|
|
||||||||
Swedish Cancer Institute, Seattle, Washington
Accepted for publication May 13, 2009.
* Address correspondence to Dr Vallières, Swedish Cancer Institute, 850-1101 Madison, Seattle, WA (Email: eric.vallieres{at}swedish.org).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
| Abstract |
|---|
|
|
|---|
Methods: The study comprised a retrospective cohort of 46 patients treated with induction chemotherapy, followed by EPP, during a 10-year period. Of these, 24 completed adjuvant external beam radiotherapy (EBRT), and 14 had intensity-modulated radiotherapy (IMRT).
Results: Mean follow-up was 20.6 months (range, 0.5 to 75 months). Operative mortality after EPP was 4.3% (n = 2). Pathologic stage was p0, 4.3%; pII, 23.9%; pIII, 56.5%; and pIV, 15.2%. Median overall survival was 24 months. On univariate analysis and Cox proportional hazards model, only nodal metastases (hazard ratio, 3.7; 95% confidence interval, 1.6 to 8.7; p = 0.002) was a significant predictor of survival. First site of recurrence was local in 12, the contralateral chest in 5, abdominal in 8, and distant in 5. The incidence of local recurrence was 14.3% with IMRT vs 41.7% with EBRT (p = 0.03). The time to local recurrence with the use of IMRT was 12 months vs 7 for EBRT (p = 0.19).
Conclusions: Induction chemotherapy, followed by EPP and adjuvant radiotherapy for selected patients with mesothelioma, is safe, with acceptable operative mortality. Adjuvant IMRT may be more effective in terms of local control than EBRT.
| Introduction |
|---|
|
|
|---|
We reviewed the clinical outcomes of treating MPM using this sequence of induction chemotherapy, followed by EPP and adjuvant radiotherapy, and analyzed prognostic variables for survival.
| Material and Methods |
|---|
|
|
|---|
Patients
This was a retrospective review of a consecutive cohort of patients diagnosed with MPM and treated with induction chemotherapy, followed by EPP, at the University of Washington Medical Center and at the Swedish Cancer Institute between February 1997 and June 2008.
After diagnosis of MPM, patients were initially staged with computed tomography and positron emission tomography. Inclusion criteria for consideration of trimodality therapy were histologic confirmation of MPM confined to the ipsilateral hemithorax, Eastern Cooperative Oncology Group performance status 0 to 1, predicted postoperative forced expiratory volume 1 second of 40% or more, normal serum creatinine level, and no major systemic comorbidities precluding chemotherapy, operation, or radiotherapy. Sarcomatoid or mixed histologic subtypes were not contraindications to trimodality therapy.
Induction Chemotherapy
The choice of chemotherapy agent, dose, and schedule were at the discretion of the treating medical oncologist. During the last 4 years, the chemotherapy regimen has been standardized to cisplatin and pemetrexed (Table 1).
|
Surgical Resection
In addition to radiologic restaging after completion of chemotherapy, pulmonary function tests, a ventilation-perfusion scan, and a stress echocardiogram were performed. Within 3 to 5 weeks after completion of chemotherapy, all patients deemed fit to withstand EPP underwent mediastinoscopy. Patients with evidence of a pleural effusion also underwent a single-port pleuroscopy and talc pleurodesis under the same anesthetic. Histologically confirmed malignant nodes identified at mediastinoscopy were considered an absolute contraindication to EPP, which was typically scheduled 1 week later.
EPP was performed by a posterolateral thoracotomy through the bed of the excised sixth rib. En bloc resection of the parietal pleura, lung, ipsilateral pericardium, and hemidiaphragm was performed. The diaphragm was reconstructed using a 2-mm-thick polytetrafluoroethylene patch, and the pericardium with a polyglactin 910 Vicryl mesh (Ethicon Inc, Somerville, NJ).
Radiotherapy
After recovery from the operation, usually at 6 to 8 weeks, postoperative radiotherapy was initiated. Areas with residual gross disease at resection were marked with metallic clips at the time of thoracotomy and noted in the operative dictation to guide radiation treatment. An extensive map of the sampled parietal margins was also obtained according to pathology protocol and reviewed with the radiation oncologist. During the treatment period, radiation was delivered in two modalities: conventional external beam radiotherapy (EBRT) or intensity-modulated radiotherapy (IMRT). Choice of radiation type was by preference of the treating radiation oncologist and was not based on stage, results of the operation, or patient factors.
EBRT was delivered to a median total dose of 30 Gy (range, 18 to 70 Gy) in a daily fractionated dose of 1.8 to 2 Gy delivered through anterior and posterior fields. Scars and areas with gross or microscopic positive resection margins, as noted by the surgeon at the time of operation or by the pathologist on examination of the specimen, were treated with a supplemental dose of 9 to 18 Gy.
During the period of observation, IMRT was delivered in a median dose of 50.4 Gy (range, 49 to 56 Gy) for 6 weeks using conventional fractionation. Surgical scars received a mean supplemental dose of 24 Gy using direct electrons and customized bolus. This dose was used to account for the inherent uncertainty of skin dose from the IMRT treatment plan. By exploiting the sophisticated IMRT treatment planning tools and the inherent heterogeneity in dose, the areas of gross or microscopic residual disease received higher dose up to a total of 60 Gy.
We measured overall survival from the time of diagnosis to date of the last follow-up appointment or death. Postoperative mortality included all deaths within 30 days of EPP or within the same hospitalization. Positive margins were subdivided into macroscopic (R2, gross disease left in hemithorax at time of operation) or microscopic (R1, microscopic involvement of inked surgical resection margins). Disease recurrence was confirmed with histologic or cytologic sampling when appropriate. Otherwise, a clinical diagnosis of recurrence was made using imaging.
Statistics
The primary end points were overall and disease-free survival. We performed univariate analysis of demographic and pathologic variables to determine the effect on survival, with the log-rank test to determine significant differences (defined as p < 0.05) in survival curves. Significant variables on univariate analysis were included in a Cox multivariate proportional hazards model with stepwise backward elimination to determine independent predictors for survival. Statistical analysis was performed using SPSS 16 software (SPSS Inc, Chicago, IL).
| Results |
|---|
|
|
|---|
Of the 55 patients who underwent mediastinoscopy, 2 had positive nodes and were deemed inoperable. Of 53 patients who underwent exploration for resection, 7 were found to be unresectable. Reasons for not completing EPP included presence of multistation N2-positive nodes beyond the field of mediastinoscopy in 3 and extensive tumor invasion of the esophagus in 1, aorta in 2, or full-thickness chest wall infiltration in 3. Thus, 46 patients underwent EPP after induction chemotherapy. Demographic data are reported in Table 2. Mean duration of follow-up was 20.6 months (range, 0.5 to 75 months).
|
|
|
|
|
Nodal and R2 status was evaluated by multivariate regression model analysis. Nodal status (hazard ratio [(HR)], 3.7; 95% CI, 1.6 to 8.7; p = 0.002) was an independent predictor for survival. R2 status was eliminated from the regression model (HR, 1.09; 95% CI, 0.4 to 2.7; p = 0.85).
In the cohort of 38 patients who completed chemotherapy, surgical resection, and radiotherapy, overall median survival was 25 months. Univariate log-rank tests showed nodal positivity (median survival 17 vs 37 months if node-negative; p = 0.0005) and R2 margin (median survival 17 vs 33 months if no R2 margin; p = 0.04) were significant predictors of worse survival. On Cox multivariate regression, N stage was the only independent prognostic variable for survival (HR, 4.0; 95% CI, 1.5 to 10.8; p = 0.006). R2 margin was eliminated from the model (HR, 1.1; 95% CI, 0.4 to 3.2; p = 0.9).
Adjuvant radiotherapy with curative intent was administered to 38 patients. Six patients did not undergo further adjuvant treatment (not including the 2 who died postoperatively). One patient was lost to follow-up, and early postoperative empyema developed in 1 patient and radiotherapy was deferred. Brain metastases were found in 1 patient on the day of radiation simulation. No additional information from the medical record was available for the other 3 patients, who were from out of state.
Radiotherapy consisted of EBRT in 24 patients and IMRT in 14. Of the EBRT group, clinically significant anorexia or dehydration, or both, developed in 14 patients and mild esophagitis in 1. In the IMRT group, significant dehydration developed in 4 patients, mild esophagitis in 2, and radiation pneumonitis in 2. There were no unplanned treatment interruptions. At 8 months after completion of IMRT, a late empyema developed in 1 patient that required surgical drainage and serial packing.
Recurrent disease developed in 24 of the 38 patients (63%) who completed adjuvant radiation. Patterns of failure are outlined in Table 5. Half of the recurrences occurred within the ipsilateral hemithorax. Of the 12 local recurrences, 5 had associated disease relapse outside the ipsilateral hemithorax; therefore, the incidence of isolated local recurrence for the patients who completed trimodality therapy was 7 of 38 (18%). The IMRT group had a lower incidence of local recurrence of 14.3% compared with 41.7% for EBRT (p = 0.03). The isolated local recurrences were in 5 R1 patients and 2 R2 patients who had EBRT after surgical resection; no isolated local recurrences were noted in the patients who received IMRT.
|
| Comment |
|---|
|
|
|---|
Early on, our recommended induction chemotherapy was a combination of cisplatin, methotrexate, and vinblastine based on our own institutional experience in patients with advanced disease [7]. Since the approval in 2004 of pemetrexed in combination with cisplatin or carboplatin in the treatment of patients with higher-stage disease, the combination of cisplatin and pemetrexed has been our preferred and current induction regimen [8].
Data regarding the initiation of chemotherapy were often missing because most patients received their treatment at other centers. We therefore used time of histologic diagnosis as time 0 for calculating survival and recurrence time. Chemotherapy start dates were available in 21 of 46 patients, and the median time from diagnosis to start of chemotherapy for these patients was 2 months (range, 0 to 5 months).
EPP is a technically demanding procedure, with a reported operative mortality rate of up to 31% in earlier reports [9]. EPP after induction chemotherapy has the potential additional operative challenges of dissecting through inflamed, obliterated tissue planes and immunosuppression [5, 10]. In addition, many patients received talc pleurodesis at the time of their diagnosis. By the time these individuals recovered from induction chemotherapy, up to 20 weeks had elapsed and the fibrotic reaction from the pleurodesis made it virtually impossible to differentiate scar from malignant involvement of the extrapleural plane.
Our preference is single-port thoracoscopy at the time of diagnosis. We also recommend delaying any chemical pleurodesis until the postchemotherapy mediastinoscopy assessment. There less extrapleural scarring at thoracotomy 1 week later, and the acute inflammatory response from the recent pleurodesis often facilitates the extrapleural dissection (personal communication from Dr D. L. Miller). Despite these concerns, our operative mortality rate of 4% compares favorably among other contemporary series of trimodality treatment for mesothelioma (Table 6). [2–6, 11–13]
|
Disease recurrence has been an ongoing problem after surgical resection for MPM. To improve local control, adjuvant radiotherapy has been the mainstay of therapy after EPP. One advantage of EPP over pleurectomy/decortication is the ability to maximize the delivery of postoperative radiotherapy. Rusch and colleagues [14] reported a 13% locoregional recurrence rate using high-dose hemithoracic EBRT after EPP. A more recent update from the same group, however, reported a 37% local failure using the same protocol [15]. Difficulties with EBRT include the need for a high radiation dose to achieve therapeutic efficacy, with concomitant locoregional toxicity.
Although EBRT was used in our early experience, our current preference is to administer IMRT after EPP. Compared with EBRT, IMRT can specifically target the irregular conformations of the hemithorax, which maximizes the tumoricidal effect and spares adjacent organs from exposure. Postoperative IMRT was well tolerated in our cohort, with no treatment interruptions. Other institutions have reported a significant incidence of major pulmonary toxicity and an incidence of up to 46% of fatal pneumonitis after IMRT in post-EPP patients [16–18]. Radiation pneumonitis did developed in 2 patients in our study, but treatment was not interrupted. Nevertheless, it is imperative to carefully plan the dose of radiation and minimize volume of exposure to the contralateral lung. In the aforementioned studies, it was noted that V20, or percentage of lung volume receiving 20 Gy or more, was a risk factor for fatal pneumonitis. The Allen report [18] described a median V20 of 17.6% and 10.9% in the patients who did and did not develop post-IMRT pneumonitis, respectively. The median V20 in our cohort of IMRT patients was 7%. Thus, with careful control of radiation exposure to the adjacent lung, IMRT can be well tolerated in the post-EPP setting.
We noted a lower incidence in local disease recurrence as well as a modest trend toward longer time to local recurrence in the IMRT cohort compared with the EBRT group. Our local recurrence of 14% with adjuvant IMRT after EPP is comparable with other reports [19]. As other series have shown, nodal status is an important predictor of survival [2, 12–14]. We believe that preoperative mediastinoscopy should be mandated in all MPM patients and that superior mediastinal nodal involvement is a contraindication for EPP. Despite this, approximately 44% of patients were node-positive at resection.
Nodal involvement was a poor prognostic finding, with a median survival of 11 months, compared with 34 months if node-negative. Other studies have noted the inherent difficulty of accurately staging mediastinal nodes with cervical mediastinoscopy in MPM due to different patterns of lymphatic drainage, inaccessible nodes, and heterogeneity of involvement [11]. Unfortunately, access to the mediastinum is a relatively late step of EPP. We pursue resection even if we document minimal nodal involvement at thoracotomy unless other nonfavorable T factors are identified. We do, however, considered multistation mediastinal nodal involvement—identified early during the dissection—as a contraindication to completing EPP.
Sugarbaker and colleagues [2] identified resection margin positivity as an independent predictor of worse survival, citing a median survival of 15 months in margin-positive patients compared with 23 months if margins were clear. Margin status is an important prognostic factor in other series [20]. In our study, there appears to be delineation between macroscopically (R2) and microscopically (R1) positive margins.
Although margin status was an independent predictor for survival in our study, some points should be noted. The prognosis of patients with R2 margins was poor, with median survival of 11 months. Patients with R1 margins did not fare as badly, however, with a 25-month median survival. Obviously, negative margin status is the goal for any curative resection, but this may not be attainable for most patients with mesothelioma. Thus, although grossly positive margins should be avoided, microscopic disease does not necessarily predict a uniformly poor prognosis after trimodality therapy.
Our study has inherent limitations. Our initial encounter with many referred patients was after they had already started on or completed chemotherapy at other institutions. This pattern of referral made it impossible for us to determine the true denominator of patients who may have started chemotherapy in the hope of making it to resection but who were never referred. Obviously, the cohort of patients we present is highly selective; patients who had progressed on chemotherapy or who may have had serious treatment-related complications were likely never referred. Although we cannot provide a concrete number for the number of MPM patients turned down for resection, for every patient deemed acceptable for trimodality therapy, about another 4 MPM patients were turned down because of advanced disease or comorbidities. It is possible that the most aggressive malignancies were excluded from our cohort; however, 71% of patients were stage III or IV. Thus, although patient selection was important, we did not limit our treatment to patients with very early-stage disease.
In the absence of a control group, some would claim that the apparent survival benefit of treatment merely reflects patient selection and the variable natural history of the disease. Untreated MPM, however, typically is a rapidly fatal disease. Merritt and colleagues [21] reviewed 101 MPM patients treated palliatively who had a median overall survival of 7 months. The question of whether radical surgical resection offers any benefit over nonoperative or less aggressive procedures can only be answered in the context of a randomized trial. The results of the British Mesothelioma and Radical Surgery trial, randomizing patients to EPP or not after chemotherapy may provide some answers to these questions [22].
MPM is a challenging disease to treat. Our data show that an aggressive trimodality approach of induction chemotherapy, EPP, and adjuvant radiotherapy is feasible and can be performed with acceptable morbidity and low operative mortality in selected patients.
| Discussion |
|---|
|
|
|---|
Secondly, do you have any PET [positron emission tomography] data on the changes that occur of the max SUV [standardized uptake values] before and after chemotherapy? If a patient's SUV dropped significantly, is that an indication for improved survival?
Thirdly, if you do a mediastinoscopy and you find N2 disease, do you give definitive chemotherapy or neoadjuvant chemotherapy and proceed with resection if there is no activity within the mediastinum?
DR BUDUHAN: In terms of whether or not we gave chemotherapy in patients with minimal tumor burden, our approach was to treat all of these patients with chemotherapy as long as they could withstand it.
The question about the PET, this was very hard to elucidate because we had PET scans from different centers. Some reported SUV changes and some did not. In addition, because some patients got talc pleurodesis, this also influenced the SUV uptake. So we tried to quantify it as best we could. It did not come up as significant on univariate analysis, but we had limited data. And the third question?
DR ROBERT CERFOLIO (Birmingham, AL): Occult N2, surprise N2.
DR BUDUHAN: We would typically perform the mediastinoscopy after completion of the chemotherapy. So we did not do a mediastinoscopy up front. All patients that we saw had chemotherapy first and then were evaluated.
DR DAVID C. RICE (Houston, TX): I really enjoyed that series, and congratulations for having such a low mortality with a difficult patient subset.
My question, and I think you touched on it, is what were your indications for EPP in patients who had undergone induction chemotherapy? Do you know the number of patients who may been prescribed an induction regimen but then either progressed or for some other performance status reason maybe didn't get surgery, and therefore would not have been included your survival analysis. I think before we can say that this is a safe and feasible technique, we would need to know what the denominator population is.
DR BUDUHAN: First of all, after completion of their chemotherapy, we operated on patients even if they had high tumor burden, and even if they progressed in the ipsilateral hemithorax. So as long as they did not develop extrathoracic disease, metastases, as long as they had preserved performance status, and disease remained confined to the hemithorax and were deemed operable, then we offered it to them.
DR MARK KRASNA (Towson, MD): I congratulate you on using the mediastinoscopy information in deciding who should get surgery. I think that's really an important step forward.
I have one question about the data on intensity-modulated radiation therapy [IMRT] results. This is now the third large series, including the one from Anderson and the Brigham. I believe it was the Boston series that actually did have a higher perioperative morbidity in the patients who had the radiation. How can you explain that?
My last question is a general question. How did you treat the bronchial stump? Our group for non-small cell lung cancer has routinely now been using the serratus anterior muscle on our pneumonectomy patients post high-dose chemorads. A very large series recently from Europe also talked about using muscle flaps in patients, specifically in extrapleural pneumonectomies, postchemotherapy. What have you guys been doing? I enjoyed your paper.
DR BUDUHAN: In answer to your question about the bronchial stump, if we found local tissue, pericardial fat or something that we could cover up, we would try to. We did not routinely harvest a muscle flap.
DR CERFOLIO: Aren't you taking the pericardial fat out? I mean isn't that gone?
DR VALLIERES: Basically if you can still find anything in that area on the right, you use it. On the left it doesn't matter. On the right, we usually find enough tissue to bring over the stump and we try to cover it as much as we can. Often, we put a couple of superficial sutures in the esophageal wall, bring it across, and tack it to the back of the pericardium which sits right in front of the right main stump. Cerf, you are right, there's not much to cover with but usually there is enough. We have not brought down parietal muscle to cover our stumps.
DR MILLER: An azygous vein patch does quite well.
DR VALLIERES: Yes, Dan is right.
DR BUDUHAN: The question about the IMRT, we are aware of the almost 50% incidence of fatal pneumonitis in the Brigham group. We did note a higher mean V20, or the volume of lung exposed to greater than 20 Gy was up to 15% in their series, and in our series, although I didn't have the time to report this, it was much lower, in the range of 7%. So with minimization of exposure of the contralateral lung, I think you can do IMRT with minimal morbidity and good local control.
DR KRASNA: I have just a quick comment. Given the paper we heard from Cerfolio earlier, perhaps doing pulmonary function tests postinduction, preresection, and then looking at that data, I think that is going to help predict those patients.
DR MICHAEL J. LIPTAY (Chicago, IL): Others have shown that biphasic tumors and sarcomatous tumors have an abysmal prognosis with aggressive treatment. Your numbers were small. Over 80% were epithelial tumors. Did you find the same results with the biphasic and the sarcomatous?
DR BUDUHAN: Thanks for your question. Because our numbers were so small, we couldn't find a significant difference overall, but, the nature of our referral was that the majority of patients that we had had epithelioid histology, so we couldn't discern a difference in our series.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. Lindenmann, V. Matzi, N. Neuboeck, U. Anegg, A. Maier, J. Smolle, and F. M. Smolle-Juettner Multimodal therapy of malignant pleural mesothelioma: is the replacement of radical surgery imminent? Interact CardioVasc Thorac Surg, March 1, 2013; 16(3): 237 - 243. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. PASELLO, G. MARULLI, V. POLO, C. BREDA, L. BONANNO, L. LOREGGIAN, F. REA, and A. FAVARETTO Pemetrexed plus Carboplatin or Cisplatin as Neoadjuvant Treatment of Operable Malignant Pleural Mesothelioma (MPM) Anticancer Res, December 1, 2012; 32(12): 5393 - 5399. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. S. Farivar, B. E. Louie, R. W. Aye, M. P. Horton, and E. Vallieres Extrapleural Pneumonectomy for Primary Pleural Mullerian Tumor in a Young Woman Ann. Thorac. Surg., January 1, 2012; 93(1): e1 - e2. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. YOSHIDA, T. EBARA, Y. SATO, T. KAMINUMA, T. TAKAHASHI, T. ASAO, and T. NAKANO Interaction of Radiation and Pemetrexed on a Human Malignant Mesothelioma Cell Line In Vitro Anticancer Res, September 1, 2011; 31(9): 2847 - 2851. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Price What Is the Role of Radiotherapy in Malignant Pleural Mesothelioma? Oncologist, March 1, 2011; 16(3): 359 - 365. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. M. Rudd Malignant mesothelioma Br. Med. Bull., March 1, 2010; 93(1): 105 - 123. [Abstract] [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 |