peritoneal cytology does not increase

8/4/2019 Peritoneal Cytology Does Not Increase

1/6

Peritoneal Cytology Does Not Increasethe Prognostic Information Provided by TNMin Gastric Cancer

G. de Manzoni, MD,1 G. Verlato, MD,2 A. Di Leo, MD,1 A. Tomezzoli, MD,3

C. Pedrazzani, MD,1 F. Pasini, MD,4 Q. Piubello, MD,3 C. Cordiano, MD1

1First Division of General Surgery, University of Verona, 37126 Verona, Italy2Unit of Epidemiology and Medical Statistics, University of Verona, 37126 Verona, Italy3

Division of Pathology, Borgo Trento Hospital, Verona, 37126 Verona, Italy4Division of Medical Oncology, University of Verona, 37126 Verona, Italy

Abstract

Background: This study aimed at verifying whether peritoneal cytology could improve the prog-

nostic information provided by TNM staging in gastric cancer patients.

Method: The presence of free peritoneal tumor cells was investigated in 168 patients who

underwent curative resection for gastric cancer from January 1992 to July 2002 in Verona, Italy.

The influence of peritoneal cytology on survival was evaluated by a Cox regression model, con-

trolling for potential confounders.

Results: Twenty-three patients (14%) had positive peritoneal cytology. Patients with positive la-vage were more likely to present serosal infiltration (100 vs. 46%) and nodal metastases (91 vs.

67%; P < 0.001). Positive lavage was associated with a very poor prognosis: 3-year survival was

only 9% (95% CI 227%) when peritoneal cancer cells had been detected, whereas survival

reached 50% (95% CI 4259%) in patients with a negative cytology. In multivariate survival

analysis, peritoneal cytology was an independent predictor of mortality when controlling for sex,

age, site, histology, and nodal metastases, but not when adjusting also for depth of tumor invasion

(RR of positive versus negative = 1.2, 95% CI 0.72.0). Similarly, the influence of peritoneal

cytology on survival was no longer significant when univariate analysis was restricted to T3/T4

patients (RR = 1.5, 0.92.5).

Conclusions: Positive peritoneal cytology was a marker of poor prognosis in gastric cancer pa-

tients. Nevertheless, peritoneal lavage did not increase the prognostic information already pro-

vided by the TNM staging system in this Italian series.

P eritoneal metastases are the most frequent type ofrecurrence in patients with gastric cancer, evenafter potentially curative resection, especially for diffuse

neoplasms with serosal invasion.1 The causal relation-

ship between infiltration of serosal layer, dissemination of

malignant cells, and their peritoneal implantation has

been well described, and free cancer cells identified by

cytology of intraoperative peritoneal lavage could be a

strong indicator of future peritoneal metastases.24

However, many patients with negative cytology will de-

velop peritoneal metastases, and some authors deny that

cytological findings in peritoneal lavage can be useful in

predicting peritoneal recurrence.5

This study aimed at verifying whether peritoneal cytol-

ogy could improve the prognostic information provided byCorrespondence to: G. de Manzoni, MD, e-mail: nadaffona@inter-

free.it; [email protected]

2006 by the Societe Internationale de Chirurgie World J Surg (2006) 30: 579584

Published Online: 10 March 2006 DOI: 10.1007/s00268-005-7901-2


2/6

the TNM staging in gastric cancer patients, with respect

to peritoneal recurrence and disease-related survival.

METHODS

Patients

Intraoperative cytological examination was carried out

before dealing with tumors in 196 patients who underwent

gastrectomy for gastric cancer in the First Department of

General Surgery, University of Verona, from 1 January

1992 to July 2002. These patients represented 55% of all

the patients operated on for the disease in that period

(n = 359). After excluding 8 patients whose lavage was

classified as inadequate, 16 patients who underwent R2

resections or presented microscopic residual disease at

the resection margins, and 4 patients who died of post-

operative complications, 168 patients were recruited for

the study. In all these patients at least D2 lymphaden-

ectomy had been performed.

Tumors were classified according to the 1997 patho-

logic classification (pTNM) of the International Union

Against Cancer,6 and histological classification followed

the criteria of Lauren.7

Peritoneal Lavage Cytology

Immediately after the abdominal cavity was opened,

200 ml of physiological saline was introduced in the in-

framesocolic region, and after manually stimulated dis-

persion of the saline, a sample of about 50 ml was

recovered from the Douglas pouch.

In all the 168 cases, the collected samples underwent

centrifugation (2500 rpm per 5 minutes), and four slides

of samples were prepared for extemporary cytological

examination, always carried out by the same expert

cytologist (A.T.) through rapid fixing with alcohol 95% and

rapid coloration with haematoxylin-eosin. In cases of

negative extemporary cytological examination, two moreslides were prepared with the same procedure regarding

centrifugation and cytocentrifugation, but using the Pa-

panicolau stain for the final examination. Each slide was

classified as inadequate, negative, or positive. Slide

preparations were considered inadequate if there was

insufficient cellular material for a diagnosis, if blood col-

oration was so heavy that it prevented cellular examina-

tion, or if there was a fault in fixation or coloration.

The cytological samples already prepared with stan-

dard coloration were re-examined with immunocyto-

chemistry by another expert cytologist (Q.P.), who was

unaware of the previous cytological response. Two dif-

ferent antibodies were used: one for epithelial cells (Ber-

EP4 1:50) and one for mesothelial cells (anti-Calretinin

1:300) as negative control.

Follow-up

All patients, after being discharged from hospital, were

scheduled for outpatient follow-up examinations every 6

months. None of the patients was treated with perioper-

ative or postoperative chemotherapy. Follow-up was

completed in June 2005. Patients classified as disease-

free at the last follow-up examination presented com-

pletely normal results from all studies of images as well

as normal tumor marker levels. The median follow-up for

surviving patients was 64 (range: 35159) months.

Statistical Analysis

Significance of differences between patients with and

without free cancer cells after the cytological examination

of the peritoneal lavage were assessed by a t-test for

continuous variables (age), by the v2 test or Fisher exact

test for nominal variables (sex, site, histology, lympha-

denectomy, relapse), and by thev2 test for trend for ordinal

variables (depth of tumor invasion, node metastasis).

Only deaths from gastric cancer were considered as

events in survival analysis, whereas deaths from other

causes were considered as censored observations at thetime of occurrence. Survival curves were computed

according to the Kaplan-Meier method and compared by

the log-rank test. The Cox regression model was used to

evaluate the prognostic significance of positive lavage,

controlling for sex, age, site, histology, depth of tumor

invasion (T), node metastasis (N). The assumptions of

proportional hazard over time made in the Cox model

were met for all the variables tested according to graph-

ical methods.8

A multivariate logistic regression model was used to

evaluate the influence of peritoneal cytology on peritoneal

recurrence, controlling for depth of tumor invasion andnodal metastases. Few explanatory variables were in-

cluded in the logistic models, as the number of events

(peritoneal recurrence) was rather low (n = 33).

RESULTS

In 23 of the 168 cases analyzed (14%), free cancer

cells were detected: 20 positive cases were identified by

immediate cytological examination and 3 additional cases

were detected by immunocytochemistry. Patients with

580 De Manzoni et al.: Peritoneal Lavage in Gastric Cancer


3/6

positive lavage were more likely to present serosal infil-

tration (100 vs. 46%; P < 0.001) and nodal metastases

(91 vs. 67%; P < 0.001; Table 1).

During the follow-up 21 patients with positive lavage

and 73 patients with negative lavage died from gastric

cancer. Among the latter group, 13 patients died from

other causes. Disease-related survival after 3 years was

50% (95% C.I. 42%59%) in patients with negative lavage

and 9% (95% C.I. 2%27%) in patients with positive la-

vage (P < 0.001; Fig. 1).Mortality risk increased more than threefold in the

presence of peritoneal free cancer cells both in univariate

survival analysis (RR of positive versus negative free

cancer cells = 3.3, 95% CI 2.05.3; P < 0.001) and after

controlling for sex, age, histology, and site of the tumor

(RR = 3.2, 95%CI 1.95.4; P< 0.001). The significance of

this novel prognostic factor could also be appreciated after

controlling for nodal involvement (RR = 2.0, 95% CI 1.2

3.4; P = 0.014), but it disappeared after adjusting for

depth of tumor invasion (RR = 1.2, 95% CI 0.72.0;

P = 0.519; Table 2). Survival analysis was repeated on

patients with serosal invasion (n = 90), who comprised all

cases with positive cytology and most of the disease-re-

lated deaths (75 of 94 patients). In this subgroup perito-

neal lavage lost its statistical significance in the univariate

analysis (RR = 1.5, 95% CI 0.92.5; P = 0.119).

Disease recurrence was observed in 21 patients (91%)

with positive immunocytology and in 74 patients (51%) with

negative lavage(P< 0.001). Relapse during the firstyearof

follow-up was observed in most of the patients with positive

immunocytology (17/23 = 74%) but only in one third of

those with negative lavage (50/145 = 34.5%). In the first

group 16 patients (70%) showed a diffused peritonealrelapse as compared to 19 (13%) in the second group

(P < 0.001). In turn, patients with negative lavage pre-

sented a slightly higher proportion of recurrence at gastric

bed level (n = 27, 19%) and in the liver (n = 19, 13%) with

respect to patients with positive lavage (n = 4, 17% at

gastric bed level and n = 2, 9% at both the hepatic and

peritoneal levels). The sensitivity of peritoneal lavage in

predicting peritoneal recurrence was rather low (16/

35 = 46%), whereas the specificity was much higher (126/

133 = 95%); the positive and negative predictive values

were, respectively, 70% (16/23) and 87% (126/145).

0

10

20

30

40

50

60

70

80

90

100

0 12 24 36 48 60

time (months)

survival(%)

Lavage

Neg. 145 114 77 68 53

P


4/6

As shown in Table 3, peritoneal recurrence was mostly

confined to patients with T3-T4 disease, with positivelavage approximately doubling the risk of recurrence. In

the logistic regression model, peritoneal immunocytology

emerged as the most important predictor of peritoneal

recurrence (Table 4). Peritoneal immunocytology retained

its prognostic value even in the subsample of patients with

serosal invasion (OR of positive versus negative free

cancer cells = 6.4, 95% CI 2.218.8; P< 0.001).

DISCUSSION

The most important findings of the present investigation

are the following:

Positive immunocytology of the peritoneal lavage fluid

was confined to patients with gastric cancer invading

the serosa (pT3/pT4).

Positive peritoneal lavage was the most important

predictor of peritoneal recurrence of thetumor.However,

it is questionable whether this finding represents a real

predictive achievement or a kind of quality control.

Patients with positive peritoneal lavage had a very poor

prognosis. Their cumulative survival dropped to 48%

after 1 year, to 13% after 2 years, and to 9% after 3

years, whereas in patients with negative immunocytol-

ogy, survival was 80%, 55%, and 50%, respectively.

In multivariate survival analysis, peritoneal immunocy-

tology was an independent predictor of mortality when

controlling for sex, age, site, histology, and nodal

metastases, but not when adjusting also for depth of

tumor invasion. Similarly, when considering only the

subgroup of patients with serosal invasion, the influ-

ence of peritoneal cytology on survival was no longer

significant.

The poor prognosis of patients suffering from advanced

gastric adenocarcinoma who undergo curative surgical

operation is still an unsolved problem both in Eastern and

Western countries, because of the high incidence of re-

lapse, especially peritoneal recurrence.1,912 A great

number of studies24,1319 maintain that cytological

examination of peritoneal lavage, when performed

immediately after laparotomy and before every surgical

maneuver, is useful in individuating high-risk patients. In

our survey a positive immunocytology was associated

with a low long-term survival in univariate analysis

Table 2.

Relative risks (with 95% confidence intervals in parentheses) of

death from gastric cancer in the 168 patients who underwent

curative resection (R0). Relative risks and significance of dif-

ferences (Likelihood Ratio Test) were derived from Cox

regression model, controlling for all other variables. Calculation

of the relative risk for age was based on an increase in thevalues of 1 SD

Variables

Relative risk Adjusted

for all other variables P Value

Sex (women versus men) 1.0 (0.71.7) 0.862

Age (SD = 11.8 years) 1.3 (1.01.6) 0.024

Site

Middle third versus

upper third

0.9 (0.61.6) = 0.381

Lower third versus

upper third

0.7 (0.41.2)

Histology

Diffuse versus intestinal 1.0 (0.61.6) 0.906

Nodal involvement

pN1 versus pN0 3.8 (1.68.6)

pN2 versus pN0 4.9 (2.111.5)


5/6

(Fig. 1) and relapsing patients with positive immunocy-

tology showed a progression of illness in most cases, at

the peritoneal level. This examination, even if it is char-

acterized by a high specificity in predicting peritoneal

relapse, had a low sensitivity (46%), in agreement with a

large Japanese study.4

Therefore the technique is oftenused in various centers with some variations aiming at

improving its sensitivity, as immunocytochemistry,20,21 or

together with further analysis for indirect research of

hidden metastasis in the peritoneal lavage fluid, like

carcinoembryonic antigen (CEA) dosage2224 and mRNA

of trypsinogen25 or CEA.26

In the current literature the prevalence of positive

cytology in the peritoneal lavage fluid ranges from 5% to

55%.24,1321 These large fluctuations in positive cyto-

logical diagnoses largely reflect methodological differ-

ences in series recruitment, as some studies considered

only patients with R0 resection whereas others focused

on patients with macroscopic serosal invasion. Some

investigations also included patients with peritoneal

carcinosis and/or ascites, whereas others included early

gastric cancer. Indeed the fluctuations in the percentage

of positive cytology are somewhat blunted (from 12% to

34%) if one takes into account only patients with serosal

invasion.4,14,1618,21 In the present series the prevalence

of positive immunocytology was in the middle of the range

reported in the scientific literature, being 13.7% in the

overall series (n = 168) and 25.5% in patients with

serosal involvement (n = 90). When considering alsopatients with residual tumor or those who died in the

postoperative period, the prevalence of positive immun-

ocytology rose to 17.6% (33/188).

A statistically significant correlation was found between

positive lavage and advanced stage (P< 0.001) and nodal

involvement (P < 0.001), as already reported by some

authors.4,11,15,17,19,21 The most important observation

was that no patients with T1 or T2 tumors had positive

washings, confirming previous studies.3,18,19 However,

one should keep in mind that T1 and T2 subgroups had a

limited size (n = 26 and n = 52, respectively). Positivecytology was always associated with neoplasms infiltrat-

ing the serosa (T3) or adjacent organs (T4), and this

finding seems to attribute once again a crucial role to

infiltration and crossing of gastric serosa as a prerequisite

for the transcelomatic metastasising of gastric carcinoma.

In the present investigation, peritoneal immunocytolo-

gy was a strong predictor of mortality in univariate sur-

vival analysis, in agreement with the current

literature.14,17 However, in multivariate survival analysis

this variable did not emerge as an independent prog-

nostic factor, in contrast with other studies. It should be

pointed out, however, that some of these studies4 in-

cluded among the patients with positive cytology some

patients who also had peritoneal dissemination, whose

prognosis is very poor.

In conclusion, peritoneal lavage is a marker of tumor

progressioni.e., serosal invasionand a strong pre-dictor of mortality in gastric cancer. However, its prog-

nostic significance is already provided by the TNM staging

system in this Italian series. Hence, the insertion of this

novel marker (M1cy+) in the TNM UICC staging, proposed

by some authors,27,28 seems to be premature. Neverthe-

less, as positive immunocytology predicts to a certain

extent peritoneal relapse of the tumor, in spite of its low

sensitivity, this diagnostic technique may be useful in

identifying patients who could benefit from intraperitoneal

treatments, such as peritoneal chemohyperthermia.29,30

ACKNOWLEDGMENTS

This research was supported by a grant from the Italian

Association for Cancer Research (AIRC), regione Veneto.

REFERENCES

1. Marrelli D, Roviello F, de Manzoni G, et al. for the Italian

Research Group for Gastric Cancer (IRGGC). Different

patterns of recurrence in gastric cancer depending onLauren histological type: a longitudinal study. World J Surg

2002;26:11601165.

2. Nakajima T, Harashima S, Hirata M, et al. Prognostic and

therapeutic values of peritoneal cytology in gastric cancer.

Acta Cytol 1978;22:225229.

3. Iitsuka Y, Kaneshima S, Tanida O, et al. Intraperitoneal free

cancer cells and their viability in gastric cancer. Cancer

1979;44:14761480.

4. Bando E, Yonemura Y, Takeshita Y, et al. Intraoperative

lavage for cytological examination in 1,297 patients with

gastric carcinoma. Am J Surg 1999;178:256262.

5. Abe S, Yoshimura H, Tabara H, et al. Curative resection of

gastric cancer: limitation of peritoneal lavage cytology in

predicting the outcome. J Surg Oncol 1995;59:226229.

6. Hermanek P, Sobin LJ. UICC TNM Classification of Malig-

nant Tumors, 4th ed., 2nd rev, Berlin, Springer-Verlag, 1992.

7. Lauren P. The two histological main types of gastric carci-

noma: diffuse and so called intestinal-type carcinoma. Acta

Pathol Microbiol Scand 1965;64:3149.

8. Marubini E, Valsecchi MG. Analysing Survival Data from

Clinical Trials and Observational Studies, Chichester, UK,

John Wiley & Sons, 1995.

9. Ohno S, Maehara Y, Ohiwa H, et al. Peritoneal dissemi-

nation after a curative gastrectomy in patients with undif-

De Manzoni et al.: Peritoneal Lavage in Gastric Cancer 583


6/6

ferentiated adenocarcinoma of the stomach. Semin Surg

Oncol 1994;10:117120.

10. Kaibara N, Iitsuka Y, Kimura A, et al. Relationship between

area of serosal invasion and prognosis in patients with

gastric carcinoma. Cancer 1987;60:136139.

11. Maehara Y, Kabashima A, Tokunaga E, et al. Recurrences

and tumor growth potential and local immune response in

gastric node-negative advanced gastric cancer. Oncology

1999;56:322327.

12. Haraguchi M, Watanabe A, Kakeji Y, et al. Prognostic sig-

nificance of serosal invasion in carcinoma of the stomach.

Surg Gynecol Obstet 1991;172:2932.

13. Boku Y, Nakane Y, Minoura T, et al. Prognostic significance

of serosal invasion and free intraperitoneal cancer cells in

gastric cancer. Br J Surg 1990;77:436439.

14. Bonenkamp JJ, Songun I, Hermans J, et al. Prognostic

value of positive cytology findings from abdominal washings

in patients with gastric cancer. Br J Surg 1996;83:672674.

15. Schott A, Vogel I, Krueger U, et al. Isolated tumor cells arefrequently detectable in the peritoneal cavity of gastric and

colorectal cancer patients and serve as a new prognostic

marker. Ann Surg 1998;3:372379.

16. Suzuki T, Ochiai T, Hayashi H, et al. Importance of positive

peritoneal lavage cytology findings in the stage grouping of

gastric cancer. Jpn J Surg 1999;29:111115.

17. Hayes N, Wayman J, Wadehra V, et al. Peritoneal cytology

in the surgical evaluation of gastric carcinoma. Br J Cancer

1999;79:520524.

18. Burke EC, Karpeh MS, Conlon KC, et al. Peritoneal lavage

cytology in gastric cancer: an independent predictor of

outcome. Ann Surg Oncol 1998;5:411415.19. Boku Y, Nakane Y, Minoura T, et al. Prognostic significance

of serosal invasion and free intraperitoneal cancer cells in

gastric cancer. Br J Surg 1990;77:436439.

20. Benevolo M, Mottolese M, Cosimelli M, et al. Diagnostic

and prognostic value of peritoneal immunocytology in gas-

tric cancer. J Clin Oncol 1998;16:34063411.

21. Nekarda H, Geb C, Stark M, et al. Immunocytochemically

detected free peritoneal tumour cells (FPTC) are a strong

prognostic factor in gastric carcinoma. Br J Cancer

1999;79:611619.

22. Asao T, Fukuda T, Yazawa S, et al. Carcinoembryonic

antigen levels in peritoneal washing can predict peritoneal

recurrence after curative resection of gastric cancer. Can-

cer 1991;68:4447.

23. Nishiyama M, Takashima I, Tanaka T, et al. Carcinoem-

bryonic antigen levels in peritoneal cavity: useful guide to

peritoneal recurrence and prognosis for gastric cancer.

World J Surg 1995;19:133137.

24. Bold RJ, Ota DM, Ajani JA, et al. Peritoneal and serum

tumour markers predict recurrence and survival of patients

with resectable gastric cancer. Gastric Cancer 1999;2:17.

25. Fujimura T, Ohta T, Kitagawa H, et al. Trypsinogen

expression and early detection for peritoneal dissemination

in gastric cancer. J Surg Oncol 1998;69:7175.

26. Kodera Y, Nakanishi H, Yamamura Y, et al. Prognosticvalue and clinical implications of disseminated cancer cells

in the peritoneal cavity detected by reverse transcriptase-

polymerase chain reaction and cytology. Int J Cancer

1998;79:429433.

27. Hermanek P, Hutter RVP, Sobin LH, et al. Classification of

the isolated tumor cells and micrometastasis. Cancer

1999;86:26682673.

28. Wittekind Ch, Henson DE, Hutter RVP, et al. TNM Sup-

plement, 2nd Edition, Wiley-Liss; New York, 2001.

29. Fujimoto S, Takahashi M, Mutou T, et al. Improved mortality

rate of gastric carcinoma patients with peritoneal carcino-

matosis treated with intraperitoneal hyperthermic chemop-erfusion combined with surgery. Cancer 1997;79:884891.

30. Hirose K, Katayama K, Iida A, et al. Efficacy of continous

hyperthermic peritoneal perfusion for the prophylaxis and

treatment of peritoneal metastasis of advanced gastric

cancer: evaluation by multivariate regression analysis.

Oncology 1999;57:106114.

584 De Manzoni et al.: Peritoneal Lavage in Gastric Cancer

peritoneal cytology does not increase

Documents