UICC and Japanese stage classifications for carcinoma of the pancreas

Int. J. Pancreatol. �9 Copyright 1991 by The Humana Press Inc. All rights of any nature whatsoever reserved. 0169-4197/91/8(3): 205-214/$2.00

UICC and Japanese Stage Classifications for Carcinoma of the Pancreas

Tsukasa Tsunoda, * K a z u h i d e Ura, T o s h i f u m i Eto , Teiji M a t s u m o t o , and R y o i c h i Tsuchiya

Second Department of Surgery, Nagasaki University School of Medicine, 7-1, Sakamoto, Nagasaki, 852, Japan

Received April 15, 1990; Accepted September I0, 1990

Summary

The stage classification (SC) for carcinoma of the pancreas recommended by UICC (UICC-SC) was compared with that of Japan Pancreas Society (JPN-SC) using 229 patients encountered consecutively at the Second Department of Surgery, Nagasaki University School of Medicine over the past 20 yr. By UICC-SC, 51 ~ of the patients belonged to stage IV and 38~ to Stage III. By JPN-SC, 82% of the patients belonged to stage IV. Curative resection rates in JPN stage II and III were significantly higher than those in UICC-SC by the chi-squared test. In 60 patients undergoing resectional surgery, postoperative cumulative survival (PCS) curves and rates by each staging criterion (tumor size IT], lymph node metastasis [N], distant metastasis [M], serosal invasion IS], retroperitoneal invasion [Rp], and invasion to the portal venous systems IV]) were calculated by the Kaplan-Meier method. Among these prognostic factors, significant differences in the PCS curves were demonstrated only between Rp(-) and Rp( + ), and between V( - ) and V( + ) according to the generalized Wilcoxon's text. In UICC-SC, the underestimation of these factors leads to a tendency to classify the patients in a less advanced stage than in JPN-SC. JPN-SC is more complex than UICC-SC. Continuing efforts are necessary to estab- hsh a more practical, simple, and universal staging system for the disease.

Key Words: Pancreatic ductal carcinoma; prognosis; resectability; UICC stage classification; Japanese stage classification.

INTRODUCTION

The stage of cancer should accurately reflect the prognosis of patients suf- fering f rom the disease, as well as resectability of tumors. UICC has recommended a stage classification of pancreatic cancer based on TNM factors

*Author to whom all correspondence and reprint requests should be addressed.

International Journal of Pancreatology 205 Volume 8, 1991

206 Tsunoda et aL

(UICC stage classification, UICC-SC) (1). Japan Pancreas Society has pro- posed a more complex stage classification plan that includes certain other factors (JPN stage classification, JPN-SC) (2)~ Unfortunately, it is often the case that a patient is classified under different stages in different classification systems, for example, stage II in one system and stage I or stage III in another. This confusion can be a serious obstacle to the international ex- change of scientific ideas on carcinoma of the pancreas. The purpose of this article is to analyze the outcome of pancreatic carcinoma cases in view of UICC-SC and JPN-SC, to clarify the differences, and especially to deter- mine the most influencial factors and the most effective grouping of these factors for staging of the disease~

PATIENTS AND METHODS

A total of 229 patients with pancreatic ductal carcinoma were encountered consecutively at the Second Department of Surgery, Nagasaki University School of Medicine, between October 1969 and September 1989o There were 138 males and 91 females. The ages ranged from 2-87 yr, with a mean of 61.5.

Resectional operations, namely, 17 total pancreatectomies, 34 pancreato- duodenectomies, and 9 distal pancreatectomies, were performed in 60 patients. Palliative by-pass operations, such as choledochojejunostomy and gastroje- junostomy, were done in 95 patients. Exploratory laparotomy was performed in 30 patients. Percutaneous external biliary drainage was established in 15 patients. Twenty-nine patients went without invasive treatment of any kind.

The UICC-SC and JPN-SC used in this study are described in Tables 1 and 2. The disease is divided into four stages in both classifications. The staging factors employed in UICC-SC are primary tumor (T-factor), lymph node metastasis (N-factor), and distant metastasis (M-factor), whereas those in JPN-SC include serosal invasion (S-factor), retroperitoneal invasion (Rp- factor), and invasion to portal vein systems (V-factor). T-factor refers only to tumor size in JPN-SC. The expressions of severity and groupings of each factor differ in the two stage classifications. To avoid confusion, therefore, the UICC stage and equivalent JPN staging factors assigned in this study are shown in Table 3.

Patient distribution, the resectability, and curative resection rates of all patients were reviewed according to each staging system. In addition, stage correlation of patient distribution and postoperative cumulative survival (PCS) curves and rates by each staging factor were analyzed in 60 resected cases for which all staging factors, except tumor size and distant metastasis, had been confirmed microscopically. PCS curves and rates were calculated by the Kaplan-Meier method. The generalized Wilcoxon's test was used to compare the survival curves. Statistical analysis of PCS rates at each year was done by the Z-test. Chi-squared analysis was employed for statistical comparison of the data concerning distribution of patients, resection rates, and curative resection rates according to the two stage classifications. The term "curative resection" was used to mean no residual tumor macro- and microscopically. Cystadenocarcinoma, islet cell tumor, ampullary carcinoma, and carcinoid tumor were excluded from this study.

International Journal of Pancreatology Volume 8, 1991

Staging for Pancreatic Cancer

Table 1 UICC Stage Classification(/)

207

Primary Tumor (T)

TX Primary tumor cannot be assessed

TO No evidence of primary tumor

T1 Tumor limited to the pancreas

Tla Tumor 2 cm or less in greatest dimension

Tlb Tumor more than 2 cm in greatest dimension

T2 Tumor extends directly to the duodenum, bile duct,

or peripancreatic tissues

T3 Tumor extends directly to the stomach, spleen,

colon, or adjacent large vessels

R e ~ o ~ h Nodes (N)

NX Regional lymph nodes cannot be assessed

NO No regional lymph node metastasis

N1 Regional lymph node metastasis

Distant Metastasis (M)

MX Presence of distant metastasis cannot be assessed

MO No distant metastasis

M1 Distant metastasis

STAGE GROUPING

Stage I T1 NO MO

T2 NO MO

Stage ~ T3 NO MO

Stage ~ Any T N1 MO

Stage ~ Any T Any N M1

RESULTS

Patient Distribution, Resectabiiity, and Curative Resection Rates (Table 4)

By UICC-SC, 51070 of the patients belonged to stage IV and 38% to stage III. By JPN-SC, 82~ of the patients belonged to stage IV. Resectional surgery was not done in one patient of UICC stage I because of suspected portal vein invasion, and in three patients of UICC stage II because of definite or severe invasion to surrounding organs. On the other hand, 100% resection rates were achieved in JPN stages I and II. Resection rates in stage III and IV were significantly higher by JPN-SC than by UICC-SC. In JPN stage I, a 100~ curative resection rate was achieved, and the curative resection rates in stages II and III were also significantly higher than those in the UICC stages. There were no curative resections in either UICC or JPN stage IV.


208 Tsunoda et al.

Table 2 JPN Stage Classification (2)

Stage I Tl(0-2cm) No So Rp0 V0

Stage K T2(2ol-4~ N~ $I RPl Vl

Stage ~ T3(4.1-6o0Cm) N2 $2 Rp2 V2

Stage ~ T4(>6~ N3 $3 Rp3 Va

Distant metastasis including hepatic metastasis or

peritoneal dissemination is allocated to stage ~o

T: Size of tumor, N: Lymph node metastasis (N0:No

Lymph node involvement, N~: Involvement of primary

group of lymph nodes situated closely to che tumor.

N2: Involvement of secondary group of lymph nodes

between N~ and N3; N3: Involvement of tertiary

group of lymph nodes regarded as juxta-regional

lymph nodes, S: Serosal invasion, Rp: retro-

peritoneal invasion, V: Invasion to portal venous

systems, So, Rp0, V0: Absence of invasion, Sl, Rpl,

VI: Suspected invasion, Sz, Rp2, V2: Definite in-

Vasion, $3, Rp3, V3: Severe invasion

Table3 Co~elationofUICC Stage andJPN Staging Factors

UICC Stage JPN staging factors assigned to UICC stage

I (TI-2, NO, MO) TI-4, N(-), M(-), S0-e, Rp0-2~ V0-1

(T3, NO, MO) Ti-~, N(-), M(-), S0-3, Rp0-3, V0-3

(Any T, NI, MO) TI-~, NI-3, M(-), So-~, RP0-3, V0-~

(Any T, Any N, M1) T 1 - 4 , N~-a , M(+), S 0 - 3 , RPo-3, V0-3

Table 4 Distribution of the Patients, Resection Rates,

and Curative Resection.Rates According to the Two Different Stage Classifications

Stage I Stage I[ Stage ]K Stage IV Total

Patient UICC

Distribution% JPN.,

Resection UICC

rate,% .... JPN

Curative Resec- UICC

6(12) 4(9) 38(77)n e 51(I04)~. 100(202)

I(2) ....... 5(I0) ........ .13(27) -I 82(163) ~ i 100(202)

92(11/12) 67(6/9) 51(39/77) n ~ 4(4/i04)~ ~ 30(60/202)

100(2/2) I00(i0/i0) 93(215~27) ~ 14(23/163) __ 30(60/202)

82(7/9) 0(0/6)'-]~ 33(13/39)] ~ 0(0/4) lOO(2/2) so(8/ lo) ~ ~ 1 2 / t 5 ) -~ ~ o(oL23)

UICC: UICC stage classification, JPN: JPN stage classification

Parenthesis: ntmaber of patients, ~: P<0.05, ~: P<0o01

Twenty-seven patients with unknown stage were excluded.

37(22/60)

37(22/60)


Staging for Pancreatic Cancer 209

Table 5 Correlation of Patients Distribution

in UICC and JPN Stage Classifications in Resected Cases

JPN I JPN ]I JPN ~I JPN IV Total

UICC I 2 4 4 1 II

UICC K 0 0 3 3 6

UICC I~ 0 6 18 15 39

UICC IV 0 0 0 4 4

Total 2 i0 25 23 60

Stage Correlation of Patient Distribution in Resected Cases

The patient distribution among resected cases in each staging system is shown in Table 5. The stage corresponded in the two systems in 24 of 60 patients (40~ Except for six patients belonging to UICC stage III and JPN stage II, all patients were in the same or a less advanced stage by UICC-SC than by JPN-SC. In the above six patients, N-factor (N1) was responsible for the staging of the patients as UICC stage III.

The PCS Curves and Rates According to Stage in Resected Cases

According to UICC-SC~ there were no significant differences in the PCS curves among stages I, II, III, and IV (Fig. 1). However, significant differences were observed in the 1-yr PCS rates between stage I and II, in the 2-yr PCS rates between stage I and III, and in the 3-yr PCS rates between stage I and III.

By JPN-SC, significant differences were observed in the PCS curves between stages I and III (P< 0.05), I and IV (P< 0.05), and III and IV ( P < 0.05) (Fig. 2)~ Significant differences were also observed in the 1- and 2-yr PCS rates between stages I and II. However, no significant differences were present in the PCS curves or rates between stages II and III, and II and IV.

The PCS Curves and Rates by Each Staging Factor in Resected Cases

T-Factor (Fig. 3a, b) In UICC-SC, significant differences were observed in the PCS curves be-

tween T l a and T2 (P< 0.05), T l a and T3 (P< 0.05), T lb and T2 (P<0.05) , and T lb and T3 (P< 0.05). T-factor in JPN-SC means tumor size alone. There were no significant differences in the PCS curves or rates among JPN T~, Tz, T3, and T4 in this series.

N-Factor (Fig. 4) Although no significant difference was present in the PCS curves between

N( - ) and N( + ), a significant difference was present in the 2-yr PCS rate. By JPN-SC, there was no significant difference in the PCS curves among No, N1, N2, and N3, but a significant difference was observed in the 1-yr PCS rate between No and N2.

International Journal of Pancreatotogy Volume 8, 1991

210 Tsunoda et al.

fgJ

1"3

$3 o)

I00-~

80- i 60+ ~ ~ i + 40

2 0 S t a g e

; Stage I ( n = 11) ........... ~ S t a g e / I ( n = 6)

�9 �9 Stage ill (n = 3 g ) . . . . . . . . . . . . ~ Stage IV (n = 4)

i Stage I

Stage m

o i } ~ ~

Year After Resection

Fig. i. Cumulative survival after resectional surgery according to UICC stage classification. There were significant differences in the l-yr PCS rate between stages I and I!, in the 2-yr PCS rate between stages I and III, and in the 3-yr PCS rate between stages I and III (P<0.05).

z

100

80 ~

60-

40-

20-

: e Stage t n=2 !i ~21 . . . . . . . Stage ]~ n=lO

~ . ',~_. stage I

;~.: ~-~ ; ~ ...........................

stage IV ~ ...... :~ stage m .... J L

1 ~ ~ s


Fig. 2. Cumulative survival after resectional surgery according to the Japanese stage classification. There are significant differences between the curves of stages I and III, I and IV, and III and IV (P< 0.05).

M-Factor There was no significant difference in the PCS curves between M ( - ) and

M ( + ) , which may be because of the small number of M( + ) patients. How- ever, there was a significant difference in the 6-too PCS rate, namely, 71~ in M ( - ) and 2507o in M ( + ) . No M ( + ) patient survived more than 7 mo.

S-Factor (Fig. 5) There was a significant difference in the 1-yr ( P < 0.05) PCS rate between

S ( - ) and S( + ), although no significant difference was observed in the PCS curves. By JPN-SC, there were significant differences in the PCS curves between So and $2 (P<0.05) and in the l-yr PCS rate between So and S~ (P<O.05).

International Journal o f Pancreatology Volume 8, 1991

Staging for Pancreatic Cancer

100

8o 6J n,-

- - 6 0

>

> 4 0 09

20

L~

!-: L

i ,~ T 2

T3 ~-~ L7

i

T l b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - T l a ( n = 2 )

Tla . . . . . . . T1b(n = 3 ) I . . . . . . . T 2 ( n = 2 4 )

............ T3 (n = 3 1 )

211

o - i ~ ....... ~ 4 ........

Year A f t e r Resect ion

Fig. 3a. Cumulative survival after resectional surgery for primary tumor by UICC-SC. There are significant differences between Tla and T2, Tlb and T3, Tlb and T2 and Tlb and T3 (P< 0.05).

lob

8 0 03

n.-

- - 6 0 03 > i>

4o Go

2O

i . ' 7 ~L

" L- ~L__ q T, L L_.__,

: '#]- ...... :L"::~ ........ ~ Tz I -~ ,

T1 (n = 8)

. . . . . . . T2 (n = 25)

. . . . . . T3 (n = 20)

........... T4 (n = 7)

1 2 3 4 5


Fig. 3b. Cumulative survival after resectional surgery according to tumor sizes by JPN-SC. The results are not significant among T~, T2, T3, and T 4 .

100

80 03

r ~

- - 60 o~ >

> 40 03

20

0 T - - 5

,1

= = N { - 1 , N 0 (n = t 8 )

! LT~ I . . . . . . . N, ( n = 2 2 ) i ~ q . . . . . . N z ( n = 15)

L - ' L ~ L . I ' - ] N,- , ,No

! L~ ~ - ~ " 1

N~ L...2m-I-CL--#, N~ { N~+~'-} ~l:; ..= -"

1 2 3

Year A f te r Resect ion

Fig. 4. Cumulative survival after resectional surgery according to N-factoro There is a significant difference in the 2-yr survival rate between N ( - ) and N(+) (P<0.05). A significant difference in the 1-yr survival rate is found between No and N2 (P< 0.05).


212 Tst, t noda et al.

.> >

co

,~176 t 80

6 0 -

40 t 20

0

r ' t ' - ' 7

$1

= ~ S (-~.So(n = 33 ) = = S {+} (n = 27 )

1 . . . . . . . S ~ ( n = 8 ) ::Sa . . . . . . S 2 ( n = 15)

........... S 3 ( R = 4 )

~+1: S ! - l ,So [


Fig. 5. Cumulative survival after resectional surgery according to S-factor. There are significant differences in the 1-yr survival rate between S ( - ) and S( + ), So and Sl, and So and $2 (P<0.05). A significant difference was shown between the curves of So and $2 (P<0.05).

I g

>

>

O9

100 -

8 0

6 0 -

4 0 "

20 "

0

= -" R p ' , - l , R p 0 ( n = 27) = = Rp/+>(n = 33 ) . . . . . . . Rpl (n = 9) . . . . . . Rpe(n = 17) ........... Rp3 (n = 7)

R p ~ i , R p o ]


Fig. 6. Cumulative survival after resectional surgery according to Rp-factor. There are significant differences between the curves of Rp( - ) and Rp( + ), Rpo and Rp2, and Rpo and Rp3 (P<O.05).

r r

2> >

r

100

80

60

40

20

0

, - - - ' - 1

!~ i "7

t 2

-" -- V ~ - ; . V 0 ( n = 2 7 ) = -- V { ~ , ( n = 3 3 ) . . . . . . . V~ (n = 5) . . . . . . V 2 ( n = 15) ........... v 3 (n = 13)

. . . . l ,~ V - ~ . VO


Fig. 7. Cumulative survival after resectional surgery according to V-factor. There are significant differences between the curves of V( - ) and V(+) (P< 0.01), Vo and V3 (P<0.001), and V2 and V3 (P<0.05).


Staging for Pancreatic Cancer 213

Rp-Factor (Fig. 6) There was a significant difference in the PCS curves between R p ( - ) and

Rp( + ) (P< 0.05). Significant differences were also observed in the PCS curves between Rpo and Rp2 (P< 0.05), and Rpo and Rp3 (P< 0.05) by JPN-SC.

V-Factor (Fig. 7) There was a significant difference in the PCS curves between V ( - ) and

V(+ ) (P< 0.01). Significant differences were also observed in the PCS curves between Vo and V3 (P< 0.001), and Vz and V3 (P<0.05) by JPN-SC.

DISCUSSION

In general, UICC-SC tends to classify patients in a less advanced stage than JPN-SC. Prognostic factors employed for staging and stage grouping differ, leading to the complexities of comparison of the two classification systems. It is clear, therefore, that the prognostic value of each staging factor should be evaluated independently from the analysis of survival rates, and that stage grouping of the factors should then be discussed.

With regard to tumor size, no significant difference was observed in the survival rates in this series. However, we have reported a significant difference in survival rates among tumor sizes (3,4). It is considered that tumor size itself should not be excluded from the staging factors. T-factor in UICC- SC does not mean simply tumor size. This will be discussed later.

With regard to N-factor, Mannell et al. (5) reported that regional lymph node metastasis had no significant influence on survival, whereas in our previous report, N-factor had a significant influence on the 1-yr survival rate (3). In the present series, a significant difference was observed in the 2-yr survival rate between N ( - ) and N( + ) patients. N-factor seemed to have some influence on prognosis.

With regard to M-factor, M ( + ) patients survived no longer than 7 too. It is considered, therefore, that M-factor may be an independent prognostic factor.

Unlike JPN-SC, UICC-SC gives no clear descriptions concerning S-, Rp-, and V-factors. However, the definition of T2 in UICC-SC is that a tumor extends directly to the duodenum, bile duct, or peripancreatic tissues, which in this article is judged as corresponding to S0-2, Rp0-2, and V0-~ in JPN-SC. T3 in UICC-SC is similarly judged as corresponding to S~, Rp3, and V2-3 in JPN-SC. There was a significant difference in the 1-yr survival rate between S ( - ) and S(+) . However, significant differences were demonstrated in the PCS curves between Rp( - ) and Rp( + ), and between V( - ) and V( + ) according to the generalized Wilcoxon's test. Patients without invasions to retroperitoneal tissues (Rp-factor) or portal venous systems (V-factor) showed a better prognosis than those with definite or severe Rp and/or V factors.

Among the prognostic factors employed in each stage classification, significant differences were demonstrated in the PCS curves only on the basis of the presence or absence of Rp- and V-factors, even though significant differences were sometimes observed in the PCS rates on the basis of N-, M-, and S-factors. These results indicate that Rp- and V-factors exert a strong influence on the postoperative prognosis, as has been shown in the previous reports


214 Tsunoda et al.

(6, 7). The analyses of resection rates and curative resection rates indicate that they also reflect the resectability of the tumors.

With regard to stage grouping, UICC-SC reveals certain underestimations of the stage, as follows~ T2 patients, i.e., patients with definite involvement of peripancreatic tissues corresponding to RP2 in JPN-SC, are allocated to stage I. T3 patients, who have invasions to the surrounding organs or adjacent large vessels corresponding to Rp~ and Vz-3 in JPN-SC, belong to stage II. By JPN-SC, the patients with these conditions ($3, Rpj, and/or Vj) belong to stage IV. When severe portal vein invasion is present, for example, the tumor is usually unresectable unless the portal vein is resected. $3, Rpj, and V3 factors, therefore, should be grouped in the most advanced stage. The above underestirnations resulted in a lowering of the resection rate and curative resection rate in the earlier stages of UICC-SC.

In UICC-SC, positive regional lymph node metastasis is the most impor- tant factor in the definition of stage III regardless of tumor size, invasions to surrounding organs, or large vessels. Significant differences in the PCS rates between N ( - ) and N( + ) were revealed only at certain times. In addition, it has been reported that long-term survivors were obtained by complete dissec- tion of the regional lymph nodes (4). N-factor is perhaps overestimated in UICC-SC as compared to Rp- and V-factors.

In JPN-SC, the severity of each factor (T, N, S, Rp, and V) was divided into four grades, but significant differences were not always demonstrated in the PCS curves or rates among the grades in each factor. In the future, the grading system could be simplified and stage assignment of gradings of each factor better organized. Moreover, M ( + ) patients with diseases, such as peritoneal dissemination and hepatic metastasis, might be subgrouped under the most advanced stage, for example, stage IV-B. In conclusion, continuing efforts are necessary to establish a more practical, simple, and universal stag~ ing system for carcinoma of the pancreas.

REFERENCES

1 UICC. TNM Classification of Malignant Tumors, 4th ed., Springer-Verlag, 1987. 2 Japan Pancreas Society. General Rules for Surgical and Pathological Studies on Cancer

of the Pancreas, 3rd ed. Kanehara Publishing, Tokyo, 1987 (in Japanese). 3 Tsuchiya R, Oribe T, Noda T. Size of the tumor and other factors influencing prognosis

of carcinoma of the head of the pancreas. Am. J. Gastroenterol. 1985; 80: 459-462. 4 Tsuchiya R, Tomioka T, Izawa K, Noda T, Yamamoto K, Tsunoda T, Harada N,

Yamaguchi T, Yoshino R, Miyamoto T, Eto T. Collective review of small carcinoma of the pancreas. Ann. Surg. 1986; 203: 77-81.

5 Mannell A, van Heerden JA, Weitand LH, Ilstrup DM. Factors influencing survival after resection of ductal adenocarcinoma of the pancreas. Ann. Surg. 1986; 203: 403-407.

6 Tsuchiya R, Harada N, Tsunoda T, Miyamoto T, Ura K. Long-term survivors after operation on carcinoma of the pancreas. Int. J. Pancreatology 1988; 3: 491-496.

7 Manabe T, Baba N, Nonaka A, Asano N, Yamaki K, Shibamoto Y, Takahashi M, Abe M, Tobe T. Combined treatment using radiotherapy for carcinoma of the pancreas in- volving the adjacent vessels. Int. Surg. t988; 73: t53-156.


Documents

UICC and Japanese stage classifications for carcinoma of the pancreas