- Email: [email protected]
Surgical outcomes and survival after multiorgan resection for locally advanced gastric cancer
The American Journal of Surgery (2009) 198, 25-30
Clinical Surgery-International
Surgical outcomes and survival after multiorgan resection for locally advanced gastric cancer Ilter Ozer, M.D.*, E. Birol Bostanci, M.D., Taner Orug, M.D., Yusuf B. Ozogul, M.D., Murat Ulas, M.D., Metin Ercan, M.D., Can Kece, M.D., Fuat Atalay, M.D., Musa Akoglu, M.D. Turkiye Yuksek Ihtisas Training and Research Hospital, Demirlibahce Mah Doganbahcesi Sok. 6/3, 06340, Mamak, Ankara, Turkey KEYWORDS: Gastric cancer; Locally advanced; T4; Resection
Abstract BACKGROUND: Multiple organ resection for locally advanced (assumed T4) gastric cancer is associated with high morbidity and mortality. Our aim was to evaluate the efficacy of these surgeries with regard to surgical morbidity, mortality, and survival. METHODS: Fifty-six patients underwent potentially radical gastrectomy combined with invaded organ resection. Early and late results of multiorgan resection and clinicopathologic factors influencing these results were evaluated. RESULTS: Forty patients (71.4%) received 1 additional organ resection and 16 patients (28.6%) received 2 or more additional organ resections. Postoperative morbidity and mortality was 37.5% and 12.5%, respectively. Resection of 2 or more additional organs increased postoperative morbidity and advanced age increased mortality. The 1- and 3-year survival rates were 53.3% and 28.1%, respectively. Advanced age, lymph node metastasis, and resection of more than 1 additional organ were significant prognostic factors for survival. CONCLUSIONS: For patients with locally advanced gastric carcinoma, multiple organ resection is worthwhile with careful patient selection. © 2009 Elsevier Inc. All rights reserved.
Radical resection is the most effective treatment modality for gastric cancer and R0 resection is the most important indicator of long-term survival for patients with gastric cancer.1 Despite improvements in the surgical technique, long-term survival of patients with invasion to adjacent organs has been reported to be poor.2 For locally advanced gastric cancer with adjacent organ infiltration, extended resection including the invaded organ is required to achieve R0 resection with negative surgical margins. However, multiorgan resection has been associated with increased post-
operative morbidity and mortality.1,3 Aggressive surgical treatment and appropriate patient selection for locally advanced gastric cancer still remain controversial. The aim of this study was to evaluate the short- and long-term results of multiorgan resection for locally advanced gastric cancer and to indicate which patients might benefit from en bloc resection.
Patients and Methods * Corresponding author. Tel.: ⫹903123201802; fax: ⫹903123116351. E-mail address: [email protected] Manuscript received March 6, 2008; revised manuscript June 25, 2008
0002-9610/$ - see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2008.06.031
Among 722 gastric adenocarcinoma patients who underwent gastrectomy between January 1998 and December
26
The American Journal of Surgery, Vol 198, No 1, July 2009
2006, 56 patients underwent multiorgan resection for locally advanced gastric cancer. Data from these patients were obtained retrospectively from a prospectively recorded database. Early and late results of these patients were analyzed. Clinicopathologic features of the patients also were analyzed. All patients underwent potentially curable gastrectomy with combined resection of invaded organs and with histopathologically confirmed clear surgical margins. Multiorgan resection was defined as resection of the stomach with directly invaded organs. Resection of the distal esophagus for proximal expansion of the tumor, splenectomy for lymph node dissection (LND), and additional procedures owing to disorders other than tumoral invasion were not regarded as additional organ resection and were not included in this study. Patients with peritoneal carcinomatosis and distant metastasis also were excluded. Surgical morbidity was defined as any complication that occurred in the 30-day postoperative period. Complications were reported in a graded fashion on a scale of 1–5, as previously described.4 A grade 1 complication required basic monitoring, oral antibiotics, bowel rest, or supportive care (bedside management). A grade 2 complication required intravenous medication (antibiotics), transfusions, chest tubes, prolonged tube feedings, or total parenteral nutrition. A grade 3 complication required interventional radiology, reoperation, intensive care unit admission, intubation, or bronchoscopy. A grade 4 complication resulted in permanent disability (renal failure requiring dialysis) or organ resection, and a grade 5 complication resulted in the patient’s death. Postoperative mortality was defined as deaths within 30 days after the surgery. A distal subtotal gastrectomy or total gastrectomy (TG) was performed depending on the location of the primary tumor. The most common preferred lymphadenectomy was D2 LND. However, D1 or D3 LND also was performed in selected cases. Staging was performed according to the 1997 TNM staging system of the Union Internationale Contra le Cancer. All of our patients after curative gastrectomy were directed to the medical oncology clinic to be evaluated for adjuvant therapy. The patients who needed to receive adjuvant therapy and the treatment protocols were determined by medical oncologists. Our patients in this series did not receive neoadjuvant treatment. Follow-up information was obtained from outpatient clinical visits and telephone interviews. Statistical analysis was performed using SPSS software (SPSS, Chicago, IL). Statistical comparisons were performed using the Mann–Whitney U test, the chi-square test, or the Fisher exact test. In addition, logistic regression analysis was performed to estimate the influence of clinicopathologic factors on the development of postoperative morbidity and mortality. Odds ratios (ORs) with 95% confidence intervals (95% CIs) were calculated as a measure of association. Survival was calculated using the Kaplan– Meier method and statistical significance was assessed by the log-rank test. Cox regression analysis was performed to
estimate the influence of patient-related parameters on survival.
Results Fifty-six patients underwent gastrectomy with additional organ resection for locally advanced (assumed T4) gastric cancer. All of the surgeries were performed with curative intent and clear surgical margins were confirmed by histopathologic examination. Details concerning the clinicopathologic factors are shown in Table 1. Forty patients (71.4%) were men and 16 (28.6%) were women. Their mean age was 57.84 ⫾ 12.0 years, ranging between 31 and 81 years. Fifteen patients (26.8%) underwent distal subtotal gastrectomy and 41 (73.2%) underwent TG. Distal esophagectomy was added to TG for 6 patients and total esophagectomy was added for 1 patient. Twenty-six patients
Table 1
Patient characteristics
Mean age, y Sex Male Female Mean tumor size, cm Resection type TG Subtotal gastrectomy Primary tumor site Proximal Middle Distal Number of organs resected 1 ⱖ2 Comorbidity** ⫺ ⫹ Differentiation Well–moderate Poor–undifferentiated Borrmann I–II III–IV Lymph node involvement ⫺ ⫹ LND D1 D2 D3 Depth of invasion T3 T4 Stage II III IV *Mean ⫾ standard deviation. **Cardiac, respiratory, and/or metabolic diseases.
57.8 ⫾ 12.0* 40 (71.4%) 16 (28.6%) 91.3 ⫾ 37.1* 41 (73.2%) 15 (26.8%) 11 (19.6%) 26 (46.5%) 19 (33.9%) 40 (71.4%) 16 (28.6%) 34 (60.7%) 22 (39.3%) 17 (30.3%) 39 (69.7%) 22 (39.3%) 34 (60.7%) 11 (19.6%) 45 (80.4%) 16 (28.6%) 26 (46.4%) 14 (25.0%) 19 (33.9%) 37 (66.1%) 4 (7.1%) 17 (30.4%) 35 (62.5%)
I. Ozer et al.
Multiorgan resection for T4 gastric cancer
(46.4%) underwent D2, 16 (28.6%) underwent D1, and 14 (25%) underwent D3 LND. The majority of the patients had lymph node involvement (45; 80.4%). Histopathologic examination did not confirm T4 disease in 19 patients (33.9%). These patients did not reveal microscopic adjacent organ resection. T4 disease was confirmed by histopathologic examination in 37 patients (66.1%). Forty patients (71.4%) received one additional organ resection and 16 (28.6%) received 2 or more additional organ resections. The most common organ resected was the pancreas and the most common type of pancreatic resection was a distal pancreatectomy. The colon was the second most common site of additional organ resection (Table 2). Intraoperative complications occurred in 5 patients. Pleural injury occurred in 2 patients. The other intraoperative complications were splenic artery injury, incomplete stapler ring, and vena porta and middle colic artery injury, each of which occurred in 1 patient. A total of 21 patients had postoperative complications. These complications are listed in Table 3. Reoperations were required for 4 of these patients. When complications were classified according to severity, grade 1 complications were seen in 5 (8.9%) patients, grade 2 complications were seen in 7 (12.5%), grade 3 complications were seen in 2 (3.6%), and grade 5 complications (mortality) were seen in 7 (12.5%) patients. We did not observe any grade 4 complications. Bivariate analysis of individual risk factors revealed that resection of 2 or more organs with gastrectomy was associated with higher surgical morbidity (P ⫽ .008). Age older than 70 years (P ⫽ .009.) was associated with higher surgical mortality. Logistic regression analysis showed that 2 or more additional organ resections was a significant risk factor for morbidity (OR, 5.0; 95% CI, 1.45–17.27) and age
Table 2
27 Table 3
Complications seen in 21 patients
Complication
n
Cardiac Sepsis/MRSA infection Intra-abdominal abscess Psychiatric Respiratory Radiologic leakage Ascites/serous drainage Fever (transient and with unknown etiology) Clinical leakage of anastomosis Wound dehiscence Wound infection Duodenal stump leakage Organ perforation Intra-abdominal hemorrhage Gastrointestinal hemorrhage Total
3 3 3 3 2 2 2 2 1 1 1 1 1 1 1 27
MRSA ⫽ methicillin-resistant Staphylococcus aureus.
older than 70 was a significant risk factor for mortality (OR, 11.75; 95% CI, 1.73–79.63). The median follow-up period was 10.8 months (range, 1–98.7 mo). The overall survival of the patients was 53.3% at 1 year, 36.0% at 2 years, and 28.1% at 3 years (Fig. 1). The median survival was 13.3 months (95% CI, 8.6 –18.1). Survival of the patients with T3 tumors was 58.7% at 1 year and 51.3% at 3 years. Beyond a follow-up period of 13.8 months, no deaths occurred within this group of patients. Survival of the patients with T4 tumors was 51.9% at 1 year and 18.8% at 3 years. The difference between these 2 groups was not statistically significant (Fig. 2). There were only 4 patients with stage 2 (T3N0M0) gastric cancer and these patients are still alive as of this writing. Survival of stage 3 patients was 74.9% at 1 year and
Additional organ resections
Additional organ resection
n
Distal pancreas Colon* Liver† Distal pancreas ⫹ colon Partial diaphragma ⫹ crus Whipple Liver ⫹ colon Distal pancreas ⫹ liver Liver ⫹ gall bladder Gall bladder ⫹ choledoc Liver ⫹ partial diaphragma Total pancreaticoduodenectomy Total pancreaticoduodenectomy ⫹ colon Whipple ⫹ colon Distal pancreas Pancreas wedge resection Total esophagus ⫹ surrenal gland ⫹ distal pancreas Total
20 11 4 3 3 2 2 2 1 1 1 1 1 1 1 1 1 56
*Colon resections were performed as segmental resections. †Liver resections were segment or wedge resections.
Figure 1
Survival of the patients.
28
The American Journal of Surgery, Vol 198, No 1, July 2009
53.9% at 3 years. Survival of stage 4 patients was 39.9% at 1 year and 12.3% at 3 years. The difference was statistically significant. The number of organs resected, presence of comorbidities, age older than 70 years, and lymph node metastasis were associated with poor patient survival according to univariate analysis (Table 4). The prognostic parameters in the univariate analysis were fitted into the multivariate analysis model. By using a backward likelihood ratio (LR) procedure, age older than 70 years, presence of lymph node metastasis, and number of organs resected were found to have a significant influence on survival (Table 5).
Comments Multiorgan resection for locally advanced gastric cancer with adjacent organ involvement is performed for the purpose of achieving an R0 resection. However, only a small proportion of patients with gastric cancer has the chance of multiorgan resection. Colen et al5 reported that only 5% of patients with gastric cancer had undergone en bloc multiorgan resection. They also stated that multiorgan resection was performed without a significant increase in morbidity and mortality compared with gastrectomy alone. However, it generally is accepted that the morbidity and mortality of these patients increases significantly after multiorgan resection.1,6,7 With early detection of gastric cancer and appropriate surgical technique, the prognosis of gastric cancer patients has been improving. However, the prognosis of patients with locally advanced gastric cancer still remains poor.8 Survival after resection of tumors has been shown to increase when compared with that after noncurable surgeries.8 Even palliative gastrectomy has been proposed in advanced incurable gastric cancer in selected cases.9 How-
Table 4
Univariate analysis of prognostic factors Survival rate, %
Parameters Age, y ⱕ70 ⬎70 Sex Male Female Tumor size, mm ⱕ80 ⬎80 Resection type TG Subtotal gastrectomy Tumor location Upper third Middle third Lower third Number of organs resected 1 ⱖ2 Comorbidity ⫺ ⫹ Differentiation Well–moderate Poor–undifferentiated Borrmann I–II III–IV Lymph node metastasis ⫺ ⫹ LND 1 2 3 Depth of invasion T3 T4
1y
3y
P
58.1 27.8
32.4 0
.029
51.6 59.3
30.3 14.8
.710
67.7 39.8
30.5 25.1
.334
51.4 58.2
29.3 27.3
.842
50.0 57.7 46.9
37.5 29.6 25.1
.868
62.3 30.0
40.8 6.4
.003
66.9 25.8
39.0 0
.003
58.8 50.4
46.3 18.3
.212
54.0 52.7
32.2 25.3
.942
90.0 41.4
77.1 17.1
⬍.001
60.6 49.4 51.4
33.7 26.9 25.7
.729
58.7 51.9
51.3 18.8
.265
ever, the clinical efficacy of these radical surgeries has remained unclear. Extended multiorgan resections may increase the risk of intraoperative complications. In our series, 5 patients (8.9%) were subjected to intraoperative complications. The patient who was subjected to vena porta and middle colic artery injury during TG, distal pancreatectomy, and colon resection died after duodenal stump leakage on the 28th
Table 5 Prognostic factors determined by multivariate analysis
Figure 2
Survival according to histologic depth of invasion.
Variables
OR
95% CI
P
Age ⬎70 y Lymph node metastasis Number of organs resected
3.32 11.48 2.65
1.37–8.08 2.48–53.10 1.30–5.42
.008 .002 .007
I. Ozer et al.
Multiorgan resection for T4 gastric cancer
postoperative day. The other 4 patients with intraoperative complications were discharged, although 1 patient showed complications, which were treated successfully. Increased morbidity and mortality may be the result of these intraoperative complications. We generally prefer an aggressive surgical approach including multiorgan resection when we are faced with locally advanced gastric tumors. This aggressive approach may result in such intraoperative complications. All of the intraoperative complications were managed successfully during the surgeries and none of the patients was reoperated directly because of these complications. Various postoperative morbidity and mortality rates have been reported in the literature. It has been generally accepted that the morbidity and mortality of these patients increases significantly after multiorgan resection.1,10 Bloechle et al11 reported a morbidity and mortality rates of 25 and 3.6% respectively. Yong et al12 reported a higher mortality rate of 15%. However, a very low mortality rate of 1.2% was reported by Kobayashi et al8 and their morbidity rate was 28%. A complication rate of 75% also was reported13 in the literature. Several studies have reported morbidity rates of more than 25% 1,5,7,14 and most of these studies reported mortality rates of 10% or more.5,14 Different complication rates may be owing to medical recording and definition of complications. We graded the complications with a grading system according to severity and recorded even a minimal problem such as transient tachycardia or transient fever with unknown etiology. Postoperative complications occurred in 21 (37.5%) of our patients. Most of these complications were grade 1 and 2, which mainly required only short-term simple medical treatment (oral antibiotics, intravenous antibiotics, transfusion, and so forth). We observed a mortality rate of 12.5% and this rate is similar to previously reported rates mentioned earlier. These relatively high morbidity rates may have been affected by various comorbidities. Our hospital is an important tertiary center to which patients with advanced age having various comorbidities are referred and 39.3% of our patients had preoperative comorbidities. Eight of these 21 patients showed postoperative complications without any gastrointestinal problem specific to the surgery. More than half of the patients had extra gastrointestinal complications such as cardiac or respiratory problems with or without gastrointestinal complications directly related to the surgery. According to logistic regression analysis, patients with comorbidities tended to show more complications, but this tendency did not reach statistical significance (P ⫽ .09). Resection of 2 or more additional organs increased postoperative morbidity. Age older than 70 years was an independent risk factor for postoperative mortality. The mortality rate tended to be higher after 2 or more additional organ resections, however, this tendency did not reach statistical significance. Morbidity and mortality rates of our patients who underwent gastrectomy alone were 20.5% and 3.3%, respectively, and these rates were significantly lower compared with the patients who underwent multiple organ resection during the same time
29 interval. However, morbidity and mortality rates after gastrectomy with one additional organ resection were similar to those after gastrectomy alone. Macroscopic presumed T4 disease during surgery was not confirmed histopathologically in 19 of 56 patients (33.9%). In some cases dense inflammatory adhesions to adjacent organs without actual invasion may be present and this may occur in up to 55% of patients.1,15 Distinguishing the presence of T3 and T4 lesions may be challenging and intraoperative assessment of true invasion is inaccurate. Colen et al5 reported that 13 of 21 of their patients had histopathologic T3 disease. This inflammatory adhesion without actual tumor invasion may cause false overstaging of the disease and this may result in higher survival rates in this group of patients. Kobayashi et al8 reported that histopathologic T factor did not affect prognosis in their study. We also compared the survival rates of patients with T3 and T4 diseases and there was not a statistically significant difference between these 2 subgroups. This probably is owing to the small number of patients in the T3 groups. Interestingly, all of the deaths in the T3 group occurred within 13.8 months after the surgeries. We did not observe any death after that time. Multiorgan resection may be more beneficial for this group of patients. Whether these patients were overtreated may be controversial, but during the surgery it is almost impossible to detect actual microscopic T4 disease and the decision for resection should be made during the surgery. When a presumed T4 disease is present during surgery, potentially curative surgery with the removal of adjacent organs is advocated. Varying survival rates after multiorgan resection have been reported in the literature. Kobayashi et al8 reported survival rates of 59.8%, 40.9%, and 31.1% at 1, 3, and 5 years, respectively, which seems to be higher than the results of our study, although they also performed R2 resections. We performed all of the surgeries with curative intent and excluded all of the R1 and R2 resections. Yong et al12 reported a median survival of 17 months and stated that there was a poor survival benefit and multiorgan resection should be performed in a selected group of patients. Bloechle et al11 found 1- and 2-year survival rates of 64% and 44%, respectively. Carboni et al15 reported 5-year survival rates of 34.1% for T3 and 17.9% for T4 patients. The difference was found to be significant. In our series, there were only 4 patients with stage 2 gastric cancer (T3, N0) and these patients are still alive after 4, 7, 17.8, 44.4, and 98.7 months of follow-up evaluation. However, this small number of patients was not available for statistical analysis. Patients with stage 3 disease had higher survival rates than those with stage 4 disease. This finding confirms the effect of stage on survival. Age older than 70 years, presence of lymph node metastasis, and number of organs resected seem to affect survival in our series. The presence of lymph node metastasis was found to be the most important prognostic factor in this study. However, it is difficult to detect the actual status of lymph node involvement during the surgery.
30
The American Journal of Surgery, Vol 198, No 1, July 2009
Kodama et al6 reported that single-organ invasion showed better prognosis than cases with multiple organ invasion because these cases were associated more frequently with other incurable factors. Although our study excluded patients with incurable factors, patients who underwent 2 or more additional organ resections showed poorer survival. The survival of the patients who underwent potentially curative gastrectomy without additional organ resection was 74.8%, 58.7%, and 53.5% at 1, 2, and 3 years, respectively. Survival in this group of patients was significantly better compared with the patients who underwent multiple organ resection during the same time interval. Survival after gastrectomy with one additional organ resection was similar to that after gastrectomy alone. However, survival worsened when more than one additional organ resection was performed. There is increasing evidence that patients may benefit from neoadjuvant chemotherapy.16 Neoadjuvant treatment potentially downstages the tumor and therefore may improve the resectability rate with negative surgical margins. Lordick and Siewert16 stated that the end point of neoadjuvant treatment must remain the rate of R0 resection and overall survival. Our study already included patients who underwent potentially curative resections with negative surgical margins. Neoadjuvant chemotherapy may especially be beneficial in patients in whom the probability of performing a potentially curative (R0) resection is low. In conclusion, lymph node involvement, advanced age, and resection of 2 or more additional organs were associated with poorer survival. In addition, resection of 2 or more organs and advanced age increased postoperative morbidity and mortality, respectively. These factors should be kept in mind during selection of patients who are more likely to benefit from aggressive surgery. However, we must be aware of the fact that the only proven effective therapy is surgery and each patient should be given the chance of R0 resection, although multiorgan resection has a limited survival advantage because of advanced stage.
References 1. Martin RGC, Jaques DP, Brennan MF, et al. Achieving R0 resection for locally advanced gastric cancer: is it worth the risk of multiorgan resection. J Am Coll Surg 2002;194:568 –77. 2. Isozaki H, Tanaka N, Tanigawa N, et al. Prognostic factors in patients with advanced gastric cancer with macroscopic invasion to adjacent organs treated with radical surgery. Gastric Cancer 2000;3:202–10. 3. Sasako M. Risk factors for surgical treatment in the Dutch gastric cancer trial. Br J Surg 1997;84:1567–71. 4. Martin RC, Brennan MF, Jaques DP. Quality of complication reporting in the surgical literature. Ann Surg 2002;235:803–13. 5. Colen KL, Marcus SG, Newman E, et al. Multiorgan resection for gastric cancer: intraoperative and computed tomography assessment of locally advanced disease is inaccurate. J Gastrointest Surg 2004;8: 897–900. 6. Kodama I, Takamiya H, Mizutani K, et al. Gastrectomy with combined resection of other organs for carcinoma of the stomach with invasion to adjacent organs: clinical efficacy in a retrospective study. J Am Coll Surg 1997;184:16 –22. 7. Dhar DK, Kubota H, Tachibana M, et al. Prognosis of T4 gastric carcinoma patients: an appraisal of aggressive surgical treatment. J Surg Oncol 2001;76:278 – 82. 8. Kobayashi A, Nakagohri T, Konishi M, et al. Aggressive surgical treatment for T4 gastric cancer. J Gastrointest Surg 2004;8:464 –70. 9. Kunisaki C, Shimada H, Akiyama H, et al. Survival benefit of palliative gastrectomy in advanced incurable gastric cancer. Anticancer Res 2003;23:1853– 8. 10. Martin RC, Jaques DP, Brennan MF, et al. Extended local resection for advanced gastric cancer: increased survival versus increased morbidity. Ann Surg 2002;236:159 – 65. 11. Bloechle C, Izbicki JR, Limmer J, et al. Multi-visceral resection for locally advanced gastric cancer. Acta Chir Belg 1995;95:72–5. 12. Yong WS, Wong WK, Chan HS, et al. Extended resection of locally advanced (T4) stomach cancer. Ann Acad Med Singapore 2000;29: 723– 6. 13. Kasakura Y, Fujii M, Mochizuki F, et al. Is there a benefit of pancreaticosplenectomy with gastrectomy for advanced gastric cancer? Am J Surg 2000;179:237– 42. 14. Shchepotin IB, Chorny VA, Nauta RJ, et al. Extended surgical resection in T4 gastric cancer. Am J Surg 1998;175:123– 6. 15. Carboni F, Lepiane P, Santoro R, et al. Extended multiorgan resection for T4 gastric carcinoma: 25-year experience. J Surg Oncol 2005;90: 95–100. 16. Lordick F, Siewert JR. Recent advances in multimodal treatment for gastric cancer: a review. Gastric Cancer 2005;8:78 – 85.
Clinical Surgery-International
Surgical outcomes and survival after multiorgan resection for locally advanced gastric cancer Ilter Ozer, M.D.*, E. Birol Bostanci, M.D., Taner Orug, M.D., Yusuf B. Ozogul, M.D., Murat Ulas, M.D., Metin Ercan, M.D., Can Kece, M.D., Fuat Atalay, M.D., Musa Akoglu, M.D. Turkiye Yuksek Ihtisas Training and Research Hospital, Demirlibahce Mah Doganbahcesi Sok. 6/3, 06340, Mamak, Ankara, Turkey KEYWORDS: Gastric cancer; Locally advanced; T4; Resection
Abstract BACKGROUND: Multiple organ resection for locally advanced (assumed T4) gastric cancer is associated with high morbidity and mortality. Our aim was to evaluate the efficacy of these surgeries with regard to surgical morbidity, mortality, and survival. METHODS: Fifty-six patients underwent potentially radical gastrectomy combined with invaded organ resection. Early and late results of multiorgan resection and clinicopathologic factors influencing these results were evaluated. RESULTS: Forty patients (71.4%) received 1 additional organ resection and 16 patients (28.6%) received 2 or more additional organ resections. Postoperative morbidity and mortality was 37.5% and 12.5%, respectively. Resection of 2 or more additional organs increased postoperative morbidity and advanced age increased mortality. The 1- and 3-year survival rates were 53.3% and 28.1%, respectively. Advanced age, lymph node metastasis, and resection of more than 1 additional organ were significant prognostic factors for survival. CONCLUSIONS: For patients with locally advanced gastric carcinoma, multiple organ resection is worthwhile with careful patient selection. © 2009 Elsevier Inc. All rights reserved.
Radical resection is the most effective treatment modality for gastric cancer and R0 resection is the most important indicator of long-term survival for patients with gastric cancer.1 Despite improvements in the surgical technique, long-term survival of patients with invasion to adjacent organs has been reported to be poor.2 For locally advanced gastric cancer with adjacent organ infiltration, extended resection including the invaded organ is required to achieve R0 resection with negative surgical margins. However, multiorgan resection has been associated with increased post-
operative morbidity and mortality.1,3 Aggressive surgical treatment and appropriate patient selection for locally advanced gastric cancer still remain controversial. The aim of this study was to evaluate the short- and long-term results of multiorgan resection for locally advanced gastric cancer and to indicate which patients might benefit from en bloc resection.
Patients and Methods * Corresponding author. Tel.: ⫹903123201802; fax: ⫹903123116351. E-mail address: [email protected] Manuscript received March 6, 2008; revised manuscript June 25, 2008
0002-9610/$ - see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2008.06.031
Among 722 gastric adenocarcinoma patients who underwent gastrectomy between January 1998 and December
26
The American Journal of Surgery, Vol 198, No 1, July 2009
2006, 56 patients underwent multiorgan resection for locally advanced gastric cancer. Data from these patients were obtained retrospectively from a prospectively recorded database. Early and late results of these patients were analyzed. Clinicopathologic features of the patients also were analyzed. All patients underwent potentially curable gastrectomy with combined resection of invaded organs and with histopathologically confirmed clear surgical margins. Multiorgan resection was defined as resection of the stomach with directly invaded organs. Resection of the distal esophagus for proximal expansion of the tumor, splenectomy for lymph node dissection (LND), and additional procedures owing to disorders other than tumoral invasion were not regarded as additional organ resection and were not included in this study. Patients with peritoneal carcinomatosis and distant metastasis also were excluded. Surgical morbidity was defined as any complication that occurred in the 30-day postoperative period. Complications were reported in a graded fashion on a scale of 1–5, as previously described.4 A grade 1 complication required basic monitoring, oral antibiotics, bowel rest, or supportive care (bedside management). A grade 2 complication required intravenous medication (antibiotics), transfusions, chest tubes, prolonged tube feedings, or total parenteral nutrition. A grade 3 complication required interventional radiology, reoperation, intensive care unit admission, intubation, or bronchoscopy. A grade 4 complication resulted in permanent disability (renal failure requiring dialysis) or organ resection, and a grade 5 complication resulted in the patient’s death. Postoperative mortality was defined as deaths within 30 days after the surgery. A distal subtotal gastrectomy or total gastrectomy (TG) was performed depending on the location of the primary tumor. The most common preferred lymphadenectomy was D2 LND. However, D1 or D3 LND also was performed in selected cases. Staging was performed according to the 1997 TNM staging system of the Union Internationale Contra le Cancer. All of our patients after curative gastrectomy were directed to the medical oncology clinic to be evaluated for adjuvant therapy. The patients who needed to receive adjuvant therapy and the treatment protocols were determined by medical oncologists. Our patients in this series did not receive neoadjuvant treatment. Follow-up information was obtained from outpatient clinical visits and telephone interviews. Statistical analysis was performed using SPSS software (SPSS, Chicago, IL). Statistical comparisons were performed using the Mann–Whitney U test, the chi-square test, or the Fisher exact test. In addition, logistic regression analysis was performed to estimate the influence of clinicopathologic factors on the development of postoperative morbidity and mortality. Odds ratios (ORs) with 95% confidence intervals (95% CIs) were calculated as a measure of association. Survival was calculated using the Kaplan– Meier method and statistical significance was assessed by the log-rank test. Cox regression analysis was performed to
estimate the influence of patient-related parameters on survival.
Results Fifty-six patients underwent gastrectomy with additional organ resection for locally advanced (assumed T4) gastric cancer. All of the surgeries were performed with curative intent and clear surgical margins were confirmed by histopathologic examination. Details concerning the clinicopathologic factors are shown in Table 1. Forty patients (71.4%) were men and 16 (28.6%) were women. Their mean age was 57.84 ⫾ 12.0 years, ranging between 31 and 81 years. Fifteen patients (26.8%) underwent distal subtotal gastrectomy and 41 (73.2%) underwent TG. Distal esophagectomy was added to TG for 6 patients and total esophagectomy was added for 1 patient. Twenty-six patients
Table 1
Patient characteristics
Mean age, y Sex Male Female Mean tumor size, cm Resection type TG Subtotal gastrectomy Primary tumor site Proximal Middle Distal Number of organs resected 1 ⱖ2 Comorbidity** ⫺ ⫹ Differentiation Well–moderate Poor–undifferentiated Borrmann I–II III–IV Lymph node involvement ⫺ ⫹ LND D1 D2 D3 Depth of invasion T3 T4 Stage II III IV *Mean ⫾ standard deviation. **Cardiac, respiratory, and/or metabolic diseases.
57.8 ⫾ 12.0* 40 (71.4%) 16 (28.6%) 91.3 ⫾ 37.1* 41 (73.2%) 15 (26.8%) 11 (19.6%) 26 (46.5%) 19 (33.9%) 40 (71.4%) 16 (28.6%) 34 (60.7%) 22 (39.3%) 17 (30.3%) 39 (69.7%) 22 (39.3%) 34 (60.7%) 11 (19.6%) 45 (80.4%) 16 (28.6%) 26 (46.4%) 14 (25.0%) 19 (33.9%) 37 (66.1%) 4 (7.1%) 17 (30.4%) 35 (62.5%)
I. Ozer et al.
Multiorgan resection for T4 gastric cancer
(46.4%) underwent D2, 16 (28.6%) underwent D1, and 14 (25%) underwent D3 LND. The majority of the patients had lymph node involvement (45; 80.4%). Histopathologic examination did not confirm T4 disease in 19 patients (33.9%). These patients did not reveal microscopic adjacent organ resection. T4 disease was confirmed by histopathologic examination in 37 patients (66.1%). Forty patients (71.4%) received one additional organ resection and 16 (28.6%) received 2 or more additional organ resections. The most common organ resected was the pancreas and the most common type of pancreatic resection was a distal pancreatectomy. The colon was the second most common site of additional organ resection (Table 2). Intraoperative complications occurred in 5 patients. Pleural injury occurred in 2 patients. The other intraoperative complications were splenic artery injury, incomplete stapler ring, and vena porta and middle colic artery injury, each of which occurred in 1 patient. A total of 21 patients had postoperative complications. These complications are listed in Table 3. Reoperations were required for 4 of these patients. When complications were classified according to severity, grade 1 complications were seen in 5 (8.9%) patients, grade 2 complications were seen in 7 (12.5%), grade 3 complications were seen in 2 (3.6%), and grade 5 complications (mortality) were seen in 7 (12.5%) patients. We did not observe any grade 4 complications. Bivariate analysis of individual risk factors revealed that resection of 2 or more organs with gastrectomy was associated with higher surgical morbidity (P ⫽ .008). Age older than 70 years (P ⫽ .009.) was associated with higher surgical mortality. Logistic regression analysis showed that 2 or more additional organ resections was a significant risk factor for morbidity (OR, 5.0; 95% CI, 1.45–17.27) and age
Table 2
27 Table 3
Complications seen in 21 patients
Complication
n
Cardiac Sepsis/MRSA infection Intra-abdominal abscess Psychiatric Respiratory Radiologic leakage Ascites/serous drainage Fever (transient and with unknown etiology) Clinical leakage of anastomosis Wound dehiscence Wound infection Duodenal stump leakage Organ perforation Intra-abdominal hemorrhage Gastrointestinal hemorrhage Total
3 3 3 3 2 2 2 2 1 1 1 1 1 1 1 27
MRSA ⫽ methicillin-resistant Staphylococcus aureus.
older than 70 was a significant risk factor for mortality (OR, 11.75; 95% CI, 1.73–79.63). The median follow-up period was 10.8 months (range, 1–98.7 mo). The overall survival of the patients was 53.3% at 1 year, 36.0% at 2 years, and 28.1% at 3 years (Fig. 1). The median survival was 13.3 months (95% CI, 8.6 –18.1). Survival of the patients with T3 tumors was 58.7% at 1 year and 51.3% at 3 years. Beyond a follow-up period of 13.8 months, no deaths occurred within this group of patients. Survival of the patients with T4 tumors was 51.9% at 1 year and 18.8% at 3 years. The difference between these 2 groups was not statistically significant (Fig. 2). There were only 4 patients with stage 2 (T3N0M0) gastric cancer and these patients are still alive as of this writing. Survival of stage 3 patients was 74.9% at 1 year and
Additional organ resections
Additional organ resection
n
Distal pancreas Colon* Liver† Distal pancreas ⫹ colon Partial diaphragma ⫹ crus Whipple Liver ⫹ colon Distal pancreas ⫹ liver Liver ⫹ gall bladder Gall bladder ⫹ choledoc Liver ⫹ partial diaphragma Total pancreaticoduodenectomy Total pancreaticoduodenectomy ⫹ colon Whipple ⫹ colon Distal pancreas Pancreas wedge resection Total esophagus ⫹ surrenal gland ⫹ distal pancreas Total
20 11 4 3 3 2 2 2 1 1 1 1 1 1 1 1 1 56
*Colon resections were performed as segmental resections. †Liver resections were segment or wedge resections.
Figure 1
Survival of the patients.
28
The American Journal of Surgery, Vol 198, No 1, July 2009
53.9% at 3 years. Survival of stage 4 patients was 39.9% at 1 year and 12.3% at 3 years. The difference was statistically significant. The number of organs resected, presence of comorbidities, age older than 70 years, and lymph node metastasis were associated with poor patient survival according to univariate analysis (Table 4). The prognostic parameters in the univariate analysis were fitted into the multivariate analysis model. By using a backward likelihood ratio (LR) procedure, age older than 70 years, presence of lymph node metastasis, and number of organs resected were found to have a significant influence on survival (Table 5).
Comments Multiorgan resection for locally advanced gastric cancer with adjacent organ involvement is performed for the purpose of achieving an R0 resection. However, only a small proportion of patients with gastric cancer has the chance of multiorgan resection. Colen et al5 reported that only 5% of patients with gastric cancer had undergone en bloc multiorgan resection. They also stated that multiorgan resection was performed without a significant increase in morbidity and mortality compared with gastrectomy alone. However, it generally is accepted that the morbidity and mortality of these patients increases significantly after multiorgan resection.1,6,7 With early detection of gastric cancer and appropriate surgical technique, the prognosis of gastric cancer patients has been improving. However, the prognosis of patients with locally advanced gastric cancer still remains poor.8 Survival after resection of tumors has been shown to increase when compared with that after noncurable surgeries.8 Even palliative gastrectomy has been proposed in advanced incurable gastric cancer in selected cases.9 How-
Table 4
Univariate analysis of prognostic factors Survival rate, %
Parameters Age, y ⱕ70 ⬎70 Sex Male Female Tumor size, mm ⱕ80 ⬎80 Resection type TG Subtotal gastrectomy Tumor location Upper third Middle third Lower third Number of organs resected 1 ⱖ2 Comorbidity ⫺ ⫹ Differentiation Well–moderate Poor–undifferentiated Borrmann I–II III–IV Lymph node metastasis ⫺ ⫹ LND 1 2 3 Depth of invasion T3 T4
1y
3y
P
58.1 27.8
32.4 0
.029
51.6 59.3
30.3 14.8
.710
67.7 39.8
30.5 25.1
.334
51.4 58.2
29.3 27.3
.842
50.0 57.7 46.9
37.5 29.6 25.1
.868
62.3 30.0
40.8 6.4
.003
66.9 25.8
39.0 0
.003
58.8 50.4
46.3 18.3
.212
54.0 52.7
32.2 25.3
.942
90.0 41.4
77.1 17.1
⬍.001
60.6 49.4 51.4
33.7 26.9 25.7
.729
58.7 51.9
51.3 18.8
.265
ever, the clinical efficacy of these radical surgeries has remained unclear. Extended multiorgan resections may increase the risk of intraoperative complications. In our series, 5 patients (8.9%) were subjected to intraoperative complications. The patient who was subjected to vena porta and middle colic artery injury during TG, distal pancreatectomy, and colon resection died after duodenal stump leakage on the 28th
Table 5 Prognostic factors determined by multivariate analysis
Figure 2
Survival according to histologic depth of invasion.
Variables
OR
95% CI
P
Age ⬎70 y Lymph node metastasis Number of organs resected
3.32 11.48 2.65
1.37–8.08 2.48–53.10 1.30–5.42
.008 .002 .007
I. Ozer et al.
Multiorgan resection for T4 gastric cancer
postoperative day. The other 4 patients with intraoperative complications were discharged, although 1 patient showed complications, which were treated successfully. Increased morbidity and mortality may be the result of these intraoperative complications. We generally prefer an aggressive surgical approach including multiorgan resection when we are faced with locally advanced gastric tumors. This aggressive approach may result in such intraoperative complications. All of the intraoperative complications were managed successfully during the surgeries and none of the patients was reoperated directly because of these complications. Various postoperative morbidity and mortality rates have been reported in the literature. It has been generally accepted that the morbidity and mortality of these patients increases significantly after multiorgan resection.1,10 Bloechle et al11 reported a morbidity and mortality rates of 25 and 3.6% respectively. Yong et al12 reported a higher mortality rate of 15%. However, a very low mortality rate of 1.2% was reported by Kobayashi et al8 and their morbidity rate was 28%. A complication rate of 75% also was reported13 in the literature. Several studies have reported morbidity rates of more than 25% 1,5,7,14 and most of these studies reported mortality rates of 10% or more.5,14 Different complication rates may be owing to medical recording and definition of complications. We graded the complications with a grading system according to severity and recorded even a minimal problem such as transient tachycardia or transient fever with unknown etiology. Postoperative complications occurred in 21 (37.5%) of our patients. Most of these complications were grade 1 and 2, which mainly required only short-term simple medical treatment (oral antibiotics, intravenous antibiotics, transfusion, and so forth). We observed a mortality rate of 12.5% and this rate is similar to previously reported rates mentioned earlier. These relatively high morbidity rates may have been affected by various comorbidities. Our hospital is an important tertiary center to which patients with advanced age having various comorbidities are referred and 39.3% of our patients had preoperative comorbidities. Eight of these 21 patients showed postoperative complications without any gastrointestinal problem specific to the surgery. More than half of the patients had extra gastrointestinal complications such as cardiac or respiratory problems with or without gastrointestinal complications directly related to the surgery. According to logistic regression analysis, patients with comorbidities tended to show more complications, but this tendency did not reach statistical significance (P ⫽ .09). Resection of 2 or more additional organs increased postoperative morbidity. Age older than 70 years was an independent risk factor for postoperative mortality. The mortality rate tended to be higher after 2 or more additional organ resections, however, this tendency did not reach statistical significance. Morbidity and mortality rates of our patients who underwent gastrectomy alone were 20.5% and 3.3%, respectively, and these rates were significantly lower compared with the patients who underwent multiple organ resection during the same time
29 interval. However, morbidity and mortality rates after gastrectomy with one additional organ resection were similar to those after gastrectomy alone. Macroscopic presumed T4 disease during surgery was not confirmed histopathologically in 19 of 56 patients (33.9%). In some cases dense inflammatory adhesions to adjacent organs without actual invasion may be present and this may occur in up to 55% of patients.1,15 Distinguishing the presence of T3 and T4 lesions may be challenging and intraoperative assessment of true invasion is inaccurate. Colen et al5 reported that 13 of 21 of their patients had histopathologic T3 disease. This inflammatory adhesion without actual tumor invasion may cause false overstaging of the disease and this may result in higher survival rates in this group of patients. Kobayashi et al8 reported that histopathologic T factor did not affect prognosis in their study. We also compared the survival rates of patients with T3 and T4 diseases and there was not a statistically significant difference between these 2 subgroups. This probably is owing to the small number of patients in the T3 groups. Interestingly, all of the deaths in the T3 group occurred within 13.8 months after the surgeries. We did not observe any death after that time. Multiorgan resection may be more beneficial for this group of patients. Whether these patients were overtreated may be controversial, but during the surgery it is almost impossible to detect actual microscopic T4 disease and the decision for resection should be made during the surgery. When a presumed T4 disease is present during surgery, potentially curative surgery with the removal of adjacent organs is advocated. Varying survival rates after multiorgan resection have been reported in the literature. Kobayashi et al8 reported survival rates of 59.8%, 40.9%, and 31.1% at 1, 3, and 5 years, respectively, which seems to be higher than the results of our study, although they also performed R2 resections. We performed all of the surgeries with curative intent and excluded all of the R1 and R2 resections. Yong et al12 reported a median survival of 17 months and stated that there was a poor survival benefit and multiorgan resection should be performed in a selected group of patients. Bloechle et al11 found 1- and 2-year survival rates of 64% and 44%, respectively. Carboni et al15 reported 5-year survival rates of 34.1% for T3 and 17.9% for T4 patients. The difference was found to be significant. In our series, there were only 4 patients with stage 2 gastric cancer (T3, N0) and these patients are still alive after 4, 7, 17.8, 44.4, and 98.7 months of follow-up evaluation. However, this small number of patients was not available for statistical analysis. Patients with stage 3 disease had higher survival rates than those with stage 4 disease. This finding confirms the effect of stage on survival. Age older than 70 years, presence of lymph node metastasis, and number of organs resected seem to affect survival in our series. The presence of lymph node metastasis was found to be the most important prognostic factor in this study. However, it is difficult to detect the actual status of lymph node involvement during the surgery.
30
The American Journal of Surgery, Vol 198, No 1, July 2009
Kodama et al6 reported that single-organ invasion showed better prognosis than cases with multiple organ invasion because these cases were associated more frequently with other incurable factors. Although our study excluded patients with incurable factors, patients who underwent 2 or more additional organ resections showed poorer survival. The survival of the patients who underwent potentially curative gastrectomy without additional organ resection was 74.8%, 58.7%, and 53.5% at 1, 2, and 3 years, respectively. Survival in this group of patients was significantly better compared with the patients who underwent multiple organ resection during the same time interval. Survival after gastrectomy with one additional organ resection was similar to that after gastrectomy alone. However, survival worsened when more than one additional organ resection was performed. There is increasing evidence that patients may benefit from neoadjuvant chemotherapy.16 Neoadjuvant treatment potentially downstages the tumor and therefore may improve the resectability rate with negative surgical margins. Lordick and Siewert16 stated that the end point of neoadjuvant treatment must remain the rate of R0 resection and overall survival. Our study already included patients who underwent potentially curative resections with negative surgical margins. Neoadjuvant chemotherapy may especially be beneficial in patients in whom the probability of performing a potentially curative (R0) resection is low. In conclusion, lymph node involvement, advanced age, and resection of 2 or more additional organs were associated with poorer survival. In addition, resection of 2 or more organs and advanced age increased postoperative morbidity and mortality, respectively. These factors should be kept in mind during selection of patients who are more likely to benefit from aggressive surgery. However, we must be aware of the fact that the only proven effective therapy is surgery and each patient should be given the chance of R0 resection, although multiorgan resection has a limited survival advantage because of advanced stage.
References 1. Martin RGC, Jaques DP, Brennan MF, et al. Achieving R0 resection for locally advanced gastric cancer: is it worth the risk of multiorgan resection. J Am Coll Surg 2002;194:568 –77. 2. Isozaki H, Tanaka N, Tanigawa N, et al. Prognostic factors in patients with advanced gastric cancer with macroscopic invasion to adjacent organs treated with radical surgery. Gastric Cancer 2000;3:202–10. 3. Sasako M. Risk factors for surgical treatment in the Dutch gastric cancer trial. Br J Surg 1997;84:1567–71. 4. Martin RC, Brennan MF, Jaques DP. Quality of complication reporting in the surgical literature. Ann Surg 2002;235:803–13. 5. Colen KL, Marcus SG, Newman E, et al. Multiorgan resection for gastric cancer: intraoperative and computed tomography assessment of locally advanced disease is inaccurate. J Gastrointest Surg 2004;8: 897–900. 6. Kodama I, Takamiya H, Mizutani K, et al. Gastrectomy with combined resection of other organs for carcinoma of the stomach with invasion to adjacent organs: clinical efficacy in a retrospective study. J Am Coll Surg 1997;184:16 –22. 7. Dhar DK, Kubota H, Tachibana M, et al. Prognosis of T4 gastric carcinoma patients: an appraisal of aggressive surgical treatment. J Surg Oncol 2001;76:278 – 82. 8. Kobayashi A, Nakagohri T, Konishi M, et al. Aggressive surgical treatment for T4 gastric cancer. J Gastrointest Surg 2004;8:464 –70. 9. Kunisaki C, Shimada H, Akiyama H, et al. Survival benefit of palliative gastrectomy in advanced incurable gastric cancer. Anticancer Res 2003;23:1853– 8. 10. Martin RC, Jaques DP, Brennan MF, et al. Extended local resection for advanced gastric cancer: increased survival versus increased morbidity. Ann Surg 2002;236:159 – 65. 11. Bloechle C, Izbicki JR, Limmer J, et al. Multi-visceral resection for locally advanced gastric cancer. Acta Chir Belg 1995;95:72–5. 12. Yong WS, Wong WK, Chan HS, et al. Extended resection of locally advanced (T4) stomach cancer. Ann Acad Med Singapore 2000;29: 723– 6. 13. Kasakura Y, Fujii M, Mochizuki F, et al. Is there a benefit of pancreaticosplenectomy with gastrectomy for advanced gastric cancer? Am J Surg 2000;179:237– 42. 14. Shchepotin IB, Chorny VA, Nauta RJ, et al. Extended surgical resection in T4 gastric cancer. Am J Surg 1998;175:123– 6. 15. Carboni F, Lepiane P, Santoro R, et al. Extended multiorgan resection for T4 gastric carcinoma: 25-year experience. J Surg Oncol 2005;90: 95–100. 16. Lordick F, Siewert JR. Recent advances in multimodal treatment for gastric cancer: a review. Gastric Cancer 2005;8:78 – 85.