• 
• 

Three Part Question

In [patients with non-small cell lung cancer] can [radical lymph node dissection] improve [survival]?

Clinical Scenario

You are a Specialist Registrar in Cardiothoracic surgery. You meet a senior trainee from Japan who is an enthusiastic proponent of radical lymph node dissection in early lung cancer. Your boss is an exponent of VATS lobectomy and you are aware that the application of systematic lymph node dissection is near impossible with this technique. You wonder whether in general conducting a simpler staging technique, such as sampling, impacts on the accuracy of staging and the overall survival.

Search Strategy

Medline 1966-12/03 using the Ovid interface.
[exp lymph node excision OR lymph node dissection.af OR lymphadenectomy.af] AND [exp lung neoplasms OR lung cancer.af OR lobectomy.af OR exp pneumonectomy OR pneumonectomy.af] AND [exp mortality OR exp survival OR survival.af] LIMIT to human AND English language.

Search Outcome

Out of the 305 papers found, 20 were deemed relevant. Eleven were out of scope and one was rejected on the basis of poor methodology. Eight papers were reviewed in full. These are listed in the table.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Izbicki J et al, 1995, Germany	201 patients with operable NSCLC 100 patients with radical systematic mediastinal lymph node dissection (LA) 82 patients with mediastinal lymph node exploration and removal of suspicious nodes only (LS) After randomisation 19 patients (LS n=1, LA n=18) were excluded from analysis due to residual tumour or classification as small cell lung cancer	Controlled prospective RCT (level 2b)	Effect of redical lymphadenectomy on nodal staging of NSCLC	Procedure: LS n=1 wef00 N0 55%, N1 20%, N2 23%, N3 2% LA n=82, N0 58.5%, N1 10.9%, N2 26.8%, N3 3.7% No significant difference	Disproportionately high number excluded from the LA group Small patient numbers in each subgroup Median follow-up period was only 26.8 months LS involved extensive sampling Nodes from stations 4,5 and 7 were excised from all patients, nodes in regions 2-9 were explored and removed if suspicious
			Survival/local recurrence	Type of lymphadectomy did not influence the risk of local tumour recurrence, distant mediastasis or survival
Izbicki JR et al, 1998, Germany	201 patients with operable NSCLC 32 patients excluded from analysis LA n = 76 LS n = 93 Same patient population as Izbicki et al 1995 Median follow up 47.5 months (range 25-67)	Controlled prospective randomized clinical trial (level 2b)	Survival	LA 5 yr survival 70.6%* LS 5 yr survival 47.9%* LA prolonged relapse free survival p=0.037 with a borderline effect on overall survival (p = 0.058) in patients with limited lymph node involvement (*pN1 disease or pN2 disease with involvement of only one lymph node level); In patients with pN0 disease, no survival benefit was observed	Median follow up 47.5 months (25-67 months) Small numbers within subgroups reducing statistical power More squamous cell carcinomas in the LA group than in the LS group (52.7 vs. 31.6%) Study was powered to detect a 20% survival benefit of LA over LS only if there were 100 patients in each group Stage migration
			Morbidity	Post operative course was more or less the same in the 2 groups Higher need for post op transfusion and higher incidence of prolonged air leakage in patients with systematic lymph node dissection No difference in Hospital or ICU stay
Passlick B et al, 2002, Germany	94 patients with pathological stage I-IIIA NSCLC Stage I NSCLC n = 73 Radical systematic en-bloc-mediastinal lymphadencetomy (LA) n = 42 Mediastinal Lymph node sampling (LS) n = 31 Samples were screened by immunohistochemistry for disseminated tumour cells using antibody Ber-Ep4	Prospective Randomized Trial (level 2b)	Survival	Overall survival was 54.7% 5 year survival in the LA group and 41% in the LS group p=0.27 Stage I LA 62% 5 yr survival, LS 42% 5 year survival p=0.044 Stage II to IV LA 27% 5 yr survival, LS 30% 5 yr survival, p=0.39	Small numbers involved in the stage II to IV group (only 21 patients in total)
			Immuno-histochemistry evidence of micrometastases	In patients without nodal micrometastases LA was associated with significantly improved long-term outcome p = 0.0044 In patients with nodal micrometastases the prognosis was not influenced by the type of lymphadenectomy
Wu Y et al, 2002, China	Resectable clinical stage I-IIIA NSCLC 268 patients assigned to lung cancer resection combined with systematic lymph node dissection (SND) 264 patients were assigned to lung cancer combined with mediastinal lymph node sampling (MLS) 471 patients were eligible for follow up	Prospective Randomized Trial (level 1b)	Survival	Stage I n=156, SND 82.2% 5 yr survival, MLS 57.5% 5 yr survival p=0.0104 Stage II n=136, SND 50% 5 yr survival, MLS 34.1% 5 year survival p=0.052 Stage III n=179, SND 27% 5 yr survival, MLS 6.2% 5 yr survival p=0.284	More stage I and less stage IIIA patients in those undergoing MLS, suggesting that MLS is less accurate in staging the disease and that stage migration may influence survival in this trial Unequal follow up between groups
			Post operative recurrence	SND 7 (2.9%) MLS 11 (4.8%) No statistical significance specified
			Metastases	SND 54 (22.5%) MLS 71 (30.7%) No statistical significance specified
Keller SM et al, 2000, USA	373 patients with non small cell lung cancer 186 patients underwent systematic sampling (SS) of the mediastinal lymph nodes 187 patients underwent complete mediastinal lymph node dissection (MLND)	Non randomized controlled study (level 2b)	Survival	N1 disease SS 5 yr survival 57%, MLND 5 year survival 48%, p=0.04. N2 disease SS 5 yr survival 41%, MLND 5 yr survival 35%, p=0.035.	Not randomized 192 surgeons involved in the study, some exclusively performed one technique or the other 131 surgeons entered only one patient
Garja A et al, 2003, USA	442 patients with stage I NSCLC 246 patients had random sampling. 115 patients had systematic sampling (SS) 81 patients had complete mediastinal lymphadenectomy (MLND)	Retrospectve Cohort Study (level 4)	Overall survival	56%, 83% and 86% (p<0.0001) for random sampling, SS and MLND respectively	Retrospective data Spurious downstaging of patients with inadequate sampling Large number of patients excluded from the original cohort due to inadequate follow up and incomplete data
			Disease free survival	51%, 80% and 80% (p<0.0001) for random sampling, SS and MLND respectively
Sugi K et al, 1998, Japan	115 patients with peripheral non small cell lung cancers < 2cm diameter 59 patients had lymph node sampling 56 patients had radical lymph node dissection	Randomized Controlled Trial (level 2b)	Survival	Lymph Node Sampling Group 83.9% 5 yr survival Lymph Node Dissection Group 81.4% 5 year survival No statistically significant difference	Small numbers in each group Study limited to 2cm diameter tumours, therefore excluding some stage I tumours
			Morbidity	Morbidity was significantly higher in the dissection group 26.8% vs. 3.4% p value not stated Comparable ICU and hospital stay
Wu YC et al, 2003 Taiwan	321 patients undergoing surgery for Stage I NSCLC	Retrospective Cohort Study (level 2b)	Survival	Removed no. of lymph nodes > 15 5 yr survival 57.1% 10 yr survival 46.3% Removed no. of lymph nodes < 15 5 yr survival 45.5% 10 yr survival 31.5% p<0.01	13 patients with inadequate pulmonary reserve had sub lobar resections 18 patients lost to follow up Retrospective study Spurious downstaging of patients with inadequate sampling

Comment(s)

Izbicki et al [4,5], Passlick et al [9], Wu et al [8], and Keller et al [7] found a survival benefit to the use of radical lymph node dissection. Gargi [11] and Sugi [6] could not identify a survival benefit. Izbicki et al [4,5] and Passlick et al [9] all describe outcomes regarding the same cohort of patients. These papers in common with Wu et al [10] were prospective randomized controlled trials. Sugi et al [6] is a randomized controlled trial, however, the findings are limited to a small subset of stage I lung cancer in which tumours are peripheral and less than 2 cm in diameter. Gargi et al [11] conducted a retrospective case note review and acknowledge the potential for spurious down-staging of patients with inadequate sampling, a large number of patients from the original cohort were excluded due to inadequate follow up and incomplete data. In a number of patients in whom only a limited lymphadenectomy is performed, the true N stage remains unrecognised because the relevant lymph nodes are not removed and consequently not examined by a histopathologist. In the Izbicki cohort 5.5% of patients in the LA group had N2 disease that was detected only at lymph node levels that would not have been routinely included in the lymph node sampling group. Consequently the hypothetical benefit of LA in patients with limited mediastinal lymph node involvement might be due at least to an imbalance within the groups with respect to the number of patients with lymph node involvement at multiple levels of the N2 region. The phenomenon of stage migration as a source of misleading statistics for survival in cancer has been called the Will Rogers phenomenon [1]. Of note, the British Thoracic Society guidelines [2] with regard to lymph node management are based on the interim work of Izbicki et al in 1994. At this juncture the results had not shown a statistical significance between their two groups of patients. According to the BTS Guidelines lymph node dissection is essential at the time of lung resection to achieve accurate staging, however, extensive lymph node resection is not advised for its therapeutic value. Assessment of a greater number of lymph nodes correlates with improved survival in patients with colorectal, breast and bladder cancer. Wu et al [10] found that the number of lymph node metastases was an independent predictor of survival. Vansteenkiste et al [1] analysed 18 articles published between 1980 and 1995 and also concluded that N status is the most important prognostic factor. Hypothetically there exists a cohort of patients with metastatic disease that is truly limited to the regional lymph nodes. These are the patients who would benefit from the aggressive resection of the intrathoracic lymph nodes. Proponents of the radical approach claim better staging of the tumour and improved prognosis. Opponents of radical lymph node dissection have claimed higher morbidity and mortality rates owing to the extent of the operation and even a negative effect on the long term prognosis because of an impaired local immune response. Both Izbicki et al [5] and Sugi et al [6] found increased morbidity in the lymphadenectomy groups, however, neither found any increase in hospital or ICU stay.

Clinical Bottom Line

The small numbers of patients involved in these studies and the problems with study design have not yielded a clear cut answer as to the survival benefit of mediastinal lymphadenectomy. If a cohort does exist where metastatic disease is limited to the regional nodes then these are the patients who would benefit in terms of survival. This group of patients may be very small requiring a highly powered study to detect them.

References

1. Izbicki J, Passlick B, Pantel K et al. Effectiveness of radical systematic mediastinal lymphadectomy in patients with resectable non-small cell lung cancer. Ann Surg 1998;227(1):138-44.
2. Izbicki JR, Passlick B, Karg O, et al. Impact of radical systematic mediastinal lymphadenectomy on tumor staging in lung cancer. Ann Thorac Surg 1995;59(1):209-14.
3. Passlick B, Kubuschock B, Sienel W, et al. Mediastinal Lymphadenectomy in non-small cell lung cancer: effectiveness in patients with or without nodal micrometastases – results of a preliminary study. European Journal of Cardiothoracic Surgery 2002;21(3):520-526.
4. Wu Y, Huang ZF, Wang SY, et al. A randomized trial of systematic nodal dissection in resectable non-small cell lung cancer. Lung Cancer 2002;36(1):1-6.
5. Keller SM, Adak S, Wagner H, et al. Mediastinal Lymph Node Dissection Improves Survival in Patients With Stages II and IIIa Non Small Cell Lung Cancer. Ann Thorac Surg 2000;70(2):358-66.
6. Gajra A, Newman N, Gamble GP et al. Effect of Lymph Nodes Sampled on Outcome in Patients With Stage I Non-Small-Cell Lung Cancer. J Clin Oncol 2003;21(6):1029-1034.
7. Sugi K, Nawata K, Fujita N, et al. Systematic Lymph Node Dissection for Clinically Diagnosed Peripheral Non-Small-Cell Lung Cancer Less Than 2 cm in Diameter. World J Surg 1998:22(3);290-295.
8. Wu YC, Lin C, Hsu W, et al. Long-term results of pathological stage I non-small cell lung cancer: validation of using the number of totally removed lymph nodes as a staging control. European Journal of Cardiothoracic Surgery 2003;24(6):994-1001.
9. Feinstein AR, Sosin DM, Wells CK. The Will Rogers phenomenon. Stage migration and new diagnostic techniques as a source of misleading statistics for survival in cancer. N Engl J Med 1985;312(25):1604-8.
10. Fountain et al. BTS guidelines: guidelines on the selection of patients with lung cancer for surgery. Thorax 2001;56(2):89-108.
11. Vansteenkiste JF, De Leyn PR, Deneffe GJ, et al. Clinical prognostic factors in surgically treated stage IIIA-N2 non-small cell lung cancer: analysis of the literature. Lung Cancer 1998;19(3):219.