Abstract
Background
Surgical resection is the mainstay of potential cure for patients with pancreatic cancer, however, local recurrence is frequent. Previously, we have described an extended resection technique for pancreatoduodenectomy aiming at a radical resection of the nerve and lymphatic tissue between celiac artery, superior mesenteric artery and mesenteric–portal axis (TRIANGLE operation). Until now, data on postoperative outcome have not been reported, yet.
Methods
Patients who underwent either partial (PD) or total pancreatoduodenectomy (TP) applying the TRIANGLE procedure were identified. These cohorts were compared to matched historic cohorts with standard resections.
Results
Overall, 330 patients were analysed (PDTRIANGLE and PDSTANDARD, each n = 108; TPTRIANGLE and TPSTANDARD, each n = 57). More lymph nodes were harvested in TRIANGLE compared to standard resection (PD: 27.5 (21–35) versus 31.5 (24–40); P = 0.0187, TP: 33 (28–49) versus 44 (29–53); P = 0.3174) and the rate of tumour positive resections margins, R1(direct), dropped. Duration of operation was significantly longer and blood loss higher. Postoperative mortality and complications did not differ significantly.
Conclusion
Pancreatoduodenectomy according to the TRIANGLE protocol can be performed without increased morbidity and mortality at a high-volume centre. Long-term survival and quality of life need to be investigated in prospective clinical trials with adequate sample size.
Introduction
Pancreatic cancer (PDAC) is the fourth most frequent cause of cancer-related death in the western world
1
and the global incidence and mortality of PDAC is predicted to rise by almost 80% by 2040.2
It has a dismal prognosis with a 5-year overall survival of approximately 5%.WHO International Agency for Research in Cancer
Cancer tomorrow.
Cancer tomorrow.
http://gco.iarc.fr/tomorrow/home
Date accessed: November 15, 2020
1
In patients with resectable tumours surgery and adjuvant chemotherapy improve survival to >30 months in favourable subgroups.
3
,- Neoptolemos J.P.
- Palmer D.H.
- Ghaneh P.
- Psarelli E.E.
- Valle J.W.
- Halloran C.M.
- et al.
Comparison of adjuvant gemcitabine and capecitabine with gemcitabine monotherapy in patients with resected pancreatic cancer (ESPAC-4): a multicentre, open-label, randomised, phase 3 trial.
Lancet. 2017; 389: 1011-1024
4
However, disease progression is frequent, with approximately 70% of patients resected having a relapse within 2 years.5
,6
Furthermore, local recurrence as well as distant metastases are associated with poor prognosis.7
A recent meta-analysis of recurrence patterns showed that local recurrence was the sole recurrence site in 20.8% of patients.8
Another 27.8% of patients exhibited recurrence at multiple sites, almost always including local recurrence.8
Risk factors for local recurrence after resection are borderline resectable and locally advanced tumours,9
lymph node involvement10
and a positive pathological resection margin.11
Incomplete tumour clearance most frequently affects the medial soft tissue margins along the celiac artery (CA) and superior mesenteric artery (SMA).12
, 13
, 14
Negative resection margin status (R0) along the posterior and medial margins has been associated with reduced local recurrence rate and disease-free survival (DFS) in a large observational study.15
However, resection of the CA and SMA – even if technically feasible – is not generally recommended16
as it increases perioperative morbidity and mortality and compromises overall oncological outcomes.17
,18
Yet, the negative impact of local recurrence on outcome warrants efforts to reduce the risk for it. To this end, in 2017, we introduced a more radical dissection of soft tissue, lymphatics and nerve plexus along the CA, SMA and mesenteric–portal axis (MPA), i.e. the TRIANGLE region.
19
However, the anticipated oncological advantage of the TRIANGLE operation with better margin clearance and expected reduced local recurrence and better survival must be weighed against the risk of a more radical surgical approach, i.e. higher morbidity. The TRIANGLE procedure was piloted in 15 patients with acceptable morbidity,19
however data from a larger cohort are lacking. As parts of the autonomous nerve system as well as all lymphatic tissue in the TRIANGLE region are removed, postoperative diarrhoea as well as chyle leak may be increased.20
,21
The aim of the present study is therefore to evaluate the safety and postoperative outcome of the TRIANGLE procedure in a large single-centre cohort compared to standard PD and TP.
Methods
Study design and patient cohort
Data collection and analysis was approved by the institutional ethics committee (S-011/2015) and follows the STROBE recommendations for observational studies.
22
Informed consent was obtained from all patients. The prospectively maintained electronic pancreas database of the Department of Surgery at Heidelberg University Hospital, Germany was searched and all consecutive patients undergoing PD and TP according to the TRIANGLE procedure (PDTRIANGLE and TPTRIANGLE) for suspected PDAC between 03/2016 and 10/2019 were included. To prevent selection bias a historical matched cohort was chosen consisting of patients operated between 01/2013 and 02/2016 with standard PD and TP without extended lymphadenectomy in the TRIANGLE region (PDSTANDARD and TPSTANDARD). Matching criteria for the study population were age, ASA status, BMI, neoadjuvant chemotherapy (yes/no), CA19-9 (<37/37–400/400) and grading (X/I/II/III). A matching list of the total cohort, PDTRIANGLE, TPTRIANGLE and patients undergoing PDSTANDARD and TPSTANDARD between 01/2013 and 02/2016 for suspected PDAC, sorted by the matching criteria according to the above order was generated. No selection conflict of two or more PDSTANDARD and TPSTANDARD with the same six criteria was observed.Data collection and outcome parameters
The following clinical parameters were extracted from the prospectively maintained database: Baseline data acquisition included patients’ gender, age, body mass index (BMI) and American Association of Anaesthesiologists (ASA) physical status classification. Preoperative serum values of the tumour marker carbohydrate antigen 19-9 (CA 19-9) were electronically extracted from the hospital laboratory information system. Operative outcome parameters included type of pancreatic resection (PD/TP), extent of surgery regarding other adjacent organs, number of resected lymph nodes and type of vascular resection (type 1: regular resection, type 2: additional venous resection, type 3: multi-visceral resection, type 4: additional arterial resection
33
), insertion of intraabdominal drainage, operation time and estimated intraoperative blood loss. All postoperative complications within 30 days after surgery were classified according to the Clavien-Dindo Classification.23
Only complications graded > II were recorded. Furthermore, the following classifications were applied: Postpancreatectomy haemorrhage (PPH),24
postoperative pancreatic fistula (POPF),25
chyle leak,26
delayed gastric emptying (DGE),27
and bile leakage28
were classified according to ISGPS definitions and clinically relevant complications (grade B and C according to ISGPS) were reported. Diarrhoea is not regularly recorded in the prospectively maintained database and could only be analysed for the period of the postoperative hospital stay. Diarrhoea defined as defecation more than three times a day and regular use of anti-diarrheal medication in postoperative week two were recorded.Reoperations as well as length of hospital stay, length of intensive care unit stay and mortality were assessed. Pathological work-up included tumour grading and pTNM tumour stage according to the 7th (before January 2017) and 8th edition (after January 2017) of the TNM Staging Manual, American Joint Committee on Cancer (AJCC).
29
The pTNM tumour stages categorized before January 2017 were reclassified according to the current version for this analysis.The resection margin (R) status was assessed along with the new definition of R0 requiring a 1 mm tumour-free margin (corresponds to R0 (CRM-)).
11
“R1 (<1 mm)” was used for tumours classified as R0 (CRM+) and “R1 (direct)” was used for tumours with direct microscopic involvement of the resection margin (R1 according to TNM 8th edition).TRIANGLE intervention (PDTRIANGLE/TPTRIANGLE)
The experimental/TRIANGLE intervention is a modification of the extent of the resection. Both groups (TRIANGLE and standard) start with the same exploratory phase. Following exploration, the TRIANGLE operation is a combination of two previously described surgical steps
19
,21
:- A.)Dissection of the SMA according to the level 3 described by Inoue et al.21. This step includes a dissection of the nerve plexus around the SMA from at least 5 to 11 o'clock (180°). A wider resection (≥180°) up to a circular (360°) resection of the lymph and nerve plexus around the SMA was sometimes performed at the discretion of the surgeon. Performing this step results in a circular (360°) dissection of the superior mesenteric vein (SMV).
- B.)Complete dissection of the soft tissue in the TRIANGLE between CA, SMA and MPA.19
Besides that, lymphadenectomy was performed as recommended by the ISGPS
30
and current German S3-guidelines.31
Control intervention (PDSTANDARD/TPSTANDARD)
Patients in the control groups received standard PD or TP with dissection of the SMA according to Inoue level 1 or 2
21
and standard lymphadenectomy according to the ISGPS30
and German S3-guidelines31
, but no TRIANGLE dissection.Treatments in both groups
Reconstruction was similar in both groups, i.e. pancreatojejunostomy (in PD cases), hepatojejunostomy and duodeno-/gastrojejunostomy on a single jejunal loop. If necessary for complete tumour removal, venous resections (portal vein or superior mesenteric vein) or multi-visceral resections were performed. Likewise, arterial resection or periarterial divestment
32
was done in both groups if needed for tumour clearance.Our standard oncological treatment strategy in the study period was upfront resection and adjuvant chemotherapy for all tumours considered resectable, even in borderline or locally advanced cases. Adjuvant chemotherapy was recommended to all patients with upfront surgery of PDAC irrespective of tumour stage. However, some patients were only admitted after receiving neoadjuvant chemotherapy according to decision of external tumour board or oncologist. 5-FU/folinic acid or gemcitabine or FOLFIRINOX were given as standard regimens for both neoadjuvant and adjuvant therapy.
Statistical analysis
Data management and statistical analysis were carried out by SAS® software release 9.4 (SAS Institute, Cary, North Carolina, USA).
For analysis of perioperative outcomes, all matched patients were included. Patient characteristics were summarized using absolute and relative frequencies for categorical variables and the median with the interquartile range for continuous variables. Comparisons between the PD and TP subgroups were performed using the Fisher's exact test for categorical variables and the Mann–Whitney u test for continuous variables. For short-term oncological analysis, only pairs of patients with histologically proven PDAC were included. In the TP groups oncological outcomes were not analysed due to the insufficient sample size for this sub-cohort.
Two-sided P values were computed and a difference was considered statistically significant at P < 0.05. Because of the exploratory character of the analyses performed all results were interpreted cautiously and the p-values were used descriptively.
Results
Study cohort and clinical characteristics
A total of 330 patients were included: Between March 2016 and October 2019, a total of 165 patients underwent PDTRIANGLE (n = 108) or TPTRIANGLE (n = 57). These patients were compared to a matched historical cohort with PDSTANDARD or TPSTANDARD operated between January 2013 and February 2016.
An overview of demographic, clinicopathological and surgical characteristics of the cohort is provided in Table 1. Baseline characteristics were comparable between the groups. Of all patients, the mean age was 65.5 (range: 34.5–87.6) years, mean BMI was 24.3 (range: 17.4–41.8) kg/m2 and ASA-Status was I-III (ASA I: n = 16; ASA II: n = 193; ASA III: n = 121). The study population consisted of 142 women (43.0%) and 188 men (57.0%). Overall, 89 patients received neoadjuvant chemotherapy including 49 patients in the TRIANGLE group (PDTRIANGLE n = 27 and TPTRIANGLE n = 22) and 40 patients in the standard resection group (PDSTANDARD n = 23 and TPSTANDARD n = 17; P = 0.6287 and P = 0.4299, respectively). In total, 166 patients (50.3%) underwent pancreatic resections only (type 1 pancreatic resection), 120 patients (36.4%) had additional venous resection (type 2 pancreatic resection), 29 patients (8.8%) had multi-visceral resection (type 3 pancreatic resection) and 15 patients (4.5%) underwent additional arterial resection (type 4 pancreatic resection).
33
Histopathology confirmed PDAC in 266 patients (80.6%). The remaining patients were diagnosed with periampullary carcinoma (n = 30), neuroendocrine tumour (NET, n = 19), intraductal papillary mucinous neoplasm (IPMN, n = 8) or other pancreatic pathologies (n = 7).Table 1Demographic, clinicopathological and surgical characteristics according to type of resection
| Pancreatoduodenectomy | Total pancreatectomy | |||||
|---|---|---|---|---|---|---|
| Standard | TRIANGLE | P-value | Standard | TRIANGLE | P-value | |
| N = 108 | N = 108 | N = 57 | N = 57 | |||
| Age (years) | 66.1 (58.8–74.3) | 65.8 (58.7–73.3) | 0.9687 | 64.8 (57.4–72.4) | 64.8 (59.7–71.8) | 0.9684 |
| Sex ratio (male/female)) | 55:53 (50.9:49.1) | 68:40 (63.0:37.0) | 0.0989 | 33:24 (57.9:42.1) | 32:25 (56.1:43.9) | 1.0 |
| BMI (kg/m2) | 24.1 (22.4–26.8) | 24.5 (22.0–27.2) | 0.9844 | 24.4 (22.5–26.2) | 24.3 (22.1–27.4) | 0.8784 |
| ASA classification | 0.8780 | 0.8342 | ||||
| ASA 1 | 3 (2.8) | 4 (3.7) | 4 (7.0) | 5 (8.8) | ||
| ASA 2 | 62 (57.4) | 59 (54.6) | 35 (61.4) | 37 (64.9) | ||
| ASA 3 | 43 (39.8) | 45 (41.7) | 18 (31.6) | 15 (26.3) | ||
| CA 19-9 (U/m) | 50.0 (17.2–288.4) | 73.5 (19.9–324.0) | 0.5665 | 71.3 (16.4–356.2) | 116.9 (20.9–267.5) | 0.6874 |
| Neoadjuvant therapy | 23 (21.3) | 27 (25.0) | 0.6287 | 17 (29.8) | 22 (38.6) | 0.4299 |
| Type of resection | 0.1923 | 0.7933 | ||||
| 1 | 67 (62.0) | 56 (51.9) | 23 (40.4) | 20 (35.1) | ||
| 2 | 34 (31.5) | 47 (43.5) | 17 (29.8) | 22 (38.6) | ||
| 3 | 6 (5.6) | 3 (2.8) | 11 (19.3) | 9 (15.8) | ||
| 4 | 1 (0.9) | 2 (1.9) | 6 (10.5) | 6 (10.5) | ||
| Histology | 0.6339 | 0.1000 | ||||
| PDAC | 82 (75.9) | 89 (82.4) | 45 (78.9) | 50 (87.7) | ||
| Periampullary carcinoma | 15 (13.9) | 10 (9.3) | 1 (1.8) | 4 (10.3) | ||
| IPMN | 1 (0.9) | 0 (0.0) | 6 (10.5) | 1 (1.8) | ||
| NET | 7 (6.5) | 8 (7.4) | 2 (3.5) | 2 (3.5) | ||
| Other | 3 (2.89) | 1 (0.9) | 3 (5.3) | 0 (0.0) | ||
| Duration of operation (min) | 322 (263–380) | 360 (302–421) | 0.0014 | 367 (315–445) | 434 (385–490) | 0.0002 |
| Intraoperative blood loss (ml) | 600 (500–1000) | 1000 (650–1600) | < 0.0001 | 800 (500–1500) | 1650 (1150–2550) | < 0.0001 |
| Intraabdominal drain | 91 (84.3) | 77 (71.3) | 0.0327 | 51 (89.5) | 41 (71.9) | 0.0311 |
Values in parentheses are percentages unless indicated otherwise. Significant P values (P < 0.05) are presented in bold.
Grading and R classification are given for the entire cohort including PDAC and all other tumor types. ASA, american society of anesthesiologists; ELN, examined lymph nodes; PDAC, pancreatic periductal adenocarcinoma; IPMN, intraductal papillary mucinous neoplasia; NET, neuroendocrine tumour.
a Values are median (IQR).
Surgical data
Duration of operation was significantly longer in the TRIANGLE groups compared to standard resections (Table 1) (PDSTANDARD: 322 (263–380) min versus PDTRIANGLE: 359.5 (301.5–420.5) min, P = 0.0014; TPSTANDARD: 367 (315–445) min versus TPTRIANGLE: 434 (385–490) min, P = 0.0002). Intraoperative blood loss was significantly higher in the TRIANGLE than in the standard groups (PDSTANDARD: 600 (500–1000) ml versus PDTRIANGLE: 1000 (650–1600) ml, P < 0.0001; TPSTANDARD 800 (500–1500) ml versus TPTRIANGLE: 1650 (1150–2550) ml, P < 0.0001). An intraabdominal drain was placed more often in standard resections than after TRIANGLE resections (PDSTANDARD: 91 (84.3%) versus PDTRIANGLE: 77 (71.3%), P = 0.0327; TPSTANDARD: 51 (89.5%) versus TPTRIANGLE: 41 (71.9%), P = 0.0311).
Postoperative morbidity and mortality
An overview of postoperative morbidity and mortality is provided in Table 2. The 30-day, 90-day and in-hospital mortality rates did not differ significantly between standard and TRIANGLE resection (Table 2). Overall 30-day mortality rate in the entire cohort was 1.8%, 90-day mortality 3.0% and in-hospital mortality 3.6%.
Table 2Postoperative morbidity and mortality
| Pancreatoduodenectomy | Total pancreatectomy | |||||
|---|---|---|---|---|---|---|
| Standard | TRIANGLE | P-value | Standard | TRIANGLE | P-value | |
| N = 108 | N = 108 | N = 57 | N = 57 | |||
| Mortality | ||||||
| 30–day | 3 (2.8) | 0 (0.0) | 0.2465 | 2 (3.5) | 1 (1.8) | 1.0 |
| 90-day | 4 (3.7) | 1 (0.9) | 0.3691 | 4 (7.0) | 1 (1.8) | 0.3638 |
| In-hospital | 5 (4.6) | 1 (0.9) | 0.2122 | 4 (7.0) | 2 (3.5) | 0.6790 |
| Number of complications | 62 | 70 | 0.6427 | 30 | 50 | 0.2705 |
| IIIa | 29 | 37 | 15 | 26 | ||
| IIIb | 18 | 23 | 7 | 18 | ||
| IVa | 15 | 10 | 6 | 6 | ||
| IVb | 0 | 0 | 2 | 0 | ||
| Surgical Morbidity | 48 (44.4) | 51 (47.2) | 0.7848 | 31 (54.4) | 28 (49.1) | 0.7080 |
| POPF B/C | 17 (15.7) | 12 (11.1) | 0.4251 | – | – | – |
| Grade B | 11 (10.2) | 8 (7.4) | – | – | – | |
| Grade C | 6 (5.6) | 4 (3.7) | – | – | – | |
| Bile leakage B/C | 2 (1.9) | 7 (6.5) | 0.1705 | 4 (7.0) | 2 (3.5) | 0.6790 |
| Grade B | 1 (0.9) | 5 (4.6) | 2 (3.5) | 2 (3.5) | ||
| Grade C | 1 (0.9) | 2 (1.9) | 2 (3.5) | 0 (0.0) | ||
| Chyle leak B/C | 4 (3.7) | 6 (5.6) | 0.7480 | 5 (8.8) | 3 (5.3) | 0.7165 |
| Grade B | 4 (3.7) | 5 (4.6) | 3 (5.3) | 2 (3.5) | ||
| Grade C | 0 (0.0) | 1 (0.9) | 2 (3.5) | 1 (1.8) | ||
| PPH B/C | 3 (2.8) | 5 (4.6) | 0.7214 | 2 (3.5) | 5 (8.8) | 0.4382 |
| Grade B | 0 (0.0) | 1 (0.9) | 2 (3.5) | 2 (3.5) | ||
| Grade C | 3 (2.8) | 4 (3.8) | 0 (0.0) | 3 (5.3) | ||
| DGE B/C | 10 (9.3) | 10 (9.3) | 1.0 | 8 (14.0) | 8 (14.0) | 1.0 |
| Grade B | 5 (4.6) | 6 (5.6) | 7 (12.3) | 5 (8.8) | ||
| Grade C | 5 (4.6) | 4 (3.7) | 1 (1.8) | 3 (5.3) | ||
| Abscess | 14 (13.0) | 15 (13.9) | 1.0 | 9 (15.8) | 12 (21.1) | 0.6297 |
| Non-surgical morbidity | 33 (30.6) | 21 (19.4) | 0.0834 | 23 (40.4) | 26 (45.6) | 0.7054 |
| pulmonary | 14 (13.0) | 7 (6.5) | 0.1670 | 9 (15.8) | 9 (15.8) | 1.0 |
| cardiac | 6 (5.6) | 4 (3.7) | 0.7480 | 2 (3.5) | 5 (8.8) | 0.4382 |
| Relaparotomy | 13 (12.0) | 18 (16.7) | 0.4380 | 7 (12.3) | 13 (22.80) | 0.2177 |
| Completion pancreatectomy | 3 (2.8) | 4 (3.7) | 1.0 | – | – | – |
| ICU stay ≥ 2 days | 12 (11.1) | 33 (30.6) | 0.0007 | 8 (14.0) | 26 (45.6) | 0.0004 |
| Postoperative hospital stay (days) | 12 (10–20) | 13 (9.5–20.5) | 0.8229 | 15 (13–20) | 17 (13–25) | 0.5342 |
| Diarrhoea | 13 (14.4) | 31 (34.4) | 0.0029 | 14 (29.2) | 14 (29.2) | 1.0 |
| Anti-diarrheal medication | 5 (5.6) | 15 (16.7) | 0.0307 | 4 (8.3) | 9 (18.8) | 0.2321 |
Values in parentheses are percentages unless indicated otherwise. Significant P values (P < 0.05) are presented in bold.
POPF: postoperative pancreatic fistula, PPH: postpancreatectomy haemorrhage, DGE: delayed gastric emptying, ICU: Intensive Care Unit.
a Values are median (IQR).
b Data are missing in 18 PDSTANDARD and 9 TPSTANDARD patients and are only given for complete matched pairs.
There was no difference in overall surgical and non-surgical morbidity between standard and TRIANGLE resection. Clinically relevant POPF (grade B and C) occurred in 15.7% (PDSTANDARD) and 11.1% (PDTRIANGLE) of patients (P = 0.4251). There were also no significant differences in clinically relevant bile leakage (PDSTANDARD: 1.9% versus PDTRIANGLE: 6.5%, P = 0.1705; TPSTANDARD: 7% versus TPTRIANGLE: 3.5%, P = 0.6790), chyle leak (PDSTANDARD: 3.7% versus PDTRIANGLE: 5.6%, P = 0.7480; TPSTANDARD: 8.8% versus TPTRIANGLE: 5.3%, P = 0.7165), postpancreatectomy haemorrhage (PDSTANDARD: 2.8% versus PDTRIANGLE: 4.6%, P = 0.7214; TPSTANDARD: 3.5% versus TPTRIANGLE: 8.8%, P = 0.4382) and clinically relevant delayed gastric emptying (PDSTANDARD: 9.3% versus PDTRIANGLE: 9.3%, P = 1.0; TPSTANDARD: 14.0% versus TPTRIANGLE: 14%, P = 1.0).
In the TRIANGLE groups, more patients had to stay on Intensive Care Unit (ICU) ≥ 2 days than in the standard groups (PDSTANDARD: 12 (11.1%) versus PDTRIANGLE: 33 (30.6%), P = 0.0007; TPSTANDARD: 8 (14.0%) versus TPTRIANGLE: 26 (45.6%), P = 0.0004). Rate of relaparotomy and median length of hospital stay did not differ significantly between standard and TRIANGLE resections with a median length of hospital stay of 13 days after PD and 16 days after TP.
Data on the occurrence of diarrhoea in postoperative week two were missing in 18 PDSTANDARD and 9 TPSTANDARD patients and are only presented for complete matched pairs. Number of patients suffering of diarrhoea differed significantly between PD groups (PDSTANDARD: 13 (14.4%) versus PDTRIANGLE: 31 (34.4%), P = 0.0029) and significantly more patients required anti-diarrheal medication in the PDTRIANGLE group compared to PDSTANDARD (PDSTANDARD: 5 (5.6%) versus PDTRIANGLE: 15 (16.7%), P = 0.0307). In the TP cohorts no significant difference was seen regarding diarrhoea and anti-diarrheal medication.
Short-term oncological outcome in pancreatic cancer patients
Oncological outcomes were assessed in the subgroup of patients with histopathological confirmation of PDAC (PDTRIANGLE and PDSTANDARD, each n = 70; TPTRIANGLE and TPSTANDARD, each n = 39; Table 3). There were no significant differences in preoperative CA 19-9 values, number of patients with neoadjuvant chemotherapy, tumour grading and T, N and M stage between TRIANGLE and standard resections. Significantly more lymph nodes were harvested in the PDTRIANGLE cohort compared to PDSTANDARD (PDSTANDARD: 27.5
7
,21
, 22
, 23
, 24
, 25
, 26
, 27
, 28
, 29
, 30
, 31
, 32
, 33
, 34
versus PDTRIANGLE: 31.57
,24
, 25
, 26
, 27
, 28
, 29
, 30
, 31
, 32
, 33
, 34
, 35
, 36
, 37
, 38
, 39
; P = 0.0187). This difference did not reach significance in the TP cohorts (TPSTANDARD: 33 (28–49) versus TPTRIANGLE: 44 (29–53); P = 0.3174).Table 3Clinicopathological characteristics of patients with confirmed PDACs
| PD | TP | |||||
|---|---|---|---|---|---|---|
| Standard | Triangle | P-value | Standard | Triangle | P-value | |
| N = 70 | N = 70 | N = 39 | N = 39 | |||
| CA 19-9 (U/mL) | 85.7 (22.5–508.5) | 132.8 (27.5–661.6) | 0.5709 | 124.2 (24.9–625.5) | 159.0 (22.2–578.2) | 0.7643 |
| Neoadjuvant therapy | 17 (24.3) | 19 (27.1) | 0.8469 | 14 (35.9) | 19 (48.7) | 0.3594 |
| T status (8th) | 0.8252 | 0.3611 | ||||
| T0 | 1 (1.4) | 0 (0.0) | 4 (10.3) | 4 (10.3) | ||
| T1 | 10 (14.3) | 8 (11.4) | 13 (33.3) | 19 (48.7) | ||
| T2 | 42 (60.0) | 46 (65.7) | 20 (51.3) | 16 (41.0) | ||
| T3 | 17 (24.3) | 16 (22.9) | 2 (5.1) | 0 (0.0) | ||
| N status (8th) | 0.7429 | 0.2162 | ||||
| N0 | 23 (32.9) | 19 (27.1) | 10 (25.6) | 6 (15.4) | ||
| N1 (1–3 PLN) | 15 (21.4) | 18 (25.7) | 9 (23.1) | 16 (41.0) | ||
| N2 (≥4 PLN) | 32 (45.7) | 33 (47.1) | 20 (51.3) | 17 (43.6) | ||
| M status | 0.2746 | 0.1153 | ||||
| M0 | 64 (91.4) | 68 (97.1) | 35 (89.7) | 39 (100.0) | ||
| M1 | 6 (8.6) | 2 (2.9) | 4 (10.3) | 0 (0.0) | ||
| Grading | 0.5065 | 0.2732 | ||||
| 1 | 0 (0.0) | 1 (1.8) | 18 (64.3) | 14 (63.6) | ||
| 2 | 31 (55.4) | 29 (52.7) | 10 (35.7) | 6 (27.3) | ||
| 3 | 23 (41.1) | 25 (45.5) | 0 (0.0) | 2 (9.1) | ||
| 4 | 2 (3.6) | 0 (0.0) | 11 | 17 | ||
| X | 14 | 15 | ||||
| R classification | 0.2721 | 0.0606 | ||||
| R0 | 20 (28.6) | 20 (28.6) | 4 (10.5) | 12 (30.8) | ||
| R1 (<1 mm) | 20 (28.6) | 28 (40.0) | 11 (29.0) | 12 (30.8) | ||
| R1 (direct) | 30 (42.9) | 22 (31.4) | 23 (60.5) | 15 (38.5) | ||
| Number of ELN | 27.5 (21–35) | 31.5 (24–40) | 0.0187 | 33 (28–49) | 44 (29–53) | 0.3174 |
Values in parentheses are percentages unless indicated otherwise. Significant P values (P < 0.05) are presented in bold.
PLN, positive lymph nodes; ELN, examined lymph nodes.
a Values are median (IQR).
For PD the rate of R0 resections (>1 mm resection margin; R0(CRM-)) did not differ between groups. However, R1 (<1 mm) (R0 CRM+) was >10% more frequent in the TRIANGLE group compared to standard resection. Vice versa R1(direct) was more frequent in the standard PD group (Table 3). In the TP cohort, R0 resections were significantly more frequent in the TPTRIANGLE group than in the TPSTANDARD group (TPTRIANGLE 30.8% vs. TPSTANDARD 10.5%; p = 0.0475), while the rate of R1 (<1 mm) was comparable (TPTRIANGLE 30.8% vs. TPSTANDARD 29.0%; p = 1.0).
Discussion
We show in a large series of patients that pancreatoduodenectomy and total pancreatectomy with the so-called TRIANGLE procedure is feasible and safe with comparable postoperative morbidity as standard surgery at a high-volume institution. Nevertheless, compared to standard pancreatic resection, duration of surgery is longer, intraoperative blood loss higher, number of harvested lymph nodes greater and ICU stay longer. However, the overall number and severity of postoperative complications and 30-day as well as 90-day and in-hospital mortality do not differ significantly.
Local recurrence is frequent after surgical resection of PDAC and prognosis is worse for patients with local recurrence than without.
5
, 6
, 7
, 8
The question whether the extent of resection (including tissue along the major vessels) influences survival and health-related quality of life is among the top 10 research priorities in pancreatic cancer.34
The TRIANGLE procedure goes beyond previously described extended lymphadenectomies by additionally removing all tissue between the CA, the SMA and the MPA.19
This operation leads to clearance of all neural and lymphatic tissue which drains the pancreatic head. Owing to the perineural growth pattern of PDAC, R1 resections and local recurrence are often found in the “TRIANGLE” region3
,- Neoptolemos J.P.
- Palmer D.H.
- Ghaneh P.
- Psarelli E.E.
- Valle J.W.
- Halloran C.M.
- et al.
Comparison of adjuvant gemcitabine and capecitabine with gemcitabine monotherapy in patients with resected pancreatic cancer (ESPAC-4): a multicentre, open-label, randomised, phase 3 trial.
Lancet. 2017; 389: 1011-1024
7
,13
,14
,35
which supports a more radical dissection techniques along the SMA and the CA.Techniques of radical tumour clearance with removal of nerve tissue and lymphatics in pancreatic surgery have been described for different arteries separately. Inoue et al. reported similar rates of perioperative complications and length of hospital stay in a retrospective study of extended resection along the SMA compared to standard resection.
21
In 2019, the group published long-term results showing that extended resection resulted in a higher rate of postoperative diarrhoea requiring opioid treatment. Compliance to adjuvant chemotherapy and survival did not differ significantly compared to standard resections.36
Similar results have been demonstrated for radical tumour clearance along the common hepatic artery.37
,38
It is noteworthy, that the TRIANGLE operation is not another type of extended lymphadenectomy, but rather focuses on the site of microscopic tumour spread and frequent local tumour recurrence (i.e. the TRIANGLE region) by removing all soft tissue, lymphatics and nerve plexus. The increase in dissected lymph nodes observed in the TRIANGLE groups (Table 3) can therefore been seen as a by-product or surrogate parameter, but not as the primary goal of the procedure. This is important as extended lymphadenectomies alone failed to show a survival benefit in meta-analyses of randomized trials, and may even be associated with increased morbidity.
39
,40
In our study baseline characteristics did not differ between the standard and the TRIANGLE group, which confirms successful matching. Nevertheless, higher blood loss and longer duration of surgery reflect more extended resections in the TRIANGLE groups. Complication rates including POPF, bile leakage, postpancreatectomy haemorrhage and DGE did not differ between groups, confirming the safety of the TRIANGLE procedure in experienced hands. The number of complications in the standard group were in line with results in the literature.
36
,41
,42
The non-significantly increased number of complications graded as Clavien-Dindo IIIa and IIIb in the TRIANGLE cohorts might be reflected by the prolonged stay in the intensive care unit.As expected, more patients in the PDTRIANGLE group suffered of postoperative diarrhoea compared to the control group. Besides, more PDTRIANGLE patients required anti-diarrheal medication. This finding is in line with previous data and can be explained by the more radical resection of the periarterial nerve plexus.
21
In the TP cohort, rate of patients suffering from diarrhoea was approximately 30% in both groups (accordingly comparable to rate in the PDTRIANGLE cohort), with twice as many patients in the TPTRIANGLE group in need of anti-diarrheal medication compared to TPSTANDARD.The rate of lymphatic fistula, which we would have expected higher in TRIANGLE patients, did not differ significantly (PDSTANDARD 3.7% vs. PDTRIANGLE 5.6; P = 0.7480; TPSTANDARD 8.8% vs. TPTRIANGLE 5.3; P = 0.7165). Several factors may have contributed to this unexpected result. First, we only recorded chyle leaks type B and C according to the ISGPS definition in our database.
26
Thus, the total number of lymphatic fistulas (including type A) was not assessed. Second, the ISGPS definition of chyle leaks was introduced in 2016. Therefore, we performed a retrospective classification of patient data before 2016, including the entire control group, which may have distorted results compared to the prospective data acquisition of chyle leak in the TRIANGLE group from 2016 onwards. Finally, based on results from recent randomized trials,43
drains following PD and TP were more frequently avoided in later years, leading to significant differences between the groups (see Table 1). This change in drain management over time may have blurred the diagnosis of (mild) lymphatic fistulas in the TRIANGLE group.Although our study focused on perioperative morbidity and mortality, we performed some exploratory analyses of oncological short-term outcome parameters. The TNM classification only focusses on the number of lymph nodes and presence of perineural invasion but not on the anatomical site. The more radical extent of surgery in the TRIANGLE region is therefore inadequately reflected in the TNM classification. The observed significant increase in the number of resected lymph nodes in the PD cohort, as well as a clinically relevant decrease in R1 (direct) resections in the TRIANGLE groups of 11.5% after PD and 22.5% after TP, however, reflect a more radical surgery. Involvement of surgical margins has been shown to be a prognostic marker of local recurrence and overall survival.
7
Therefore, the observed differences might be clinically relevant for long-term outcome as a large cohort study including 561 PDAC patients with R0 (≥1 mm margin) and R1 (<1 mm) resections showed a better overall survival than patients with R1 (direct) margin involvement.11
This independent association was confirmed by multivariate analysis11
and confirmed in a recent meta-analysis which showed that R1 resections significantly increase the risk for local recurrence (OR 3.18, 95% CI 1.16–8.71).8
As the periarterial tissue dissected along the SMA, the CA and the MPA is frequently removed separately from the main pathologic specimen, true margin clearance can be difficult to establish on pathological work-up. Consequently, one might hypothesize that the observed resection margin status on standard pathologic work-up underestimates the true radicality of the TRIANGLE operation.There are several limitations to this study. First, the dissection extent (TRIANGLE versus standard) was determined individually for each patient according to the surgical report and intraoperative photo documentation (if available) following standardized criteria and in consensus between two surgeons. Regardless, the dissection extent may have been influenced by individual factors like surgeon's experience and patient anatomy, tumour size and comorbidities, leading to a risk of intergroup migration. To reduce both the risk of intergroup migration and a potential selection bias, we chose a historical matched cohort for comparison. Before March 2016, no TRIANGLE procedures were performed at our institution. Second, this is a single-centre study, so the generalizability of the results may be limited and future multicentre studies will be needed to verify the results. Third, given the retrospective study design we may not have been able to control for all known and unknown confounding factors. Fourth, comprehensive data on adjuvant chemotherapy are lacking. Fifth, the severity and frequency of diarrhoea, which was reported to be increased in patients with nerve plexus resection along the SMA
39
,44
could not be analysed in a long-term follow up. Data about diarrhoea in postoperative week two is not that critical as many patients suffer from gastrointestinal symptoms directly after PD and TP. Hence, a profound analysis of the degree and frequency of diarrhoea would need to comprise more than just the immediate postoperative period and include evaluation of malnutrition and the effect on the application of adjuvant chemotherapy as well. Finally, our report focuses on postoperative mortality and morbidity only, therefore the effect of the TRIANGLE operation on local recurrence, survival, nutritional status and health-related quality of life and other patient-reported outcome measures needs to be evaluated in future studies.In summary, PD and TP according to the radical TRIANGLE protocol can be performed without increased morbidity and mortality at a high-volume centre. Therefore, the TRIANGLE operation is feasible and safe in patients with pancreatic head and body tumours. It may have the potential to improve local recurrence and disease-free survival. The efficacy of the TRIANGLE operation as well as its effect on patient well-being and health-related quality of life would merit evaluation in a well-designed multicentre randomized-controlled trial which is currently underway.
Author’s contribution
Rosa Klotz: Conceptualization; Data curation; Methodology; Formal analysis; Writing - original draft.
Thilo Hackert: Conceptualization; Methodology; Supervision; Visualization; Writing - review & editing.
Patrick Heger: Conceptualization; Data curation; Methodology; Writing - review & editing.
Pascal Probst: Conceptualization; Data curation; Methodology; Writing - review & editing.
Ulf Hinz: Conceptualization; Data curation; Methodology; Formal analysis; Visualization; Writing - original draft.
Martin Loos: Methodology; Supervision; Writing - review & editing.
Christoph Berchtold: Conceptualization; Methodology; Supervision; Writing - review & editing.
Arianeb Mehrabi: Conceptualization; Methodology; Supervision; Writing - review & editing.
Martin Schneider: Conceptualization; Methodology; Supervision; Writing - review & editing.
Beat P. Müller-Stich: Conceptualization; Methodology; Supervision; Writing - review & editing.
Oliver Strobel: Conceptualization; Methodology; Supervision; Writing - review & editing.
Markus Diener: Conceptualization; Methodology; Supervision; Writing - review & editing.
André L. Mihaljevic: Conceptualization; Methodology; Supervision; Writing - original draft & editing.
Markus W. Büchler: Conceptualization; Methodology; Visualization; Supervision; Writing - review & editing.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflict of interest
None.
References
- Gesellschaft der epidemiologischen Krebregister in Deutschland e. V. Krebs in Deutschland 2011/2012. E.2015
- Cancer tomorrow.http://gco.iarc.fr/tomorrow/homeDate accessed: November 15, 2020
- Comparison of adjuvant gemcitabine and capecitabine with gemcitabine monotherapy in patients with resected pancreatic cancer (ESPAC-4): a multicentre, open-label, randomised, phase 3 trial.Lancet. 2017; 389: 1011-1024
- FOLFIRINOX or gemcitabine as adjuvant therapy for pancreatic cancer.N Engl J Med. 2018; 379: 2395-2406
- Adjuvant chemotherapy with fluorouracil plus folinic acid vs gemcitabine following pancreatic cancer resection: a randomized controlled trial.J Am Med Assoc. 2010; 304: 1073-1081
- Adjuvant chemotherapy with gemcitabine vs observation in patients undergoing curative-intent resection of pancreatic CancerA randomized controlled trial.J Am Med Assoc. 2007; 297: 267-277
- Impact of resection margin status on recurrence and survival in pancreatic cancer surgery.Br J Surg. 2019; 106: 1055-1065
- Meta-analysis of recurrence pattern after resection for pancreatic cancer.Br J Surg. 2019 Nov; 106: 1590-1601
- Recurrence after neoadjuvant therapy and resection of borderline resectable and locally advanced pancreatic cancer.Eur J Surg Oncol. 2019; 45: 1674-1683
- Actual five-year survival after upfront resection for pancreatic ductal adenocarcinoma: who beats the odds?.Eur J Surg Oncol. 2019 Sep; 45: 1674-1683
- Pancreatic cancer surgery: the new R-status counts.Ann Surg. 2017; 265: 565-573
- A systematic review of the role of periadventitial dissection of the superior mesenteric artery in affecting margin status after pancreatoduodenectomy for pancreatic adenocarcinoma.HPB. 2016; 18: 305-311
- Most pancreatic cancer resections are R1 resections.Ann Surg Oncol. 2008; 15: 1651-1660
- Redefining the R1 resection in pancreatic cancer.Br J Surg. 2006; 93: 1232-1237
- Detailed analysis of margin positivity and the site of local recurrence after pancreaticoduodenectomy.Ann Surg Oncol. 2021 Jan; 28: 539-549
- Borderline resectable pancreatic cancer: a consensus statement by the international study group of pancreatic surgery (ISGPS).Surgery. 2014; 155: 977-988
- Indications and perioperative outcomes for pancreatectomy with arterial resection.J Am Coll Surg. 2018; 227: 255-269
- Arterial resection during pancreatectomy for pancreatic cancer: a systematic review and meta-analysis.Ann Surg. 2011; 254: 882-893
- The TRIANGLE operation - radical surgery after neoadjuvant treatment for advanced pancreatic cancer: a single arm observational study.HPB. 2017; 19: 1001-1007
- Incidence, risk factors and clinical implications of chyle leak after pancreatic surgery.Br J Surg. 2017; 104: 108-117
- Pancreatoduodenectomy with systematic mesopancreas dissection using a supracolic anterior artery-first approach.Ann Surg. 2015; 262: 1092-1101
- The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.Ann Intern Med. 2007; 147: 573-577
- The Clavien-Dindo classification of surgical complications: five-year experience.Ann Surg. 2009; 250: 187-196
- Postpancreatectomy hemorrhage (PPH): an international study group of pancreatic surgery (ISGPS) definition.Surgery. 2007; 142: 20-25
- The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after.Surgery. 2017; 161: 584-591
- Definition and classification of chyle leak after pancreatic operation: a consensus statement by the International Study Group on Pancreatic Surgery.Surgery. 2017; 161: 365-372
- Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS).Surgery. 2007; 142: 761-768
- Bile leakage after hepatobiliary and pancreatic surgery: a definition and grading of severity by the International Study Group of Liver Surgery.Surgery. 2011; 149: 680-688
- The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM.Ann Surg Oncol. 2010; 17: 1471-1474
- Definition of a standard lymphadenectomy in surgery for pancreatic ductal adenocarcinoma: a consensus statement by the International Study Group on Pancreatic Surgery (ISGPS).Surgery. 2014; 156: 591-600
- AWMF. S3-Leitlinie zum exokrinen Pankreaskarzinom. AWMF-Registernummer: 032/010OL, 2013
- Periarterial divestment in pancreatic cancer surgery.Surgery. 2021 May; 169: 1019-1025
- Not all Whipple procedures are equal: proposal for a classification of pancreatoduodenectomies.Surgery. 2021 Jun; 169: 1456-1462
- Top ten research priorities for pancreatic cancer therapy.Lancet Oncol. 2020; 21: e295-e296
- Patterns of recurrence after resection of pancreatic ductal adenocarcinoma: a secondary analysis of the ESPAC-4 randomized adjuvant chemotherapy trial.JAMA Surg. 2019; 154: 1038-1048
- Optimal extent of superior mesenteric artery dissection during pancreaticoduodenectomy for pancreatic cancer: balancing surgical and oncological safety.J Gastrointest Surg. 2019; 23: 1373-1383
- Prognostic significance of dissecting the nerve plexus around the common hepatic artery in pancreatic cancer.Langenbeck’s Arch Surg. 2021 May; 406: 679-689
- A novel classification and staged approach for dissection along the celiac and hepatic artery during pancreaticoduodenectomy.World J Surg. 2018; 42: 2963-2967
- Systematic review and meta-analysis of standard and extended lymphadenectomy in pancreaticoduodenectomy for pancreatic cancer.Br J Surg. 2007; 94: 265-273
- Standard and extended lymphadenectomy for adenocarcinoma of the pancreatic head: a meta-analysis and systematic review.J Gastroenterol Hepatol. 2014; 29: 453-462
- Arterial resection in pancreatic cancer surgery: effective after a learning curve.Ann Surg. 2020; ([Publish Ahead of Print])
- Multivisceral resection for pancreatic malignancies: risk-analysis and long-term outcome.Ann Surg. 2009; 250: 81-87
- No need for routine drainage after pancreatic head resection: the dual-center, randomized, controlled PANDRA trial (ISRCTN04937707).Ann Surg. 2016; 264: 528-537
- Standard versus extended lymphadenectomy in radical pancreatoduodenectomy for ductal adenocarcinoma of the head of the pancreas.J Hepato-Biliary-Pancreatic Sci. 2012; 19: 230-241
Article info
Publication history
Published online: July 06, 2021
Accepted:
June 28,
2021
Received:
December 15,
2020
Identification
Copyright
© 2021 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd.
User license
Elsevier user license | How you can reuse 
Elsevier's open access license policy
Elsevier user license
Permitted
For non-commercial purposes:
- Read, print & download
- Text & data mine
- Translate the article
Not Permitted
- Reuse portions or extracts from the article in other works
- Redistribute or republish the final article
- Sell or re-use for commercial purposes
Elsevier's open access license policy
