Pain Relief by Surgery in Chronic Pan cr eatit i s ? Relationship between Pain Relief, Pancreatic Dysfunction, and Alcohol Withdrawal R. W. AMMANN, F. LARGIADER & A. AKOVBIANTZ
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
Dept. of Internal Medicine, Gastroenterology Service, Dept. of Surgery A, University Hospital, and Dept. of Surgery, City Hospital Waid, Zurich, Switzerland Ammann. R. W., Largiader, F. & Akovbiantz, A. Pain relief by surgery in chronic pancreatitis? Relationship between pain relief. pancreatic dysfunction, and alcohol withdrawal. Scand. J . Gastroent. 1979, 14, 209-215. Since 1963, 57 consecutive patients with chronic pancreatitis, 44 of them alcoholics who had been operated upon for recurrent severe pain, have been controlled regularly for an average of 6 years. Thirty-two of them had a cyst drainage procedure (group A), and 25 had a ductal drainage procedure and/or distal pancreatectomy (group B). Ten patients died within 2 years (group A, n = 5). Lasting pain relief by surgery occurred in 19 patients only. Of 28 patients with pain relapses after surgery (group A, n = 15), however, 22 (78.6%) obtained late pain relief 1-8 years after surgery in association with marked increase of pancreatic dysfunction (group A, n = 12). Pain relief was associated with pancreatic calcifications in 7 1-86% of the alcoholics. Cyst drainage procedures were successful in preventing pain relapses mainly in patients with either advanced pancreatic dysfunction or in non-alcoholic pancreatitis. The data suggest that in chronic pancreatitis lasting pain relief is more often due to marked pancreatic dysfunction than to surgery. Alcohol abstinence after surgery was probably an additional factor for lasting pain relief in some patients.
Key-words: Disease history, natural; pain relief; pancreas, calcifications; pancreas, dysfunction; pancreatitis, chronic; surgery, long-term results Prof. Dr. R . W. Ammann. Division of Gastroenterology, Dept. of Internal Medicine, University Hospital, CH-8091Zurich, Switzerland
Many different surgical procedures for the treatment of severe recurrent pain in chronic relapsing pancreatitis (CP) have been proposed (12, 16, 19,28). Despite wide experience no agreement on tactic and technic of surgery in C P has been achieved (8, 11, 12, 18, 22, t 5 , 227).Empfricid experience has not solved this problem so far, because our knowledge of the natural history and particularly of the factors responsible for pain of CP is still limited. The natural history of C P is characterized by recurrent pain episodes and by progressive exocrine and endocrine insufficiency (3, 22). In association with marked pancreatic dysfunction, uncomplicated C P has a tendency to become painless spontaneously ('burning out') (2, 5). It seemed, therefore, interesting to study the problem of pain relief after pancreatic surgery in relation to pancreatic function.
In addition, other factors that may influence longterm results of surgery were analyzed, particularly etiology, postoperative alcohol intake, and the presence of pancreatic calcifications.
PATIENTS AND METHODS Since 1963, 5 7 consecutive patients with C P have been regularly controlled after pancreatic surgery. Recurrent severe pain was the main indication for surgery. This study is part of a larger long-term investigation of the natural history of CP with special interest in the relationship between the pain profile and pancreatic dysfunction (3, 5). The diagnostic criteria of CP, which are in accordance with the definition of Marseille, have been outlined previously (3, 5).
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
210
R . W . Ammann, F. Largiad6r & A . Akovbiantz
The regular postoperative controls, on the average once a year, included clinical evaluation, routine blood tests, function tests (glucose tolerance test (GTT)), fecal chymotrypsin estimation (3), chemical fecal fat determination, and radiology, particularly plain films of the pancreas in three projections. Alcohol intake was carefully evaluated at each visit by interrogation of the patient and family members and with the help of social workers. Patients with a daily alcohol intake of 8 0 g or more were considered to have alcoholic CP. CP was interpreted as idiopathic after repeated exclusion of all known factors. The patients were divided into two groups according to the surgical procedure. Group A comprised 32 patients with a pseudocyst drainage procedure (n = 27) (cystojejunostomy) or sphincterotomy (n = 5). Group B included 25 patients with a ductojejunostomy (n = 16) (combined with a distal resection of 40-60% in 6) or with distal resection alone ( n = 9 ) . All patients with a duct drainage procedure had a markedly dilated pancreatic duct, and a longitudinal pancreatico-jejunostomy was performed in 1 1 cases. Thirty-nine of the 5 7 patients were operated either by two of us (F.L. and A.A.) or by senior staff members working at or trained in Dept. of Surgery A, University Hospital. A ‘function index’ (FI) has been introduced for quantitative assessment of the severity of pancreatic dysfunction. Endocrine function was graded as fol-
lows: 0 (= normal GTT), 1 (pathological GTT), 2 (diabetes with need of diet), 3 (diabetes with need of antidiabetics). Exocrine function was graded as follows: 0 (= fecal chymotrypsin at least twice above 120 &g), 1 (= fecal chymotrypsin repeatedly below 120 pg!g), 2 (= fecal chymotrypsin repeatedly below 40 pg/g), 3 (= fecal fat above 7 g/24 h and fecal chymotrypsin below 40 pg/g or fecal chymotrypsin output below 5 mg/24 h). Thus a function index of 6 means pancreatic steatorrhea and severe diabetes. Results are given as means standard error. RESULTS The relevant clinical data are summarized in Table I. The mean total and postoperative follow-up are comparable in both groups. Pancreatic calcifications were demonstrated only after surgery in 8 of 22 cases of group A and in 6 of 20 cases of group B. The postoperative results will be discussed separately for alcoholic and non-alcoholic CP of both groups. Group A Alcoholics (n = 26). Four died within 2 years, two of them postoperatively. Of the remaining 22 patients 7 obtained pain relief from the operation (cyst drainage in 6) in association with marked pancreatic dysfunction (FI 4.6 t 0.58) (Fig. 1). Pancreatic calcifications at the operation (n = 5 ) or
Table I. Clinical data of surgical group A (cyst drainage) and group B (ductal drainage and/or distal resection)
No. of patients Sex Age (mean) (at time of op.) Range EtiolOgy Alcoholic C P Non-alcoholic C P (idiopathic) Calcifications of pancreas Mean follow-up (in years) Total (from onset of symptoms) Postoperatively
Group A
Group B
32 29 M, 3 F 38.9 years 24-58
25 23M, 2 F 35.4 years 18-63
26 6 22 (68.8%)
18
11.5
(4-29) 6.7 (1-29)
I 20 (80.0%) 10.25 (3-19) 5.9 (1-16)
Pain Relief in Chronic Pancreatitis
21 1
FI
*-.
LASTING PAIN RELIEF CALCIFICATIONS OF PANCREAS AT. ALCOHOL ABSTINENCE A ALCOHOLIC CP I IDIOPATHIC CP
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
@
e
-1 I
I
2
I
I
4
I
1
6
I
1
8
I
I
10
I
I
!!-
12 r13 YEARS (postoperatively)
Fig. 1. Group A (cyst drainage, surgical success). Lasting pain relief was obtained by 12 patients, 7 of them alcoholics (* 2 ) In idiopathic . CP, pain relief by surgery seemed unrelated to the degree of pancreatic dysfunction (FI ,< 4.0 in all 5 cases). In alcoholic CP, however, pain relief was associated either with marked pancreatic dysfunction (in 4 cases FI 4.0) and/or alcohol abstinence. Six of 7 alcoholics also had pancreatic calcifications at the operation ( 0 )or later in the course
>
(a).
postoperatively (n = 1 ) were demonstrated in six of follow-up (Fig. 1). Pancreatic calcifications were seven patients (85.7%). Four of the seven patients demonstrated in only two of five patients. discontinued alcohol abuse postoperatively. The other 15 patients with a low postoperative F I Group B (2.04 f 0.26), four of them with calcifications Alcoholics (n = 18). Three patients died post(26.7%) and all with continuous alcohol abuse, operatively. Of the remaining 15 patients, 7 obsuffered recurrent pain attacks ( 3 had had a sphinc- tained lasting pain relief from the operation despite a terotomy). Twelve of these patients did, however, rather low postoperative FI (2.57 ? 0.48) (Fig. 3). obtain late pain relief within an average of 3.9 years Six of the seven patients discontinued alcohol intake postoperatively in association with marked pan- postoperatively. Pancreatic calcifications at the opcreatic dysfunction (FI 4.79 ? 0.22) (Fig. 2). Pan- eration were found in six cases (85.7%).During the creatic calcifications at the operation (n = 4) or follow-up period six of seven patients showed a postoperatively (n = 7 ) were demonstrated in 1 1 of progressive increase of pancreatic dysfunction the 15 patients or in 9 of the 12 patients (75%) who (Fig. 3). obtained late pain relief. The three patients who The remaining eight patients Suffered postoperacontinued to have pain relapses had a shorter follow- tively from recurrent pain episodes; all continued up and less marked pancreatic dysfunction (Fig. 2). alcohol abuse, their FI being low (2.31 ? 0.47). Non-alcoholics (n = 6). One patient died within Only two had pancreatic calcifications (25%). 1.5 years after sphincterotomy. The remaining five Seven of the eight patients did, however, obtain late patients obtained pain relief from the operation 'pain relief within an average of 2.7 years postoperadespite a rather low FI (2.33 ? 0.88) (Fig. 1). The tively in association with severe pancreatic dysfuncFI increased in four of five patients during the tion (FI 5.0 0.48) (Fig. 4). Pancreatic calcifica-
R . W. Ammann, F. Largiader & A . Akovbiantz
212
FI 61
4i I 4
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
@ -
-
2
PAIN CONTINUED PAIN RELAPSES LASTING PAIN RELIEFCALCIFICATIONS OF PANCREAS @@
-*
2
4
8
6
10 12 YEARS (postoperatively)
Fig. 2. Group A (cyst drainage, surgical failure). Postoperative pain relapses in relation to pancreatic dysfunction. Fifteen patients with alcoholic C P and persisting alcohol abuse who had a low postoperative FI (g4.0)relapsed after surgery ( L O , L*). Twelve of them obtained late pain relief within 1-7 years postoperatively in association with marked pancreatic dysfunction (FI 4.0) (* 2 ) Nine . of the 12 had calcifying CP (0, 0 ) .The time relationship between operation, degree of dysfunction and pain indicates that pain relief occurred probably secondary to the increase of pancreatic dysfunction and despite alcohol abuse (see also Fig. 4).
>
FI
I-
1
I
I
2
I
I
4
I
I
6
I
I
8
I
//-----
10
-13 YEARS (postoperatively)
Fig. 3. Group B (ductal drainage, surgical success). Lasting pain relief (* 2 ) in relation to alcohol abstinence. Six of the 7 patients with lasting pain relief abstained from alcohol postoperatively. Pain relief was obtained by 5 patients despite a low FI (Q3.0), indicating that alcohol withdrawal was probably the major factor for pain relief (see also Fig. 1). Six of the 7 patients also had pancreatic calcifications at the operation (0).
Pain Relief in Chronic Pancreatitis
213
FI mA/R
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
I/R
-
PAIN CONTINUED PAIN RELAPSES LASTING PAIN RELIEF*CALCIFICATIONS OF PANCREAS @8
-*
A ALCOHOLIC CP I IDIOPATHIC CP R RESECTION (40-60%1 0 DUCTAL DRAINAGE *l/D
I
I
2
I
I
I
4
I
I
6
I
8
I
I
I
I
10 12 YEARS (postoperatively)
Fig. 4. Group B (ductal drainage and/or distal resection, surgical failure). Postoperative pain relapses in relation to pancreatic dysfunction. Thirteen patients, 8 of them with alcoholic C P and continuing alcohol abuse (with a low postoperative FI ($3.0) in all but 4 cases) relapsed after surgery ( L O . 0*). Late pain relief was obtained by 10 of 13 patients, 7 of them alcoholics, within 1-8 years postoperatively in association with an increase of pancreatic dysfunction (* 2). The time relationship between operation, degree of pancreatic dysfunction, and pain indicates that pain relief occurred probably secondary to the increase in pancreatic dysfunction (see also Fig. 2). Pancreatic resection probably enhanced pancreatic dysfunction in some cases. Pancreatic calcifications were present in 5 of the 7 alcoholics at the operation ( 0 )or later in the course ( 0 ,0).
tions were demonstrated in five of seven patients (71.4%) who obtained late pain relief. Non-alcoholics (n = 7). Two patients died postoperatively. None of the remaining five patients obtained pain relief from the operation, the FI being low (2.8 k 0.8) (Fig. 4). Three of the five patients, however, obtained late pain relief in association with severe pancreatic dysfunction (5.3 k 0.67) within an average of 3.3 years. Pancreatic calcifications at the operation (n = 3) or postoperatively (n = 2) were demonstrated in all five patients.
DISCUSSION The principal aim of surgery in C P is to relieve pain while preserving as much functioning pancreatic tissue as possible (1 2, 28). Since the factors responsible for pain are poorly defined, the surgical method of choice is still debated ( 1 1, 12, 18, 19, 28). Pancreatico-jejunostomy does not markedly increase
pancreatic dysfunction but achieves pain relief only in about tyo-thirds of cases (14). The better results with reg&d to pain relief by resecting procedures are negatively influenced by the marked increase of pancreatic dysfunction (and by the late mortality) (12, 19). Thus the question arises whether the increase of pancreatic dysfunction induced by surgery may not promote pain relief. This hypothesis is supported by our findings. The time relationship between surgery, degree of pancreatic dysfunction (k alcohol abstinence), and pain will be discussed separately in order to evaluate the relative effect of these factors on pain relief after surgery. No attempts have been made so far to correlate pancreatic dysfunction with the postoperative pain profile. There are several reasons for the lack of adequate data on postoperative exocrine function. The CCK-secretin test often cannot be performed properly because of the deranged anatomy after surgery. Steatorrhea, on the other hand, is a rather insensitive and non-specific criterion, since it occurs
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
214
R. W. Ammann, F. LargiadPr & A . Akovbiantz
only in advanced CP with a decrease of lipase secretion by more than 90% (7, 17) and since other factors, particularly gastric resections, increase the fecal fat excretion ( 15, 29). The diagnostic value of the fecal chymotrypsin method, which is particularly useful for long-term studies of exocrine function in CP (3, 5), has been thoroughly evaluated recently (3, 5,9,26).A comparable high percentage of pathological results with the fecal chymotrypsin method and with the CCK-secretin test was found in CP with advancing insufficiency (3). Since the endocrine function may be affected disproportionally more than the exocrine function after distal pancreatectomy (1 2,25), the FI, which expresses the overall pancreatic dysfunction in one figure, appears to be a valuable criterion.
ingly, the good results with regard to pain by pancreatic duct ligation in CP support our thesis of a direct relationship between pancreatic dysfunction and pain relief (6, 16, 21). The current concept that the good results observed after resection of the mostly inflamed segment of the pancreas are due to elimination of peripancreatic and perineural inflammation has to be questioned (8, 1 1 , 12, 19). All available data suggest rather that marked pancreatic dysfunction is a major factor in lasting pain relief in uncomplicated CP. Pancreatic calciJicationand pain relief after surgery Alcoholic CP with calcifications seems to do better after surgery in our experience than noncalcifying CP, which is in accordance with results first reported by Duval & Enquist (10) and confirmed by others (1 8, 19). That is not surprising because there is a close relationship between progressive pancreatic dysfunction and pancreatic calcifications in alcoholic CP (22) (Figs. 1-4). In nonalcoholic CP the relationship between pancreatic calcifications and operative results is less evident, since in group A only two of five patients with postoperative pain relief had calcifications and in group B all five patients relapsed after surgery despite pancreatic calcifications.
Pancreatic dysfunction and pain relief after surgery If it is assumed that pain in CP is mainly due to stasis in the cysts (group A) or in the ductal system (group B), lasting pain relief should be obtained by appropriate drainage procedures that do not markedly affect pancreatic function. In alcoholic CP over 60%of patients of groups A + B (23 of 3 7) had pain relapses after surgery, but 19 of them obtained late pain relief 1-8 years postoperatively in association with marked increase of pancreatic dysfunction (Figs. 2 and 4). In group B all five non-alcoholic patients relapsed postoperatively, but three of them Pseudocysts and results of drainage procedures Successful cyst drainage procedures in alcoholic obtained late pain relief 1-8 years after surgery parallel with the increase of the FI (Fig. 4). Lasting CP usually do not prevent recurrent episodes of pain relief was also achieved with the operation in pancreatitis (8, 12,25) except in patients either with group A in four alcoholic CP patients with marked advanced pancreatic dysfunction (see above) or pancreatic dysfunction (Fig. 1). These data strongly with alcohol withdrawal (see below). In contrast, all suggest that credit for pain relief cannot be given to five patients with non-alcoholic CP of group A surgery in these groups of patients. The results, on obtained lasting pain relief from the operation the contrary, seem to confirm that spontaneous pain (Fig. I). In this group pain relief must be attributed relief parallel with progressive pancreatic dysfunc- directly to the cyst drainage procedure, since.at a tion is part of the natural history of CP (2, 4, 5). A similar low postoperative FI, patients with alcoholic steady increase of pancreatic dysfunction over the CP continued to relapse. years is also seen in most operated cases of alcoholic and idiopathic CP (Figs. 1-4). Alcohol abstinence and surgical results Spontaneous disappearance of pain in CP has As shown above, pain relief through surgery is been reported previously but without reference to likely to occur in alcoholic CP with advanced panpancreatic dysfunction (1, 13, 20, 23, 24). It is creatic dysfunction (high FI). Do all patients with obvious that in complicated CP (e.g. with pseudo- alcoholic CP and a low FI develop postoperative cysts), pain may persist or recur over years despite pain relapses? Obviously not. In our experience marked pancreatic dysfunction (12, 22). Interest- alcohol abstinence postoperatively was probably
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
Pain Relief in Chronic Pancreatitis
important for pain relief in three cases in group A with low FI (Fig. 1 ) and in seven alcoholic C P patients in group B without postoperative pain relapses despite a low FI after surgery (Fig. 3). Six of these seven patients discontinued alcohol abuse, while eight alcoholics in group B with a similar low FI who continued alcohol abuse suffered pain relapses (Fig. 4). Since pain tends to subside spontaneously with alcohol abstinence in uncomplicated C P in 30-50% or more of the patients (18,22,2527), many experts reject surgery in uncomplicated alcoholic C P before the effect of complete abstinence on pain has been evaluated (12, 19, 22, 25). Results of ductal decompressionand distal resection Ductal stasis seems to be an important factor of pain in C P (8, 18, 19, 28). Thirteen patients in group B had a ductal drainage procedure, five of them combined with distal resection (Figs. 3 and 4). Eight of the 13 patients had pain relapses after surgery (Fig. 4), but 6 of them obtained late pain relief 1 year or more postoperatively in association with marked increase of pancreatic dysfunction (Fig. 4). Three of these six patients also had a distal pancreatectomy. Of seven patients with distal resection alone, two became pain-free by surgery (Fig. 3), and four obtained late pain relief 2-8 years postoperatively in association with an increase of pancreatic dysfunction (Fig. 4). In our experience the operative result of ductojejunostomy and/or distal resection seems to be infiuenced primarily by the severity of pancreatic dysfunction, which is enhanced by partial pancreatectomy. In conclusion, our data suggest that the outcome with regard to pain relief by surgery in uncomplicated C P depends predominantly on pancreatic dysfunction, presence of pancreatic calcifications, and alcohol abstinence. In complicated C P (e.g. pseudocysts), the indication for surgery is not debated, but for lasting pain relief these factors are also important, at least in alcoholic CP. ACKNOWLEDGEMENTS Presented in part at the meeting of the European Pancreatic Club, 8 October 1977, in Dublin. SupReceived 16 May 1978 Accepted 1 November 1978
215
ported in part by a grant from the Hartmann-Muller Foundation for Medical Research. REFER ENC ES 1. Adharn, N. F., Dyce, B. & Haverback, B. J. Amer. J . Dig. Dis. 1968, I S , 8-15 2. Arnrnann, R. Schweiz. Med. Wschr. 1970, 95, 1-7 3. Amrnann, R. Leber Magen Darm 1976,6,217-226 4. Ammann, R. Z . Gastroent. 1978, Suppl. 13, 42-48 5. Ammann, R., Hammer, B. & Fumagalli, I. Digestion 1973, 9, 404-415 6. Cannon, J. A. Amer. J . Surg. 1955, 90, 266-279 7. Di Magno, E. P., Go, V. L. W. & Surnmerskill, W. H. J. New. Engl. J . Med. 1973, 288, 813-815 8. Dixon, J. A. & Englert, E. Gastroenterology 1971, 61, 375-381 9. Diirr, H. K., Schneider, R. & Bode, J. Ch. pp. 17-29 in Bartelheimer, H., Classen, M. & Ossenberg, F. W. (eds.) Die Untersuchung der Bauchspeicheldruse. Thieme, Stuttgart, 1976 10. Duval, M. K. & Enquist. I. F. Surgery 1961, 50, 965-969 11. Editorial. Lancet 1977. I , 460-462 12. Frey, Ch. F., Child, Ch. G. & Fry, W. Ann. Surg. 1976. 184, 403-412 13. Gambill, E. E., Baggenstoss, A. H. & Priestley, J. T. Gastroenterology 1960. 39, 404-41 I 14. Govaerts, J. P., Kestens, P. J. & Kiekens, R. et al. Acta Gastroent. Belg. 1976, 39. 595-602 15. Hillrnann, H. Gut 1968, 9, 576-584 16. Hoffrnann, E., Usmiani, J. & Gebhardt, Ch. Dtsch. Med. Wschr. 1977, 102, 392-395 17. Hotz, J., Goberna,R. & Clodi,Ph. Digestion 1973,9, 212-223 18. Jordan, G. L., Strug, B. S . & Crowder, W. E. Amer. J. Surg. 1977, 133, 46-50 19. Leger, L., Lenriot, J. P. & Lemaigre, G. Ann. Surg. 1974, 180, 185-191 20. Levrat, L., Moulinier, B. et al. Arch. Mal. Appar. Dig. 1970, 59, 5-18 2 1. Maddig, G. F. & Kennedy, P. A. Amer. J . Surg. 1973, 125, 538-541 22. Marks, I. N., Bank. S. & Barbezat, G. 0. Leber Magen Darm 1976, 6, 257-270 23. Mc Elroy, R. & Christiansen, Ph. A. Amer. J. Med. 1972, 52, 228-24 1 24. Sarles, H. & Gerolmi-Santandrea, A. Clin. Gastroent. 1972, I , 167-193 25. Sarles, J. C. & Sarles, H. LeberMagen Darm 1976,6, 294-299 26. Schneider, R., Durr, H. K. & Bode, J. Ch. Dtsch. Med. Wschr. 1974. 99, 1449-1454 27. Strum, W. B. & Spiro, H. M. Ann. Intern. Med. 197 1, 74, 264-277 28. White, Th. T. & Keith, R. G. Surg. Gynec. Obstet. 1973, 136, 353-358 29. Wollaeger, E. E., Waugh, J. M. & Power, M. H. Gastroenterology 1963, 44, 25-32
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
Dept. of Internal Medicine, Gastroenterology Service, Dept. of Surgery A, University Hospital, and Dept. of Surgery, City Hospital Waid, Zurich, Switzerland Ammann. R. W., Largiader, F. & Akovbiantz, A. Pain relief by surgery in chronic pancreatitis? Relationship between pain relief. pancreatic dysfunction, and alcohol withdrawal. Scand. J . Gastroent. 1979, 14, 209-215. Since 1963, 57 consecutive patients with chronic pancreatitis, 44 of them alcoholics who had been operated upon for recurrent severe pain, have been controlled regularly for an average of 6 years. Thirty-two of them had a cyst drainage procedure (group A), and 25 had a ductal drainage procedure and/or distal pancreatectomy (group B). Ten patients died within 2 years (group A, n = 5). Lasting pain relief by surgery occurred in 19 patients only. Of 28 patients with pain relapses after surgery (group A, n = 15), however, 22 (78.6%) obtained late pain relief 1-8 years after surgery in association with marked increase of pancreatic dysfunction (group A, n = 12). Pain relief was associated with pancreatic calcifications in 7 1-86% of the alcoholics. Cyst drainage procedures were successful in preventing pain relapses mainly in patients with either advanced pancreatic dysfunction or in non-alcoholic pancreatitis. The data suggest that in chronic pancreatitis lasting pain relief is more often due to marked pancreatic dysfunction than to surgery. Alcohol abstinence after surgery was probably an additional factor for lasting pain relief in some patients.
Key-words: Disease history, natural; pain relief; pancreas, calcifications; pancreas, dysfunction; pancreatitis, chronic; surgery, long-term results Prof. Dr. R . W. Ammann. Division of Gastroenterology, Dept. of Internal Medicine, University Hospital, CH-8091Zurich, Switzerland
Many different surgical procedures for the treatment of severe recurrent pain in chronic relapsing pancreatitis (CP) have been proposed (12, 16, 19,28). Despite wide experience no agreement on tactic and technic of surgery in C P has been achieved (8, 11, 12, 18, 22, t 5 , 227).Empfricid experience has not solved this problem so far, because our knowledge of the natural history and particularly of the factors responsible for pain of CP is still limited. The natural history of C P is characterized by recurrent pain episodes and by progressive exocrine and endocrine insufficiency (3, 22). In association with marked pancreatic dysfunction, uncomplicated C P has a tendency to become painless spontaneously ('burning out') (2, 5). It seemed, therefore, interesting to study the problem of pain relief after pancreatic surgery in relation to pancreatic function.
In addition, other factors that may influence longterm results of surgery were analyzed, particularly etiology, postoperative alcohol intake, and the presence of pancreatic calcifications.
PATIENTS AND METHODS Since 1963, 5 7 consecutive patients with C P have been regularly controlled after pancreatic surgery. Recurrent severe pain was the main indication for surgery. This study is part of a larger long-term investigation of the natural history of CP with special interest in the relationship between the pain profile and pancreatic dysfunction (3, 5). The diagnostic criteria of CP, which are in accordance with the definition of Marseille, have been outlined previously (3, 5).
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
210
R . W . Ammann, F. Largiad6r & A . Akovbiantz
The regular postoperative controls, on the average once a year, included clinical evaluation, routine blood tests, function tests (glucose tolerance test (GTT)), fecal chymotrypsin estimation (3), chemical fecal fat determination, and radiology, particularly plain films of the pancreas in three projections. Alcohol intake was carefully evaluated at each visit by interrogation of the patient and family members and with the help of social workers. Patients with a daily alcohol intake of 8 0 g or more were considered to have alcoholic CP. CP was interpreted as idiopathic after repeated exclusion of all known factors. The patients were divided into two groups according to the surgical procedure. Group A comprised 32 patients with a pseudocyst drainage procedure (n = 27) (cystojejunostomy) or sphincterotomy (n = 5). Group B included 25 patients with a ductojejunostomy (n = 16) (combined with a distal resection of 40-60% in 6) or with distal resection alone ( n = 9 ) . All patients with a duct drainage procedure had a markedly dilated pancreatic duct, and a longitudinal pancreatico-jejunostomy was performed in 1 1 cases. Thirty-nine of the 5 7 patients were operated either by two of us (F.L. and A.A.) or by senior staff members working at or trained in Dept. of Surgery A, University Hospital. A ‘function index’ (FI) has been introduced for quantitative assessment of the severity of pancreatic dysfunction. Endocrine function was graded as fol-
lows: 0 (= normal GTT), 1 (pathological GTT), 2 (diabetes with need of diet), 3 (diabetes with need of antidiabetics). Exocrine function was graded as follows: 0 (= fecal chymotrypsin at least twice above 120 &g), 1 (= fecal chymotrypsin repeatedly below 120 pg!g), 2 (= fecal chymotrypsin repeatedly below 40 pg/g), 3 (= fecal fat above 7 g/24 h and fecal chymotrypsin below 40 pg/g or fecal chymotrypsin output below 5 mg/24 h). Thus a function index of 6 means pancreatic steatorrhea and severe diabetes. Results are given as means standard error. RESULTS The relevant clinical data are summarized in Table I. The mean total and postoperative follow-up are comparable in both groups. Pancreatic calcifications were demonstrated only after surgery in 8 of 22 cases of group A and in 6 of 20 cases of group B. The postoperative results will be discussed separately for alcoholic and non-alcoholic CP of both groups. Group A Alcoholics (n = 26). Four died within 2 years, two of them postoperatively. Of the remaining 22 patients 7 obtained pain relief from the operation (cyst drainage in 6) in association with marked pancreatic dysfunction (FI 4.6 t 0.58) (Fig. 1). Pancreatic calcifications at the operation (n = 5 ) or
Table I. Clinical data of surgical group A (cyst drainage) and group B (ductal drainage and/or distal resection)
No. of patients Sex Age (mean) (at time of op.) Range EtiolOgy Alcoholic C P Non-alcoholic C P (idiopathic) Calcifications of pancreas Mean follow-up (in years) Total (from onset of symptoms) Postoperatively
Group A
Group B
32 29 M, 3 F 38.9 years 24-58
25 23M, 2 F 35.4 years 18-63
26 6 22 (68.8%)
18
11.5
(4-29) 6.7 (1-29)
I 20 (80.0%) 10.25 (3-19) 5.9 (1-16)
Pain Relief in Chronic Pancreatitis
21 1
FI
*-.
LASTING PAIN RELIEF CALCIFICATIONS OF PANCREAS AT. ALCOHOL ABSTINENCE A ALCOHOLIC CP I IDIOPATHIC CP
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
@
e
-1 I
I
2
I
I
4
I
1
6
I
1
8
I
I
10
I
I
!!-
12 r13 YEARS (postoperatively)
Fig. 1. Group A (cyst drainage, surgical success). Lasting pain relief was obtained by 12 patients, 7 of them alcoholics (* 2 ) In idiopathic . CP, pain relief by surgery seemed unrelated to the degree of pancreatic dysfunction (FI ,< 4.0 in all 5 cases). In alcoholic CP, however, pain relief was associated either with marked pancreatic dysfunction (in 4 cases FI 4.0) and/or alcohol abstinence. Six of 7 alcoholics also had pancreatic calcifications at the operation ( 0 )or later in the course
>
(a).
postoperatively (n = 1 ) were demonstrated in six of follow-up (Fig. 1). Pancreatic calcifications were seven patients (85.7%). Four of the seven patients demonstrated in only two of five patients. discontinued alcohol abuse postoperatively. The other 15 patients with a low postoperative F I Group B (2.04 f 0.26), four of them with calcifications Alcoholics (n = 18). Three patients died post(26.7%) and all with continuous alcohol abuse, operatively. Of the remaining 15 patients, 7 obsuffered recurrent pain attacks ( 3 had had a sphinc- tained lasting pain relief from the operation despite a terotomy). Twelve of these patients did, however, rather low postoperative FI (2.57 ? 0.48) (Fig. 3). obtain late pain relief within an average of 3.9 years Six of the seven patients discontinued alcohol intake postoperatively in association with marked pan- postoperatively. Pancreatic calcifications at the opcreatic dysfunction (FI 4.79 ? 0.22) (Fig. 2). Pan- eration were found in six cases (85.7%).During the creatic calcifications at the operation (n = 4) or follow-up period six of seven patients showed a postoperatively (n = 7 ) were demonstrated in 1 1 of progressive increase of pancreatic dysfunction the 15 patients or in 9 of the 12 patients (75%) who (Fig. 3). obtained late pain relief. The three patients who The remaining eight patients Suffered postoperacontinued to have pain relapses had a shorter follow- tively from recurrent pain episodes; all continued up and less marked pancreatic dysfunction (Fig. 2). alcohol abuse, their FI being low (2.31 ? 0.47). Non-alcoholics (n = 6). One patient died within Only two had pancreatic calcifications (25%). 1.5 years after sphincterotomy. The remaining five Seven of the eight patients did, however, obtain late patients obtained pain relief from the operation 'pain relief within an average of 2.7 years postoperadespite a rather low FI (2.33 ? 0.88) (Fig. 1). The tively in association with severe pancreatic dysfuncFI increased in four of five patients during the tion (FI 5.0 0.48) (Fig. 4). Pancreatic calcifica-
R . W. Ammann, F. Largiader & A . Akovbiantz
212
FI 61
4i I 4
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
@ -
-
2
PAIN CONTINUED PAIN RELAPSES LASTING PAIN RELIEFCALCIFICATIONS OF PANCREAS @@
-*
2
4
8
6
10 12 YEARS (postoperatively)
Fig. 2. Group A (cyst drainage, surgical failure). Postoperative pain relapses in relation to pancreatic dysfunction. Fifteen patients with alcoholic C P and persisting alcohol abuse who had a low postoperative FI (g4.0)relapsed after surgery ( L O , L*). Twelve of them obtained late pain relief within 1-7 years postoperatively in association with marked pancreatic dysfunction (FI 4.0) (* 2 ) Nine . of the 12 had calcifying CP (0, 0 ) .The time relationship between operation, degree of dysfunction and pain indicates that pain relief occurred probably secondary to the increase of pancreatic dysfunction and despite alcohol abuse (see also Fig. 4).
>
FI
I-
1
I
I
2
I
I
4
I
I
6
I
I
8
I
//-----
10
-13 YEARS (postoperatively)
Fig. 3. Group B (ductal drainage, surgical success). Lasting pain relief (* 2 ) in relation to alcohol abstinence. Six of the 7 patients with lasting pain relief abstained from alcohol postoperatively. Pain relief was obtained by 5 patients despite a low FI (Q3.0), indicating that alcohol withdrawal was probably the major factor for pain relief (see also Fig. 1). Six of the 7 patients also had pancreatic calcifications at the operation (0).
Pain Relief in Chronic Pancreatitis
213
FI mA/R
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
I/R
-
PAIN CONTINUED PAIN RELAPSES LASTING PAIN RELIEF*CALCIFICATIONS OF PANCREAS @8
-*
A ALCOHOLIC CP I IDIOPATHIC CP R RESECTION (40-60%1 0 DUCTAL DRAINAGE *l/D
I
I
2
I
I
I
4
I
I
6
I
8
I
I
I
I
10 12 YEARS (postoperatively)
Fig. 4. Group B (ductal drainage and/or distal resection, surgical failure). Postoperative pain relapses in relation to pancreatic dysfunction. Thirteen patients, 8 of them with alcoholic C P and continuing alcohol abuse (with a low postoperative FI ($3.0) in all but 4 cases) relapsed after surgery ( L O . 0*). Late pain relief was obtained by 10 of 13 patients, 7 of them alcoholics, within 1-8 years postoperatively in association with an increase of pancreatic dysfunction (* 2). The time relationship between operation, degree of pancreatic dysfunction, and pain indicates that pain relief occurred probably secondary to the increase in pancreatic dysfunction (see also Fig. 2). Pancreatic resection probably enhanced pancreatic dysfunction in some cases. Pancreatic calcifications were present in 5 of the 7 alcoholics at the operation ( 0 )or later in the course ( 0 ,0).
tions were demonstrated in five of seven patients (71.4%) who obtained late pain relief. Non-alcoholics (n = 7). Two patients died postoperatively. None of the remaining five patients obtained pain relief from the operation, the FI being low (2.8 k 0.8) (Fig. 4). Three of the five patients, however, obtained late pain relief in association with severe pancreatic dysfunction (5.3 k 0.67) within an average of 3.3 years. Pancreatic calcifications at the operation (n = 3) or postoperatively (n = 2) were demonstrated in all five patients.
DISCUSSION The principal aim of surgery in C P is to relieve pain while preserving as much functioning pancreatic tissue as possible (1 2, 28). Since the factors responsible for pain are poorly defined, the surgical method of choice is still debated ( 1 1, 12, 18, 19, 28). Pancreatico-jejunostomy does not markedly increase
pancreatic dysfunction but achieves pain relief only in about tyo-thirds of cases (14). The better results with reg&d to pain relief by resecting procedures are negatively influenced by the marked increase of pancreatic dysfunction (and by the late mortality) (12, 19). Thus the question arises whether the increase of pancreatic dysfunction induced by surgery may not promote pain relief. This hypothesis is supported by our findings. The time relationship between surgery, degree of pancreatic dysfunction (k alcohol abstinence), and pain will be discussed separately in order to evaluate the relative effect of these factors on pain relief after surgery. No attempts have been made so far to correlate pancreatic dysfunction with the postoperative pain profile. There are several reasons for the lack of adequate data on postoperative exocrine function. The CCK-secretin test often cannot be performed properly because of the deranged anatomy after surgery. Steatorrhea, on the other hand, is a rather insensitive and non-specific criterion, since it occurs
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
214
R. W. Ammann, F. LargiadPr & A . Akovbiantz
only in advanced CP with a decrease of lipase secretion by more than 90% (7, 17) and since other factors, particularly gastric resections, increase the fecal fat excretion ( 15, 29). The diagnostic value of the fecal chymotrypsin method, which is particularly useful for long-term studies of exocrine function in CP (3, 5), has been thoroughly evaluated recently (3, 5,9,26).A comparable high percentage of pathological results with the fecal chymotrypsin method and with the CCK-secretin test was found in CP with advancing insufficiency (3). Since the endocrine function may be affected disproportionally more than the exocrine function after distal pancreatectomy (1 2,25), the FI, which expresses the overall pancreatic dysfunction in one figure, appears to be a valuable criterion.
ingly, the good results with regard to pain by pancreatic duct ligation in CP support our thesis of a direct relationship between pancreatic dysfunction and pain relief (6, 16, 21). The current concept that the good results observed after resection of the mostly inflamed segment of the pancreas are due to elimination of peripancreatic and perineural inflammation has to be questioned (8, 1 1 , 12, 19). All available data suggest rather that marked pancreatic dysfunction is a major factor in lasting pain relief in uncomplicated CP. Pancreatic calciJicationand pain relief after surgery Alcoholic CP with calcifications seems to do better after surgery in our experience than noncalcifying CP, which is in accordance with results first reported by Duval & Enquist (10) and confirmed by others (1 8, 19). That is not surprising because there is a close relationship between progressive pancreatic dysfunction and pancreatic calcifications in alcoholic CP (22) (Figs. 1-4). In nonalcoholic CP the relationship between pancreatic calcifications and operative results is less evident, since in group A only two of five patients with postoperative pain relief had calcifications and in group B all five patients relapsed after surgery despite pancreatic calcifications.
Pancreatic dysfunction and pain relief after surgery If it is assumed that pain in CP is mainly due to stasis in the cysts (group A) or in the ductal system (group B), lasting pain relief should be obtained by appropriate drainage procedures that do not markedly affect pancreatic function. In alcoholic CP over 60%of patients of groups A + B (23 of 3 7) had pain relapses after surgery, but 19 of them obtained late pain relief 1-8 years postoperatively in association with marked increase of pancreatic dysfunction (Figs. 2 and 4). In group B all five non-alcoholic patients relapsed postoperatively, but three of them Pseudocysts and results of drainage procedures Successful cyst drainage procedures in alcoholic obtained late pain relief 1-8 years after surgery parallel with the increase of the FI (Fig. 4). Lasting CP usually do not prevent recurrent episodes of pain relief was also achieved with the operation in pancreatitis (8, 12,25) except in patients either with group A in four alcoholic CP patients with marked advanced pancreatic dysfunction (see above) or pancreatic dysfunction (Fig. 1). These data strongly with alcohol withdrawal (see below). In contrast, all suggest that credit for pain relief cannot be given to five patients with non-alcoholic CP of group A surgery in these groups of patients. The results, on obtained lasting pain relief from the operation the contrary, seem to confirm that spontaneous pain (Fig. I). In this group pain relief must be attributed relief parallel with progressive pancreatic dysfunc- directly to the cyst drainage procedure, since.at a tion is part of the natural history of CP (2, 4, 5). A similar low postoperative FI, patients with alcoholic steady increase of pancreatic dysfunction over the CP continued to relapse. years is also seen in most operated cases of alcoholic and idiopathic CP (Figs. 1-4). Alcohol abstinence and surgical results Spontaneous disappearance of pain in CP has As shown above, pain relief through surgery is been reported previously but without reference to likely to occur in alcoholic CP with advanced panpancreatic dysfunction (1, 13, 20, 23, 24). It is creatic dysfunction (high FI). Do all patients with obvious that in complicated CP (e.g. with pseudo- alcoholic CP and a low FI develop postoperative cysts), pain may persist or recur over years despite pain relapses? Obviously not. In our experience marked pancreatic dysfunction (12, 22). Interest- alcohol abstinence postoperatively was probably
Scand J Gastroenterol Downloaded from informahealthcare.com by University of Newcastle on 12/27/14 For personal use only.
Pain Relief in Chronic Pancreatitis
important for pain relief in three cases in group A with low FI (Fig. 1 ) and in seven alcoholic C P patients in group B without postoperative pain relapses despite a low FI after surgery (Fig. 3). Six of these seven patients discontinued alcohol abuse, while eight alcoholics in group B with a similar low FI who continued alcohol abuse suffered pain relapses (Fig. 4). Since pain tends to subside spontaneously with alcohol abstinence in uncomplicated C P in 30-50% or more of the patients (18,22,2527), many experts reject surgery in uncomplicated alcoholic C P before the effect of complete abstinence on pain has been evaluated (12, 19, 22, 25). Results of ductal decompressionand distal resection Ductal stasis seems to be an important factor of pain in C P (8, 18, 19, 28). Thirteen patients in group B had a ductal drainage procedure, five of them combined with distal resection (Figs. 3 and 4). Eight of the 13 patients had pain relapses after surgery (Fig. 4), but 6 of them obtained late pain relief 1 year or more postoperatively in association with marked increase of pancreatic dysfunction (Fig. 4). Three of these six patients also had a distal pancreatectomy. Of seven patients with distal resection alone, two became pain-free by surgery (Fig. 3), and four obtained late pain relief 2-8 years postoperatively in association with an increase of pancreatic dysfunction (Fig. 4). In our experience the operative result of ductojejunostomy and/or distal resection seems to be infiuenced primarily by the severity of pancreatic dysfunction, which is enhanced by partial pancreatectomy. In conclusion, our data suggest that the outcome with regard to pain relief by surgery in uncomplicated C P depends predominantly on pancreatic dysfunction, presence of pancreatic calcifications, and alcohol abstinence. In complicated C P (e.g. pseudocysts), the indication for surgery is not debated, but for lasting pain relief these factors are also important, at least in alcoholic CP. ACKNOWLEDGEMENTS Presented in part at the meeting of the European Pancreatic Club, 8 October 1977, in Dublin. SupReceived 16 May 1978 Accepted 1 November 1978
215
ported in part by a grant from the Hartmann-Muller Foundation for Medical Research. REFER ENC ES 1. Adharn, N. F., Dyce, B. & Haverback, B. J. Amer. J . Dig. Dis. 1968, I S , 8-15 2. Arnrnann, R. Schweiz. Med. Wschr. 1970, 95, 1-7 3. Amrnann, R. Leber Magen Darm 1976,6,217-226 4. Ammann, R. Z . Gastroent. 1978, Suppl. 13, 42-48 5. Ammann, R., Hammer, B. & Fumagalli, I. Digestion 1973, 9, 404-415 6. Cannon, J. A. Amer. J . Surg. 1955, 90, 266-279 7. Di Magno, E. P., Go, V. L. W. & Surnmerskill, W. H. J. New. Engl. J . Med. 1973, 288, 813-815 8. Dixon, J. A. & Englert, E. Gastroenterology 1971, 61, 375-381 9. Diirr, H. K., Schneider, R. & Bode, J. Ch. pp. 17-29 in Bartelheimer, H., Classen, M. & Ossenberg, F. W. (eds.) Die Untersuchung der Bauchspeicheldruse. Thieme, Stuttgart, 1976 10. Duval, M. K. & Enquist. I. F. Surgery 1961, 50, 965-969 11. Editorial. Lancet 1977. I , 460-462 12. Frey, Ch. F., Child, Ch. G. & Fry, W. Ann. Surg. 1976. 184, 403-412 13. Gambill, E. E., Baggenstoss, A. H. & Priestley, J. T. Gastroenterology 1960. 39, 404-41 I 14. Govaerts, J. P., Kestens, P. J. & Kiekens, R. et al. Acta Gastroent. Belg. 1976, 39. 595-602 15. Hillrnann, H. Gut 1968, 9, 576-584 16. Hoffrnann, E., Usmiani, J. & Gebhardt, Ch. Dtsch. Med. Wschr. 1977, 102, 392-395 17. Hotz, J., Goberna,R. & Clodi,Ph. Digestion 1973,9, 212-223 18. Jordan, G. L., Strug, B. S . & Crowder, W. E. Amer. J. Surg. 1977, 133, 46-50 19. Leger, L., Lenriot, J. P. & Lemaigre, G. Ann. Surg. 1974, 180, 185-191 20. Levrat, L., Moulinier, B. et al. Arch. Mal. Appar. Dig. 1970, 59, 5-18 2 1. Maddig, G. F. & Kennedy, P. A. Amer. J . Surg. 1973, 125, 538-541 22. Marks, I. N., Bank. S. & Barbezat, G. 0. Leber Magen Darm 1976, 6, 257-270 23. Mc Elroy, R. & Christiansen, Ph. A. Amer. J. Med. 1972, 52, 228-24 1 24. Sarles, H. & Gerolmi-Santandrea, A. Clin. Gastroent. 1972, I , 167-193 25. Sarles, J. C. & Sarles, H. LeberMagen Darm 1976,6, 294-299 26. Schneider, R., Durr, H. K. & Bode, J. Ch. Dtsch. Med. Wschr. 1974. 99, 1449-1454 27. Strum, W. B. & Spiro, H. M. Ann. Intern. Med. 197 1, 74, 264-277 28. White, Th. T. & Keith, R. G. Surg. Gynec. Obstet. 1973, 136, 353-358 29. Wollaeger, E. E., Waugh, J. M. & Power, M. H. Gastroenterology 1963, 44, 25-32
