|
|
||||||||
a Department of Thoracic Surgery, Schillerhöhe Hospital, Gerlingen, Germany
b Department of Thoracic, Cardiac and Vascular Surgery, Tübingen University, Tübingen, Germany
c Department of Mathematics, Stuttgart University, Stuttgart, Germany
Accepted for publication March 25, 2009.
* Address correspondence to Dr Steger, Department of Thoracic Surgery, Schillerhöhe Hospital, Solitudestr 18, Gerlingen, 70839, Germany (Email: vrsteger{at}gmx.de).
| Abstract |
|---|
|
|
|---|
Methods: Included were all patients with potentially resectable, previously untreated stage III non–small cell lung cancer operated on between February 1999 and May 2006 in the General Thoracic Surgery Unit of the Schillerhöhe Clinic following the same neoadjuvant protocol. Treatment-related morbidity, recurrence, survival after R0 resection, and risk factors for survival (pN0 after trimodal therapy, downstaging of International Union Against Cancer stage, T downstaging, N downstaging, regression rate, and histologic type of tumor) were analyzed.
Results: From 107 patients with stage III non–small cell lung cancer, 55 patients with mediastinoscopy-positive N2 or N3 were eligible for this study. Forty patients (72%) had the effect of International Union Against Cancer downstaging. Treatment-related comorbidity was 54% with hospital and 120-day mortality of 3.6% and 5.4%, respectively. Overall mean survival (Kaplan-Meier) was 43 months (95% confidence interval, 35 to 52) with an estimated 5-year survival rate of 49%. In multivariate testing, International Union Against Cancer downstaging after trimodal therapy achieved a level of significance (p = 0.031), and patients with UICC-downstaging after trimodal therapy had a mean survival of 53 months (95% confidence interval, 44 to 63) with an estimated 5-year survival rate of 60%.
Conclusions: Neoadjuvant trimodal treatment for histologically proven N2 or N3 stage III non–small cell lung cancer is promising and can, like no other approach at present time, considerably improve 5-year survival rates up to 63% in selected patients.
| Introduction |
|---|
|
|
|---|
| Material and Methods |
|---|
|
|
|---|
Study Population
Included in this study were all patients with potentially resectable, previously untreated stage III NSCLC operated on between February 1999 and May 2006 in the General Thoracic Surgery Unit of the Schillerhöhe Clinic following the same neoadjuvant protocol. Treatment decisions were made on the basis of a recommendation by the institutional multidisciplinary tumor board. Inclusion criteria were (1) pathologically confirmed stage III NSCLC, (2) World Health Organization performance scale of 0 to 1, and (3) potentially resectable disease after chemoradiotherapy. No discrimination was made among single, multilevel, or bulky mediastinal effect. Patients with contralateral hilar mediastinal N3 or both contralateral bulky N3 and bulky N2 were excluded.
Staging
Oncologic staging included (1) chest computed tomography (CT) scan to determine primary tumor localization and extension, (2) cranial CT or magnetic resonance imaging, (3) 99m-technetium bone scan, and (4) abdominal CT or ultrasound to rule out distant metastasis as well as bronchoscopy to obtain cytologic or histologic specimen and to rule out endobronchial disease. All patients underwent standard cervical mediastinoscopy to confirm or rule out mediastinal nodal disease. Functional staging included (1) spirometry, (2) whole body plethysmography, and (3) electrocardiography. Since 2003 additional pulmonary diffusion capacity (DLCO) measurements were performed. The entire set of staging procedures except mediastinoscopy was repeated after chemoradiotherapy and before surgery. Patients were eligible for surgery if forced expiratory volume after 1 second (FEV1) was greater than 70% and DLCO greater than 60%. A perfusion scan of the lung was done to estimate postoperative FEV1 and DLCO if FEV1 was less than 70% or DLCO was less than 60%. Operation was declined if postoperative FEV1 or postoperative DLCO was less than 40%. Since 2005, spiroergometry was routinely applied to determine peak oxygen consumption, with peak oxygen consumption less than 10 mL · kg–1
· min–1 resulting in declining the operation, and a peak oxygen consumption ranging between 10 and 20 mL · kg–1
· min–1 limiting patient acceptance to those necessitating only lobectomies.
Neoadjuvant Protocol
The trimodality protocol consisted of four courses of polychemotherapy applying carboplatin to the area under the curve times two and paclitaxel (100 mg/m2) once weekly. Subsequently, accelerated hyperfractionated radiotherapy with two treatments per day (each 1.5 Gy) on a daily basis (five times per week) and a cumulative dose of 45 Gy (representing 50 to 90 Gy standard fractionations) was started on week 6 targeting tumor, mediastinum, and the supraclavicular region with a margin of 1.5 cm. Finally, two additional concurrent cycles of chemotherapy (carboplatin [area under the curve times two] and paclitaxel [50 mg/m2]) were administered (Fig 1). Prophylactic cerebral radiation (30 Gy cumulative dose) was offered to 30 patients. Operations were then carried out if R0 resection was deemed technically feasible.
|
Outcome Measures
For each patient, treatment-related comorbidities, relapse, survival, and risk factors for survival after resection were analyzed. The disease-free interval was defined as the time from R0 resection to histologically proven relapse or unambiguous radiologic findings with consecutive appropriate therapy. In case of no recurrence, the date of the last contact was used. The survival end point was the date of death defined as the time between the date of operation and the date of death. Follow-up was obtained from our own outpatient files, the attending family doctor's files, or the central death registry. Risk factors analyzed were (1) sterile mediastinal lymph nodes (ypN0), (2) formal T and N downstaging, (3) Junker regression score [20], (4) International Union Against Cancer (UICC) downstaging, (5) histologic type of primary tumor (adenocarcinoma, squamous cell carcinoma), and (6) pneumonectomy.
Statistical Analysis
All statistical calculations were performed with the SPSS software package for Windows (SPSS for Windows 13, SPSS Inc, Chicago, IL). The Kaplan-Meier method was used to analyze survival after operation. Risk factors for survival were first analyzed using univariate log-rank test. Factors with probability values less than or equal to 0.05 were considered to be significant. Multivariate testing using Cox regression was performed with factors reaching level of significance in univariate testing. Survival plots with 95% confidence intervals (CI) were created with R 2.7.0 Software (R Foundation for Statistical Computing, Vienna, Austria).
| Results |
|---|
|
|
|---|
|
|
Pathologic Response
A formal T downstaging was achieved in 44 patients (80%), whereas 7 patients (13%) maintained T stage and 4 patients (7%) had formal upstage. The latter 4 patients were initially graded T2 and finally classified as pathologic T3 after trimodal therapy (ypT3; n = 2) and pathologic T4 after trimodal therapy (ypT4; n = 2), respectively. All upstaged ypT4 cases were satellite tumors within the same lobe. Formal N downstaging was present in 38 patients (69%), whereas 17 patients (31%) remained in their initial N stage. N upstaging was not observed. A sterile mediastinum (pathologic N0 after trimodal therapy; ypN0) was present in 28 (51%) patients. UICC downstaging was achieved in 40 patients (72%), whereas 10 patients (18%) maintained UICC stage and 5 patients (9%) were upstaged. Two were upgraded from IIIa to IIIb and 3 were finally classified as stage IV with separate ipsilateral tumor in another lobe. In our cohort 19 patients (35%) were staged pathologic UICC stage 0 after trimodal therapy (ypUICC) stage 0, indicating absence of viable tumor tissue after chemoradiotherapy (Table 3). Applying the Junker regression score [20], 1 patient was regression score I (1.8%) representing no regression at all by chemoradiotherapy, 17 patients (31%) were classified regression score IIa, 16 patients (29%) were score IIb, indicating less than 10% vital tumor tissue, and 19 patients (35%) were regression score III, indicating no vital tumor tissue detectable (Table 4).
|
|
Survival
Follow-up was complete for all 55 patients (100%). Perioperative morbidity was 54.5% which included rhythm disturbances, heart attack, stump insufficiency, pneumonia, adult respiratory distress syndrome, acute lung injury, pulmonary embolism, respiratory insufficiency, atelectasis, prolonged air leakage, and pleural empyema (Table 5). The mean follow-up time was 30 months. Overall hospital and 120-day mortality was 3.6% (2 of 55 patients) and 5.4% (3 of 55 patients), respectively. The Kaplan-Meier method with 28 censored cases (49%) showed an overall estimated mean and median survival of 43 months (95% CI, 35 to 52) and 48 months (95% CI, 9 to 87), respectively (Fig 2). Persistent N2/N3 after trimodal therapy (ypN2/N3) resulted in a mean survival of 27 months (95% CI, 16 to 38), 19 months (95% CI, 8 to 30) median survival, and a 5-year-survival-rate of 16%. In cases of nominal T downstaging, regression score at least IIb, N downstaging, UICC downstaging, and ypN0, mean survival improved to 47 months (95% CI, 37 to 56), 49 months (95% CI, 38 to 59), 51 months (95% CI, 41 to 61), 53 months (95% CI, 44 to 62), and 54 months (95% CI, 42 to 66), respectively (Figs 3, 4).
With univariate log-rank test, risk factors for survival were UICC downstaging (p < 0.001), ypN0 (p = 0.014), and ypN downstaging (p = 0.004). In multivariate testing only UICC downstaging remained significant (p = 0.031; 95% CI, 0.075 to 0.883). In our analysis, formal T downstaging (p = 0.072) as well as regression scores of at least IIb (p = 0.072), histologic type of cancer (squamous cell carcinoma, p = 0.484; adenocarcinoma, p = 0.162), and pneumonectomy failed to be significant predictors (Table 6).
|
|
|
|
|
| Comment |
|---|
|
|
|---|
In 1999, an aggressive trimodal therapy protocol was implemented at the Schillerhöhe Clinic. For our protocol, we applied carboplatin and paclitaxel [26, 27], being established agents in treatment of advanced NSCLC and the known radiosensitizing effect of paclitaxel. Weekly administration has demonstrated similar efficiency with slightly fewer side effects and foremost better tolerability than traditional applications [28, 29]. To maximize radiosensitization and numbers of completed radiochemotherapy before surgery, sequential application of chemoradiotherapy was chosen. Accelerated hyperfractionated radiotherapy with 1.5 Gy twice daily up to 45 Gy was selected to minimize time delay by neoadjuvant protocol and enhance toxicity, representing 50 to 90 Gy if converted to standard fractionation of 2 Gy once daily. To identify treatment efficiency and impact on patient survival, we analyzed the subgroup of patients with mediastinoscopy-positive pN2 or pN3 NSCLC operated on between 1999 and 2006.
The clinical applicability and patient tolerability of our aggressive multimodal regimen is illustrated by the high rate of completed neoadjuvant protocols (72%). Neoadjuvant chemoradiotherapy took an average of 137 days from first diagnosis. While intentionally the operation should have followed chemoradiotherapy within 30 days, it took on average 40 days until surgery because of radiotherapy-induced esophagitis. This delay may account for the detected upstaging of 3 patients from stage III to stage IV owing to ipsilateral tumor in another lobe.
Neoadjuvant chemoradiotherapy resulted in major pathologic response rates with less than 10% vital tumor tissue in almost two thirds (n = 35, 63%) of our patients, with 19 patients (35%) having no detectable tumor tissue at time of surgery. Moreover, 28 patients (51%) achieved a sterile mediastinum (ypN0). Therefore, the applied regimen is superior to most reported less-aggressive neoadjuvant protocols [9, 30–33]. However, because of the chance of viable tumor residues after chemoradiotherapy, surgery is still mandatory to improve long-term patient survival [34–36].
After neoadjuvant chemoradiotherapy, we could not prevent pneumonectomy in 47% of patients despite the high pathologic response rate of 72% UICC downstaging and 80% T downstaging. This high proportion is well founded by numerous T4 tumors (44%) and cicatrization after radiation therapy of the hilar structures. Therefore, our treatment-related perioperative comorbidity is 54% and mirrors the impact of neoadjuvant therapy and surgery on the patients' general condition. However, meticulous postoperative care including intensive physiotherapeutic support, nursing, and precocious bronchoscopy in case of postoperative secretion retention resulted in a perioperative mortality rate of 3.6% (3.8% after pneumonectomy and 3.5% after lobectomy/bilobectomy). As a consequence, postoperative hospital stay was long (21 ± 9.7 and 23 ± 22 days after pneumonectomy and lobectomy or bilobectomy, respectively).
In our study, the overall mean survival was 43 months (calculated from 50% censored events and 28 patients still alive). Based on these data, the estimated overall 5-year survival rate for stage III NSCLC is 46%, which at present time surpasses any single-modality treatment protocol for comparable stage III NSCLC. In the literature, overall 5-year survival rates after neoadjuvant chemotherapy have been reported to be 35% but with less local control [23, 24, 31, 37]. Patients with persistent mediastinal effects (ypN2 or ypN3) showed less favorable results with a mean survival of 27 months and an estimated 5-year survival of 16%. Formal UICC downstaging emerged as a major risk factor for survival, corroborating both a cleared mediastinum and a responding primary lesion as a pillar in neoadjuvant therapy. For the more than two thirds (38 of 55) of patients with at least formal N downstaging, a remarkable improvement in survival can be expected: according to our data patients who achieved formal N downstaging, sterile mediastinum (ypN0), or UICC downstaging after trimodal therapy showed a mean survival up to 53 months and a 5-year survival rate up to 63%, which is at the upper level of known survival rates after histologically confirmed N2 or N3 NSCLC.
We are not able to predict pathologic response rate after neoadjuvant chemoradiotherapy in the individual patient. However, we did show that we are able to achieve complete pathologic response (yUICC0) in 35% of patients, resulting in superior survival rates compared with those obtained using multimodal protocols without surgery [35, 38, 39]. Therefore, patients with histologically confirmed pN2- or pN3-positive UICC stage III NSCLC should be offered multimodal therapies including surgery. Because of the complexity involved in patient selection, logistics of the multimodal concept, surgical approach, and postoperative care, these multimodal treatments will only be realizable in specialized thoracic surgery units.
This study is limited by its retrospective design and statistical power, but we do demonstrate that for selected pN2 or pN3 NSCLC this trimodal protocol is safe, promising, and superior to most other neoadjuvant approaches with chemotherapy alone [9, 30, 31, 36] or multimodal protocols without surgery.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
V. Steger, W. Spengler, J. Hetzel, S. Veit, T. Walker, M. Mustafi, G. Friedel, and T. Walles Pneumonectomy: calculable or non-tolerable risk factor in trimodal therapy for Stage III non-small-cell lung cancer? Eur J Cardiothorac Surg, January 10, 2012; (2012) ezr160v1. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Friedel and T. Walles Reply to R. Stupp et al J. Clin. Oncol., February 10, 2011; 29(5): e120 - e120. [Full Text] [PDF] |
||||
![]() |
L.- R. He, H.- Y. Zhao, B.- K. Li, L.- J. Zhang, M.- Z. Liu, H.- F. Kung, X.- Y. Guan, X.- W. Bian, Y.- X. Zeng, and D. Xie Overexpression of AIB1 negatively affects survival of surgically resected non-small-cell lung cancer patients Ann. Onc., August 1, 2010; 21(8): 1675 - 1681. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Friedel, W. Budach, J. Dippon, W. Spengler, S. M. Eschmann, C. Pfannenberg, F. Al-Kamash, T. Walles, H. Aebert, T. Kyriss, et al. Phase II Trial of a Trimodality Regimen for Stage III Non-Small-Cell Lung Cancer Using Chemotherapy As Induction Treatment With Concurrent Hyperfractionated Chemoradiation With Carboplatin and Paclitaxel Followed by Subsequent Resection: A Single-Center Study J. Clin. Oncol., February 20, 2010; 28(6): 942 - 948. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Hampel, I. Dally, T. Walles, V. Steger, S. Veit, T. Kyriss, and G. Friedel Impact of neo-adjuvant radiochemotherapy on bronchial tissue viability Eur J Cardiothorac Surg, February 1, 2010; 37(2): 461 - 466. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Walles, M. Hampel, I. Dally, and G. Friedel eComment: Defective bronchial tissue homeostasis following neoadjuvant therapy for lung cancer Interact CardioVasc Thorac Surg, October 1, 2009; 9(4): 575 - 575. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |