Keywords pancreatic cancer - pancreaticoduodenectomy - pancreatic fistula - octreotide - meta-analysis
Pancreaticoduodenectomy (PD) is the only curative option available for the treatment
of pancreatic and periampullary cancers. PD has always been considered to be associated
with a high perioperative morbidly and mortality. With the advancements in the perioperative
critical care and the refinements in the surgical techniques, the postoperative morbidity
following PD has drastically come down to less than 5% in the high volume centers.[1 ] Postoperative pancreatic fistula (POPF) remains a major cause of the morbidity and
the mortality following PD. Because of the lack of uniformity in the definition of
the POPF, there has been a wide variation in its reported incidence ranging from 2%
to more than 20% (2). In 2005, an international study group on pancreatic fistula
(ISGPF), an international panel of pancreatic surgeons working in well-known high-volume
centers, formulated an acceptable and an objective definition of the POPF to decrease
the interobserver variability. They defined the POPF as a drain output of any measurable
volume of fluid on or after the postoperative day 3 with an amylase content of greater
than three times the serum amylase activity.[2 ]
How the POPF can be prevented continues to remain a challenge for the surgeons. Several
surgical modifications have been proposed to reduce the risk of the POPF: (1)the type
of pancreaticointestinal anastomosis (pancreaticogastrostomy versus pancreaticojejunostomy),
(2) the different methods of doing anastomosis (duct to mucosa, invagination), (3)
the use of stents (internal vs. external vs. no stent), and (4) the application of
various topical sealants over the anastomosis.[3 ] A pharmacologic approach using perioperative octreotide, a long-acting somatostatin
analog having an inhibitory action on the pancreatic exocrine secretions, was also
proposed to reduce the incidence of the POPF. Despite contradictory results in various
randomized controlled trials (RCTs), the prophylactic octreotide has been widely used
in the last two decades to reduce the POPF. The present meta-analysis aims to assess
the effectiveness of the prophylactic octreotide in preventing the POPF following
a PD.
Methods
A literature search was performed in the PubMed for the RCTs that compared the prophylactic
octreotide with the placebo following a PD published prior to October 2016. Review
manager (Cochrane Collaboration's software) version RevMan 5.2 was used for the analysis.
Those RCTs, which had compared the prophylactic octreotide with placebo to reduce
the POPF following PD, were considered eligible for the meta-analysis. Low quality
(Jadad score of two or less) RCTs or those including mixed pancreatic resections without
reporting specific PD outcomes were excluded. The effect size for dichotomous and
continuous data was displayed as odds ratio (OR) and weighted mean difference (WMD),
respectively, with their corresponding 95% confidence intervals (CI). A fixed effect
or random effects model was used to pool the data according to the result of a statistical
heterogeneity test. Heterogeneity between studies was evaluated using the Cochran
Q statistic and the I
2 test, with p < 0.05 indicating significant heterogeneity.
Result
There were eight RCTs available for the analysis ([Table 1 ]).[4 ]
[5 ]
[6 ]
[7 ]
[8 ]
[9 ]
[10 ]
[11 ] A total of 959 patients were included in the meta-analysis–492 received the prophylactic
perioperative octreotide and 467 patients received the placebo. The perioperative
octreotide was not found to significantly decrease the total number of the POPF (OR,
1.03; 95% CI: 0.73–1.45; p -value 0.85) or the clinically significant POPF (OR, 0.76; 95% CI: 0.35–1.65; p -value 0.49) compared with the placebo. There was also no difference in the duration
of hospital stay (WMD, 1.19; 95% CI:1.84–4.23; p -value 0.44) or the postoperative mortality (OR, 2.04; 95% CI: 0.87–4.78; p -value 0.10) in two groups. The perioperative octreotide was also not found to significantly
delay the gastric emptying (OR, 0.76; 95% CI: 0.41–1.40 p -value 0.38). [Fig. 1 ] displays the forest plots of the eight RCTs comparing the role of the prophylactic
octreotide to prevent the POPF compared with the placebo.
Fig. 1 Displays the forest plots of eight RCTs comparing the prophylactic role of octreotide
to prevent the POPF compared with placebo. CI, confidence interval; df, degrees of
freedom; M-H, Mantel-Haenszel.
Table 1
Characteristics of the included studies
Author
Y of publication
Sample size
Jadad score
Dose and route of octreotide
Definition of the POPF
THE POPF (octreotide vs. control)
Kurumboor et al,
2015
109
3
100 microgram octreotide SC, 8 h for 5 d
As per ISGPF definition
33/55 vs. 34/54 (p = 0.626)
Fernández-Cruz et al
2013
62
3
100 microgram octreotide SC, 8 h for 10 d
As per ISGPF definition
2/32 vs. 3/30, (p = 0.769)
Kollmar et al
2008
65
5
100 microgram octreotide SC, 8 h for 7 d
As per ISGPF definition
9/35 vs. 6/32 (p = NS)
Hesse et al
2005
80
3
100 microgram octreotide SC, 8 h for 7 d
>100 mL/day of amylase rich fluid (> 5 times upper limit of normal serum amylase after
d 3, persisting after d 7 with rising temperature and preseptic condition
5/41 vs. 3/39 (p = NS)
Suc et al
2004
177
4
100 microgram octreotide SC, 8 h for 10 d
Amylase rish fluid (> 4 times upper limit of normal serum amylase after d 3
18/92 vs. 18/85 (p = NS)
Yeo et al
2000
211
5
250 microgram octreotide SC, 8 h for 7 d
> 50 mL/day of amylase rich fluid (> 3 times upper limit of normal serum amylase after
d 10, or pancreatic anastomotic disruption demonstrated radiologically
11/104 vs. 10/107 (p = NS)
Lowy et al
1997
110
3
150 microgram octreotide SC, 8 h for 5 d
Amylase rich fluid (> 2.5 times upper limit of normal serum amylase after d 3
16/57 vs. 11/53 (p = 0.23)
Montorsi et al
1995
143
4
100 microgram octreotide SC, 8 h for 7 d
> 10 mL/day of amylase rich fluid (> 3 times of normal serum amylase after d 3
8/76 vs. 10/67 (p = NS)
Abbreviations: ISGPF, International Study Group for Pancreatic Fistula; NS, not significant;
POPF, postoperative pancreatic fistula; SC, subcutaneously.
Discussion
The rationale for using somatostatin in the prevention of the POPF is based on their
ability to decrease the output of secretions from the pancreas as a high pancreatic
juice output in a soft pancreas is an important risk factor for the POPF.[12 ] As the half-life of the somatostatin is short at approximately 2 minutes, the synthetic
analogues of the somatostatin with longer half-lives, such as octreotide, have been
developed and used in pancreatic surgery in an attempt to decrease the POPF, with
the hypothesis that decreased pancreatic juice secretion would allow for an improved
healing of the pancreatic ductal anastomoses and consequently would decrease the leak
rates. The use of the octreotide has been studied in multiple randomized prospective
trials in the United States and Europe; however, the results have been contradictory.
The current meta-analysis fails to support the prophylactic role of octreotide in
reducing the chances of the POPF, delayed gastric emptying (DGE), duration of hospital
stay, or mortality. However, the results of this meta-analysis should be viewed in
the light of potential confounding factors–a nonuniform definition of the POPF, a
different dosing regimen of the octreotide employed, the surgical technique and experience
of different surgeons, and other high-risk factors for the POPF. A nonuniform definition
of the POPF has been a major barrier in evaluating the results of various studies
addressing the issue of the POPF; however, a widespread acceptance of the ISGPF definition
of the POPF has resulted in bringing some homogeneity in the studies. All three of
the eight RCTs published after the year 2005 used the ISGPF definition of the POPF.
Various dosing regimen is another concern while comparing the results of the prophylactic
octreotide across the studies. Both the dose (100–250 micrograms) and duration (5–10
days) of the prophylactic octreotide were used by authors differently in various studies.
Yeo et al[4 ] used a high dose of octreotide (250 micrograms three times a day) in their RCT of
211 patients. They also reported the ineffectiveness of the prophylactic octreotide
following PD in reducing the POPF (9 vs. 11%), the overall complication rates (34
vs. 40%), the in-hospital death rates (0 vs. 1%), and the duration of hospital stay
(9 vs. 9 days). The authors admitted that they had used a higher dose intentionally
so as not be blamed for using an inadequate dose of the prophylactic octreotide had
their RCT turned out to be a negative one.
Various RCTs have been criticized for not assessing the role of the prophylactic octreotide
in those patients having high-risk factors for the POPF. Though several factors have
been implicated in the formation of the POPF, nondilated pancreatic duct and soft
pancreas are widely reported as high-risk factors for the POPF.[6 ]
[13 ]
[14 ] A study of 109 patients undergoing PD with soft pancreas and nondilated pancreatic
duct did not show any statistically significant difference in rates of the pancreatic
fistula with the use of the prophylactic octreotide (Total POPF, 60 vs. 63%).[6 ]
Pasireotide is a newer generation long-acting somatostatin analog which is being evaluated
to decrease the POPF following PD. However, the results of the prophylactic pasireotide
have also been variable till now. A single-center, randomized, double-blind trial
of perioperative subcutaneous pasireotide in patients undergoing PD showed a significant
reduction in the rates of grade three or higher postoperative pancreatic fistula,
leak, or abscess among patients who received pasireotide compared with the patients
who received placebo (10 vs. 21%; relative risk, 0.49; 95% CI: 0.25–0.95).[15 ] However, another recently published prospective case-control study revealed no difference
in the pancreatic fistula, the overall complications, the 90-day readmission or the
90-day mortality in the patients who received the pasireotide.[16 ] However, it would be too early to arrive to a conclusion regarding the efficacy
of the prophylactic pasireotide following PD.
Interestingly, the present meta-analysis suggests a trend toward a higher mortality
in the patients receiving prophylactic octreotide though it failed to attain a statistical
significance (OR, 2.04; 95% CI: 0.87–4.78; p -value 0.10). However, this may prove to be a final nail in the coffin of use of prophylactic
octreotide following PD. Several reasons may be attributed to this unexpected outcome
as octreotide causes: (1) a reduction in the splanchnic blood flow; (2) suppression
of the secretion of anabolic and tropic hormones such as pituitary growth hormone
(GH), insulin-like growth factor 1, and epidermal growth factor; and (3) a reduction
in the volume of the pancreatic juice resulting in large fluctuations in the enzyme
concentration.[17 ]
Despite many limitations, the accumulated evidence through the present meta-analysis
indicates that the prophylactic octreotide has failed to show any demonstrable benefit
in reducing the POPF. Moreover, elimination of the prophylactic octreotide would also
be helpful as a considerable cost saving measure. The time has come to look for other
newer modalities including both pharmacological and nonpharmacological to address
the issue of the POPF which continues to haunt the surgeons performing PD.
Conclusion
The present meta-analysis does not support the role of the prophylactic octreotide
to prevent the POPF following PD.
