Extended Sleeve Lobectomy: One More StepToward Avoiding Pneumonectomy in CentrallyLocated Lung CancerJean-Philippe Berthet, MD, Marina Paradela, MD, Maria Jose Jimenez, MD, PhD,Laureano Molins, MD, PhD, and Abel G�omez-Caro, MD, PhDGeneral Thoracic Surgery Department, Hospital Clinic, Universidad de Barcelona, Barcelona, Spain; and U1046, INSERM, Universit�eMontpellier 1, Universit�e Montpellier 2, Montpellier, France

Background. The purpose of this study was to evaluatesurgical outcomes of extended sleeve lobectomy(ESL) in centrally located non–small-cell lung cancer(NSCLC), sparing lung tissue and aggressively avoidingpneumonectomy.

Methods. Patients who underwent ESL betweenJanuary 2006 and January 2013 were included prospec-tively. An atypical bronchial anastomosis was used forsleeve lobectomy involving additional lobes or segments.

Results. We included 27 patients, aged 62.7 ± 8.2 years(range, 49–83 years), with a forced expiratory volume in1 second (FEV1) of 2.27 ± 0.6 (range, 1.6–2.7). According tothe Okada classification, 16 cases were type A (right up-per lobe D middle lobe ± segment 6), 7 cases were type B(left upper lobe D segment 6), and 2 cases were type C(left lower lobe D segments 4–5); we additionally classi-fied 2 patients with right lower lobe tumors involving theright main bronchus as type D (right lower lobe D mid-dle lobe). Anastomosis was performed between the rightsuperior and right main bronchial stumps. Eleven

Accepted for publication July 1, 2013.

Address correspondence to Dr Berthet, Department of Thoracic Surgery,Hospital Clinic, 170 Villarroel, Barcelona, Spain; e-mail: jeanphilippe.berthet@gmail.com.

� 2013 by The Society of Thoracic SurgeonsPublished by Elsevier Inc

patients underwent combined pulmonary angioplasties.Complete resection was achieved in all cases. There wereno operative deaths. Mean segment reimplantation was4.5 ± 0.84 (range, 3–6), resulting in a mean FEV1

improvement of 0.620 ± 0.16 (right-sided ESL) and 0.393 ±0.21 (left-sided ESL). The complication rate was 25% (noimmediate anastomosis-related complications; 1 case ofdelayed bronchial stenosis). No local recurrence was re-ported. At 6 months, mean FEV1 was 1.5 ± 0.4 (right-sidedESL) and 1.4 ± 0.3 (left-sided ESL). Mean follow-up timewas 28 ± 19 months (range, 7–72 months). Overall 5-yearsurvival was 62%.Conclusions. In patients with centrally located NSCLC,

lung-sparing ESL, whose safety and reliability rival thatof pneumonectomy, should be considered. Functionaleffectiveness is higher with right-sided than with left-sided ESL.

(Ann Thorac Surg 2013;-:-–-)� 2013 by The Society of Thoracic Surgeons

esections of centrally located non–small-cell lung

Rcancer (NSCLC) are frequently related to massiveparenchymal extirpation and poor prognosis because ofhigh rates of interlobar and mediastinal lymph nodemetastases. Pneumonectomy causes a substantial declinein lung function and quality of life, precluding adjuvanttreatment [1, 2] or further lung resection when NSCLCdevelops in the remaining lung. In centrally locatedNSCLC, surgical intervention is thus a double-edgedsword for multidisciplinary teams, raising doubts aboutthe role of pneumonectomy in patient prognosis andhealth benefits.

Several teams performing pulmonary surgical pro-cedures [1, 3] have developed aggressive parenchyma-sparing policies, with a reported pneumonectomy tosleeve resection ratio of at least 1:2.5. Classic sleeveresection involves the resection of 1 lobe and end-to-endbronchial anastomosis. Management of centrally locatedNSCLC may combine various surgical techniques to

avoid pneumonectomy, including pulmonary arteryreconstruction and, in some cases, resection of more than1 lobe and airway anastomoses in segmental bronchi[4, 5]. When parenchymal sparing is technically feasibleand R0 is possible but resection involving more than1 lobe is required for oncologic or functional reasons (eg,tumors involving fissures or a section of the segmentalpulmonary arteries of the remnant lobe, respectively),extended sleeve lobectomy (ESL) may represent a last-ditch effort to avoid pneumonectomy. Three series ofESL have been reported so far, by Okada and associates[6], Chida and colleagues [7] and Yamamoto and co-workers [8], and have demonstrated feasibility. Here wedescribe our institutional experience with ESL in 27 pa-tients with centrally located NSCLC and assess our sur-gical results in terms of lung function, postoperativecomplications, and long-term survival.

Patients and Methods

Patient PopulationData were analyzed in 763 consecutive major lung re-sections for NSCLC between January 2006 and January

0003-4975/$36.00http://dx.doi.org/10.1016/j.athoracsur.2013.07.011

2 BERTHET ET AL Ann Thorac SurgEXTENDED SLEEVE LOBECTOMY 2013;-:-–-

2013. Sleeve resection was considered the first option toavoid pneumonectomy in all cases if it was possible toperform complete local resection that resulted in suffi-cient spared parenchyma to occupy the thoracic cavity.Patients who underwent ESL, defined as an alternativeprocedure to resect more than 1 lobe using an atypicalextended bronchoplasty, were selected from this series.These patients had a large central NSCLC (cT2b/T3)located at the origin of the main bronchus, invasive N1disease, or extended circumferential submucosal endo-bronchial spread of a tumor initially staged as cT1/cT2a.Resections included right upper þ middle bilobectomy �segment 6, left upper lobectomy with segment 6, and leftlower lobectomy with segments 4 and 5, which were thenprospectively assigned to Okada types A, B, and C,respectively, according to the level of the bronchialanastomosis [6]. Bronchial anastomoses between the rightsuperior bronchial and the right main bronchial stumpsafter the resection of a tumor originating in the rightinferior lobe and involving the right main bronchus werealso included, and were classified as type D (Fig 1).

Fig 1. Modified Okada classification. (LLL ¼ left lower lobe; ML¼ middle

Central NSCLCs with positive resection margins orremnant lung tissue that was insufficient or too unhealthyto fill the thoracic cavity were considered unsuitable forESL.The institutional review board approved the study,

and written informed consent was obtained for allpatients. Pathologic TNM staging was performed ac-cording to the 2009 International Staging System forNSCLC.

Preoperative AssessmentLung cancer staging included examination; chest roent-genologic analysis; bronchoscopy; thoracic, abdominal,and cerebral computed tomography (CT); and 18F-fluo-rodeoxyglucose positron emission tomography. Thesuggestion of N2 disease on CT or 18F-fluorodeox-yglucose positron emission tomography (short axis>1cm and standardized uptake value >2.5, respectively)was confirmed by mediastinal histologic examinationusing the least invasive method. In N2 disease, simul-taneous neoadjuvant platinum-based chemotherapy and

lobe; RLL ¼ right lower lobe; RUL ¼ right upper lobe.)

3Ann Thorac Surg BERTHET ET AL2013;-:-–- EXTENDED SLEEVE LOBECTOMY

radiotherapy, 45 to 60 Gy, were delivered. Patients with aclinical response seen on computed tomographic scanwere considered operable. In case of persistent N2 dis-ease at mediastinal restaging, surgical intervention wasruled out.

Demonstration of functional tolerance of pneumonec-tomy was required before any attempt at ESL [9]. Thepredicted postoperative forced expiratory volume in1 second (FEV1) was calculated according to the19-segment method [10] or by scanning with isotopeswhen needed.

Systematic bronchoscopy was performed in the oper-ating theater by the surgeon to verify intraluminal tumorspread to macroscopically anticipate potential sites foranastomosis [11].

An intercostal flap was harvested and preserved duringstandard posterolateral thoracotomy through the fifthintercostal space. The technical and oncologic feasibilityof ESL was determined in situ by the surgeon during theprocedure. Once a technique for bronchoplasty wasconsidered, circumferential resection of the bronchuswith or without the pulmonary artery was performedwith a cold scalpel. The distal bronchotomy always fellclose to the origin of the segmental bronchi. Bronchialand arterial margins were routinely assessed by frozensection analysis [12]. Larger resections or completepneumonectomies were considered when faced withfurther submucosal NSCLC spread or macroscopic find-ings that were inconsistent with the viability of the sparedlung or anastomotic sites. If needed, vascular recon-struction was performed as previously described [2], al-ways after bronchial anastomosis.

Bronchial AnastomosisDuring airway exposure, the proximal bronchial arterieswere carefully preserved and a minimal distal bronchialstump dissected, avoiding electrocoagulation. Caliberdiscrepancies between proximal and distal bronchialstumps were technically amended: the proximal stumpdiameter was reduced by inserting 4-0 absorbablemonofilament stitches (PDS II; Ethicon Inc, Somerville,NJ) through the membranous portion and adjacent endsof the stump’s cartilaginous ring to achieve plication and

substantial narrowing (Fig 2). These end-to-end anasto-moses were carried out using a hybrid technique. Acontinuous running 4-0 absorbable monofilament suturewas applied from the cartilaginous membranous junctureto the middle of the cartilaginous wall (Fig 2). The rest ofthe anastomosis was performed using interrupted su-tures. Each suture was inserted through the full thicknessof the bronchial wall, and all knots were tied outside.Traction sutures inserted into the stumps were gentlypulled to reduce tension and were also used as landmarksto prevent bronchial twisting. The pulmonary ligamentswere released. In ESL types A and B, a U-shaped peri-cardial incision was made just below the pulmonary veinsto minimize tension in the anastomoses. The anastomoseswere checked for air leaks (airway pressure of 30 mm Hgafter filling the thoracic cavity with serum). The bronchialsuture line was routinely wrapped with the intercostalflap.

Management of the Remnant ParenchymaWhen segmentectomy was required (types A, B, and C),the intersegmental plane was identified, with or withoutjet ventilation to inflate the segment to be resected. Thesegment was removed en bloc while developing theintersegmental plane with electric scissors, sequentiallydividing individual vein branches. No mechanical staplerwas used. The raw lung parenchymal surface waschecked for bleeding and air leaks and was reinforcedwith pulmonary sealant.

Postoperative Care, Anastomosis Control, and SurvivalThe discrepancy between the pleural space and sparedsegments was managed by adjusting drainage time andsuction intensity (applied gently during mechanicalventilation and discontinued as soon as possible).A bronchoscopy was performed in the operating the-

ater to check the anastomosis. A routine bronchoscopywas performed on the seventh postoperative day orbefore discharge. Bronchoscopic data were used to clas-sify bronchial anastomosis [13]. Routine bronchoscopy(eg, in the absence of symptoms or abnormalities on chestroentgenogram) was scheduled at 3 and 6 months post-operatively. To evaluate pulmonary function, FEV1 was

Fig 2. (A and B) Bronchial anastomosisusing a modified end-to-end technique.

4 BERTHET ET AL Ann Thorac SurgEXTENDED SLEEVE LOBECTOMY 2013;-:-–-

measured more than 3 months (10.4 � 3.4 (range,4–21 months) after operation.

Statistical AnalysisSPSS software was used for analysis (SPSS Inc, Chicago,IL). All results are expressed as means � standard de-viations. Operative mortality included all deaths relatedto the surgical procedure regardless of postoperative in-tervals. No patient was lost to follow-up, and all wereincluded in the survival analysis. Survival probabilitieswere calculated using the Kaplan-Meier method.

Results

Patient ProfilesBetween January 2006 and January 2013, 27 patients witha mean age of 62.7 � 8.2 years (range, 49–83 years) un-derwent ESL. They represented 26% of the 101 broncho-plastic lobectomies performed during this period.

Eleven patients were older than 70 years, 19 weresmokers, and 12 had a history of pulmonary disease. Thefeasibility of ESL was predicted preoperatively in 20 (74%)patients by CT and was considered at the time of opera-tion in the 7 (26%) other patients. During the same period,16 pneumonectomies were finally performed despiteattempted ESL because of positive bronchial margins infrozen sections, but none was performed because oftechnical issues.

Mean FEV1 and FEV1 percentage were 2.2 � 0.59 L and72.4% � 14.8%, respectively; 8 patients had an FEV1 ofless than 60%, but all patients met the functional criteriato tolerate pneumonectomy [8].

Preoperative clinical staging identified 6 cIB, 6 cIIA,7 cIIB, 6 cIIIA, and 2 cIIIB tumors. cT4 disease was iden-tified in 3 patients (patients 6, 9, and 16) because ofmediastinal invasion or an associated nodule in themiddle lobe. In 4 patients (patients 12, 15, 17, and 19),cT3 disease was diagnosed because tumor size wasgreater than 7 cm. The mean tumor size was 4.4 � 1.9 cm(range, 1.9–9.4 cm). Preoperative N staging revealed12 cN0, 8 cN1, and 7 cN2 cases; the indication for induc-tion chemotherapy was N2 disease (25.9%). CT after in-duction chemotherapy showed partial responses in allcases. Surgical procedures and TNM scores are summa-rized in Table 1.

Type of Procedure and StagingComplete R0 resection was achieved in all patients. Re-sections were performed on the right side in 71% of pa-tients (Figs 3, 4). According to the modified Okadaclassification, 16 (59%) procedures were type A, 7 (25%)were type B, 2 (7%) were type C, and 2 (7%) were type D.There were 16 (59%) single bronchial reconstructions and11 (42%) associated pulmonary angioplasties. Dependingon pulmonary artery involvement, reconstructionsranged from direct suture and end-to-end anastomosis topulmonary artery replacement by cryopreserved arterialallografts. Double reconstruction was most frequently

carried out in type A procedures (9 of 11). Mean operativetime was 235 � 37 minutes (range, 182–290 minutes).Final staging identified 2 yIA, 4 pIB, 1 yIB, 9 pIIA, 1

yIIA, 4 pIIB, 3 pIIIA, and 3 yIIIA stage cases. Final stage Idisease included patients with a tumor response tochemoradiotherapy and patients with extended sub-mucosal bronchial invasion. A final pN0 stage wasidentified in 7 (26%) patients, pN1 in 17 (63%) patients,and pN2 in 3 (11%) patients, requiring appropriateadjuvant chemotherapy according to current guidelines.According to the Ludwig classification [13], the quality

of bronchial anastomosis on the seventh postoperativeday was satisfactory (grades 1 and 2) in 85.2% of patientsand critical (grade 3) in 14.8% of patients, leading tosystematic antibiotherapy and more closely spacedbronchoscopic controls. No patient had a dehiscentanastomosis.

Follow-UpThere were no intraoperative deaths, and morbidity was25.9%, including major (n ¼ 4) and minor (n ¼ 3) com-plications. All patients required an intensive care unitstay (2.7 � 0.4 days (range, 1–12 days) and the duration ofhospitalization was 10.3 � 2.2 (range, 7–26 days). Com-plications were treated medically without any need forreoperation. Four patients experienced complicationsassociated with pneumonia (14.8%), 1 each experiencedsegmental atelectasis, partial pulmonary artery throm-bosis, and prolonged air leaks, and 2 experienced woundinfection (Table 1). Three of these complications weredelayed, including 2 cases of recurrent pneumonia(1 related to partial stenosis of the bronchial anastomosis).The partial pulmonary artery thrombosis (patient 19),which was considered to be related to slight kinking aftertangential suturing, was asymptomatic and successfullytreated by systemic heparinization. The bronchial steno-sis was successfully treated by dilation and laser treat-ment of the stricture.The morphologic and functional sparing of lung pa-

renchyma was successful in all patients, as assessed byfollow-up CT (the color-coded iodine perfusion mapshowed that the spared lobes were well perfused andexpanded). Postoperative FEV1 was 1.50 � 0.4 L and1.41 � 0.3 L for right-sided and left-sided ESL, respec-tively, in disease-free survivors. During follow-up, 2 pa-tients underwent pulmonary resection of the oppositelung because of a solitary pulmonary nodule or a secondNSCLC tumor.At final clinical evaluation in January 2013, 6 patients

had died. Five deaths were related to extrathoracicNSCLC (bone metastasis, n ¼ 4; liver metastasis, n ¼ 1).No evidence of local NSCLC recurrence was found. Meanfollow-up time was 28 � 19 months (range, 7–72 months).The overall 5-year survival rate was 62% (Fig 5) and meandisease-free survival time was 26 � 18 (range,3–72 months). Bivariate Cox analysis did not identify anyvariables studied (age > 70 years, side, N2 disease, sex,histologic type, double reconstruction, neoadjuvanttreatment, low FEV1, pathologic stages I/II versus III/IV)as individual risk factors affecting survival.

Table 1. Patient Characteristics, Surgical Procedure, Tumor-Node-Metastasis (TNM) Staging

Patient Age/SexOkadaClass Sleeve Type

VascularReconstruction

cTNMy/pTNMHistologicType/Stage Complications

HospitalStay (d)

Outcome(mo) Recurrence

1 68/M A RUL þ Seg 6 PA tangential suture cT2b, N0, M0pT2N1M0SCC

Sputum retention 8 Alive47.0

No

2 64/M A RUL þ ML þSeg 6

. . . cT2a, N2, M0yT2aN0M0Adenocarcinoma

. . . 8 Alive36.0

No

3 66/M C LLL þ Seg 4/5 PA tangential suture cT2aN0M0pT2aN1M0SCC

Sputum retentionAtelectasis

8 Alive63.0

No

4 79/M A RUL þ ML PA end-to-endanastomosis

SVC

cT2aN0M0pT2aN2M0SCC

. . . 12 Dead32.0

Distant metastasis

5 49/M A RUL þ ML . . . cT2aN0M0pT2aN0M0SCC

. . . 8 Alive69.0

No

6 62/M A RUL þ Seg 6 . . . cT4N2M0yT1aN0M0Adenocarcinoma

Sputum retentionWound infection

10 Alive39.0

Distant metastasis

7 78/M A RUL þ ML . . . cT2aN0M0pT2aN1M0Adenocarcinoma

. . . 9 Dead17.0

Distant metastasis

8 66/M A RUL þ ML . . . cT2aN2M0yT2aN1M0SCC

Sputum retentionEarly pneumoniaWound infection

11 Dead26.0

Distant metastasis

9 65/F A RUL þ Seg 6 PA tangentialsutureSVC

cT4N0M0pT4N1M0Adenocarcinoma

. . . 11 Alive39.0

No

10 72/M B LUL þ Seg 6 PA end-to-endanastomosis

cT2aN1M0pT1aN1M0SCC

Atrial fibrillationSputum retention

7 Dead32.0

No

11 73/M A RUL þ ML þSeg 6

. . . cT2aN2M0yT2aN0M0yT2aN0M0Adenocarcinoma

. . . 7 Alive18.0

No

12 62/M A RUL þ Seg 6 . . . cT3N0M0pT3N1M0SCC

. . . 8 Alive59.0

No

13 61/M B LUL þ Seg 6 . . . cT2aN0M0pT2aN1M0SCC

. . . 9 Alive34.0

No

14 75/M A RUL þ ML þSeg 6

. . . cT2aN1M0pT2aN1M0SCC

Air leak 9 Dead8.0

Distant metastasisbone

(Continued) 5AnnThorac

Surg

BERTHETETAL

2013;-:-

–-

EXTENDED

SLEEVELOBECTOMY

Table 1. Continued

Patient Age/SexOkadaClass Sleeve Type

VascularReconstruction

cTNMy/pTNMHistologicType/Stage Complications

HospitalStay (d)

Outcome(mo) Recurrence

15 55/F A RUL þ Seg 6 . . . cT3N0M0pT3N1M0Adenocarcinoma

. . . 7 Alive59.0

No

16 56/M C LLL þ Seg 4/5 PA tangential suture cT4N2M0yT4N1M0Adenocarcinoma

. . . 8 Alive7.0

No

17 83/M B LUL þ Seg 6 PA tangential suture cT3N0M0pT3N0M0Adenocarcinoma

Early pneumonia 14 Dead18.0

No

18 73/M D RLL þ ML . . . cT2aN1M0yT2aN2M0Adenocarcinoma

. . . 9No

Alive24.0

No

19 61/M A RUL þ ML PA tangential sutureSVC

cT3N2M0yT3N1M0SCC

Delayed PA partialthrombosis

8 Alive15.0

No

20 67/M B LUL þ Seg 6 . . . cT2aN0M0pT2aN1M0Adenocarcinoma

. . . 7 Alive8.0

No

21 72/M A RUL þ ML PA end-to-endanastomosis

cT2aN0M0pT2aN0M0;IB Submucosal

extended invasionSCC

. . . 7 Alive22.0

No

22 66/M D RLL þ ML . . . cT1bN1M0pT2aN0M0SCC Submucosal

extended invasion

Early pneumoniaWound infection

14 Alive9.0

No

23 62/M B LUL þ Seg 6 . . . cT2aN1M0pT2aN1M0SCC

AtelectasisDelayed bronchialstenosis

12 Alive8.0

No

24 82/M A RUL þ ML Cryopreservedarterial allograft

cT2bN1M0pT2aN1M0SCC

. . . 8 Alive9.0

No

25 69/M A RUL þ ML Cryopreservedarterial allograft

cT2bN1M0pT2bN1M0Adenocarcinoma

. . . 10 Alive13.0

No

26 79/M B LUL þ Seg 6 . . . cT2bN1M0pT2bN1M0SCC

. . . 10 Alive18.0

No

27 62/F B LUL þ Seg 6 . . . cT1bN1M0pT1bN1M0Adenocarcinoma

Early pneumonia 11 Alive24.0

No

LLL ¼ left lower lobe; LUL ¼ left upper lobe; ML ¼ middle lobe; PA ¼ pulmonary artery; RLL ¼ right lower lobe; RUL ¼ right upper lobe; Seg ¼ segment; SCC ¼ squamous cellcarcinoma; SVC ¼ superior vena cava.

6BERTHETETAL

AnnThorac

Surg

EXTENDED

SLEEVELOBECTOMY

2013;-:-

–-

Fig 3. Patient 18, adenocarcinoma involvingbronchus intermedius beyond origin of rightupper bronchus (RUB) at mediastinal face.(A) Computed tomographic scan showingendobronchial tumor involving main bron-chus immediately beneath origin of RUB(arrow), precluding bilobectomy. (B)Segment 6 tumor involving carinal divisionbetween bronchus intermedius and RUB. (C)Reimplantation of RUB into main bronchus(type D) was carried out.

7Ann Thorac Surg BERTHET ET AL2013;-:-–- EXTENDED SLEEVE LOBECTOMY

Comment

According to our results, ESL is feasible and results inlocal disease control similar to that achieved with pneu-monectomy but with mortality and major complicationrates similar to those of classic sleeve lobectomies [8] andsubstantially lower than those for pneumonectomy [2,14–17]. The functional value of the spared lung in ourreport and the analysis of our surgical technique partic-ularly identify right-sided ESL (types A and D) withmodified end-to-end bronchial anastomosis as a reliableoption to avoid right pneumonectomy in selectedindications.

The absence of operative mortality in our study iscomparable to that of other reports so far [6, 7, 8] andsupports the use of ESL to replace pneumonectomy. Themost frequent major complication management proce-dure was reoperation for completion pneumonectomy,and endoscopic treatment of anastomosis-related com-plications (Table 2).

Frequently reported complications of sleeve resectioninclude bronchial anastomosis dehiscence or arterialfistulization, essentially related to different surgical ap-proaches to amend bronchial mismatches. In sleeveresection series [6, 7, 8, 18–20], various bronchial anas-tomosis techniques have been reported to amend caliberand shape discrepancies, and changes in bronchial di-rection between the proximal and distal stumps. In casesof notable diameter mismatch, Hollaus and colleagues[18] applied a telescoping anastomosis technique (n ¼15), bringing the distal bronchus up into the lumen ofthe proximal stump. One-third of patients requirediterative interventional bronchoscopy, and 6% displayedanastomotic dehiscence. To limit the risk of partialdehiscence of a telescoping anastomosis when insertingthe basal or middle bronchus into the right main bron-chus, Kamiyoshihara and associates [19] used a “bron-chofolding” technique, whereas Miyoshi and coworkers[20] modified the telescoping technique by adding

Fig 4. (A) Patient 13, squamous tumorarising from left upper bronchus (LUB) andinvolving segment 6. Computed tomographicscan, view of the tumor extending fromorigin of LUB with endobronchial invasion(arrow) to the origin of the superior segmentof the inferior lobe. (B) Involvement ofcarinal division of both lobes (arrow);proximity of segment 6 seen by endoscopyprecluded anastomosis sparing the entireinferior lobe. (C) Bronchial anastomosisbetween main bronchus and basal segmentsof lower left lobe (type B).

Fig 5. Overall 5-year survival rate.

8 BERTHET ET AL Ann Thorac SurgEXTENDED SLEEVE LOBECTOMY 2013;-:-–-

transverse stitches to the membranous part of the mainbronchus. The thinness and frailty of the distal stumpare never as evident as when trying to adapt a segmentalbronchial cap with a small cylindrical cross section tothe horseshoe-shaped main bronchus, and this is evenmore obvious with the right main bronchus. Since thebeginning of our experience, similar to Okada and co-workers [6], we have adapted the proximal bronchialstump diameter to match the distal one by insertingtransverse stitches through the membranous portion.Both in our series and in the 3 previously reported ESLseries, only telescoping [7, 8] or modified end-to-end [6,current series] anastomoses were used, with slightlydifferent complication rates, but the only reportedtransmural necrosis occurred when using the tele-scoping technique.

Rather than being a choice of surgical teams, the extentand quality of the resection determine the reconstructiontype that meets functional and oncologic requirements.To be reliable, ESL should be associated with R0 resection[21] and requires frozen section analysis of bronchialstumps. Okada and colleagues [6] chose resection pointswith a margin of at least 1 cm of macroscopically unaf-fected bronchus. We have rarely had this opportunitybecause of endobronchial or extrabronchial spreading ofthe tumor. Proximally, we deliberately chose to leave theshortest possible stump to protect bronchial vasculariza-tion [22]. Distal bronchotomies were usually performedclose to the branching point of the first disease-freesegmental bronchus to avoid excessive handling of thesegmental bronchus and preserve pulmonary circulation(accounting for 75% of airway blood supply at segmentallevels). Such aggressive management is unfeasible withtelescoping anastomosis, and we speculate that the 100%R0 resection rate and reduced incidence of structuralairway weakness observed in our series and the of Okadaand associates [6] (contrary to the 8.7% to 10% R1 re-ported in other series) simply reflect an effectivecompromise between oncologic and functional issues.

Although the data reported regarding ESL do not lendthemselves to comparisons, Okada type, neoadjuvant

treatment, and combined pulmonary artery angioplastyhave not been associated with increased risk of anasto-motic complications, reoperation, or local recurrence.To permit the complete reexpansion of the spared pa-

renchyma and prevent pleural and bronchial complica-tions, the remaining segments should be sufficient inquantity and quality. In contrast to Chida and associates[7], phrenic palsy was not used to adapt thoracic cavityvolume to the remaining lobe because of functional im-pairments and sputum retention in spared segments. Ourstrategy to manage residual pleural space was (1) pres-ervation of spared segment expansion by meticulousblunt dissection of the intersegmental plane and mainte-nance of reexpansion by regular endoscopic cleaning ofairways, (2) careful intraoperative checking for air leaks,and (3) close postoperative management of chest tubesuction. We consider partial thoracoplasty or pleural tentapproaches to be interesting when the remaining seg-ments are of poor quality, especially in type A with middlelobe and segment 6 involvement, and in type C becausethese are the largest parenchymal resections. However,without implementing these techniques, we did notobserve more bronchopleural complications in these casesin our series.To maintain good lung reexpansion may require the

endoscopic cleaning of airways. The other role of bron-choscopy is to assess the quality of the anastomosis beforethe patient is discharged [13]. Our results agree withthose of Ludwig and colleagues [13], who report no sec-ondary bronchial dehiscence when endoscopy performed7 days after operation revealed satisfactory results.Finally, bronchoscopy allowed us to treat 1 case ofdelayed bronchial stenosis using a less aggressive man-agement strategy than repeated surgery, which wouldlikely have led to completion pneumonectomy.Parenchymal sparing justifies ESL; in our experience

4.5 � 0.84 (range, 3–6) segments were reimplanted,including 5 � 0.7 (range, 3–6) segments on the right sideand 3.8 � 0.42 (3–4) segments on the left side. Reim-plantation of appropriately reexpanded segments leads toa mean FEV1 saving of 0.620 � 0.163 L and 0.393 � 0.124 Lfor right- and left-sided ESL, respectively. Avoidance ofthe high morbidity and mortality rates and poor post-operative function associated with right pneumonectomysuggests that right-sided ESL is particularly more effec-tive in saving the parenchyma (1.6 times more thanleft-sided ESL). Contrary to the studies of Okada andcolleagues, Chida and associates, and Yamamoto andcoworkers [6, 7, 8], a majority of our ESL procedures wereright-sided (63% versus 40%, 30.4%, and 20%, respec-tively), which enhanced overall mean postoperative lungfunction but did not affect long-term survival. Thisproven functional merit of right ESL has led us to add atype D to the Okada classification [6], which could save 3segments (Fig 1).In conclusion, ESL should be among the routine tech-

niques used to avoid pneumonectomy because of itsreliability and effectiveness. However, more aggressivemonitoring of patients is required. ESL techniques arecritical in right-sided lung resections.

Table 2. Bronchial Healing and Local Recurrence According to Type of ESL in Current and Previous Studies

Author (date) n CRT n, %

Type of Sleevea

Associated AngioplastyAnastomosis TypeBronchial Healingb

R0 LocalRecurrence

Early BronchialNecrosis (Grade 5)/Delayed Bronchial

StenosisOperativeMortality

CompletionPneumonectomy Follow-Up (mo)

A B C Atypical

Okada et al.(1999)

15 426

6 4 5 0 End-to-end100%G X

R0 100% . . . . . . . . . 6.6% 9–106

9 LS, 4 LL, 2 LLSCombined arterioplasty 53%

Chida et al.(2009)

23 416

2 4 10 5 Telescoping91%G4 (4.5%)G5 (4.5%)

R0 95.6%8.7% LR

2 (8.7%)/1 (4.3%) . . . . . . 3–36

15 LS, 7 LL, 1 LLSCombined angioplasty 78%

Yamamotoet al. (2008)

20 . . . 5 8 7 0 TelescopingEnd-to-end95%G5 (5%)

R0 90%. . .

1(5%)/ . . . . . . 5% 13–113

18 LS, 0 LL, 2 LLSCombined angioplasty 45%

Current study(2013)

27 726

16 7 2 2 End-to-end G 1/2(85.2%)

G3 (7.4%)G4 (7.4%)

R0 100%. . .

. . ./1 (7.4%) . . . . . . 7–69

13 LS, 10 LL, 4 LLSCombined angioplasty 41.7%

a Type of sleeve according to the Okada classification. Amount of resected parenchyma, (LS ¼ 1 lobe plus segment, LL ¼ 2 lobes, LLS ¼ 2 lobes plus segment). b Grades (G) of bronchial healing reportedaccording to the Ludwig classification.

CRT ¼ chemoradiotherapy; G ¼ grade; LR ¼ local recurrence.

9AnnThorac

Surg

BERTHETETAL

2013;-:-

–-

EXTENDED

SLEEVELOBECTOMY

10 BERTHET ET AL Ann Thorac SurgEXTENDED SLEEVE LOBECTOMY 2013;-:-–-

References

1. Deslauriers J, Tronc F, Gr�egoire J. History and current statusof bronchoplastic surgery for lung cancer. Gen Thorac Car-diovasc Surg 2009;57:3–9.

2. G�omez-Caro A, Garcia S, Reguart N, et al. Determining theappropriate sleeve lobectomy versus pneumonectomy ratioin central non-small cell lung cancer patients: an audit of anaggressive policy of pneumonectomy avoidance. Eur J Car-diothorac Surg 2011;39:352–9.

3. Yildizeli B, Fadel E, Mussot S, Fabre D, Chataigner O,Dartevelle PG. Morbidity, mortality, and long-term survivalafter sleeve lobectomy for non-small cell lung cancer. Eur JCardiothorac Surg 2007;31:95–102.

4. Rendina EA, De Giacomo T, Venuta F, Ciccone AM,Coloni GF. Lung conservation techniques: bronchial sleeveresection and reconstruction of the pulmonary artery. SeminSurg Oncol 2000;18:165–72.

5. Rendina EA, Venuta F, de Giacomo T, Rossi M, Coloni GF.Parenchymal sparing operations for bronchogenic carci-noma. Surg Clin North Am 2002;82:589–609.

6. Okada M, Tsubota N, Yoshimura M, et al. Extended sleevelobectomy for lung cancer: the avoidance of pneumonec-tomy. J Thorac Cardiovasc Surg 1999;118:710–3.

7. Chida M, Minowa M, Miyoshi S, Kondo T. Extended sleevelobectomy for locally advanced lung cancer. Ann ThoracSurg 2009;87:900–5.

8. Yamamoto K, Miyamoto Y, Ohsumi A, et al. Sleeve lungresection for lung cancer: analysis according to the type ofprocedure. J Thorac Cardiovasc Surg 2008;136:1349–56.

9. Brunelli A, Charloux A, Bolliger CT, et al. The ERS and ESTSclinical guidelines for evaluating fitness for radical treatment(surgery and chemoradiotherapy) in patients with lungcancer. Eur J Cardiothorac Surg 2009;36:181–4.

10. Zeiher BG, Gross TJ, Kern JA, Lanza LA, Peterson MW.Predicting postoperative pulmonary function in patientsundergoing lung resection. Chest 1995;108:68–72.

11. Wright CD. Sleeve lobectomy in lung cancer. Semin ThoracCardiovasc Surg 2006;18:92–5.

12. Wain JC. Bronchoplastic resections. In: Kaiser LR, ed.Mastery of Cardiothoracic Surgery. Philadelphia: Lippincott-Raven, 1998:68–76.

13. Ludwig C, Stoelben E. A new classification of bronchialanastomosis after sleeve lobectomy. J Thorac CardiovascSurg 2012;144:808–12.

14. Deslauriers J, Gr�egoire J, Jacques LF, Piraux M, Guojin L,Lacasse Y. Sleeve lobectomy versus pneumonectomy forlung cancer: a comparative analysis of survival and sites orrecurrences. Ann Thorac Surg 2004;77:1152–6.

15. Predina JD, Kunkala M, Aliperti LA, Singhal AK, Singhal S.Sleeve lobectomy: current indications and future directions.Ann Thorac Cardiovasc Surg 2010;16:310–8.

16. Fadel E, Yildizeli B, Chapelier AR, Dicenta I, Mussot S,Dartevelle PG. Sleeve lobectomy for bronchogenic cancers:factors affecting survival. Ann Thorac Surg 2002;74:851–8.

17. Powell ES, Cook D, Pearce AC, et al. A prospective, mul-ticentre, observational cohort study of analgesia andoutcome after pneumonectomy. Br J Anaesth 2011;106:364–70.

18. Hollaus PH, Janakiev D, Pridun NS. Telescope anastomosisin bronchial sleeve resections with high-caliber mismatch.Ann Thorac Surg 2001;72:357–61.

19. Kamiyoshihara M, Ibe T, Takise A, Takeyoshi I. Atypicalbronchoplasty to preserve the lung parenchyma: the bron-chofolding technique. J Thorac Cardiovasc Surg 2008;135:1184–5.

20. Miyoshi S, Tamura M, Araki O, et al. Telescoping bronchialanastomosis for extended sleeve lobectomy. J Thorac Car-diovasc Surg 2006;132:978–80.

21. Yu JA, Weyant MJ. Techniques of bronchial sleeve resection.Semin Cardiothorac Vasc Anesth 2012;16:196–202.

22. Fr�echette E, Deslauriers J. Surgical anatomy of the bronchialtree and pulmonary artery. Semin Thorac Cardiovasc Surg2006;18:77–84.