- Email: [email protected]
A phase II trial of chemotherapy and surgery for non-small cell lung cancer patients with a synchronous solitary metastasis
Lung Cancer 38 (2002) 193 /197 www.elsevier.com/locate/lungcan
A phase II trial of chemotherapy and surgery for non-small cell lung cancer patients with a synchronous solitary metastasis Robert J. Downey a,*, Kenneth K. Ng b, Mark G. Kris b, Manjit S. Bains a, Vincent A. Miller b, Robert Heelan c, Mark Bilsky d, Robert Ginsberg a,1, Valerie W. Rusch a a
Thoracic Services, Memorial Sloan-Kettering Cancer Center, Division of Thoracic Surgery, 1275 York Avenue, New York, NY 10021, USA b Department of Surgery, The Thoracic Oncology Service, Memorial Sloan-Kettering Cancer Center, New York, NY, USA c Department of Medicine, and The Department of Radiology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA d Neurosurgery Services, Memorial Sloan-Kettering Cancer Center, New York, NY, USA Received 26 February 2002; received in revised form 6 May 2002; accepted 6 May 2002
Abstract Purpose: Retrospective reports suggest that selected patients with non-small cell lung cancer (NSCLC) and a solitary synchronous site of M1 disease may be effectively treated by resection of all disease sites. The feasibility and potential benefit of combining surgery and chemotherapy in this setting are unclear. Therefore, we performed a prospective trial to test this therapeutic approach. Methods: Patients with solitary synchronous M1 NSCLC with or without N2 disease were to receive three cycles of mitomycin, vinblastine, cisplatin (MVP) chemotherapy, followed by resection of all disease sites, and then two cycles of VP chemotherapy. Solitary brain metastases were to be resected before chemotherapy. Results: From 10/92 /2/99, 23 patients (12 men, 11 women, median age /55 years) were enrolled. Mediastinoscopy, performed in 22 patients, showed involved N2 nodes in 12. The M1 sites included brain (14 patients) adrenal (3), bone (3), spleen (1), lung (1), and colon (1). Of 12 patients who completed all three induction therapy cycles, 8 underwent R0 resections. Another 5 patients had R0 resections without completing induction therapy. Eight of the 13 patients undergoing R0 resections completed postoperative chemotherapy. The median survival was 11 months; 2 patients survived to 5 years without disease. Conclusions: (1) The number of patients with solitary M1 disease who qualified for this combined modality therapy was small; (2) MVP was poorly tolerated as induction chemotherapy in this patient population; (3) Compared to historical experience with surgery alone, overall survival does not appear to be superior with this treatment strategy. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Non-small cell lung cancer; Mitomycin vinblastine cisplatin; Surgery
1. Introduction The standard treatment for fit patients with Stage IV non-small cell lung cancer (NSCLC) is chemotherapy. However, retrospective studies suggest that a small number of patients with metachronous or synchronous solitary M1 sites may benefit from resection of both the primary and the metastatic site of disease [1 /15]. The feasibility and potential benefit of combined chemother-
* Corresponding author. Tel.: /1-212-639-8124/201-891-8507; fax: /1-646-422-2078 E-mail address: [email protected] (R.J. Downey). 1 Present address: Division of Thoracic Surgery, Toronto General Hospital, Toronto, Ont., Canada.
apy and surgery for any patient with solitary M1 disease is unclear. Therefore, we performed a prospective Phase II trial to test the feasibility of combined modality therapy for patients with NSCLC and a solitary synchronous M1 site, with or without involved N2 nodes.
2. Methods The protocol was approved by the Institutional Review Board of Memorial Sloan-Kettering Cancer Center (MSKCC). Written informed consent was obtained from all patients. Patients were eligible if they had biopsy proven, previously untreated NSCLC with potentially respectable intrathoracic disease (T1 3N0 2)
0169-5002/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 9 - 5 0 0 2 ( 0 2 ) 0 0 1 8 3 - 6
194
R.J. Downey et al. / Lung Cancer 38 (2002) 193 /197
and a solitary, synchronous, resectable, metastatic lesion. A Karnofsky Performance Status (KPS) of 70% or greater was required and patients also had to have sufficient cardiopulmonary function to undergo safely the planned resection of the primary lung tumor. In addition to a history and physical exam, the pretreatment evaluation included a CT scan of the chest and upper abdomen, a CT or MR scan of the brain with contrast, a bone scan, pulmonary function tests, and a bronchoscopy and mediastinoscopy. All brain metastases were to be resected prior to chemotherapy, with the decision to give postoperative whole brain irradiation being left to the discretion of the neurosurgeon. All other patients had to have needle biopsies of the metastatic disease site for histologic proof. Induction chemotherapy was administered intravenously with mitomycin 8 mg/m2 on days 1, 29, 71; vinblastine 4.0 mg/m2 on days 1, 2.0 mg/m2 on day 8, 4.5 mg/m2 on days 15, 22, 29, then every 2 weeks; and cisplatin 120 mg/m2 or 3.0 mg/kg (whichever was less) on days 1, 29, and 71. Dose attenuation was planned for leukopenia, alteration in creatinine clearance, neurotoxicity, and thrombocytopenia. After completion of chemotherapy, reevaluation for resection was performed that included all the elements of the initial evaluation, except that repeat mediastinoscopy was not required. Resection of all remaining sites of disease was then to be performed. Lung resection was to be by lobectomy or pneumonectomy, if possible by pulmonary function tests, with a mediastinal lymph node dissection. Postoperatively, patients with completely resected tumors were to receive two cycles of vinblastine and cisplatin (mitomycin not included because of concerns about toxicity). Standard definitions of complete and partial response, stable disease, and progression of disease were used [16].
Table 1 Pretreatment patient characteristics Number of patients enrolled Men Women
23 12 11
Age in years: median (range)
55 (35 /74)
Histology Adenocarcinoma Squamous cell Large cell
17 1 5
Location of primary tumor Right upper lobe Right middle lobe Right lower lobe Left upper lobe Left lower lobe
12 1 7 1 2
T status (pretreatment) T1N0 or T2N0 T3N0 T1N1 or T2N1 T3N1 T1N2 or T2N2 T3N2
5 4 1 0 11 2
M1 site Brain Adrenal Bone Spleen Lung Colon
14 3 3 1 1 1
3. Results From October 1992 through February 1999, 23 patients were enrolled. Patient characteristics are summarized in Table 1. All patients had a KPS of 70% or greater, but 7 patients had weight loss at presentation (median 8%, range 4 /13%). The site of the bone metastases in the 3 patients presenting with M1 bone was the humerus, rib, and spine in one each. Mediastinoscopy was performed in 22 patients and involved N2 nodes found in 12; the remaining patient had mediastinal adenopathy on CT thought to be highly suspicious for malignant involvement. Fig. 1 delineates how patients completed the intended steps of the protocol. In summary, although all enrolled patients received some chemotherapy, only 12 patients completed the intended three cycles of induction mito-
Fig. 1. MSKCC Protocol 92-090: progression of patients through algorithm.
mycin, vinblastine, cisplatin (MVP). Of the other 11 patients, the chemotherapy was not completed because of progression of disease during treatment in 6, and in the remaining patients because of peripheral neuropathy (1), progressive pulmonary insufficiency (3), nadir fever (1), and severe nausea and vomiting (1). The median total dose of mitomycin administered was 24 mg/m2 (range 8 /24). The median total dose of vinblastine administered was 31 mg/m2 (range 10 /57). Fourteen patients received three cycles of cisplatin. The
R.J. Downey et al. / Lung Cancer 38 (2002) 193 /197
median total dose of cisplatin administered was 319 mg/ m2 (range 102 /362). The radiographic response of the primary site of disease to induction therapy was no response in 9 patients, minor response in 3, partial response in 8, and a complete response in none. The operation to resect the primary lung tumor was a pneumonectomy in one, a bilobectomy in 1, and a lobectomy in 12. Thirteen of these resections were R0 (grossly complete resection with microscopically negative margins) and one was R1 (grossly complete resection with microscopically positive margins). There were no operative deaths. Complications of thoracotomy included mitomycin-related respiratory insufficiency requiring steroids (1 patient), re-intubation (1 patient), and prolonged air leaks (3 patients). The pathologic T status in the 11 resected patients (either R0 or R1) was T1 in 5 patients, and T2 in 9; the pathologic N status was N0 in 6 patients, N1 in 1, and N2 in 7. A lung resection was not undertaken in the remaining 9 patients because of a brain recurrence in 5, progression of disease in other sites during chemotherapy in 4. The surgery for the M1 site was a craniotomy in 13, adrenalectomy in 1, splenectomy in 1, partial colectomy in 1, segmental bone resection in 2, and lung resection in 1. One patient had a cerebral metastases treated with stereotactic irradiation without craniotomy. The 13 craniotomies and the one colectomy were performed prior to administration of chemotherapy; the other 5 resections were performed after induction therapy. Three patients did not have resection of the M1 site, because of progression of disease during chemotherapy. Six of the 10 patients who had undergone complete resections of both primary and M1 sites received postoperative chemotherapy: five cisplatin and vinblastine, one vinblastine alone. The median total dose of cisplatin administered in the 5 patients was 115 mg/m2 (range 59/ 221). The median total dose of vinblastine administered was 26 mg/m2 (range 4.2 /27). None of the 10 patients who had undergone complete resections of both primary and M1 sites received mediastinal radiation therapy. Overall, 20 patients had definitive treatment of the M1 site, and 13 patients had complete resections of the primary site of disease. Taken together, 10 patients had complete resections of both the primary and M1 sites of disease, 8 of whom had completed three cycles of chemotherapy. The site of first recurrence in the 10 patients who had undergone complete resections of both the primary and M1 sites was the brain in 4, the mediastinum in 1, the soft tissues of the chest wall in 1, and multiple simultaneous sites in 3; 1 patient has not had a recurrence. Of the patients with M1 brain, the site of first recurrence during or after treatment was the brain in 7, soft tissues in 1, and none in 6. Of the patients with M1 sites other than brain, the site of first recurrence
195
during or after treatment was the lung in 3, mediastinum in 1, bone in 1 and none in 4. The median survival for all patients entered into the study was 11 months (range 1/104 months). At last follow-up, 3 patients were alive, with status and length of follow-up as follows: 1 patient who is free of disease at 104 months, and 2 patients who are alive with disease at 31 and 77 months after thoracotomy.
4. Discussion Retrospective studies suggest that some patients with Stage IV lung cancer limited to the primary organ and one site of extrathoracic metastatic disease are effectively treated by resection of both the primary tumor and the metastasis with or without systemic therapy [1 / 15,17]. Evidence that such a solitary metastatic disease state can be effectively treated by surgery exists for malignancies other than lung cancer. Removal of the primary site and pulmonary metastases of a sarcoma appears to be an effective treatment [18,19]. For selected patients with colorectal cancer, resection of the primary and synchronous or metachronous liver and lung resections may offer prolonged survival [20]. For lung cancer, most studies have focused on cerebral and adrenal metastases, although anecdotal reports describing management of other M1 sites have been published, including small bowel [1 /3], spleen [4,5], lymph nodes, skeletal muscle, and bone [6]. Several retrospective reviews suggest that resection of synchronous lung and brain lesions may be associated with prolonged survival. In 1976, Magilligan et al. [21] published the first series of patients undergoing combined resection of a synchronous solitary cerebral metastasis and a primary NSCLC, which was updated to include a total of 41 patients in 1986 [8], some of whom received cranial radiation and/or chemotherapy as well as surgery. Overall survival was 55% at 1 year, 21% at 5 years, and 15% at 10 years. Similarly, in 1989, Read et al.[9] reported that if both a pulmonary resection and a brain resection could be performed, overall survival of 52% at 1 year, 35% at 2 years, and 21% at 5 years could be achieved. A retrospective analysis of the MSKCC experience with brain metastatectomy was published by Burt et al. in 1992 [10], and later updated in a report [11] which included 185 patients with NSCLC who experienced a median survival of 27 months if the intrathoracic disease was resected, and 11 months if it was not. Neither Read et al. [9] nor Burt et al. [10] separated synchronous from metachronous presentations in their analysis In 1996, Mussi et al.[12] reported 19 patients with synchronous and 33 with metachronous isolated cerebral metastases who were treated with surgical resection. Actuarial 5year survivals were 6.6% in the synchronous group, and
196
R.J. Downey et al. / Lung Cancer 38 (2002) 193 /197
19% in the metachronous group. Most recently, in 2001, the Mayo Clinic reported 28 patients with NSCLC and a synchronous solitary brain metastases treated over 22 years [13]. Overall survival at 1, 2, and 5 years was 64, 54, and 21%, respectively. Taken together, these retrospective reports suggest that if a complete resection of both the primary lung tumor and the cerebral metastasis can be performed, survivals at 1 year of 50% and at 5 years of 10 /30% can be expected. Patients with metachronous disease may have a better prognosis than patients with synchronous disease. There have been several published reports of patients undergoing combined resection of a primary lung cancer NSCLC and a solitary adrenal metastasis, but no prospective data. A review [14] compiling these anecdotal reports indicates that the adrenal metastasis was synchronous in 59%, and that the loco-regional (primary tumor) stage was Stage I in 22%, II in 16%, III in 43%, and not specified in 18% of patients. The median survival was 24 months and approximately one-third of patients survived 5 years. The MSKCC experience [15] (which was included in the above review [14]) suggests that the median survival of patients with adrenal metastases treated with chemotherapy alone is approximately 8.5 months, with all patients dead by 21 months, but the survival of patients treated with chemotherapy and surgical resection of all known disease is 31 months (with two long-term survivors). Our most recent retrospective review of patients treated at MSKCC with induction chemotherapy and surgery for NSCLC [22] included 43 patients with solitary site M1 disease; some of these patients were enrolled on the prospective protocol that is the subject of this paper. The sites of M1 disease were all proven by histopathology, and included the brain in 16, the lung in 9, the adrenals in 7, the bone in 7, and the colon, an inguinal node, the spleen, and the subcutaneous tissues in 1 patient each. The survival of patients with M1 disease detected preoperatively was 18.8 months, and of patients with M1 detected intraoperatively, 7 months. With such data suggesting both a role for surgery but also that survival with surgery alone was limited, we designed a prospective Phase II study combining chemotherapy and surgical resection of all sites of disease for patients with NSCLC and a solitary metastasis. By being prospective, our study has provided unique information on the feasibility of combined medical and surgical therapies: However, the small number of patients and the heterogeneity of their clinical presentation does not permit subgroup analyses by either M1 site or nodal status. Our results suggest first, that the sequence of induction therapy-surgical resection of primary and metastatic sites-adjunctive therapy is difficult to complete, both because intolerance and progression of disease during treatment. Second, survival was poor, with only
2 patients out of 23 alive without disease at 5 years. This result, despite being poorer than prior reports, is not inconsistent with the many retrospective studies that have been previously published. If our experience had been reviewed retrospectively by a search of our databases for patients who had undergone complete resections of a solitary M1 site and intrathoracic locoregional disease, 10 would have been found with 3 patients alive at last follow-up (30%) and two true 5year survivors (20%). These are the results found in our retrospective review of all patients undergoing exploration with the goal of curative resection after induction therapy by Martin et al.[22] discussed above and similar to the retrospective report from the Mayo Clinic [13]. The prospective nature of our current study defines ‘the denominator’ and allows a more accurate prognosis for the patient initiating therapy, specifically, an expectation of a 4/8% chance of being alive and disease free at 5 years. However, because of the small number of patients enrolled, this study does not provide information to definitively support or refute prior case reports suggesting a role for the resection of all sites of disease as complete therapy for patients with solitary metastases. For now, it appears reasonable to recommend that highly selected patients with a solitary resectable metastasis from NSCLC undergo surgical resection of all evident disease, especially if the primary lung tumor would otherwise be Stage I or II. However, it is difficult on the basis of this trial to recommend the regimen of combined medical and surgical therapies used in this protocol in the management of solitary M1 disease. Consideration could be given to trials combining surgery and newer, less toxic chemotherapy.
Acknowledgements The editorial assistance of Ms Phyllis Byczek, the review of statistical issues by E.S. Venkatraman, Ph.D., and Mr Alain Vincent for data management.
References [1] Hinoshita E, Nakahashi H, Wakasugi K, et al. Duodenal metastasis from large cell carcinoma of the lung: report of a case. Surg Today 1999;29(8):799 /802. [2] Berger A, Cellier C, Daniel C, et al. Small bowel metastases from primary carcinoma of the lung: clinical findings and outcome. Am J Gastroenterol 1999;94(7):1884 /7. [3] Mosier DM, Bloch RS, Cunningham PL, et al. Small bowel metastases from primary lung carcinoma: a rarity waiting to be found? Am Surg 1992;58:677 /82. [4] Macheers SK, Mansour KA. Management of isolated splenic metastases from carcinoma of the lung: a case report and review of the literature. Am Surg 1992;58:683 /5.
R.J. Downey et al. / Lung Cancer 38 (2002) 193 /197 [5] Edelman AS, Rotterdam H. Solitary splenic metastasis of an adenocarcinoma of the lung. Am J Clin Pathol 1990;94:326 /8. [6] Luketich JD, Martini N, Ginsberg RJ, et al. Successful treatment of solitary extracranial metastases from non-small cell lung cancer. Ann Thorac Surg 1995;60:1609 /11. [7] Saitoh Y, Fujisawa T, Shiba M, et al. Prognostic factors in surgical treatment of solitary brain metastasis after resection of non-small-cell lung cancer. Lung Cancer 1999;24(2):99 /106. [8] Magilligan DJ, Jr., Duvernoy C, Malik G, et al. Surgical approach to lung cancer with solitary cerebral metastasis: twenty-five years’ experience. Ann Thor Surg 1986;42(4):360 /4. [9] Read RC, Boop WC, Yoder G, et al. Management of non-small cell lung carcinoma with solitary brain metastasis. J Thorac Cardiovasc Surg 1989;98:884 /90. [10] Burt M, Wronski M, Arbit E, et al. Resection of brain metastases from non-small cell lung carcinoma. J Thorac Cardiovasc Surg 1992;103:399 /410. [11] Wronski M, Arbit E, Burt M, et al. Survival after surgical treatment of brain metastases from lung cancer: a follow-up study of 231 patients treated between 1976 and 1991. J Neurosurg 1995;83:605 /16. [12] Mussi A, Pistolesi M, Lucchi M, et al. Resection of a single brain metastasis in non-small-cell lung cancer: prognostic factors. J Thorac Cardiovasc Surg 1996;112(1):146 /53. [13] Billing PS, Miller DL, Allen MS, et al. Surgical treatment of primary lung cancer with synchronous brain metastases. J Thorac Cardiovasc Surg 2001;122:548 /53.
197
[14] Beitler AL, Urschel JD, Velagapudi SR, et al. Surgical management of adrenal metastases from lung cancer. J Surg Oncol 1998;69:54 /7. [15] Luketich JD, Burt ME. Does resection of adrenal metastases from non-small cell lung cancer improve survival? Ann Thorac Surg 1996;62:1614 /6. [16] Green S, Weiss GR. Southwest Oncology Group standard response criteria, endpoint analysis definitions, and toxicity criteria. Invest New Drugs 1992;10:239 /52. [17] Hellman S, Weichselbaum RR. Oligometastases. J Clin Oncol 1995;13(1):8 /10. [18] Billingsley KG, Burt ME, Jara E, et al. Pulmonary metastases from soft tissue sarcoma: analysis of patterns of diseases and postmetastasis survival. Ann Surg 1999;229(5):602 /10. [19] Billingsley KG, Lewis JJ, Leung DH, et al. Multifactorial analysis of the survival of patients with distant metastasis arising from primary extremity sarcoma. Cancer 1999;85:389 /95. [20] Fong Y, Salo J. Surgical therapy of hepatic colorectal metastasis. Semin Oncol 1999;26:514 /23. [21] Magilligan DJ, Jr., Rogers JS, Knighton RS, et al. Pulmonary neoplasm with solitary cerebral metastasis: results of combined excision. J Thorac Cardiovasc Surg 1976;72(5):690 /8. [22] Martin J, Ginsberg RJ, Venkatraman ES, et al. Long-term results of combined modality therapy in resectable non-small cell lung cancer. J Clin Oncol 2002;20(8):1989 /95.
A phase II trial of chemotherapy and surgery for non-small cell lung cancer patients with a synchronous solitary metastasis Robert J. Downey a,*, Kenneth K. Ng b, Mark G. Kris b, Manjit S. Bains a, Vincent A. Miller b, Robert Heelan c, Mark Bilsky d, Robert Ginsberg a,1, Valerie W. Rusch a a
Thoracic Services, Memorial Sloan-Kettering Cancer Center, Division of Thoracic Surgery, 1275 York Avenue, New York, NY 10021, USA b Department of Surgery, The Thoracic Oncology Service, Memorial Sloan-Kettering Cancer Center, New York, NY, USA c Department of Medicine, and The Department of Radiology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA d Neurosurgery Services, Memorial Sloan-Kettering Cancer Center, New York, NY, USA Received 26 February 2002; received in revised form 6 May 2002; accepted 6 May 2002
Abstract Purpose: Retrospective reports suggest that selected patients with non-small cell lung cancer (NSCLC) and a solitary synchronous site of M1 disease may be effectively treated by resection of all disease sites. The feasibility and potential benefit of combining surgery and chemotherapy in this setting are unclear. Therefore, we performed a prospective trial to test this therapeutic approach. Methods: Patients with solitary synchronous M1 NSCLC with or without N2 disease were to receive three cycles of mitomycin, vinblastine, cisplatin (MVP) chemotherapy, followed by resection of all disease sites, and then two cycles of VP chemotherapy. Solitary brain metastases were to be resected before chemotherapy. Results: From 10/92 /2/99, 23 patients (12 men, 11 women, median age /55 years) were enrolled. Mediastinoscopy, performed in 22 patients, showed involved N2 nodes in 12. The M1 sites included brain (14 patients) adrenal (3), bone (3), spleen (1), lung (1), and colon (1). Of 12 patients who completed all three induction therapy cycles, 8 underwent R0 resections. Another 5 patients had R0 resections without completing induction therapy. Eight of the 13 patients undergoing R0 resections completed postoperative chemotherapy. The median survival was 11 months; 2 patients survived to 5 years without disease. Conclusions: (1) The number of patients with solitary M1 disease who qualified for this combined modality therapy was small; (2) MVP was poorly tolerated as induction chemotherapy in this patient population; (3) Compared to historical experience with surgery alone, overall survival does not appear to be superior with this treatment strategy. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Non-small cell lung cancer; Mitomycin vinblastine cisplatin; Surgery
1. Introduction The standard treatment for fit patients with Stage IV non-small cell lung cancer (NSCLC) is chemotherapy. However, retrospective studies suggest that a small number of patients with metachronous or synchronous solitary M1 sites may benefit from resection of both the primary and the metastatic site of disease [1 /15]. The feasibility and potential benefit of combined chemother-
* Corresponding author. Tel.: /1-212-639-8124/201-891-8507; fax: /1-646-422-2078 E-mail address: [email protected] (R.J. Downey). 1 Present address: Division of Thoracic Surgery, Toronto General Hospital, Toronto, Ont., Canada.
apy and surgery for any patient with solitary M1 disease is unclear. Therefore, we performed a prospective Phase II trial to test the feasibility of combined modality therapy for patients with NSCLC and a solitary synchronous M1 site, with or without involved N2 nodes.
2. Methods The protocol was approved by the Institutional Review Board of Memorial Sloan-Kettering Cancer Center (MSKCC). Written informed consent was obtained from all patients. Patients were eligible if they had biopsy proven, previously untreated NSCLC with potentially respectable intrathoracic disease (T1 3N0 2)
0169-5002/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 9 - 5 0 0 2 ( 0 2 ) 0 0 1 8 3 - 6
194
R.J. Downey et al. / Lung Cancer 38 (2002) 193 /197
and a solitary, synchronous, resectable, metastatic lesion. A Karnofsky Performance Status (KPS) of 70% or greater was required and patients also had to have sufficient cardiopulmonary function to undergo safely the planned resection of the primary lung tumor. In addition to a history and physical exam, the pretreatment evaluation included a CT scan of the chest and upper abdomen, a CT or MR scan of the brain with contrast, a bone scan, pulmonary function tests, and a bronchoscopy and mediastinoscopy. All brain metastases were to be resected prior to chemotherapy, with the decision to give postoperative whole brain irradiation being left to the discretion of the neurosurgeon. All other patients had to have needle biopsies of the metastatic disease site for histologic proof. Induction chemotherapy was administered intravenously with mitomycin 8 mg/m2 on days 1, 29, 71; vinblastine 4.0 mg/m2 on days 1, 2.0 mg/m2 on day 8, 4.5 mg/m2 on days 15, 22, 29, then every 2 weeks; and cisplatin 120 mg/m2 or 3.0 mg/kg (whichever was less) on days 1, 29, and 71. Dose attenuation was planned for leukopenia, alteration in creatinine clearance, neurotoxicity, and thrombocytopenia. After completion of chemotherapy, reevaluation for resection was performed that included all the elements of the initial evaluation, except that repeat mediastinoscopy was not required. Resection of all remaining sites of disease was then to be performed. Lung resection was to be by lobectomy or pneumonectomy, if possible by pulmonary function tests, with a mediastinal lymph node dissection. Postoperatively, patients with completely resected tumors were to receive two cycles of vinblastine and cisplatin (mitomycin not included because of concerns about toxicity). Standard definitions of complete and partial response, stable disease, and progression of disease were used [16].
Table 1 Pretreatment patient characteristics Number of patients enrolled Men Women
23 12 11
Age in years: median (range)
55 (35 /74)
Histology Adenocarcinoma Squamous cell Large cell
17 1 5
Location of primary tumor Right upper lobe Right middle lobe Right lower lobe Left upper lobe Left lower lobe
12 1 7 1 2
T status (pretreatment) T1N0 or T2N0 T3N0 T1N1 or T2N1 T3N1 T1N2 or T2N2 T3N2
5 4 1 0 11 2
M1 site Brain Adrenal Bone Spleen Lung Colon
14 3 3 1 1 1
3. Results From October 1992 through February 1999, 23 patients were enrolled. Patient characteristics are summarized in Table 1. All patients had a KPS of 70% or greater, but 7 patients had weight loss at presentation (median 8%, range 4 /13%). The site of the bone metastases in the 3 patients presenting with M1 bone was the humerus, rib, and spine in one each. Mediastinoscopy was performed in 22 patients and involved N2 nodes found in 12; the remaining patient had mediastinal adenopathy on CT thought to be highly suspicious for malignant involvement. Fig. 1 delineates how patients completed the intended steps of the protocol. In summary, although all enrolled patients received some chemotherapy, only 12 patients completed the intended three cycles of induction mito-
Fig. 1. MSKCC Protocol 92-090: progression of patients through algorithm.
mycin, vinblastine, cisplatin (MVP). Of the other 11 patients, the chemotherapy was not completed because of progression of disease during treatment in 6, and in the remaining patients because of peripheral neuropathy (1), progressive pulmonary insufficiency (3), nadir fever (1), and severe nausea and vomiting (1). The median total dose of mitomycin administered was 24 mg/m2 (range 8 /24). The median total dose of vinblastine administered was 31 mg/m2 (range 10 /57). Fourteen patients received three cycles of cisplatin. The
R.J. Downey et al. / Lung Cancer 38 (2002) 193 /197
median total dose of cisplatin administered was 319 mg/ m2 (range 102 /362). The radiographic response of the primary site of disease to induction therapy was no response in 9 patients, minor response in 3, partial response in 8, and a complete response in none. The operation to resect the primary lung tumor was a pneumonectomy in one, a bilobectomy in 1, and a lobectomy in 12. Thirteen of these resections were R0 (grossly complete resection with microscopically negative margins) and one was R1 (grossly complete resection with microscopically positive margins). There were no operative deaths. Complications of thoracotomy included mitomycin-related respiratory insufficiency requiring steroids (1 patient), re-intubation (1 patient), and prolonged air leaks (3 patients). The pathologic T status in the 11 resected patients (either R0 or R1) was T1 in 5 patients, and T2 in 9; the pathologic N status was N0 in 6 patients, N1 in 1, and N2 in 7. A lung resection was not undertaken in the remaining 9 patients because of a brain recurrence in 5, progression of disease in other sites during chemotherapy in 4. The surgery for the M1 site was a craniotomy in 13, adrenalectomy in 1, splenectomy in 1, partial colectomy in 1, segmental bone resection in 2, and lung resection in 1. One patient had a cerebral metastases treated with stereotactic irradiation without craniotomy. The 13 craniotomies and the one colectomy were performed prior to administration of chemotherapy; the other 5 resections were performed after induction therapy. Three patients did not have resection of the M1 site, because of progression of disease during chemotherapy. Six of the 10 patients who had undergone complete resections of both primary and M1 sites received postoperative chemotherapy: five cisplatin and vinblastine, one vinblastine alone. The median total dose of cisplatin administered in the 5 patients was 115 mg/m2 (range 59/ 221). The median total dose of vinblastine administered was 26 mg/m2 (range 4.2 /27). None of the 10 patients who had undergone complete resections of both primary and M1 sites received mediastinal radiation therapy. Overall, 20 patients had definitive treatment of the M1 site, and 13 patients had complete resections of the primary site of disease. Taken together, 10 patients had complete resections of both the primary and M1 sites of disease, 8 of whom had completed three cycles of chemotherapy. The site of first recurrence in the 10 patients who had undergone complete resections of both the primary and M1 sites was the brain in 4, the mediastinum in 1, the soft tissues of the chest wall in 1, and multiple simultaneous sites in 3; 1 patient has not had a recurrence. Of the patients with M1 brain, the site of first recurrence during or after treatment was the brain in 7, soft tissues in 1, and none in 6. Of the patients with M1 sites other than brain, the site of first recurrence
195
during or after treatment was the lung in 3, mediastinum in 1, bone in 1 and none in 4. The median survival for all patients entered into the study was 11 months (range 1/104 months). At last follow-up, 3 patients were alive, with status and length of follow-up as follows: 1 patient who is free of disease at 104 months, and 2 patients who are alive with disease at 31 and 77 months after thoracotomy.
4. Discussion Retrospective studies suggest that some patients with Stage IV lung cancer limited to the primary organ and one site of extrathoracic metastatic disease are effectively treated by resection of both the primary tumor and the metastasis with or without systemic therapy [1 / 15,17]. Evidence that such a solitary metastatic disease state can be effectively treated by surgery exists for malignancies other than lung cancer. Removal of the primary site and pulmonary metastases of a sarcoma appears to be an effective treatment [18,19]. For selected patients with colorectal cancer, resection of the primary and synchronous or metachronous liver and lung resections may offer prolonged survival [20]. For lung cancer, most studies have focused on cerebral and adrenal metastases, although anecdotal reports describing management of other M1 sites have been published, including small bowel [1 /3], spleen [4,5], lymph nodes, skeletal muscle, and bone [6]. Several retrospective reviews suggest that resection of synchronous lung and brain lesions may be associated with prolonged survival. In 1976, Magilligan et al. [21] published the first series of patients undergoing combined resection of a synchronous solitary cerebral metastasis and a primary NSCLC, which was updated to include a total of 41 patients in 1986 [8], some of whom received cranial radiation and/or chemotherapy as well as surgery. Overall survival was 55% at 1 year, 21% at 5 years, and 15% at 10 years. Similarly, in 1989, Read et al.[9] reported that if both a pulmonary resection and a brain resection could be performed, overall survival of 52% at 1 year, 35% at 2 years, and 21% at 5 years could be achieved. A retrospective analysis of the MSKCC experience with brain metastatectomy was published by Burt et al. in 1992 [10], and later updated in a report [11] which included 185 patients with NSCLC who experienced a median survival of 27 months if the intrathoracic disease was resected, and 11 months if it was not. Neither Read et al. [9] nor Burt et al. [10] separated synchronous from metachronous presentations in their analysis In 1996, Mussi et al.[12] reported 19 patients with synchronous and 33 with metachronous isolated cerebral metastases who were treated with surgical resection. Actuarial 5year survivals were 6.6% in the synchronous group, and
196
R.J. Downey et al. / Lung Cancer 38 (2002) 193 /197
19% in the metachronous group. Most recently, in 2001, the Mayo Clinic reported 28 patients with NSCLC and a synchronous solitary brain metastases treated over 22 years [13]. Overall survival at 1, 2, and 5 years was 64, 54, and 21%, respectively. Taken together, these retrospective reports suggest that if a complete resection of both the primary lung tumor and the cerebral metastasis can be performed, survivals at 1 year of 50% and at 5 years of 10 /30% can be expected. Patients with metachronous disease may have a better prognosis than patients with synchronous disease. There have been several published reports of patients undergoing combined resection of a primary lung cancer NSCLC and a solitary adrenal metastasis, but no prospective data. A review [14] compiling these anecdotal reports indicates that the adrenal metastasis was synchronous in 59%, and that the loco-regional (primary tumor) stage was Stage I in 22%, II in 16%, III in 43%, and not specified in 18% of patients. The median survival was 24 months and approximately one-third of patients survived 5 years. The MSKCC experience [15] (which was included in the above review [14]) suggests that the median survival of patients with adrenal metastases treated with chemotherapy alone is approximately 8.5 months, with all patients dead by 21 months, but the survival of patients treated with chemotherapy and surgical resection of all known disease is 31 months (with two long-term survivors). Our most recent retrospective review of patients treated at MSKCC with induction chemotherapy and surgery for NSCLC [22] included 43 patients with solitary site M1 disease; some of these patients were enrolled on the prospective protocol that is the subject of this paper. The sites of M1 disease were all proven by histopathology, and included the brain in 16, the lung in 9, the adrenals in 7, the bone in 7, and the colon, an inguinal node, the spleen, and the subcutaneous tissues in 1 patient each. The survival of patients with M1 disease detected preoperatively was 18.8 months, and of patients with M1 detected intraoperatively, 7 months. With such data suggesting both a role for surgery but also that survival with surgery alone was limited, we designed a prospective Phase II study combining chemotherapy and surgical resection of all sites of disease for patients with NSCLC and a solitary metastasis. By being prospective, our study has provided unique information on the feasibility of combined medical and surgical therapies: However, the small number of patients and the heterogeneity of their clinical presentation does not permit subgroup analyses by either M1 site or nodal status. Our results suggest first, that the sequence of induction therapy-surgical resection of primary and metastatic sites-adjunctive therapy is difficult to complete, both because intolerance and progression of disease during treatment. Second, survival was poor, with only
2 patients out of 23 alive without disease at 5 years. This result, despite being poorer than prior reports, is not inconsistent with the many retrospective studies that have been previously published. If our experience had been reviewed retrospectively by a search of our databases for patients who had undergone complete resections of a solitary M1 site and intrathoracic locoregional disease, 10 would have been found with 3 patients alive at last follow-up (30%) and two true 5year survivors (20%). These are the results found in our retrospective review of all patients undergoing exploration with the goal of curative resection after induction therapy by Martin et al.[22] discussed above and similar to the retrospective report from the Mayo Clinic [13]. The prospective nature of our current study defines ‘the denominator’ and allows a more accurate prognosis for the patient initiating therapy, specifically, an expectation of a 4/8% chance of being alive and disease free at 5 years. However, because of the small number of patients enrolled, this study does not provide information to definitively support or refute prior case reports suggesting a role for the resection of all sites of disease as complete therapy for patients with solitary metastases. For now, it appears reasonable to recommend that highly selected patients with a solitary resectable metastasis from NSCLC undergo surgical resection of all evident disease, especially if the primary lung tumor would otherwise be Stage I or II. However, it is difficult on the basis of this trial to recommend the regimen of combined medical and surgical therapies used in this protocol in the management of solitary M1 disease. Consideration could be given to trials combining surgery and newer, less toxic chemotherapy.
Acknowledgements The editorial assistance of Ms Phyllis Byczek, the review of statistical issues by E.S. Venkatraman, Ph.D., and Mr Alain Vincent for data management.
References [1] Hinoshita E, Nakahashi H, Wakasugi K, et al. Duodenal metastasis from large cell carcinoma of the lung: report of a case. Surg Today 1999;29(8):799 /802. [2] Berger A, Cellier C, Daniel C, et al. Small bowel metastases from primary carcinoma of the lung: clinical findings and outcome. Am J Gastroenterol 1999;94(7):1884 /7. [3] Mosier DM, Bloch RS, Cunningham PL, et al. Small bowel metastases from primary lung carcinoma: a rarity waiting to be found? Am Surg 1992;58:677 /82. [4] Macheers SK, Mansour KA. Management of isolated splenic metastases from carcinoma of the lung: a case report and review of the literature. Am Surg 1992;58:683 /5.
R.J. Downey et al. / Lung Cancer 38 (2002) 193 /197 [5] Edelman AS, Rotterdam H. Solitary splenic metastasis of an adenocarcinoma of the lung. Am J Clin Pathol 1990;94:326 /8. [6] Luketich JD, Martini N, Ginsberg RJ, et al. Successful treatment of solitary extracranial metastases from non-small cell lung cancer. Ann Thorac Surg 1995;60:1609 /11. [7] Saitoh Y, Fujisawa T, Shiba M, et al. Prognostic factors in surgical treatment of solitary brain metastasis after resection of non-small-cell lung cancer. Lung Cancer 1999;24(2):99 /106. [8] Magilligan DJ, Jr., Duvernoy C, Malik G, et al. Surgical approach to lung cancer with solitary cerebral metastasis: twenty-five years’ experience. Ann Thor Surg 1986;42(4):360 /4. [9] Read RC, Boop WC, Yoder G, et al. Management of non-small cell lung carcinoma with solitary brain metastasis. J Thorac Cardiovasc Surg 1989;98:884 /90. [10] Burt M, Wronski M, Arbit E, et al. Resection of brain metastases from non-small cell lung carcinoma. J Thorac Cardiovasc Surg 1992;103:399 /410. [11] Wronski M, Arbit E, Burt M, et al. Survival after surgical treatment of brain metastases from lung cancer: a follow-up study of 231 patients treated between 1976 and 1991. J Neurosurg 1995;83:605 /16. [12] Mussi A, Pistolesi M, Lucchi M, et al. Resection of a single brain metastasis in non-small-cell lung cancer: prognostic factors. J Thorac Cardiovasc Surg 1996;112(1):146 /53. [13] Billing PS, Miller DL, Allen MS, et al. Surgical treatment of primary lung cancer with synchronous brain metastases. J Thorac Cardiovasc Surg 2001;122:548 /53.
197
[14] Beitler AL, Urschel JD, Velagapudi SR, et al. Surgical management of adrenal metastases from lung cancer. J Surg Oncol 1998;69:54 /7. [15] Luketich JD, Burt ME. Does resection of adrenal metastases from non-small cell lung cancer improve survival? Ann Thorac Surg 1996;62:1614 /6. [16] Green S, Weiss GR. Southwest Oncology Group standard response criteria, endpoint analysis definitions, and toxicity criteria. Invest New Drugs 1992;10:239 /52. [17] Hellman S, Weichselbaum RR. Oligometastases. J Clin Oncol 1995;13(1):8 /10. [18] Billingsley KG, Burt ME, Jara E, et al. Pulmonary metastases from soft tissue sarcoma: analysis of patterns of diseases and postmetastasis survival. Ann Surg 1999;229(5):602 /10. [19] Billingsley KG, Lewis JJ, Leung DH, et al. Multifactorial analysis of the survival of patients with distant metastasis arising from primary extremity sarcoma. Cancer 1999;85:389 /95. [20] Fong Y, Salo J. Surgical therapy of hepatic colorectal metastasis. Semin Oncol 1999;26:514 /23. [21] Magilligan DJ, Jr., Rogers JS, Knighton RS, et al. Pulmonary neoplasm with solitary cerebral metastasis: results of combined excision. J Thorac Cardiovasc Surg 1976;72(5):690 /8. [22] Martin J, Ginsberg RJ, Venkatraman ES, et al. Long-term results of combined modality therapy in resectable non-small cell lung cancer. J Clin Oncol 2002;20(8):1989 /95.