Choung-Soo Kim, Department of Urology, Asan Medical Center, 388-1 Pungnap 2-dong, Songpa-gu, Seoul 138-736, Korea. e-mail: firstname.lastname@example.org
Study Type – Prognosis (case series)
Level of Evidence 4
What's known on the subject? and What does the study add?
In patients with pRCC, the presence of venous tumour thrombus is known to be a predictor of poorer outcomes. However, a paucity of data is available regarding the prognostic significance of histology in patients with RCC and IVC thrombus.
In our series, we found that patients with type II pRCC had significantly poorer outcomes when compared to those with cRCC. Although the lack of effective treatment for patients with metastatic pRCC may have contributed to these adverse outcomes, type II papillary histology was independent predictor not only of CSS but also of RFS.
• To analyze the prognostic impact of papillary histology on oncological outcomes in patients with renal cell carcinoma (RCC) and inferior vena cava (IVC) thrombus.
PATIENTS AND METHODS
• We reviewed the medical records of 74 patients who underwent radical nephrectomy and IVC thrombectomy between 1990 and 2010 for clear cell or papillary RCC.
• We compared the clinicopathological features and clinical outcomes of 62 patients with clear cell RCC (cRCC) and 12 with papillary RCC (pRCC).
• All cases of pRCC were subdivided into type I or type II.
• The prognostic role of papillary histology on recurrence-free survival (RFS) and cancer-specific survival (CSS) was estimated using Cox's regression models.
• Upon reclassification of the pRCC subtype, all 12 patients with pRCC had type II tumours.
• Patients with type II pRCC were significantly younger (P = 0.028) and were more probably women (P = 0.025) than those with cRCC
• The 2- and 5-year CSS rates were 81.1% and 53.6% in cRCC patients, and 28.1% and 0% in type II pRCC patients, respectively. All eight patients with non-metastatic type II pRCC developed disease recurrence at a median interval of 6 months after surgery, whereas 25 of 44 (56.8%) patients with non-metastatic cRCC experienced such recurrence at a median interval of 10 months after surgery.
• Patients with type II pRCC showed significantly lower CSS (P < 0.001) and RFS (P = 0.002) than those with cRCC.
• On multivariate analysis, type II papillary histology was an independent predictor of CSS (hazard ratio, 3.73; P = 0.003) and RFS (hazard ratio, 3.15; P = 0.015).
• Type II papillary histology appears to be predominant in cases of pRCC with IVC thrombus.
• Patients with type II pRCC who presented with IVC thrombus had significantly worse outcomes than those with cRCC, and histology is an important prognostic factor in patients with RCC and IVC thrombus.
RCC tends to invade the renal venous system and form a tumour thrombus that can extend into the inferior vena cava (IVC). Earlier reviews found that a venous thrombus involving the IVC was present in 4–10% of RCC patients, and, in 1% of cases, the tumour thrombus extended into the right atrium [1,2]. RCCs with IVC thrombus are aggressive tumours that are associated with poor prognosis and higher Fuhrman grade, as well as lymph node involvement or distant metastasis at the time of surgery. Aggressive surgical management has significantly improved the survival of patients with RCC and IVC thrombus and the 5-year survival rate is currently 40–65% in patients without metastasis [3–6].
Several studies have aimed to identify prognostic factors in patients with RCC and IVC thrombus. However, most studies focused on the prognostic significance of the cephalad extent of the tumour thrombus and validation of T3 disease classification [6–10]. Margulis et al.  reported that the presence of venous tumour thrombus was an independent prognostic factor in papillary RCC (pRCC) patients but not in patients with clear cell RCC (cRCC), and Mancilla-Jimenez et al.  reported that venous tumour thrombus at surgery resulted in a drastic decrease in survival in patients with pRCC.
Previous studies examining the prognostic significance of tumour histology did not reach consistent conclusions. Some studies have shown that patients with cRCC had significantly lower survival rates than patients with pRCC [13–15], whereas other studies failed to do so . The general assumption is that pRCC is associated with a favourable outcome in localized disease, whereas metastatic pRCC is characterized by resistance to systemic therapy and poor patient prognosis. However, less is known about the prognostic significance of histological subtype in patients with RCC and IVC thrombus. In the present study, we evaluated the clinicopathological data of patients with RCC and IVC thrombus, and explored the prognostic impact of papillary histology on oncological outcomes in these patients.
MATERIALS AND METHODS
The present study was carried out after receiving institutional review board approval. We retrospectively reviewed the medical records of 81 patients who underwent surgery for RCC and IVC thrombosis at our institute from 1989 to 2010. The histological subtypes were clear cell in 63 patients, papillary in 12 patients, chromophobic in two patients and unclassified in four patients. The six patients with chromophobe and unclassified histology, and one patient with cRCC who died from postoperative pulmonary thromboembolism, were excluded from the present study. Thus, 62 patients with cRCC and 12 patients with pRCC were included in the present study.
Staging evaluation at the time of diagnosis included chest X-ray, CT of the abdomen and pelvis, and a bone scan. Based on signs and symptoms, some patients underwent CT of the chest or brain imaging. Preoperative assessment of the extent of the tumour thrombus was determined either by CT or MRI. The thrombus extent was graded using the Nevus classification: level I (tumour thrombus extending ≤2 cm above the ostium of the renal vein into the IVC); level II (thrombus extending ≥2 cm above the ostium of the renal vein and infrahepatic); level III (thrombus extending into the intrahepatic portion of the vena cava but remaining below the diaphragm); and level IV (thrombus extending into the right atrium) .
The surgical procedure has been described in detail previously . All patients underwent radical nephrectomy, and laparotomy was usually performed employing either an inverted T or a chevron incision, both of which can be extended to the thorax when cardiopulmonary bypass is required. Cardiopulmonary bypass was performed in 11 (14.8%) patients and preoperative embolization of the renal artery was conducted in 34 (45.9%) patients.
All pathology slides were retrospectively reviewed by a single pathologist (Y.M.C.) and pathological staging was performed using the 2010 TNM staging system . Histological subtype (papillary or clear cell) was determined according to the 2004 WHO classification  and grading was performed using the Fuhrman nuclear grading system . All pRCC samples were subdivided into type I and type II pRCC . IVC tumour thrombus was defined as the presence of a pathologically-confirmed gross malignant tumour extending into the IVC.
Follow-up included biannual abdominal CT and annual chest X-ray for the first 2 years, and annual abdominal imaging and chest X-ray thereafter. Disease recurrence was diagnosed by biopsy and/or radiologically, and was correlated with clinical condition. Survival data were obtained from medical records or by contacting the families of patients, or by reviewing death certificates. The duration of follow-up was calculated from the date of surgery to the date of death or last follow-up visit.
The clinicopathological features of patients with pRCC and cRCC were compared, and the treatments chosen for patients with synchronous and asynchronous metastases were evaluated. Qualitative variables were compared using a chi-squared test and quantitative variables were compared using the Mann–Whitney U-test. Recurrence-free survival (RFS) and cancer-specific survival (CSS) were estimated by the Kaplan–Meier method and compared using the log-rank test. RFS was calculated after excluding patients with metastatic disease. Cox's proportional hazard regression models were used to identify the independent parameters associated with RFS and CSS. In multivariate analysis, variables significant (P < 0.2) upon univariate analysis were included. SPSS, version 12.0 (SPSS, Chicago, IL) was used for all statistical analyses. P < 0.05 (two-sided) was considered statistically significant.
A total of 74 patients were included in the present study; their clinicopathological characteristics are shown in Table 1. Of the 74 patients, 62 (83.7%) had cRCC and 12 (16.2%) had pRCC. Upon reclassification of the pRCC subtype, all 12 patients with pRCC had type II tumours. Mean (range; median) age at surgery was 55.1 (29–86; 56) years and 57 (77.0%) patients were men. With respect to pathological stage, three (4.1%) patients presented with pT4 disease and 22 (29.7%) with M1 disease. The extent of tumour thrombus was level III in 17 (23.0%) patients and level IV in seven (9.5%) patients. Patients with type II pRCC were significantly younger than those with cRCC (46.6 vs 56.8 years; P = 0.028) and were more probably women (50.0% vs 17.7%; P = 0.025). There were no significant between-group differences with respect to tumour size, pathological stage, lymph node involvement, the presence of distant metastasis, tumour grade or tumour thrombus level.
Table 1. Clinicopathological characteristics of patients
IVC, inferior vena cava.
Patients, n (%)
Age (years), mean (sd)
Tumour size (cm), mean (sd)
T stage, n (%)
Lymph node involvement, n (%)
Distant metastases, n (%)
Nuclear grade, n (%)
COMPARISON OF CSS RATES IN ALL PATIENTS
The median (range) postoperative follow-up period was 29 (2–179) months. At the time of analysis, 34 (45.9%) patients had died from disease, and the CSS rate in the entire patient cohort was 72.3% at 2 years and 46.7% at 5 years. Of 62 patients with cRCC, 25 (40.3%) died from RCC at a median (range) interval of 29 (2–130) months after surgery, and the 2- and 5-year CSS rates were 81.1% and 53.6%, respectively. Of 12 patients with type II pRCC, nine (75%) died from RCC at a median (range) interval of 13 (3–56) months after surgery, and the 2- and 5-year CSS rates were 28.1% and 0%, respectively. Patients with type II pRCC showed a significantly lower CSS than did those with cRCC (P < 0.001) (Fig. 1A). There were three patients with type II pRCC who were alive at last follow-up (11, 24 and 55 months after surgery). The median (range) follow-up interval for cRCC patients who did not die was 37 (2–179) months.
By univariate analysis, female sex (P = 0.018), the presence of synchronous metastasis (P = 0.007), thrombus level (P = 0.041) and type II papillary histology (P = 0.001) were statistically significant factors influencing CSS. On multivariate analysis, the presence of synchronous metastasis (hazard ratio, HR, 3.59; 95% CI, 1.68–7.69; P = 0.001), the level of tumour thrombus (HR, 2.33; 95% CI, 1.06–5.12; P = 0.034) and type II papillary histology (HR, 3.73; 95% CI, 1.55–8.95; P = 0.003) were independent prognostic factors (Table 2).
COMPARISON OF RFS RATES IN NON-METASTATIC PATIENTS
Of 52 patients who were free of distant metastases at the time of surgery, 33 (63.4%) experienced disease recurrence. The estimated RFS rate in non-metastatic patients was 54.9% at 2 years and 29.7% at 5 years. Of 44 patients with non-metastatic cRCC, 25 (56.8%) developed disease recurrence at a median (range) interval of 10 (1–78) months after surgery. The 2- and 5-year RFS rates in the cRCC group were 63.6% and 36.1%, respectively. However, all eight patients with non-metastatic type II pRCC developed metastases at a median (range) interval of 6 (3–36) months after surgery, and the estimated 2- and 5-year RFS rates were 12.5% and 0%, respectively. By the log-rank test, RFS was significantly lower in the type II pRCC compared to the cRCC group (P = 0.002) (Fig. 1B).
By univariate analysis, female sex (P = 0.011), thrombus level (P = 0.010) and type II papillary histology (P = 0.004) were significantly associated with RFS. On multivariate analysis, thrombus level (HR, 2.31; 95% CI, 1.05–5.07; P = 0.037) and type II papillary histology (HR, 3.15; 95% CI, 1.25–7.95; P = 0.015) were independent risk factors for RFS (Table 3).
TREATMENT OF SYNCHRONOUS AND ASYNCHRONOUS METASTATIC RCC
There were four patients with type II pRCC who had synchronous metastases; two received immunotherapy and the other two received targeted therapy. All eight patients with type II pRCC who were free of distant metastases at the time of surgery experienced disease recurrence. Immunotherapy and targeted therapy were conducted in three and four patients, respectively; only one patient was treated conservatively.
Of 43 patients with either synchronous or asynchronous metastatic cRCC, targeted therapy and immunotherapy were given to 15 and 17 patients, respectively. There were five patients who underwent metastasectomy and six patients were treated conservatively.
The present study examined the outcomes for 74 patients with RCC and IVC thrombus. All 12 patients with pRCC were classified as type II, and patients with type II papillary histology had a poorer prognosis than those with cRCC. The findings remained consistent after adjusting for other prognostic variables, such as age, sex, stage and thrombus level.
Although prognostic factors for patients with RCC and IVC thrombus have been well studied, the prognostic impact of histological subtype in these patients remains unclear. Recently, Ciancio et al.  reported a single-centre experience with 87 patients who underwent surgery for RCC with IVC thrombus. They found that ‘non-clear cell’ histology was an independent prognostic factor for poorer CSS, and emphasized the prognostic significance of histology in patients with RCC and IVC thrombus.
Previous studies suggest that pRCC is associated with a lower risk of renal vein or IVC extension  and that patients with pRCC of pT3 or greater at baseline have an unfavourable prognosis . Another study of 2157 patients, including 245 with pRCC and 1912 with cRCC, found that pRCC patients with a venous tumour thrombus experienced significantly lower CSS rates compared to cRCC patients, and also that the presence of a venous tumour thrombus was an independent prognostic factor for pRCC but not for cRCC . In that study, however, multivariate analysis was not used to evaluate the prognostic value of papillary histology in patients with a venous tumour thrombus. In addition, pRCC was not further classified into type I or II, as was the case in most previous studies.
Histological classification of pRCC into two distinctive subtypes (type I or II) is based on specific cytological and architectural features . Type II pRCC comprises large eosinophilic cells arranged in an irregular or pseudostratified manner, whereas type I consists of small cuboid cells with scant cytoplasm. Previous studies report that type II pRCC is associated with aggressive pathological features and a worse outcome than type I; however, the results of multivariate analyses assessing the prognostic significance of type II disease are inconsistent [25–27]. These inconsistencies may be a result of the relatively few patients enrolled, the infrequency of death related to RCC or the different prognostic variables included in the multivariate analysis models. Moreover, given that pRCC often shows heterogeneous features and a mixed type I and type II phenotype, heterogeneity may also explain the different results .
Many studies have shown that patients with pRCC have better outcomes than those with cRCC [13–15,29]; however, other studies suggest that histological subtype is not an independent prognostic factor [16,30]. It is questionable whether this is a result of not distinguishing between type I and type II pRCC. Waldert et al.  reported that, although histological subtype (pRCC vs cRCC) had no impact on survival, type II pRCC had a significantly worse outcome than type I pRCC despite similar clinical and pathological features. In the present study cohort with IVC thrombus, no patients with pRCC were classified as type I, and only type II pRCC patients were included. Although the worse outcome for pRCC reported in the present study could have resulted from focusing on type II disease, cRCC has not been compared with type II pRCC, especially in advanced cases such as vena cava extension. Despite the few patients with type II pRCC enrolled in the present study, we noted that patients with type II pRCC who presented with IVC thrombus had significantly lower CSS and RFS compared to those with cRCC.
Recent data suggest that pRCC is associated with an unfavourable prognosis when metastasis develops [11,32,33], possibly because pRCC is resistant to immunotherapy. Most patients with metastatic pRCC in previous studies were treated before the era of targeted therapy and only an insignificant number of patients received such therapy. Tyrosine kinase inhibitors and mammalian target of rapamycin inhibitors were shown to be effective not only in patients with clear cell, but also in those with non-clear cell histology [34,35]. In the present study, all 12 patients with type II pRCC had either synchronous or asynchronous metastases, and six received targeted therapy. All three patients who were alive at last follow-up had been treated with targeted therapy. Although the lack of an effective treatment for metastatic non-clear cell RCC before the era of targeted therapy may have contributed to our results, we found that patients with type II pRCC also had a significantly lower RFS than those with cRCC, and that a type II papillary histological subtype was an independent predictor of RFS.
In the present study, we found that the level of the tumour thrombus was also an independent prognostic factor. We divided patients into two groups (levels III–IV vs levels I–II) because the number of patients with level IV thrombus was small. The prognostic significance of tumour thrombus level has proved controversial. In recent study of 1122 patients with RCC and IVC thrombus, which contains the largest cohort of patients with supradiaphragmatic disease, the level of the thrombus was an independent predictor of survival. However, although 36 patients with pRCC were included in the study cohort, the prognostic significance of histological subtype was not evaluated .
To the best of our knowledge, no previous study has analyzed the prognostic significance of histology in patients with RCC and IVC thrombus. Previous studies have simply noted that the presence of a venous tumour thrombus was associated with poorer outcomes in patients with pRCC [11,12]. Recently, Bertini et al.  reported that the presence of a friable thrombus was an independent predictor of CSS in patients with RCC and IVC thrombus. In their study, they did not analyze the histological subtypes of friable thrombus. However, a tumour thrombus classified as friable showed thin papillary features . A tumour thrombus with type II papillary histology may be associated with particular pathognomonic features rendering tumour cells more prone to dissemination, thus exposing patients to a high risk of systemic disease progression. More research should be conducted aiming to better understand the biological behaviour of tumour thrombus as a determinant of survival.
The present study had several limitations. As with all retrospective analyses, an inherent selection bias is present, and the relatively small numbers of patients enrolled in the present study may limit the ability to draw meaningful conclusions with respect to cancer-related outcomes, and may also affect the weight to be attached to the independently obtained prognostic information. Moreover, patients with type I pRCC were not included in the present study; thus, it is unclear whether this finding is also applicable to type I pRCC. Also, large numbers of patients with metastatic RCC did not have access to targeted therapy and the present study did not consider the effect of targeted therapy in the analysis of patient outcomes.
Nevertheless, we consider that the present study is the first to analyze the prognostic significance of papillary histology in patients with RCC and IVC thrombus. It appears that patients with pRCC who presented with IVC thrombus have mostly type II tumours, and also that type II papillary histology is an independent prognostic factor of both CSS and RFS in patients with RCC and IVC thrombus. We recommend the close surveillance of these patients and note that histological subtype is an important prognostic factor in patients with RCC and IVC thrombus.
CONFLICT OF INTEREST
The authors declare that there are no conflicts of interest.