[Show abstract][Hide abstract] ABSTRACT: Current chemotherapeutic regimens have only modest benefit for non-small cell lung cancer (NSCLC) patients. Cumulative toxicities/drug resistance limit chemotherapy given after the first-line regimen. For personalized chemotherapy, clinically relevant NSCLC models are needed for quickly predicting the most effective regimens for therapy with curative intent. In this study, first generation subrenal capsule xenografts of primary NSCLCs were examined for (a) determining responses to conventional chemotherapeutic regimens and (b) selecting regimens most effective for individual patients.
Pieces (1x3x3 mm(3)) of 32 nontreated, completely resected patients' NSCLCs were grafted under renal capsules of nonobese diabetic/severe combined immunodeficient mice and treated with (A) cisplatin+vinorelbine, (B) cisplatin+docetaxel, (C) cisplatin+gemcitabine, and positive responses (treated tumor area <or=50% of control, P < 0.05) were determined. Clinical outcomes of treated patients were acquired.
Xenografts from all NSCLCs were established (engraftment rate, 90%) with the retention of major biological characteristics of the original cancers. The entire process of drug assessment took 8 weeks. Response rates to regimens A, B, and C were 28% (9 of 32), 42% (8 of 19), and 44% (7 of 16), respectively. Certain cancers that were resistant to a particular regimen were sensitive to others. The majority of responsive tumors contained foci of nonresponding cancer cells, indicative of tumor heterogeneity and potential drug resistance. Xenografts from six of seven patients who developed recurrence/metastasis were nonresponsive.
Models based on first generation NSCLC subrenal capsule xenografts have been developed, which are suitable for quick assessment (6-8 weeks) of the chemosensitivity of patients' cancers and selection of the most effective regimens. They hold promise for application in personalized chemotherapy of NSCLC patients.
Clinical Cancer Research 02/2010; 16(5):1442-51. DOI:10.1158/1078-0432.CCR-09-2878 · 8.72 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Cervical esophagogastric anastomoses are commonly used for reconstruction after esophagectomy because of the lower mortality rate associated with an anastomotic leak compared to intrathoracic anastomoses. However, cervical esophagogastric anastomoses have been criticized for their higher leak rates, stricture formation and greater need for later dilatations when compared with intrathoracic anastomoses. Multiple studies have looked at varying techniques to improve the outcome of the cervical esophagogastric anastomosis. This study was performed to determine whether a partially stapled (posterior stapled wall and anterior hand-sewn wall) anastomosis reliably reduced leaks and the need for later dilatation. From January 2001 to March 2006, 168 patients who underwent cervical esophagogastric anastomosis following esophagectomy (transhiatal or three-hole) for cancer were identified. Beginning in September 2003, the partially stapled technique was introduced and used in 79 patients. Clinical outcomes were compared to patients in whom hand-sewn technique was used (n = 89). Outcomes related to anastomotic leak, other hospital complications, length of stay, postoperative dilatations and survival were compared using Student's t-tests and chi-square tests (P < 0.05), as well as multiple regression analyses. An anastomotic leak occurred in 10 (12.7%) patients who received a partially stapled anastomosis. A hand-sewn anastomosis was complicated by an anastomotic leak in 24 patients (27.0%). This difference was statistically significant (P = 0.021). This lowered incidence of leak was associated with an earlier initiation of oral feeds (median 7 vs. 9.5 days, P < 0.001) and a reduction in hospital stay (median 10 vs. 15 days, P < 0.001). Furthermore, dysphagia associated with stricture requiring postoperative dilatations was markedly diminished in the stapled anastomosis [23 (31.3%) vs. 49 (55.1%), P = 0.001]. The partially stapled cervical esophagogastric anastomosis significantly decreased the incidence of postoperative anastomotic leaks and the need for postoperative dilatation to treat strictures compared to the hand-sewn anastomosis.
Diseases of the Esophagus 08/2008; 21(5):422-9. DOI:10.1111/j.1442-2050.2007.00792.x · 1.78 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Disruptions of beta-catenin and the canonical Wnt pathway are well documented in cancer. However, little is known of the non-canonical branch of the Wnt pathway. In this study, we investigate the transcript level patterns of genes in the Wnt pathway in squamous cell lung cancer using reverse-transcriptase (RT)-PCR. It was found that over half of the samples examined exhibited dysregulated gene expression of multiple components of the non-canonical branch of the WNT pathway. In the cases where beta catenin (CTNNB1) was not over-expressed, we identified strong relationships of expression between wingless-type MMTV integration site family member 5A (WNT5A)/ frizzled homolog 2 (FZD2), frizzled homolog 3 (FZD3) / dishevelled 2 (DVL2), and low density lipoprotein receptor-related protein 5 (LRP5)/ secreted frizzled-related protein 4 (SFRP4). This is one of the first studies to demonstrate expression of genes in the non-canonical pathway in normal lung tissue and its disruption in lung squamous cell carcinoma. These findings suggest that the non-canonical pathway may have a more prominent role in lung cancer than previously reported.
[Show abstract][Hide abstract] ABSTRACT: The purpose of this study was to assess the validity of using a pulsatile, pressure waveform transduced from the epidural space through an epidural needle or catheter to confirm correct placement for maximal analgesia and to compare 3 different types of catheters' ability to transduce a waveform.
A single-center, prospective, randomized trial.
A tertiary-referral hospital.
Eighty-one patients undergoing posterolateral thoracotomy who required a thoracic epidural catheter for postoperative pain management.
Each epidural needle and each epidural catheter was transduced to determine if there was a pulsatile waveform exhibited.
Sensitivity of the pulsatile waveform transduced through an epidural needle to identify correct placement of the epidural needle and the sensitivity of each catheter type to identify placement were compared. In 79 of 81 cases (97.5%), the waveform transduced directly through the epidural needle had a pulsatile characteristic as determined by blinded observers. In a total of 53 of 81 epidural catheters (65.4%), the transduced waveform displayed pulsations. Twenty-four of 27 catheters in group S-P/Sims Portex (Smiths Medical MD, Inc, St Paul, MN) (88.9%) transduced a pulsatile tracing from the epidural space, a significantly greater percentage than in the other 2 groups (p = 0.02).
The technique of transducing the pressure waveform from the epidural needle inserted in the epidural space is a sensitive and reliable alternative to other techniques for confirmation of correct epidural catheter placement. The technique is simple, sensitive, and inexpensive and uses equipment available in any operating room.
Journal of Cardiothoracic and Vascular Anesthesia 11/2006; 20(5):659-63. DOI:10.1053/j.jvca.2006.02.022 · 1.46 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The purpose of this study was to describe the nature of diaphragm injury, to quantify the injury and number of macrophages at the light microscopic level, and to determine their association with airflow obstruction in humans. Partial-thickness diaphragm biopsies were obtained from 21 subjects going for thoracotomy surgery (FEV(1): 74 +/- 34% predicted; range: 16 to 122% predicted). Cross sections cut from frozen diaphragm were processed with H&E or processed for immunohistochemistry using the monoclonal antibody Ber-MAC3 (DAKO Corp., Carpinteria, CA) to label macrophages. Area fractions (A(A)) or the proportions of the cross- sectional area were determined by point counting all viable fields of H&E-stained diaphragm cross sections. A(A) were 66.2 +/- 9.0% for normal muscle, 17.6 +/- 7.2% for abnormal muscle, and 16.3 +/- 4.2% for connective tissue. Percent predicted FEV(1) was inversely related to the A(A) of abnormal muscle (r = -0.53, p < 0.01) and directly related to the A(A) of normal muscle (r = 0.37, p < 0.05). The number of macrophages was not related to % predicted FEV(1) (mean +/- SD: 0.41 +/- 0.18/fiber; 52 +/- 19/mm(2)). We conclude that increasing severity of airflow obstruction is associated with an increased A(A) of abnormal diaphragm and a decreased A(A) of normal diaphragm.
American Journal of Respiratory and Critical Care Medicine 06/2001; 163(7):1654-9. DOI:10.1164/ajrccm.163.7.2001042 · 13.00 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Background. The optimal approach to the investigation of possible distant metastases in patients with apparently operable non-small cell lung cancer who do not have symptoms suggesting metastatic disease is controversial. Methods. We conducted a randomized, controlled trial in thoracic surgery services at mainly academic tertiary- and secondary-care general hospitals. We recruited 634 patients with apparently operable, suspected or proven non-small cell carcinoma of the lung without findings on history, physical examination, laboratory testing, or imaging suggesting extrathoracic metastases. Patients were randomly allocated to receive either mediastinoscopy and computed tomography of the chest and then, depending on the results, immediate thoracotomy or bone scintigraphy and computed tomographic scanning of the head, liver, and adrenal glands. Results. The relative risk of thoracotomy without cure (the combination of open and closed thoracotomy, incomplete resection, and thoracotomy with subsequent recurrence) in the full investigation group versus the limited investigation group was 0.80 (95% confidence interval [CI], 0.56 to 1.13; p = 0.20). Forty-three patients in the full investigation group and 61 patients in the limited investigation group underwent a thoracotomy but subsequently had recurrence (relative risk, 0.70; 95% CI, 0.47 to 1.03; p = 0.07). Patients in the full investigation group were more likely to have avoided thoracotomy because of extrathoracic metastatic disease than those in the limited investigation group (22 patients versus 10 patients, respectively; relative risk, 2.19; 95% CI, 1.04 to 4.59; p value = 0.04). The total number of negative invasive tests was six in the full investigation group and one in the limited investigation group (relative risk, 6.1; 95% CI, 0.72 to 51.0; p = 0.10) and the total number of invasive tests, 11 versus six, respectively (relative risk, 1.84; 95% CI, 0.68 to 4.98; p = 0.23). The full investigation strategy cost $823 less per patient (95% CIs 2,482 to -725). Conclusions. Full investigation for metastatic disease in patients with non-small cell lung cancer without symptoms or signs of metastatic disease may reduce the number of thoracotomies without cure. The higher the threshold for considering symptoms to suggest metastatic disease, the more likely it is that investigation will spare patients futile thoracotomy. (C) 2001 by The Society of Thoracic Surgeons.
The Annals of Thoracic Surgery 02/2001; 71(2):425-433. DOI:10.1016/S0003-4975(00)02359-6 · 3.85 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: We conducted a prospective study to evaluate whether lack of an adequate increase in diffusing capacity for carbon monoxide (DL(CO)) during exercise is associated with a greater postoperative complication rate after lung resection. We used the three-equation method (3EQ-DL(CO)), a modification of the single breath DL(CO) technique to determine DL(CO) during exercise in 57 patients undergoing lung resection at Vancouver General Hospital from October 1998 to May 1999. 3EQ-DL(CO) was determined during steady-state exercise at 35% and 70% of the maximal workload reached in a progressive exercise test. Maximal oxygen uptake (VO(2)max), DL(CO) at rest, and the increase in DL(CO) during exercise were compared in relation to postoperative complications. Patients with complications had lower resting values of DL(CO) (R-DL(CO)), a smaller increase in DL(CO) from rest to 70% of maximal workload expressed as a percent of the predicted DL(CO) at rest ([70% - R]-DL(CO)%), and a lower VO(2)max than did patients without complications. Results suggested that (70% - R)-DL(CO)% was the best preoperative predictor of postoperative complications; a cutoff limit of 10% was the best index to identify complications, yielding a complication rate of 100% in patients with (70% - R)-DL(CO)% < 10% as compared with a complication rate of 10% in patients with (70% - R)-DL(CO)% >/= 10% (sensitivity = 78%, specificity = 100%). Patients who do not increase their DL(CO) sufficiently during exercise ([70% - R]-DL(CO)% < 10%) have a greater complication rate after lung resection.
American Journal of Respiratory and Critical Care Medicine 11/2000; 162(4 Pt 1):1435-44. DOI:10.1164/ajrccm.162.4.2001117 · 13.00 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Bronchioloalveolar cell adenomas (BAAs) have been described in up to 10% of patients with bronchogenic carcinoma. Their prognostic significance is unknown. The purpose of this study was to determine the prognostic implications of finding BAAs coexisting in specimens resected for primary bronchogenic carcinoma and to determine how frequently BAAs can be detected radiologically.
Follow-up information for a mean of 30 months was obtained on 28 patients with a single primary bronchogenic carcinoma and one or more coexistent BAAs. Preoperative chest radiographs (n=27) and CT scans (n=24) were retrospectively reviewed to assess the number of patients in whom BAAs could be detected radiologically.
There was no significant difference between percentage survival of patients with a primary bronchogenic carcinoma and coexistent BAAs when compared with the percentage predicted survival of these patients based on their primary bronchogenic carcinoma alone. BAAs could be detected retrospectively in 1 of 27 (4%) preoperative radiographs and 11 of 24 (46%) CT scans.
On standard preoperative imaging for bronchogenic carcinoma, BAAs were retrospectively detected in more than one third of patients in whom they were detected pathologically. However, the presence of coexistent BAAs with bronchogenic carcinoma does not affect short- and medium-term prognosis.
[Show abstract][Hide abstract] ABSTRACT: The frequency and causes of gastrointestinal complications following esophagectomy for malignancy are unknown.
We reviewed 295 esophagectomies performed for malignancy between January 1980 and September 1994 in order to determine the frequency and causes of early and late gastrointestinal complications.
Compared to transhiatal and left thoracoabdominal esophagectomies, esophagectomies carried out through a right posterolateral thoracotomy with cervical esophagogastric anastomosis had a higher incidence of delayed gastric emptying (11%), pneumonia (26%), and hospital death (9%). The same operation had a higher incidence of gastroesophageal reflux (20%) and dysphagia requiring esophageal dilatation (53%). We found no independent effect of gastric drainage procedures, feeding jejunostomy, preoperative radiotherapy, pathology, or age on these outcomes. Women had no operative mortality, but a higher incidence of gastroesophageal reflux and diarrhea following esophagectomy.
Surgical techniques aimed at improving gastric emptying following esophagectomy for cancer should improve operative morbidity and mortality.
The American Journal of Surgery 06/1995; 169(5):471-5. DOI:10.1016/S0002-9610(99)80197-4 · 2.29 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: We reviewed our experience from 1979 to 1990 with 160 cases of transhiatal esophagectomy for carcinoma of the lower esophagus and cardia to evaluate trends in patient selection, management, and outcome. Patients treated in the past 6 years (n = 110) and those treated before 1985 (n = 50) were similar in terms of age and sex distribution, medical history, and weight loss. The majority of tumors seen were adenocarcinoma, with patients in the latter group having significantly lower stages. Significant decreases in anesthetic time, units of blood transfusions, chest tube insertions, length of postoperative ventilation, incidence of postoperative pneumonia, and length of hospital stay were seen during the past 6 years. Wound infections increased significantly during the same period. The decrease in the 30-day mortality rate from 6% to 0.9% was not significant. Survival rates did not differ between groups, with overall rates of 62%, 40%, and 21% at 1, 2, and 5 years, respectively.
Archives of Surgery 11/1992; 127(10):1164-7; discussion 1167-8. · 4.93 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: A variant left hepatic artery occurs at a rate of approximately 10%. In standard esophagogastrectomy and some proximal gastric operations this variant artery is sacrificed, which has led to reported fatalities secondary to hepatic necrosis. We report our method of esophagogastrectomy in the presence of an aberrant left hepatic artery.
The Annals of Thoracic Surgery 08/1992; 54(1):166-8. DOI:10.1016/0003-4975(92)91173-7 · 3.85 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Following esophagectomy, restoration of swallowing by gastric tube interposition with cervical esophagogastric anastomosis reduces morbidity and mortality associated with intrathoracic anastomoses at the expense of an increased incidence of both anastomotic leak and stricture formation. A retrospective study of 165 patients with either squamous cell carcinoma or adenocarcinoma of the distal esophagus or gastric cardia undergoing esophagogastrectomy with gastric tube interposition and cervical anastomosis at Vancouver, British Columbia, or London, Ontario, was undertaken. Forced-entry multiple logistic regression analysis of factors believed to influence anastomotic outcome was performed. Anastomotic leak occurred in 17% of patients; statistically significant correlation with low preoperative serum albumin (p = 0.005), running suture technique (p = 0.029), high intraoperative blood loss (p = 0.038), and the occurrence of postoperative delayed gastric emptying (p = 0.045) was found. Anastomotic strictures occurred in 31% of patients; a statistically significant correlation was found with preceding anastomotic leak (p = 0.001) and intraoperative blood loss (p = 0.042). Factors including preoperative radiotherapy and diabetes mellitus were not found to be significant.
The American Journal of Surgery 06/1992; 163(5):484-9. DOI:10.1016/0002-9610(92)90393-6 · 2.29 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Thirty-four patients undergoing thoracotomy were entered into a randomized, double-blind, placebo-controlled study to compare the effects of patient-controlled, lumbar epidural (PCA-E) fentanyl with patient-controlled intravenous (PCA-i.v.) fentanyl with respect to drug requirements, analgesic efficacy and respiratory function. Prior to chest closure patients received fentanyl 2 micrograms.kg-1 by the epidural or i.v. route. In the recovery room further doses of epidural or i.v. fentanyl, 50 micrograms, were administered by the patients who controlled two PCA pumps. Background fentanyl infusion rates were increased by 10 micrograms.hr-1 each time the patient administered a drug bolus and were decreased by 10 micrograms.hr-1 whenever visual analogue scale (VAS) pain scores were less than 2 on a maximum 10 scale. Twenty-nine patients completed the study. Patients in the PCA-E group (n = 14) required less total fentanyl than those in the PCA-i.v. (n = 15) group (1857 +/- 693 micrograms vs 2573 +/- 890 micrograms respectively, P less than 0.05). Fentanyl infusion rates were lower in the PCA-E group at most measurement times. There were no differences between groups in respiratory rates, PaCO2, VAS pain scores or changes in pulmonary function as measured by FVC and FEV1. It is concluded that satisfactory patient-controlled analgesia can be achieved with both epidural and i.v. fentanyl after thoracotomy but that fentanyl requirements are less when given via the epidural route. This supports a direct spinal cord site of action for lumbar epidural fentanyl.
Canadian Journal of Anaesthesia 04/1992; 39(3):214-9. DOI:10.1007/BF03008779 · 2.53 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Between August 1984 and October 1988, 7 women and 16 men underwent chest-wall resection. The 23 patients ranged in age from 17 to 79 years. Resection was done for benign lesions in 9 patients, for recurrent chest-wall sarcoma in 4 and for carcinoma involving the chest wall in 10. The number of ribs resected ranged from none to six. Prosthetic material was required for reconstruction in eight patients. There were no operative deaths and no flail segments developed postoperatively. Three patients have since died of metastatic disease, one has died of unrelated causes but with no residual disease and the remainder were alive and well at follow-up intervals ranging from 11 to 60 months. Aggressive resection, including a wide margin of healthy tissue, provides the best chance for recurrence-free survival for patients with many types of chest-wall tumour. Resection can be performed with low morbidity and satisfactory cosmetic results.
Canadian journal of surgery. Journal canadien de chirurgie 07/1990; 33(3):229-32. · 1.51 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The effectiveness of fibrin glue as a sealant to reduce postoperative air leaks after pulmonary lobectomy was evaluated in 28 consecutive patients between November 1988 and May 1989. A fibrin glue spray was used in 14 patients, and 14 patients served as controls. Assignment of either group was made before thoracotomy. Nine male and 5 female patients with a mean age of 63.8 years were in the fibrin glue experimental group, and 8 male and 6 female patients with a mean age of 59 years, in the control group. An equal number of complete and incomplete fissures were in each group. All fissures were handled in the same way (stapled). Two milliliters of fibrin glue was applied through a double-syringe delivery system and sprayed on the staple line and any cut surface of the inflated lung just before thoracotomy closure. The fibrin glue-treated group had a mean air leak duration of 2.3 +/- 3.7 days, chest tube drains for 6 +/- 4.1 days, and a postoperative hospitalization of 9.8 +/- 3.1 days. The control group had a mean air leak duration of 3.3 +/- 3.3 days (p = 0.94), chest tube drains for 5.9 +/- 3.9 days (p = 0.95), and a postoperative hospitalization of 11.5 +/- 3.9 days (p = 0.21). We conclude that the routine use of a fixed quantity of fibrin glue is not effective in reducing the duration of air leaks, chest tube drainage, or hospitalization after uncomplicated pulmonary lobectomy.
The Annals of Thoracic Surgery 02/1990; 49(1):133-4. DOI:10.1016/0003-4975(90)90371-C · 3.85 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: We determined single breath diffusing capacity (DLCO) and pulmonary capillary blood volume (Vc) in a total of 110 patients, who were being evaluated for resectional lung surgery for a localized tumor or lesion. Pathologic assessment of emphysema was obtained in 55 cases who had resection of a lung or an upper lobe, based on a standard reference panel for emphysema grading. In 86 cases, the extent of emphysema was quantitated by computed tomography (CT) of the chest. There was a significant negative correlation between Vc and emphysema assessed by either pathology or CT (r = about -0.5, p less than 0.001) similar to the correlation between DLCO and the extent of emphysema. Results of Vc were significantly lower in cases with moderate emphysema (pathologic grade greater than or equal to 30) than those with no emphysema (grade less than or equal to 5) (p less than 0.001) or mild emphysema (grade 10 to 25) (p less than 0.05), and they were significantly lower (p less than 0.05) in the group with mild emphysema compared with the group with no emphysema on pathologic assessment similar to DLCO results. Although Vc was reduced in emphysema, determination of Vc did not result in improved discrimination in separating cases with emphysema from those without emphysema when compared with DLCO.
The American review of respiratory disease 02/1990; 141(1):53-61. DOI:10.1164/ajrccm/141.1.53 · 10.19 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: To evaluate the sensitivity of diffusing capacity (DLCO) and pressure-volume (P-V) curves in the detection of emphysema, these tests were compared with pathologic assessment of emphysema in patients undergoing lung resection for a localized tumor, and with the overall extent of emphysema as assessed by computed tomography (CT). The resected lung specimens were fixed in the inflated state and cut at 1-cm intervals in the horizontal plane. The pathologic extent of emphysema was quantitated by comparison with a standard reference panel of emphysema grading. The overall extent of emphysema on CT was assessed by a visual scoring system in a total of 55 patients, 19 undergoing lung resection and 36 not undergoing lung resection. Analysis of 37 patients by pathology scores revealed 18 with no or trivial emphysema (emphysema grades less than or equal to 5; mean grade, 2.2 +/- SD 2.6) and 19 with emphysema (grades greater than or equal to 10; mean grade, 33.2 +/- SD 24.2). Diffusing capacity, the ratio of DLCO to alveolar volume (DLCO/VA), maximal lung elastic recoil (PLmax), and lung elastic recoil at 90% of total lung capacity (PL90) were significantly different between the two groups, whereas K (the exponential constant describing the shape of the P-V curve) was not. The pathology grade of emphysema showed a significant correlation with (DLCO) (r = -0.53) and DLCO/VA (r = -0.55), which was greater than the correlation with PLmax (r = -0.42) and PL90 (r = -0.43).(ABSTRACT TRUNCATED AT 250 WORDS)
The American review of respiratory disease 06/1989; 139(5):1179-87. DOI:10.1164/ajrccm/139.5.1179 · 10.19 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: A review of 33 months' experience with primary anterior mediastinal masses in 31 patients disclosed that in 9 (29%) the tumour was benign. Nineteen (86%) of the 22 malignant tumours and only 2 of the 9 benign tumours were symptomatic. Diagnosis was established by histopathologic examination of a biopsy specimen or resected tissue and serum radioimmunoassay for alpha-fetoprotein or human chorionic gonadotropin, beta-subunit. The authors present an investigational algorithm, using modern procedures, to facilitate the diagnosis of this relatively uncommon and challenging condition.
Canadian journal of surgery. Journal canadien de chirurgie 04/1989; 32(2):139-42. · 1.51 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Neuroendocrine carcinomas of the lung are characterized by differentiation toward Kulchitsky cells and are classified as Kulchitsky-cell carcinoma (KCC) I (classic carcinoid), KCC II (atypical carcinoid), and KCC III (small-cell carcinoma). The clinical, computed tomographic (CT), and pathologic findings in 31 patients with KCC were reviewed. KCC I lesions generally occurred in younger (56 years +/- 18) nonsmoking women, were small (1.8 cm +/- 0.7 in diameter on CT scans), and were associated with lymphadenopathy in one of ten patients. KCC II tumors were found predominantly in older (66 years +/- 12) smoking men and were larger (3.9 cm +/- 1.3 in diameter, P less than .001); four of ten patients had CT evidence of lymphadenopathy. KCC III tumors occurred in older (66 years +/- 8) smoking men and were large (4.2 cm +/- 1.0); 11 of 11 patients had massive lymphadenopathy. Clinical, radiologic, or pathologic overlap was noted in three patients. Sputum cytologic and fine-needle and bronchoscopic biopsy findings were often nondiagnostic or misleading, particularly for KCC II lesions. CT of the chest provides additional discriminating information in the preoperative diagnosis of neuroendocrine lung carcinomas.