Author
Year
# Resected (%)
% Pneum
Ch/Ch-R
# pN2 (%) +
Median f/u ++
Median surv
5 years surv
Quality
Albain et al. [2]
1995
57 (76)
33
Ch-R
28 (50)
29 months
10 months
18 %
High
Mathisen et al. [3]
1996
35 (78)
18
Ch-R
21 (60)
4–83 months
26 months
High
Bueno et al. [4]
2000
103
37
Both
49 (48)
Min 4 years
15.9 months
9 %
Moderate
Betticher et al. [5]
2003
75 (83)
49
Ch
26 (35)
32 months
16.2 months
11 % 3 years
High
Steger et al. [6]
2009
55 (100)
47
Ch-R
20 (36)
30 months
19 months
16 %
Moderate
Decaluwe [7]
2009
85
24
Ch
49 (53)
51 months
27 %
High
Albain et al. [8]
2009
155 (77)
35
Ch-R
85 (55)
69.3 months
26.4 months
25 %
High
Stefani et al. [9]
2010
175
45
Ch
107 (61)
32.5 months
24.9 month
22 %
Moderate
Freidel et al. [10]
2010
62
58
Both
18 (18)
93 months
28 months
33 %
Moderate
Paul et al. [11]
2011
136
23
Ch, 10 % R
65 (48)
42 months
21 % L; 19 % P
Moderate
Meacci et al. [12]
2011
161
15
Ch-R
40 (25)
24 months
19 %
High
Shitani et al. [13]
2012
52
17
Ch-R
23 (44)
58 months
0
Moderate
Perhaps the gold standard for comparing each of these studies is the Intergroup 0139 study which has the largest number of patients [13]. They were treated in both the academic and community setting by a large number of oncologists, radiation therapists and surgeons. The control arm was definitive chemo-radiation. The median survival of patients with pN2 disease following resection in that report was 26.4 months. The average median survival in the other studies in Table 13.1 was 21 months with a range of 10–28 months. The predicted 5 year survival in INT0139 was 25 % versus an average of 18 % in the other studies. These numbers give us our best approximation of the overall anticipated benefit of operating on patients with pN2 NSCLC after induction therapy. This does not answer the question of what is the best induction chemotherapy regimen, whether radiation should be included in the regimen, or whether postoperative adjuvant therapy should be given. In INT 0139 patients received two cycles of adjuvant chemotherapy. In seven of the 12 studies the induction therapy included radiation.
Survival Following Resection of pN2 Subsets
While the studies in Table 13.1 can be interpreted as showing limited benefit to some patients with pN2 NSCLC, it is important to look at potential subsets of these patients who might have a better or worse prognosis. Studies that exclude patients with known post induction preoperative pN2 disease and report only the results of patients with recalcitrant pN2 have noted significantly better results. In the prospective phase II trimodality trial RTOG 0229, where the induction regimen included concurrent chemotherapy with full-dose (60 Gy) radiation, six of 56 (11 %) patients with mediastinoscopy or thoracoscopy positive pN2 disease after induction therapy were excluded from surgery [14]. Ten additional patients had recalcitrant mediastinal nodal disease discovered after resection. They had a median survival of 33 months but only a progression free survival of 9 months. These results, however, included the six patients with pN2 disease not offered resection. In another study of patients with Stage IIIA-N2 NSCLC, 198 patients completed neoadjuvant chemo-radiation and presented for restaging [15]. Forty-nine (25 %) patients had persistent N2 disease and were excluded from surgery. Fourteen patients who were found to have recalcitrant N2 disease on pathology had a 42 % 5 year survival. In this study, however, patients with upstaged nodal disease discovered at surgery were also not resected. These studies suggest that patients with persistent microscopic disease do better.
Meacci reported on their 40 patients who had persistent N2 involvement following induction therapy and demonstrated the most significant factor associated with mortality was the relative macroscopic vs. microscopic residual tumor in the lymph nodes [11]. Patients with macroscopic residual N2 disease had 2.8 times the risk of death. Their 5 year survival rate in these patients was 12.6 %. Conversely it was 54.6 % with microscopic involvement defined as 1–10 % viable tumor cells in the lymph node. Patients with single level involvement had a better survival than patients with multilevel involvement (39.5 % vs. 10.6 %; p = 0.10). No patient with extracapsular spread survived. Decaluwe demonstrated a difference in survival between patients with single versus multilevel involvement [12]. Patients with single level pN2 involvement had a 37 % 5 year survival whereas patients with multilevel pN2 involvement had a 0 % 5 year survival. Similarly, Tokeda observed a 34.8 % 5 year survival in 17 resected patients with single station pN2 disease versus no 5 year survivors in 16 patients with multi-station pN2 disease [16]. These few studies suggest that patients with persistent multi-station or bulky mediastinal lymph node involvement should not be offered resection. On the other hand, patients identified with single station or microscopic disease in the mediastinal lymph nodes will likely benefit from surgery.
Post Induction Patient Evaluation
One of the problems in managing patients with Stage IIIA-N2 NSCLC is determining who had a satisfactory response to induction therapy. A minimum requirement for most surgical series is an absence of disease progression by CT and PET. A more important determinant is whether a patient’s tumor has sufficiently responded to induction therapy to permit a lobectomy as opposed to a pneumonectomy, and most importantly what the status of the mediastinal lymph nodes is.
Evaluation of Lobectomy vs. Pneumonectomy
Not surprising, in most series reporting on surgery for patients with involved mediastinal lymph nodes, the requirement for pneumonectomy in order to achieve an R0 resection is high even after induction therapy. Very often these patients have central tumors or involved mediastinal lymph nodes adherent to central structures. In Table 13.1 the percentage of patients with N2 NSCLC requiring a pneumonectomy varied from a low of 15 % to as high as 58 %, the average 33 % and the weighted average 36 %. It is important to know the mortality in patients with N2 disease as opposed to, for example, patients with T3N0 or T4N0 tumors; disease stage impacts operative mortality apart from the technical issues involved, and this has been our own experience. Series reporting on operative mortality for patients undergoing lobectomy or pneumonectomy, who received induction therapy, and who had histologic proof of N2 disease prior to resection are shown in Table 13.2 [2, 3, 5–8, 12, 17]. In this table the operative mortality for lobectomy averages 2.37 % but is 10.23 % for pneumonectomy, a fivefold difference. For patients undergoing induction therapy that included radiation in the induction regimen, the mortality was slightly higher with a range from 0 to 26 %. The highest mortality was observed in the randomized controlled study reported by Albain (INT0139) [2]. This study, as we have already mentioned, was performed in many academic and community hospitals which has the strength of accumulating results from a more diverse population of patients and medical professionals. At the same time it begs the question, particularly with regards to pneumonectomy, whether experience in this type of surgery affects mortality. In this study, the operative mortality for pneumonectomy was sufficiently high as to make operation no better than non-operative management with definitive chemotherapy and radiotherapy, whereas the converse was true for lobectomy. While it is generally accepted that the addition of radiation to the induction regimen improves tumor response and sterilization of the mediastinal lymph nodes, this has not necessarily meant improved survival, particularly when one incorporates the mortality associated with pneumonectomy. When the radiation dose escalates to 60 Gy, the operative mortality associated with pneumonectomy generally increases. In addition, while most series report on 30 day and in-hospital mortality, it is becoming clear that this is arbitrary and does not truly reflect the operative risk. The 90 day mortality may be a better approximation of risk associated with pneumonectomy, the 90 day mortality being approximately twice the 30 day mortality [18, 19]. Furthermore, one needs to factor in the reduced 5 year survival associated with pneumonectomy patients and NSCLC in general to appreciate the true risk associated with the decision to operate on any patient with pN2 NSCLC. It is, therefore, imperative to determine whether an R0 resection can be performed by lobectomy and it also important to make a preoperative decision on whether to abandon resection if the need for pneumonectomy is only determined at operation.
Table 13.2
Operative mortality following induction therapy for N2 NSCLC
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
