Nephrectomy was performed in 1124 patients, renal resection in 340 patients, and nephrectomy with extirpation of tumour thrombus in 186 patients (11.3%), respectively.

Among the latter subgroup, 21 patients (1.3%) who had a TTE into the right atrium (Nesbitt IV), are the focus of this paper. There were 12 men with a mean (range) age of 57 (42–74) years and nine women with a mean (range) age of 59 (35–72) years. The patients' demographic data are given in the Table 1.

Table 1. The patients' characteristics and perioperative features
  1. BMI, body mass index.

Patients, n (%)21 (100)
 Men12 (57)
 Women9 (43)
Median (range) age, years60 (35–74)
Median (range) BMI, kg/m223.4 (18.4–33.3)
Diabetes, n (%)2 (10)
Hypertension, n (%)8 (38)
Coronary artery disease, n (%)4 (19)
Mean (sd): 
 Creatinine, mmol./L98 (30)
 Procedural time, min362 (56)
 Hypothermia, °C20.4 (2.8)
 Circulatory arrest, min16 (6)
Median (range): 
 Postoperative blood loss, mL350 (150–3600)
 Duration of intubation, h19 (5–96)

Diagnosis of the tumour, its extent and staging were confirmed by abdominal sonography and CT angiography (Fig. 1A,B), echocardiography (Fig. 1C), MRI, chest X-ray, skeletal scintigraphy, and pulmonary and brain CT. Radical surgery was suggested to the patients only when distant metastases had been excluded.

Figure 1.

Intra-atrial tumour thrombus diagnosed by: A, Contrast coronal reconstruction multi-slice CT: obliteration of the inferior vena cava by tumour thrombus protruding into the right atrium (arrow). B, Contrast CT, axial scan: terminal portion of a large tumour thrombus reaching up into the right atrium (arrow). C, Large tumour thrombus of the right atrium in transthoracic echocardiographic apical four chamber view (arrow).


Radical nephrectomy is performed as a first step. The approach for a left-sided tumour is via a chevron incision, whereas in right-sided RCC a liberal sub-costal incision is chosen. After nephrectomy, the stump of the renal vein is secured with a ligature to prevent tumour thrombus displacement (Fig. 2).

Figure 2.

Scheme of complex urological-cardiosurgical procedure involving radical nephrectomy, isolation of the renal vein and vena cava obturated by the tumour thrombus, and finally, in DHCA, extirpation of the voluminous thrombus from cavotomy and right atriotomy.

Next, a median sternotomy and pericardiotomy is performed. CPB is instituted via cannulation of the ascending aorta (arterial line), and superior vena cava and the right atrium (venous drainage). On CPB, the patient is cooled down to a core temperature of 20 °C (urinary bladder and rectal temperature measurement being more decisive than oesophageal temperature). This is required for patients with obturating thrombus that adheres to the wall of the inferior vena cava without circumfluence and leads to its dilatation. In the presence of clear pre- and perioperative signs of good thrombus circumfluence and non-adherence to the caval wall, the cooling is stopped earlier, as the thrombus extirpation should be easy and longer circulation arrest not required. CPB is then stopped and blood is drained into the venous reservoir. During DHCA the thrombus is extirpated in the bloodless field via cavotomy and right atriotomy (dumb-bell shaped thrombus) or cavotomy only (non-obturating thrombus). Cavotomy starts with the circumcision of the renal vein stump and is prolonged cranially within the accessible extent (Fig. 2). Extirpation of adherent obturating thrombus with extension into the hepatic veins is more difficult and sometimes can only be accomplished in a piece-meal manner. Removal of the thrombus is followed by a meticulous search for any tumour thrombus remnants left in the right atrium or vena cava. The cavotomy is closed with running suture still in DHCA. If the thrombectomy was performed quickly and without complication, the atriotomy is closed also within the period of circulatory arrest. Otherwise, the right atrial venous cannula is inserted into the inferior vena cava, total CPB is reinstituted, and the closure of the right atrium is performed during rewarming. Duration of circulatory arrest can be considered safe for 20–25 min at the core temperature of 20 °C. Should circulation for any reasons fail to be reinstituted within 25 min, brain protection is secured by retrograde cerebral perfusion started via a venous cannula in the superior vena cava. After rewarming, CPB is weaned-off and cannulae are removed. Meticulous control of haemostasis and placement of tubes precedes the closure of sternotomy and laparotomy.

For recorded data, means or medians were calculated according to values distribution. Standard deviation (sd) and range were used for statement of the associated uncertainty. The postoperative cumulative overall and cancer-specific survival was estimated using the Kaplan–Meier method.


Complete removal of the tumour thrombus was accomplished in all patients within a mean (sd, range) period of circulatory arrest of 16 (6, 5–33) min, at a mean core temperature of 20.4 (2.8, 17–28) °C. The mean duration of the complex surgery (skin-to-skin) was 6 h; postoperative blood loss in all but two patients was <850 mL (Table 1).

RCC was right-sided in 11, and left-sided, in 10 cases. According to TNM classification, the patients were in T3c (21 patients), N0 (four.), N1 (three), NX (13), and M0 (21). The mean (sd, range) longitudinal dimension of the extirpated RCC was 97 (35, 50–180) mm. The histological types were: clear cell RCC in 19 cases (Fuhrman nuclear grade II in two, grade III in 16, and grade IV in one), papillary RCC in one case (Fuhrman nuclear grade III), and unclassified RCC in one case (Fuhrman nuclear grade III) (Table 2).

Table 2. Tumour characteristics, survival and cause of death
Patient numberSideDimensions, mmHistologyFuhrman gradeTNM classificationSurvival, monthsCause of death/status
  1. NA, not available.

 1RightNAClear cellIIIT3cNX097Stroke
 2Right180Clear cellIIIT3cNX040Stroke
 3Right75Clear cellIIIT3cNX01Multi-organ failure
 4Left65Clear cellIIIT3cNX025Metastases
 5Left75Clear cellIIIT3cNX018Metastases
 6LeftNAClear cellIIIT3cNX069Metastases
 7Left105Clear cellIIIT3cNO029Metastases
 8Left115Clear cellIIIT3cN1070Metastases
 9Right100Clear cellIIIT3cN106Metastases
10Right50Clear cellIIIT3cNX00Multi-organ failure
11Left100Clear cellIIIT3cNX021Metastases
12Left75Clear cellIIT3cNO03Unknown
13Right150Clear cellIVT3cN1014Metastases
14Right70Unclassified RCCIIIT3cNX07Myocardial infarction
15Left70Clear cellIIIT3cNX05Pancreatitis
16Right130Clear cellIIIT3cNX025Metastases
17Right100Clear cellIIIT3cNX077Alive
18Right140Clear cellIIIT3cNO065Alive
19Left140Papillary RCCIIIT3cNO055Alive
20Left65Clear cellIIT3cNX042Alive
21Right90Clear cellIIIT3cN0039Alive

In-hospital mortality was 9.5% (two patients). The two patients (aged 61 and 64 years) died from multi-organ failure, preceded by complicated postoperative course, on postoperative day 10 and 32, respectively. In all, 19 patients entered the long-term follow-up of a median (range) duration of 29 (3–97) months. Malignant generalization was the cause of death in nine patients after a median (range) of 31 (6–70) months. Five patients died from other causes (pancreatitis [one], myocardial infarction [one], stroke [two], unknown [one]) within 3–97 months. Five patients are still alive at 39, 42, 55, 65, and 77 months postoperatively (median 55 months) (Table 2). The median survival after the operation (including operative mortality) was 25 months. Cumulative survival of the whole cohort is shown in a Kaplan–Meier plot (Fig. 3) with 2- and 5-year survival rate of 57 (95% CI 36–78)%, and 37 (95% CI 15–58)%. Cancer-specific cumulative survival was 68 (95% CI 49–89)% at 2 years and 51 (95% CI 28–74)% at 5 years.

Figure 3.

Kaplan–Meier overall and cancer-specific cumulative survival estimates according to the study group (including operative mortality).


Radical and extensive surgery, nephrectomy with extraction of the tumour thrombus, seems to be the only chance for survival for patients with RCC and TTE into the right atrium (level IV). Substantial controversy exists about the advantages and disadvantages of various strategies in this demanding procedure. In the present series of 21 patients, we performed the operation with use of CPB and DHCA. This technique enables safe and exact extraction of the tumour thrombus in a bloodless operative field and eliminates the risk of leaving residual tumour fragments adherent to the wall of the inferior vena cava or even free fragments inside the right atrium. The same conclusion is supported by published case reports [3–6] and series of four to 21 patients operated with use of CPB and DHCA [7–16].

In an effort to avoid DHCA, the procedure can be performed on normothermic CPB and the use of the Pringle manoeuvre (series of 4–13 patients) [17–21]. Alternatively, CPB with mild hypothermia and intermittent supra-celiac abdominal aortic occlusion may also be used [22]. A risky and hazardous technique, without CPB and sternotomy, was described in a series of 12 patients with supra-diaphragmatic thrombus in whom the tumour thrombus was blindly ‘milked’ downward from the right atrium and the intra-pericardial inferior vena cava [23]. The ‘blind milking’ manoeuvre may result in embolic complications [24].

The benefit of using DHCA has been confirmed by a multi-institutional retrospective study in patients with RCC/TTE IV, operated on between 1983 and 2007. Perioperative mortality was significantly higher (37.5%) in 36 patients operated with no DHCA, compared with 8.3% mortality in 24 patients operated with the use of DHCA (P= 0.006) [25]. Granberg et al.[8] reported 10.7% 30-day mortality in 28 patients operated using DHCA between 1970 and 2005 (21 patients with RCC/TTE IV together with seven with RCC/TTE II-III). These results corroborate the 9.5% in-hospital mortality of the present single-institution series of 21 patients, operated between 1993 and 2010. However, an overall comparison with reported percentage mortality rates is confusing and methodically incorrect, as most papers include only a few patients with 0–2 early deaths observed. Currently, better surgical results than those published can be expected because of the enormous improvement in perioperative and postoperative care. A risk of 5–10% can currently be hypothesised for an operation with the use of DHCA in patients with RCC/TTE IV.

Long-term results are of equal importance and also justify the correctness of such a complex and demanding surgical procedure. Analysis of the long-term results based on published single-institution series remains disputable because of the few patients, heterogeneous methods, and short follow-ups. In 2009 and 2011, two large multi-institutional studies were published focusing on the long-term results of operations for RCC with TTE into the venous system [26,27].

The first was a retrospective study including 1192 patients operated for RCC with tumour thrombus between 1982 and 2003 at 13 European institutions [27]. The median survival was 52 months for 933 patients, with a renal vein tumour thrombus, 25.8 months for 196 patients with a subdiaphragmatic intracaval tumour thrombus and only 18 months for 63 patients with supradiaphragmatic tumour thrombus. Independent prognostic factors in multivariate analysis were tumour size, perinephric fat invasion, lymph node invasion, tumour thrombus level and distant metastasis [27]. There is no improvement in postoperative survival of patients with multiple distant metastases when compared with the natural history of untreated RCC [21].

The second was a multi-institutional study that collected retrospective data from 11 USA and European academic institutions between 1970 and 2006 [26]. Of 1215 patients who underwent radical nephrectomy and complete tumour thrombectomy, 77 patients had tumour thrombus extending into the right atrium; with median survival time of 12 months and 22% 5-year cumulative survival. In a multivariate analysis, tumour size, Fuhrman grade, nodal metastasis and tumour thrombus level correlated independently with the survival [26].

In our single-institution series of 21 patients operated for RCC/TTE IV, the median survival time was 25 months and the 2- and 5-year cumulative survival rates was 57% and 37%, respectively (including in-hospital mortality). The cancer-specific cumulative survival was 68% at 2 years and 51% at 5 years. Currently, five patients are alive for >39 months (range, 39–77). The longest documented survival in the present series was 97 months. These data compare favourably with the results from multi-institutional studies, and support an active surgical approach in the treatment of this advanced malignant disease. The present results also support the chosen technique of CPB and DHCA which, on one hand, adds to the complexity of the procedure but, on the other hand, enables safe and precise surgical management of the tumour extirpation in a bloodless field, as well as control over potential bleeding.

Despite the advantage of a uniform surgical protocol in the reported series the main methodological limitation of the study remains to be the small size of the cohort treated over a large timespan reflecting the low prevalence of this diagnosis. Over the period of 18 years, major improvements in diagnostics and perioperative care can be seen, which beneficially influence both early and late results of RCC/TTE IV surgery. In recent years, advances have also been made in the adjuvant therapy of metastatic renal cancer. Solitary metastasis should be treated by surgery, while in multiple metastases a biological therapy is preferred. The tyrosine kinase inhibitors are recommended as a first-line therapy for most patients [28]. Efficacy of this therapy can be shown in the present patient number 17, who had developed multiple lung metastases 4 years after surgery. Therapy with sorafenib was successful and at present the patient has been in complete remission for 2 years (confirmed by repeated positron-emission tomography/CT).

In conclusion, radical nephrectomy with extirpation of the tumour thrombus in DHCA may be considered the safest surgical strategy for patients with RCC and TTE into the right atrium. The bloodless field obtained with this technique gives the surgeon optimal control over complete thrombus extirpation. The long-term results achieved in the present series justify radical surgical treatment of patients with RCC and TTE into the right atrium.


None declared.