- Email: [email protected]
Evaluation of minimally invasive surgical staging for esophageal cancer
The American Journal of Surgery 182 (2001) 702–706
Evaluation of minimally invasive surgical staging for esophageal cancer Ninh T. Nguyen, M.D.a,*, Peter F. Roberts, M.D.a, David M. Follette, M.D.a, Derek Lau, M.D.b, John Lee, M.D.b, Shiro Urayama, M.D.b, Bruce M. Wolfe, M.D.a, James E. Goodnight, Jr., M.D., Ph.D.a a
Department of Surgery, University of California, Davis, Medical Center, 2221 Stockton Blvd., 3rd Flr., Sacramento, CA 95817-2214, USA b Department of Internal Medicine, University of California, Davis, Medical Center, Sacramento, CA, USA Manuscript received July 31, 2001; revised manuscript September 2, 2001 Presented at the 53rd Annual Meeting of the Southwestern Surgical Congress, Cancun, Mexico, April 29 –May 2, 2001.
Abstract Background: Conventional imaging studies (computed tomography and endoscopic esophageal ultrasonography) used for preoperative evaluation of patients with esophageal cancer can be inaccurate for detection of small metastatic deposits. We evaluated the efficacy of minimally invasive surgical (MIS) staging as an additional modality for evaluation of patients with esophageal cancer. Methods: Between December 1998 and February 2001, 33 patients with esophageal cancer were evaluated for surgical resection. Conventional imaging studies demonstrated operable disease in 31 patients and equivocal findings in 2 patients. All patients then underwent MIS staging (laparoscopy, bronchoscopy, and ultrasonography of the liver). We compared the results from surgical resection and MIS staging with those from conventional imaging. Results: MIS staging altered the treatment plan in 12 (36%) of 33 patients; MIS staging upstaged 10 patients with operable disease and downstaged 2 patients with equivocal findings. MIS staging accurately determined resectability in 97% of patients compared with 61% of patients staged by conventional imaging. The specificity and negative predictive value for detection of unsuspected metastatic disease in MIS staging were 100% and 96%, respectively, compared with 91% and 65%, respectively, for conventional imaging studies. Conclusion: In addition to conventional imaging studies, MIS staging should be included routinely in the preoperative work-up of patients with esophageal cancer. © 2002 Excerpta Medica, Inc. All rights reserved. Keywords: Esophageal cancer; Laparoscopic ultrasonography; Thoracoscopic staging; Laparoscopic staging
The incidence of esophageal cancer is increasing in the United States; the estimated number of new patients diagnosed during 2000 was 12,300 and the projected number of new cases in 2001 is expected to be 13,200 [1]. Currently, surgical resection is the treatment of choice for patients presenting with stage I, II, or III disease. Patients with distant metastatic disease, local invasion of adjacent structures by the primary tumor (T4 disease), or advanced localregional nodal involvement of the celiac axis have stage IV disease. The primary aim in any preoperative evaluation of patients with esophageal cancer is to identify patients with stage IV cancer who would not benefit from surgical resection. Patients with stage IV disease have a limited life
* Corresponding author. Tel.: ⫹1-916-734-4596; fax: 916-734-3951. E-mail address: [email protected]
expectancy and are best treated palliatively with an esophageal stent, endoscopic laser ablation, or photodynamic therapy [2]. Selection of patients for resection has been based primarily upon conventional imaging modalities such as computed tomography (CT) of the chest and abdomen and endoscopic esophageal ultrasonography (EUS). However, CT and EUS have been shown to be inaccurate for identifying small sites of metastatic disease in patients with esophageal cancer. EUS is sensitive for detecting the depth of tumor invasion, but it is a poor indicator of distant metastatic disease [3]. CT scan of the chest and abdomen evaluates patients for suspected distant metastatic disease and locally advanced disease, but CT is inaccurate for evaluation of small metastatic deposits and is also a poor indicator of advanced local-regional nodal involvement [3]. In this study, we evaluated minimally invasive surgical (MIS) staging as an additional, more accurate staging modality for
0002-9610/01/$ – see front matter © 2002 Excerpta Medica, Inc. All rights reserved. PII: S 0 0 0 2 - 9 6 1 0 ( 0 1 ) 0 0 8 0 4 - 2
N.T. Nguyen et al. / The American Journal of Surgery 182 (2001) 702–706
patients shown to have resectable esophageal cancer by preoperative conventional imaging studies. Patients and methods Between December 1998 and February 2001, MIS staging was performed in 33 patients with esophageal carcinoma. All patients had a preoperative CT scan of the chest and abdomen. EUS was also performed on 27 of the 33 patients. All CT scans and EUS were reviewed by a radiologist and staged accordingly. MIS staging consisted of laparoscopic staging, bronchoscopy, esophagoscopy, and laparoscopic ultrasonography of the liver. Fiberoptic bronchoscopy was performed to determine possible tumor infiltration of the tracheobronchial tree in patients with proximal and middle-third esophageal cancer. Flexible esophagoscopy was performed to evaluate the proximal and distal extent of tumor involvement. MIS staging was performed before the surgical resection procedure to evaluate patients for enrollment into a neoadjuvant chemotherapy protocol. Minimally invasive surgical staging technique Laparoscopic staging was performed using five abdominal trocars (four 5-mm and one 12-mm). All quadrants of the abdominal cavity as well as the liver surface were systematically inspected for any suspicious peritoneal metastatic deposits, and a biopsy was performed on all suspicious lesions. Intraoperative ultrasonography of the liver was performed using a flexible tip 7.5-MHz linear array laparoscopic ultrasonography probe (Aloka Co., Wallingford, Connecticut). A core biopsy was performed for any suspicious intraparenchymal liver lesions. The celiac axis region was evaluated by dividing the lesser omentum to expose the left gastric vessels. Findings of matted, firm, enlarged, or fixed lymphadenopathy in the celiac axis region was considered as locally advanced unresectable disease and biopsy was obtained for confirmation when possible. Lymphadenopathy at the celiac axis region without fixation was considered resectable disease and routine lymph node biopsy was not performed in these patients. Thoracoscopic staging was performed using two 10-mm and two 5-mm trocars in the single patient who had a suspicious left upper lobe lesion identified on chest CT. The upper lobe lesion corresponding with the abnormality seen on CT of the chest was identified and a wedge resection was performed using multiple applications of the endoscopic stapler (ENDO GIA; United States Surgical Corp., Norwalk, Connecticut). Statistical analysis Accuracy, specificity, and negative predictive value for detection of patients with unsuspected stage IV disease were calculated using the following formulas based on the findings of MIS staging and surgical exploration. Accuracy was
703
calculated by the sum of the number of true-positive cases plus the number of true-negative cases and divided by the number of all cases. Specificity was calculated by the number of true-negative cases detected by MIS divided by the number of true-negative cases plus the number of false-positive cases. Negative predictive value was calculated by the number of true-negative cases detected by MIS divided by the sum of the true-negative cases plus the false-negative cases. Results Thirty-three patients (24 females) with a mean age of 64 years (range 34 to 83) underwent MIS staging. Tumor location was in the distal esophagus in 26 patients, mid esophagus in 6 patients, and proximal esophagus in 1 patient. Tumor histology consisted of adenocarcinoma in 24 patients and squamous cell carcinoma in 9 patients. CT scan of the chest and abdomen identified resectable disease in 31 patients and equivocal findings in 2 patients (a possible single focus of liver metastasis and a possible single focus of left upper lobe pulmonary metastasis). EUS was performed on 27 of 33 patients. Three (11%) of the 27 patients had an incomplete EUS examination due to severe luminal obstruction that prohibited the passage of the endoscope. EUS demonstrated resectable T3 or less disease in the 24 patients who had a complete EUS examination. Findings at MIS staging MIS staging was technically feasible in all patients. The additional findings from MIS staging altered the treatment plan in 12 (36%) of 33 patients (Fig. 1). Ten patients were found to have unresectable disease on MIS staging. Laparoscopic staging demonstrated unresectable disease in 8 patients; 4 patients had tumor encasement of celiac vessels, 2 patients had liver surface metastases, 1 patient had peritoneal surface metastases, and 1 patient had unsuspected macronodular cirrhosis. Laparoscopic ultrasonography examination of the liver was also performed on all patients. Intraoperative ultrasonography of the liver did not demonstrate any additional findings of intraparanchymal occult liver metastases. Bronchoscopic staging identified 2 patients with tracheal infiltration of the tumor. Both of these patients had severe luminal obstruction at presentation and EUS examination had not been completed. In the 2 patients with equivocal findings on CT imaging, laparoscopic ultrasonography of the liver confirmed the suspected metastatic liver lesion to be a liver cyst and thoracoscopic staging was used to perform a wedge resection of the suspected metastatic pulmonary lesion that confirmed a pseudotumor with central necrosis. Operative results Mean operative time was 88 ⫾ 36 minutes (range 60 to 140). Conversion to open laparotomy occurred in 1 (3%) of
704
N.T. Nguyen et al. / The American Journal of Surgery 182 (2001) 702–706 Table 1 Comparison of minimally invasive surgical staging and conventional imaging staging for detection of unresectable metastatic disease in patients with esophageal cancer Type of staging
Accuracy Specificity Negative (%) (%) predictive value (%)
Minimally invasive surgical 97 staging* Conventional imaging staging† 61
100
96
91
65
* Minimally invasive surgical staging includes laparoscopy, bronchoscopy, and laparoscopic ultrasound of the liver. † Conventional imaging staging includes computed tomography and endoscopic esophageal ultrasound.
modalities. The specificity and negative predictive value for MIS staging were 100% and 96%, respectively, compared with 91% and 65% for conventional imaging (Table 1). Comments
Fig. 1. Outcomes of 33 patients with esophageal cancer who were evaluated by conventional imaging and minimally invasive surgical (MIS) staging.
33 patients due to a bladder perforation. The perforation occurred during lysis of adhesion in a patient who had extensive lower abdominal adhesions from previous pelvic surgery. There were no major postoperative complications. Minor postoperative complications included one port site wound infection. Mean hospital stay was 0.9 ⫾ 0.5 days. At a mean follow-up of 11 months (range 1 to 23), there was no port site tumor recurrence. Treatment plans after MIS staging Ten of the 23 patients with potentially resectable disease confirmed by MIS staging underwent neoadjuvant radiation and chemotherapy. The mean time interval between the staging procedure and surgical resection in these 10 patients was 83 ⫾ 12 days. Of the 23 patients who underwent surgical resection, 1 patient had unresectable disease missed by MIS staging. This patient was found to have unresectable mediastinal pleural involvement. Eight of the 10 patients with unresectable disease identified by MIS staging received definitive radiation and chemotherapy treatment. None of the 10 patients underwent palliative resection. Two patients underwent photodynamic therapy for palliation of dysphagia symptoms. Accuracy of MIS staging MIS staging accurately determined surgical resectability in 97% of patients compared with 61% for conventional imaging
Patients with esophageal cancer frequently have advanced disease at initial presentation. Accurate preoperative staging can prevent unnecessary surgical resection for patients with stage IV disease. We found that the accuracy of MIS staging for detecting occult metastases was 97% compared with 61% for conventional imaging. MIS staging achieved a higher specificity for detection of occult metastatic disease than conventional imaging studies. The 2 patients with equivocal findings suggestive of metastatic disease reflected the lower specificity of CT imaging. MIS staging accurately identified these equivocal findings as benign disease and both patients subsequently underwent successful surgical resection. Luketich et al [4] reported similar findings; six of eight equivocal CT findings in their study proved to be benign disease when assessed by MIS staging. In our study, MIS staging achieved a higher negative predictive value than conventional imaging (96% versus 65%, respectively). The lower negative predictive value of conventional imaging studies reflects its high false-negative findings. In our study, findings at MIS staging avoided unnecessary laparotomy in 10 (30%) of 33 patients. Eight (80%) of these 10 patients proceeded to have definitive radiation and chemotherapy. Velanovich et al [5] reported that patients with unresectable disease discovered at laparoscopic staging are more likely to receive postoperative chemotherapy than patients with unresectable malignancies who underwent exploratory laparotomy. Laparoscopic staging has been shown by several investigators to be superior to CT imaging for detection of intraabdominal occult metastases [4,6 – 8]. Laparoscopic staging detects unsuspected metastatic disease in 7% to 38% of patients with operable esophageal cancer [4,6 – 8]. Laparoscopic staging can identify unsuspected small liver surface
N.T. Nguyen et al. / The American Journal of Surgery 182 (2001) 702–706
metastases, peritoneal metastases, local-regional advanced lymphadenopathy of the celiac axis, and tumor invasion of surrounding structures. In our study, laparoscopic staging spared 8 patients from unnecessary laparotomies: 2 patients had small superficial liver metastases that were missed by conventional imaging studies, 4 patients had locally advanced nodal disease at the celiac axis region, 1 patient had multiple small peritoneal surface metastases, and 1 patient had unsuspected liver cirrhosis. These 8 patients were treated nonoperatively. Smith et al [8] reported similar results. In their study, 18 (19%) of 93 patients with esophageal cancer had occult metastases discovered during laparoscopy. Molloy et al [7] also demonstrated that laparoscopic staging for esophageal cancer detected inoperable disease in 38% of their patients. However, only 72% of patients in Molloy’s study who underwent exploratory laparotomy after laparoscopy had a successful resection. The reason for failure of laparoscopy to detect unresectable disease in their study included fixed lymphadenopathy at the celiac axis region in 7 patients. We routinely perform a complete examination of the celiac axis region by opening the gastrohepatic omentum to examine the lesser sac for fixed, matted, and enlarged lymphadenopathy. Occasionally it is necessary to mobilize the greater curvature of the gastric fundus to inspect for involvement of the posterior aspect of the gastric cardia and diaphragmatic hiatus. In our study, 22 (96%) of 23 patients underwent successful resection after the MIS staging evaluation. The only false-negative finding of MIS staging occurred in one patient with metastatic mediastinal pleural disease, which could have been detected preoperatively only by thoracoscopic staging. Bronchoscopic staging is an important tool for evaluation of patients with a proximal or mid-esophageal tumor for detection of possible posterior membranous involvement of the trachea or main stem bronchus. Slight compression of the tracheobronchial tree at bronchoscopic examination was not a criterion for unresectability. Baisi et al [9] reported that all patients in their study with a finding of only slight compression on the posterior wall of the trachea and normal trachea mucosa had successful surgical resections. In our study, we identified carcinoma involvement of the posterior membranous portion of the trachea in 2 patients: 1 patient had carcinoma of the proximal third esophagus and the other had carcinoma of the middle third esophagus. EUS certainly would have detected T4 disease in these patients; however, due to severe luminal obstruction, the results of the EUS examination were inconclusive in both of these patients. Intraoperative ultrasonography of the liver did not demonstrate any additional liver metastases in our study; both patients with unsuspected liver metastases had surface lesions that were easily detected by laparoscopic inspection alone. Our findings concurred with results from Luketich et al [4] who found that intraoperative ultrasonography of the liver did not increase the detection of liver metastases. They no longer routinely perform intraoperative liver ultrasonography. Velasco et al [10] demonstrated that intraoperative
705
ultrasonography of the liver altered the management of patients with pancreatic and hepatobiliary tumors but did not change the management of patients with gastric or esophageal cancer. Conversely, Hunerbein et al [11] reported that intraoperative ultrasonography of the liver improved the staging in 29% of their patients, but their study consisted primarily of patients with gastric cancer. Smith et al [8] also reported that intraoperative ultrasonography of the liver in addition to laparoscopy identified unresectable disease in 4 of 93 esophageal cancer patients; in all 4 patients, enlarged celiac and para-aortic lymphadenopathies were the reason for unresectability. We believe that these enlarged lymphadenopathies would have been detected without intraoperative ultrasonography of the liver if a complete surgical evaluation of the celiac axis had been performed. In our study, MIS staging failed to identify 1 patient with unsuspected mediastinal pleural metastatic disease, which would have been detected if routine thoracoscopic staging had been performed. However, the benefit of routine thoracoscopic staging for esophageal cancer remains controversial. Thoracoscopic staging has been advocated as a tool for sampling of periesophageal lymph nodes and to allocate patients to appropriate neoadjuvant therapy protocols or as a means to avoid unnecessary thoracotomy in patients with unsuspected intrathoracic metastatic disease. Krasna et al [12] reported 46 patients who were thoracoscopically staged and none of the 46 patients had unresectable disease identified by thoracoscopy. Thoracoscopy was accurate only for detection of lymph node involvement. Luketich et al [4] also reported a combined laparoscopic and thoracoscopic staging technique with lymph node sampling in 53 patients. Thoracoscopic staging identified positive thoracic nodes in 36% of their study patients, but none of the 53 patients had locally advanced unresectable intrathoracic disease. In addition, the disadvantages of thoracoscopic staging include longer operative time (average length was more than 3 hours in the Luketich study [4]), the need for single lung ventilation, necessity for hospitalization from placement of a chest tube, and the potentially increased difficulty from adhesions at the time of definitive esophageal resection. In summary, MIS staging is a highly accurate modality for detection of unsuspected metastases in patients with esophageal cancer. MIS staging altered the treatment plans of more than one third of patients after conventional staging. MIS staging prevented unnecessary laparotomies in patients with occult metastases and accurately downstaged patients with equivocal findings on conventional imaging. Therefore, MIS staging should be a part of the preoperative evaluation of patients with potentially resectable esophageal cancer as well as CT and EUS. References [1] Greenlee RT, Hill-Harmon MB, Murray T, Thun M. Cancer statistics, 2001. CA Cancer J Clin 2001;51:15–36.
706
N.T. Nguyen et al. / The American Journal of Surgery 182 (2001) 702–706
[2] Nguyen NT, Luketich JD. Palliation of dysphagia in advanced esophageal cancer. In: Carrau RL, Murray T, editors. Comprehensive management of swallowing disorders. San Diego: CV Singular, 1998, p 377– 81. [3] Stell DA, Carter CR, Stewart I, Anderson JR. Prospective comparison of laparoscopy, ultrasonography and computed tomography in the staging of gastric cancer. Br J Surg 1996;83:1260 –2. [4] Luketich JD, Meehan M, Nguyen NT, et al. Minimally invasive surgical staging for esophageal cancer. Surg Endosc 2000;14:700 –2. [5] Velanovich V, Wollner I, Ajlouni M. Staging laparoscopy promotes increased utilization of postoperative therapy for unresectable intraabdominal malignancies. J Gastrointest Surg 2000;4:542– 6. [6] Heath EI, Kaufman HS, Talamini MA, et al. The role of laparoscopy in preoperative staging of esophageal cancer. Surg Endosc 2000;14: 495–9. [7] Molloy RG, McCourtney JS, Anderson JR. Laparoscopy in the management of patients with cancer of the gastric cardia and oesophagus. Br J Surg 1995;82:352– 4. [8] Smith A, John TG, Garden OJ, Brown SP. Role of laparoscopic ultrasonography in the management of patients with oesophagogastric cancer. Br J Surg 1999;86:1083–7. [9] Baisi A, Bonavina L, Peracchia A. Bronchoscopic staging of squamous cell carcinoma of the upper thoracic esophagus. Arch Surg 1999;134:140 –3. [10] Velasco JM, Rossi H, Hieken TJ, Fernandez M. Laparoscopic ultrasound enhances diagnostic laparoscopy in the staging of intra-abdominal neoplasms. Am Surg 2000;66:407–11. [11] Hunerbein M, Rau B, Hohenberger P, Schlag PM. The role of staging laparoscopy for multimodal therapy of gastrointestinal cancer. Surg Endosc 1998;12:921–5. [12] Krasna MJ. Minimally invasive staging for esophageal cancer. Chest 1997;112(suppl 4):191– 4.
Discussion Dr. Patrick J. Offner (Denver, CO): Dr. Nguyen and colleagues present us with 33 patients with esophageal cancer who underwent minimally invasive surgical staging prior to definitive therapy. Their results, in combination with other published reports, provide mounting evidence that minimally invasive staging techniques are accurate and may alter the treatment plan in a significant number of patients. The authors have defined minimally invasive staging as including laparoscopy, laparoscopic ultrasound, and bronchoscopy. One could question, however, the inclusion of bronchoscopy, as many of us would consider this to be part of standard preoperative evaluation for large tumors of the middle and upper esophagus. Did you perform bronchoscopy on every patient regardless of tumor location or CT scan findings and if not, how were patients selected for bronchoscopy? If you were to consider the patients staged with bronchoscopy as undergoing conventional staging procedures, would this have influenced your performance analysis at all? Four patients were noted to be unresectable owing to tumor encasement of the celiac axis on laparoscopy. It was unclear to me whether these were biopsy proved as metastatic disease; and this leads to my final and perhaps most important question: what was the gold standard to which conventional staging procedures and minimally invasive staging procedures were compared during your analysis?
Dr. Ninh T. Nguyen (Sacramento, CA): We perform bronchoscopy routinely on all patients at the time of laparoscopic staging. It is very simple to place the bronchoscope while the patient is under general anesthesia. Therefore we perform bronchoscopy and esophagoscopy routinely, followed by laparoscopic staging. Many surgeons do not perform their own preoperative upper endoscopy. Because I approach esophagectomy using a minimally invasive method, I must know the proximal and distal extent of the tumor. Tactile feedback is lost during laparoscopic and thorascopic operations; we therefore must rely on other diagnostic tests such as bronchoscopy and esophagoscopy. In the 4 patients with tumor encasement, we identified matted or fixed lymph nodes in the region of the celiac axis. These patients underwent biopsy only if it was feasible. Patients with extensive tumor encasement of the celiac axis did not undergo biopsy. The gold standard for comparison in our study was based on findings at minimally invasive surgical staging and findings at the time of surgical resection. Dr. Wynn: The CT is always done prior to the laparoscopic exploration, and you have that knowledge when you do that laparoscopic staging. So, given that, aren’t you influencing your results? Aren’t you more likely to have better results in the laparoscopic evaluation if you already have the findings of the CT in your databank? Also, given your data, are you now suggesting you don’t need to do the CT or that this needs to be done together? Dr. Ninh T. Nguyen: No, CT scanning is important to provide an overall survey for possibly liver and pulmonary metastases. It is a good evaluator for distant metastatic disease, as well as local regional involvement of the descending thoracic aorta. Minimally invasive surgical staging should be performed in addition to but not in lieu of CT scanning. Dr. David Easter (San Diego, CA): I’m wondering why you include your jejunostomy data, as it didn’t affect the staging, and I’m wondering if you’ve overutilized it. How many actually use their jejunostomies? It seems you’re turning a clean case into a clean-contaminated case. Dr. Ninh T. Nguyen: If you look at our complication rate, there was only 1 patient who developed a wound infection. We perform laparoscopic jejunostomy because some of our patients elected to be enrolled in our neoadjuvant chemotherapy/radiotherapy protocol. This regimen lasts approximately 8 to 10 weeks. During this interval, if the patients develop dysphagia, the jejunostomy tube is used for nutritional support. Dr. David Easter: Does everybody get the jejunostomy tube? Dr. Ninh T. Nguyen: Yes, everyone gets the J-tube. Dr. David Easter: Does everybody get the new adjuvant therapy? Closing Dr. Ninh T. Nguyen: No. Not everybody gets the adjuvant chemotherapy/radiotherapy treatment.
Evaluation of minimally invasive surgical staging for esophageal cancer Ninh T. Nguyen, M.D.a,*, Peter F. Roberts, M.D.a, David M. Follette, M.D.a, Derek Lau, M.D.b, John Lee, M.D.b, Shiro Urayama, M.D.b, Bruce M. Wolfe, M.D.a, James E. Goodnight, Jr., M.D., Ph.D.a a
Department of Surgery, University of California, Davis, Medical Center, 2221 Stockton Blvd., 3rd Flr., Sacramento, CA 95817-2214, USA b Department of Internal Medicine, University of California, Davis, Medical Center, Sacramento, CA, USA Manuscript received July 31, 2001; revised manuscript September 2, 2001 Presented at the 53rd Annual Meeting of the Southwestern Surgical Congress, Cancun, Mexico, April 29 –May 2, 2001.
Abstract Background: Conventional imaging studies (computed tomography and endoscopic esophageal ultrasonography) used for preoperative evaluation of patients with esophageal cancer can be inaccurate for detection of small metastatic deposits. We evaluated the efficacy of minimally invasive surgical (MIS) staging as an additional modality for evaluation of patients with esophageal cancer. Methods: Between December 1998 and February 2001, 33 patients with esophageal cancer were evaluated for surgical resection. Conventional imaging studies demonstrated operable disease in 31 patients and equivocal findings in 2 patients. All patients then underwent MIS staging (laparoscopy, bronchoscopy, and ultrasonography of the liver). We compared the results from surgical resection and MIS staging with those from conventional imaging. Results: MIS staging altered the treatment plan in 12 (36%) of 33 patients; MIS staging upstaged 10 patients with operable disease and downstaged 2 patients with equivocal findings. MIS staging accurately determined resectability in 97% of patients compared with 61% of patients staged by conventional imaging. The specificity and negative predictive value for detection of unsuspected metastatic disease in MIS staging were 100% and 96%, respectively, compared with 91% and 65%, respectively, for conventional imaging studies. Conclusion: In addition to conventional imaging studies, MIS staging should be included routinely in the preoperative work-up of patients with esophageal cancer. © 2002 Excerpta Medica, Inc. All rights reserved. Keywords: Esophageal cancer; Laparoscopic ultrasonography; Thoracoscopic staging; Laparoscopic staging
The incidence of esophageal cancer is increasing in the United States; the estimated number of new patients diagnosed during 2000 was 12,300 and the projected number of new cases in 2001 is expected to be 13,200 [1]. Currently, surgical resection is the treatment of choice for patients presenting with stage I, II, or III disease. Patients with distant metastatic disease, local invasion of adjacent structures by the primary tumor (T4 disease), or advanced localregional nodal involvement of the celiac axis have stage IV disease. The primary aim in any preoperative evaluation of patients with esophageal cancer is to identify patients with stage IV cancer who would not benefit from surgical resection. Patients with stage IV disease have a limited life
* Corresponding author. Tel.: ⫹1-916-734-4596; fax: 916-734-3951. E-mail address: [email protected]
expectancy and are best treated palliatively with an esophageal stent, endoscopic laser ablation, or photodynamic therapy [2]. Selection of patients for resection has been based primarily upon conventional imaging modalities such as computed tomography (CT) of the chest and abdomen and endoscopic esophageal ultrasonography (EUS). However, CT and EUS have been shown to be inaccurate for identifying small sites of metastatic disease in patients with esophageal cancer. EUS is sensitive for detecting the depth of tumor invasion, but it is a poor indicator of distant metastatic disease [3]. CT scan of the chest and abdomen evaluates patients for suspected distant metastatic disease and locally advanced disease, but CT is inaccurate for evaluation of small metastatic deposits and is also a poor indicator of advanced local-regional nodal involvement [3]. In this study, we evaluated minimally invasive surgical (MIS) staging as an additional, more accurate staging modality for
0002-9610/01/$ – see front matter © 2002 Excerpta Medica, Inc. All rights reserved. PII: S 0 0 0 2 - 9 6 1 0 ( 0 1 ) 0 0 8 0 4 - 2
N.T. Nguyen et al. / The American Journal of Surgery 182 (2001) 702–706
patients shown to have resectable esophageal cancer by preoperative conventional imaging studies. Patients and methods Between December 1998 and February 2001, MIS staging was performed in 33 patients with esophageal carcinoma. All patients had a preoperative CT scan of the chest and abdomen. EUS was also performed on 27 of the 33 patients. All CT scans and EUS were reviewed by a radiologist and staged accordingly. MIS staging consisted of laparoscopic staging, bronchoscopy, esophagoscopy, and laparoscopic ultrasonography of the liver. Fiberoptic bronchoscopy was performed to determine possible tumor infiltration of the tracheobronchial tree in patients with proximal and middle-third esophageal cancer. Flexible esophagoscopy was performed to evaluate the proximal and distal extent of tumor involvement. MIS staging was performed before the surgical resection procedure to evaluate patients for enrollment into a neoadjuvant chemotherapy protocol. Minimally invasive surgical staging technique Laparoscopic staging was performed using five abdominal trocars (four 5-mm and one 12-mm). All quadrants of the abdominal cavity as well as the liver surface were systematically inspected for any suspicious peritoneal metastatic deposits, and a biopsy was performed on all suspicious lesions. Intraoperative ultrasonography of the liver was performed using a flexible tip 7.5-MHz linear array laparoscopic ultrasonography probe (Aloka Co., Wallingford, Connecticut). A core biopsy was performed for any suspicious intraparenchymal liver lesions. The celiac axis region was evaluated by dividing the lesser omentum to expose the left gastric vessels. Findings of matted, firm, enlarged, or fixed lymphadenopathy in the celiac axis region was considered as locally advanced unresectable disease and biopsy was obtained for confirmation when possible. Lymphadenopathy at the celiac axis region without fixation was considered resectable disease and routine lymph node biopsy was not performed in these patients. Thoracoscopic staging was performed using two 10-mm and two 5-mm trocars in the single patient who had a suspicious left upper lobe lesion identified on chest CT. The upper lobe lesion corresponding with the abnormality seen on CT of the chest was identified and a wedge resection was performed using multiple applications of the endoscopic stapler (ENDO GIA; United States Surgical Corp., Norwalk, Connecticut). Statistical analysis Accuracy, specificity, and negative predictive value for detection of patients with unsuspected stage IV disease were calculated using the following formulas based on the findings of MIS staging and surgical exploration. Accuracy was
703
calculated by the sum of the number of true-positive cases plus the number of true-negative cases and divided by the number of all cases. Specificity was calculated by the number of true-negative cases detected by MIS divided by the number of true-negative cases plus the number of false-positive cases. Negative predictive value was calculated by the number of true-negative cases detected by MIS divided by the sum of the true-negative cases plus the false-negative cases. Results Thirty-three patients (24 females) with a mean age of 64 years (range 34 to 83) underwent MIS staging. Tumor location was in the distal esophagus in 26 patients, mid esophagus in 6 patients, and proximal esophagus in 1 patient. Tumor histology consisted of adenocarcinoma in 24 patients and squamous cell carcinoma in 9 patients. CT scan of the chest and abdomen identified resectable disease in 31 patients and equivocal findings in 2 patients (a possible single focus of liver metastasis and a possible single focus of left upper lobe pulmonary metastasis). EUS was performed on 27 of 33 patients. Three (11%) of the 27 patients had an incomplete EUS examination due to severe luminal obstruction that prohibited the passage of the endoscope. EUS demonstrated resectable T3 or less disease in the 24 patients who had a complete EUS examination. Findings at MIS staging MIS staging was technically feasible in all patients. The additional findings from MIS staging altered the treatment plan in 12 (36%) of 33 patients (Fig. 1). Ten patients were found to have unresectable disease on MIS staging. Laparoscopic staging demonstrated unresectable disease in 8 patients; 4 patients had tumor encasement of celiac vessels, 2 patients had liver surface metastases, 1 patient had peritoneal surface metastases, and 1 patient had unsuspected macronodular cirrhosis. Laparoscopic ultrasonography examination of the liver was also performed on all patients. Intraoperative ultrasonography of the liver did not demonstrate any additional findings of intraparanchymal occult liver metastases. Bronchoscopic staging identified 2 patients with tracheal infiltration of the tumor. Both of these patients had severe luminal obstruction at presentation and EUS examination had not been completed. In the 2 patients with equivocal findings on CT imaging, laparoscopic ultrasonography of the liver confirmed the suspected metastatic liver lesion to be a liver cyst and thoracoscopic staging was used to perform a wedge resection of the suspected metastatic pulmonary lesion that confirmed a pseudotumor with central necrosis. Operative results Mean operative time was 88 ⫾ 36 minutes (range 60 to 140). Conversion to open laparotomy occurred in 1 (3%) of
704
N.T. Nguyen et al. / The American Journal of Surgery 182 (2001) 702–706 Table 1 Comparison of minimally invasive surgical staging and conventional imaging staging for detection of unresectable metastatic disease in patients with esophageal cancer Type of staging
Accuracy Specificity Negative (%) (%) predictive value (%)
Minimally invasive surgical 97 staging* Conventional imaging staging† 61
100
96
91
65
* Minimally invasive surgical staging includes laparoscopy, bronchoscopy, and laparoscopic ultrasound of the liver. † Conventional imaging staging includes computed tomography and endoscopic esophageal ultrasound.
modalities. The specificity and negative predictive value for MIS staging were 100% and 96%, respectively, compared with 91% and 65% for conventional imaging (Table 1). Comments
Fig. 1. Outcomes of 33 patients with esophageal cancer who were evaluated by conventional imaging and minimally invasive surgical (MIS) staging.
33 patients due to a bladder perforation. The perforation occurred during lysis of adhesion in a patient who had extensive lower abdominal adhesions from previous pelvic surgery. There were no major postoperative complications. Minor postoperative complications included one port site wound infection. Mean hospital stay was 0.9 ⫾ 0.5 days. At a mean follow-up of 11 months (range 1 to 23), there was no port site tumor recurrence. Treatment plans after MIS staging Ten of the 23 patients with potentially resectable disease confirmed by MIS staging underwent neoadjuvant radiation and chemotherapy. The mean time interval between the staging procedure and surgical resection in these 10 patients was 83 ⫾ 12 days. Of the 23 patients who underwent surgical resection, 1 patient had unresectable disease missed by MIS staging. This patient was found to have unresectable mediastinal pleural involvement. Eight of the 10 patients with unresectable disease identified by MIS staging received definitive radiation and chemotherapy treatment. None of the 10 patients underwent palliative resection. Two patients underwent photodynamic therapy for palliation of dysphagia symptoms. Accuracy of MIS staging MIS staging accurately determined surgical resectability in 97% of patients compared with 61% for conventional imaging
Patients with esophageal cancer frequently have advanced disease at initial presentation. Accurate preoperative staging can prevent unnecessary surgical resection for patients with stage IV disease. We found that the accuracy of MIS staging for detecting occult metastases was 97% compared with 61% for conventional imaging. MIS staging achieved a higher specificity for detection of occult metastatic disease than conventional imaging studies. The 2 patients with equivocal findings suggestive of metastatic disease reflected the lower specificity of CT imaging. MIS staging accurately identified these equivocal findings as benign disease and both patients subsequently underwent successful surgical resection. Luketich et al [4] reported similar findings; six of eight equivocal CT findings in their study proved to be benign disease when assessed by MIS staging. In our study, MIS staging achieved a higher negative predictive value than conventional imaging (96% versus 65%, respectively). The lower negative predictive value of conventional imaging studies reflects its high false-negative findings. In our study, findings at MIS staging avoided unnecessary laparotomy in 10 (30%) of 33 patients. Eight (80%) of these 10 patients proceeded to have definitive radiation and chemotherapy. Velanovich et al [5] reported that patients with unresectable disease discovered at laparoscopic staging are more likely to receive postoperative chemotherapy than patients with unresectable malignancies who underwent exploratory laparotomy. Laparoscopic staging has been shown by several investigators to be superior to CT imaging for detection of intraabdominal occult metastases [4,6 – 8]. Laparoscopic staging detects unsuspected metastatic disease in 7% to 38% of patients with operable esophageal cancer [4,6 – 8]. Laparoscopic staging can identify unsuspected small liver surface
N.T. Nguyen et al. / The American Journal of Surgery 182 (2001) 702–706
metastases, peritoneal metastases, local-regional advanced lymphadenopathy of the celiac axis, and tumor invasion of surrounding structures. In our study, laparoscopic staging spared 8 patients from unnecessary laparotomies: 2 patients had small superficial liver metastases that were missed by conventional imaging studies, 4 patients had locally advanced nodal disease at the celiac axis region, 1 patient had multiple small peritoneal surface metastases, and 1 patient had unsuspected liver cirrhosis. These 8 patients were treated nonoperatively. Smith et al [8] reported similar results. In their study, 18 (19%) of 93 patients with esophageal cancer had occult metastases discovered during laparoscopy. Molloy et al [7] also demonstrated that laparoscopic staging for esophageal cancer detected inoperable disease in 38% of their patients. However, only 72% of patients in Molloy’s study who underwent exploratory laparotomy after laparoscopy had a successful resection. The reason for failure of laparoscopy to detect unresectable disease in their study included fixed lymphadenopathy at the celiac axis region in 7 patients. We routinely perform a complete examination of the celiac axis region by opening the gastrohepatic omentum to examine the lesser sac for fixed, matted, and enlarged lymphadenopathy. Occasionally it is necessary to mobilize the greater curvature of the gastric fundus to inspect for involvement of the posterior aspect of the gastric cardia and diaphragmatic hiatus. In our study, 22 (96%) of 23 patients underwent successful resection after the MIS staging evaluation. The only false-negative finding of MIS staging occurred in one patient with metastatic mediastinal pleural disease, which could have been detected preoperatively only by thoracoscopic staging. Bronchoscopic staging is an important tool for evaluation of patients with a proximal or mid-esophageal tumor for detection of possible posterior membranous involvement of the trachea or main stem bronchus. Slight compression of the tracheobronchial tree at bronchoscopic examination was not a criterion for unresectability. Baisi et al [9] reported that all patients in their study with a finding of only slight compression on the posterior wall of the trachea and normal trachea mucosa had successful surgical resections. In our study, we identified carcinoma involvement of the posterior membranous portion of the trachea in 2 patients: 1 patient had carcinoma of the proximal third esophagus and the other had carcinoma of the middle third esophagus. EUS certainly would have detected T4 disease in these patients; however, due to severe luminal obstruction, the results of the EUS examination were inconclusive in both of these patients. Intraoperative ultrasonography of the liver did not demonstrate any additional liver metastases in our study; both patients with unsuspected liver metastases had surface lesions that were easily detected by laparoscopic inspection alone. Our findings concurred with results from Luketich et al [4] who found that intraoperative ultrasonography of the liver did not increase the detection of liver metastases. They no longer routinely perform intraoperative liver ultrasonography. Velasco et al [10] demonstrated that intraoperative
705
ultrasonography of the liver altered the management of patients with pancreatic and hepatobiliary tumors but did not change the management of patients with gastric or esophageal cancer. Conversely, Hunerbein et al [11] reported that intraoperative ultrasonography of the liver improved the staging in 29% of their patients, but their study consisted primarily of patients with gastric cancer. Smith et al [8] also reported that intraoperative ultrasonography of the liver in addition to laparoscopy identified unresectable disease in 4 of 93 esophageal cancer patients; in all 4 patients, enlarged celiac and para-aortic lymphadenopathies were the reason for unresectability. We believe that these enlarged lymphadenopathies would have been detected without intraoperative ultrasonography of the liver if a complete surgical evaluation of the celiac axis had been performed. In our study, MIS staging failed to identify 1 patient with unsuspected mediastinal pleural metastatic disease, which would have been detected if routine thoracoscopic staging had been performed. However, the benefit of routine thoracoscopic staging for esophageal cancer remains controversial. Thoracoscopic staging has been advocated as a tool for sampling of periesophageal lymph nodes and to allocate patients to appropriate neoadjuvant therapy protocols or as a means to avoid unnecessary thoracotomy in patients with unsuspected intrathoracic metastatic disease. Krasna et al [12] reported 46 patients who were thoracoscopically staged and none of the 46 patients had unresectable disease identified by thoracoscopy. Thoracoscopy was accurate only for detection of lymph node involvement. Luketich et al [4] also reported a combined laparoscopic and thoracoscopic staging technique with lymph node sampling in 53 patients. Thoracoscopic staging identified positive thoracic nodes in 36% of their study patients, but none of the 53 patients had locally advanced unresectable intrathoracic disease. In addition, the disadvantages of thoracoscopic staging include longer operative time (average length was more than 3 hours in the Luketich study [4]), the need for single lung ventilation, necessity for hospitalization from placement of a chest tube, and the potentially increased difficulty from adhesions at the time of definitive esophageal resection. In summary, MIS staging is a highly accurate modality for detection of unsuspected metastases in patients with esophageal cancer. MIS staging altered the treatment plans of more than one third of patients after conventional staging. MIS staging prevented unnecessary laparotomies in patients with occult metastases and accurately downstaged patients with equivocal findings on conventional imaging. Therefore, MIS staging should be a part of the preoperative evaluation of patients with potentially resectable esophageal cancer as well as CT and EUS. References [1] Greenlee RT, Hill-Harmon MB, Murray T, Thun M. Cancer statistics, 2001. CA Cancer J Clin 2001;51:15–36.
706
N.T. Nguyen et al. / The American Journal of Surgery 182 (2001) 702–706
[2] Nguyen NT, Luketich JD. Palliation of dysphagia in advanced esophageal cancer. In: Carrau RL, Murray T, editors. Comprehensive management of swallowing disorders. San Diego: CV Singular, 1998, p 377– 81. [3] Stell DA, Carter CR, Stewart I, Anderson JR. Prospective comparison of laparoscopy, ultrasonography and computed tomography in the staging of gastric cancer. Br J Surg 1996;83:1260 –2. [4] Luketich JD, Meehan M, Nguyen NT, et al. Minimally invasive surgical staging for esophageal cancer. Surg Endosc 2000;14:700 –2. [5] Velanovich V, Wollner I, Ajlouni M. Staging laparoscopy promotes increased utilization of postoperative therapy for unresectable intraabdominal malignancies. J Gastrointest Surg 2000;4:542– 6. [6] Heath EI, Kaufman HS, Talamini MA, et al. The role of laparoscopy in preoperative staging of esophageal cancer. Surg Endosc 2000;14: 495–9. [7] Molloy RG, McCourtney JS, Anderson JR. Laparoscopy in the management of patients with cancer of the gastric cardia and oesophagus. Br J Surg 1995;82:352– 4. [8] Smith A, John TG, Garden OJ, Brown SP. Role of laparoscopic ultrasonography in the management of patients with oesophagogastric cancer. Br J Surg 1999;86:1083–7. [9] Baisi A, Bonavina L, Peracchia A. Bronchoscopic staging of squamous cell carcinoma of the upper thoracic esophagus. Arch Surg 1999;134:140 –3. [10] Velasco JM, Rossi H, Hieken TJ, Fernandez M. Laparoscopic ultrasound enhances diagnostic laparoscopy in the staging of intra-abdominal neoplasms. Am Surg 2000;66:407–11. [11] Hunerbein M, Rau B, Hohenberger P, Schlag PM. The role of staging laparoscopy for multimodal therapy of gastrointestinal cancer. Surg Endosc 1998;12:921–5. [12] Krasna MJ. Minimally invasive staging for esophageal cancer. Chest 1997;112(suppl 4):191– 4.
Discussion Dr. Patrick J. Offner (Denver, CO): Dr. Nguyen and colleagues present us with 33 patients with esophageal cancer who underwent minimally invasive surgical staging prior to definitive therapy. Their results, in combination with other published reports, provide mounting evidence that minimally invasive staging techniques are accurate and may alter the treatment plan in a significant number of patients. The authors have defined minimally invasive staging as including laparoscopy, laparoscopic ultrasound, and bronchoscopy. One could question, however, the inclusion of bronchoscopy, as many of us would consider this to be part of standard preoperative evaluation for large tumors of the middle and upper esophagus. Did you perform bronchoscopy on every patient regardless of tumor location or CT scan findings and if not, how were patients selected for bronchoscopy? If you were to consider the patients staged with bronchoscopy as undergoing conventional staging procedures, would this have influenced your performance analysis at all? Four patients were noted to be unresectable owing to tumor encasement of the celiac axis on laparoscopy. It was unclear to me whether these were biopsy proved as metastatic disease; and this leads to my final and perhaps most important question: what was the gold standard to which conventional staging procedures and minimally invasive staging procedures were compared during your analysis?
Dr. Ninh T. Nguyen (Sacramento, CA): We perform bronchoscopy routinely on all patients at the time of laparoscopic staging. It is very simple to place the bronchoscope while the patient is under general anesthesia. Therefore we perform bronchoscopy and esophagoscopy routinely, followed by laparoscopic staging. Many surgeons do not perform their own preoperative upper endoscopy. Because I approach esophagectomy using a minimally invasive method, I must know the proximal and distal extent of the tumor. Tactile feedback is lost during laparoscopic and thorascopic operations; we therefore must rely on other diagnostic tests such as bronchoscopy and esophagoscopy. In the 4 patients with tumor encasement, we identified matted or fixed lymph nodes in the region of the celiac axis. These patients underwent biopsy only if it was feasible. Patients with extensive tumor encasement of the celiac axis did not undergo biopsy. The gold standard for comparison in our study was based on findings at minimally invasive surgical staging and findings at the time of surgical resection. Dr. Wynn: The CT is always done prior to the laparoscopic exploration, and you have that knowledge when you do that laparoscopic staging. So, given that, aren’t you influencing your results? Aren’t you more likely to have better results in the laparoscopic evaluation if you already have the findings of the CT in your databank? Also, given your data, are you now suggesting you don’t need to do the CT or that this needs to be done together? Dr. Ninh T. Nguyen: No, CT scanning is important to provide an overall survey for possibly liver and pulmonary metastases. It is a good evaluator for distant metastatic disease, as well as local regional involvement of the descending thoracic aorta. Minimally invasive surgical staging should be performed in addition to but not in lieu of CT scanning. Dr. David Easter (San Diego, CA): I’m wondering why you include your jejunostomy data, as it didn’t affect the staging, and I’m wondering if you’ve overutilized it. How many actually use their jejunostomies? It seems you’re turning a clean case into a clean-contaminated case. Dr. Ninh T. Nguyen: If you look at our complication rate, there was only 1 patient who developed a wound infection. We perform laparoscopic jejunostomy because some of our patients elected to be enrolled in our neoadjuvant chemotherapy/radiotherapy protocol. This regimen lasts approximately 8 to 10 weeks. During this interval, if the patients develop dysphagia, the jejunostomy tube is used for nutritional support. Dr. David Easter: Does everybody get the jejunostomy tube? Dr. Ninh T. Nguyen: Yes, everyone gets the J-tube. Dr. David Easter: Does everybody get the new adjuvant therapy? Closing Dr. Ninh T. Nguyen: No. Not everybody gets the adjuvant chemotherapy/radiotherapy treatment.