- Email: [email protected]
Prognostic factors after surgical treatment of lung cancer invading the diaphragm
Prognostic Factors After Surgical Treatment of Lung Cancer Invading the Diaphragm Gaetano Rocco, MD, Erino A. Rendina, MD, Alberto Meroni, MD, Federico Venuta, MD, Claudio Della Pona, MD, Tiziano De Giacomo, MD, Mario Robustellini, MD, Gerolamo Rossi, MD, Fabio Massera, MD, Giuseppe Vertemati, MD, Adriano Rizzi, MD, and Giorgio F. Coloni, MD Division of General Thoracic Surgery, Azienda Ospedaliera “E. Morelli,” Sondalo (Sondrio), and Division of General Thoracic Surgery, University “La Sapienza,” Rome, Italy
Background. Diaphragmatic invasion from lung cancer (T3-diaphragm) is a rare occurrence reported to portend a poor prognosis. Methods. Fifteen patients with T3-diaphragm (14 males, 1 female; median age, 64 years) were surgically treated over a twenty-year period by en bloc resection (14 patients). One patient was only explored. Pathologic stage IIB (T3N0) was found in 11 patients. A partial infiltration of the diaphragm was observed in 3 patients, whereas full-depth invasion was found in 12. Diaphragmatic reconstruction was done primarily in 9 patients, and, by prosthetic material in 5.
Results. Two patients are still alive without evidence of disease at 88, and, 114 months from surgery. Overall median survival was 23 months (range, 3 to 168). The actuarial 5-year survival was 20%, when all patients were considered, and, 27%, for T3N0 patients. Univariate analysis showed that prosthetic replacement of the muscle (p ⴝ 0.018) was significantly related to survival. Conclusions. T3-diaphragm is best treated with en bloc resections with wide tumor-free margins and prosthetic replacement of the diaphragm. (Ann Thorac Surg 1999;68:2065– 8) © 1999 by The Society of Thoracic Surgeons
I
bilobectomy in 2, and, a pneumonectomy in 1. In the two bilobectomy patients, the middle lobe was resected due to the involvement of the major fissure by the pulmonary tumor. One patient underwent an exploratory thoracotomy only. Complete hilar and mediastinal lymph node dissection was done in all resected patients. Nine patients had an epidermoid histotype and 6 had an adenocarcinoma. Tumors were graded according to Broder’s classification as grade I (4 patients), II (5 patients), and, III (4 patients). Tumor grade was unknown in 2 patients. Pathologic stage was stage IIB in 11 patients, and, stage IIIA in 4. Eleven patients showed no nodal involvement (T3N0). On the other hand, N1 and N2 disease was found in 1 and 3 patients, respectively. Three patients had only a partial infiltration of the diaphragm whereas, in the remaining 12, full-depth extension of the pulmonary neoplasm through the muscular structure was observed. On pathologic examination, all patients had tumor-free margins on the diaphragm resected en bloc with the pulmonary neoplasm. The diaphragm was reconstructed primarily in 9 cases, and, using prosthetic materials in 5. Reconstruction was done using a Marlex mesh (Davol, Covington, GA) in 3 patients, and, a Gore-Tex (W.L. Gore, Flagstaff, AZ) patch in 2. No patients died within 30 days of operation. Minor bleeding and atrial fibrillation complicated the postoperative course of 2 patients. Postoperatively, four patients (numbers 1, 3, 8, and, 9) received an adjuvant treatment. The 11 patients with T3N0 disease were all males, all but one had full-depth muscle invasion, and, all under-
nvasion of the diaphragm is a rare occurrence demonstrated in less than 0.5% of the patients with locally advanced lung cancer [1, 2]. Few reports have addressed the issue of exactly determining the position of diaphragmatic infiltration into the staging system for lung cancer. The revised version of this staging system [3] still classifies diaphragmatic invasion as T3. However, limited series of surgically treated T3 lung cancers invading the diaphragm have included no 3-year survivors [2].
Material and Methods Between January, 1978, and December, 1997, 15 patients with lung cancer invading the diaphragm were surgically treated at our institutions (Table 11). This figure represents 0.4% of all resected patients for lung cancer in the same period. There were 14 males and 1 female, with a median age of 64 years (range, 34 to 79). Preoperative work-up, including computed tomography of the chest in all instances, had suggested the invasion of the diaphragm in only 1 patient (cT3). Mediastinoscopy was done in four patients (numbers 10, 12, 13, and, 14) and yielded negative results. One patient (number 1) underwent induction chemotherapy following video-assisted thoracoscopic demonstration of N2 disease. At operation, a lobectomy was done in 11 patients, a Accepted for publication May 27, 1999. Address reprint requests to Dr Rocco, Via Agricoltura, 20 23037, Tirano (Sondrio), Italy; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
0003-4975/99/$20.00 PII S0003-4975(99)01121-2
2066
ROCCO ET AL T3-DIAPHRAGM
Ann Thorac Surg 1999;68:2065– 8
Table 1. Clinical Features of 15 Patients With Lung Cancer Invading the Diaphragm (T3-Diaphragm) Depth of Histology Patient Age (y) Sex cT3 Invasion P Stage Grade Type 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
57 71 79 69 67 70 67 46 62 64 34 60 53 66 53
F M M M M M M M M M M M M M M
no no no yes no no no no no no no no no no no
partial full full full partial full partial full full full full full full full full
T3N2 T3N0 T3N1 T3N0 T3N0 T3N0 T3N2 T3N2 T3N0 T3N0 T3N0 T3N0 T3N0 T3N0 T3N0
III II I ... ... III III II I II I III I II II
Ad Sq Ad Ad Ad Sq Sq Sq Sq Sq Sq Ad Ad Sq Sq
Operation Pneumonect Lobect Exploration Lobect Bilobect Lobect Lobect Lobect Lobect Lobect Lobect Bilobect Lobect Lobect Lobect
Reconstruction Status primary primary ... primary primary primary primary primary primary prosthesis prosthesis primary prosthesis prosthesis prosthesis
DOD DOD DOD DOD Alive DOD DOUC DOD DOD DOUC Alive DOD DOD DOD DOD
Ad ⫽ adenocarcinoma; Bilobect ⫽ bilobectomy; DFI ⫽ disease-free interval; DOD ⫽ dead of disease; condition; Lobect ⫽ lobectomy; Pneumonect ⫽ pneumonectomy; Sq ⫽ squamous cell carcinoma.
went resection for cure as the primary treatment modality. In this group, the diaphragm was reconstructed primarily in 6 patients and using a prosthesis in 5. The only patient with a superficial infiltration of the diaphragm underwent limited circular resection and primary reconstruction of the diaphragm. Only 1 patient in this group (number 9) received radiotherapy in an adjuvant setting. Statistical analysis was done using Statview, version 4.5 (Abacus Concepts, Berkeley, CA). Actuarial survival was calculated by the Kaplan-Meier method and differences between groups were assessed by the Gehan-Wilcoxon test, as indicated. Statistical significance was set at a p value less than 0.05.
Results Upon completion of the study, the median follow-up is 23 months (range, 3 to 168 months), and, is complete in all patients. Two patients (13%) are alive without recurrent disease at 88, and, 114 months from operation. These two patients had T3N0 disease. For the other patients, the cause of death was tumor progression in 11 (73%), and, it was unrelated to the tumor in 2 (1 with T3N0 disease). Commonly involved sites by metastatic disease were the liver (8 patients), and, the lung (2 patients). Brain metastases were found in one patient only. Local recurrences in the ipsilateral hemithorax occurred in 6 patients, mostly in combination with liver metastases. Overall median survival was 23 months (range, 3 to 168). The actuarial 5-year survival was 20% when all patients were evaluated (Fig 1). The median disease-free interval was 18 months (range, 1 to 168). Overall recurrence rates at 1, 2, and, 5 years were 47%, 61%, and, 77%, respectively. When only the patients with T3N0 subset were consid-
Survival DFI Recurrence (mo) (mo) Site 13 9 13 3 114 3 23 6 27 168 88 14 37 27 44
7 3 10 1 114 1 23 3 18 168 88 12 36 24 40
Liver Liver/local ... Lung/local ... Liver/local ... Liver/brain Liver ... ... Liver/local Local Liver/lung Liver/local
DOUC ⫽ dead of unrelated
ered, the median survival was 27 months (range, 3 to 168), and the actuarial 5-year survival was 27% (Fig 2). In this group of patients, the median disease-free interval was 24 months (range, 1 to 168) and the recurrence rates at 1, 2, and, 5 years were 36%, 54%, and, 73%, respectively. Despite the limited number of patients available for evaluation, univariate analysis was conducted to assess the impact of each variable on survival (Table 24). The only variable predicting a favorable postsurgical outcome was the prosthetic replacement of the diaphragm. In fact, stratification of survival times according to the technique used for diaphragmatic reconstruction showed that median survival after primary reconstruction was 13 months (range, 3 to 114), compared to 44 months (range, 27 to 168) following prosthetic replacement of the diaphragm. The only long survivor (114 months) after primary
Fig 1. Overall actuarial survival according to the Kaplan-Meier method.
Ann Thorac Surg 1999;68:2065– 8
ROCCO ET AL T3-DIAPHRAGM
2067
replacement of the diaphragm. In addition, early tumor grade showed a borderline favorable prognostic significance (p ⫽ 0.065). The only patient with preoperative evidence of cT3-diaphragm had a survival of only 3 months after en bloc resection of the lower lobe and the diaphragm. In this patient, recurrent disease was demonstrated by chest CT in the ipsilateral lung and in the site of the primary reconstruction of the diaphragm.
Comment
Fig 2. Actuarial survival curve for patients with T3N0 disease.
reconstruction of the diaphragm had only a partial infiltration of the muscle by the pulmonary tumor. The recurrent rates at 1, 2, and 5 years were significantly higher for patients undergoing primary reconstruction (67%, 78%, and, 78%, respectively) compared to patients subjected to prosthetic replacement of the diaphragm (0%, 20%, and 60%, respectively). Also, there was a trend towards better survival for patients with pathologic stage IIB (T3N0) compared to patients with pIIIA (T3N1-2) (p ⫽ 0.065). When the N factor was studied, no statistically significant difference was found between N0, and, N1/ N2 involvement found at thoracotomy. However, survival stratification showed that N0 and N1/N2 subsets had a median survival of 27 and 13 months, respectively, despite the partial invasion of the diaphragm seen in 2 of the 3 patients with N2 disease. Univariate analysis of the eleven T3N0 patients confirmed the favorable prognostic impact of the prosthetic Table 2. Impact on Survival by Univariate Analysis
Variable Age Sex Diaphragm invasion Pathologic stage Grade Histologic type Operation Reconstruction
Groups
Median Survival (mo)
⬍ 65 years ⬎ 65 years Male Female Partial Full IIB IIIA I/II III Epidermoid Adenocarcinoma Lobectomy ⬎ Lobectomy Primary Prosthetic
32 13 25 13 23 20.5 27 13 27 13.5 27 13.5 27 14 13 44
p Value 0.15 0.54 0.62 0.16 0.25 0.67 0.74 0.018
According to the most recent revision of the International Staging System for lung cancer, patients with T3N0 subset are classified as having stage IIB disease due to the comparable survival with T1-2N1 subsets [3]. The reported 5-year survival figures for pathologic stages IIB and IIIA are 39% and 23%, respectively [3], with no mention being made of the subgroup of patients with T3 invading the diaphragm (T3-diaphragm). More recently, the inclusion of T3-diaphragm into the T3N0 subset has been questioned due to a worse prognosis after surgical resection [1, 2], with a mean survival time of 53 weeks [1] and no 3-year survivors [2]. The most important limiting factor in the evaluation of lung cancer invading the diaphragm is its rare occurrence (less than 0.5% of all lung cancers) which makes it difficult to accrue a sufficient number of patients from a single institution [2]. Although the same quantitative limitations as above apply, our series includes almost twice the number of patients previously reported with this entity [1]. Diaphragmatic invasion is rarely suspected preoperatively [1]. When this is the case, alternative treatments should be instituted either in a neoadjuvant setting or as a definitive therapy [1]. At operation, en bloc resection of the diaphragm with the tumor should be attempted whenever possible [1]. As with chest wall invasion from T3 lung cancers [4], wide tumor-free margins should be obtained on the diaphragm. We now tend to excise up to 5 centimeters of macroscopically uninvolved diaphragm away from all tumor borders. This could preclude primary reconstruction of the muscle. However, when diaphragmatic reconstruction is done using prosthetic materials, a better survival may be expected for the increased likelihood of completeness of resection [5]. This could be explained by the partial extirpation, through wider muscular resection, of those diaphragmatic lymphatic channels conveying cancer cells to the abdominal lymph system [2, 6]. The reduced recurrence rates after prosthetic replacement of the diaphragm observed in our series further supports wide excision of the muscle en bloc with the pulmonary tumor. The observation of the direct liver involvement in 4 out of 6 patients with local recurrences (in 3 of these after primary reconstruction of the diaphragm) seems to indicate the propensity for T3-diaphragm to spread transdiaphragmatically. In the 2 patients with local recurrences
2068
ROCCO ET AL T3-DIAPHRAGM
following prosthetic replacement of the diaphragm, the disease-free survival exceeded 3 years. In one series of eight patients from the Memorial Sloan-Kettering Cancer Center [1], all but one patient underwent primary reconstruction of the diaphragm. Besides diaphragmatic infiltration from lung cancer, half of these patients also had N2 disease [1]. At follow-up, 1 patient was alive after 17 months, whereas 3 (37%) had died of conditions unrelated to the pulmonary tumor. The remaining 4 patients had died of disease progression within 6, 7, 26, and 42 months of operation. The reported mean survival for the eight patients was almost 13 months. Although definitive conclusions about the prognostic impact of resection of T3-diaphragm cannot be drawn from this series— especially in light of the 3 patients who died for causes other than their lung cancers—it seems reasonable to exclude from surgery as the primary treatment modality patients with concurrent N2 disease found at mediastinoscopy [1, 7]. In conclusion, we support the view according to which resection of T3N0-diaphragm carries a worse impact on prognosis than T3N0-chest wall invading lung cancer [2]. The 27% 5-year survival figure for T3N0-diaphragm observed in our series is significantly lower than the 39% reported by Mountain for stage IIB [3], the 40% figure reported by McCaughan and coworkers for tumors invading the chest wall [8], and the 54% 5-year survival for patients with T3N0 reported by Piehler and associates [9]. The difference becomes strikingly significant when the involvement of the parietal pleura only is considered [10]. Likewise, the 20% 5-year overall survival is also lower compared to the 23% reported for stage IIIA [3]. It should be noted that a 20% 5-year survival is also anticipated after resection of T4 tumors invading the carina in good surgical candidates [10]. In addition, 70% of our patients ran the risk to die from recurrent disease within 36 months of the operation— considerably higher than expected for stage IIB [3]. In order to establish whether a T3-diaphragm should be considered as a specific subtype of T3 tumors [11], or
Ann Thorac Surg 1999;68:2065– 8
even as a T4 [2], a multicentric study is needed to enroll an adequate number of patients and to confirm the operative indications, the extent of diaphragmatic resection and modalities for subsequent reconstruction, and, the survival benefit from resection [1, 2].
References 1. Weksler B, Bains M, Burt M, et al. Resection of lung cancer invading the diaphragm. J Thorac Cardiovasc Surg 1997;114: 500–1. 2. Inoue K, Sato M, Fujimura S, et al. Prognostic assessment of 1310 patients with non-small cell lung cancer who underwent complete resection from 1980 to 1993. J Thorac Cardiovasc Surg 1998;116:407–11. 3. Mountain CF. Revisions in the international system for staging lung cancer. Chest 1997;111:1710–7. 4. Pairolero PC. Management of lung cancer with chest wall invasion. Proceedings of the 31st Postgraduate Program of The Society of Thoracic Surgeons, New Orleans 1998; 24 –5. 5. Kao CC, Rand RP, Stridde BC, Marchioro TL. Techniques in the composite reconstruction of extensive thoracoabdominal tumor resections. J Am Coll Surg 1995;180:146 –9 6. Nomori H, Horio H, Kobayashi R, Morinaga S. Long survival after extrapleural pneumonectomy for pleural malignant mesothelioma with metastasis to infradiaphragmatic lymph node. Scand Cardiovasc J 1997;31:237–9. 7. Tatsumura T, Kasajima M, Koyama S, et al. Problems in combined resection of adjacent organ in lung cancer: significance of preoperative and intraoperative ultrasonic examination. Kyobu Geka 1992;45:1156– 62. 8. McCaughan BC, Martini N, Bains MS, McCormack P. Chest wall invasion of carcinoma of the lung: therapeutic and prognostic implications. J Thorac Cardiovasc Surg 1985;89: 836– 41. 9. Piehler JM, Pairolero PC, Weiland LH, Offord KP, Payne WS, Bernatz PE. Bronchogenic carcinoma with chest wall invasion: factors affecting survival following en bloc resection. Ann Thorac Surg 1982;34:684–91. 10. Martini N, Ginsberg RJ. Surgical management—lung cancer. In: Pearson FG, ed. Thoracic surgery. New York: ChurchillLivingstone, 1995:690 –705. 11. Detterbeck FC, Socinski MA. IIB or not IIB: the current question in staging non-small cell lung cancer. Chest 1997; 112:229–34.
Background. Diaphragmatic invasion from lung cancer (T3-diaphragm) is a rare occurrence reported to portend a poor prognosis. Methods. Fifteen patients with T3-diaphragm (14 males, 1 female; median age, 64 years) were surgically treated over a twenty-year period by en bloc resection (14 patients). One patient was only explored. Pathologic stage IIB (T3N0) was found in 11 patients. A partial infiltration of the diaphragm was observed in 3 patients, whereas full-depth invasion was found in 12. Diaphragmatic reconstruction was done primarily in 9 patients, and, by prosthetic material in 5.
Results. Two patients are still alive without evidence of disease at 88, and, 114 months from surgery. Overall median survival was 23 months (range, 3 to 168). The actuarial 5-year survival was 20%, when all patients were considered, and, 27%, for T3N0 patients. Univariate analysis showed that prosthetic replacement of the muscle (p ⴝ 0.018) was significantly related to survival. Conclusions. T3-diaphragm is best treated with en bloc resections with wide tumor-free margins and prosthetic replacement of the diaphragm. (Ann Thorac Surg 1999;68:2065– 8) © 1999 by The Society of Thoracic Surgeons
I
bilobectomy in 2, and, a pneumonectomy in 1. In the two bilobectomy patients, the middle lobe was resected due to the involvement of the major fissure by the pulmonary tumor. One patient underwent an exploratory thoracotomy only. Complete hilar and mediastinal lymph node dissection was done in all resected patients. Nine patients had an epidermoid histotype and 6 had an adenocarcinoma. Tumors were graded according to Broder’s classification as grade I (4 patients), II (5 patients), and, III (4 patients). Tumor grade was unknown in 2 patients. Pathologic stage was stage IIB in 11 patients, and, stage IIIA in 4. Eleven patients showed no nodal involvement (T3N0). On the other hand, N1 and N2 disease was found in 1 and 3 patients, respectively. Three patients had only a partial infiltration of the diaphragm whereas, in the remaining 12, full-depth extension of the pulmonary neoplasm through the muscular structure was observed. On pathologic examination, all patients had tumor-free margins on the diaphragm resected en bloc with the pulmonary neoplasm. The diaphragm was reconstructed primarily in 9 cases, and, using prosthetic materials in 5. Reconstruction was done using a Marlex mesh (Davol, Covington, GA) in 3 patients, and, a Gore-Tex (W.L. Gore, Flagstaff, AZ) patch in 2. No patients died within 30 days of operation. Minor bleeding and atrial fibrillation complicated the postoperative course of 2 patients. Postoperatively, four patients (numbers 1, 3, 8, and, 9) received an adjuvant treatment. The 11 patients with T3N0 disease were all males, all but one had full-depth muscle invasion, and, all under-
nvasion of the diaphragm is a rare occurrence demonstrated in less than 0.5% of the patients with locally advanced lung cancer [1, 2]. Few reports have addressed the issue of exactly determining the position of diaphragmatic infiltration into the staging system for lung cancer. The revised version of this staging system [3] still classifies diaphragmatic invasion as T3. However, limited series of surgically treated T3 lung cancers invading the diaphragm have included no 3-year survivors [2].
Material and Methods Between January, 1978, and December, 1997, 15 patients with lung cancer invading the diaphragm were surgically treated at our institutions (Table 11). This figure represents 0.4% of all resected patients for lung cancer in the same period. There were 14 males and 1 female, with a median age of 64 years (range, 34 to 79). Preoperative work-up, including computed tomography of the chest in all instances, had suggested the invasion of the diaphragm in only 1 patient (cT3). Mediastinoscopy was done in four patients (numbers 10, 12, 13, and, 14) and yielded negative results. One patient (number 1) underwent induction chemotherapy following video-assisted thoracoscopic demonstration of N2 disease. At operation, a lobectomy was done in 11 patients, a Accepted for publication May 27, 1999. Address reprint requests to Dr Rocco, Via Agricoltura, 20 23037, Tirano (Sondrio), Italy; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
0003-4975/99/$20.00 PII S0003-4975(99)01121-2
2066
ROCCO ET AL T3-DIAPHRAGM
Ann Thorac Surg 1999;68:2065– 8
Table 1. Clinical Features of 15 Patients With Lung Cancer Invading the Diaphragm (T3-Diaphragm) Depth of Histology Patient Age (y) Sex cT3 Invasion P Stage Grade Type 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
57 71 79 69 67 70 67 46 62 64 34 60 53 66 53
F M M M M M M M M M M M M M M
no no no yes no no no no no no no no no no no
partial full full full partial full partial full full full full full full full full
T3N2 T3N0 T3N1 T3N0 T3N0 T3N0 T3N2 T3N2 T3N0 T3N0 T3N0 T3N0 T3N0 T3N0 T3N0
III II I ... ... III III II I II I III I II II
Ad Sq Ad Ad Ad Sq Sq Sq Sq Sq Sq Ad Ad Sq Sq
Operation Pneumonect Lobect Exploration Lobect Bilobect Lobect Lobect Lobect Lobect Lobect Lobect Bilobect Lobect Lobect Lobect
Reconstruction Status primary primary ... primary primary primary primary primary primary prosthesis prosthesis primary prosthesis prosthesis prosthesis
DOD DOD DOD DOD Alive DOD DOUC DOD DOD DOUC Alive DOD DOD DOD DOD
Ad ⫽ adenocarcinoma; Bilobect ⫽ bilobectomy; DFI ⫽ disease-free interval; DOD ⫽ dead of disease; condition; Lobect ⫽ lobectomy; Pneumonect ⫽ pneumonectomy; Sq ⫽ squamous cell carcinoma.
went resection for cure as the primary treatment modality. In this group, the diaphragm was reconstructed primarily in 6 patients and using a prosthesis in 5. The only patient with a superficial infiltration of the diaphragm underwent limited circular resection and primary reconstruction of the diaphragm. Only 1 patient in this group (number 9) received radiotherapy in an adjuvant setting. Statistical analysis was done using Statview, version 4.5 (Abacus Concepts, Berkeley, CA). Actuarial survival was calculated by the Kaplan-Meier method and differences between groups were assessed by the Gehan-Wilcoxon test, as indicated. Statistical significance was set at a p value less than 0.05.
Results Upon completion of the study, the median follow-up is 23 months (range, 3 to 168 months), and, is complete in all patients. Two patients (13%) are alive without recurrent disease at 88, and, 114 months from operation. These two patients had T3N0 disease. For the other patients, the cause of death was tumor progression in 11 (73%), and, it was unrelated to the tumor in 2 (1 with T3N0 disease). Commonly involved sites by metastatic disease were the liver (8 patients), and, the lung (2 patients). Brain metastases were found in one patient only. Local recurrences in the ipsilateral hemithorax occurred in 6 patients, mostly in combination with liver metastases. Overall median survival was 23 months (range, 3 to 168). The actuarial 5-year survival was 20% when all patients were evaluated (Fig 1). The median disease-free interval was 18 months (range, 1 to 168). Overall recurrence rates at 1, 2, and, 5 years were 47%, 61%, and, 77%, respectively. When only the patients with T3N0 subset were consid-
Survival DFI Recurrence (mo) (mo) Site 13 9 13 3 114 3 23 6 27 168 88 14 37 27 44
7 3 10 1 114 1 23 3 18 168 88 12 36 24 40
Liver Liver/local ... Lung/local ... Liver/local ... Liver/brain Liver ... ... Liver/local Local Liver/lung Liver/local
DOUC ⫽ dead of unrelated
ered, the median survival was 27 months (range, 3 to 168), and the actuarial 5-year survival was 27% (Fig 2). In this group of patients, the median disease-free interval was 24 months (range, 1 to 168) and the recurrence rates at 1, 2, and, 5 years were 36%, 54%, and, 73%, respectively. Despite the limited number of patients available for evaluation, univariate analysis was conducted to assess the impact of each variable on survival (Table 24). The only variable predicting a favorable postsurgical outcome was the prosthetic replacement of the diaphragm. In fact, stratification of survival times according to the technique used for diaphragmatic reconstruction showed that median survival after primary reconstruction was 13 months (range, 3 to 114), compared to 44 months (range, 27 to 168) following prosthetic replacement of the diaphragm. The only long survivor (114 months) after primary
Fig 1. Overall actuarial survival according to the Kaplan-Meier method.
Ann Thorac Surg 1999;68:2065– 8
ROCCO ET AL T3-DIAPHRAGM
2067
replacement of the diaphragm. In addition, early tumor grade showed a borderline favorable prognostic significance (p ⫽ 0.065). The only patient with preoperative evidence of cT3-diaphragm had a survival of only 3 months after en bloc resection of the lower lobe and the diaphragm. In this patient, recurrent disease was demonstrated by chest CT in the ipsilateral lung and in the site of the primary reconstruction of the diaphragm.
Comment
Fig 2. Actuarial survival curve for patients with T3N0 disease.
reconstruction of the diaphragm had only a partial infiltration of the muscle by the pulmonary tumor. The recurrent rates at 1, 2, and 5 years were significantly higher for patients undergoing primary reconstruction (67%, 78%, and, 78%, respectively) compared to patients subjected to prosthetic replacement of the diaphragm (0%, 20%, and 60%, respectively). Also, there was a trend towards better survival for patients with pathologic stage IIB (T3N0) compared to patients with pIIIA (T3N1-2) (p ⫽ 0.065). When the N factor was studied, no statistically significant difference was found between N0, and, N1/ N2 involvement found at thoracotomy. However, survival stratification showed that N0 and N1/N2 subsets had a median survival of 27 and 13 months, respectively, despite the partial invasion of the diaphragm seen in 2 of the 3 patients with N2 disease. Univariate analysis of the eleven T3N0 patients confirmed the favorable prognostic impact of the prosthetic Table 2. Impact on Survival by Univariate Analysis
Variable Age Sex Diaphragm invasion Pathologic stage Grade Histologic type Operation Reconstruction
Groups
Median Survival (mo)
⬍ 65 years ⬎ 65 years Male Female Partial Full IIB IIIA I/II III Epidermoid Adenocarcinoma Lobectomy ⬎ Lobectomy Primary Prosthetic
32 13 25 13 23 20.5 27 13 27 13.5 27 13.5 27 14 13 44
p Value 0.15 0.54 0.62 0.16 0.25 0.67 0.74 0.018
According to the most recent revision of the International Staging System for lung cancer, patients with T3N0 subset are classified as having stage IIB disease due to the comparable survival with T1-2N1 subsets [3]. The reported 5-year survival figures for pathologic stages IIB and IIIA are 39% and 23%, respectively [3], with no mention being made of the subgroup of patients with T3 invading the diaphragm (T3-diaphragm). More recently, the inclusion of T3-diaphragm into the T3N0 subset has been questioned due to a worse prognosis after surgical resection [1, 2], with a mean survival time of 53 weeks [1] and no 3-year survivors [2]. The most important limiting factor in the evaluation of lung cancer invading the diaphragm is its rare occurrence (less than 0.5% of all lung cancers) which makes it difficult to accrue a sufficient number of patients from a single institution [2]. Although the same quantitative limitations as above apply, our series includes almost twice the number of patients previously reported with this entity [1]. Diaphragmatic invasion is rarely suspected preoperatively [1]. When this is the case, alternative treatments should be instituted either in a neoadjuvant setting or as a definitive therapy [1]. At operation, en bloc resection of the diaphragm with the tumor should be attempted whenever possible [1]. As with chest wall invasion from T3 lung cancers [4], wide tumor-free margins should be obtained on the diaphragm. We now tend to excise up to 5 centimeters of macroscopically uninvolved diaphragm away from all tumor borders. This could preclude primary reconstruction of the muscle. However, when diaphragmatic reconstruction is done using prosthetic materials, a better survival may be expected for the increased likelihood of completeness of resection [5]. This could be explained by the partial extirpation, through wider muscular resection, of those diaphragmatic lymphatic channels conveying cancer cells to the abdominal lymph system [2, 6]. The reduced recurrence rates after prosthetic replacement of the diaphragm observed in our series further supports wide excision of the muscle en bloc with the pulmonary tumor. The observation of the direct liver involvement in 4 out of 6 patients with local recurrences (in 3 of these after primary reconstruction of the diaphragm) seems to indicate the propensity for T3-diaphragm to spread transdiaphragmatically. In the 2 patients with local recurrences
2068
ROCCO ET AL T3-DIAPHRAGM
following prosthetic replacement of the diaphragm, the disease-free survival exceeded 3 years. In one series of eight patients from the Memorial Sloan-Kettering Cancer Center [1], all but one patient underwent primary reconstruction of the diaphragm. Besides diaphragmatic infiltration from lung cancer, half of these patients also had N2 disease [1]. At follow-up, 1 patient was alive after 17 months, whereas 3 (37%) had died of conditions unrelated to the pulmonary tumor. The remaining 4 patients had died of disease progression within 6, 7, 26, and 42 months of operation. The reported mean survival for the eight patients was almost 13 months. Although definitive conclusions about the prognostic impact of resection of T3-diaphragm cannot be drawn from this series— especially in light of the 3 patients who died for causes other than their lung cancers—it seems reasonable to exclude from surgery as the primary treatment modality patients with concurrent N2 disease found at mediastinoscopy [1, 7]. In conclusion, we support the view according to which resection of T3N0-diaphragm carries a worse impact on prognosis than T3N0-chest wall invading lung cancer [2]. The 27% 5-year survival figure for T3N0-diaphragm observed in our series is significantly lower than the 39% reported by Mountain for stage IIB [3], the 40% figure reported by McCaughan and coworkers for tumors invading the chest wall [8], and the 54% 5-year survival for patients with T3N0 reported by Piehler and associates [9]. The difference becomes strikingly significant when the involvement of the parietal pleura only is considered [10]. Likewise, the 20% 5-year overall survival is also lower compared to the 23% reported for stage IIIA [3]. It should be noted that a 20% 5-year survival is also anticipated after resection of T4 tumors invading the carina in good surgical candidates [10]. In addition, 70% of our patients ran the risk to die from recurrent disease within 36 months of the operation— considerably higher than expected for stage IIB [3]. In order to establish whether a T3-diaphragm should be considered as a specific subtype of T3 tumors [11], or
Ann Thorac Surg 1999;68:2065– 8
even as a T4 [2], a multicentric study is needed to enroll an adequate number of patients and to confirm the operative indications, the extent of diaphragmatic resection and modalities for subsequent reconstruction, and, the survival benefit from resection [1, 2].
References 1. Weksler B, Bains M, Burt M, et al. Resection of lung cancer invading the diaphragm. J Thorac Cardiovasc Surg 1997;114: 500–1. 2. Inoue K, Sato M, Fujimura S, et al. Prognostic assessment of 1310 patients with non-small cell lung cancer who underwent complete resection from 1980 to 1993. J Thorac Cardiovasc Surg 1998;116:407–11. 3. Mountain CF. Revisions in the international system for staging lung cancer. Chest 1997;111:1710–7. 4. Pairolero PC. Management of lung cancer with chest wall invasion. Proceedings of the 31st Postgraduate Program of The Society of Thoracic Surgeons, New Orleans 1998; 24 –5. 5. Kao CC, Rand RP, Stridde BC, Marchioro TL. Techniques in the composite reconstruction of extensive thoracoabdominal tumor resections. J Am Coll Surg 1995;180:146 –9 6. Nomori H, Horio H, Kobayashi R, Morinaga S. Long survival after extrapleural pneumonectomy for pleural malignant mesothelioma with metastasis to infradiaphragmatic lymph node. Scand Cardiovasc J 1997;31:237–9. 7. Tatsumura T, Kasajima M, Koyama S, et al. Problems in combined resection of adjacent organ in lung cancer: significance of preoperative and intraoperative ultrasonic examination. Kyobu Geka 1992;45:1156– 62. 8. McCaughan BC, Martini N, Bains MS, McCormack P. Chest wall invasion of carcinoma of the lung: therapeutic and prognostic implications. J Thorac Cardiovasc Surg 1985;89: 836– 41. 9. Piehler JM, Pairolero PC, Weiland LH, Offord KP, Payne WS, Bernatz PE. Bronchogenic carcinoma with chest wall invasion: factors affecting survival following en bloc resection. Ann Thorac Surg 1982;34:684–91. 10. Martini N, Ginsberg RJ. Surgical management—lung cancer. In: Pearson FG, ed. Thoracic surgery. New York: ChurchillLivingstone, 1995:690 –705. 11. Detterbeck FC, Socinski MA. IIB or not IIB: the current question in staging non-small cell lung cancer. Chest 1997; 112:229–34.