Surg Endosc DOI 10.1007/s00464-015-4152-7
and Other Interventional Techniques
Minimally invasive gastrectomy for cancer: current utilization in US academic medical centers Jason A. Glenn1 • Kiran K. Turaga1 • T. Clark Gamblin1 • Samuel F. Hohmann2 Fabian M. Johnston1
•
Received: 3 December 2014 / Accepted: 2 March 2015 Ó Springer Science+Business Media New York 2015
Abstract Background Internationally, the utilization of minimally invasive techniques for gastric cancer resection has been increasing since first introduced in 1994. In the USA, the feasibility and safety of these techniques for cancer have not yet been demonstrated. Methods The University HealthSystem Consortium database was queried for gastrectomies performed between 2008 and 2013. Any adult patient with an abdominal visceral malignancy that necessitated gastric resection was included in the cohort. Clinicopathological and in-hospital outcome metrics were collected for open, laparoscopic, and robotic procedures. Results Open gastrectomies comprised 89.5 % of the total study group, while 8.2 % of procedures were performed laparoscopically, and 2.3 % were performed with robotic assistance. When accounting for disparities in patient severity of illness and risk of mortality subclass designations, there were no significant differences in mean length of stay, 30-day readmission, and in-hospital mortality between the three groups; however, mean total cost was highest in the robotic-assisted group (P = 0.017). Overall, complication rates were also similar; however, there was a higher incidence of superficial infection in the laparoscopic
& Fabian M. Johnston [email protected] Jason A. Glenn [email protected] 1
Division of Surgical Oncology, Medical College of Wisconsin, 9200 W. Wisconsin Ave, Milwaukee, WI 53226, USA
2
University HealthSystem Consortium, 155 N Upper Wacker Dr, Chicago, IL 60606, USA
group (P = 0.013) and a higher incidence of venous thromboembolism in the robotic group (P = 0.038). Conclusion Despite widespread adoption for benign indications, minimally invasive gastrectomy for cancer remains underutilized in the USA. In these patients, laparoscopic and robot-assisted gastrectomies appear to be comparable to open resection with respect to overall complications, length of stay, 30-day readmission, and inhospital mortality. However, when employing minimally invasive techniques, infection and thromboembolism risk reduction strategies should be emphasized in the operative and postoperative periods. Keywords Surgical
GI Cancer Digestive Human/robotic
Background The National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program estimates that over twenty-two thousand new gastric cancers and almost eleven thousand gastric cancer-related deaths will occur in the USA in 2014 [1]. While this incidence is almost tenfold less than what is reported in Asia, the relative late stage of presentation and poorer prognosis makes gastric cancer in the USA an important topic of discussion [1–4]. Even though the role of surgical resection for gastric cancer has been well established, the optimal operative approach remains less well defined. The fundamental surgical approach to gastric pathology has remained largely unchanged for over a hundred years, dating back to Billroth’s first successful gastrectomy in 1881 [5]. It was not until 1994 that we would first see minimally invasive techniques employed as a treatment for
123
Surg Endosc
gastric malignancy [6]. Twenty years later, minimally invasive techniques are now utilized in over 25 % of all gastric surgeries performed in Korea, where minimally invasive gastrectomy (MIG) has become the standard of care for early gastric cancer [7–10]. Asian centers have led the pursuit of published evidence on the technical feasibility and procedural safety of MIG. However, since the age of presentation, tumor pathogenesis, tumor location, and stage at time of diagnosis differs between the USA and much of Asia, and their extensive experience with gastric malignancy may not be immediately translatable to a Western patient population. Gastric cancer in Europe more closely resembles that which is encountered in the USA, in terms of its patient population, clinicopathological factors, and disease presentation [11–13]. There have been several prospective studies from Europe, reporting comparable MIG outcomes to Asian cohorts [12, 13]. Additionally, a recent multiinstitutional retrospective analysis in an American cohort notes similar results [3]. These reports indicate acceptable perioperative outcomes, but consist of relatively small patient samples (largest reporting 130 laparoscopic gastrectomies) [3, 12, 13]. Representative national data are necessary to adequately assess the feasibility and safety of MIG in the USA. National databases, such as the University HealthSystem Consortium (UHC), will be instrumental in this initial evaluation as they provide a large population for analysis. The aim of this study was to describe the utilization of MIG for malignancies in the USA. We also sought to examine the feasibility and safety of MIG by evaluating in-hospital outcomes, using the largest patient cohort to date of MIG performed in the USA.
Methods Database The UHC clinical database is comprised of 120 academic medical centers and 307 of their affiliated hospitals. Available data represents approximately 90 % of the nonprofit academic medical centers in the USA [14]. The database includes patient-level insurance claim data (UB92) and abstracted discharge data for the purposes of comparative data analysis between academic institutions [14]. All patients undergo risk adjustment, which involves the assignment of an All Payer Refined-Diagnosis-Related Group defined risk of mortality (ROM) and severity of illness (SOI) subclass [15]. These subclass designations are largely dependent on the patient’s underlying or primary diagnosis. The interactions of multiple illnesses are
123
accounted for with these subclass designations, which then recognize specific cohorts with unique challenges and potential worse outcomes [15]. Regression models provide postoperative outcome metrics based on normative patient populations. Patient selection Using the UHC database, patients who underwent gastrectomy between January 2008 and December 2013 were identified. Appropriate diagnosis and procedural codes as specified by the International Classification of Diseases, 9th-Edition Clinical Modification (ICD-9-CM) were identified. Principal diagnosis codes for malignant neoplasms of the abdominal viscera requiring gastrectomy were included. All partial, subtotal, and total gastrectomy procedure codes were assessed. Procedures that included a conversion to an open technique, as well as bariatric, antireflux, ulcer related, and other procedures for benign indications were excluded from the study population. Patients were classified into three operative groups: open, laparoscopic (including hand-assisted and combined laparoscopic/endoscopic), and robot assisted. Data on patient outcome metrics such as in-hospital complications, mean length of stay (LOS), 30-day readmission rate, total cost, and in-hospital mortality were collected. The UHC does not provide oncologic or long-term follow-up data [15]. Data analysis Continuous variables were evaluated using paired t tests. Categorical variables were analyzed with Chi-square tests as appropriate. Univariable and multivariable logistic regression models determined the association of relevant clinicopathological factors. All analyses were carried out with STATA version 13.0 (StataCorp, College Station, TX), and a P value of \0.05 was considered statistically significant.
Results From January 2008 to December 2013, a total of 9597 patients underwent gastrectomy at UHC academic medical centers in the USA (Table 1). Indications for gastrectomy included gastric (n = 8314; 86.6 %), esophageal (n = 1237; 12.9 %), and other abdominal (n = 46; 0.5 %) malignancies (Table 2). Of the 427 hospitals included in the UHC database, 207 hospitals recorded at least one gastrectomy within the study period. The ten hospitals with the highest gastrectomy volumes accounted for approximately 26 % of all procedures performed. From 2010 to 2011, there was a significant rise in the number of MIGs
Surg Endosc Table 1 Demographic information of patients who underwent gastrectomy at US academic medical centers, 2008–2013
Variable
Open
Laparoscopic
Robotic
All
Patients
8585
789
223
9597
Male
5899 (68.7)
499 (63.2)
153 (68.6)
6551 (68.33)
Female
2686 (31.3)
290 (36.8)
70 (31.4)
3046 (31.7)
Gender
P value
0.007
Age (y)
0.689
18–30
79 (0.9)
9 (1.1)
2 (0.9)
87 (0.9)
31–50
1119 (13.0)
114 (14.5)
25 (11.2)
1258 (13.1)
51–64
3038 (35.4)
264 (33.5)
87 (39.0)
3389 (35.3)
[65
4349 (50.7)
402 (51.0)
109 (48.9)
4860 (50.7)
Race
0.002
White
5941 (69.2)
500 (63.4)
163 (73.1)
6604 (68.8)
Black
1043 (12.2)
101 (12.8)
19 (8.5)
1163 (12.1)
Asian
526 (6.1)
83 (10.5)
13 (5.8)
622 (6.5)
Other
1075 (12.5)
105 (13.3)
28 (12.6)
1208 (12.6)
Risk of mortality 1 (Minor)
2909 (33.9)
312 (39.5)
76 (34.1)
3297 (34.4)
2 (Moderate)
3017 (35.1)
285 (36.1)
76 (34.1)
3378 (35.2)
3 (Major)
1775 (20.7)
132 (16.7)
36 (16.1)
1943 (20.3)
4 (Extreme)
884 (10.3)
60 (7.6)
35 (15.7)
979 (10.2)
1 (Minor)
31 (0.4)
13 (1.7)
0 (0)
44 (0.5)
2 (Moderate)
3433 (40.0)
376 (47.7)
78 (35.0)
3887 (40.5)
3 (Major)
3815 (44.4)
309 (39.2)
103 (46.2)
4227 (44.1)
4 (Extreme)
1306 (15.2)
91 (11.5)
42 (18.8)
1439 (15.0)
\0.001
\0.001
Severity of illness
Percentages noted in parentheses
performed; however, this trend plateaued over the final 3 years of the study period (Fig. 1). Gastrectomies were performed as open procedures 89.5 % of the time, whereas 8.2 % were performed laparoscopically, and 2.3 % were performed with robot-assistance. Males accounted for 68.3 % of study participants, and patients over 50-year old comprised 86.0 % of the cohort. Caucasians accounted for 68.8 % of patients, Blacks 12.2 %, and Asians 6.5 %. About 70 % of the cohort had a minor to moderate ROM, and about 85 % had a moderate to major SOI (Table 1). Overall complication rates, as well as the occurrence of pulmonary, cardiac, cerebrovascular, deep space infection (abscess, bacteremia, sepsis), bowel (obstruction, anastomotic failure), and wound-related events were similar between the three operative techniques, without statistically significant differences. The laparoscopic group had a higher incidence of superficial infection (7.6 %, compared to 6.0 % in the open group and 5.8 % in the robotic group, P = 0.013). Also, the robotic group had a higher incidence of venous/pulmonary thromboembolism (7.6 %, compared to 4.9 % in the open group and 5.3 % in the laparoscopic group, P = 0.038). There was a higher rate of acute renal failure in the robotic group; however, this difference did not reach statistical significance (P = 0.307). The robotic
group had the highest percentage of patients with an extreme SOI (robotic 18.8 %, open 15.2 %, laparoscopic 11.5 %, P \ 0.001) or ROM (robotic 15.7 %, open 10.3 %, laparoscopic 7.6 %, P \ 0.001) subclass designation. In-hospital mortality was 2.6 % for the open group, 1.9 % for the laparoscopic group, and 4.9 % for the robotic group; however, these differences were not significantly different (P = 0.064). Differences in LOS and 30-day readmission did not reach statistical significance (Table 3). The robotic group did have a higher mean total cost (54,345 USD, compared to approximately 41,000 USD for both the open and laparoscopic groups, P = 0.017). Univariable and multivariable logistic regression models were used to demonstrate significant predictors of adverse outcomes. Race, ROM, SOI, LOS, and mean total costs were found to be independently associated with an increased risk of complications. Gender and race were independently associated with increased risk of readmission. Age, ROM, SOI, LOS, mean total costs, and in-hospital complications were found to be independently associated with an increased risk of in-hospital mortality. The choice of surgical technique (open, laparoscopic, robotic assisted) was not associated with complications, readmission, or death (Table 4).
123
Surg Endosc Table 2 Indications for gastrectomy based on ICD-9 malignant principle diagnosis codes
ICD-9 category
Open
Laparoscopic
Robotic
All
Patients
8585 (89.5)
789 (8.2)
223 (2.3)
9597
Esophagus
1138 (92.0)
52 (4.2)
47 (3.8)
1237
Upper 1/3
3 (100)
0 (0)
0 (0)
Middle 1/3
3 (60.0)
1 (20.0)
1 (20.0)
Lower 1/3
1122 (92.1)
Other, unspecified site
50 (4.1)
46 (3.8)
5 1218
10 (90.9)
1 (9.1)
Stomach
7403 (89.0)
735 (8.8)
176 (2.1)
8314
Cardia
2777 (90.5)
199 (6.5)
91 (3.0)
3067
Pylorus/antrum
1117 (88.5)
129 (10.2)
16 (1.3)
1262
163 (81.1)
32 (15.9)
6 (3.0)
201
Fundus
0 (0)
3
11
Body
521 (86.0)
76 (12.5)
9 (1.5)
606
Lesser curvature
348 (87.9)
43 (10.9)
5 (1.3)
396
Greater curvature Other, unspecified site Other abdominal Small intestine Colon
155 (91.2)
12 (7.1)
3 (1.8)
170
2322 (88.9)
244 (9.3)
46 (1.8)
2612
44 (95.7) 4 (100)
2 (4.3) 0 (0)
0 (0) 0 (0)
46 4
17 (100)
17
0 (0)
0 (0)
Rectum
5 (83.3)
1 (16.7)
0 (0)
6
Appendix/peritoneum
3 (100)
0 (0)
0 (0)
3
Retroperitoneum
6 (100)
0 (0)
0 (0)
6
Other, unspecified sue
9 (90.0)
1 (10.0)
0 (0)
10
Percentages noted in parentheses
Fig. 1 Gastrectomies performed at US academic medical centers, 2008–2013
Discussion There have been multiple large prospective studies examining MIG performed in countries such as Japan, Korea, and China, where the incidence of gastric cancer reaches 60 to 80 cases per 100,000 people [2, 4, 7–10, 16–20]. Asian patients with gastric cancer tend to present at a younger age, have a lower BMI, have less comorbidity, and have distal tumors that are discovered at a relatively early
123
stage [2, 4, 21]. This is likely attributable in part to Helicobacter pylori eradication and atrophic gastritis screening programs that have been established in Asia [2, 22]. Tumor biology, genetic predisposition, environment, and dietary factors have also been implicated in this discrepancy [2, 22]. Conversely, gastric cancer in the USA tends to present later in life (mean age of 69 years), at a more advanced stage, and in a relatively unhealthy patient population. Due to these clinicopathological dissimilarities, we must also
Surg Endosc Table 3 Outcomes of patients who underwent gastrectomy at US academic medical centers, 2008–2013
Variable
Open
Laparoscopic
Robotic
P value
Patients
8585 (89.5)
789 (8.2)
223 (2.3)
Overall complication
2719 (31.7)
219 (27.8)
75 (33.0)
0.13
Pulmonary
77 (0.9)
2 (0.3)
3 (1.3)
0.257
Cardiac
485 (5.7)
39 (4.9)
12 (5.4)
0.835
CVA
91 (1.1)
8 (1.0)
1 (0.4)
0.795
VTE/PE
419 (4.9)
42 (5.3)
17 (7.6)
0.038
Renal failure
604 (7.0)
44 (5.6)
29 (13.0)
0.307
Bowl related
404 (4.7)
37 (4.7)
8 (3.6)
0.838
Wound dehiscence
158 (1.8)
12 (1.5)
3 (1.3)
0.62
Superficial infection
512 (6.0)
60 (7.6)
13 (5.8)
0.013
Intra-abdominal abscess
158 (1.8)
16 (2.0)
3 (1.3)
0.652
Bacteremia/sepsis
562 (6.5)
40 (5.1)
21 (9.4)
0.975
Severe sepsis/shock
436 (5.1)
31 (3.9)
20 (9.0)
0.562
Mean length of stay (days)
13.0 ± 12.0
12.1 ± 12.8
12.2 ± 11.2
0.085
30-day readmission In-hospital mortality
1640 (19.1) 225 (2.6)
146 (18.5) 15 (1.9)
43 (19.3) 11 (4.9)
0.916 0.064
Mean rural cost (USD)
41,326
40,633
54,345
0.017
Percentages noted in parentheses CVA cerebrovascular accident, VTE/PE venous of pulmonary thromboembolism
Table 4 Multivariate analysis of gastrectomies performed at US academic medical centers Complications
Readmission
Age
X
Gender Race
Death
X X
X
Type of surgery ROM
X
X
SOI
X
X
LOS
X
X
Total cost
X
X
Complications Death
X X
Boxes with an ‘‘X’’ indicate statistically significant associations, P \ 0.05 ROM risk of mortality, SOI severity of illness, LOS mean length of stay
study gastric resections for cancer in the US population to determine whether conclusions from Asian studies are generalizable [3, 11]. The current study suggests that MIG may be underutilized in the USA, despite a complication rate, LOS, readmission, and mortality similar to that of open resection. Gastric cancer in the USA occurs with a frequency of 7–10 cases per 100,000 people [1, 3, 22]. The current study suggests that US academic medical centers’ use of MIG accounts for only 10.5 % of resections for cancer,
compared with up to 25 % in Eastern nations [2, 7, 17]. The overall utilization of minimally invasive techniques in the USA increased during the study period (Fig. 1). However, over the last 3 years of the study period, a plateau occurred at approximately 250 cases per year. This would suggest that despite increasing national interest in MIG for cancer, it has not been widely implemented. This pattern of delayed acceptance was similar to that of other minimally invasive oncologic procedures in the early 1990s, such as laparoscopic colectomy for colorectal cancer [23–25]. Initially reported in 1991, the outcomes for laparoscopic colectomy were initially debated [26]. However, due to a unified national commitment by surgeons and a high prevalence of colorectal cancer in the USA, multiple prospective randomized controlled trials were able to be performed, demonstrating comparable outcomes for laparoscopic and open colectomies [23–25]. Using the US Nationwide Inpatient Sample database, Juo et al. reported that from 2008 to 2010, greater than 244,000 colectomies were performed, of which 47.6 % were performed laparoscopically [25]. This was a dramatic increase in the utilization of minimally invasive techniques from just a few years prior, when from 2005 to 2007, of the 240,000 colectomies performed for cancer, only 5.7 % were laparoscopic [27]. This rapid acceleration in utilization of laparoscopic techniques may suggest that MIG for cancer may also experience an upward trend, once technical feasibility and procedural safety have been demonstrated through focused investigation.
123
Surg Endosc
Multiple studies suggest that there are no differences in morbidity or mortality between open and MIG for cancer after a learning curve of 30–60 laparoscopic gastrectomies has been satisfied [8, 16, 28]. Due to the relative paucity of gastric cancer in the USA, it may be hard for some surgeons to reach 30–60 laparoscopic gastrectomies. The question and challenge then becomes, how should we satisfy this learning curve, when except for at high-volume institutions, surgeons typically do not encounter this volume of gastric cancer during their career. The answer may lie in the volume of minimally invasive procedures for benign gastric pathology [29–32]. A study evaluating Agency for Healthcare Research and Quality (AHRQ) data reports that from 1998 to 2008, the annual number of bariatric procedures performed in the USA increased from 12,775 to 124,838. The percentage of these that were performed laparoscopically also increased from 2.1 to 90.2 % [29]. As the study period progressed, patients were noted to be older and have more comorbidity; however, the length of stay and in-hospital mortality significantly decreased [29]. Similar to bariatric surgery, the annual number of minimally invasive anti-reflux procedures performed in the USA increased from 9000 in 1993 to [32,000 by 2000 [31, 32]. As a result, there was also a rise in the number of redo operations, necessitating the development of a complex technical acumen for laparoscopic foregut surgery. It is then possible that with gained experience in foregut surgery, surgeons are acquiring the necessary skills to consider minimally invasive approaches for oncologic indications. Alternatively, as was observed in early laparoscopic colectomy experience, there is opportunity for surgeons to be mentored by more experienced laparoscopic surgeons in order to combine technical and oncological expertise [33, 34]. Cost is a frequent criticism of minimally invasive surgery. Data regarding the differences in total cost for the surgical techniques were analyzed. Mean total cost was approximately 41,000 USD for both the open and laparoscopic groups, but was significantly higher in the robotic group at approximately 54,000 USD (P = 0.017). Several studies assessing other surgical procedures have reported that higher costs incurred with longer minimally invasive operative time and more expensive instruments may be somewhat negated by a decreased complication rate and LOS [3, 19, 20, 35, 36]. The current study would suggest significant room for improvement with MIG, as neither LOS nor overall complication rates were significantly different between the groups, yet the total costs were similar for open and laparoscopic procedures. As MIG experience increases and we progress up the laparoscopic foregut learning curve, we should also experience shorter LOS and less complications, yielding lower overall costs [29, 30].
123
Data analysis demonstrates that the choice of surgical technique does not have any independent associations with complications, readmission, or death (Table 4). This suggests the relative safety of minimally invasive techniques for gastrectomy in the USA. However, the concept of procedural safety requires the analysis of a complex interaction of multiple variables. Surgeon experience, operative time, estimated blood loss, extent of lymphadenectomy, and methods of reconstruction are not accounted for in the UHC database; however, all of these may affect a comprehensive evaluation of procedural safety. Future studies should seek to comprehensively assess these clinicopathological factors using large nationally representative clinical databases to be able to make further conclusions about the relative safety of minimally invasive procedures for gastric malignancy. In-hospital outcome metrics play important roles in the determination of the success of a particular procedure. We found that LOS, 30-day readmission, and in-hospital mortality were similar between the three techniques (Table 3). The 30-day readmission and mortality rates correspond with multiple prospective randomized trials from Asia [8–10, 17, 18, 20]. MIG LOS is reported to be less than open gastrectomy LOS in other studies; however, differences in LOS for this cohort did not reach significance. As was noted in the laparoscopic bariatric literature, our LOS for MIG in the USA may continue to decrease as experience with these techniques improves [29, 30, 37]. Overall complication rates for our cohort were very similar between the three groups. The increased incidence of superficial infection in the laparoscopic group and venous thromboembolism in the robotic group may be attributable to presumed longer operative times, although this variable is not collected by the UHC database. As another possible explanation, in the laparoscopic colectomy literature, venous thromboembolism has been shown to be independently associated with obesity and increasing age [38]. This may be extrapolated to our gastrectomy data, as gastric cancer patients in the USA tend to be older and have a higher body mass index when compared to Asian cohorts, which have not previously demonstrated differences in thromboembolic rates [1–4]. This study has several limitations. The UHC database is comprised of abstracted discharge data that is limited to inhospital events without outpatient follow-up. As such, no information is available to identify confounding factors that might necessitate an open operation (i.e., previous abdominal operations, locally advanced cancer, concurrent procedures, operative take backs, emergencies, or palliative procedures). Likewise, factors relevant to short-term outcomes (i.e., surgeon experience/training, extent of disease, and nutritional status) and long-term outcomes (i.e., extent
Surg Endosc
of lymphadenectomy, chemotherapy, and radiation) are not available. Additionally, the coding of certain complications may be inaccurate, as postoperative adverse events may be coded differently [39, 40]. Despite these limitations, the UHC database has been utilized to evaluate multiple other surgical procedures and serves the purposes of assessing current ‘‘status of care’’ and perioperative safety-related data [37, 41–43].
Conclusion Despite adoption for benign indications, minimally invasive gastrectomy for cancer remains underutilized in the USA. Within the context of UHC database, laparoscopic and robot-assisted gastrectomies appear to be comparable to open resection with respect to overall complications, LOS, 30-day readmission, and in-hospital mortality. Despite increasing data from Europe and Asia describing minimally invasive techniques as being analogous to open resection, US practice habits remained largely unchanged over the last 3 years of the study period. Similar to what has occurred with gastrectomies in Asia and with colectomies in the USA, as the incidence of gastric cancer continues to rise and interest in MIG grows, prospective randomized controlled trials will be required to further evaluate the feasibility, safety, oncologic equivalency, and long-term outcomes of MIG in the USA. Lastly, when performing MIG for cancer, infectious and thromboembolic risk reduction strategies should be emphasized in the operative and postoperative periods. Disclosures Jason Glenn, Kiran Turaga, T. Clark Gamblin, Samuel Hohmann, and Fabian Johnston have no financial or commercial relationships to disclose that might pose a conflict of interest.
References 1. SEER cancer statistics factsheets: stomach cancer. National Cancer Institute, Bethesda, MD, 2013. http://seer.cancer.gov/ statfacts/html/stomach.html 2. Inoue M, Tsugane S (2005) Epidemiology of gastric cancer in Japan. Postgrad Med J 81(957):419–424 3. Spolverato G, Kim Y, Ejaz A, Valero V, 3rd, Squires MH, Poultsides G, et al (2014) A Multi-institutional analysis of open versus minimally-invasive surgery for gastric adenocarcinoma: results of the US gastric cancer collaborative. J Gastrointest Surg. PubMed PMID: 24912915 4. Jung KW, Won YJ, Kong HJ, Oh CM, Seo HG (2013) Lee JS (2013) Prediction of cancer incidence and mortality in Korea. Cancer Res Treat 45(1):15–21 5. de Steur WO, Dikken JL, Hartgrink HH (2013) Lymph node dissection in resectable advanced gastric cancer. Dig Surg 30(2):96–103 6. Kitano S, Iso Y, Moriyama M, Sugimachi K (1994) Laparoscopyassisted Billroth I gastrectomy. Surg Laparosc Endosc 4(2):146–148
7. Kim HH, Han SU, Kim MC, Hyung WJ, Kim W, Lee HJ et al (2013) Prospective randomized controlled trial (phase III) to comparing laparoscopic distal gastrectomy with open distal gastrectomy for gastric adenocarcinoma (KLASS 01). J Korean Surg Soc 84(2):123–130 8. Kim HH, Han SU, Kim MC, Hyung WJ, Kim W, Lee HJ et al (2014) Long-term results of laparoscopic gastrectomy for gastric cancer: a large-scale case-control and case-matched Korean multicenter study. J Clin Oncol 32(7):627–633 9. Hayashi H, Ochiai T, Shimada H, Gunji Y (2005) Prospective randomized study of open versus laparoscopy-assisted distal gastrectomy with extraperigastric lymph node dissection for early gastric cancer. Surg Endosc 19(9):1172–1176 10. Nakamura K, Katai H, Mizusawa J, Yoshikawa T, Ando M, Terashima M et al (2013) A phase III study of laparoscopyassisted versus open distal gastrectomy with nodal dissection for clinical stage IA/IB gastric Cancer (JCOG0912). Jpn J Clin Oncol 43(3):324–327 11. Markar SR, Karthikesalingam A, Jackson D, Hanna GB (2013) Long-term survival after gastrectomy for cancer in randomized, controlled oncological trials: comparison between West and East. Ann Surg Oncol 20(7):2328–2338 12. Azagra JS, Ibanez-Aguirre JF, Goergen M, Ceuterick M, BordasRivas JM, Almendral-Lopez ML et al (2006) Long-term results of laparoscopic extended surgery in advanced gastric cancer: a series of 101 patients. Hepatogastroenterology 53(68):304–308 13. Iba´n˜ez-Aguirre FJ, Azagra JS, Erro-Azcarate ML (2006) Laparoscopic gastrectomy for gastric adenocarcinoma. Long-term results. REV ESP ENFERM DIG (Madrid) 98(7):491–500 14. Reavis KM (2009) Hospital volume is not a predictor of outcomes after gastrectomy for neoplasm. Am Surg 75(10):932–936 15. Meurer SJ (2009) Mortality Measurement: mortality risk adjustment methodology for University Health System Consortium. Agency for Healthcare Research and Quality, Rockville, MD. http://archive.ahrq.gov/professionals/quality-patient-safety/qualityresources/tools/mortality/Meurer.html 16. Chen K, Xu XW, Mou YP, Pan Y, Zhou YC, Zhang RC, et al (2013) Systematic review and meta-analysis of laparoscopic and open gastrectomy for advanced gastric cancer. World J Surg Oncol 11(182):12 17. Hosono S (2006) Meta-analysis of short-term outcomes after laparoscopy-assisted distal gastrectomy. R World J Gastroenterol 12(47):7676–7683 18. Lee JH, Han HS, Lee JH (2005) A prospective randomized study comparing open vs laparoscopy-assisted distal gastrectomy in early gastric cancer: early results. Surg Endosc 19(2):168–173 19. Mochiki E, Toyomasu Y, Ogata K, Andoh H, Ohno T, Aihara R et al (2008) Laparoscopically assisted total gastrectomy with lymph node dissection for upper and middle gastric cancer. Surg Endosc 22(9):1997–2002 20. Ohtani H, Tamamori Y, Noguchi K, Azuma T, Fujimoto S, Oba H et al (2011) Meta-analysis of laparoscopy-assisted and open distal gastrectomy for gastric cancer. J Surg Res 171(2):479–485 21. Strong VE, Devaud N, Karpeh M (2009) The role of laparoscopy for gastric surgery in the West. Gastric Cancer 12(3):127–131 22. Wong J, Jackson P (2011) Gastric cancer surgery: an American perspective on the current options and standards. Curr Treat Options Oncol 12(1):72–84 23. Nelson H, Sargent Daniel J, Sam Wieand H, Fleshman James, Anvari Mehran, Stryker Steven J et al (2004) A comparison of laparoscopically assisted and open colectomy for colon cancer. N Engl J Med 350(20):2050–2059 24. Lacy AM, Garcı´a-Valdecasas JC, Delgado S, Castells A, Taura´ P, Pique´ JM et al (2002) Laparoscopy-assisted colectomy versus open colectomy for treatment of non-metastatic colon cancer: a randomised trial. Lancet 359(9325):2224–2229
123
Surg Endosc 25. Juo YY, Hyder O, Haider AH, Camp M, Lidor A, Ahuja N (2014) Is minimally invasive colon resection better than traditional approaches?: first comprehensive national examination with propensity score matching. JAMA Surg 149(2):177–184 26. Cooperman AM, Katz V, Zimmon D, Botero G (1991) Laparoscopic colon resection: a case report. J Laparoendosc Surg 1(4):221–224 27. Robinson C, Chen GJ, Balentine C, Sansgiry S, Marshall C, Anaya D et al (2011) Minimally invasive surgery is underutilized for colon cancer. Ann Surg Oncol 18(5):1412–1418 (English) 28. El-Sedfy A, Brar SS, Coburn NG (2014) Current role of minimally invasive approaches in the treatment of early gastric cancer. World J Gastroenterol 20(14):3880–3888 29. Nguyen NT, Masoomi H, Magno CP, Nguyen XM, Laugenour K, Lane J (2011) Trends in use of bariatric surgery, 2003–2008. J Am Coll Surg 213(2):261–266 30. Dumon KR, Murayama KM (2011) Bariatric surgery outcomes. Surg Clin North Am 91(6):1313–1338 31. Carlson MA, Frantzides CT (2001) Complications and results of primary minimally invasive antireflux procedures: a review of 10,735 reported cases. J Am Coll Surg 193(4):428–439 32. Kim D, Velanovich V (2014) Surgical treatment of GERD: Where have we been and where are we going? Gastroenterol Clin North Am 43(1):135–145 33. Schlachta CM, Lefebvre KL, Sorsdahl AK, Jayaraman S (2010) Mentoring and telementoring leads to effective incorporation of laparoscopic colon surgery. Surg Endosc 24(4):841–844 34. Sebajang H, Trudeau P, Dougall A, Hegge S, McKinley C, Anvari M (2006) The role of telementoring and telerobotic assistance in the provision of laparoscopic colorectal surgery in rural areas. Surg Endosc 20(9):1389–1393
123
35. Huscher CGS, Mingoli A, Sgarzini G, Sansonetti A, Di Paola M, Recher A et al (2005) Laparoscopic versus open subtotal gastrectomy for distal gastric cancer. Ann Surg 241(2):232–237 36. Xiong B, Ma L, Zhang C (2012) Robotic versus laparoscopic gastrectomy for gastric cancer: a meta-analysis of short outcomes. Surg Oncol 21(4):274–280 37. Tiwari MM, Reynoso JF, High R, Tsang AW, Oleynikov D (2011) Safety, efficacy, and cost-effectiveness of common laparoscopic procedures. Surg Endosc 25(4):1127–1135 38. Becattini C, Rondelli F, Vedovati MC, Camporese G, Giustozzi M, Boncompagni M, et al (2004) Incidence and risk factors for venous thromboembolism after laparoscopic surgery for colorectal cancer. Haematologica. PubMed PMID: 25216684 39. Bruce J, Russell EM, Mollison J, Krukowski ZH (2001) The measurement and monitoring of surgical adverse events. Health Technol Assess (Winchester, England) 5(22):1–194 40. Davenport DL, Holsapple CW, Conigliaro J (2009) Assessing surgical quality using administrative and clinical data sets: a direct comparison of the University HealthSystem Consortium Clinical Database and the National Surgical Quality Improvement Program data set. Am J Med Qual 24(5):395–402 41. Cauble S, Abbas A, Balart L, Bazzano L, Medvedev S, Shores N (2013) United States women receive more curative treatment for hepatocellular carcinoma than men. Dig Dis Sci 58(10):2817–2825 42. Jalisi S, Bearelly S, Abdillahi A, Truong MT (2013) Outcomes in head and neck oncologic surgery at academic medical centers in the United States. Laryngoscope 123(3):689–698 43. Stack BC Jr, Spencer H, Moore E, Medvedev S, Bodenner D (2012) Outpatient parathyroid surgery data from the University Health System Consortium. Otolaryngol Head Neck Surg 147(3):438–443
and Other Interventional Techniques
Minimally invasive gastrectomy for cancer: current utilization in US academic medical centers Jason A. Glenn1 • Kiran K. Turaga1 • T. Clark Gamblin1 • Samuel F. Hohmann2 Fabian M. Johnston1
•
Received: 3 December 2014 / Accepted: 2 March 2015 Ó Springer Science+Business Media New York 2015
Abstract Background Internationally, the utilization of minimally invasive techniques for gastric cancer resection has been increasing since first introduced in 1994. In the USA, the feasibility and safety of these techniques for cancer have not yet been demonstrated. Methods The University HealthSystem Consortium database was queried for gastrectomies performed between 2008 and 2013. Any adult patient with an abdominal visceral malignancy that necessitated gastric resection was included in the cohort. Clinicopathological and in-hospital outcome metrics were collected for open, laparoscopic, and robotic procedures. Results Open gastrectomies comprised 89.5 % of the total study group, while 8.2 % of procedures were performed laparoscopically, and 2.3 % were performed with robotic assistance. When accounting for disparities in patient severity of illness and risk of mortality subclass designations, there were no significant differences in mean length of stay, 30-day readmission, and in-hospital mortality between the three groups; however, mean total cost was highest in the robotic-assisted group (P = 0.017). Overall, complication rates were also similar; however, there was a higher incidence of superficial infection in the laparoscopic
& Fabian M. Johnston [email protected] Jason A. Glenn [email protected] 1
Division of Surgical Oncology, Medical College of Wisconsin, 9200 W. Wisconsin Ave, Milwaukee, WI 53226, USA
2
University HealthSystem Consortium, 155 N Upper Wacker Dr, Chicago, IL 60606, USA
group (P = 0.013) and a higher incidence of venous thromboembolism in the robotic group (P = 0.038). Conclusion Despite widespread adoption for benign indications, minimally invasive gastrectomy for cancer remains underutilized in the USA. In these patients, laparoscopic and robot-assisted gastrectomies appear to be comparable to open resection with respect to overall complications, length of stay, 30-day readmission, and inhospital mortality. However, when employing minimally invasive techniques, infection and thromboembolism risk reduction strategies should be emphasized in the operative and postoperative periods. Keywords Surgical
GI Cancer Digestive Human/robotic
Background The National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program estimates that over twenty-two thousand new gastric cancers and almost eleven thousand gastric cancer-related deaths will occur in the USA in 2014 [1]. While this incidence is almost tenfold less than what is reported in Asia, the relative late stage of presentation and poorer prognosis makes gastric cancer in the USA an important topic of discussion [1–4]. Even though the role of surgical resection for gastric cancer has been well established, the optimal operative approach remains less well defined. The fundamental surgical approach to gastric pathology has remained largely unchanged for over a hundred years, dating back to Billroth’s first successful gastrectomy in 1881 [5]. It was not until 1994 that we would first see minimally invasive techniques employed as a treatment for
123
Surg Endosc
gastric malignancy [6]. Twenty years later, minimally invasive techniques are now utilized in over 25 % of all gastric surgeries performed in Korea, where minimally invasive gastrectomy (MIG) has become the standard of care for early gastric cancer [7–10]. Asian centers have led the pursuit of published evidence on the technical feasibility and procedural safety of MIG. However, since the age of presentation, tumor pathogenesis, tumor location, and stage at time of diagnosis differs between the USA and much of Asia, and their extensive experience with gastric malignancy may not be immediately translatable to a Western patient population. Gastric cancer in Europe more closely resembles that which is encountered in the USA, in terms of its patient population, clinicopathological factors, and disease presentation [11–13]. There have been several prospective studies from Europe, reporting comparable MIG outcomes to Asian cohorts [12, 13]. Additionally, a recent multiinstitutional retrospective analysis in an American cohort notes similar results [3]. These reports indicate acceptable perioperative outcomes, but consist of relatively small patient samples (largest reporting 130 laparoscopic gastrectomies) [3, 12, 13]. Representative national data are necessary to adequately assess the feasibility and safety of MIG in the USA. National databases, such as the University HealthSystem Consortium (UHC), will be instrumental in this initial evaluation as they provide a large population for analysis. The aim of this study was to describe the utilization of MIG for malignancies in the USA. We also sought to examine the feasibility and safety of MIG by evaluating in-hospital outcomes, using the largest patient cohort to date of MIG performed in the USA.
Methods Database The UHC clinical database is comprised of 120 academic medical centers and 307 of their affiliated hospitals. Available data represents approximately 90 % of the nonprofit academic medical centers in the USA [14]. The database includes patient-level insurance claim data (UB92) and abstracted discharge data for the purposes of comparative data analysis between academic institutions [14]. All patients undergo risk adjustment, which involves the assignment of an All Payer Refined-Diagnosis-Related Group defined risk of mortality (ROM) and severity of illness (SOI) subclass [15]. These subclass designations are largely dependent on the patient’s underlying or primary diagnosis. The interactions of multiple illnesses are
123
accounted for with these subclass designations, which then recognize specific cohorts with unique challenges and potential worse outcomes [15]. Regression models provide postoperative outcome metrics based on normative patient populations. Patient selection Using the UHC database, patients who underwent gastrectomy between January 2008 and December 2013 were identified. Appropriate diagnosis and procedural codes as specified by the International Classification of Diseases, 9th-Edition Clinical Modification (ICD-9-CM) were identified. Principal diagnosis codes for malignant neoplasms of the abdominal viscera requiring gastrectomy were included. All partial, subtotal, and total gastrectomy procedure codes were assessed. Procedures that included a conversion to an open technique, as well as bariatric, antireflux, ulcer related, and other procedures for benign indications were excluded from the study population. Patients were classified into three operative groups: open, laparoscopic (including hand-assisted and combined laparoscopic/endoscopic), and robot assisted. Data on patient outcome metrics such as in-hospital complications, mean length of stay (LOS), 30-day readmission rate, total cost, and in-hospital mortality were collected. The UHC does not provide oncologic or long-term follow-up data [15]. Data analysis Continuous variables were evaluated using paired t tests. Categorical variables were analyzed with Chi-square tests as appropriate. Univariable and multivariable logistic regression models determined the association of relevant clinicopathological factors. All analyses were carried out with STATA version 13.0 (StataCorp, College Station, TX), and a P value of \0.05 was considered statistically significant.
Results From January 2008 to December 2013, a total of 9597 patients underwent gastrectomy at UHC academic medical centers in the USA (Table 1). Indications for gastrectomy included gastric (n = 8314; 86.6 %), esophageal (n = 1237; 12.9 %), and other abdominal (n = 46; 0.5 %) malignancies (Table 2). Of the 427 hospitals included in the UHC database, 207 hospitals recorded at least one gastrectomy within the study period. The ten hospitals with the highest gastrectomy volumes accounted for approximately 26 % of all procedures performed. From 2010 to 2011, there was a significant rise in the number of MIGs
Surg Endosc Table 1 Demographic information of patients who underwent gastrectomy at US academic medical centers, 2008–2013
Variable
Open
Laparoscopic
Robotic
All
Patients
8585
789
223
9597
Male
5899 (68.7)
499 (63.2)
153 (68.6)
6551 (68.33)
Female
2686 (31.3)
290 (36.8)
70 (31.4)
3046 (31.7)
Gender
P value
0.007
Age (y)
0.689
18–30
79 (0.9)
9 (1.1)
2 (0.9)
87 (0.9)
31–50
1119 (13.0)
114 (14.5)
25 (11.2)
1258 (13.1)
51–64
3038 (35.4)
264 (33.5)
87 (39.0)
3389 (35.3)
[65
4349 (50.7)
402 (51.0)
109 (48.9)
4860 (50.7)
Race
0.002
White
5941 (69.2)
500 (63.4)
163 (73.1)
6604 (68.8)
Black
1043 (12.2)
101 (12.8)
19 (8.5)
1163 (12.1)
Asian
526 (6.1)
83 (10.5)
13 (5.8)
622 (6.5)
Other
1075 (12.5)
105 (13.3)
28 (12.6)
1208 (12.6)
Risk of mortality 1 (Minor)
2909 (33.9)
312 (39.5)
76 (34.1)
3297 (34.4)
2 (Moderate)
3017 (35.1)
285 (36.1)
76 (34.1)
3378 (35.2)
3 (Major)
1775 (20.7)
132 (16.7)
36 (16.1)
1943 (20.3)
4 (Extreme)
884 (10.3)
60 (7.6)
35 (15.7)
979 (10.2)
1 (Minor)
31 (0.4)
13 (1.7)
0 (0)
44 (0.5)
2 (Moderate)
3433 (40.0)
376 (47.7)
78 (35.0)
3887 (40.5)
3 (Major)
3815 (44.4)
309 (39.2)
103 (46.2)
4227 (44.1)
4 (Extreme)
1306 (15.2)
91 (11.5)
42 (18.8)
1439 (15.0)
\0.001
\0.001
Severity of illness
Percentages noted in parentheses
performed; however, this trend plateaued over the final 3 years of the study period (Fig. 1). Gastrectomies were performed as open procedures 89.5 % of the time, whereas 8.2 % were performed laparoscopically, and 2.3 % were performed with robot-assistance. Males accounted for 68.3 % of study participants, and patients over 50-year old comprised 86.0 % of the cohort. Caucasians accounted for 68.8 % of patients, Blacks 12.2 %, and Asians 6.5 %. About 70 % of the cohort had a minor to moderate ROM, and about 85 % had a moderate to major SOI (Table 1). Overall complication rates, as well as the occurrence of pulmonary, cardiac, cerebrovascular, deep space infection (abscess, bacteremia, sepsis), bowel (obstruction, anastomotic failure), and wound-related events were similar between the three operative techniques, without statistically significant differences. The laparoscopic group had a higher incidence of superficial infection (7.6 %, compared to 6.0 % in the open group and 5.8 % in the robotic group, P = 0.013). Also, the robotic group had a higher incidence of venous/pulmonary thromboembolism (7.6 %, compared to 4.9 % in the open group and 5.3 % in the laparoscopic group, P = 0.038). There was a higher rate of acute renal failure in the robotic group; however, this difference did not reach statistical significance (P = 0.307). The robotic
group had the highest percentage of patients with an extreme SOI (robotic 18.8 %, open 15.2 %, laparoscopic 11.5 %, P \ 0.001) or ROM (robotic 15.7 %, open 10.3 %, laparoscopic 7.6 %, P \ 0.001) subclass designation. In-hospital mortality was 2.6 % for the open group, 1.9 % for the laparoscopic group, and 4.9 % for the robotic group; however, these differences were not significantly different (P = 0.064). Differences in LOS and 30-day readmission did not reach statistical significance (Table 3). The robotic group did have a higher mean total cost (54,345 USD, compared to approximately 41,000 USD for both the open and laparoscopic groups, P = 0.017). Univariable and multivariable logistic regression models were used to demonstrate significant predictors of adverse outcomes. Race, ROM, SOI, LOS, and mean total costs were found to be independently associated with an increased risk of complications. Gender and race were independently associated with increased risk of readmission. Age, ROM, SOI, LOS, mean total costs, and in-hospital complications were found to be independently associated with an increased risk of in-hospital mortality. The choice of surgical technique (open, laparoscopic, robotic assisted) was not associated with complications, readmission, or death (Table 4).
123
Surg Endosc Table 2 Indications for gastrectomy based on ICD-9 malignant principle diagnosis codes
ICD-9 category
Open
Laparoscopic
Robotic
All
Patients
8585 (89.5)
789 (8.2)
223 (2.3)
9597
Esophagus
1138 (92.0)
52 (4.2)
47 (3.8)
1237
Upper 1/3
3 (100)
0 (0)
0 (0)
Middle 1/3
3 (60.0)
1 (20.0)
1 (20.0)
Lower 1/3
1122 (92.1)
Other, unspecified site
50 (4.1)
46 (3.8)
5 1218
10 (90.9)
1 (9.1)
Stomach
7403 (89.0)
735 (8.8)
176 (2.1)
8314
Cardia
2777 (90.5)
199 (6.5)
91 (3.0)
3067
Pylorus/antrum
1117 (88.5)
129 (10.2)
16 (1.3)
1262
163 (81.1)
32 (15.9)
6 (3.0)
201
Fundus
0 (0)
3
11
Body
521 (86.0)
76 (12.5)
9 (1.5)
606
Lesser curvature
348 (87.9)
43 (10.9)
5 (1.3)
396
Greater curvature Other, unspecified site Other abdominal Small intestine Colon
155 (91.2)
12 (7.1)
3 (1.8)
170
2322 (88.9)
244 (9.3)
46 (1.8)
2612
44 (95.7) 4 (100)
2 (4.3) 0 (0)
0 (0) 0 (0)
46 4
17 (100)
17
0 (0)
0 (0)
Rectum
5 (83.3)
1 (16.7)
0 (0)
6
Appendix/peritoneum
3 (100)
0 (0)
0 (0)
3
Retroperitoneum
6 (100)
0 (0)
0 (0)
6
Other, unspecified sue
9 (90.0)
1 (10.0)
0 (0)
10
Percentages noted in parentheses
Fig. 1 Gastrectomies performed at US academic medical centers, 2008–2013
Discussion There have been multiple large prospective studies examining MIG performed in countries such as Japan, Korea, and China, where the incidence of gastric cancer reaches 60 to 80 cases per 100,000 people [2, 4, 7–10, 16–20]. Asian patients with gastric cancer tend to present at a younger age, have a lower BMI, have less comorbidity, and have distal tumors that are discovered at a relatively early
123
stage [2, 4, 21]. This is likely attributable in part to Helicobacter pylori eradication and atrophic gastritis screening programs that have been established in Asia [2, 22]. Tumor biology, genetic predisposition, environment, and dietary factors have also been implicated in this discrepancy [2, 22]. Conversely, gastric cancer in the USA tends to present later in life (mean age of 69 years), at a more advanced stage, and in a relatively unhealthy patient population. Due to these clinicopathological dissimilarities, we must also
Surg Endosc Table 3 Outcomes of patients who underwent gastrectomy at US academic medical centers, 2008–2013
Variable
Open
Laparoscopic
Robotic
P value
Patients
8585 (89.5)
789 (8.2)
223 (2.3)
Overall complication
2719 (31.7)
219 (27.8)
75 (33.0)
0.13
Pulmonary
77 (0.9)
2 (0.3)
3 (1.3)
0.257
Cardiac
485 (5.7)
39 (4.9)
12 (5.4)
0.835
CVA
91 (1.1)
8 (1.0)
1 (0.4)
0.795
VTE/PE
419 (4.9)
42 (5.3)
17 (7.6)
0.038
Renal failure
604 (7.0)
44 (5.6)
29 (13.0)
0.307
Bowl related
404 (4.7)
37 (4.7)
8 (3.6)
0.838
Wound dehiscence
158 (1.8)
12 (1.5)
3 (1.3)
0.62
Superficial infection
512 (6.0)
60 (7.6)
13 (5.8)
0.013
Intra-abdominal abscess
158 (1.8)
16 (2.0)
3 (1.3)
0.652
Bacteremia/sepsis
562 (6.5)
40 (5.1)
21 (9.4)
0.975
Severe sepsis/shock
436 (5.1)
31 (3.9)
20 (9.0)
0.562
Mean length of stay (days)
13.0 ± 12.0
12.1 ± 12.8
12.2 ± 11.2
0.085
30-day readmission In-hospital mortality
1640 (19.1) 225 (2.6)
146 (18.5) 15 (1.9)
43 (19.3) 11 (4.9)
0.916 0.064
Mean rural cost (USD)
41,326
40,633
54,345
0.017
Percentages noted in parentheses CVA cerebrovascular accident, VTE/PE venous of pulmonary thromboembolism
Table 4 Multivariate analysis of gastrectomies performed at US academic medical centers Complications
Readmission
Age
X
Gender Race
Death
X X
X
Type of surgery ROM
X
X
SOI
X
X
LOS
X
X
Total cost
X
X
Complications Death
X X
Boxes with an ‘‘X’’ indicate statistically significant associations, P \ 0.05 ROM risk of mortality, SOI severity of illness, LOS mean length of stay
study gastric resections for cancer in the US population to determine whether conclusions from Asian studies are generalizable [3, 11]. The current study suggests that MIG may be underutilized in the USA, despite a complication rate, LOS, readmission, and mortality similar to that of open resection. Gastric cancer in the USA occurs with a frequency of 7–10 cases per 100,000 people [1, 3, 22]. The current study suggests that US academic medical centers’ use of MIG accounts for only 10.5 % of resections for cancer,
compared with up to 25 % in Eastern nations [2, 7, 17]. The overall utilization of minimally invasive techniques in the USA increased during the study period (Fig. 1). However, over the last 3 years of the study period, a plateau occurred at approximately 250 cases per year. This would suggest that despite increasing national interest in MIG for cancer, it has not been widely implemented. This pattern of delayed acceptance was similar to that of other minimally invasive oncologic procedures in the early 1990s, such as laparoscopic colectomy for colorectal cancer [23–25]. Initially reported in 1991, the outcomes for laparoscopic colectomy were initially debated [26]. However, due to a unified national commitment by surgeons and a high prevalence of colorectal cancer in the USA, multiple prospective randomized controlled trials were able to be performed, demonstrating comparable outcomes for laparoscopic and open colectomies [23–25]. Using the US Nationwide Inpatient Sample database, Juo et al. reported that from 2008 to 2010, greater than 244,000 colectomies were performed, of which 47.6 % were performed laparoscopically [25]. This was a dramatic increase in the utilization of minimally invasive techniques from just a few years prior, when from 2005 to 2007, of the 240,000 colectomies performed for cancer, only 5.7 % were laparoscopic [27]. This rapid acceleration in utilization of laparoscopic techniques may suggest that MIG for cancer may also experience an upward trend, once technical feasibility and procedural safety have been demonstrated through focused investigation.
123
Surg Endosc
Multiple studies suggest that there are no differences in morbidity or mortality between open and MIG for cancer after a learning curve of 30–60 laparoscopic gastrectomies has been satisfied [8, 16, 28]. Due to the relative paucity of gastric cancer in the USA, it may be hard for some surgeons to reach 30–60 laparoscopic gastrectomies. The question and challenge then becomes, how should we satisfy this learning curve, when except for at high-volume institutions, surgeons typically do not encounter this volume of gastric cancer during their career. The answer may lie in the volume of minimally invasive procedures for benign gastric pathology [29–32]. A study evaluating Agency for Healthcare Research and Quality (AHRQ) data reports that from 1998 to 2008, the annual number of bariatric procedures performed in the USA increased from 12,775 to 124,838. The percentage of these that were performed laparoscopically also increased from 2.1 to 90.2 % [29]. As the study period progressed, patients were noted to be older and have more comorbidity; however, the length of stay and in-hospital mortality significantly decreased [29]. Similar to bariatric surgery, the annual number of minimally invasive anti-reflux procedures performed in the USA increased from 9000 in 1993 to [32,000 by 2000 [31, 32]. As a result, there was also a rise in the number of redo operations, necessitating the development of a complex technical acumen for laparoscopic foregut surgery. It is then possible that with gained experience in foregut surgery, surgeons are acquiring the necessary skills to consider minimally invasive approaches for oncologic indications. Alternatively, as was observed in early laparoscopic colectomy experience, there is opportunity for surgeons to be mentored by more experienced laparoscopic surgeons in order to combine technical and oncological expertise [33, 34]. Cost is a frequent criticism of minimally invasive surgery. Data regarding the differences in total cost for the surgical techniques were analyzed. Mean total cost was approximately 41,000 USD for both the open and laparoscopic groups, but was significantly higher in the robotic group at approximately 54,000 USD (P = 0.017). Several studies assessing other surgical procedures have reported that higher costs incurred with longer minimally invasive operative time and more expensive instruments may be somewhat negated by a decreased complication rate and LOS [3, 19, 20, 35, 36]. The current study would suggest significant room for improvement with MIG, as neither LOS nor overall complication rates were significantly different between the groups, yet the total costs were similar for open and laparoscopic procedures. As MIG experience increases and we progress up the laparoscopic foregut learning curve, we should also experience shorter LOS and less complications, yielding lower overall costs [29, 30].
123
Data analysis demonstrates that the choice of surgical technique does not have any independent associations with complications, readmission, or death (Table 4). This suggests the relative safety of minimally invasive techniques for gastrectomy in the USA. However, the concept of procedural safety requires the analysis of a complex interaction of multiple variables. Surgeon experience, operative time, estimated blood loss, extent of lymphadenectomy, and methods of reconstruction are not accounted for in the UHC database; however, all of these may affect a comprehensive evaluation of procedural safety. Future studies should seek to comprehensively assess these clinicopathological factors using large nationally representative clinical databases to be able to make further conclusions about the relative safety of minimally invasive procedures for gastric malignancy. In-hospital outcome metrics play important roles in the determination of the success of a particular procedure. We found that LOS, 30-day readmission, and in-hospital mortality were similar between the three techniques (Table 3). The 30-day readmission and mortality rates correspond with multiple prospective randomized trials from Asia [8–10, 17, 18, 20]. MIG LOS is reported to be less than open gastrectomy LOS in other studies; however, differences in LOS for this cohort did not reach significance. As was noted in the laparoscopic bariatric literature, our LOS for MIG in the USA may continue to decrease as experience with these techniques improves [29, 30, 37]. Overall complication rates for our cohort were very similar between the three groups. The increased incidence of superficial infection in the laparoscopic group and venous thromboembolism in the robotic group may be attributable to presumed longer operative times, although this variable is not collected by the UHC database. As another possible explanation, in the laparoscopic colectomy literature, venous thromboembolism has been shown to be independently associated with obesity and increasing age [38]. This may be extrapolated to our gastrectomy data, as gastric cancer patients in the USA tend to be older and have a higher body mass index when compared to Asian cohorts, which have not previously demonstrated differences in thromboembolic rates [1–4]. This study has several limitations. The UHC database is comprised of abstracted discharge data that is limited to inhospital events without outpatient follow-up. As such, no information is available to identify confounding factors that might necessitate an open operation (i.e., previous abdominal operations, locally advanced cancer, concurrent procedures, operative take backs, emergencies, or palliative procedures). Likewise, factors relevant to short-term outcomes (i.e., surgeon experience/training, extent of disease, and nutritional status) and long-term outcomes (i.e., extent
Surg Endosc
of lymphadenectomy, chemotherapy, and radiation) are not available. Additionally, the coding of certain complications may be inaccurate, as postoperative adverse events may be coded differently [39, 40]. Despite these limitations, the UHC database has been utilized to evaluate multiple other surgical procedures and serves the purposes of assessing current ‘‘status of care’’ and perioperative safety-related data [37, 41–43].
Conclusion Despite adoption for benign indications, minimally invasive gastrectomy for cancer remains underutilized in the USA. Within the context of UHC database, laparoscopic and robot-assisted gastrectomies appear to be comparable to open resection with respect to overall complications, LOS, 30-day readmission, and in-hospital mortality. Despite increasing data from Europe and Asia describing minimally invasive techniques as being analogous to open resection, US practice habits remained largely unchanged over the last 3 years of the study period. Similar to what has occurred with gastrectomies in Asia and with colectomies in the USA, as the incidence of gastric cancer continues to rise and interest in MIG grows, prospective randomized controlled trials will be required to further evaluate the feasibility, safety, oncologic equivalency, and long-term outcomes of MIG in the USA. Lastly, when performing MIG for cancer, infectious and thromboembolic risk reduction strategies should be emphasized in the operative and postoperative periods. Disclosures Jason Glenn, Kiran Turaga, T. Clark Gamblin, Samuel Hohmann, and Fabian Johnston have no financial or commercial relationships to disclose that might pose a conflict of interest.
References 1. SEER cancer statistics factsheets: stomach cancer. National Cancer Institute, Bethesda, MD, 2013. http://seer.cancer.gov/ statfacts/html/stomach.html 2. Inoue M, Tsugane S (2005) Epidemiology of gastric cancer in Japan. Postgrad Med J 81(957):419–424 3. Spolverato G, Kim Y, Ejaz A, Valero V, 3rd, Squires MH, Poultsides G, et al (2014) A Multi-institutional analysis of open versus minimally-invasive surgery for gastric adenocarcinoma: results of the US gastric cancer collaborative. J Gastrointest Surg. PubMed PMID: 24912915 4. Jung KW, Won YJ, Kong HJ, Oh CM, Seo HG (2013) Lee JS (2013) Prediction of cancer incidence and mortality in Korea. Cancer Res Treat 45(1):15–21 5. de Steur WO, Dikken JL, Hartgrink HH (2013) Lymph node dissection in resectable advanced gastric cancer. Dig Surg 30(2):96–103 6. Kitano S, Iso Y, Moriyama M, Sugimachi K (1994) Laparoscopyassisted Billroth I gastrectomy. Surg Laparosc Endosc 4(2):146–148
7. Kim HH, Han SU, Kim MC, Hyung WJ, Kim W, Lee HJ et al (2013) Prospective randomized controlled trial (phase III) to comparing laparoscopic distal gastrectomy with open distal gastrectomy for gastric adenocarcinoma (KLASS 01). J Korean Surg Soc 84(2):123–130 8. Kim HH, Han SU, Kim MC, Hyung WJ, Kim W, Lee HJ et al (2014) Long-term results of laparoscopic gastrectomy for gastric cancer: a large-scale case-control and case-matched Korean multicenter study. J Clin Oncol 32(7):627–633 9. Hayashi H, Ochiai T, Shimada H, Gunji Y (2005) Prospective randomized study of open versus laparoscopy-assisted distal gastrectomy with extraperigastric lymph node dissection for early gastric cancer. Surg Endosc 19(9):1172–1176 10. Nakamura K, Katai H, Mizusawa J, Yoshikawa T, Ando M, Terashima M et al (2013) A phase III study of laparoscopyassisted versus open distal gastrectomy with nodal dissection for clinical stage IA/IB gastric Cancer (JCOG0912). Jpn J Clin Oncol 43(3):324–327 11. Markar SR, Karthikesalingam A, Jackson D, Hanna GB (2013) Long-term survival after gastrectomy for cancer in randomized, controlled oncological trials: comparison between West and East. Ann Surg Oncol 20(7):2328–2338 12. Azagra JS, Ibanez-Aguirre JF, Goergen M, Ceuterick M, BordasRivas JM, Almendral-Lopez ML et al (2006) Long-term results of laparoscopic extended surgery in advanced gastric cancer: a series of 101 patients. Hepatogastroenterology 53(68):304–308 13. Iba´n˜ez-Aguirre FJ, Azagra JS, Erro-Azcarate ML (2006) Laparoscopic gastrectomy for gastric adenocarcinoma. Long-term results. REV ESP ENFERM DIG (Madrid) 98(7):491–500 14. Reavis KM (2009) Hospital volume is not a predictor of outcomes after gastrectomy for neoplasm. Am Surg 75(10):932–936 15. Meurer SJ (2009) Mortality Measurement: mortality risk adjustment methodology for University Health System Consortium. Agency for Healthcare Research and Quality, Rockville, MD. http://archive.ahrq.gov/professionals/quality-patient-safety/qualityresources/tools/mortality/Meurer.html 16. Chen K, Xu XW, Mou YP, Pan Y, Zhou YC, Zhang RC, et al (2013) Systematic review and meta-analysis of laparoscopic and open gastrectomy for advanced gastric cancer. World J Surg Oncol 11(182):12 17. Hosono S (2006) Meta-analysis of short-term outcomes after laparoscopy-assisted distal gastrectomy. R World J Gastroenterol 12(47):7676–7683 18. Lee JH, Han HS, Lee JH (2005) A prospective randomized study comparing open vs laparoscopy-assisted distal gastrectomy in early gastric cancer: early results. Surg Endosc 19(2):168–173 19. Mochiki E, Toyomasu Y, Ogata K, Andoh H, Ohno T, Aihara R et al (2008) Laparoscopically assisted total gastrectomy with lymph node dissection for upper and middle gastric cancer. Surg Endosc 22(9):1997–2002 20. Ohtani H, Tamamori Y, Noguchi K, Azuma T, Fujimoto S, Oba H et al (2011) Meta-analysis of laparoscopy-assisted and open distal gastrectomy for gastric cancer. J Surg Res 171(2):479–485 21. Strong VE, Devaud N, Karpeh M (2009) The role of laparoscopy for gastric surgery in the West. Gastric Cancer 12(3):127–131 22. Wong J, Jackson P (2011) Gastric cancer surgery: an American perspective on the current options and standards. Curr Treat Options Oncol 12(1):72–84 23. Nelson H, Sargent Daniel J, Sam Wieand H, Fleshman James, Anvari Mehran, Stryker Steven J et al (2004) A comparison of laparoscopically assisted and open colectomy for colon cancer. N Engl J Med 350(20):2050–2059 24. Lacy AM, Garcı´a-Valdecasas JC, Delgado S, Castells A, Taura´ P, Pique´ JM et al (2002) Laparoscopy-assisted colectomy versus open colectomy for treatment of non-metastatic colon cancer: a randomised trial. Lancet 359(9325):2224–2229
123
Surg Endosc 25. Juo YY, Hyder O, Haider AH, Camp M, Lidor A, Ahuja N (2014) Is minimally invasive colon resection better than traditional approaches?: first comprehensive national examination with propensity score matching. JAMA Surg 149(2):177–184 26. Cooperman AM, Katz V, Zimmon D, Botero G (1991) Laparoscopic colon resection: a case report. J Laparoendosc Surg 1(4):221–224 27. Robinson C, Chen GJ, Balentine C, Sansgiry S, Marshall C, Anaya D et al (2011) Minimally invasive surgery is underutilized for colon cancer. Ann Surg Oncol 18(5):1412–1418 (English) 28. El-Sedfy A, Brar SS, Coburn NG (2014) Current role of minimally invasive approaches in the treatment of early gastric cancer. World J Gastroenterol 20(14):3880–3888 29. Nguyen NT, Masoomi H, Magno CP, Nguyen XM, Laugenour K, Lane J (2011) Trends in use of bariatric surgery, 2003–2008. J Am Coll Surg 213(2):261–266 30. Dumon KR, Murayama KM (2011) Bariatric surgery outcomes. Surg Clin North Am 91(6):1313–1338 31. Carlson MA, Frantzides CT (2001) Complications and results of primary minimally invasive antireflux procedures: a review of 10,735 reported cases. J Am Coll Surg 193(4):428–439 32. Kim D, Velanovich V (2014) Surgical treatment of GERD: Where have we been and where are we going? Gastroenterol Clin North Am 43(1):135–145 33. Schlachta CM, Lefebvre KL, Sorsdahl AK, Jayaraman S (2010) Mentoring and telementoring leads to effective incorporation of laparoscopic colon surgery. Surg Endosc 24(4):841–844 34. Sebajang H, Trudeau P, Dougall A, Hegge S, McKinley C, Anvari M (2006) The role of telementoring and telerobotic assistance in the provision of laparoscopic colorectal surgery in rural areas. Surg Endosc 20(9):1389–1393
123
35. Huscher CGS, Mingoli A, Sgarzini G, Sansonetti A, Di Paola M, Recher A et al (2005) Laparoscopic versus open subtotal gastrectomy for distal gastric cancer. Ann Surg 241(2):232–237 36. Xiong B, Ma L, Zhang C (2012) Robotic versus laparoscopic gastrectomy for gastric cancer: a meta-analysis of short outcomes. Surg Oncol 21(4):274–280 37. Tiwari MM, Reynoso JF, High R, Tsang AW, Oleynikov D (2011) Safety, efficacy, and cost-effectiveness of common laparoscopic procedures. Surg Endosc 25(4):1127–1135 38. Becattini C, Rondelli F, Vedovati MC, Camporese G, Giustozzi M, Boncompagni M, et al (2004) Incidence and risk factors for venous thromboembolism after laparoscopic surgery for colorectal cancer. Haematologica. PubMed PMID: 25216684 39. Bruce J, Russell EM, Mollison J, Krukowski ZH (2001) The measurement and monitoring of surgical adverse events. Health Technol Assess (Winchester, England) 5(22):1–194 40. Davenport DL, Holsapple CW, Conigliaro J (2009) Assessing surgical quality using administrative and clinical data sets: a direct comparison of the University HealthSystem Consortium Clinical Database and the National Surgical Quality Improvement Program data set. Am J Med Qual 24(5):395–402 41. Cauble S, Abbas A, Balart L, Bazzano L, Medvedev S, Shores N (2013) United States women receive more curative treatment for hepatocellular carcinoma than men. Dig Dis Sci 58(10):2817–2825 42. Jalisi S, Bearelly S, Abdillahi A, Truong MT (2013) Outcomes in head and neck oncologic surgery at academic medical centers in the United States. Laryngoscope 123(3):689–698 43. Stack BC Jr, Spencer H, Moore E, Medvedev S, Bodenner D (2012) Outpatient parathyroid surgery data from the University Health System Consortium. Otolaryngol Head Neck Surg 147(3):438–443
