Total Gastrectomy

ANNALS OF SURGERY Vol. 224, No. 1, 37-42 C 1996 Lippincott-Raven Publishers
Total Gastrectomy Updated Operative Mortality and Long-Term Survival with Particular Reference to Patients Older than 70 Years of Age
Reinhard Bittner, M.D.,* Michael Butters, M.D.,* Martin Ulrich, M.D.,* Stefan Uppenbrink, M.D.,* and Hans G. Beger, M.D.t
From the Department of General Surgery, Marienhospital Stuttgart,* Stuttgart, Germany; and the Department of General Surgery, University of Ulm, t Ulm, Germany
Objective The authors conducted a study of patients who underwent total gastrectomy for gastric malignancy to elucidate contributing factors that lead to successful management of this disease in geriatric patients.
Summary Background Data The average mortality rate for patients undergoing stomach surgery due to carcinoma is 7.8% according to the literature overview, still relatively high. Even higher mortality rates are observed for geriatric patients after a total gastrectomy. Because of epidemiologic changes, a total gastrectomy is required with growing frequency in these high-risk patients.
Methods The study involved 380 patients with a gastric malignancy. Risks and benefits of a total gastrectomy with radical lymphadenectomy at an advanced age were analyzed retrospectively in 163 patients older than 70 years of age. The results achieved in these patients were compared with those observed in 217 younger patients.
Results The 30-day mortality and morbidity rates for the elderly patients were 3% and 33.7%, respectively; for the younger patients, they were 0.46% and 21.2%, respectively. A statistically significant correlation was found between the presence of risk factors, the occurrence of complications, and the mortality rate. No difference was seen between the two age groups when risk factors were absent. The 5-year survival rate was 30%, with no difference between young and elderly patients.
Conclusions The data prove that a total gastrectomy with a radical lymphadenectomy can be carried out safely in older patients, with long-term results comparable to those achieved in younger patients.
37
Patient Age
Male 150(69.1) 95 (58.3) Female 67 (30.9) 68 (41.7)
Age (yr) 57 ± 10 75.5 ± 5 pTNM
IA 21(9.8) 16 (10.1) IB 23 (10.8) 19 (12.1) 11 35 (16.4) 31 (19.6) IIIA 39 (18.3) 34 (21.5) IIIB 33 (15.5) 24 (15.2) IV 62 (29.2) 34 (21.5)
Nonepithelial 9 (2.4)
*N = 380. Values in parentheses are percentages.
Gastric carcinoma continues to be of major clinical importance, although during the past few decades, a
worldwide decrease in its incidence has been observed.' Two main reasons are responsible for this:
1. A growing number of geriatric patients are being admitted, frequently with tumors at an advanced stage, as a result ofthe increased average life expec- tancy.
2. An altered pathoanatomic behavior of gastric car- cinoma with increasing localization in the upper sections of the stomach and a predominance of the diffuse growth type, as defined by Lauren, is ob- served in western countries.2 Therefore, to achieve a curative oncologic resection, total gastrectomy is required with growing frequency.
Because of the epidemiologic change of gastric carci- noma, there appears to be an urgent need to investigate whether advanced age still should be considered a con-
traindication for a total gastrectomy. In a retrospective study, the risks and advantages oftotal gastrectomy at an advanced age are analyzed in 163 patients older than 70 years of age.
PATIENTS AND METHODS Gastric resection due to a malignant tumor was per-
formed in 604 patients (resection rate 85%) at the De-
Presented in part at the First International Gastric Cancer Congress; March 29-April 1, 1995; Kyoto,Japan.
Address reprint requests to Reinhard Bittner, M.D., Marienhospital Stuttgart, Allgemeinchirurgische Klinik, Arztlicher Direktor, B1- heimstr, 37, 70199 Stuttgart, Germany.
Accepted for publication August 30, 1995.
partment of General Surgery, University of Ulm (May 1982-December 1989) and the Marienhospital in Stutt- gart (January 1990-December 1994) during a 13-year period. For 380 patients (63%), a total gastrectomy in the sense of a gastrectomy "of necessity" was required. This included 46 patients (12%) with a gastric stump carci- noma after a previous Billroth-II resection for a benign ulcer. At the time of surgery, 163 patients (42.9%) were older
than 70 years of age. These patients were compared with 217 younger ones (57.1 %; Table 1). The gender distribu- tion shows an increase in women among the older pa- tients. With regard to the TNM stages, there was no sig- nificant difference between the two age groups. An R2 (D2) total gastrectomy3 with the corresponding lymph- adenectomy of the compartments I and II (lymph node groups 1- 12) was performed in all patients, with the ex-
ception ofthose with distant metastases observed during surgery.4 An extended procedure was necessary in 40% ofthe patients ofboth age groups (Table 2). Reconstruc- tion of the esophagus was performed in 73% of the pa- tients regardless of age, according to Hunt-Lawrence- Rodino (formation of a pouch but bypassing the duode- num). The duodenal passage was preserved by forming an interposition in only 14% ofthe younger patients. Re- construction was performed with an omega loop accord- ing to Graham in 11% ofthese patients. The correspond- ing percentages for the older patients were 0.6% and 26.4%, respectively. A Roux-en-Y reconstruction was performed in only
six patients. The esophagojejunal anastomosis always was sewn as an end-to-side anastomosis, two-row single with polyglactin 910 3-0 and 4-0 suture material, and more recently, exclusively with polydioxanone suture 4.0. All anastomoses were radiologically checked on the fifth postoperative day by administering meglumine dia-
Table 2. COMPARISON OF TOTAL GASTRECTOMY EXTENSION
Patient Age
<70 yr >70 yr Gastrectomy Extension (n = 217) (n = 163)
Splenectomy 19 14 Colon resection 5 9 Liver resection 7 4 Pancreatic resection 6 2 Resection of gastric remnant 28 18 Abdominal thoracic approach 2 1 Other* 20 14 Total 87 (40%) 62(38%)
* Cholecystectomy, ovariectomy, small bowel resection, pleura, diaphragm.
Ann. Surg. -July 1996
Table 3. COMPARISON OF MAJOR POSTOPERATIVE COMPLICATIONS
Patient Age
<70 yr >70 yr Total No. Complication (n = 217) (n = 163) (n = 380)
Pulmonary 17(7.8) 21 (12.8) 38(10) Local infection 5 (2.3) 5 (3.0) 10 (2.6) Bleeding 4 (1.8) 5 (3.0) 9 (2.4) Pancreatitis/fistula 3 (1.4) 5 (3.0) 8 (2.1) Anastomotic leak 6 (2.8) 2 (1.2) 8 (2.1) Urinary infection - 6 (3.6) 6 (1.6) Cardial 3 (1.4) 3 (1.8) 6 (1.6) Intra-abdominal abscess 3 (1.4) 3 (1.8) 6 (1.6) Cerebral 1 (0.46) 2(1.2) 3(0.8) Pulmonary embolism 2 (0.92) - 2 (0.5) Enterocutaneous fistula - 2 (1.2) 2 (0.5) Ileus 1 (0.46) 1 (0.6) 2(0.5) Insufficiency of
duodenal stump 1 (0.46) - 1 (0.25)
Total 46 (21.2) 55 (33.7) 101 (26.6)
Reoperation 4 (1.8) 8 (4.9) 12 (3.1)
Values in parentheses are percentages.
trizoate (Gastrografin, Squibb, Munich, Germany) and the oral diet was started after this. The following risk factors were defined: hemoglobin
level < 10 g/dL, reduced vital lung capacity < 80%, FEVI < 60%, chronic renal insufficiency (creatinine level > 1.5 mg/dL), diabetes mellitus, and electrocardiogram abnor- malities.
Statistics The Kaplan-Meier method was employed for the cal-
culation of survival rates. Comparisons of survival were performed with the log-rank test and tabulated data were analyzed using the chi-square test.
between the operation and death (3 patients died later than 3 months after surgery, and 2 patients died later than 2 months and 1 month afterward, respectively). Only 3 of the 217 younger patients died (1.4%), and the 30-day mortality rate was 0.46%. In the older group, seven patients (4.3%) died, and the 30-day mortality rate was 3%. The difference between the two age groups was statistically negligible. Causes of death for the patients older than 70 years ofage were pneumonia (2), sepsis (2), heart failure (1), bleeding (1), and cachexia (1). With the exception of one patient who died of pneumonia, all other patients who died had tumors at an advanced stage (pTNM III and IV) and had at least two ofthe described risk factors. Two of the younger patients who died also had tumors at an advanced stage (pTNM IV), and they died of protracted sepsis due to anastomotic insuffi- ciency. The third young patient died acutely on the 11th postoperative day from a massive pulmonary embolism.
Morbidity and Risk Factors with Regard to Age
Altogether, 101 complications (corresponding to a complication rate of 26.6%) were observed (Table 3). The number of complications was statistically signifi- cantly higher for the older patients than for the younger ones (p < 0.05). The most frequent complication con- cerned the lungs; all others tended to be of secondary importance. An anastomotic insufficiency or fistula was observed in six younger patients (2.8%) and in only one patient older than 70 years of age (1.2%). Only one of these patients had to undergo surgery again. A total of eight older patients (4.9%) and four younger patients
Table 4. COMPARISON OF RISK FACTORS
Patient Age
<70 yr >70 yr Total No. Risk Factor (n = 217) (n = 163) (n = 380)
RESULTS
Surgical Mortality and Causes of Death with Regard to Age
The average operating time was 276 ± 70 minutes for the younger patients and 241 ± 67 minutes for the older patients. Average postoperative hospital stay was 17.5 days (range, 7-53 days) in the younger group and 21 days (range, 7-115 days) in the older group. There were no
significant statistical differences between the two groups. Ten patients (2.6%) died as a result ofthe operation. This includes all patients whose deaths were directly or indi- rectly due to the operation, regardless ofthe time interval
Hemoglobin < 10 g/dL
Chronic pulmonary disease*
insufficiency ECG abnormalities
7(3.1) 31 (14.4)
3 (1.8) 50 (30.6)
10 (2.6) 81 (21.3)
89 (41.0) 158 (96.9) 247 (65.0)
ECG = electrocardiogram. * Vital capacity (VC) < 80% or forced expiratory volume in the first second < 60% (FeV,). Values in parentheses are percentages.
Vol. 224 - No. 1
40 Bittner and Others
Table 5. COMPARISON OF RISK FACTORS AND FREQUENCY OF COMPLICATIONS
Risk Factors
Patient Age
<70 yr (n = 217) >70 yr (n = 163) Total No. (n = 380)
(+)(n = 47) (-) (n = 170) (+)(n = 83) (-) (n = 80) (+)(n = 130) (-)(n = 250) (21.6%) (78.4%) (50.9) (49.1%) (34.2) (65.8)
Frequency of complications* 13 (27.6) 28 (16.6) 26 (31.3) 17 (20.7) 39 (30.0) 45 (18.0) Nonsurgical complications 5 (10.6) 18 (10.7) 20 (24.1) 9 (11.0) 25 (19.2) 27 (10.8) Mortality 2(4.2) 1 (0.6) 5(6.0) 2(2.4) 7(5.4) 3(1.2)
Values in parentheses are percentages. * Bleeding, anastomotic leak fistula, ileus, pancreatitis.
(1.8%) underwent surgery again; however, the difference was not statistically significant. The risk factors ofboth groups were statistically different
(p < 0.01). Previous damage to the lungs (p < 0.01), hypo- albuminemia (p < 0.01), diabetes mellitus (p < 0.01), and electrocardiogram abnormalities (p < 0.01) were observed more frequently in the older patients (Table 4). With the exception of electrocardiogram changes, the
complication rate was statistically different among pa- tients with and without risk factors (Table 5), regardless of age. Complications and mortality rates were signifi- cantly more frequent and higher for the patients with risk factors than for the patients without (p < 0.01 and p < 0.05). Complication and mortality rates in patients with- out risk factors were not discernible in both groups; how- ever, the frequency of nonsurgical complications for the older patients with risk factors was more than twice as high as that for the corresponding younger patients (24.1% vs. 10.6%, insignificant). The distribution of the TMN stages III and IV and of
the substitute stomach formation according to Hunt- Lawrence-Rodino was nearly identical in all groups, with 60% and 70%, respectively.
Long-Term Survival with Regard to the Patient's Age A complete follow-up was achieved for 358 patients
(94.2% of all patients on whom a total gastrectomy was performed). The operative deaths were included in the 5- year survival rate calculated in accordance with Kaplan- Meier (Fig. 1). The 5-year survival rate totaled 30%. A higher rate of approximately 10% was observed for the younger patients; however, the difference to their older counterparts was not statistically significant.
DISCUSSION The average life expectancy in Germany has more
than doubled over the last 100 years and has increased
by more than 5 years in the last 15 years.5 Accordingly, more and more elderly people with gastric carcinoma are observed compared with the 1970s.6 Moreover, the al- tered pathoanatomic behavior of gastric carcinoma2'7 has resulted in a considerable rise in the number of older patients requiring total gastrectomy. Whereas in past de- cades the share ofpatients with total gastrectomy did not exceed 8% to 1 0%,6,8 it currently is nearly three times as high. Because of the high morbidity and mortality rates in earlier years, an age of older than 70 years was consid- ered a contraindication for total gastrectomy in many institutions.9 Despite the enormous advances in surgical technique, anesthesia, and intensive care medicine over the last 20 years, the average mortality rate for stomach surgery due to carcinoma is 7.8%, according to the liter- ature overview, still relatively high.'0 Even higher mor- tality rates are observed for elderly patients after total gastrectomy. There are reports ofa mortality rate of20% for patients older than 70 years ofage and ofgreater than 25% for those older than 75 years of age.8 More favor- able results are obtained only in Japan'2" 3; however, even Habu and Endo'4 reported a mortality rate of 8.6% after total gastrectomy in elderly patients. Nearly all au- thors observe a higher death rate among elderly patients when compared with younger ones.8"2"4 However, it is uniformly stated that the current age of the patients is less important for the poorer results than the accompa- nying illnesses and risk factors,'4-'6 whereby it was possi- ble to identify preexisting lung damage as the main fac- tor.'6 For this reason, for example, a mortality rate of only 2% for patients older than 70 years ofage was deter- mined in colon surgery, provided that there were no risk factors or only one risk factor present. However, the mortality rate rose to 16% 17 and even to 3 1% 18 for two or more risk factors. The scope of the operation, its dura- tion, and the amount of blood loss during the operation were identified as additional factors with adverse influ- ence on the success ofthe operation. 3""4"1
Ann. Surg. *-July 1996
Figure 1. Long-term survival with regard to the patient's age.
2
Survival [%]
survival time [years] Dep. of surgery, Marienhospital Stuttgart, Germany
- < 70 years (n=217)
+ > 70 years (n=141)
4 5
For the most part, our results confirm reports in the literature; however, it was possible for the first time to analyze a very large number ofelderly patients with total gastrectomy. In our experience, the mortality rate after a total gastrectomy in patients older than 70 years of age still was very high (32%), even up to the end of the 1970s.6 Improvements in surgical technique, introduc- tion of standardized perioperative antibiotics (one shot preoperatively), and thromboembolic prophylaxis led to the drastic reduction of the mortality rate to 4.4%.2° In addition, a systemic analysis of the risk factors and their thorough preoperative therapy also was done, with par- ticular attention paid to intensive physical and breathing therapy. Patients in poor nutritional condition received preoperative high-caloric nourishment for approxi- mately 5 to 7 days. Careful routine observance of these principles led to a reduction ofthe 30-day mortality rate
to 3% for the elderly and to 0.5% for younger patients. Technical complications seldom occur currently, e.g., none of the elderly patients died because of suture in- sufficiency. Our results clearly indicate that age alone is not a significant risk factor; the complication and mor- tality rates for the two age groups without risk factors do not differ significantly. However, if risk factors are pres- ent, the complication and mortality rates significantly in- crease, regardless of the patient's age. Nevertheless, the results show a tendency for more nonsurgical complica- tions for elderly patients with risk factors compared with younger patients, so that a reduced tolerance must in fact be assumed for elderly patients. In particular, the combi- nation of advanced age, advanced tumor stage, and multiple risk factors proved to be especially unfavorable for the operative outcome, so that a decision in favor of a total gastrectomy in these patients can only be made
Gastric Carcinoma
Figure 2. Algorithm detailing our decision-mak- ing process in gastric carcinoma.
+ I
Billroth II Total Gastrectomy Lymphadenekt. + Lymphadenekt.
I,, ~
Middle, Upper Lower Third Third
Billroth II + Lymphadenekt.
Interposition - pouch
HLR * - pouch
42 Bittner and Others
with extreme caution. From these experiences, the algo- rithm demonstrated in Figure 2 was developed for the decision-making process. To get the best results, we have to focus on the age and risk factors of a patient, the his- tologic growth pattern according to Lauren,7 and the lo- calization of the cancer. In a current literature overview, the 5-year survival rate for patients with gastric carci- noma in the western countries is stated to be 15% to 17%.2' The total collective 5-year survival rate is 30%. There is no significant difference between elderly and young patients. The observed tendency for a better sur- vival in the younger patients can be explained with their lower operative mortality, possibly also with their lower comorbidity. Thus, only 6.8% of the younger patients with the curable TNM stage I carcinoma die within the first 18 months postoperatively versus 14.3% in the elder group. Compared with the patients on whom we operated un-
til the beginning of the 1980s,20 the current survival rate has nearly doubled. This quite considerable improve- ment in the 5-year survival rate may result from the lymphadenectomy of compartments I and II, systemati- cally performed since that time.
SUMMARY
During a 13-year period, gastric resection for a malig- nant tumor was performed in 604 patients. Total gastrec- tomy was necessary in 380 patients (63%). In this group, 163 patients (42.9%) had reached or exceeded the age of 70 years, whereas 217 patients (57.1%) were younger. The 30-day mortality and morbidity rates for the elderly patients were 3% and 33.7%, respectively; for the youn- ger patients, they were 0.46% and 21.2%, respectively. A statistically significant correlation was found between the presence of risk factors, the occurrence of complica- tions, and the mortality rate. There was no statistically demonstrable difference between elderly and young pa- tients without risk factors, so that age alone cannot be considered a risk factor. However, elderly patients with tumors at an advanced stage and multiple risk factors represented a high-risk group in which total gastrectomy should be avoided whenever possible. The total collective 5-year survival rate was 30%, with
no significant difference between young and elderly pa- tients existed. The analysis of our large number of cases shows that a
total gastrectomy with a radical lymphadenectomy also can
Ann. Surg. -July 1996
be performed safely in older patients with long-term results comparable to those achieved in younger patients.
References 1. Howson CP, Hujama T, Wynder EL. The decline in gastric cancer
epidemiology ofan unplanned triumph. Epidemiol Rev 1986; 8:1-27. 2. Meyers WC, Damiano RJ, Rotolo FS, Postlethwait RW. Adeno-
carcinoma of the stomach: changing patterns over the last 4 de- cades. Ann Surg 1987; 205:1-8.
3. Dent DM, Madden MV, Price SK. Randomized comparison ofR I and R2 gastrectomy for gastric carcinoma. Br J Surg 1988; 75:1 110- 112.
4. Japanese Research Society for Gastric Cancer. The general rules for the gastric cancer study in surgery and pathology. Jpn J Surg 1981; 11: 127-139.
5. Statistisches Bundesamt Wiesbaden Allgemeine Verlag. Wirt- schaft und Statistik. 1991; 6:371-381.
6. Bittner R, Beger HG, Kraas E, Gogler H. Surgical treatment of carcinoma of the stomach in geriatric patients. Front Gastrointest Res…