Posterior mediastinal goiter.

Because of their tendency to progressively enlarge with compression of adjacent structures, as well as the small chance of malignancy, most intrathoracic goiters should be excised surgically. Most anterior substernal goiters and some ipsilateral posterior mediastinal goiters can be removed safely through a cervical incision. Large posterior mediastinal goiters, contralateral retrotracheal or retroesophageal posterior mediastinal goiters, and isolated mediastinal goiters with no significant cervical connection are best removed through a combined cervical and thoracic approach. The Lahey Clinic experience with three patients with posterior mediastinal goiter is described.

Analysis of morbidity and mortality rates in right hepatectomy with the preoperative APACHE II score.

Morbidity and mortality rates in 70 patients who underwent major liver resection for liver tumors (primary and metastatic) were determined and correlated with the preoperative APACHE II score. Patients were divided into three groups according to their preoperative APACHE II score: low (0 to 3), mid (4 to 7), and high (8 and above). A higher score was closely correlated with increased postoperative morbidity and operative mortality rates. The group with low scores had a postoperative morbidity rate of 34% and a mortality rate of 0%, the group with mid scores had a postoperative morbidity rate of 54% and a mortality rate of 3%, and the group with high scores had a postoperative morbidity rate of 80% and a mortality rate of 20%. Age did not correlate with morbidity. It was therefore postulated that morbidity and mortality rates were related to the combination of points for abnormal physiologic variables and points for chronic health, or APACHE II score minus points for age. As the combination of these points increases the postoperative morbidity and operative mortality rates increase significantly (from 24% in the 0-point group to 69% in the greater than or equal to 3-point group). Also the two deaths occurred in the group with 3 or more points. The preoperative APACHE II score may be used by clinicians to evaluate before surgery the risk of postoperative morbidity and death in elective major liver surgery.

The effect of thyroid hormone administration upon survival in patients with differentiated thyroid carcinoma.

Seven hundred sixty-one patients with operable differentiated thyroid carcinoma were treated between 1931 and 1970. Median follow-up time was 18 years and ranged from 5 to 40 years. Sixty-three percent of the patients were followed more than 15 years, and 46% were followed more than 20 years. Because resections usually spared sufficient thyroid tissue for homeostasis, thyroid hormone was not routinely prescribed after operations performed before 1960. Altogether 244 patients with papillary carcinoma and 76 patients with follicular carcinoma received thyroid hormone, while 296 patients with papillary carcinoma and 45 patients with follicular carcinoma did not receive thyroid hormone after operation. With papillary carcinoma, 14% of men greater than 40 and women greater than 50 years of age (high risk) but only 2% of men less than or equal to 40 years of age and women less than or equal to 50 years of age (low risk) died of disease (P 0.0001). Twenty-six percent of high-risk but only 4% of patients with low-risk follicular cancer died (P 0.0001). However, there was no statistically significant improvement in survival times with use of thyroid hormone when patients were categorized by risk group and pathology. These data support the importance of age and sex in previously described risk groups; these factors supersede the effects of adjuvant treatment. This absence of effect on survival times calls into question current recommendations for routine use of thyroid hormone after surgical therapy. Conceptually, such absence of adjunctive hormone effect on survival time after operation is similar to hormonal effects in other endocrine cancers, which may nevertheless provide good palliation in some cases.

Local resection for ampullary tumors. Is there a place for it?

Two of 14 patients with adenomas were without disease 25 and 43 months after ampullary resection. Two patients with an initial diagnosis of malignant neoplasm had no recurrence at 75 and 40 months; one underwent pancreatoduodenectomy at 8 months because of recurrence. Six of nine patients with initial diagnoses of villous adenoma were without disease at 1, 2, 16, 23, 46, and 51 months; three underwent conversion to pancreatoduodenectomy because of invasive carcinoma. Frozen-section studies revealed adenocarcinoma in two patients with villous adenoma but failed to show invasion in one patient. One patient with villous adenoma was mistakenly thought to have carcinoma based on results of frozen-section studies. Local ampullary resection is valuable in treating benign and selected premalignant and malignant ampullary lesions. The threshold for conversion to pancreatoduodenectomy should be low unless ampullectomy is performed with palliative intent.

Pylorus-preserving pancreatoduodenectomy. Is it an adequate cancer operation.

OBJECTIVE: To assess the adequacy of pylorus-preserving pancreatoduodenectomy (PPPD) as a definitive surgical treatment for primary malignant diseases of the periampullary region. DESIGN: Retrospective review of the clinical records of patients undergoing PPPD for malignant diseases of the periampullary region. Median and 5-year actuarial survival by type and stage of cancer were determined. Survival data from this study were compared with those of patients undergoing a conventional Whipple operation. SETTING: Lahey Clinic, Burlington, Mass. STUDY PARTICIPANTS: One hundred six patients undergoing PPPD for primary malignant disease of the periampullary region between November 1979 and June 1992. INTERVENTION: Pylorus-preserving pancreatoduodenectomy was performed with curative intent in the 106 patients. Ninety-five patients underwent proximal pancreatectomy; 11 patients, total pancreatectomy. Resection of the portal vein was performed in 10 patients. MAIN OUTCOME MEASURE: Long-term survival following PPPD was analyzed with respect to the type and stage of cancer. Median follow-up was 30 months (range, 6 to 156 months). RESULTS: Five-year actuarial survival rates were 45.4% for patients with ampullary adenocarcinoma; 6.6%, with pancreatic ductal adenocarcinoma; 33.3%, with distal bile duct adenocarcinoma; 75%, with pancreatic islet cell adenocarcinoma; and 0%, with pancreatic cystadenocarcinoma. An early cancer stage was associated with more favorable survival for ampullary and distal bile duct adenocarcinomas. For pancreatic ductal adenocarcinoma only, tumors less than 2 cm were associated with better survival. Duodenal resection margins were free of disease in all patients, while peripancreatic and retroperitoneal extension of the tumor was found in 20%. CONCLUSION: For patients with periampullary malignant disease, long-term survival following PPPD is similar to that following a conventional Whipple operation. The potential benefits of hemigastrectomy with perigastric lymphadenectomy are frequently obviated by the presence of positive margins and lymph nodes elsewhere, ie, in the retroperitoneum. We advocate PPPD as the procedure of choice for locally resectable malignant disease of the periampullary region, provided the duodenal margin is viable and tumor free.

Long-term results of pylorus-preserving pancreatoduodenectomy for chronic pancreatitis.

OBJECTIVE: To assess the long-term outcome of patients following pylorus-preserving pancreatoduodenectomy (PPPD) for chronic pancreatitis. DESIGN: Retrospective study with mean follow-up of 63 months (range, 1 month to 13.7 years). SETTING: Tertiary referral hospital. PATIENTS: Records of all patients who underwent PPPD for chronic pancreatitis at Lahey Clinic were reviewed. All patients who were alive were contacted by telephone. In cases where patients had died, information was gathered from family members and hospital records. RESULTS: Forty-five patients underwent PPPD for disabling chronic pancreatitis. The mean preoperative duration of pain was 50 months, with 32 patients (70%) requiring daily narcotics. In one patient resection of the portal vein was required. One patient died within 30 days of the operation. Forty-one patients (92%) had improvement of pain at 5 years. The mean pain score (on a scale of 0 to 10) was 9.2 preoperatively and 1.5, 0.8, 1.1, and 1.1 at 6 months, 1 year, 2 years, and 5 years, respectively. Thirty-three patients (74%) had a postoperative weight gain to an average of 92% of their pre-illness weight. New-onset diabetes occurred in six patients (14%) by 6 months and in 21 patients (46%) by 5 years. Hypoglycemia was the cause of death in one patient who underwent total pancreatectomy. Four patients died of causes unrelated to PPPD. Marginal ulcers occurred in five patients (10%). Nine patients required late operations. CONCLUSIONS: In selected patients, resection of the head of the pancreas achieves long-term pain improvement in over 90% of cases. The early development of diabetes mellitus is infrequent, but over longer follow-up periods it reaches prevalence rates similar to those described in patients who have not undergone resection. Weight gain in this group was superior to that previously reported for our patients who underwent "standard Whipple" operation for chronic pancreatitis.

Clinical and pathologic correlations in patients with periampullary tumors.

Perioperative data on 87 patients undergoing pancreatoduodenectomy for periampullary tumors were correlated with pathologic study of operative specimens to identify the accuracy of diagnosis and the factors affecting survival. Accuracy of endoscopic retrograde cholangiopancreatography and computed tomography in locating lesions was 75% and 44%, respectively. Histologic diagnosis before or at the time of resection was available in only 61% of the patients. Carcinoma was correctly diagnosed clinically by the pathologist or the surgeon in 95% (83/87) of patients with 4 patients found to have benign disease on final pathologic examination. Intraoperative diagnosis of site of origin was incorrect in 18% (16/87) of patients. In 28% (23/83) of patients, pathologists identified nodal metastatic disease missed by the surgeon. Survival correlated with nodal and margin status and tumor grade. Tumor size demonstrated no predictive capacity. Although preoperative diagnostic accuracy is less than optimal, surgeons can usually diagnose malignant lesions but more often fail to identify tumor origin and nodal disease. We continue to advocate resection for patients with periampullary lesions thought to be malignant and resectable without a positive histologic diagnosis.

Gastric adenocarcinoma. A disease in transition.

Two hundred eleven gastric adenocarcinomas diagnosed from 1967 to 1982 were analyzed. Thirty-four percent had a proximal location, a proportionate increase from previous decades that suggested a distinctive epidemiology. Diffuse histology occurred in 49% of cases overall and in 55% of unresectable cases, which were also increases from previous decades. No deaths followed curative resections, two (4%) of 50 patients with palliative resections died, and three (6%) of 54 patients who underwent exploration without resection died, indicating improved operative management. Superficial gastric cancer constituted 6% of cases; 91% were cured. Seventeen percent of cases were linitis plastica and required total gastrectomy in 77% of resections; only 13% of patients had curative operations; none were cured. Seventy-nine percent of cases were polypoid or ulcerated focal cancers. Of operable focal cancers, 72% were resected; 27 (47%) of 57 patients who underwent resection for cure survived five years, a distinct improvement from previous reports, as was the overall survival of 21%.

Liver abscess. The need for complete gastrointestinal evaluation.

The cause of liver abscess is frequently obscure at initial presentation. We reviewed the medical records of 20 patients with pyogenic liver abscess over a 6-year period from 1981 to 1987. Liver abscess was suspected in only 3 patients on admission; the most common initial diagnosis was fever of unknown origin. Subsequently, the origin of the abscess was found to be intestinal in 7 patients, pancreatobiliary in 11 patients, and cryptogenic in 2 patients. Eleven patients underwent percutaneous transhepatic drainage of the abscess as the initial treatment, while open operation was the initial treatment in 9 patients. Percutaneous transhepatic drainage was ultimately successful in only 4 patients (36%). In the absence of an obvious pathologic condition of the biliary tract, all patients should undergo full gastrointestinal evaluation.

Management of pancreatic fistulas.

Conservative management of pancreatic fistulas resulting from trauma, operation for tumor, or operation for pancreatitis has met with variable success. To assess optimal management strategies and outcome, we reviewed the records of 35 patients with external pancreatic fistulas (26 patients), pancreatic ascites (6 patients), or pancreatic pleural effusion (3 patients). Treatment included no operation in 5 patients, oversewing of the fistula in 7 patients, internal drainage in 11 patients, and resection in 12 patients. One (3%) postoperative death occurred. The overall rate of operative success was 83% (25 patients). The incidence of recurrent fistulas was about the same regardless of the procedure. Patients treated successfully without operation did not have pancreatitis as an underlying disease. Patient selection is of great importance in the decision to resect or to drain and is based in part on imaging the pancreatic duct and fistula.

Pancreatoduodenectomy in the management of chronic pancreatitis.

The records of 73 consecutive patients who underwent pancreatoduodenectomy for chronic pancreatitis between 1960 and 1985 were reviewed. The median size of the pancreatic duct was 5 mm. Two operative deaths (2.7%) occurred early in the series. Eighty-eight percent, 86%, and 79% of the patients had improvement in pain at six months, two years, and five years, respectively. Diabetes was present preoperatively in 25% of patients and postoperatively in 37%, 45%, and 69% of patients at six months, two years, and five years, respectively. Pancreatic enzyme preparations were used preoperatively by 26% of patients; this use increased to 75% by five years. Only four of 17 late deaths could be related to diabetes or malnutrition. In most patients, pancreatoduodenectomy achieves long-term pain improvement and permits return to normal activities. Selection of patients is important to decrease the late morbidity and mortality.

Cystic neoplasms of the pancreas.

From 1963 to 1983, 26 patients with cystic neoplasms of the pancreas were treated at the Lahey Clinic, Burlington, Mass. Cystadenoma (15 patients) was more common than cystadenocarcinoma (11 patients). Preoperative symptoms, such as abdominal pain, were present for as long as 18 years before diagnosis. The mean size of cysts was 7 cm. Distal pancreatectomy, the most common operation, was performed in ten patients. Eight of the 11 patients with cystadenocarcinoma had metastatic disease at the time of surgical exploration. There was one postoperative death (3.8%). Patients with cystadenocarcinoma had an adjusted median survival time after operation of 6.0 months. The long prodrome in many of the cancer patients suggests that benign cystadenomas, particularly of the mucinous type, may undergo malignant degeneration. Benign cystadenoma seems unlikely to recur after adequate resection. Whenever possible, complete excision of cystadenoma and cystadenocarcinoma is the procedure of choice.

Management of bile duct cysts in adults.

In 31 adult patients with bile duct cysts seen at the Lahey Clinic (Burlington, Mass) during a 20-year period, the median age at time of initial therapy at Lahey Clinic was 34 years. Abdominal pain was the most common presenting symptom, followed by jaundice and fever. The 31 patients underwent a total of 86 biliary tract procedures, of which 37 were performed at Lahey Clinic. Internal drainage was the most common operation, but it frequently resulted in recurrent symptoms requiring reoperation. Cyst excision was associated with a significantly lower incidence of recurrent cholangitis and need for reoperation and was not associated with increased operative mortality. Cystic disease was frequently associated with other hepatobiliary diseases. Biliary carcinoma occurred in five (16%) of our patients, and late deaths from biliary-related disease occurred in seven patients (22%). When technically possible, cyst excision is the treatment of choice.

Laparoscopic bile duct injuries. Risk factors, recognition, and repair.

Records of 11 patients undergoing biliary reconstruction after laparoscopic cholecystectomy are reviewed. Ductal injuries resulted from failure to define the anatomy of Calot's triangle. Risk factors include scarring, acute cholecystitis, and obesity. Presenting findings included anorexia, ileus, failure to thrive, pain, ascites, and jaundice. All patients required hepaticojejunostomies, which were multiple and above the hepatic bifurcation in four patients. Given the extensive nature of these injuries and the frequent need for intrahepatic anastomosis and early stenosis of repairs by referring physicians, we recommend reconstruction be undertaken by an experienced hepatobiliary surgeon. To avoid injuries, a greater appreciation of risk factors and anatomic distortion and variance and strict adherence to principles of dissection and identification of anatomic structures are suggested. The use of cholangiography and a low threshold for conversion to the open procedure are advised.

Mediastinal goiters. The need for an aggressive approach.

We reviewed the cases of 52 patients with substernal goiters to examine clinical presentation, workup, technique of removal, malignancy, and outcome. Half of the patients were asymptomatic; half had at least one compressive symptom. Chest film was the most used; computed tomography or magnetic resonance imaging was by far the most useful study. Thyroid scans often failed to show the intrathoracic goiter. Fine-needle aspiration was not helpful because of the gland's inaccessibility. Seventeen percent (nine) of the thyroids showed malignancy, 21% (11) including incidental papillary carcinomas. These were not identified by duration of goiter, symptoms, or fine-needle aspiration. Except for lymphomas, prognosis was good after resection. Removal was almost always accomplished via cervical incision, with low morbidity and no deaths. The threat of compression, the substantial chance of malignancy, and the safety of resection mean that the presence of substernal goiter is an indication for surgery.

Primary adenocarcinoma of third and fourth portions of duodenum. Favorable prognosis after resection.

Duodenal adenocarcinoma, a rare malignant lesion, is associated with a poor 5-year survival. Few series have addressed differences between resectable tumors of the proximal and distal duodenum. We reviewed records of 17 consecutive patients with adenocarcinoma of the duodenum who underwent resection: 10 had adenocarcinoma of the proximal duodenum, and seven had tumors of the distal duodenum. Most patients underwent pancreatoduodenectomy. Five patients with adenocarcinoma of the distal duodenum underwent segmental resection. No perioperative deaths occurred. Six of 10 patients with proximal tumors died of metastatic disease. Of the seven patients with tumors of the distal duodenum, five are alive without evidence of disease, and two died of unrelated causes. The survival of patients with adenocarcinoma of the distal duodenum is surprisingly good, and segmental resection is the procedure of choice.

Surgical treatment of biliary cystadenoma. A report of 15 cases.

Mucinous biliary cystadenomas are rare intrahepatic or, less commonly, extrahepatic neoplasms that may produce massive enlargement, hemorrhage, rupture, secondary infection, jaundice, or vena caval obstruction. Radiologic criteria differentiate biliary cystadenomas from more common parasitic or simple cysts. Treatment has included sclerosis, marsupialization, internal drainage, or resection, but without resection the patient is at risk for enlargement, infection, or progression of an unrecognized malignant neoplasm. We report the course of 15 patients who underwent resection for biliary cystadenoma to elucidate the clinical presentation, preoperative evaluation, and surgical treatment. Nine patients had had previous radiologic or surgical intervention other than excision, and complications of sepsis and tumor recurrence had developed. Following complete resection, however, only five postoperative complications were encountered, and no patient experienced recurrence of tumor. Thus, we recommend complete surgical resection as the preferred therapy.

Surgically incurable well-differentiated thyroid carcinoma. Prognostic factors and results of therapy.

A series of 97 consecutive patients with well-differentiated thyroid carcinoma treated between 1941 and 1970 presented with distant metastatic disease or extensive nonresectable local neck disease or had residual carcinoma after thyroid resection. Men 40 years of age or younger and women 50 years of age or younger were considered at low risk for dying of disease; older patients were considered at high risk for dying of disease. Of 17 patients with distant metastatic carcinoma, 40% of younger patients in the low-risk group and 92% of older patients in the high-risk group died. Of 80 patients with unresectable or residual local neck cancer, only 13% of younger patients but 71% of older patients died. Survival related better to risk group classification as defined by age and sex than to any details of disease presentation or management. Treatment was far more successful in patients in the low-risk group.

Unilateral hepatic duct obstruction.

Experience with unilateral hepatic duct obstruction is limited. We reviewed 33 cases of unilateral hepatic duct obstruction from a total of 500 patients with biliary reconstruction treated between 1965 and 1984. The median age of the patients reviewed was 56 years. The most common cause of unilateral hepatic duct obstruction was operative injury (73 percent of patients). Including operations for unilateral obstruction, patients in the series underwent 131 operations related to biliary tract problems. The most common presentation (73 percent of patients) was fever and pain. Obstruction was more common in the right duct than in the left duct by a ratio of 2:1 (22 patients versus 11 patients). The three types of surgical procedures used were hepaticojejunostomy (17 patients), dilatation and drainage (13 patients), and primary hepatic resection (3 patients). Atrophic hepatic lobes resulting in rotational deformity of the portal structures were resected in six patients with combined hepatic duct and arterial injury. No operative deaths occurred, although 51.5 percent of the patients had postoperative complications. Follow-up studies ranging from 1 to 16 years demonstrated that patients who had hepaticojejunostomy required less frequent reoperation compared with those who had dilatation (36 percent versus 64 percent) and had a lower postoperative mortality rate related to biliary tract problems (7 percent versus 18 percent). We conclude that unilateral hepatic duct obstruction continues to occur most commonly because of operative injury and is best treated by hepaticojejunostomy or by resection of chronically obstructed lobes when possible.

Parietal cell vagotomy for intractable and obstructing duodenal ulcer.

Parietal cell vagotomy can be accomplished with minimal morbidity and mortality. Symptoms and signs of delayed gastric emptying early after operation are common and occur more frequently in patients with preoperative gastric outlet obstruction than in those without, a difference that is statistically significant. These symptoms are generally mild and transient. Dumping and diarrhea were not problems in our series. In patients with preoperative gastric outlet obstruction, parietal cell vagotomy with pyloroduodenal dilatation achieved good or excellent results in 79 percent of patients; however, the possibility of a higher recurrence rate requires further evaluation and suggests caution and selectivity in the use of this procedure. The recurrence rate of 3 percent of these patients without gastric outlet obstruction and a very good or excellent clinical result in 91 percent of these patients appear acceptable and encourage us to continue to use parietal cell vagotomy as the procedure of choice in patients with intractable duodenal ulcer. Most patients with recurrent ulcer have been treated medically with success. Close long-term clinical follow-up studies will be required to assess better the success of this procedure.