ate-Onset Occult Pneumothorax After Lungolume-Reduction Surgery
ederico Tacconi, MD, Eugenio Pompeo, MD, and Tommaso C. Mineo, MD
ivision of Thoracic Surgery, Policlinico Tor Vergata University, Rome, Italy
Tarbvcpgl
ppseutf
Background. Lung volume-reduction surgery hasroved to be a reliable palliative surgical treatment foratients with severe emphysema. Nonetheless, late com-lications can arise after lung volume-reduction surgerylthough this matter has been poorly investigated byrevious studies.Methods. We report a series of 6 patients undergoing
nilateral lung volume-reduction surgery at our institu-ion between October 1995 and December 2004, who wereeadmitted several months after discharge because of theccurrence of occult pneumothorax mimicking acute re-piratory failure.
Results. Occult pneumothorax occurred in 3.3% of the82 patients treated with lung volume-reduction surgeryt our institution. Patients were readmitted after a meanf 94 days (range, 20 to 700 days) from the discharge.hest roentgenography was unable to detect the occur-
ence of pneumothorax, which was instead revealed by
eans of a computed tomographic scan in all patients.
2005 by The Society of Thoracic Surgeonsublished by Elsevier Inc
ats.ctsnetjournDownloaded from
he interval between admission and correct diagnosisveraged 22.4 hours. The number of air collectionsanged between two and four. Treatment entailed solelylind chest drainage placement in 2 patients and awakeideo-assisted thoracoscopic surgery in the others, in-luding placement of chest tube under direct vision in 1atient, repair of lung tears by means of cyanoacrylatelue in 2 and bovine pericardium patch plus cyanoacry-ate glue apposition in 1 patient.
Conclusions. In conclusion, we believe that occultneumothorax should be kept in mind as one of theossible late complications of lung volume-reductionurgery and should be suspected whenever sudden wors-ning of dyspnea is noticed even in the presence of anneventful chest roentgenogram. Awake video-assisted
horacoscopic surgery management can represent an ef-ective option in these instances.
ung volume-reduction surgery (LVRS) has proved tobe a reliable palliative surgical treatment for pa-
ients with heterogeneously distributed severe emphy-ema. Nonetheless, several late complications can occurfter LVRS including development of giant bullous em-hysema [1] and the occurrence of secondary pneumo-
horax. This latter in particular has been poorly investi-ated by the previous literature. More to the point, weave found that loculated air collections can occur afterVRS even several months after discharge, leading toevere impairment of respiratory function. Prompt rec-gnition of LVRS-related complications is crucial to avoidcute respiratory failure, which can be associated with aigh mortality rate in these delicate patients. Unfortu-ately, detection of this complication can be a difficultrst-line assessment owing to the presence of postoper-tive pleural adhesions resulting in complex summationrtifacts on standard chest roentgenogram. Herein, weescribe peculiar clinical features and surgical manage-ent of occult loculated pneumothorax occurring in a
arge single-center consecutive series of patients under-oing LVRS.
ccepted for publication June 7, 2005.
ddress correspondence to Dr Pompeo, Cattedra di Chirurgia Toracica,
atients and Methods
e reviewed a series of 6 patients undergoing stapledVRS at our institution between October 1995 and De-ember 2004, and presenting several months after dis-harge as critical care patients because of the occurrencef secondary pneumothorax.All patients had advanced smoking-related upper-lobe
revailing emphysema with severe airflow obstructionnd marked air trapping (mean forced expiratory volumen 1 second, 33.5%; mean residual volume, 210.8% ofredicted). No patient in this series had giant bullae, ho-ogeneous emphysema, or homozygous �1-antitrypsin de-
ciency. All patients also met standard inclusion criteria forVRS including smoking cessation of at least 6 months,oor expectoration, and absence of cor pulmonale andther concomitant diseases affecting the outcome. Addi-ional details regarding our inclusion criteria and preoper-tive assessment are described elsewhere [2]. Five pa-ients underwent unilateral videothoracoscopic LVRS,nd 1 underwent staged bilateral operation. The mostamaged portions of the lung were identified by visual
nspection as the regions retaining inflation after venti-atory exclusion. Both computed tomographic (CT) andcintigraphic scan findings were also taken into theccount. The targeted areas were resected by endoscopictaplers (Endopath 45; Ethicon Endosurgery, Pomezia,
taly). In the first 3 patients, suture lines were reinforced
2009Ann Thorac Surg TACCONI ET AL2005;80:2008–12 OCCULT PNEUMOTHORAX AFTER LVRS
GEN
ERA
LT
HO
RA
CIC
ith the use of bovine pericardial strips (Peristrips;iovascular Inc, St. Paul, MN). After resection, carefulssessment of pneumostasis was performed. If majoreaks were identified, they were ablated by additionalno-cut” stapling (Endopath NK45B; Ethicon Endosur-ery). Two chest tubes were inserted at the end of therocedure and placed under water-seal. No patient in
his series required additional suction at the end of therocedure. After operation, 4 patients were dischargedfter complete cessation of air leaks (days 7, 10, 12, and5). Two patients were discharged with a Heimlich valvewing to small persisting air leaks, and were controlleds outpatients thrice weekly until air leak cessation wasnsured (postoperative days 18 and 21). In both groups,riteria for chest drainage removal were radiologicallyocumented complete lung reexpansion and the absencef bubbling in the drainage chamber. After tube removal,atients underwent first follow-up within 7 days entail-
ng solely physical examination, and complete reassess-ent within 15 days including respiratory function tests,
-minute walking test, blood-gas assay, and in-expiratoryhest roentgenogram. Four patients underwent postop-rative respiratory rehabilitation performed either on anutpatient or on an inpatient basis, depending on indi-idual preference.
esults
he study cohort (all men, mean age, 66.4 years) accountsor 3.3% of the 182 patients treated by LVRS at ournstitution. Patients were readmitted after a mean of 168ays from discharge (range, 20 to 700 days; median, 94
ats.ctsnetjournDownloaded from
ays). In this subgroup, mean air leak duration afterVRS was 13.8 days (range, 7 to 21 days), a number thateaches statistical difference when compared with theverage duration observed in our historical cohort (5.2ays; Mann-Whitney U test, p � 0.00009).Four patients arrived directly at the emergency depart-ent of our institution because of acute respiratory
ailure. The other 2 patients were referred at our institu-ion after preliminary admission to the emergency de-artment of a peripheral hospital. Both these subjectsrrived with a misdiagnosis of exacerbated chronic ob-tructive pulmonary disease and deteriorating bullousmphysema, respectively.In all patients, the acute scenario entailed a marked
mpairment in subjective dyspnea and severe hypoxia inhe absence of fever, cough, sputum, and leukocytosis.hree patients also had cardiovascular symptoms (ie,
achycardia and hypotension). Hypercarbia was presentn 2 patients.
Preliminary chest roentgenography showed signs ofnilateral lung hyperinflation with flattening of the dia-hragmatic dome and slight mediastinal shift (2 pa-
ients), and a slight hypertranslucent area in the involvedemithorax (3 patients). In 2 patients, chest roentgeno-ram was unremarkable. All patients’ CT scan was per-ormed after a mean of 22.4 hours from admission andevealed definitive diagnosis (Figs 1, 2). Figure 3 depictshe fall in arterial oxygenation (expressed as arterialartial pressure of oxygen to fraction of inspired oxygenatio) per patient in the interval between initial admis-ion and definitive diagnosis.
Three patients underwent initial blind chest drainage
Fig 1. Figure shows roentgenogram (left) andcomputed tomographic (right) feature of aloculated pneumothorax occupying almost45% of the left hemithorax. Despite the largesize revealed by computed tomography, chestroentgenogram showed solely a minimal hy-perlucency area on the left upper field.
Fig 2. Tomographic feature of a very largeright pneumothorax (right), which was mim-icking lung hyperinflation at scout scanogram(left; slight leftward tracheal shift and com-pression of the upper lobe vessels).
2010 TACCONI ET AL Ann Thorac SurgOCCULT PNEUMOTHORAX AFTER LVRS 2005;80:2008–12
GEN
ERA
LT
HO
RA
CIC
lacement, which allowed definitive resolution of theneumothorax in 2 patients. One patient received firstlind drainage placement and was subsequently sched-led for video-assisted thoracoscopic surgery because ofersisting air collection. Three patients were immedi-tely scheduled for video-assisted thoracoscopic surgeryecause of the presence of multiple loculated pneumo-
horaces whose treatment by blind drainage was deemedangerous. To avoid the risks of general anesthesia andechanical ventilation, all the procedures were per-
ormed with the patient awake, under spontaneous ven-ilation. Analgesia entailed intercostal nerve blockadeith a mix of lidocaine 2% and ropivacaine 7.5%. Mild
edation was obtained with 2 to 5 mL of intravenousidazolam. No patient required intubation or suffered
unctional impairment throughout the procedure. Two5-mm trocar ports were sufficient for the operation in allatients but 1, who required an additional 7-mm portecause of particular anatomic complexity. The air leakas easily discovered in 5 patients. In 1 patient the leakas not found, and operative technique entailed only
ideo-assisted placement of a chest drainage. Sealing of amall lung tear lying in the vicinity of the suture line byeans of cyanoacrylate glue instillation was carried out
n 2 patients. In 1 patient, a gross bronchoalveolar leakying on the major fissure was repaired by means of ayanoacrylate layer covered with a bovine pericardiumatch. Operative time ranged from 35 to 90 minutes. Air
eak cessation was achieved after 1 to 3 days; meanospital stay was 3.2 days. Follow-up ranged from 6 to 36onths, with no patients lost to follow-up. No patient
eveloped a second pneumothorax during follow-up.ables 1 and 2 report major clinical and radiologic
ig 3. Decline in arterial blood oxygenation (arterial partial pres-ure of oxygen to fraction of inspired oxygen ratio [Pao2/Fio2]) inhe interval time between admission and diagnosis.
ndings of our series. F
ats.ctsnetjournDownloaded from
omment
he most meaningful finding of our study is that aultiloculated pneumothorax may develop late after
VRS, and that this complication may be easily missed attandard radiographic assessment even in the presencef large air collections, thus leading to erroneous misdi-gnosis of respiratory failure caused by exacerbatedhronic obstructive pulmonary disease or deterioratingullous emphysema. The causal relation of late-onsetneumothorax with previous LVRS can be exclusivelyypothetical. In some cases, an alveolar, or even bron-hoalveolar, leak might develop after several days as anffect of the mechanical stress induced by reexpansion ofhe lung surface. This mechanism supplies a better ex-lanation of cases that present after a period of relativeell-being after discharge. Alternatively, we suggest that
n some patients, a minimal air leak can persist despiteadiologically documented full lung reexpansion and thebsence of bubbling on the water seal chamber. In thesenstances, pneumothorax is likely to occur early afterischarge, although the presence of thick pleural adhe-ions might initially limit its enlargement and its radio-ogic detection even with a provocative clamping test. Inhese patients, the presence of a minuscule air collection
ay explain the lack of subjective improvement at firstutpatient visit after surgery in some patients. In these
nstances, uneventful chest roentgenogram may lead theurgeon to rule out other procedure-related complica-ions. Therefore, we now prefer to include CT scan evenn the short-term follow-up of LVRS patients presentingith unsatisfactory outcome, particularly if they hadrolonged postoperative air leaks. The basic mechanismf very late onset pneumothorax is more demanding toypothesize. We acknowledge that these pneumothora-es might simply develop as a result of a progression ofhe native disease, without reasonable relationship withhe previous surgery. Nonetheless, the regional differ-nces in elastic recoil induced by LVRS might at leastxaggerate the disruption of lung tissue and facilitate theupture of superficial bullae, even in sites distant fromhe suture lines. A similar mechanism has been proposedo explain the late development of giant bullae afterVRS [1].In our series, conventional chest roentgenography
ailed to pinpoint the exact diagnosis even in presence ofery large sized pneumothoraces, which were instead
2011Ann Thorac Surg TACCONI ET AL2005;80:2008–12 OCCULT PNEUMOTHORAX AFTER LVRS
GEN
ERA
LT
HO
RA
CIC
evealed at CT scan. The merits of CT in this setting areot surprising, and have been evidenced in both clinicalnd experimental studies [3–5]. In particular, Omert andoworkers [6] have stressed that CT scan results in aignificant change in clinical management of traumaatients with intrathoracic complications. Nonetheless,hysicians involved in primary care are sometimes reluc-
ant to include CT in first-line assessment of patientsith thoracic complications, on account of cost-saving
oncerns. This kind of behavior resulted in a remarkableiagnostic delay for our patients that translated in aignificant decrease in arterial blood oxygenation (Fig 3),articularly in those who were first referred to peripheralospitals. On the basis of this observation, we adviseVRS patients to be immediately referred to their spe-ialized center rather than be seen at primary caretructures not experienced with this kind of surgery norith its particular critical care needs.Several authors have suggested that any pneumotho-
ax occurring in critically ills patients should be treatedxpeditiously by chest tube placement because of theigh rate of progression to tension, especially in case ofentilatory support [7, 8]. However, although blind chestube placement may achieve initial stabilization of theatient or even definitive resolution in certain cases, itay be demanding or even hazardous owing to the risk
f lung injury in the presence of loculated pneumothorax.oreover, chest drainage alone may not resolve the air
eaks when these are sustained by a large lung tear,eading to prolonged hospital stay with increased mor-idity and overall costs. We therefore now prefer video-ssisted thoracoscopic surgery management with theoal of sealing pulmonary tears and reducing the risks ofriggering further complications. In addition, having fur-her experience with awake procedures for resection ofulmonary nodules [9] and LVRS [10], we are prone todopt this option even in this setting, with the aim ofeducing surgical risks in these delicate patients. Previ-us studies have focused on the advantages of thispproach to managing air leak problems [11, 12]. Differ-nt video-assisted thoracoscopic surgery techniques cane adopted to seal the air leak, including limited resec-
ion, “no-cut” stapling, instillation of biologic glues, or aombination of these [11–13]. In particular, we haveound that definitive air leak control can be rapidlychieved with the use of instilled cyanoacrylate glue,
able 2. Radiologic and Operative Data
Patient Site of Air Leak Radiographic
Staple line NoneStaple line Hyperlucency, meMajor fissure NoneLower lobe Mediastinal shiftUnknown HyperlucencyUnknown Hyperlucency
P � drainage placement; VATS � video-assisted thoracoscopic surg
hich can be also used in combination with a bovine
ats.ctsnetjournDownloaded from
ericardium patch to seal large lung tears as reported byorsley and Miller [12]. Using this technique, operative
ime was short even in the most demanding instances.espite the finding that more detailed evaluation of
ost-to-benefit analysis of minimally invasive surgeryersus only chest drainage or interventional radiologicrocedures is warranted in future studies, we believe that
his approach translates into decreased morbidity,horter hospital stay, and lower hospital costs.
In conclusion, we believe that occult pneumothoraxhould be considered as a possible late-onset life-hreatening complication of LVRS, especially in patientsith a history of prolonged postoperative air leaks. Oc-
ult pneumothorax may present even several monthsfter the procedure as sudden, progressively deteriorat-ng respiratory impairment in relatively well-faring pa-ients. Nonetheless, in some patients, the presence of a
inimal active air leak should already be suspected athort-term follow-up as exacerbating dyspnea is noticedespite uneventful chest roentgenogram. We thereforeuggest that CT scan should be routinely included in thearly postoperative management of this subgroup. More-ver, because immediate recognition and treatment oflate-onset pneumothorax is crucial to avoid the need
f mechanical ventilation, we recommend that criti-ally ill LVRS patients should be sent immediately toheir referral center for prompt diagnosis and optimal
anagement.
upported by MURST COFIN grants No. 9906274194–06, CNRo. CU0100935 2002, and Centro di Eccellenza 2001. This studyas carried out within the Research Fellowship Program Tec-ologie e Terapie Avanzate in Chirurgia awarded by the Torergata University.
eferences
1. Iqbal M, Rossoff L, McKeon K, Graver M, Scharf SM.Development of a giant bulla after lung volume reductionsurgery. Chest 1999;116:1809–11.
2. Mineo TC, Pompeo E, Mineo D, Rogliani P, Leonardis C,Nofroni I. Results of unilateral lung volume reduction sur-gery in patients with distinct heterogeneity of emphysemabetween lungs. J Thorac Cardiovasc Surg 2005;129:73–9.
3. Engdhal O, Toft T, Boe J. Chest radiography: a poor methodfor determining the size of a pneumothorax. Chest 1993;103:26–9.
ings Treatment Hospital Stay (days)
VATS 3inal shift VATS 4
DP, VATS 2VATS 4DP 3DP 5
Find
diast
4. Phillips GD, Trotman-Dickensen B, Hodson ME, GeddesDM. Role of CT in the management of pneumothorax in
patients with complex cystic lung disease. Chest 1997;112:275–8.
5. Wolfman NT, Wendell SM, Glauser SJ, Meredith JW, ChenMY. Validity of CT classification on management of occultpneumothorax: a prospective study. AJR Am J Roentgenol1998;171:1317–20.
6. Omert L, Yeaney WW, Protetch J. Efficacy of thoracic com-puterized tomography in blunt chest trauma. Am Surg2001:67:660–4.
7. Gobien RP, Reines HD, Schabel SI. Localized tension pneu-mothorax: unrecognized form of barotrauma in adult respi-ratory distress syndrome. Radiology 1982;142:15–9.
8. Streiter RM, Lynch JP. Complications in the ventilatedpatients. Clin Chest Med 1988;9:127–39.
9. Pompeo E, Mineo D, Rogliani P, Sabato AF, Mineo TC.
nline Discussion Forum
he CTSNet Discussion Forum and Web site.
tf
VtKM
TTUP5SPFe
2005 by The Society of Thoracic Surgeonsublished by Elsevier Inc
ats.ctsnetjournDownloaded from
solitary pulmonary nodules. Ann Thorac Surg 2004;78:1761–8.
0. Mineo TC, Pompeo E, Mineo D, Pampana E, Dauri M,Sabato AF. Awake non-resectional lung volume reductionsurgery. Ann Surg. In press.
1. Thistlethwaite PA, Luketich JD, Ferson PF, Keenan RJ,Jamieson SW. Ablation of persistent air-leaks after thoracicprocedures with fibrin sealant. Ann Thorac Surg 1999;67:575–7.
2. Mukaida T, Andou A, Date H, Aoe M, Shimizu N. Thoraco-scopic operation for secondary pneumothorax under localanesthesia in high-risk patients. Ann Thorac Surg 1998;65:924–6.
3. Horsley WS, Miller JI. Management of the uncontrollableair-leak with cyanoacrylate glue. Ann Thorac Surg
Feasibility and results of awake thoracoscopic resection of 1997;63:1492–3.
ach month, we select an article from the The Annals ofhoracic Surgery for discussion within the Surgeon’s Fo-um of the CTSNet Discussion Forum Section. The arti-les chosen rotate among the six dilemma topics coverednder the Surgeon’s Forum, which include: Generalhoracic Surgery, Adult Cardiac Surgery, Pediatric Car-iac Surgery, Cardiac Transplantation, Lung Transplan-
ation, and Aortic and Vascular Surgery.Once the article selected for discussion is published
n the online version of The Annals, we will post a noticen the CTSNet home page (http://www.ctsnet.org) with aREE LINK to the full-text article. Readers wishing toomment can post their own commentary in the discus-ion forum for that article, which will be informallyoderated by The Annals Internet Editor. We encourage
ll surgeons to participate in this interesting exchangend to avail themselves of the other valuable features of
For December, the article chosen for discussion underhe Adult Cardiac Dilemma Section of the Discussionorum is:
acuum-Assisted Wound Closure of Deep Sternal Infec-ions in High-Risk Patients After Cardiac Surgeryyle Northcote Cowan, MD, PhD, Laura Teague, RN,N, Sammy C. Sue, BS, and James L. Mahoney, MD
om R. Karl, MDhe Annals Internet EditorCSF Children’s Hospitalediatric Cardiac Surgical Unit05 Parnassus Ave, Room S-549an Francisco, CA 94143-0118hone: (415) 476-3501ax: (212) 202-3622
