- Email: [email protected]
Pancreatic resection in the elderly1
ORIGINAL SCIENTIFIC ARTICLES
Pancreatic Resection in the Elderly Amy M Lightner, MD, Robert E Glasgow, MD, Thomas H Jordan, BS, Alexander D Krassner, BA, Lawrence W Way, MD, Sean J Mulvihill, MD, Kimberly S Kirkwood, MD Elderly patients undergoing pancreatic resection present unique challenges in postoperative care. Although mortality rates among elderly patients after pancreatectomy at high-volume centers is known to be low, the anticipated decline in functional status and nutritional parameters has received little attention. Functional decline is an unrecognized but critically important consequence of pancreatic resection in older patients. STUDY DESIGN: This study is a retrospective review, validation cohort, of older and younger patients undergoing major pancreatic resection. The setting is the state of California (database of all hospitals in the state) and The University of California, San Francisco (UCSF; a tertiary care referral center). The study population is a consecutive sample of older (greater than or equal to 75 years) and younger (16 to 74 years) patients from California (January 1990 to December 1996; n ⫽ 3,113) and UCSF (January 1993 to November 2000; n ⫽ 218), who underwent radical pancreaticoduodenectomy, distal pancreatectomy, or total pancreatectomy for neoplasia. The main outcomes measures were length of stay, complications, mortality, discharge disposition, supplemental nutrition requirement, and readmissions. RESULTS: Elderly patients had higher mortality rates than the young statewide (10% versus 7%, p ⫽ 0.006). Although the 3% mortality at UCSF was the same for both groups, older patients were more often admitted to the ICU (47% versus 20%, p ⫽ 0.003), treated for major cardiac events (13% versus 0.5%, p ⬍ 0.001), discharged with enteral tube feedings (48% versus 16%, p ⬍ 0.001), or malnourished on readmission (17% versus 2%, p ⬍ 0.005). Older patients were more frequently discharged to skilled nursing facilities (17% versus 1% at UCSF; 24% versus 7% in California; p ⬍ 0.001, both groups). CONCLUSIONS: Older patients are more likely than younger patients to require an ICU stay, suffer a cardiac complication, and experience compromised nutritional and functional status after major pancreatic resection. ( J Am Coll Surg 2004;198:697–706. © 2004 by the American College of Surgeons) BACKGROUND:
impact of these procedures on functional status in the elderly. The recent expansion of research in geriatric medicine has brought increasing emphasis on quality of life in the elderly and on the importance of evaluating the impact of medical interventions on functional status. So it is important to determine the impact of pancreatic resection on physical independence at the time of discharge. The incidence of malnutrition among older hospitalized nonsurgical patients has been well-documented,7 but the scope of this condition among patients undergoing abdominal operations is not known. Patients with pancreatic cancer may be at especially high risk for perioperative malnutrition because cachexia frequently attends this diagnosis. Severe malnutrition adversely affects mortality rate, functional status, and the need for skilled nursing home admission after acute hospitaliza-
The aging of the western population has generated an interest in the suitability of older patients for major surgical procedures. Several studies have documented “acceptable” increases in morbidity and mortality rates for older patients undergoing major pancreatic resections.1-6 These standard end points provide little insight into the No competing interests declared.
Supported by NIH T32 DK-07573-12 (Lightner, Kirkwood) and the Claude E Welch Fellowship from the Massachusetts General Hospital (Lightner). Poster presented at the 73rd Annual Meeting of the Pacific Coast Surgical Association, Las Vegas, NV, February 2002. Received August 22, 2003; Revised December 18, 2003; Accepted December 18, 2003. From the Department of Surgery, University of California, San Francisco, CA. Correspondence address: Kimberly S Kirkwood, MD, Department of Surgery, University of California, San Francisco, 533 Parnassus Ave, Room U-372, San Francisco, CA 94143-0790. email: [email protected]
© 2004 by the American College of Surgeons Published by Elsevier Inc.
697
ISSN 1072-7515/04/$30.00 doi:10.1016/j.jamcollsurg.2003.12.023
698
Lightner et al
Pancreatic Resection in the Elderly
Abbreviations and Acronyms
ASA ⫽ American Society of Anesthesiologists OSHPD ⫽ Office of Statewide Health Planning and Development PD ⫽ pancreaticoduodenectomy PPPD ⫽ pylorus-preserving pancreaticoduodenectomy UCSF ⫽ The University of California, San Francisco
tion among older nonsurgical patients.7 So nutritional status is a potentially important determinant of postoperative outcomes. We evaluated outcomes of major pancreas resections for neoplasia using two sources of data: a large population database that includes patient discharge abstracts from every admission to every acute care hospital statewide in California (CA), and at The University of California, San Francisco Moffit-Long Hospital, a highvolume tertiary care referral center (UCSF). Within these groups, we reviewed hospital records, discharge, and readmission data for elderly and young cohorts. We hypothesized that elderly patients experience a more complicated hospital course (increased complications and higher level of care required) than do younger patients. We also predicted that older patients are more susceptible to impaired functional and nutritional status (discharge disposition, need for supplemental nutrition, readmission for malnutrition or dehydration) after pancreatic resection. METHODS The methods for assessing pancreas surgery outcomes at the UCSF and in the state of California were slightly different and are described in this section. Analysis of statewide data provided population-based information and included a broad range of regional practice patterns and hospital volumes. In contrast, a more detailed analysis of the hospital course was afforded by reviewing our single institution experience. The University of California, San Francisco
All patients who underwent major pancreatic resections at the UCSF for a presumed neoplastic process of the pancreas or periampullary region during the period January 1993 through November 2000 were included in the study. Two hundred twenty-four such patients were
J Am Coll Surg
identified through Department of Surgery and medical center coding unit records. Six patients were excluded because of unavailability of complete medical records, bringing the total study population to 218. Thirty patients were age 75 years or older at the time of operation and were considered “elderly.” One hundred eightyeight patients less than 75 years old were grouped into the “younger” cohort. Patient medical records (paper and electronic charts) were reviewed retrospectively. The study was approved by the University of California, San Francisco Committee on Human Research. For purposes of inclusion, a major pancreatic resection was a radical pancreaticoduodenectomy (Whipple procedure), distal pancreatectomy (laparoscopic or open), or total pancreatectomy. A presumed neoplastic process was a mass (other than a presumed pseudocyst) or suspicious intraductal lesion imaged by radiography, esophagogastroduodenoscopy (for ampullary tumors), or endoscopic retrograde cholangiopancreatography. The mass or lesion might have been assumed to be benign or malignant, and may or may not have been biopsied before the operation. Only nonemergent operations were included. Postoperative complications were defined broadly to thoroughly assess patients’ postoperative courses. Morbidities were included in the analysis if they were significant enough to require medical or surgical intervention or consultation. Minor complications were those that required only minimal intervention and were not life-threatening. Significant complications were those that necessitated a major procedure (eg, operation, radiologically guided abscess drainage, endotracheal intubation), close monitoring (step-down or ICU), were potentially life threatening, or delayed discharge. Postoperative mortality was defined as death that occurred during the same hospitalization as the index pancreatic procedure. Longterm survival could not be accurately assessed retrospectively and was not included in our study. Because the UCSF is a tertiary care center in a large urban area, many patients obtain longterm followup care at regional facilities. Readmissions to UCSF within 30 days of discharge from the index pancreatic procedure were included in the study. Hospitalizations were excluded if the diagnosis at readmission was completely unrelated to the operation or underlying cancer diagnosis.
Vol. 198, No. 5, May 2004
Lightner et al
Analysis
mental nutrition requirement at discharge, and readmissions) was not available in the OSHPD database.
Data were analyzed using Microsoft Access 98 and Microsoft Excel 98. Statistical analyses were conducted using the Primer statistics package (Version 3.0, McGrawHill, 1992) and InStat 3. Student’s t-test, chi-square, and Mann-Whitney rank-sum tests were used as appropriate, with statistical significance set at p ⬍ 0.05.
Pancreatic Resection in the Elderly
699
Analysis
Chi-square and t-tests were used for comparisons, and statistical significance was defined as p ⬍ 0.05. The JMP 3.2.6 statistical software program (SAS Institute Inc, Cary, NC) was used for analysis.
State of California
We also conducted a retrospective review of standardized patient discharge abstracts using a database obtained from the California Office of Statewide Health Planning and Development (OSHPD). This database contains discharge data abstracts for every patient hospitalization from each acute care hospital in California. All discharge abstracts from the years 1990 to 1996 were included in the initial database search, and from these abstracts, 3,113 patients (from 318 acute care facilities) who underwent major pancreatic resection for benign or malignant disease were selected. The OSHPD database uses diagnostic and procedural codes derived from the International Classification of Diseases, Ninth Revision, Clinical Modification, fourth edition (ICD-9), issued by the US Department of Health and Human Services.8 Codes used in the database search for operations were: total pancreatectomy (ICD-9 52.6), radical pancreaticoduodenectomy (ICD-9 52.7), proximal pancreatectomy (ICD-9 52.51), radical subtotal pancreatectomy (ICD-9 52.53), pancreatectomy unspecified (ICD-9 52.59), and distal pancreatectomy (ICD-9 52.52). To further select for malignant disease, patients with a diagnosis of cancer of the exocrine pancreas or islet cells (ICD-9 157.X), duodenum (ICD-9 152.0), bile duct (ICD-9 156.1), and the ampulla of Vater (ICD-9 156.2) were included. Patients with benign disease leading to major pancreatic resection were selected using diagnostic codes for the pancreas (ICD-9 211.67, 235.5, 239.0), duodenum (ICD-9 211.2, 235.2), ampulla of Vater (ICD-9 239.0), and bile duct (ICD-9 211.5, 235.3, 239.0). Definitions of the terms elderly, young, and mortality were the same as for the UCSF cohort. Other information available from the OSHPD database and used in this study included total length of stay and disposition after discharge. Some information obtained from UCSF charts (American Society of Anesthesiologists classifications, subsets of lengths of stay, complications, supple-
RESULTS Demographics
Among the 3,113 patients in the study who underwent major pancreatic resections for neoplasia in the state of California, 515 (17%) were 75 years or older (data not shown). Similarly, older patients comprised 14% of the 218 patients who underwent pancreatic resections at UCSF between 1993 and 2000. Approximately equal numbers of men and women underwent pancreatic resection; the older patients had slightly more women (56%) than men in the statewide cohort of 515 subjects. The ASA Physical Status Classification9,10 is commonly used to stratify comorbidity. Although these data were not available for CA patients, the elderly UCSF cohort had a higher median ASA score than did the young cohort (3 versus 2, p ⬍ 0.005). Procedures
Radical pancreaticoduodenectomy was the most common resection performed in both older and younger cohorts, representing 80% of both populations. Pyloruspreserving pancreaticoduodenectomy (PPPD) was performed more often than classic pancreaticoduodenectomy (PD) at UCSF (71% versus 29% of radical pancreaticoduodenectomies), with similar percentages in the elderly and young cohorts. Median ASA classifications between the PPPD and PD groups were not significantly different (data not shown). Diagnoses
Table 1 shows the pathologic diagnoses among both populations of patients. Malignant tumors made up the majority of diagnoses, in both the statewide and UCSF cohorts (90% and 83%, respectively). The malignancies were most often of the exocrine pancreas. Among the UCSF patients, benign and malignant neuroendocrine tumors of the pancreas were seen exclusively in the young cohort. The mean age of patients with malignant disease at UCSF was greater than that of patients with
700
Lightner et al
Pancreatic Resection in the Elderly
J Am Coll Surg
Table 1. Pathologic Diagnoses CA Primary tumor site/diagnosis
Malignant disease Pancreas Exocrine Adenocarcinoma Cystic Intraductal papillary Anaplastic Neuroendocrine Ampulla Adenocarcinoma Neuroendocrine Duodenum Adenocarcinoma Gastrointestinal stromal tumor Neuroendocrine Common bile duct Other Benign disease Pancreas Exocrine Cystic Solid Neuroendocrine Duodenum Common bile duct
All patients
2,816 (90) 1,870 (60) 1,807 (58) N/A‡ N/A N/A N/A 63 (2) 507 (16) N/A N/A 187 (6) N/A N/A N/A 216 (7) 36 (1) 297 (10) 291 (9) 214 N/A N/A 77 6 (0.2) 0 (0)
UCSF All patients
181 (83) 95 (44) 79 (36) 65 7 4 3 16 (7) 39 (18) 38 1 14 (6) 9 3 2 13 (6) 20 (9) 37 (17) 33 (15) 28 22 6 5 3 (1) 1 (0.5)
Elderly >75 y
26 (87) 10 (33) 10 (33) 7 1 0 2 0 8 (27) 8 0 3 (10) 0 2 1 2 (7) 3 (10) 4 (13) 3 (10) 3 1 2 0 1 (3) 0 (0)
Young <75 y
155 (82) 85 (45) 69 (37) 58 6 4 1 16 (9) 31 (16) 30 1 11 (6) 9 1 1 11 (6) 17 (9) 33 (18) 30 (16) 25 21 4 5 2 (1) 1 (0.5)
p Value*
⬍0.005† ⬍0.005† ⬍0.005† — — — — — — — — — — — — — ⬍0.005† ⬍0.005† ⬍0.05† — — — — — —
Data are presented as number (percentage) of patients for the state of California (CA) and The University of California, San Francisco (UCSF). *p value shown only if ⬍0.05. † CA versus all UCSF patients (chi-square test). ‡ Data not available.
benign disease (61.9 ⫾ 1.0 years versus 54.8 ⫾ 2.6 years, p ⫽ 0.005, t-test). Postoperative course Complications
Fifty-eight percent of UCSF-patients (70% of the elderly and 56% of the young) had one or more complications postoperatively (Table 2). The mean age of patients who had complications was greater than those who did not (63 ⫾ 1.1 years versus 57 ⫾ 1.6 years, p ⬍ 0.005, t-test). Patients who underwent radical pancreaticoduodenectomy were more likely to have a postoperative complication than those who underwent distal pancreatectomy (p ⬍ 0.005, chi-square). Major cardiac complications were more frequent among older patients (Table 2). Four elderly patients (13% of the cohort) and one young patient (0.5% of the
cohort; p ⬍ 0.005 by chi-square) had an acute myocardial infarction, cardiac arrest, or congestive heart failure. A peripancreatic or anastomotic leak was the second most common complication overall (10% of patients) and occurred in 13% of elderly versus 9% of young patients. Older patients had a greater likelihood of intraabdominal abscesses, pneumonia, altered mental status, urinary tract infection, Clostridium difficile colitis, sepsis, and acute renal failure, but these differences were not significant. Delayed gastric emptying after pancreaticoduodenectomy prolongs the hospital course and affects nutrition.11,12 Our retrospective analysis did not permit discrimination among delayed gastric emptying, ileus, and partial small bowel obstruction. So we used the term delayed gastrointestinal transit, to include these three conditions for which notes in the medical record indicated
Lightner et al
Vol. 198, No. 5, May 2004
Pancreatic Resection in the Elderly
701
Table 2. Postoperative Complications of The University of California, San Francisco Patients All patients Complications
All complications* Delayed gastrointestinal transit Peripancreatic or anastomotic leak Wound infection and/or cellulitis Intraabdominal abscess or fluid collection Pneumonia Abdominal reoperation Altered mental status Gastrointestinal bleed Urinary tract infection C difficile colitis Minor cardiac event† Sepsis Acute renal failure Adult respiratory distress syndrome Major cardiac event‡
Elderly (>75 y)
Young (<75 y)
n
%
n
%
n
%
127 48 21 18 14 13 12 9 9 9 8 7 7 6 6 5
58 22 10 8 6 6 6 4 4 4 4 3 3 3 3 2
21 7 4 2 3 4 0 3 0 2 2 1 2 2 1 4
70 23 13 7 10 13 0 10 0 7 7 3 7 7 3 13§
106 41 17 16 11 9 12 6 9 7 6 6 5 4 5 1
56 22 9 9 6 5 6 3 5 4 3 3 3 2 3 0.5
Data are presented as number (percentage) of all (218 total), elderly (30 total), and young (188 total) The University of California, San Francisco patients. *Complications (total number) not listed in table were: adverse drug reaction (1), anorexia requiring total parenteral nutrition (1), Candida esophagitis (2), chyle leak (3), depression (1), diarrhea (non-C difficile), disseminated intravascular coagulopathy (1), hearing loss (acute)(1), hepatic lobe infarction (1), hyperglycemia (1), hyponatremia (1), influenza A infection (1), intraabdominal hemorrhage (2), ischemic bowel (2), necrotizing wound infection (2), nerve palsy (1), operation (nonabdominal)(1), pleural effusion requiring drainage (3), pulmonary edema (1), pulmonary embolus (1), sinusitis (2), small bowel obstruction (complete)(1), superficial thrombophlebitis (1), technical misadventure (1), transfusion reaction (1), urinary retention (1), wound dehiscence (1). † Benign arrhythmia, requiring pharmacologic intervention. ‡ Acute myocardial infarction, cardiac arrest, or congestive heart failure. § p ⬍ 0.005 versus young cohort (chi-square test).
resultant impairment of enteral nutrition. Delayed gastrointestinal transit affected 23% and 22% of elderly and young patients, respectively, at UCSF (Table 2). Delayed gastrointestinal transit occurred in 29% of patients who underwent PPPD compared with 28% for PD. Intensive care unit (ICU) admission and length of stay
Older patients were more likely to require admission to the ICU (47%) than younger patients (20%, p ⬍ 0.005, chi-square Table 3). Although the elderly tended to have a greater mean length of stay in the ICU than the young (2.3 ⫾ 0.7 days versus 1.3 ⫾ 0.3 days), these differences were not significant. Postoperative and total lengths of stay
Average postoperative and total lengths of stay at the UCSF were 13.3 ⫾ 0.7 days and 15.1 ⫾ 0.7 days, respectively (Table 3). Comparison of lengths of stay between elderly and young cohorts did not yield appreciable differences. Review of the statewide data for total lengths of stay
showed that patients in CA stayed in the hospital for an average of 20.1 ⫾ 0.3 days after major pancreatic resection. Elderly patients in the CA group had longer lengths of stay than their younger counterparts (21.4 ⫾ 0.7 days versus 19.8 ⫾ 0.3 days, p ⬍ 0.05, t-test). The longer hospital stays among patients in the statewide database as compared with those at UCSF might, in part, be explained by the fact that the UCSF review included some patients treated more recently, during a time period when hospital stays have been reduced. Mortality
Three percent of patients at the UCSF and 7% in CA died in the hospital after major pancreatic resection for neoplasia (Table 3, p ⬍ 0.05 by chi-square). The mortality rate at UCSF was the same for elderly and young patients (3% of each group), but statewide, the elderly had a higher mortality rate (10% versus 7%, p ⬍ 0.05 by chi-square). Six of seven deaths at UCSF occurred after pancreaticoduodenectomies (one was after total pancreatectomy).
702
Lightner et al
Pancreatic Resection in the Elderly
J Am Coll Surg
Table 3. Postoperative Course CA Variable
Intensive care unit (ICU) stay Yes No ICU length of stay (d) Mean ⫾ SEM Range Postoperative length of stay (d) Mean ⫾ SEM Range Total length of stay (d) Mean ⫾ SEM Range Mortality§
UCSF
All patients
Elderly (>75 y)
Young (<75 y)
All patients
Elderly (>75 y)
Young (<75 y)
N/A* N/A
N/A N/A
N/A N/A
52 (24) 166 (76)
14 (47)† 16 (53)
38 (20) 150 (80)
N/A N/A
N/A N/A
N/A N/A
1.4 ⫾ 0.3 0–34
2.3 ⫾ 0.7 0–17
1.3 ⫾ 0.3 0–34
N/A N/A
N/A N/A
N/A N/A
13.3 ⫾ 0.7 2–82
13.8 ⫾ 1.5 7–39
13.2 ⫾ 0.7 2–82
20.1 ⫾ 0.3 N/A 233 (7)
21.4 ⫾ 0.7† N/A 54 (10)‡
19.8 ⫾ 0.3 N/A 179 (7)
15.1 ⫾ 0.7 3–84 7 (3)
16 ⫾ 1.5 8–41 1 (3)
15.0 ⫾ 0.7 3–84 6 (3)
Data are presented as number (percentage) of patients for the state of California (CA) and The University of California, San Francisco (UCSF). *Data not available. † p ⬍ 0.005, elderly versus young cohort, UCSF (chi-square test). ‡ p ⬍ 0.05, elderly versus young cohort, CA (t-test). § During same hospital admission as index procedure.
Five of the seven patients who died were reexplored during their hospital courses, and five eventually died of sepsis. Four patients experienced adult respiratory distress syndrome, two had intestinal infection, and two had acute renal failure. The one elderly patient who died succumbed from respiratory failure. Deaths occurred from postoperative days 2 through 82 (mean day 31). Posthospital course Discharge disposition
After pancreatectomy, surviving older patients were less likely to go home than younger patients in both the
UCSF and the CA groups (Table 4). Only 7% of younger patients in the statewide cohort recovered at a skilled nursing facility, but 26% of older patients were transferred to a skilled nursing facility after their acute hospitalization. Figure 1 shows details of the posthospital course for UCSF patients. Only 41% of surviving elderly patients were sent home independently (without services) compared with 75% of young patients (p ⬍ 0.005 versus young cohort). Note that none of the UCSF patients sent to a skilled nursing facility postoperatively had been residing in a chronic care facility preoperatively.
Table 4. Posthospital Course CA Variables
Discharged from hospital,* n (%) Disposition, n (%) Home Hospital transfer Skilled nursing facility
UCSF
All patients
Elderly (>75 y)
Young (<75 y)
All patients
Elderly (>75 y)
Young (<75 y)
2,880 (93)†
461 (90)‡
2,419 (93)
211 (97)
29 (97)
182 (97)
2,589 (90) N/A¶ 291 (10)
339 (74)§ N/A 122 (26)§
2250 (93) N/A 169 (7)
203 (96) 1 (0.5) 7 (3.5)
24 (83)㛳 0 (0) 5 (17)㛳
179 (98.5) 1 (0.5) 2 (1)
*Data are presented as number (%) of patients for the state of California (CA, 3,113 total patients) and The University of California, San Francisco (UCSF, 218 total patients). † p ⬍ 0.05 versus respective cohort in UCSF group (chi-square test). ‡ p ⬍ 0.05 versus young cohort, CA (chi-square test). § p ⬍ 0.005 versus young cohort, CA (chi-square test). 㛳 p ⬍ 0.005 versus young cohort, UCSF (chi-square test). ¶ Data not available.
Vol. 198, No. 5, May 2004
Lightner et al
Pancreatic Resection in the Elderly
703
Figure 1. Disposition of surviving University of California San Francisco (UCSF) patients at the time of discharge after pancreatic resection. Displayed are percentages of 29 elderly and 182 young patients discharged home (with and without assistance), skilled nursing facilities, or with enteral tube feedings. *p ⬍ 0.005 versus young (chi-square).
Supplemental nutrition
Overall, 22% of surviving patients at UCSF required supplemental nutrition at the time of discharge. Among the 48% of surviving elderly patients who required supplemental nutrition after discharge, tube feeds were delivered through surgically placed feeding jejunostomy tubes in every case. There were no complications attributed to jejunostomy tube placement. Younger patients were less likely than older patients to require supplemental nutrition (16%, p ⬍ 0.005 by chi-square; Fig. 1). Patients who were discharged with tube feeds had longer mean ICU, postoperative, and total lengths of stay (p ⬍ 0.05 for all, by t-test, data not shown) than patients discharged without supplemental nutrition. Patients who had a pancreaticoduodenectomy procedure were more likely than patients who underwent distal or total pancreatic resection to require supplemental nutrition at discharge (p ⬍ 0.005 by chi-square). The use of tube feeds was similar among PD and PPPD patients. Readmissions
Thirty-nine (18%) of the 211 surviving UCSF patients were readmitted within 30 days of discharge for a problem directly related to the index procedure or disease
(Table 5). The overall percentages of elderly (24%) and young (18%) who were readmitted were not significantly different. Table 5 lists the diagnoses among those readmitted to UCSF. Malnutrition or dehydration were the most common reasons for readmission in older patients, and the elderly were eight times more likely to require admission for these conditions that the younger cohort (p ⬍ 0.005). DISCUSSION By the year 2050, the number of people 85 years of age or older in the United States is expected to increase by 4.5-fold.13 Not surprisingly, a recent report of a large single-institution experience with patients undergoing pancreatic resection demonstrated a significant increase over time in the mean age of their patients.12 In this communication, we sought to determine if the experience of elderly patients undergoing pancreatectomies differed from that of younger patients with respect to nutritional and functional outcomes that can affect quality of life. We conducted a retrospective review of pancreatectomy outcomes statewide in California and at a single
704
Lightner et al
Pancreatic Resection in the Elderly
J Am Coll Surg
Table 5. Common Diagnoses* at Readmission of The University of California, San Francisco Patients All patients
Elderly (>75 y)
Young (<75 y)
Diagnosis
n
%
n
%
n
%
All readmissions Intraabdominal abscess or fluid collection Peripancreatic or anastomotic leak Malnutrition or dehydration Delayed gastrointestinal transit Diabetes mellitus, uncontrolled Urinary tract infection
39 15 13 9 7 3 2
18 7 6 4 3 1 1
7 2 2 5 1 1 2
24 7 7 17† 3 3 7‡
32 13 11 4 6 2 0
18 7 6 2 3 1 0
Data are presented as number of patients (percentage discharged from hospital) readmitted to The University of California, San Francisco (within 30 days of initial discharge) for complications related to the index procedure. *Other diagnoses were: abdominal pain, altered mental status, atrial fibrillation, cholecystitis, deep venous thrombosis, fever of unknown source, intraabdominal hemorrhage, myocardial infarction, portal vein thrombosis, recurrent or advancing cancer, small bowel obstruction, wound infection, and upper gastrointestinal bleed. † p ⬍ 0.005 versus young (chi-square test). ‡ p ⬍ 0.05 versus young (chi-square test).
university hospital. We evaluated the hospital course of elderly patients on a cohort basis by examining results for patients aged 75 years or older compared with those younger than 75 years. Our data show that elderly patients undergoing major pancreatic resection are more likely than their younger counterparts to be admitted to the ICU, to require supplemental enteral nutrition postoperatively, to be transferred to a skilled nursing facility at the time of discharge, and to be readmitted with malnutrition or dehydration. Results from the California statewide database indicated that mortality after pancreatectomy was higher among elderly patients than young patients (10% versus 7%, p ⬍ 0.05). These data are consistent with mortality figures after major pancreatic resection from other multiinstitutional studies, where the incidence of death is 6% to 11%.1,14-17 The mortality rate of 10% for elderly patients in CA was also similar to that of other published multiinstitutional studies of pancreas resection in the elderly (6% to 15%).1,15,18 Data on whether age is a risk factor for mortality after pancreatectomy are conflicting.1,15,18,19 In our single-institution, UCSF experience, mortality was low (3%) and was equal among elderly and young patients. Our results, and others based on large, single-institution analyses, suggest that age alone is not a risk factor for postoperative death.16,20,21 At UCSF, 47% of older versus 20% of young patients required an ICU stay during their hospitalization after pancreatectomy, and older patients tended to remain in the ICU longer (2.3 days versus 1.3 days). The relative contributions of increased cardiac and respiratory complications versus a perceived need for closer monitoring
in older patients are difficult to distinguish in retrospect. Our findings differ from those of Fong and colleagues,20 who studied 138 patients older than 70 years after pancreas resection and found that 19% of both the elderly and young were admitted to the ICU. Because the median age of patients in Fong’s study was 75 years, as compared to 79 years for our older cohort, their inclusion of younger patients may have negated the difference. Elderly patients were more likely than the young to suffer from a major cardiac event (13% versus 0.5%; p ⬍ 0.005). They also had a higher frequency of pneumonia, altered mental status, pancreatic leak, intraabdominal abscess, urinary tract infection, C difficile colitis, sepsis, and acute renal failure. Indeed, 70% of elderly UCSF patients required consultation for, or management of, a postoperative complication, compared with 56% of younger patients. These percentages are higher than those reported in many single-institution studies for the elderly (ages greater than or equal to 65 years to greater than or equal to 80 years; incidence 39% to 63%2-6,20,22-27). Our list was intentionally inclusive (listing, for example, depression, altered mental status, hyponatremia, transfusion reaction, and urinary retention). We sought to more fully characterize the differences in the postoperative experience of older patients. Apparent differences among hospital courses of elderly and young UCSF patients (eg, ICU admission, morbidity) did not translate into differences in postoperative or total lengths of stay. This finding corroborates earlier reports.6,20,28 Mean postoperative lengths of stay for the elderly (13.8 ⫾ 1.5 days) and all (13.3 ⫾ 0.7 days) UCSF patients fell below previous reports (19 to
Vol. 198, No. 5, May 2004
Lightner et al
21 days for the elderly, 16 to 21 days for all) in singleinstitution studies.4,29-31 To our knowledge, our results represent the first reported comparison of nutritional status among elderly and younger patients after pancreatectomy. Only 16% of the younger cohort received supplemental enteral nutrition at discharge, 48% of the elderly cohort received enteral tube feedings.1,15,18 Despite selective use of supplemental enteral nutrition, malnutrition/dehydration was the most common diagnosis among elderly patients who were readmitted to UCSF after major pancreatic resection (17%). This finding is in agreement with the recent report of Balcom and associates,12 who noted a relatively high incidence of dehydration among patients readmitted after pancreatectomy.12 Because of the limited nutritional data available, we were not able to determine if elderly patients with pancreas tumors were more severely malnourished preoperatively than their younger counterparts or whether the differences occurred in the postoperative period. Although additional studies are necessary to determine which subsets of patients are at greatest risk for postoperative malnutrition, these data indicate that surgical placement of a feeding jejunostomy tube should be considered for elderly patients undergoing pancreatectomy. This likelihood should be discussed with the patient and family preoperatively. Despite prolonged hospital stays for elderly patients undergoing pancreatic resection in the state of California, 26% were discharged to a skilled nursing facility versus 7% of younger patients. Although the principal reason for transfer cannot be determined in this retrospective analysis, the likelihood of requiring care at a skilled nursing facility after an acute hospitalization is an important index of functional status. Older patients may be at greater risk for significant functional decline after pancreatic resection than younger patients. Our data are similar to other single-institution reports of a relatively low hospital mortality rate for elderly patients undergoing major pancreatic resection (3%). But a more detailed outcomes analysis indicates that patients 75 years old (or older) are more likely than younger patients to be admitted to the ICU, experience a major cardiac complication, require either home nursing care or a skilled nursing facility, require supplemental nutrition at discharge, and be readmitted for nutritional support. These results suggest that the short-term functional and nutritional condition of many elderly patients is compromised after major pancreatic resection. This
information should be included in preoperative discussions with elderly patients considering pancreatectomy.
Pancreatic Resection in the Elderly
705
Author Contributions
Study conception and design: Lightner, Glasgow, Mulvihill, Kirkwood Acquisition of data: Lightner, Glasgow, Jordan Analysis and interpretation of data: Lightner, Glasgow, Jordan, Krassner, Kirkwood Drafting of manuscript: Lightner, Krassner, Kirkwood Critical revision: Way, Mulvihill, Kirkwood Statistical expertise: Glasgow, Kirkwood Obtaining funding: Kirkwood Supervision: Kirkwood REFERENCES 1. Baumel H, Huguier M, Manderscheid JC, et al. Results of resection for cancer of the exocrine pancreas: a study from the French Association of Surgery. Br J Surg 1994;81:102–107. 2. DiCarlo V, Balzano G, Zerbi A, et al. Pancreatic cancer resection in elderly patients. Br J Surg 1998;85:607–610. 3. Kairaluoma M, Kiviniemi H, Stahlberg M. Pancreatic resection for carcinoma of the pancreas and the periampullary region in patients over 70 years of age. Br J Surg 1987;74:116–118. 4. Sohn TA, Yeo CJ, Cameron JL, et al. Should pancreaticoduodenectomy be performed in octogenarians? J Gastrointest Surg 1998;2:207–216. 5. Bathe QF, Caldera H, Hamilton KL, et al. Radical resection of periampullary tumors in the elderly: evaluation of long-term results. World J Surg 2000;24:353–358. 6. Bottger TC, Engelmann R, Junginger T. Is age a risk factor for major pancreatic surgery?—an analysis of 300 resections. Hepato-Gastroenterology 1999;46:2589–2598. 7. Covinsky KE, Martin GE, Beyth RJ, et al. The relationship between clinical assessments of nutritional status and adverse outcomes in older hospitalized patients. J Am Geriatr Soc 1999; 47:532–538. 8. Jones MK, Castillo LA, Hopkins CA, et al. St. Anthony’s compact ICD-9-CM: code book for physician payment. 1995 ed., vols. 1 and 2. Reston, VA: St. Anthony Publishing; 1994. 9. American Society of Anesthesiologists, I. New classification of physical status. Anesthesiology 1963;24:111. 10. Saklad M. Grading of patients for surgical procedures. Anesthesiology 1941;2:281–284. 11. Patel AG, Toyama MT, Kusske AM, et al. Pylorus-preserving Whipple resection for pancreatic cancer: is it any better? Arch Surg 1995;130:838–843. 12. Balcom JH IV, Rattner DW, Warshaw AL, et al. Ten-year experience with 733 pancreatic resections: changing indications, older patients, and decreasing length of hospitalization. Arch Surg 2001;136:391–398. 13. Statistical Abstract of the United States 1998, in The National Data Book. Washington, DC: US Census Bureau; 1998. 14. Bakkevold KE, Kambestad B. Morbidity and mortality after radical and palliative pancreatic cancer surgery. Ann Surg 1993; 217:356–368.
706
Lightner et al
Pancreatic Resection in the Elderly
J Am Coll Surg
15. Edge SB, Schmieg RE Jr, Rosenlof LK, et al. Pancreas cancer resection outcome in American university centers in 1989– 1990. Cancer 1993;71:3502–3508. 16. Gordon TA, Burleyson GP, Tielsch JM, et al. The effects of regionalization on cost and outcome for one general high-risk surgical procedure. Ann Surg 1995;221:43–49. 17. Neoptolemos JP, Russell RC, Bramhall S, et al. Low mortality following resection for pancreatic and periampullary tumours in 1026 patients: UK survey of specialist pancreatic units. Br J Surg 1997;84:1370–1376. 18. Begg CB, Cramer LD, Hoskins WJ, et al. Impact of hospital volume on operative mortality for major cancer surgery. JAMA 1998;280:1747–1751. 19. Wade TP, el-Ghazzawy AG, Virgo KS, et al. The Whipple resection for cancer in U.S. Department of Veterans Affairs Hospitals. Ann Surg 1995;221:241–248. 20. Fong Y, Blumgart LH, Fortner JG, et al. Pancreatic or liver resection for malignancy is safe and effective for the elderly. Ann Surg 1995;222:426–437. 21. Glasgow RE, Mulvihill SJ. Hospital volume influences outcome in patients undergoing pancreatic resection for cancer. West J Med 1996;165:294–300. 22. Delcore R, Thomas JH, Hermreck AS. Pancreaticoduodenectomy for malignant pancreatic and periampullary neoplasms in elderly patients. Am J Surg 1991;162:532–536. 23. Bottger TC, Junginger T. Factors influencing morbidity and
mortality after pancreaticoduodenectomy: critical analysis of 221 resections. World J Surg 1999;23:164–172. Hannoun L, Christophe M, Ribeiro J, et al. A report of fortyfour instances of pancreaticoduodenal resection in patients more than seventy years of age. Surg Gynecol Obstet 1993;177:556– 560. Kayahara M, Nagakawa T, Ueno K, et al. Pancreatic resection for periampullary carcinoma in the elderly. Surg Today 1994;24: 229–233. Al Sharaf K, Andren-Sandberg A, Ihse I. Subtotal pancreatectomy for cancer can be safe in the elderly. Eur J Surg 1999;165: 230–235. Spencer MP, Sarr MG, Nagorney DM. Radical pancreatectomy for pancreatic cancer in the elderly: is it safe and justified? Ann Surg 1990;212:140–143. Brooks AD, Marcus SG, Gradek C, et al. Decreasing length of stay after pancreatoduodenectomy. Arch Surg 2000;135:823– 830. Magistrelli P, Masetti R, Coppola R, et al. Pancreatic resection for periampullary cancer in elderly patients. Hepatogastroenterology 1998;45:242–247. Mosca F, Giolianotti PC, Balestracci T, et al. Long-term survival in pancreatic cancer: pylorus-preserving versus Whipple pancreatoduodenectomy. Surgery 1997;122:553–566. Yeo CJ, Cameron JL, Sohn TA, et al. Six hundred fifty consecutive pancreaticoduodenectomies in the 1990s: pathology, complications, and outcomes. Ann Surg 1997;226:248–260.
24.
25. 26. 27. 28. 29. 30. 31.
Pancreatic Resection in the Elderly Amy M Lightner, MD, Robert E Glasgow, MD, Thomas H Jordan, BS, Alexander D Krassner, BA, Lawrence W Way, MD, Sean J Mulvihill, MD, Kimberly S Kirkwood, MD Elderly patients undergoing pancreatic resection present unique challenges in postoperative care. Although mortality rates among elderly patients after pancreatectomy at high-volume centers is known to be low, the anticipated decline in functional status and nutritional parameters has received little attention. Functional decline is an unrecognized but critically important consequence of pancreatic resection in older patients. STUDY DESIGN: This study is a retrospective review, validation cohort, of older and younger patients undergoing major pancreatic resection. The setting is the state of California (database of all hospitals in the state) and The University of California, San Francisco (UCSF; a tertiary care referral center). The study population is a consecutive sample of older (greater than or equal to 75 years) and younger (16 to 74 years) patients from California (January 1990 to December 1996; n ⫽ 3,113) and UCSF (January 1993 to November 2000; n ⫽ 218), who underwent radical pancreaticoduodenectomy, distal pancreatectomy, or total pancreatectomy for neoplasia. The main outcomes measures were length of stay, complications, mortality, discharge disposition, supplemental nutrition requirement, and readmissions. RESULTS: Elderly patients had higher mortality rates than the young statewide (10% versus 7%, p ⫽ 0.006). Although the 3% mortality at UCSF was the same for both groups, older patients were more often admitted to the ICU (47% versus 20%, p ⫽ 0.003), treated for major cardiac events (13% versus 0.5%, p ⬍ 0.001), discharged with enteral tube feedings (48% versus 16%, p ⬍ 0.001), or malnourished on readmission (17% versus 2%, p ⬍ 0.005). Older patients were more frequently discharged to skilled nursing facilities (17% versus 1% at UCSF; 24% versus 7% in California; p ⬍ 0.001, both groups). CONCLUSIONS: Older patients are more likely than younger patients to require an ICU stay, suffer a cardiac complication, and experience compromised nutritional and functional status after major pancreatic resection. ( J Am Coll Surg 2004;198:697–706. © 2004 by the American College of Surgeons) BACKGROUND:
impact of these procedures on functional status in the elderly. The recent expansion of research in geriatric medicine has brought increasing emphasis on quality of life in the elderly and on the importance of evaluating the impact of medical interventions on functional status. So it is important to determine the impact of pancreatic resection on physical independence at the time of discharge. The incidence of malnutrition among older hospitalized nonsurgical patients has been well-documented,7 but the scope of this condition among patients undergoing abdominal operations is not known. Patients with pancreatic cancer may be at especially high risk for perioperative malnutrition because cachexia frequently attends this diagnosis. Severe malnutrition adversely affects mortality rate, functional status, and the need for skilled nursing home admission after acute hospitaliza-
The aging of the western population has generated an interest in the suitability of older patients for major surgical procedures. Several studies have documented “acceptable” increases in morbidity and mortality rates for older patients undergoing major pancreatic resections.1-6 These standard end points provide little insight into the No competing interests declared.
Supported by NIH T32 DK-07573-12 (Lightner, Kirkwood) and the Claude E Welch Fellowship from the Massachusetts General Hospital (Lightner). Poster presented at the 73rd Annual Meeting of the Pacific Coast Surgical Association, Las Vegas, NV, February 2002. Received August 22, 2003; Revised December 18, 2003; Accepted December 18, 2003. From the Department of Surgery, University of California, San Francisco, CA. Correspondence address: Kimberly S Kirkwood, MD, Department of Surgery, University of California, San Francisco, 533 Parnassus Ave, Room U-372, San Francisco, CA 94143-0790. email: [email protected]
© 2004 by the American College of Surgeons Published by Elsevier Inc.
697
ISSN 1072-7515/04/$30.00 doi:10.1016/j.jamcollsurg.2003.12.023
698
Lightner et al
Pancreatic Resection in the Elderly
Abbreviations and Acronyms
ASA ⫽ American Society of Anesthesiologists OSHPD ⫽ Office of Statewide Health Planning and Development PD ⫽ pancreaticoduodenectomy PPPD ⫽ pylorus-preserving pancreaticoduodenectomy UCSF ⫽ The University of California, San Francisco
tion among older nonsurgical patients.7 So nutritional status is a potentially important determinant of postoperative outcomes. We evaluated outcomes of major pancreas resections for neoplasia using two sources of data: a large population database that includes patient discharge abstracts from every admission to every acute care hospital statewide in California (CA), and at The University of California, San Francisco Moffit-Long Hospital, a highvolume tertiary care referral center (UCSF). Within these groups, we reviewed hospital records, discharge, and readmission data for elderly and young cohorts. We hypothesized that elderly patients experience a more complicated hospital course (increased complications and higher level of care required) than do younger patients. We also predicted that older patients are more susceptible to impaired functional and nutritional status (discharge disposition, need for supplemental nutrition, readmission for malnutrition or dehydration) after pancreatic resection. METHODS The methods for assessing pancreas surgery outcomes at the UCSF and in the state of California were slightly different and are described in this section. Analysis of statewide data provided population-based information and included a broad range of regional practice patterns and hospital volumes. In contrast, a more detailed analysis of the hospital course was afforded by reviewing our single institution experience. The University of California, San Francisco
All patients who underwent major pancreatic resections at the UCSF for a presumed neoplastic process of the pancreas or periampullary region during the period January 1993 through November 2000 were included in the study. Two hundred twenty-four such patients were
J Am Coll Surg
identified through Department of Surgery and medical center coding unit records. Six patients were excluded because of unavailability of complete medical records, bringing the total study population to 218. Thirty patients were age 75 years or older at the time of operation and were considered “elderly.” One hundred eightyeight patients less than 75 years old were grouped into the “younger” cohort. Patient medical records (paper and electronic charts) were reviewed retrospectively. The study was approved by the University of California, San Francisco Committee on Human Research. For purposes of inclusion, a major pancreatic resection was a radical pancreaticoduodenectomy (Whipple procedure), distal pancreatectomy (laparoscopic or open), or total pancreatectomy. A presumed neoplastic process was a mass (other than a presumed pseudocyst) or suspicious intraductal lesion imaged by radiography, esophagogastroduodenoscopy (for ampullary tumors), or endoscopic retrograde cholangiopancreatography. The mass or lesion might have been assumed to be benign or malignant, and may or may not have been biopsied before the operation. Only nonemergent operations were included. Postoperative complications were defined broadly to thoroughly assess patients’ postoperative courses. Morbidities were included in the analysis if they were significant enough to require medical or surgical intervention or consultation. Minor complications were those that required only minimal intervention and were not life-threatening. Significant complications were those that necessitated a major procedure (eg, operation, radiologically guided abscess drainage, endotracheal intubation), close monitoring (step-down or ICU), were potentially life threatening, or delayed discharge. Postoperative mortality was defined as death that occurred during the same hospitalization as the index pancreatic procedure. Longterm survival could not be accurately assessed retrospectively and was not included in our study. Because the UCSF is a tertiary care center in a large urban area, many patients obtain longterm followup care at regional facilities. Readmissions to UCSF within 30 days of discharge from the index pancreatic procedure were included in the study. Hospitalizations were excluded if the diagnosis at readmission was completely unrelated to the operation or underlying cancer diagnosis.
Vol. 198, No. 5, May 2004
Lightner et al
Analysis
mental nutrition requirement at discharge, and readmissions) was not available in the OSHPD database.
Data were analyzed using Microsoft Access 98 and Microsoft Excel 98. Statistical analyses were conducted using the Primer statistics package (Version 3.0, McGrawHill, 1992) and InStat 3. Student’s t-test, chi-square, and Mann-Whitney rank-sum tests were used as appropriate, with statistical significance set at p ⬍ 0.05.
Pancreatic Resection in the Elderly
699
Analysis
Chi-square and t-tests were used for comparisons, and statistical significance was defined as p ⬍ 0.05. The JMP 3.2.6 statistical software program (SAS Institute Inc, Cary, NC) was used for analysis.
State of California
We also conducted a retrospective review of standardized patient discharge abstracts using a database obtained from the California Office of Statewide Health Planning and Development (OSHPD). This database contains discharge data abstracts for every patient hospitalization from each acute care hospital in California. All discharge abstracts from the years 1990 to 1996 were included in the initial database search, and from these abstracts, 3,113 patients (from 318 acute care facilities) who underwent major pancreatic resection for benign or malignant disease were selected. The OSHPD database uses diagnostic and procedural codes derived from the International Classification of Diseases, Ninth Revision, Clinical Modification, fourth edition (ICD-9), issued by the US Department of Health and Human Services.8 Codes used in the database search for operations were: total pancreatectomy (ICD-9 52.6), radical pancreaticoduodenectomy (ICD-9 52.7), proximal pancreatectomy (ICD-9 52.51), radical subtotal pancreatectomy (ICD-9 52.53), pancreatectomy unspecified (ICD-9 52.59), and distal pancreatectomy (ICD-9 52.52). To further select for malignant disease, patients with a diagnosis of cancer of the exocrine pancreas or islet cells (ICD-9 157.X), duodenum (ICD-9 152.0), bile duct (ICD-9 156.1), and the ampulla of Vater (ICD-9 156.2) were included. Patients with benign disease leading to major pancreatic resection were selected using diagnostic codes for the pancreas (ICD-9 211.67, 235.5, 239.0), duodenum (ICD-9 211.2, 235.2), ampulla of Vater (ICD-9 239.0), and bile duct (ICD-9 211.5, 235.3, 239.0). Definitions of the terms elderly, young, and mortality were the same as for the UCSF cohort. Other information available from the OSHPD database and used in this study included total length of stay and disposition after discharge. Some information obtained from UCSF charts (American Society of Anesthesiologists classifications, subsets of lengths of stay, complications, supple-
RESULTS Demographics
Among the 3,113 patients in the study who underwent major pancreatic resections for neoplasia in the state of California, 515 (17%) were 75 years or older (data not shown). Similarly, older patients comprised 14% of the 218 patients who underwent pancreatic resections at UCSF between 1993 and 2000. Approximately equal numbers of men and women underwent pancreatic resection; the older patients had slightly more women (56%) than men in the statewide cohort of 515 subjects. The ASA Physical Status Classification9,10 is commonly used to stratify comorbidity. Although these data were not available for CA patients, the elderly UCSF cohort had a higher median ASA score than did the young cohort (3 versus 2, p ⬍ 0.005). Procedures
Radical pancreaticoduodenectomy was the most common resection performed in both older and younger cohorts, representing 80% of both populations. Pyloruspreserving pancreaticoduodenectomy (PPPD) was performed more often than classic pancreaticoduodenectomy (PD) at UCSF (71% versus 29% of radical pancreaticoduodenectomies), with similar percentages in the elderly and young cohorts. Median ASA classifications between the PPPD and PD groups were not significantly different (data not shown). Diagnoses
Table 1 shows the pathologic diagnoses among both populations of patients. Malignant tumors made up the majority of diagnoses, in both the statewide and UCSF cohorts (90% and 83%, respectively). The malignancies were most often of the exocrine pancreas. Among the UCSF patients, benign and malignant neuroendocrine tumors of the pancreas were seen exclusively in the young cohort. The mean age of patients with malignant disease at UCSF was greater than that of patients with
700
Lightner et al
Pancreatic Resection in the Elderly
J Am Coll Surg
Table 1. Pathologic Diagnoses CA Primary tumor site/diagnosis
Malignant disease Pancreas Exocrine Adenocarcinoma Cystic Intraductal papillary Anaplastic Neuroendocrine Ampulla Adenocarcinoma Neuroendocrine Duodenum Adenocarcinoma Gastrointestinal stromal tumor Neuroendocrine Common bile duct Other Benign disease Pancreas Exocrine Cystic Solid Neuroendocrine Duodenum Common bile duct
All patients
2,816 (90) 1,870 (60) 1,807 (58) N/A‡ N/A N/A N/A 63 (2) 507 (16) N/A N/A 187 (6) N/A N/A N/A 216 (7) 36 (1) 297 (10) 291 (9) 214 N/A N/A 77 6 (0.2) 0 (0)
UCSF All patients
181 (83) 95 (44) 79 (36) 65 7 4 3 16 (7) 39 (18) 38 1 14 (6) 9 3 2 13 (6) 20 (9) 37 (17) 33 (15) 28 22 6 5 3 (1) 1 (0.5)
Elderly >75 y
26 (87) 10 (33) 10 (33) 7 1 0 2 0 8 (27) 8 0 3 (10) 0 2 1 2 (7) 3 (10) 4 (13) 3 (10) 3 1 2 0 1 (3) 0 (0)
Young <75 y
155 (82) 85 (45) 69 (37) 58 6 4 1 16 (9) 31 (16) 30 1 11 (6) 9 1 1 11 (6) 17 (9) 33 (18) 30 (16) 25 21 4 5 2 (1) 1 (0.5)
p Value*
⬍0.005† ⬍0.005† ⬍0.005† — — — — — — — — — — — — — ⬍0.005† ⬍0.005† ⬍0.05† — — — — — —
Data are presented as number (percentage) of patients for the state of California (CA) and The University of California, San Francisco (UCSF). *p value shown only if ⬍0.05. † CA versus all UCSF patients (chi-square test). ‡ Data not available.
benign disease (61.9 ⫾ 1.0 years versus 54.8 ⫾ 2.6 years, p ⫽ 0.005, t-test). Postoperative course Complications
Fifty-eight percent of UCSF-patients (70% of the elderly and 56% of the young) had one or more complications postoperatively (Table 2). The mean age of patients who had complications was greater than those who did not (63 ⫾ 1.1 years versus 57 ⫾ 1.6 years, p ⬍ 0.005, t-test). Patients who underwent radical pancreaticoduodenectomy were more likely to have a postoperative complication than those who underwent distal pancreatectomy (p ⬍ 0.005, chi-square). Major cardiac complications were more frequent among older patients (Table 2). Four elderly patients (13% of the cohort) and one young patient (0.5% of the
cohort; p ⬍ 0.005 by chi-square) had an acute myocardial infarction, cardiac arrest, or congestive heart failure. A peripancreatic or anastomotic leak was the second most common complication overall (10% of patients) and occurred in 13% of elderly versus 9% of young patients. Older patients had a greater likelihood of intraabdominal abscesses, pneumonia, altered mental status, urinary tract infection, Clostridium difficile colitis, sepsis, and acute renal failure, but these differences were not significant. Delayed gastric emptying after pancreaticoduodenectomy prolongs the hospital course and affects nutrition.11,12 Our retrospective analysis did not permit discrimination among delayed gastric emptying, ileus, and partial small bowel obstruction. So we used the term delayed gastrointestinal transit, to include these three conditions for which notes in the medical record indicated
Lightner et al
Vol. 198, No. 5, May 2004
Pancreatic Resection in the Elderly
701
Table 2. Postoperative Complications of The University of California, San Francisco Patients All patients Complications
All complications* Delayed gastrointestinal transit Peripancreatic or anastomotic leak Wound infection and/or cellulitis Intraabdominal abscess or fluid collection Pneumonia Abdominal reoperation Altered mental status Gastrointestinal bleed Urinary tract infection C difficile colitis Minor cardiac event† Sepsis Acute renal failure Adult respiratory distress syndrome Major cardiac event‡
Elderly (>75 y)
Young (<75 y)
n
%
n
%
n
%
127 48 21 18 14 13 12 9 9 9 8 7 7 6 6 5
58 22 10 8 6 6 6 4 4 4 4 3 3 3 3 2
21 7 4 2 3 4 0 3 0 2 2 1 2 2 1 4
70 23 13 7 10 13 0 10 0 7 7 3 7 7 3 13§
106 41 17 16 11 9 12 6 9 7 6 6 5 4 5 1
56 22 9 9 6 5 6 3 5 4 3 3 3 2 3 0.5
Data are presented as number (percentage) of all (218 total), elderly (30 total), and young (188 total) The University of California, San Francisco patients. *Complications (total number) not listed in table were: adverse drug reaction (1), anorexia requiring total parenteral nutrition (1), Candida esophagitis (2), chyle leak (3), depression (1), diarrhea (non-C difficile), disseminated intravascular coagulopathy (1), hearing loss (acute)(1), hepatic lobe infarction (1), hyperglycemia (1), hyponatremia (1), influenza A infection (1), intraabdominal hemorrhage (2), ischemic bowel (2), necrotizing wound infection (2), nerve palsy (1), operation (nonabdominal)(1), pleural effusion requiring drainage (3), pulmonary edema (1), pulmonary embolus (1), sinusitis (2), small bowel obstruction (complete)(1), superficial thrombophlebitis (1), technical misadventure (1), transfusion reaction (1), urinary retention (1), wound dehiscence (1). † Benign arrhythmia, requiring pharmacologic intervention. ‡ Acute myocardial infarction, cardiac arrest, or congestive heart failure. § p ⬍ 0.005 versus young cohort (chi-square test).
resultant impairment of enteral nutrition. Delayed gastrointestinal transit affected 23% and 22% of elderly and young patients, respectively, at UCSF (Table 2). Delayed gastrointestinal transit occurred in 29% of patients who underwent PPPD compared with 28% for PD. Intensive care unit (ICU) admission and length of stay
Older patients were more likely to require admission to the ICU (47%) than younger patients (20%, p ⬍ 0.005, chi-square Table 3). Although the elderly tended to have a greater mean length of stay in the ICU than the young (2.3 ⫾ 0.7 days versus 1.3 ⫾ 0.3 days), these differences were not significant. Postoperative and total lengths of stay
Average postoperative and total lengths of stay at the UCSF were 13.3 ⫾ 0.7 days and 15.1 ⫾ 0.7 days, respectively (Table 3). Comparison of lengths of stay between elderly and young cohorts did not yield appreciable differences. Review of the statewide data for total lengths of stay
showed that patients in CA stayed in the hospital for an average of 20.1 ⫾ 0.3 days after major pancreatic resection. Elderly patients in the CA group had longer lengths of stay than their younger counterparts (21.4 ⫾ 0.7 days versus 19.8 ⫾ 0.3 days, p ⬍ 0.05, t-test). The longer hospital stays among patients in the statewide database as compared with those at UCSF might, in part, be explained by the fact that the UCSF review included some patients treated more recently, during a time period when hospital stays have been reduced. Mortality
Three percent of patients at the UCSF and 7% in CA died in the hospital after major pancreatic resection for neoplasia (Table 3, p ⬍ 0.05 by chi-square). The mortality rate at UCSF was the same for elderly and young patients (3% of each group), but statewide, the elderly had a higher mortality rate (10% versus 7%, p ⬍ 0.05 by chi-square). Six of seven deaths at UCSF occurred after pancreaticoduodenectomies (one was after total pancreatectomy).
702
Lightner et al
Pancreatic Resection in the Elderly
J Am Coll Surg
Table 3. Postoperative Course CA Variable
Intensive care unit (ICU) stay Yes No ICU length of stay (d) Mean ⫾ SEM Range Postoperative length of stay (d) Mean ⫾ SEM Range Total length of stay (d) Mean ⫾ SEM Range Mortality§
UCSF
All patients
Elderly (>75 y)
Young (<75 y)
All patients
Elderly (>75 y)
Young (<75 y)
N/A* N/A
N/A N/A
N/A N/A
52 (24) 166 (76)
14 (47)† 16 (53)
38 (20) 150 (80)
N/A N/A
N/A N/A
N/A N/A
1.4 ⫾ 0.3 0–34
2.3 ⫾ 0.7 0–17
1.3 ⫾ 0.3 0–34
N/A N/A
N/A N/A
N/A N/A
13.3 ⫾ 0.7 2–82
13.8 ⫾ 1.5 7–39
13.2 ⫾ 0.7 2–82
20.1 ⫾ 0.3 N/A 233 (7)
21.4 ⫾ 0.7† N/A 54 (10)‡
19.8 ⫾ 0.3 N/A 179 (7)
15.1 ⫾ 0.7 3–84 7 (3)
16 ⫾ 1.5 8–41 1 (3)
15.0 ⫾ 0.7 3–84 6 (3)
Data are presented as number (percentage) of patients for the state of California (CA) and The University of California, San Francisco (UCSF). *Data not available. † p ⬍ 0.005, elderly versus young cohort, UCSF (chi-square test). ‡ p ⬍ 0.05, elderly versus young cohort, CA (t-test). § During same hospital admission as index procedure.
Five of the seven patients who died were reexplored during their hospital courses, and five eventually died of sepsis. Four patients experienced adult respiratory distress syndrome, two had intestinal infection, and two had acute renal failure. The one elderly patient who died succumbed from respiratory failure. Deaths occurred from postoperative days 2 through 82 (mean day 31). Posthospital course Discharge disposition
After pancreatectomy, surviving older patients were less likely to go home than younger patients in both the
UCSF and the CA groups (Table 4). Only 7% of younger patients in the statewide cohort recovered at a skilled nursing facility, but 26% of older patients were transferred to a skilled nursing facility after their acute hospitalization. Figure 1 shows details of the posthospital course for UCSF patients. Only 41% of surviving elderly patients were sent home independently (without services) compared with 75% of young patients (p ⬍ 0.005 versus young cohort). Note that none of the UCSF patients sent to a skilled nursing facility postoperatively had been residing in a chronic care facility preoperatively.
Table 4. Posthospital Course CA Variables
Discharged from hospital,* n (%) Disposition, n (%) Home Hospital transfer Skilled nursing facility
UCSF
All patients
Elderly (>75 y)
Young (<75 y)
All patients
Elderly (>75 y)
Young (<75 y)
2,880 (93)†
461 (90)‡
2,419 (93)
211 (97)
29 (97)
182 (97)
2,589 (90) N/A¶ 291 (10)
339 (74)§ N/A 122 (26)§
2250 (93) N/A 169 (7)
203 (96) 1 (0.5) 7 (3.5)
24 (83)㛳 0 (0) 5 (17)㛳
179 (98.5) 1 (0.5) 2 (1)
*Data are presented as number (%) of patients for the state of California (CA, 3,113 total patients) and The University of California, San Francisco (UCSF, 218 total patients). † p ⬍ 0.05 versus respective cohort in UCSF group (chi-square test). ‡ p ⬍ 0.05 versus young cohort, CA (chi-square test). § p ⬍ 0.005 versus young cohort, CA (chi-square test). 㛳 p ⬍ 0.005 versus young cohort, UCSF (chi-square test). ¶ Data not available.
Vol. 198, No. 5, May 2004
Lightner et al
Pancreatic Resection in the Elderly
703
Figure 1. Disposition of surviving University of California San Francisco (UCSF) patients at the time of discharge after pancreatic resection. Displayed are percentages of 29 elderly and 182 young patients discharged home (with and without assistance), skilled nursing facilities, or with enteral tube feedings. *p ⬍ 0.005 versus young (chi-square).
Supplemental nutrition
Overall, 22% of surviving patients at UCSF required supplemental nutrition at the time of discharge. Among the 48% of surviving elderly patients who required supplemental nutrition after discharge, tube feeds were delivered through surgically placed feeding jejunostomy tubes in every case. There were no complications attributed to jejunostomy tube placement. Younger patients were less likely than older patients to require supplemental nutrition (16%, p ⬍ 0.005 by chi-square; Fig. 1). Patients who were discharged with tube feeds had longer mean ICU, postoperative, and total lengths of stay (p ⬍ 0.05 for all, by t-test, data not shown) than patients discharged without supplemental nutrition. Patients who had a pancreaticoduodenectomy procedure were more likely than patients who underwent distal or total pancreatic resection to require supplemental nutrition at discharge (p ⬍ 0.005 by chi-square). The use of tube feeds was similar among PD and PPPD patients. Readmissions
Thirty-nine (18%) of the 211 surviving UCSF patients were readmitted within 30 days of discharge for a problem directly related to the index procedure or disease
(Table 5). The overall percentages of elderly (24%) and young (18%) who were readmitted were not significantly different. Table 5 lists the diagnoses among those readmitted to UCSF. Malnutrition or dehydration were the most common reasons for readmission in older patients, and the elderly were eight times more likely to require admission for these conditions that the younger cohort (p ⬍ 0.005). DISCUSSION By the year 2050, the number of people 85 years of age or older in the United States is expected to increase by 4.5-fold.13 Not surprisingly, a recent report of a large single-institution experience with patients undergoing pancreatic resection demonstrated a significant increase over time in the mean age of their patients.12 In this communication, we sought to determine if the experience of elderly patients undergoing pancreatectomies differed from that of younger patients with respect to nutritional and functional outcomes that can affect quality of life. We conducted a retrospective review of pancreatectomy outcomes statewide in California and at a single
704
Lightner et al
Pancreatic Resection in the Elderly
J Am Coll Surg
Table 5. Common Diagnoses* at Readmission of The University of California, San Francisco Patients All patients
Elderly (>75 y)
Young (<75 y)
Diagnosis
n
%
n
%
n
%
All readmissions Intraabdominal abscess or fluid collection Peripancreatic or anastomotic leak Malnutrition or dehydration Delayed gastrointestinal transit Diabetes mellitus, uncontrolled Urinary tract infection
39 15 13 9 7 3 2
18 7 6 4 3 1 1
7 2 2 5 1 1 2
24 7 7 17† 3 3 7‡
32 13 11 4 6 2 0
18 7 6 2 3 1 0
Data are presented as number of patients (percentage discharged from hospital) readmitted to The University of California, San Francisco (within 30 days of initial discharge) for complications related to the index procedure. *Other diagnoses were: abdominal pain, altered mental status, atrial fibrillation, cholecystitis, deep venous thrombosis, fever of unknown source, intraabdominal hemorrhage, myocardial infarction, portal vein thrombosis, recurrent or advancing cancer, small bowel obstruction, wound infection, and upper gastrointestinal bleed. † p ⬍ 0.005 versus young (chi-square test). ‡ p ⬍ 0.05 versus young (chi-square test).
university hospital. We evaluated the hospital course of elderly patients on a cohort basis by examining results for patients aged 75 years or older compared with those younger than 75 years. Our data show that elderly patients undergoing major pancreatic resection are more likely than their younger counterparts to be admitted to the ICU, to require supplemental enteral nutrition postoperatively, to be transferred to a skilled nursing facility at the time of discharge, and to be readmitted with malnutrition or dehydration. Results from the California statewide database indicated that mortality after pancreatectomy was higher among elderly patients than young patients (10% versus 7%, p ⬍ 0.05). These data are consistent with mortality figures after major pancreatic resection from other multiinstitutional studies, where the incidence of death is 6% to 11%.1,14-17 The mortality rate of 10% for elderly patients in CA was also similar to that of other published multiinstitutional studies of pancreas resection in the elderly (6% to 15%).1,15,18 Data on whether age is a risk factor for mortality after pancreatectomy are conflicting.1,15,18,19 In our single-institution, UCSF experience, mortality was low (3%) and was equal among elderly and young patients. Our results, and others based on large, single-institution analyses, suggest that age alone is not a risk factor for postoperative death.16,20,21 At UCSF, 47% of older versus 20% of young patients required an ICU stay during their hospitalization after pancreatectomy, and older patients tended to remain in the ICU longer (2.3 days versus 1.3 days). The relative contributions of increased cardiac and respiratory complications versus a perceived need for closer monitoring
in older patients are difficult to distinguish in retrospect. Our findings differ from those of Fong and colleagues,20 who studied 138 patients older than 70 years after pancreas resection and found that 19% of both the elderly and young were admitted to the ICU. Because the median age of patients in Fong’s study was 75 years, as compared to 79 years for our older cohort, their inclusion of younger patients may have negated the difference. Elderly patients were more likely than the young to suffer from a major cardiac event (13% versus 0.5%; p ⬍ 0.005). They also had a higher frequency of pneumonia, altered mental status, pancreatic leak, intraabdominal abscess, urinary tract infection, C difficile colitis, sepsis, and acute renal failure. Indeed, 70% of elderly UCSF patients required consultation for, or management of, a postoperative complication, compared with 56% of younger patients. These percentages are higher than those reported in many single-institution studies for the elderly (ages greater than or equal to 65 years to greater than or equal to 80 years; incidence 39% to 63%2-6,20,22-27). Our list was intentionally inclusive (listing, for example, depression, altered mental status, hyponatremia, transfusion reaction, and urinary retention). We sought to more fully characterize the differences in the postoperative experience of older patients. Apparent differences among hospital courses of elderly and young UCSF patients (eg, ICU admission, morbidity) did not translate into differences in postoperative or total lengths of stay. This finding corroborates earlier reports.6,20,28 Mean postoperative lengths of stay for the elderly (13.8 ⫾ 1.5 days) and all (13.3 ⫾ 0.7 days) UCSF patients fell below previous reports (19 to
Vol. 198, No. 5, May 2004
Lightner et al
21 days for the elderly, 16 to 21 days for all) in singleinstitution studies.4,29-31 To our knowledge, our results represent the first reported comparison of nutritional status among elderly and younger patients after pancreatectomy. Only 16% of the younger cohort received supplemental enteral nutrition at discharge, 48% of the elderly cohort received enteral tube feedings.1,15,18 Despite selective use of supplemental enteral nutrition, malnutrition/dehydration was the most common diagnosis among elderly patients who were readmitted to UCSF after major pancreatic resection (17%). This finding is in agreement with the recent report of Balcom and associates,12 who noted a relatively high incidence of dehydration among patients readmitted after pancreatectomy.12 Because of the limited nutritional data available, we were not able to determine if elderly patients with pancreas tumors were more severely malnourished preoperatively than their younger counterparts or whether the differences occurred in the postoperative period. Although additional studies are necessary to determine which subsets of patients are at greatest risk for postoperative malnutrition, these data indicate that surgical placement of a feeding jejunostomy tube should be considered for elderly patients undergoing pancreatectomy. This likelihood should be discussed with the patient and family preoperatively. Despite prolonged hospital stays for elderly patients undergoing pancreatic resection in the state of California, 26% were discharged to a skilled nursing facility versus 7% of younger patients. Although the principal reason for transfer cannot be determined in this retrospective analysis, the likelihood of requiring care at a skilled nursing facility after an acute hospitalization is an important index of functional status. Older patients may be at greater risk for significant functional decline after pancreatic resection than younger patients. Our data are similar to other single-institution reports of a relatively low hospital mortality rate for elderly patients undergoing major pancreatic resection (3%). But a more detailed outcomes analysis indicates that patients 75 years old (or older) are more likely than younger patients to be admitted to the ICU, experience a major cardiac complication, require either home nursing care or a skilled nursing facility, require supplemental nutrition at discharge, and be readmitted for nutritional support. These results suggest that the short-term functional and nutritional condition of many elderly patients is compromised after major pancreatic resection. This
information should be included in preoperative discussions with elderly patients considering pancreatectomy.
Pancreatic Resection in the Elderly
705
Author Contributions
Study conception and design: Lightner, Glasgow, Mulvihill, Kirkwood Acquisition of data: Lightner, Glasgow, Jordan Analysis and interpretation of data: Lightner, Glasgow, Jordan, Krassner, Kirkwood Drafting of manuscript: Lightner, Krassner, Kirkwood Critical revision: Way, Mulvihill, Kirkwood Statistical expertise: Glasgow, Kirkwood Obtaining funding: Kirkwood Supervision: Kirkwood REFERENCES 1. Baumel H, Huguier M, Manderscheid JC, et al. Results of resection for cancer of the exocrine pancreas: a study from the French Association of Surgery. Br J Surg 1994;81:102–107. 2. DiCarlo V, Balzano G, Zerbi A, et al. Pancreatic cancer resection in elderly patients. Br J Surg 1998;85:607–610. 3. Kairaluoma M, Kiviniemi H, Stahlberg M. Pancreatic resection for carcinoma of the pancreas and the periampullary region in patients over 70 years of age. Br J Surg 1987;74:116–118. 4. Sohn TA, Yeo CJ, Cameron JL, et al. Should pancreaticoduodenectomy be performed in octogenarians? J Gastrointest Surg 1998;2:207–216. 5. Bathe QF, Caldera H, Hamilton KL, et al. Radical resection of periampullary tumors in the elderly: evaluation of long-term results. World J Surg 2000;24:353–358. 6. Bottger TC, Engelmann R, Junginger T. Is age a risk factor for major pancreatic surgery?—an analysis of 300 resections. Hepato-Gastroenterology 1999;46:2589–2598. 7. Covinsky KE, Martin GE, Beyth RJ, et al. The relationship between clinical assessments of nutritional status and adverse outcomes in older hospitalized patients. J Am Geriatr Soc 1999; 47:532–538. 8. Jones MK, Castillo LA, Hopkins CA, et al. St. Anthony’s compact ICD-9-CM: code book for physician payment. 1995 ed., vols. 1 and 2. Reston, VA: St. Anthony Publishing; 1994. 9. American Society of Anesthesiologists, I. New classification of physical status. Anesthesiology 1963;24:111. 10. Saklad M. Grading of patients for surgical procedures. Anesthesiology 1941;2:281–284. 11. Patel AG, Toyama MT, Kusske AM, et al. Pylorus-preserving Whipple resection for pancreatic cancer: is it any better? Arch Surg 1995;130:838–843. 12. Balcom JH IV, Rattner DW, Warshaw AL, et al. Ten-year experience with 733 pancreatic resections: changing indications, older patients, and decreasing length of hospitalization. Arch Surg 2001;136:391–398. 13. Statistical Abstract of the United States 1998, in The National Data Book. Washington, DC: US Census Bureau; 1998. 14. Bakkevold KE, Kambestad B. Morbidity and mortality after radical and palliative pancreatic cancer surgery. Ann Surg 1993; 217:356–368.
706
Lightner et al
Pancreatic Resection in the Elderly
J Am Coll Surg
15. Edge SB, Schmieg RE Jr, Rosenlof LK, et al. Pancreas cancer resection outcome in American university centers in 1989– 1990. Cancer 1993;71:3502–3508. 16. Gordon TA, Burleyson GP, Tielsch JM, et al. The effects of regionalization on cost and outcome for one general high-risk surgical procedure. Ann Surg 1995;221:43–49. 17. Neoptolemos JP, Russell RC, Bramhall S, et al. Low mortality following resection for pancreatic and periampullary tumours in 1026 patients: UK survey of specialist pancreatic units. Br J Surg 1997;84:1370–1376. 18. Begg CB, Cramer LD, Hoskins WJ, et al. Impact of hospital volume on operative mortality for major cancer surgery. JAMA 1998;280:1747–1751. 19. Wade TP, el-Ghazzawy AG, Virgo KS, et al. The Whipple resection for cancer in U.S. Department of Veterans Affairs Hospitals. Ann Surg 1995;221:241–248. 20. Fong Y, Blumgart LH, Fortner JG, et al. Pancreatic or liver resection for malignancy is safe and effective for the elderly. Ann Surg 1995;222:426–437. 21. Glasgow RE, Mulvihill SJ. Hospital volume influences outcome in patients undergoing pancreatic resection for cancer. West J Med 1996;165:294–300. 22. Delcore R, Thomas JH, Hermreck AS. Pancreaticoduodenectomy for malignant pancreatic and periampullary neoplasms in elderly patients. Am J Surg 1991;162:532–536. 23. Bottger TC, Junginger T. Factors influencing morbidity and
mortality after pancreaticoduodenectomy: critical analysis of 221 resections. World J Surg 1999;23:164–172. Hannoun L, Christophe M, Ribeiro J, et al. A report of fortyfour instances of pancreaticoduodenal resection in patients more than seventy years of age. Surg Gynecol Obstet 1993;177:556– 560. Kayahara M, Nagakawa T, Ueno K, et al. Pancreatic resection for periampullary carcinoma in the elderly. Surg Today 1994;24: 229–233. Al Sharaf K, Andren-Sandberg A, Ihse I. Subtotal pancreatectomy for cancer can be safe in the elderly. Eur J Surg 1999;165: 230–235. Spencer MP, Sarr MG, Nagorney DM. Radical pancreatectomy for pancreatic cancer in the elderly: is it safe and justified? Ann Surg 1990;212:140–143. Brooks AD, Marcus SG, Gradek C, et al. Decreasing length of stay after pancreatoduodenectomy. Arch Surg 2000;135:823– 830. Magistrelli P, Masetti R, Coppola R, et al. Pancreatic resection for periampullary cancer in elderly patients. Hepatogastroenterology 1998;45:242–247. Mosca F, Giolianotti PC, Balestracci T, et al. Long-term survival in pancreatic cancer: pylorus-preserving versus Whipple pancreatoduodenectomy. Surgery 1997;122:553–566. Yeo CJ, Cameron JL, Sohn TA, et al. Six hundred fifty consecutive pancreaticoduodenectomies in the 1990s: pathology, complications, and outcomes. Ann Surg 1997;226:248–260.
24.
25. 26. 27. 28. 29. 30. 31.