| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Click here to read this article as a CME activity
Department of Thoracic Surgery, Albert-Ludwigs-University Freiburg, Hugstetter Str. 55, 79106 Freiburg, Germany
Received 31 July 2007; received in revised form 27 November 2007; accepted 10 December 2007.
* Corresponding author. Tel.: +49 761 270 2457; fax: +49 761 270 2499. (Email: wulf.sienel{at}uniklinik-freiburg.de).
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
Key Words: Lung neoplasm • Surgery • Prognosis • Survival rate • Local recurrence • Sublobar resection
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
Recently, Okada et al. [3] revealed that wedge resections with lymph node sampling yield a disease-free and overall survival equivalent to segmentectomies and lobectomies in patients with stage IA NSCLC 
2 cm who could even tolerate a lobectomy. In contrast to segmentectomy, wedge resections with lymph node sampling can be easily performed using video-assisted thoracic surgical (VATS) techniques. Therefore, it would be of enormous importance for the future of thoracic surgery if Okada's results [3] could be confirmed in a prospective randomised trial as suggested previously [3].
However, controversy remains between the observations of Okada et al. [3] and the LCSG [2], because the LCSG reported an increased local recurrence rate of 8.6% following wedge resection compared to 4.4% following segmentectomy and Okada et al. observed no recurrence following wedge resection and 1.3% recurrences following segmentectomy. Because of this controversy, the present study has been performed to compare cancer-related survival, local recurrence rate and occurrence of metastases following wedge resection and segmentectomy in stage IA NSCLC.
|
2. Patients and methods |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
To allow comparison between pulmonary functions, forced expiratory volume in 1 s (FEV1) was expressed in percent (%) of estimated values that had been calculated according to body weight, height and age. Cranial computed tomography or bone scintigraphy was used selectively to investigate specific preoperative symptoms potentially related to metastatic disease. Mediastinoscopy was not used in preoperative evaluation of stage IA patients. Only patients without suspicion of metastases (cM0) were entered into the study. Patients were excluded if they had synchronous or secondary primary lung cancer, had had a previous malignant tumour at any site, or had received preoperative or postoperative adjuvant therapy. In patients with preoperatively known histology, a primary thoracotomy was performed for segmentectomy and for wedge resection of NSCLC. To prevent bias by the approach, patients with VATS resections were not eligible for this study except when a conversion to thoracotomy had been undertaken (n = 2).
Postoperative pathologic staging was done after the resected specimen was reviewed histologically and all dissected lymph nodes were assessed. The tumours were classified according to the International Union Against Cancer (UICC) TNM-classification [5]. Furthermore, tumour size, tumour type, grading and histologic clearance at bronchial, parenchymal and vascular resection margins were recorded. The surgical and pathologic reports of all patients were carefully reviewed to ensure that no residual tumour was left behind, and that no involved regional nodes (N1 and N2) were present.
Follow-up studies included physical examination, chest X-ray and blood tests in a 3-month interval and an additional thoracic CT scan, abdominal ultrasound and bronchoscopy in a 6-month interval. Close follow-up was documented by contacting family practitioners with questionnaires concerning local relapse, distant metastasis and death. If possible, a relapse was confirmed at our institution and the patient was admitted for subsequent therapy. Local recurrence could be treated by reoperation (extended segmentectomy) in only four patients because pulmonary function was too limited in the other cases. All patients with local recurrence who were not reoperated received palliative chemotherapy. Local recurrence was distinguished from secondary primary lung cancer and was defined as a tumour within the same lung or in the ipsilateral mediastinum. Criteria for synchronous or secondary primary lung cancer were a different histologic type or an origin from carcinoma in situ, as suggested previously [6]. The median follow-up duration was 45 months (range 3–192 months).
2.2 Operations
The approach to wedge resection and anatomical segmentectomy in this clinical series were standard posterolateral or anterolateral muscle-sparing thoracotomy being performed as described previously [7]. Segmentectomy involved dissecting out the hilar structures and securing the branches of the pulmonary artery, pulmonary vein, and segmental bronchus individually as described previously [8]. For segmentectomy, the segmental plane was identified by the use of CPAP after the bronchus was dissected. Subsequently, the segmental plane was prepared by fingering out and blunt dissection. Wedge resections were performed by excision of the tumour using electrocautery or scissors. The resulting parenchymal defect was closed by resorbable 4/0 running sutures. Intraoperative frozen section analysis was used in both resection types to assess clearance of resection margins and a re-resection was undertaken if necessary. Segmentectomies were followed by systematic lymphadenectomy [9,10] and wedge resections by lymph node sampling as described previously [9]. Postoperative care was similar following segmentectomy and wedge resection. The patients were admitted overnight to the surgical intensive care unit for observation. All patients were extubated in the operation room immediately after reversal of anaesthesia.
2.3 Statistical analyses
Statistical analysis was performed using SPSS software version 15.0 for PC (SPSS Inc., Chicago, USA). Continuous variables such as age, FEV1 in % of estimated value, tumour size, number of dissected lymph nodes and follow-up duration are expressed as mean ± standard deviation. Differences between the groups of segmentectomies and wedge resections according to continuous variables were assessed by Student's t-test. To analyse a possible correlation between the two groups and categorical variables such as gender, ASA-score, tumour histology, grading and complication rate, the two-tailed Pearson's chi square test or Fisher's exact test in frequencies 5 were used.
For analysis of follow-up data, 5-year survival curves were calculated using the Kaplan–Meier method and were compared by log-rank statistics. For survival analyses, the primary end-point was cancer-related death. The survival time was measured from the date of surgery to death and data of patients who were still alive at the end of the study were censored. The joint effects of other prognostically relevant variables were further examined using the Cox proportional hazard model. The respective covariables were entered step-wise forward into the model to assess possible independence of the prognostic value of the resection type. The 0.05 level of significance was used for entering or removing a covariable from this model. For analyses of relapse patterns, end-points were local recurrence or distant metastasis. Recurrence-free and metastasis-free time was measured from the date of surgery to the first relevant event and data of patients without relapse at the end of the study were censored. The threshold for statistical significance was p < 0.05.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
|
The number of dissected lymph nodes was significantly higher in segmentectomies (12 ± 6) than in wedge resections (6 ± 3; p = 0.0001; t-test; Table 1). Complications occurred in 23% of segmentectomies and in 10% of wedge resections, but a significant difference between the two groups was not apparent (p = 0.15; chi square test). The 30-day perioperative mortality was 0% in both groups. There was no statistical difference in follow-up durations between segmentectomies (55 ± 33 months) and wedge resections (41 ± 13 months; p = 0.29; t-test; Table 1).
3.2 Local recurrence, metastasis and survival
The median follow-up duration was 45 months. Within the first 5 years, a total of 36 stage IA patients (41%) developed a relapse (Table 2
). Local recurrence occurred in 23 patients and distant metastasis in 13 patients (Table 2). Cancer-related death was observed in 32 (37%) patients leading to a 5-year survival rate of 61% following sublobar resection in stage IA NSCLC. Segmentectomy was associated with a significantly better cancer-related 5-year survival than wedge resection in stage IA (71% vs 48%, respectively; p
= 0.016; log-rank test) (Fig. 1
, Table 2). Occurrence of distant metastasis was equal between segmentectomy and wedge resection, however, wedge resection was associated with a significantly increased local recurrence rate compared to segmentectomy (55% vs 16%, respectively; p
= 0.001; log-rank test) (Table 2). Follow-up CT scans showed that 21 (91%) out of the 23 observed local recurrences following sublobar resection developed in the remaining lung parenchyma in the areas of the former resections. Two (9%) recurrences developed in ipsilateral mediastinal lymph nodes, one following segmentectomy and one following wedge resection. Local recurrence could be treated by reoperation (extended segmentectomy) in 4 patients only (3 with prior segmentectomy and 1 with prior wedge resections) because pulmonary function was too limited in the other cases. All patients with local recurrence who were not reoperated received palliative chemotherapy and died of cancer-related causes.
|
|
2 cm diameter, 5-year survival rates following segmentectomy or wedge resection differed significantly (80% vs 48%, respectively; p
= 0.005) (Fig. 2
, Table 3
). Furthermore, a significantly increased local recurrence rate following wedge resection was observed compared to segmentectomy in this group (40% vs 11%, respectively; p
= 0.007; log-rank test) (Table 3).
|
|
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
2 cm. One explanation for the contradiction to the results of the present investigation could be that in the study of Okada et al., wedge resections were allowed for tumours 1.5 cm and segmentectomies for tumours 2 cm. Since tumour size is of significant prognostic impact in stage IA NSCLC [3,11], their comparison between outcomes following segmentectomy or wedge resection could be biased by the above mentioned inclusion criteria. A verifying multivariate analysis for segmentectomies and wedge resections has not been shown by Okada et al. [3]. Our present results are strengthened by a previous observation in the LCSG trial [2] reporting an increased local recurrence rate of 8.6% following wedge resection compared to 4.4% following segmentectomy in stage IA NSCLC. A prognostic impact of these LCSG data has been discussed controversially because local recurrence does not automatically result in cancer-related death if the local recurrence can be treated sufficiently. In the present study, 13 out of the 14 local recurrences following wedge resection were observed in the remaining lung parenchyma in the areas of the former resections and one recurrence developed in ipsilateral mediastinal lymph nodes. In the present series, local recurrence could be treated by reoperation (extended segmentectomy) in 4 patients only (3 with prior segmentectomy and 1 with prior wedge resections) because pulmonary function was too limited in the other cases. All patients with local recurrence who were not reoperated received palliative chemotherapy and died of cancer-related causes. In contrast, all patients with local recurrence following sublobar resection in the study by Okada et al. [3] could be treated by completion lobectomy and none of them died of cancer-related causes, because one of their inclusion criteria for sublobar resection in stage IA NSCLC was that their patients had to be able to tolerate lobectomy [3]. This difference to the inclusion criteria of the present study seems to impact cancer-related survival strikingly. Thus, the most likely explanation for the contradiction between the results of the present investigation and Okada's study [3] could be the different inclusion criteria of the two studies.
In the present study, significantly more lymph nodes were dissected in segmentectomies than in wedge resections (Table 1). This difference most likely results from the different types of lymph nodes dissection used in the two groups of sublobar resections: segmentectomy was followed by systematic lymphadenectomy and wedge resection by lymph node sampling. One might argue that better survival following segmentectomy could result from the Will Rogers effect [12] because the situation here resembles that of a classic Will Rogers quote, ‘When the Okies left Oklahoma and moved to California, they raised the average intelligence level in both states.’ For this study that would mean that systematic lymphadenectomy which leads to a better identification of patients with lymph node metastases than lymph node sampling [13,14] might result in reclassifying those patients as having more severe disease. This possibility of upstaging by systematic lymphadenectomy might raise the stage-specific survival rate of pN0-patients, because former pN0-patients with an unfavourable survival might be identified as pN1/2-patients who have a poorer stage specific survival by definition. To take this possibility into account, we included the number of dissected lymph nodes into a multivariate survival analysis (Table 4). It demonstrated that the prognostic effect of the resection type was independent from a possible prognostic impact of the number of dissected lymph nodes. Furthermore, the mean number of dissected lymph nodes in wedge resections was 6 ± 3 so that the UICC criteria for accurate lymph node staging (
6 lymph nodes) [15] were even fulfilled by lymph node sampling in the wedge resection group of this study.
Although no correlation between ASA-scores [4] and the two types of resection became apparent (Table 1), there was an insignificant tendency towards higher ASA-scores (65% ASA3-4) in patients with wedge resections (Table 1). To take this tendency into account, we included the ASA-score into the multivariate survival analysis, although the ASA-score was not of prognostic impact on univariate survival analysis. Multivariate analysis demonstrated that the prognostic effect of the resection type was independent from a possible prognostic impact of the ASA-score.
One might favour wedge resections because of their tendency towards a lower complication rate of 10% compared to 23% following segmentectomy in this study. However, this difference was not statistically significant (p = 0.15; chi square test) and perioperative death was not observed in any group. Also previous studies such as the LCSG trial did not observe any statistical differences in complication rates following wedge resection or segmentectomy (p-value not shown) [2]. In one recent study, wedge resections after prior lobectomy were even associated with a tendency towards more complications (14%) compared to segmentectomies after prior lobectomy (9%; p = 0.41) [16].
Since the present study showed that a reoperation for local recurrence in stage IA patients with limited pulmonary function was often not possible, a type of sublobar resection which most likely prevents local recurrence should be chosen for the initial operation. The results of the present work indicate that if limited functional operability requires a sublobar resection of stage IA NSCLC, segmentectomy with systematic lymphadenectomy should be preferred. To verify our data, we propose to perform a prospective randomised trial comparing wedge resection to segmentectomy in stage IA NSCLC. In such a future trial, patients should be randomised according to their functional operability and especially according to their ability to tolerate lobectomy for the above-mentioned reasons. Furthermore, we recommend systematic lymphadenectomy [15] in such a future trial or that randomisation should take into concern the type of lymph node dissection to prevent the above mentioned Will Rogers effect in pN-classification [12]. Additionally, future studies investigating survival after sublobar resection of stage IA NSCLC should always distinguish between anatomical segmentectomies and wedge resections.
|
Appendix A |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
Dr H. Eid (Dubai, UEA): You didn’t show which approach you used as a minimal invasive for each segmental resection done or was there just an open technique for all patients? If you go for a safer wedge, does it have an effect on the results and for recurrence or prognoses, compared to a segmentectomy?
Dr Sienel: In this study no VATS-wedge resections were included, all patients had open thoracotomy. The wedge resections were not stapled. They were done by excision of the tumour using electrocautery of scissors. For segmentectomy, the bronchus was dissected with preparation of the pulmonary artery and veins. The next step or the consequence of the data by Okada et al. would have been to perform thorascopic wedge resections for stage Ia non-small lung cancer but our data shows that segmentectomy is much better.
Dr P. De Leyn (Leuven, Belgium): Thank you for this nice presentation. I confirm the idea that segmentectomy is better for small tumours compared to wedge excision. However, this is a retrospective study with possible bias. Could it be that patients with a higher cardiopulmonary morbidity, were intra-operatively selected rather for wedge excision? Your data show that the type of lymph node dissection was different (more sampling) in the wedge excision group. Although it is not proven that there is an advantage for survival, it could be that in patients with wedge excisions there was an understaging of the tumour and that in fact some patients with stage Ia tumours in the wedge excision group had N1 or even N2 disease and this could also explain the difference in survival.
Dr Sienel: The limited pulmonary function was evaluated preoperatively. We made the decision to perform a wedge resection or segmentectomy preoperatively. Concerning the second question I thank you very much for mentioning the Will Rogers effect that may result from a stage migration if you remove more lymph nodes. But we had made up our mind about this point. We entered the number of resected lymph nodes into the multivariate analysis and there was no effect on the prognostic impacts of the resection type. Besides that, the mean number of resected lymph nodes was 6 in the wedge resection group allowing an accurate lymph nodes staging according to the UICC classification.
Dr A. Oliaro (Turino, Italy): In the group of segmentectomy you performed a systematic lymph adenectomy, in the other group you performed a sampling lymph adenectomy. Is it correct to compare these two groups and why didn’t you perform the same lymph adenectomy?
Dr Sienel: This was a retrospective study and we considered the differences in lymph node dissection in our statistical analysis. But in a prospective study, we would perform a systematic lymph adenectomy in both groups, of course.
Dr D. Branscheid (Hamburg, Germany): I think when somebody survives 5 years his functional risk should be thought about once more; you should define what the risks are. Second segmentectomy is not like segmentectomy, I think when you dissect with an anatomical resection segment 2, it is different than when you perform a segment 6 resection, so how many segments and what was the topography of your segmentectomies?
Dr Sienel: The first question was more comment. Concerning the second question, we performed 31 segmentectomies in the upper lobes, there were 8 in the left number 4 and 5 segments, or left segment 4 and 5 together. There were 9 in segment 6 and 8 in the segments number 7, 8, 9 and 10.
Dr H.B. Ris (Lausanne, Switzerland): Thank for this nice and interesting study. The problem is that this is a retrospective study over a long period of time (15–20 years).
First question: how was the preoperative staging performed in this period of time, in other words, were all patients staged the same way or do you have a bias in this respect?
Second question: on what was the choice of the surgeon based to perform either a wedge resection or a segmentectomy?
Dr Sienel: Concerning the first question, the preoperative staging was similar, standardised and without bias. There was no PET or PET/CT. We performed ultrasound of the abdomen and a chest CT in all patients. Cranial computed tomography or bone scintigraphy was used selectively to investigate specific preoperative symptoms related to metastatic disease.
Your second question is difficult to answer in a retrospective study. We read all the operation reports and found out that it was only on the discretion of the surgeon whether he performed a segmentectomy or a wedge resection. The preoperative pulmonary functions were not different in the two groups.
Dr T. Rice (Cleveland, OH): Because this is a retrospective analysis you have definite biases, you can consider a multivariate analysis or looking at matched pairs. That would make your analyses and your comparisons much more available.
Dr Sienel: There was no difference between the two groups in gender, age, ASA-score, tumour size, histology, grading, complication rate, and follow-up duration, but I will talk to our statistician about such a matched pair analysis.
Dr Rice: But you can get information from your retrospective study by given matched pairs. Even though you can compare each individual characteristic you can do a score for a wedge resection and put that in multivariable analyses as a variable and that would help you. Or you can look at potentially matched pairs who had segmentectomy lobectomy, it will make a much stronger conclusion.
Dr Sienel: This is a good idea and I will talk to our statistician about a matched pair analysis, but the best thing would be to perform a prospective study.
Dr G. Varela (Salamanca, Spain): It has been shown that lobectomy in patients with low FEV1 is useful to improve postoperative pulmonary function especially in patients with upper lobe emphysema. A second question is: I have seen in your series 77% morbidity. Probably this is not statistically significant but it seems a rather high rate of complications. Would you like to comment something on morbidity?
Dr Sienel: Of course it is feasible to combine lung volume reduction surgery with lung cancer resection in patients with pulmonary emphysema, but this was not the topic of this study. We just wanted to compare wedge resection to segmentectomies. There were no lung volume reductions included in this series. There was no mortality in both groups. The complication rate was slightly increased following segmentectomy but this increase was not statistically significant.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Presented at the 15th European Conference on General Thoracic Surgery, Leuven, Belgium, June 3–6, 2007. |
References |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
