Radiographic appearances of esophageal stents

Among patients with esophageal carcinoma and associated dysphagia, more than 60% have unresectable disease at presentation. In such cases, the goal of treatment is primarily palliation of the dysphagia; treatment options include surgery, radiation therapy, laser ablation, and placement of stents. Beginning with the first stent-made of boxwood and silver in 1885-evolution in design led to the creation of large-diameter, rigid plastic stents placed at laparotomy and eventually placed by means of endoscopy. However, complications such as perforation, hemorrhage, dislodgment, pressure necrosis, and occlusion were frequently encountered with these stents. The development of small-diameter, expandable metal stents eliminated some of these complications. Metal stents have greatly reduced procedure-related morbidity and mortality, but complications such as perforation, malposition, migration, tumor ingrowth and overgrowth, food obstruction, and tracheoesophageal fistula persist. An efficacious and increasingly used method of treating malignant dysphagia, esophageal stent placement must undergo further improvements to reduce the frequency of complications, particularly migration and tumor ingrowth.

Complications of abdominal right-thoracic (Ivor Lewis) esophagectomy

Esophageal resection and reconstruction are associated with significant operative and postoperative morbidity and mortality. Careful evaluation of the patient's cardiopulmonary status; proper preparation of the patient with smoking cessation, exercise, and cardiopulmonary rehabilitation; assessment of the stage of disease; selection of a suitable operative technique; and meticulous attention to technical details help reduce the incidence of complications and ensure a successful outcome in this technically challenging procedure.

Herniation of the transverse colon after esophagectomy: is retrocardiac air a normal postoperative finding?

OBJECTIVE

This study describes the radiographic and CT imaging features of colonic herniation complicating esophagogastrectomy in three patients, all of whom required surgery.

CONCLUSION

After routine esophagogastrectomy with gastric pull-through for esophageal carcinoma, the intrathoracic gastric pull-through most often lies in a right paraspinal location. Once a left-sided gastric pull-through has been excluded, radiologists must be suspicious for left retrocardiac air lucency that possibly represents herniation of the nearby colon. Early diagnosis is important for the prevention of bowel strangulation, especially of the ischemia-susceptible splenic flexure.

Effects of diltiazem versus digoxin on dysrhythmias and cardiac function after pneumonectomy

BACKGROUND

This prospective study was designed to determine whether diltiazem is superior to digoxin for the prophylaxis of supraventricular dysrhythmias (SVD) after pneumonectomy or extrapleural pneumonectomy (EPP) and to assess the influence of these drugs on perioperative cardiac function.

METHODS

Seventy consecutive patients without previous SVD were randomly allocated immediately after pneumonectomy or EPP to receive diltiazem (n = 35) or digoxin (n = 35). Diltiazem-treated patients received a slow intravenous loading dose of 20 mg, followed by 10 mg intravenously every 4 hours for 24 to 36 hours, then 180 to 240 mg orally daily for 1 month. Digoxin-treated patients received a 1-mg intravenous loading in the first 24 to 36 hours, then 0.125 to 0.25 mg orally daily for 1 month. A concurrent prospective cohort of 40 patients without previous SVD, who did not participate in the study and underwent pneumonectomy or EPP without prophylaxis, served as a comparison group for SVD occurrence. Serial Doppler echocardiograms were performed to assess cardiac function and all patients were continuously monitored with Holter recorders for 3 days. Data were analyzed by intent-to-treat.

RESULTS

In patients undergoing standard or intrapericardial pneumonectomy, diltiazem prevented the overall incidence of postoperative SVD when compared with digoxin, 0 of 21 patients versus 8 of 25 patients, respectively, p < 0.005. When EPP patients were included in the analysis, diltiazem decreased the incidence of all SVD from 11 of 35 patients (31%) to 5 of 35 patients (14%) when compared with digoxin, p = 0.09. Digoxin-treated patients had a similar incidence of all SVD (31%) as concurrent controls (11 of 40 patients [28%]). The two treated groups did not differ in right or left atrial size, left ventricular ejection fraction, or right heart pressure. When all patients were combined, those in whom SVD developed were significantly older (65 +/- 12 years versus 55 +/- 11 years, p = 0.004) and had a longer median hospital stay (9 versus 6 days, p = 0.03), when compared with those in whom SVD did not develop, respectively. The subset of patients undergoing EPP had a greater incidence of atrial fibrillation and electrocardiographic changes suggestive of postoperative pericarditis than all other pneumonectomy patients.

CONCLUSIONS

Diltiazem was both safe and more effective than digoxin in reducing the overall incidence of SVD after standard or intrapericardial pneumonectomy. Digoxin therapy had no effect on the incidence of postoperative SVD and is not recommended for prophylaxis of SVD. Dysrhythmias after pneumonectomy or EPP occur in older patients and are associated with a greater length of hospital stay.

Barium sulfate: a new (old) contrast agent for diagnosis of postoperative esophageal leaks

PURPOSE

To determine whether esophagography with use of barium alone is safe for detection of postoperative leaks of the esophagus.

MATERIALS AND METHODS

A review was performed of 29 postoperative esophagograms that showed 29 leaks in 12 patients. All studies were performed with a 50% dilution of barium sulfate. Leak volumes were calculated as the product (in cubic centimeters) of the length, width, and height. The safety and efficacy of barium were determined on the basis of development of mediastinitis and retention of barium in the mediastinum that would interfere with subsequent patient care.

RESULTS

In 11 of the 12 patients, follow-up studies were performed 4-48 days (mean, 10.2 days) after diagnosis. Leaks were 0.25-375 cm3 (mean, 31.4 cm3). In five of 17 (29%) follow-up procedures, small amounts of residual barium were detectable on the scout radiograph, none of which interfered with interpretation of the new images. During a follow-up period of 7-448 days (mean, 226 days), no cases of mediastinitis were found.

CONCLUSION

Esophagography can be performed safely with barium to rule out an anastomotic esophageal leak.

Role of video-assisted thoracic surgery in the treatment of pulmonary metastases: results of a prospective trial

BACKGROUND

A retrospective review revealed a 42% error rate between computed tomographic scan reports and thoracotomy findings; therefore, a prospective study was designed to compare the value of computed tomographic scans, video-assisted thoracoscopic exploration, and open thoracotomy in the management of pulmonary metastases.

METHODS

Eligibility included any patient with only one or two ipsilateral pulmonary metastases identified on computed tomographic scan who was being considered for surgical resection. Initially video-assisted thoracic surgery was performed and all lesions identified were resected. A thoracotomy adequate for complete lung palpation was then carried out and any additional lesions found were removed.

RESULTS

Eighteen patients of a planned 50 were treated before closure of the study. Four patients (22%) had no additional lesions found at thoracotomy. The primary sites of tumor were colon (10), breast (3), and one patient each skin (squamous), cervix, kidney, melanoma, and sarcoma. Four patients (22%) did have additional lesions at thoracotomy, which were benign. In the remaining 10 patients (56%) additional malignant lesions were found at thoracotomy after video-assisted thoracoscopic exploration. After 18 patients were entered, analysis of the early results disclosed a 56% failure rate of a computed tomographic scan and video-assisted thoracic surgery to detect all lesions. Being within the 95% confidence interval (32% to 78%), the study was abandoned.

CONCLUSIONS

We conclude that video-assisted thoracic surgery should be used only as a diagnostic tool in managing lung metastasis. A thoracotomy is required to achieve complete resection, which is the major survival prognosticator for satisfactory long-term results.

Symptomatic tachydysrhythmias after esophagectomy: incidence and outcome measures

BACKGROUND

Supraventricular tachydysrhythmias (SVT) after esophageal operations for carcinoma occur frequently and may be associated with increased morbidity. Prospective data on the etiology, incidence, and importance of these dysrhythmias are sparse.

METHODS

In 100 consecutive patients undergoing esophagectomy without prior history of atrial dysrhythmias or receiving antiarrhythmics, we prospectively examined the effects of predefined risk factors by history and pulmonary function on the 30-day incidence of symptomatic postoperative SVT, need for intensive care unit admission, and mortality rate.

RESULTS

Symptomatic postoperative SVT occurred in 13 (13%) of the 100 patients studied at a median of 3 days after operation and was accompanied by hypotension in 9/13 (69%). Univariate correlates of SVT were older age (p = 0.03), perioperative use of theophylline (p = 0.044), and a low carbon monoxide diffusion capacity (measured in 56% of patients) on preoperative pulmonary function. Patients in whom SVT developed had a higher rate of intensive care unit admission (p = 0.0001) and a longer hospital stay (p = 0.036). Although patients in whom SVT developed had a higher (p = 0.013) 30-day mortality rate, SVT was not the direct cause of death.

CONCLUSIONS

These prospective data show that the true incidence of symptomatic SVT within 30 days of esophagectomy is lower than previously reported. Occurrence of SVT was associated with significant morbidity. Older age was the strongest predictor of SVT after esophagectomy. In high-risk patients, continued monitoring (48 to 72 hours) and early interventions to decrease the incidence of postoperative SVT may improve overall surgical outcomes.

Erythromycin stimulates gastric emptying after esophagectomy with gastric replacement: a randomized clinical trial

Delayed gastric emptying after esophagogastrectomy can pose a significant early postoperative problem. Because erythromycin, which stimulates the gastric antral and duodenal motilin receptor, has been shown to significantly increase gastric emptying in patients with diabetic gastroparesis, we decided to evaluate its effect on gastric emptying after esophagogastrectomy.

METHODS

Twenty-four patients (18 men and six women, age range 41 to 79 years, median 66 years) were randomized to receive either erythromycin lactobionate (200 mg in 50 ml normal saline solution intravenously) (n = 13) or placebo (50 ml normal saline solution intravenously (n = 11) 11 days after esophagogastrectomy (with pyloric drainage procedure). After erythromycin or placebo had been infused over a 15-minute period, patients ingested a solid meal (scrambled egg with bread) labeled with technetium 99m sulfur colloid (500 microCi) over approximately 15 minutes. Dynamic images of the stomach were then acquired over 90 minutes in the supine position by gamma imaging. Results were expressed as percentage of counts retained in the stomach (percent gastric retention) over time.

RESULTS

There were no side effects of erythromycin. In the placebo group, the mean percent of radiolabeled meal retained in the stomach after 90 minutes was 88%, which was significantly greater than in the erythromycin group, 37% (p < 0.001). In addition, analysis of covariance demonstrated that the rate of gastric emptying (slope of the line) was significantly greater in the erythromycin-treated group than in the placebo group (p < 0.0001).

CONCLUSION

Early satiety after esophagogastrectomy may be due to delayed gastric emptying and not due to a decrease in the gastric reservoir. Intravenous erythromycin significantly improves gastric emptying in patients after esophagogastrectomy by stimulating gastric motility.

Vocal cord medialization for unilateral paralysis associated with intrathoracic malignancies

Patients with unilateral vocal cord paralysis from intrathoracic malignancies may have significant dysfunctions of speech, swallowing, ventilation, and effective coughing as a result of inadequate compensation of the nonparalyzed cord. In patients with already compromised pulmonary function, aspiration can be a life-threatening event. Sixty-three patients with intrathoracic malignancies required surgical correction of vocal cord paralysis. Primary pathology included lung cancer (49), esophageal cancer (nine), and miscellaneous tumors (five). Symptoms included hoarseness (62), dyspnea (21), aspiration (26), weight loss (19), dysphagia (14), and pneumonia (14). The surgical procedures included medial displacement of the vocal cord with silicone elastomer (48), temporary Gelfoam injection (seven), and Teflon (polytetrafluoroethylene) injection (eight) to move the affected cord to a medial position. In 11 patients, the operation was performed in the acute postoperative setting to improve pulmonary toilet. Symptomatic improvement was noted in the following proportions of affected patients: hoarseness, 92%; dyspnea, 90%; dysphagia, 93%; aspiration, 92%; pneumonia, 93%; and weight loss, 47%. Overall success rate of the intervention was 57 of 63 patients (90%). All 11 patients treated in the acute setting had immediate improvement. A variety of complications occurred in 17% of patients. Surgical management of vocal cord paralysis in patients with intrathoracic malignancies prevents life-threatening pulmonary complications in the acute postoperative setting. In chronic situations, it provides patients with improved speech, swallowing, and pulmonary function, resulting in improved quality of life, even for patients not cured of their disease.

Predictors of survival in malignant tumors of the sternum

From 1930 to 1994, 54 patients with primary malignant tumors of the sternum were seen. Fifty patients were first seen with a mass, and one half of them also had pain in the sternal region. Two patients had no symptoms at presentation. Among 39 solid tumors were 26 chondrosarcomas, 10 osteosarcomas, 1 fibrosarcoma, 1 angiosarcoma, and 1 malignant fibrous histiocytoma. Of these, 25 were low-grade and 14 were high-grade tumors. Among 15 small cell tumors were 8 plasmacytomas, 6 malignant lymphomas, and 1 Ewing's sarcoma. Partial or subtotal sternectomy was done in 37 patients and total sternectomy in 3. Of the remaining 14 patients, 3 had local excision; 10 had external radiation or chemotherapy without operation, or both; and 1 had no treatment. All but one patient treated by wide resection (N = 40) had some form of skeletal reconstruction of the chest wall defect. Thirty-one (78%) underwent repair with Marlex mesh, and in 25 this was combined with methyl methacrylate. The skin edges were closed per primum in 32 patients; 8 required muscle, omentum, or skin flaps. Resection in chondrosarcomas yielded a 5-year survival (Kaplan-Meier) of 80% (median follow-up, 17 years). The 5-year survival in osteosarcomas was 14%. Resection was curative in 64% of low-grade sarcomas but in only 7% of high-grade sarcomas. In small cell tumors, resection and radiation were helpful for local control; all failures were a result of distant metastases. We conclude that primary sarcomas of the sternum though uncommon are potentially curable by wide surgical excision. With rigid prostheses to repair the skeletal defects, the surgical complication rates are low. Overall survival after complete surgical resection is related to tumor histologic type and grade.

Impact of clinicopathologic parameters on patient survival in carcinoma of the cervical esophagus

BACKGROUND

The survival of patients with carcinoma of the cervical esophagus remains poor in spite of multimodality treatment and technical improvements in surgical resection and reconstruction. This study was undertaken to update our experience with cervical esophageal carcinoma and to identify factors that had an impact on patient survival and quality of life.

PATIENTS AND METHODS

Clinical data encompassing 132 variables were collected on 67 patients with cervical esophageal carcinoma from 1980 to 1993. Statistical analysis was performed: independent Student's t-tests, Cox regression, Kaplan-Meier curves, and log rank analyses were used in the statistical evaluation. The mean age of the patients was 63 years (range 31 to 88). Dysphagia was the primary symptom in 86% of patients; 80% had received no prior treatment. The most common abnormal finding (21%) on physical examination was a neck mass.

RESULTS

Curative resection was performed in 22 patients, 7 had palliative procedures, and 7 were found to be inoperable at exploration and received palliative treatment. Radiation with or without chemotherapy was definitive treatment for 10 patients, whereas 4 patients were treated with chemotherapy alone for cure, and 17 patients received palliative treatment. The mean survival following diagnosis was 17 months (range 1 to 96). Cumulative 5-year survival was 12%.

CONCLUSIONS

Persistent disease, chemotherapy prior to presentation, and chemotherapy for cure remained as statistically significant parameters associated with decreased survival by multivariate analysis. There was a trend toward improved survival in patients treated with surgical resection.

Thymoma: a multivariate analysis of factors predicting survival

BACKGROUND

Despite complete surgical excision, malignant thymomas often recur with resultant death. We reviewed our series to determine which factors independently predict survival after surgical resection.

METHODS

A retrospective analysis of patients operated on for thymoma between 1949 and 1993 at Memorial Sloan-Kettering Cancer Center was performed. Clinical data were collected from chart review. Only patients with a pathology report confirming the diagnosis of thymoma were included in this analysis. Kaplan-Meier survival curves were generated and comparisons of survival analyzed by log rank test. Multivariate analysis was performed by the Cox proportional hazard model.

RESULTS

One hundred eighteen patients with thymoma underwent operation. There were 86 complete resections (73%), 18 partial resections (15%), and 14 biopsies (12%). By Masaoka staging, 25 patients were stage I (21%), 41 stage II (35%), 43 stage III (36%), and 9 stage IVa (8%). Overall survival was 77% at 5 years and 55% at 10 years. Tumor recurred in 25 (29%) of 86 completely resected thymomas. Stage of disease (p = 0.03) was the only independent prognostic factor affecting recurrence. By multivariate analysis, stage (p = 0.003), tumor size (p = 0.0001), histology (p = 0.004), and extent of surgical resection (p = 0.0006) were independent predictors of long-term survival.

CONCLUSIONS

Patients with stage I disease require no further therapy after complete surgical resection. Neoadjuvant therapy should be considered for patients with large tumors and invasive disease.

Ivor Lewis esophagectomy

Lewis described a technique for resection of cancer of the midthoracic esophagus in a staged manner in 1946. The same procedure done at one stage has remained a standard technique for resection of a carcinoma involving the thoracic esophagus.

Echoplanar MR imaging for characterization of adrenal masses in patients with malignant neoplasms: preliminary evaluation of calculated T2 relaxation values

OBJECTIVE

We undertook this study to assess the utility of echoplanar MR imaging for distinguishing benign from malignant adrenal masses in patients with known malignant neoplasms.

MATERIALS AND METHODS

Thirty consecutive patients with 31 adrenal masses and a known malignant neoplasm underwent breath-hold echoplanar MR imaging with a repetition time of 6000 msec and four echo times (40, 80, 120, 160 msec) on a 1.5-T unit before biopsy. Subsequently, 10 masses were shown to be malignant at histologic examination, 12 masses were benign at histologic examination, and nine were thought to be benign because they had not changed in size at follow-up imaging. Mean lesion size was 2.4 +/- 2.1 cm. T2 calculations using regions of interest in the liver and adrenal mass were performed in each patient.

RESULTS

The mean calculated T2 value of benign adrenal masses was 70.3 msec (SD, 11.6 msec) versus 104.6 msec (SD, 35.2 msec) for malignant adrenal masses (p = .013). Using a cutoff T2 value of 84 msec, 19 (90%) of 21 benign masses and nine (90%) of 10 malignant masses were correctly classified. The mean adrenal/liver T2 ratio was 1.4 (SD, 0.25) for benign lesions, and 2.1 (SD, 0.78) for malignant lesions (p = .017). Using a cutoff ratio of 1.60, 19 (90%) of 21 benign lesions and eight (80%) of 10 malignant lesions would have been correctly classified.

CONCLUSION

This preliminary work suggests that obtaining T2 calculations from echoplanar MR images of adrenal masses is a useful technique for distinguishing benign from malignant adrenal masses in patients at risk for adrenal metastasis.

Lung resection for colorectal metastases. 10-year results

BACKGROUND

Metastasectomy for colorectal carcinoma to the lung is controversial. We analyzed results of this approach to justify its credibility.

METHODS

We studied 144 patients by retrospective review after complete resection of lung metastases from colorectal cancer from 1965 through 1988. Patient selection and prognostic factors influencing survival were analyzed. Survival was analyzed by the Kaplan-Meier method, and comparisons were made by log-rank analysis.

RESULTS

A total of 170 thoracotomies were performed in 144 patients. The overall 5- and 10-year survival was 40% and 30%, respectively. Those patients undergoing complete resection of their metastases survived significantly longer than those undergoing incomplete resections.

CONCLUSION

It appears that resection of pulmonary metastases from colorectal carcinoma translates into long-term survival benefit.

Incidence of local recurrence and second primary tumors in resected stage I lung cancer

From 1973 to 1985, 598 patients underwent resection for stage I non-small-cell lung cancer. There were 291 T1 lesions and 307 T2 lesions. The male/female ratio was 1.9:1. The histologic type was squamous carcinoma in 233 and nonsquamous carcinoma in 365. Lobectomy was performed in 511 patients (85%), pneumonectomy in 25 (4%), and wedge resection or segmentectomy in 62 (11%). A mediastinal lymph node dissection was carried out in 560 patients (94%) and no lymph node dissection in 38 (6%). Fourteen postoperative deaths occurred (2.3%). Ninety-nine percent of the patients were observed for a minimum of 5 years or until death with an overall median follow-up of 91 months. The overall 5- and 10-year survivals (Kaplan-Meier) were 75% and 67%, respectively. Survival in patients with T1 N0 tumors was 82% at 5 years and 74% at 10 years compared with 68% at 5 years and 60% at 10 years for patients with T2 tumors (p < 0.0004). The overall incidence of recurrence was 27% (local or regional 7%, systemic 20%) and was not influenced by histologic type. Second primary cancers developed in 206 patients (34%). Of these, 70 (34%) were second primary lung cancers. Despite complete resection, 31 of 62 patients (50%) who had wedge resection or segmentectomy had recurrence. Five- and 10-year survivals after wedge resection or segmentectomy were 59% and 35%, respectively, significantly less than survivals of those undergoing lobectomy (5 years, 77%; 10 years, 70%). The 5- and 10-year survivals in the 38 patients who had no lymph node dissection were reduced to 59% and 32%, respectively. Apart from the favorable prognosis observed in this group of patients, three facts emerge as significant: (1) Systematic lymph node dissection is necessary to ensure that the disease is accurately staged; (2) lesser resections (wedge/segment) result in high recurrence rates and reduced survival regardless of histologic type; and (3) second primary lung cancers are prevalent in long-term survivors.

Management of non-small cell lung cancer with direct mediastinal involvement

The results of surgical treatment were analyzed for 102 patients with non-small cell lung cancer invading the mediastinum by direct extension (T3 and T4), but those who had N2 disease were excluded to eliminate the adverse prognostic effect of this nodal subset. The histologic type was squamous cell carcinoma in 55 patients, adenocarcinoma in 40, and large cell carcinoma in 7. There were 58 T3 tumors invading the mediastinal pleura or fat, phrenic nerve, vagus nerve, pericardium, or pulmonary vessels and 44 T4 lesions invading the aorta, vena cava, esophagus, trachea, spine, or atrium. Resection included lobectomy (33 patients), pneumonectomy (32 patients), and limited resection (6 patients). Complete resection was possible in 46 patients and incomplete or no resection was possible in 56. The interstitial implantation of radioactive sources to control residual tumor also was undertaken in 43 patients. The operative mortality was 6%. The overall survival (Kaplan-Meier) was 19% at 5 years (median survival time, 18 months). Factors found to be significantly affect survival were complete resectability and the histologic type. With complete resection, the 5-year survival was 30% (p = 0.005). The 5-year survival in patients with adenocarcinoma or large-cell carcinoma was 30%, compared with 14% in patients with squamous cell carcinoma (p = 0.002). The extent of mediastinal involvement (T3 versus T4) influenced resectability and survival, and this approached statistical significance (p = 0.055). We conclude that most patients with non-small cell carcinoma and mediastinal invasion do poorly with primary surgical treatment.

Esophageal involvement in Wegener's granulomatosis

Wegener's granulomatosis (WG) is characterized by granulomatous vasculitis, renal disease, and upper and lower respiratory tract disease. Although most organ systems can be involved, gastrointestinal (GI) manifestations are notably uncommon. We describe a patient with WG whose presentation was unique for the prominence of odynophagia. Esophagoscopy revealed erosive esophagitis, which on biopsy was shown to be due to direct involvement by the underlying vasculitis. This is first antermortem documentation of esophageal disease secondary to WG. The GI manifestations of WG are reviewed.

Randomized trial comparing postoperative chemotherapy with vindesine and cisplatin plus thoracic irradiation with irradiation alone in stage III (N2) non-small cell lung cancer

This prospective randomized trial was performed to determine whether postoperative chemotherapy with vindesine and cisplatin could lengthen time to progression and overall survival in stage III (T1-3N2M0) non-small cell lung cancer (NSCLC) patients. Seventy-two patients were entered; 36 were randomized to receive chemotherapy. Patients were stratified by extent of resection (complete vs. incomplete) and histology (squamous vs. nonsquamous). All had surgery and mediastinal irradiation 6-7 weeks post-thoracotomy. Incompletely resected patients had intraoperative 125I and/or 192Ir implantation. Vindesine (3 mg/m2) weekly x 5, then every 2 weeks x 8, and cisplatin (120 mg/m2) days 1, 29, 71, 113 were planned for those randomized to chemotherapy. No difference in time to progression (median 9.2 months for radiation + chemotherapy vs. 9.0 months for radiation, P = 0.35) or overall survival (16.3 months for radiation + chemotherapy vs. 19.1 months for radiation, P = 0.42) was found. Postoperative vindesine and cisplatin did not prolong time to progression or survival in this population of stage III NSCLC.

The clamshell incision: an improved approach to bilateral pulmonary and mediastinal tumor

Median sternotomy has been the accepted approach for dealing with mediastinal tumors or bilateral pulmonary disease, but exposure to the lower lobes and to mediastinal tumors extensively involving a hemithorax is often limited. Based on the reported experience from double-lung transplantation, we explored the use of clamshell incisions for these difficult problems. From March 1991 to December 1993, we prospectively studied the utility of clamshell incisions in 90 patients for the following indications: bilateral pulmonary metastases (62 patients), primary lung carcinoma with mediastinal involvement (13 patients), primary tumors of the mediastinum (14 patients), and mesothelioma (1 patient). Bilateral anterior thoracotomies with a transverse sternotomy (clamshell incision) were employed in 71 patients and a unilateral anterior thoracotomy with partial or complete median sternotomy (hemiclamshell incision) was used in 19 patients. For closure, we used pericostal sutures and sternal wires, usually augmented by sternal K-wire stents or Steinmann pins to prevent sternal override. Exposure to all areas of the mediastinum, pericardium, pleura, and lung was excellent. Specifically, the clamshell incision afforded markedly better access to lower lobe disease and hemithoracic extension of mediastinal disease than that possible with median sternotomy. There were no deaths or significant morbidity, and all patients tolerated the incisions well without mechanical respiratory difficulties. There was one wound infection. There was no late sternal override and the cosmetic results were found to be excellent during a follow-up of 2 to 33 months. We conclude that clamshell incisions constitute an improved surgical approach for the management of bilateral pulmonary or combined pulmonary and mediastinal disease.

Influence of surgical resection and brachytherapy in the management of superior sulcus tumor

We analyzed the results of surgical treatment of all patients presenting with untreated superior sulcus tumors between 1974 to 1991 inclusive at our institution. Most patients received preoperative radiotherapy. We attempted to analyze the influence of surgical resection and intraoperative brachytherapy in obtaining locoregional control and disease-free survival. One hundred twenty-four patients underwent thoracotomy and 100 patients underwent resection. The overall 5-year survival rate was 26% for all patients and 30% for resected patients. Those patients receiving a complete resection achieved a 41% 5-year survival. The best single group were those patients undergoing a lobectomy (versus wedge resection) and en-bloc chest wall resection (60% 5-year survival). We were unable to demonstrate an advantage for the use of intraoperative brachytherapy in those patients with complete resection. For those patients with incomplete resection, the use of brachytherapy combined with preoperative or postoperative external radiation therapy resulted in a 9% 5-year survival. Locoregional failure was significant both in patients with complete resection and in patients with incomplete resection. Adverse prognostic factors included Horner's syndrome, N2 and N3 disease, T4 disease, and incomplete resections. In superior sulcus tumors, every attempt to completely resect the tumor by en-bloc chest wall resection combined with lobectomy and adequate nodal staging remains the surgical treatment of choice together with either preoperative, postoperative, or "sandwich" external radiation therapy.

Contribution of videothoracoscopy to the management of the cancer patient

BACKGROUND

Videothoracoscopy has rapidly become a popular procedure, but its technical feasibility has been emphasized without critically evaluating its role in the management of thoracic disease. To assess the value of videothoracoscopy in the diagnosis and staging of the cancer patient and to determine if it has added to our previous standard approach of thoracoscopy performed with a mediastinoscope without video, we established a prospective database when we initiated videothoracoscopy in January 1992.

METHODS

Patients were offered videothoracoscopy as an alternative to thoracotomy only if other standard approaches (e.g., needle biopsy, mediastinoscopy) were inadequate to diagnose or stage cancer or to restage patients after therapy. Parameters entered and analyzed in a prospective database were patient name; age; sex; past history; indications for videothoracoscopy; procedure type; surgical technique; whether conversion to thoracotomy was necessary, and if so, why; complications; and pathology. A complete case list of thoracoscopies performed in 1991 before videothoracoscopy was available provided historical comparison.

RESULTS

From January 1 to December 31, 1991, 82 patients underwent thoracoscopy using a mediastinoscope for diagnosis and therapy of pleural disease. From January 1 to July 31, 1992, 160 patients (male:female = 81:79; mean age 56 years) had videothoracoscopy; 72 of 160 patients (44%) had procedures that previously would have required thoracotomy: 69 lung wedge resections, one pericardial window, one pleurectomy, one mediastinal node sampling. No major resectional procedures (e.g., lobectomy, esophagectomy) were performed by videothoracoscopy. Twenty-two percent of all patients (35 of 160), and 23% of wedge resection patients (16 of 69) required conversion to thoracotomy because videothoracoscopy was inadequate for diagnosis or staging. Reasons for conversion (multiple reasons in five patients) included further resection required in 23 patients; inability to evaluate lesion in 11; adhesions in five; and inability to tolerate one lung ventilation in two. The chest tube was in place postoperatively for a mean of 2.3 days. Thirty-day postoperative complications included ventilation for > 48 h in one patient; prolonged air leak in one; pneumonia in one; arrhythmia in one; and death from progressive disease in two.

CONCLUSIONS

Although the role of videothoracoscopy in the treatment of primary thoracic malignancies and pulmonary metastases is still undefined, this early experience indicates that videothoracoscopy often enhances the ability to diagnose and stage patients by obviating thoracotomy.

Prospective evaluation of unilateral adrenal masses in patients with operable non-small-cell lung cancer. Impact of magnetic resonance imaging

We designed a prospective study to evaluate the accuracy of magnetic resonance imaging in distinguishing a benign from a malignant adrenal mass in patients with otherwise operable non-small-cell lung cancer.

METHODS

Potentially operable non-small-cell lung cancer was prospectively staged. If a unilateral adrenal mass was found by computed tomographic scanning, respiratory compensated and cardiac gated thin section magnetic resonance imaging of the adrenal glands was done. One radiologist interpreted the magnetic resonance imaging scan blinded and, on the basis of the relative signal strengths of the T1- and T2-weighted images, judged whether the adrenal mass was benign or malignant. The patients then underwent a percutaneous needle biopsy of the adrenal mass, if technically feasible. If the result of the needle biopsy was nondiagnostic or if the biopsy was not feasible, an adrenalectomy through a posterior approach was performed.

RESULTS

Twenty-seven patients with a unilateral adrenal mass entered the study-11 men and 16 women whose ages ranged from 42 to 75 years (median 58 years). Four patients had epidermoid and 23 adenocarcinoma of the lung. The clinical locoregional stage was I in 9, II in 1, IIIA in 16, and IIIB in 1. Twenty-five completed the magnetic resonance imaging procedure. Five adrenal masses (19%) were metastatic non-small-cell lung cancer (adenocarcinoma = 4, epidermoid = 1); 22 masses (81%) were benign (adenoma = 20, hyperplasia = 2). There were no significant differences in age, sex, histologic type, or locoregional stage between those with a benign versus a malignant mass. However, the malignant masses were significantly larger (3.8 +/- 1.9 cm; range 2.5 to 7.1; median 3.1) than the benign masses (2.0 +/- 0.4 cm, range 1.2 to 2.8; median 2.0) (p < 0.001). Among those having magnetic resonance imaging (n = 25), the technique correctly predicted a malignant mass in the four patients with a histologically confirmed metastasis from non-small-cell lung cancer. However, among the 21 histologically benign masses, the magnetic resonance imaging was interpreted as benign in 5, malignant in 14, and indeterminate in 2. Therefore, although the false-negative rate was 0%, the false-positive rate was 67%.

CONCLUSION

Most adrenal masses in patients with otherwise operable non-small-cell lung cancer are benign. Currently available magnetic resonance imaging methods cannot replace biopsy.

Treatment and prognosis in bronchial carcinoids involving regional lymph nodes

From 1953 to 1992, 25 patients were surgically treated for bronchial carcinoids with metastases to regional lymph nodes (N1 or N2). The tumors were located centrally, involving main or lobar bronchi in 12 patients and were peripheral in 13. Histologically, 12 of the carcinoids were classified as typical and 13 as atypical (neuroendocrine carcinoma). Pneumonectomy was performed in 11 patients, sleeve lobectomy in one, lobectomy in seven and bilobectomy in six. A formal mediastinal lymph node dissection was done in 20 patients. At final staging, 10 had N1 disease and 15 had N2. No adjuvant treatment was given to the 10 patients with N1 disease. External radiation therapy was given after the operation to 9 of 15 patients with N2 disease. The overall 5-year survival (Kaplan-Meier) was 75% (median 62 months). No difference in survival was found between patients with N1 or N2 disease. However, survival and recurrence rate differed between typical and atypical carcinoids. In typical carcinoids, the 5-year survival was 92% and, in atypical carcinoids, it was 60% (p = 0.02). We conclude that complete resection for bronchial carcinoids results in long-term survival despite the presence of regional lymph node metastases. Recurrence appears to depend more on cell type than nodal status. Postoperative radiation therapy does not appear to be beneficial.

Accuracy of lung imaging in metastases with implications for the role of thoracoscopy

Thoracoscopy for wedge resection of lung metastases is rapidly increasing in frequency. This technique precludes bimanual palpation of the lung to locate additional lesions not seen on the surface. Finger palpation is inadequate. Implications regarding the failure to identify all metastases and the negative impact on long-term survival led us to review retrospectively the correlation between pathologic findings and imaging reports. One hundred forty-four patients who had resection of lung metastases from colorectal cancer were studied. All had chest roentgenograms and 72 had computed tomographic scans as well. Chest roentgenogram and computed tomographic reports differed in the number of nodules reported in 17 of 72 patients (24%). In 3 of 17 patients chest roentgenogram showed more nodules than computed tomography. Chest roentgenogram differed from pathologic findings at surgery in 57 of 144 patients (39%). Twenty-six of 57 patients (46%) had more lesions than chest roentgenogram detected and 31 had fewer. Computed tomographic scans differed from pathologic findings in 30 of 72 patients (42%). If one or two lesions were imaged, 12 patients had fewer cancers (some lesions were benign) and 18 had more cancers than computed tomography reported; computed tomographic scans erred 28% of the time. The inability to adequately palpate the entire lung using the thoracoscope alone markedly impairs the surgeon's ability to know if a resection of all lesions has been done. The validity of using thoracoscopy resection in the management of metastatic disease is seriously questioned other than for diagnosis.