Is lobectomy really more effective than sublobar resection in the surgical treatment of second primary lung cancer?^†

Abstract

OBJECTIVES

Sublobar resection for early-stage lung cancer is still a controversial issue. We sought to compare sublobar resection (segmentectomy or wedge resection) with lobectomy in the treatment of patients with a second primary lung cancer.

METHODS

From January 1995 to December 2010, 121 patients with second primary lung cancer, classified by the criteria proposed by Martini and Melamed, were treated at our Institution. We had 23 patients with a synchronous tumour and 98 with metachronous. As second treatment, we performed 61 lobectomies (17 of these were completion pneumonectomies), 38 atypical resections and 22 segmentectomies. Histology was adenocarcinoma in 49, squamous in 38, bronchoalveolar carcinomas in 14, adenosquamous in 8, large cells in 2, anaplastic in 5 and other histologies in 5.

RESULTS

Overall 5-year survival from second surgery was 42%; overall operative mortality was 2.5% (3 patients), while morbidity was 19% (22 patients). Morbidity was comparable between the lobectomy group, sublobar resection and completion pneumonectomies (12.8, 27.7 and 30.8%, respectively, P = 0.21). Regarding the type of surgery, the lobectomy group showed a better 5-year survival than sublobar resection (57.5 and 36%, respectively, P = 0.016). Compared with lobectomies, completion pneumonectomies showed a significantly less-favourable survival (57.5 and 20%, respectively, P = 0.001).

CONCLUSIONS

From our experience, lobectomy should still be considered as the treatment of choice in the management of second primary lung cancer, but sublobar resection remains a valid option in high-risk patients with limited pulmonary function. Completion pneumonectomy was a negative prognostic factor in long-term survival.

INTRODUCTION

Currently, lung cancer is the leading cause of cancer-related deaths in developed countries; moreover, although the overall incidence of lung cancer showed a decrease in the last 15 years [1], the widespread use of accurate follow-up methods such as spiral multislice computed tomography (CT) and positron emission tomography (PET) and the better survival results after resection of primary lung carcinoma, led to an increased rate of early diagnosis of second primary lung cancer, classified as synchronous or metachronous [2].

Lobectomy still remains the standard treatment for patients with resectable non-small-cell lung cancer (NSCLC) [3], but the most efficacious and correct surgical approach for second primary lung cancer has not been identified yet.

Optimal management is, in fact, affected by many factors, that should be carefully considered in deciding the extent of parenchymal resection: age, cardiopulmonary reserve, associated medical comorbidities and stage of second lung cancer represent the most substantial of them [4].

Many patients presenting with early-stage disease may be unable to tolerate a lobar resection due to poor cardiopulmonary function, and in this group, the use of sublobar resection, when feasible in terms of anatomical setting and wide surgical margins, could lead to comparable survival results, reducing the loss of pulmonary reserve [5–9].

The purpose of our study was to analyse and compare the results of lobectomy with those of sublobar resection (segmentectomy and atypical resection) in those patients who underwent surgical treatment for second primary lung cancer.

METHODS

A retrospective study was performed analysing data of 121 consecutive patients with second primary lung cancer, treated at our Institution from January 1995 to December 2010.

During this period, 4480 patients underwent complete resection for primary lung cancer, and the study group represented an incidence of 2.7% among the overall series.

Tumours were designated as ‘synchronous’ when detected or resected simultaneously and ‘metachronous’ when the second tumour was found some time later: our criteria for the definition of a synchronous or metachronous cancer were those proposed by Martini and Melamed [2].

Twenty-three (19%) patients were classified as synchronous and 98 (81%) as metachronous: there were 105 (86.8%) males and 16 (13.2%) females, ranging in age from 50 to 83 years, with an overall median age of 68 years.

Standard preoperative evaluation included a detailed clinical history and physical examination, chest radiography, bronchoscopy, chest and upper abdomen CT scan and PET scan; when indicated from standard evaluating procedure, from the stage or clinical situation, brain magnetic resonance and/or bone scintigraph were added.

Hilar and mediastinal lymph nodes were sampled or systematically dissected during the lobectomy or segmentectomy, while only enlarged or PET-positive nodes were dissected in patients submitted to atypical resection.

All patients were restaged according to the seventh edition of the TNM classification [10]; interval time between tumours was considered to be from the date of surgery of the first tumour to the date of radiological diagnosis of the second tumour.

Among patients submitted to sublobar resections, the main reasons why this kind of surgery was selected in place of lobectomy were an impaired pulmonary function, usually related to the previous surgery or chronic lung diseases and a reduced cardiovascular function. Other different parameters, like the presence of comorbidities or size and anatomical site of the tumour were considered in the pre- and intraoperative evaluation.

Operative mortality included patients who died within the first 30 days after surgery or beyond 30 days during the same hospitalization, while postoperative complications were defined as major when potentially life-threatening and requiring an interventional treatment, and minor when requiring a non-interventional therapy and prolonging the hospital stay.

Disease recurrence was considered local when it arose in the same lobe or in hilar/mediastinal lymph nodes; all other recurrence was defined as distant metastasis.

Survival was calculated from the date of the second tumour resection to the date of the last follow-up or death.

Statistical analysis

The comparative analysis to study differences in patients and tumour characteristics was performed with the χ²-test and Fisher's exact test on categorical variables and Student's unpaired t-test on continuous variables. Survival analysis was performed by the Kaplan–Meier method and survival differences were analysed by the log-rank test. A P-value of <0.05 was considered statistically significant. All analyses was performed with SPSS version 18.0 for Windows (SPSS, Inc., Chicago, IL, USA).

RESULT

Patient characteristics

Table 1 summarizes the overall patient characteristics.

Median interval between the first tumour resection and diagnosis of the second tumour in the metachronous group, was 66 months (range 8–402).

The comparison between patients and tumour characteristics in the two groups was performed and did not show any significant statistical difference.

Histology

Adenocarcinoma was found in 49 patients (40.6%), squamous cell carcinoma in 38 (31.5%), former broncho-alveolar carcinoma in 14 (11.5%), adenosquamous cell carcinoma in 8 (6.6%), large cell carcinoma in 7 (5.7%), other histology in 5 (4.1%).

T status was T1 in 70 patients (57.8%), T2 in 38 patients (31.4%) and T3 in 13 patients (10.8%), while nodal status was N0 in 107 patients (88.4%), N1 in 9 (8.1%) and N2 in 5 (3.5%).

Tumour site and type of surgery

In the metachronous group, tumour was contralateral in 62 patients (63%) and ipsilateral in 36 (37%), while in synchronous group, tumour site was contralateral in 9 (39%) and ipsilateral in 14 (61%).

The surgical procedures for the first tumour (or for the larger tumour in the synchronous group) consisted in 106 (88%) lobectomies, 9 segmentectomies (7%), 5 pneumonectomies (4%) and 1 atypical resection (1%).

For the second tumour (or for the smaller tumour in the synchronous group), 61 patients (50.4%) underwent lobectomy (17 of these were completion pneumonectomies), 38, atypical resection (31.4%), 5 of whom were performed in those patients who had pneumonectomy as first operation and 22 segmentectomies (18.2%).

Postoperative mortality and morbidity

Overall postoperative mortality was 2.5% (3 patients). The causes of death were pneumonia in 1 patient (after right pneumonectomy), postoperative haemorrhage in 1 (after atypical resection) and myocardial infarction in 1 (after right lower lobectomy), with no evidence of significant difference between the groups.

Overall major morbidity was 19% (22 patients) (Table 2); 12.8% of lobectomy group patients experienced a postoperative complication against 27.7% of sublobar resection and 30.8% of completion pneumonectomy group: though a favourable trend was evident towards lobar resection, no statistical significant differences were found (P = 0.21).

Survival analysis

The overall 5-year survival rate from the first surgery was 76% (Fig. 1) and from the second surgery, 42% (Fig. 2).

Regarding type of surgery, lobectomy was associated with a significantly better 5-year survival (57.5%; 95% confidence interval [CI]: 48.6–67.7) compared with overall sublobar resection (36%; 95% CI: 26.8–45.2) (P = 0.016), however the differences were not significant when the lobectomy group was separately analysed with segmentectomies (P = 0.49) (Fig. 3).

When the second resection consisted in a completion pneumonectomy, comparison with lobectomy showed a significantly less favourable 5-year survival (20% [95% CI: 10.1–27.9] and 57.5%, respectively; P = 0.001).

In this group, causes of worse survival were mostly related to distant metastases and cardiopulmonary failure, and these two factors were often combined.

We found three local recurrences (5%) in the sublobar resection group and 2 (3.3%) in the lobectomy group, but the difference was not significant.

Analysis of prevalence of recurrences, considering T and N and type of resection, did not show any significant correlation. This result was probably linked to the small number, but 3 of 5 patients were T2 and 2 patients were T3.

Five-year survival from the time of initial diagnosis for patients with metachronous lesions was significantly better compared with those with synchronous disease (83% [95% CI: 83.8–115.2] and 40% [95% CI: 36.7–52.4], respectively; P = 0.02) (Fig. 4).

Regarding interval time, 5-year survival from the time of the first diagnosis in patients in whom metachronous disease developed at an interval ≥2 years was 91% (95% CI: 101.9–142.4), while survival was 52% (95% CI: 42.2–79.4) when the time gap was <2 years (P = 0.002).

Univariate analysis revealed that sublobar resection, completion pneumonectomy, age >75 years, nodal status and presence of more than one major comorbidity were significant, independent negative predictive factors for overall long-term survival.

Multivariate analysis confirmed the presence of more than one major comorbidity as the only identified negative prognostic factor in long-term outcome.

DISCUSSION

The risk of the development of a second primary lung cancer in those patients successfully treated for their first NSCLC is estimated to be ∼1–2% per patient per year [11, 12].

In clinical studies it may be difficult to absolutely define whether the second tumour is a second primary lesion, a recurrence or a metastasis.

The first description of the parameters indicating multiple primary lung cancer was by Warren and Gates [13] in 1932, followed by several authors who revised these criteria [14, 15].

Nowadays, the most recognized criteria are still those of Martini and Melamed [2], recently revised in the American College of Chest Physicians Lung Cancer Guidelines [16], where the extension of the interval time and the analysis of molecular genetic characteristics might be expected to improve the specificity of these criteria.

However, in many cases the question of whether a newly discovered lung lesion after resection is actually a second primary tumour or a recurrence may remain a theoretical one, because most thoracic surgeons would agree that the new cancer should be resected if the lesion is solitary in lung parenchyma, the local and distant staging is negative and the patient will tolerate lung resection [17, 18].

Patients with synchronous diseases did experience significantly reduced survival compared with those with metachronous diseases, similar to the results in other published reports [19], but in contrast with others [20].

Although the decreased survival noted in these patients cannot be related to a specific determinant responsible for this poor outcome, the possibility that some cases of solitary pulmonary metastasis always exist within the group of patients with synchronous and early metachronous lung cancers could be included.

In any case, our 5-year survival of 52% for ‘early’ metachronous cancer and 40% for synchronous cancers is far better than what would be expected for metastatic or local recurrent disease.

Regarding the extent of surgery, many studies, comprehending our previous experience also, have reported that sublobar resection, like segmentectomy or atypical resection, may be feasible in second primary lung cancer [21–24], while other authors showed a negative impact of atypical, non-anatomical resection [25].

In our study, we found that lobectomy, compared with all sublobar resections, had a statistically significant positive association on survival rate, while the operative risk, in terms of morbidity and mortality, showed just a slightly more favourable trend towards lobar resections.

Moreover, when compared in terms of recurrence rate, there were no differences between the two groups.

However, when lobar resection led to a completion pneumonectomy, this was an independent, negative prognostic factor for long-term survival.

Trying to further study the effective influence of sublobar resection, we separately analysed the segmentectomy and the atypical resection group.

We found that segmentectomies may produce survival comparable with lobectomies, despite a trend to better results towards in the lobectomy group.

From our experience in the surgical treatment of second primary lung cancer, we conclude that an anatomical lobectomy does confer a significant, overall survival advantage compared with sublobar resection.

However, anatomical segmentectomy has been shown to have better survival results compared with atypical, non-anatomical resections and may more closely approach the results of lobectomy.

Segmental resection should, therefore, represent the ideal surgical procedure to minimize recurrence in such situations where a parenchymal-sparing technique is indicated or recommended.

Completion pneumonectomy, instead, should only be performed in carefully selected patients.

The main limitation of this study is represented by its retrospective nature and lack of randomization, which may lead to potential, probable bias in patient selection and surgeon choice of the operative approach.

Prospective, randomized studies will be necessary to fully understand the feasibility of segmental resection compared with lobectomy, also in this field of second primary lung cancer.

Conflict of interest: none declared.

REFERENCES

APPENDIX. CONFERENCE DISCUSSION

Dr P. De Leyn(Leuven, Belgium) There remains a lot of discussion in the literature about whether there is a place for sublobar resection in the treatment of lung cancer. I think we should stress that this is a retrospective study in which you looked at the outcomes after different types of resection for second primary lung cancer. In your study, 121 patients were treated and in 80% there were metachronous tumours. Nearly half of the resections were performed by lobectomy and the others were atypical resections or segmentectomy. Your results showed that lobectomy has a better survival compared to atypical resections but not when compared to segmentectomy. However, I think it is difficult to draw too many conclusions from a retrospective study. As you know, it is not clear how patients were selected for lobectomy or for atypical resections, and I think that tumour diameter may be one factor, and this is, of course, prognostically important. Also, pulmonary function will have influenced the selection of patients for lobectomy or sublobar resection, and we all know that this has an effect on survival. However, I think your study shows, as many studies have also shown, that there is a role for sublobar resection in the treatment of these patients, especially in those with second primary lung cancer.

I have two brief questions. Can you tell us how patients were selected for either lobar resection, sublobar resection, or wedge resection? My second question relates to completion pneumonectomy which I think is a major intervention. In your study, the major morbidity was 30% and 5-year survival was only 20%. Can you tell us, do we still have to do this type of reintervention, or can we do chemoradiotherapy or stereotactic radiotherapy? What do you think about completion pneumonectomy when looking at your results?

Dr Zuin: You are perfectly right. The retrospective nature is maybe the major pitfall of the study, and this is related also to the selection of the patients. It is not a randomized study, so the selection was made by the surgeon, particularly in regard to pulmonary function. When feasible, usually lobectomy was performed, and in patients with maybe limited lung function or with a favourable anatomical site, we selected sublobar resection, and especially segmentectomy. So it really was a personal choice of the surgeon for each patient, and this is obviously the pitfall of a retrospective study. Completion pneumonectomy, that is a completely separate group. We think that this is not a prohibited technique, but it should be used only in very, very selected patients.

Dr M. Dusmet(London, United Kingdom): I have one question for you, if I may. The patients who have sublobar resections have limited lung function. We all know that COPD has definite 3-year and 5-year attrition/mortality, as does lung cancer. So were the deaths in the sublobar resection group due to their COPD or to lung cancer?

Dr Zuin: Most of the deaths were related to lung cancer, but also to cardiac problems. Not really many patients died from COPD, but it is also a very important risk factor.

Dr Dusmet: When you say cardiac problems, right or left?

Dr Zuin: We had a high number of myocardial infarction or heart failure.

Dr M. Zielinski(Zakopane, Poland): Staging for this repeated resection might be difficult because some of these patients already had mediastinoscopy or other studies for the first procedure. How did you stage the patients for this second operation?

Dr Zuin: Usually we perform CT scan and PET scan now. In this group we did not perform any re-mediastinoscopy, but usually for a first primary lung cancer.

Dr Zielinski: EBUS, EUS?

Dr Zuin: We have EBUS, but just in more recent years.

Dr S. Griffo(Naples, Italy): You say that long-term follow-up plays a prominent role in this field. Did you change your follow-up programme for your patients after this experience?

Dr Zuin: Yes. We changed our follow-up to using CT scan every year after about the first 4 years, usually changing from chest X-ray to CT scan in almost all of these patients.

Author notes