In lung cancer patients where a malignant pleural effusion is found at operation could resection ever still be justified?

Abstract

A best evidence topic in thoracic surgery was written according to a structured protocol. The question addressed was whether surgery could ever be justified in non-small cell lung cancer patients with an unexpected malignant pleural effusion at surgery. Eight papers were chosen to answer the question. The authors, journal, date and country of publication, patient group studied, study type, relevant outcomes and results of these papers were tabulated. Study limitations included a lack of retrospective studies, the heterogeneous patient population and various treatments applied. Three papers found that surgery—compared to exploratory thoracotomy—was associated with a survival advantage in cases of minimal pleural disease. One paper showed that the median survival time of 58.8 months in patients with pleural effusion was better than that of patients with more extensive pleural dissemination as pleural nodule (10 months; P = 0.0001) or pleural nodule with effusion (19.3 months; P = 0.019). Another study showed that pleural effusion patients with N0–1 status had a median survival time more than 5 years longer than patients with similar or more extensive pleural dissemination but with N2–N3 status. A further study showed a better 5-year survival time in patients with pleural effusion, than in patients with pleural nodule (22.9% vs 8.9%, respectively; P = 0.45). In two papers, surgery vs exploratory thoracotomy had better survival in cases of N0 status and of complete tumour resection independently of pleural dissemination. Different strategies were employed to obtain freedom from macroscopic residual tumour, including pneumonectomy, lobar resection or, to a lesser extent, pleurectomy in patients having pleural dissemination. Only one paper reported a worse median survival time after pneumonectomy than for more limited resections (12.8 vs 24.1 months, respectively; P = 0.0018). In the remaining papers, no comparison between the different resections was made. In all studies except one, surgery was a component of multimodal treatment. Intrapleural chemotherapy was largely applied with systemic adjuvant chemotherapy and/or radiotherapy. The study period and/or year of publication of most papers was 10 years or more, this may explain the different chemotherapy regimens used in the various studies. No current guidelines support surgery over conservative therapy and the identified studies in this review are not strong enough to change this recommendation.

INTRODUCTION

A best evidence topic was constructed according to a structured protocol. This is fully described in the ICVTS [1].

THREE-PART QUESTION

In [non-small cell lung cancer (NSCLC) patients with unexpected malignant pleural effusion (MPE)] could [resection] ever be [justified]?

CLINICAL SCENARIO

A 55-year-old patient comes to your practice presenting a resectable NSCLC. During surgery, an unexpected PE is found; a sample is send to the pathology unit for cytology. Despite the tumour looking fully resectable and the N2 nodes being negative on frozen sections, the cytology turns out to be positive. As stipulated by standard protocols, the surgeon decides against resection. The patient undergoes chemo-radiotherapy, surviving 10 months. Considering the poor prognosis, you check the literature to see whether proceeding with the resection could have improved his prognosis.

SEARCH STRATEGY

Medline 1950 to May 2012 using OVID interface [resection] AND [NSCLC.mp OR exp non-small cell lung cancer/] AND [MPE.mp OR exp malignant pleural effusion/].

SEARCH OUTCOME

Forty papers were found using the reported research strategy. Six papers covering patients with malignant pleural nodule (MPN) but not MPE were excluded. Finally, eight papers were identified that provided the most applicable evidence to answer the question. They are presented in Table 1.

Table 1:

Best evidence papers

Author, date & country, journal Study type (level of evidence)	Patient group	Outcomes	Key results						Comments
Okamoto et al. (2012) Japan [2];Eur J Cardiothorac Surg Retrospective single-centre case series (level 3)	Between 1990 and 2007, 73 patients with pleural involvement were evaluated Subgroups MPE (n = 32) MPN (n = 41) Distant metastases (M1b) (n = 25) Surgery Lobectomy (n = 73) Pneumonectomy (n = 5) Bilobectomy (n = 11) Partial resection resection (n = 9) Other treatment Intrapleural chemotherapy: 50% Postoperative systemic chemotherapy (mitomycin-C or cisplatin): 33%	MST (months)	Variables	n	MST	P			M1a resected patients had better survival than M1b patients Among M1a patients, surgery may be indicated for MPE patients with N0–1 status
			M1a/M1b N0–1/N2–3 MPE-N0–1/MPE-N2–3 MPN-N0–1/MPN-N2–3 Others/Pn	73/25 7/39 14/18 20/21 70/3	25.9/8.7 37.7/24.4 75.4/24.4 33.7/24.1 26.1/12.8	0.014 0.004 0.011 0.15 0.001
Ichinose et al, (2001) Japan [3];Surgery Today Retrospective multicentre study (level 3)	Between January 1985 and December 1994, 227 patients were evaluated Subgroups MPE (n = 55) MPN (n = 89) MPE + MPN (n = 83) Surgery Resection (R) Lobectomy (n = 139) Pneumonectomy: (n = 29) Limited resection (n = 25) Exploratory thoracotomy (ET) (n = 34) Other treatment Intrapleural chemotherapy (cisplatin; adriamycin; mitomycin): 44% Intrapleural sclerosing agent: 3% Postoperative chemotherapy (platinum based combination): 47% Radiotherapy with or without chemotherapy: 9%	YST (%)	Variables	n	3-YST	5-YST	P		Surgery is associated with longer survival in case of histology of adenocarcinoma N0 status and freedom from macroscopical residual tumour
			R/ET Ad/oth. pN0/1/2 MRT−/MRS+	193/34 178/49 58/31/97 155/61	29/11 29/16 47/35/14 31/13	15/0 14.5/9 28/15/5 18/6	0.04 0.004 <0.0001 0.001
Fukuse et al. (2001) Japan [4]; Lung Cancer Retrospective single-centre case series (level 3)	From January 1981 to December 1997, 49 patients were evaluated	Operative mortality	1 patient (2%)						Surgery may be indicated in patients with T1-2 primary tumour and with MPE only
	Subgroups MPE (n = 16) MPN (n = 17) MPE + MPN (n = 16) Surgery Complete resection (CR): Lobectomy with/or without Pl (n = 27) Pleuropneumonectomy (n = 5) Partial resection (PR) (n = 7) Exploratory thoracotomy (ET) (n = 10)	MST (months)	Variables	n	MST	P
			CR/PR/ET MPE/MPN/ MPE + MPN T1/2/3 pN0/N2	32/7/10 16/17/16 9/28/12 –	37.9/23.2/6.2 58.8/10a/19.3b 37/15c/9.9d 23/10.4	– a0.0001; b0.019 c0.0004; d0.011 0.09
	Other treatment Intrapleural chemotherapy (cisplatin; adriamycin; mitomycin; alone or in combination): 100% Postoperative chemotherapy: 100% (cisplatin in 41%) Adjuvant radiotherapy: 6%	YST (%)		All patients	MPE	MPN	MPE + MPN
			3-YST 5-YST	26.7 15.7	60 45	6.3 0	9.1 0
Shiba et al. (2001) Japan [5];Ann Thorac Surg Retrospective single-centre case series (level 3)	From 1985 to 1995, 65 patients were valuated Subgroups MPE or D0 (n = 25) MPN with/or without MPE or D1 (n = 40) Surgery Lobectomy (L) (n = 55) Partial resection (PR) or exploratory thoracotomy (ET) (n = 10) In all specimens Ki-67, a tumour proliferative marker was evaluated Other treatment Intrapleural chemotherapy (mytomicin-C): 100%	YST (%)	Variables	n	5-YST	P			Surgery appears to be beneficial in patients with N0 status and low Ki-67 index
			L/PR + ET D0/D1 Ad/oth. N0/N1-2 Ki-67low/Ki-67high	55/10 25/40 58/7 11/46 21/36	14/0 22.9/8.9 13.2/0 46.7/5.8 28.6/4.1	– 0.45 0.02 0.01 <0.0001
Kodama et al. (1993) Japan [6]; Cancer Retrospective single-centre case series (level 3)	From April 1985 to December 1991, 31 patients were valuated Subgroups MPE: 6 MPN with or without MPE: 25 (7 clinically diagnosed) Surgery Pleuropneumonectomy (n = 5) Pneumonectomy (n = 3) Lobectomy (n = 18) Segmentectomy or wedge (n = 5) Other treatment Intrathoracic chemotherapy (cisplatin) with radio frequency hyperthermia: 100%	MST (months) YST (%)	Variables	n	MST	3-YST	5-YST	P	Surgery may be indicated in patients with N0 status
			N0–1/N2	14/17	43/16	68.4/22.7	48.7/0	0.01
Shigemura et al. (2003) Japan [7]; Interact CardioVasc Thorac Surg Retrospective single-centre case series (level 3)	Five patients with MPE + MPN were evaluated Four were clinically diagnosed	Operative mortality MST (months)	0%						Multimodality treatment seems to be effective for radical local tumour control in very selected cases
	Surgery Panpleuropneumonectomy (n = 5)		19
	Other treatment Intrapleural hyperthermic chemotherapy before and after operation (cisplatin): 100% Postoperative systemic chemotherapy: 100%		The poorest survival was 8 months (T4N2M0) and the longest was 32 months (T4N0M0)
Kimura et al. (2010) Japan [8]; Interact CardioVasc Thorac Surg Retrospective single-centre non-consecutive case series (level 3)	From May 2001 to July 2005 and from October 2006 to November 2008, 19 patients with pathologically MPE with/or without MPN were evaluated	Operative mortality	0%						Intrapleural hyperthermic therapy alone or with chemotherapy might be beneficial in the prevention of pleural effusion
	Subgroups Group A (n = 7): Intraoperative intrathoracic hyperthermotherapy Group B (n = 5): Intraoperative intrathoracic hyperthermo-chemotherapy Group C (n = 7): no additional therapy Each patient received chemotherapy after surgery (different regimen) Surgery Lobectomy (n = 10) Segmentectomy or wedge (n = 6) Probe thoracotomy (n = 3) all classified in Group A	MST (months) Recurrence of pleural effusion (RPE) (%)	Variables	Group A	Group B	Group C	P
			MTS RPE	19.4 0	41 20	25 57*	NS *0.02
Sawabata et al. (2002) Japan [9]; Ann Thorac Surg Retrospective single-centre case series (level 3)	Between 1980 and 1994, 43 patients were evaluated Subgroups MPE (n = 22) MPN (n = 21) Surgery Complete resection (CR) (n = 11) Incomplete resection (IR) (n = 14) Exploratory thoracotomy (ET) (n = 18) No chemotherapy	MST (months) YST (%)	Variables	n	MST	5-YST	P		Surgery is not beneficial
			CR/IR/ET	11/14/18	13/34/17	9/10/0	–

^aMPN vs MPE.

^bMPE + MPN vs MPE.

^cT2 vs T1.

^dT3 vs T1.

*Group C vs Group A.

RESULTS

Okamoto et al. [2] reported 73 patients staged as M1a (32 had MPE and 41 MPN) and 25 as M1b. The median survival time (MST) of resected M1a vs M1b patients was 25.9 vs 8.7 months, respectively, P < 0.01. Fifty percent of patients had intrapleural chemotherapy; 33% also received adjuvant chemotherapy. The MST of MPE patients with N0–1 status (75.4 months) was better than of MPE with N2–3 (24.4 months; P = 0.011), of MPN with N0–1 (33.7 months) and of MPN with N2–3 (24.1 months). Pneumonectomy had worse MST than lesser resection (12.8 vs 24.1 months, respectively, P = 0.0018).

Ichinose et al. [3] enrolled 227 patients: 55 had MPE, 89 MPN and 83 MPE = MPN. Resection (n = 193) vs exploratory thoracotomy (n = 34) had better 3-year survival time (3-YST) (28.8% vs 10.9%, respectively) and 5-year survival time (5-YST) (14.9% vs 0%, respectively; P = 0.04). Six modalities of chemotherapy were registered. Histology of adenocarcinoma (P < 0.028), N0 status (0.006) and complete resection (0.045) were favourable prognostic factors.

Fukuse et al. [4] evaluated 49 patients: 16 presented MPE, 17 MPN and 16 MPN + MPE. Resected patients vs exploratory thoracotomy had a longer survival (37.9 vs 6.2 months, respectively). All patients received intrapleural and adjuvant chemotherapy. The MST of MPE patients (58.8) was better than MPN (10; P = 0.0001) and than MPE + MPN (19.3; P = 0.019) independently of other factors.

Shiba et al. [5] evaluated 65 patients; 25 had MPE (microscopic pleural involvement or D0) and 40 MPN with/or without MPE (macroscopic involvement or D1). A longer 5-YST was found in resected vs exploratory thoracotomy patients (14.3 vs 0%, respectively) and in D0 vs D1 patients (22.9 vs 8.9%, respectively, P = 0.45). All patients had only intrapleural chemotherapy. Histology of adenocarcinoma (P = 0.02), N0 status (P = 0.01) and low Ki-67 expression (P < 0.0001) were favourable prognostic factors.

Kodama et al. [6] operated on 31 patients with pleural involvement; six had MPE and 25 MPN with/or without MPE. Of these, seven were clinically diagnosed. After surgery, all patients received intrathoracic hyperthermic chemotherapy. N0–N1 vs N2 status had better 3-YST (68.4 vs 22.7%, respectively, P < 0.01) and 5-YST (48.7% vs 0%, respectively).

Shigemura et al. [7] operated on five patients with MPE with/or without MPN (four clinically diagnosed). All patients received a multimodality treatment with intrapleural hyperthermic chemotherapy and panpleuropneumonectomy. The MST was 19; the poorest survival 8 months (T4N2M0) and the longest 32 months (T4N0M0).

Kimura et al. [8] operated on 19 patients with MPE with/or without MPN. Before operation, seven patients received intrapleural hyperthermotherapy (Group A), five intrapleural hyperthermo-chemotherapy (Group B) and seven no additional therapy (Group C). All patients received systemic adjuvant chemotherapy. No survival difference was found between groups; Group C presented an earlier and higher recurrence of PE (57%) than Group B (20%) and Group A (0%; P = 0.02).

Sawabata et al. [9] found that surgery was not beneficial for MPE patients, also in the absence of pleural dissemination. The MST and 5-YST, respectively, were 13 and 9% for completely resected patients (n = 11), 34 and 10% for incompletely resected patients (n = 14; P = 0.3) and 17 and 0% for exploratory thoracotomy patients (n = 18; P = 0.8). No patient received adjuvant treatment, which may partially explain the poor prognosis.

CLINICAL BOTTOM LINE

The management of patients with positive cytology of PE is palliative [10]. The limitations of this analysis included the lack of retrospective studies, patient selection bias (enrolling patients with MPE only, patients with MPN and/or MPE + MPN) and various treatments applied. Three papers found that surgery vs exploratory thoracotomy had a survival advantage in cases of minimal pleural disease [2, 4, 5]. One paper showed that MPE patients had a better MST (58.8 months) than MPN (10 months; P = 0.0001) and MPE + MPN patients (19.3 months; P = 0.019), independently from other factors [4]. Another showed that MPE patients with N0–1 status had an MST longer than patients with similar or more extensive pleural involvement but with N2–N3 status [2]. A further study demonstrated a 5-YST benefit of MPE vs MPN patients (22.9 vs 8.9%; respectively, P = 0.45) [5].

In two papers, surgery vs exploratory thoracotomy had better survival in case of N0 status and complete tumour resection [3, 6]; different pleural dissemination was not analysed.

Different strategies, including pneumonectomy and lobar or lesser resection, were employed to obtain freedom from macroscopic residual tumour associated with pleurectomy in patients having pleural dissemination. Only one paper reported a worse MST after pneumonectomy vs more limited resections (12.8 vs 24.1 months; respectively; P = 0.0018) [2]. In the remaining papers, no comparison between different resections was attended [3–6, 8]. Only three studies included the operative mortality [4, 7, 8]. In all studies [2–8], with the exception of one [9], surgery was a component of multimodality treatment. Intrapleural chemotherapy was generally applied [2–8], with systemic adjuvant chemotherapy [2–4, 7, 8] and/or radiotherapy [3, 4]. The study period and/or year of publication of most papers was 10 years or more, resulting in different chemotherapy regimens adopted [2–8]. No current guidelines support surgery over conservative therapy and the identified studies in this review are not strong enough to change this recommendation.

Conflict of interest: none declared