Short- and long-term outcomes of pneumonectomy in a tertiary center

Introduction

Lung cancer is a major cause of death worldwide. It is currently the leading cause of cancer death in both men and women in the United States. ¹ It was predicted that 169,400 individuals in United States would be diagnosed with lung cancer in 2002 (90,200 men and 79,000 women) and 154,900 individuals would succumb to this disease (89,000 men and 65,000 women) during the same year.^1,2 The mortality of 154,900 sufferers is approximately equivalent to the death toll from a jumbo jet crashing every day of the year, year after year. ² More people die due to lung cancer than cancer of the colon, breast, and prostate combined, despite lung cancer accounting for only 14% of all new cancer diagnoses. For patients diagnosed with non-small-cell lung cancer, treatment involves a combination of surgery, chemotherapy, and radiation, while surgery remains the most important therapeutic modality. Pneumonectomy is a high-risk surgical procedure, which represents less than 15% of all lung cancer resections. It is associated with substantial morbidity and mortality and worse long-term survival compared to lobectomy.^3–8 The National Thoracic Surgery Activity & Outcomes Report 2011 for Great Britain and Ireland noted 6.5% mortality in pneumonectomy patients with lung cancer compared to 2% in lobectomy patients, and 7.9% vs. 2.4% mortality for non-lung cancer surgery. ⁹ However, pneumonectomy is still an important and valuable surgical procedure to achieve complete tumor resection in centrally located or locally advanced lung cancer in which lesser lung resections are not effective surgical options. Several factors have been associated with adverse effects on long-term survival, including the side of pneumonectomy, age, sex, pathological stage, and adjuvant therapy. Surgical outcomes are normally reported as 30-day or 90-day mortality or 5-year survival; 10-year survival is rarely mentioned in the national data or international studies. The aim of this study was to review our lung resection rate, operative mortality, and long-term outcomes in terms of 5- and 10-year survival in patients who underwent pneumonectomy.

Patients and methods

We retrospectively reviewed 306 consecutive patients who underwent pneumonectomy, mainly for lung cancer, from January 1998 to February 2014 in Castle Hill Hospital, Hull, UK. They were predominately male (79%) with mean age of 64 ± 10 years (range 22–82 years) and 24% were ≥70 years of age (Table 1). Data were collected from the departmental database and hospital records, and the patients were followed up until September 2014. We were unable to collect all the information regarding recurrences, complications, and adjuvant therapy for patients treated in the early half of study because they were referred back to their local hospital for adjuvant treatment and rehabilitation.

OPEN IN VIEWER

Table 1.

Preoperative and operative characteristics of 306 patients undergoing pneumonectomy.

Variable	No. of patients
Age (years)
Median	64 ± 9
Range	22–82
≥70 years	24%
Male	242 (79%)
Female	64 (21%)
Diagnosis
Benign	6 (2%)
Squamous cell carcinoma	167 (55%)
Adenocarcinoma	133 (43%)
Left pneumonectomy	177 (58%)
Right pneumonectomy	129 (42%)
Extrapleural pneumonectomy	30 (10%)

All patients were assessed for fitness for surgery according to our departmental protocol. We analyzed their demographic and pathological characteristics for short- and long-term outcomes, and used the Kaplan-Meier method to analyze survival. The risk of postoperative hospital mortality was measured retrospectively using Thoracoscore, a validated model for predicting postoperative risk: the mean predicted risk of mortality was 8%.

Survival data were collected from hospital records and the patients’ general physicians. Survival was observed up to September 2014 and measured from the date of surgery to the date of last contact or death due to any cause. Statistical analysis of survival was undertaken using a landmark of 30-day mortality after surgical resection. Univariate analysis was performed using the Kaplan-Meier life-table method to determine the effects of demographic and pathologic variables. The log-rank test was used to determine statistical significance in comparisons among survival curves. Results with a p value less than 0.05 were considered significant.

Results

Mean hospital stay was 10 ± 8 days. The median follow-up interval was 29 months (range 2 months to 16 years). Pneumonectomy was performed for lung cancer in 98% of patients, 10% of these (30 patients) had an extrapleural pneumonectomy for malignant mesothelioma, and 58% had a left pneumonectomy. Retrospectively predicted hospital mortality by Thoracoscore was 8%, and the national mortality of pneumonectomy for lung cancer was reported as 6.5% by the Society for Cardiothoracic Surgery of Great Britain and Ireland Second National Thoracic Surgery Database Report 2011. However, our observed hospital operative mortality was 4.5% (14 patients) for all pathologies while our operative mortality for cancer patients was only 3.3% and there were only 2 (3%) hospital deaths in the last 5 years. Mortality was significantly higher in patients aged ≥70 years, males, right pneumonectomies, and urgent surgery for nonmalignant pathology (Table 2). There was no hospital mortality in patients who had an extrapleural pneumonectomy for malignant mesothelioma. Mean survival after pneumonectomy was 4.8 years with a 95% confidence interval (95%CI) of 4.1–5.4 years, and median survival was 2.2 years (95%CI: 1.7–2.7 years). Overall 1-, 2-, 3-, 5- and 10-year survival rates were 70%, 54%, 42%, 32% and 20%, which were similar to those in international studies (Figure 1). Female patients had significantly better median survival of 3.6 years (95%CI: 4.5–7.6 years) compared to 2 years (95%CI: 2.3–5 years) in males, and 5- and 10-year survival was higher in females than males (43% and 28% in females vs. 29% and 17% in males; Figure 2). Patients with squamous cell carcinoma survived longer (median 3.4 years, 95%CI: 2.4–4.4 years) than those with adenocarcinoma (1.3 years, 95%CI: 0.6–2 years), and long-term survival was also significantly longer in squamous call lung cancer patients (40% and 29% vs. 16% and 13% at 5 and 10 years; Figure 3). Similarly, median survival was better after a left pneumonectomy (2.7 years, 95%CI: 1.8–3.6) than a right pneumonectomy (1.9 years, 95%CI: 1.3–2.4 years). Improved long-term survival at 5 and 10 years was observed for a left pneumonectomy (37% and 25%) vs. a right pneumonectomy (25% and 11%; Figure 4). Median survival of 2.5 years (95%CI: 1.9–3 years) was observed in patients aged <70 years, which was not significantly better than 2.1 years (95%CI: 1.2–3 years) in those ≥70 years. The long-term survival at 5 and 10 years was better in patients <70 years (34% and 23%) vs. 24% and 9% in those ≥70 years (Figure 5). We also compared long-term survival according to the surgeon and found no significant differences (Figure 6). Six patients underwent pneumonectomy for non-cancerous pathology; all of these were very sick and had undergone urgent surgery. The underlying pathology was botryomycosis, bronchiectasis, pneumoconiosis, tuberculosis empyema, lung abscess with empyema, and inflammatory myofibroblastic tumor. Two patients died in hospital postoperatively due to uncontrolled sepsis and very poor lung compliance, and one died due to postoperative myocardial infarction.

OPEN IN VIEWER

Figure 1.

Overall median and mean survival with 95% confidence intervals and overall long-term survival of 306 patients.

OPEN IN VIEWER

Figure 2.

Survival in female vs. male patients.

OPEN IN VIEWER

Figure 3.

Survival according to squamous cell lung cancer or adenocarcinoma.

OPEN IN VIEWER

Figure 4.

Survival according to left or right pneumonectomy.

OPEN IN VIEWER

Figure 5.

Survival in patients aged ≥70 vs. <70 years.

OPEN IN VIEWER

Figure 6.

Survival according to surgeon.

OPEN IN VIEWER

Table 2.

Mortality data in 306 patients undergoing pneumonectomy.

Mortality	Incidence	No. of patients
Predicted (Thoracoscore)	8%
Observed	4.6%	14
In males	4.5%	11
In females	4.7%	3
In patients ≥70-years old	9%	7
In patients <70-years old	3%	7
In left pneumonectomies	2.8%	5
In right pneumonectomies	7%	9
In lung cancer	3.7%	11
In nonmalignant disease	50%	3

Discussion

Surgery for lung cancer depends on multiple variables such as patient factors, invasiveness, size and position of the cancer, and it ranges from wedge excision to pneumonectomy. Pneumonectomy is an extensive resection sometimes required for centrally located or locally advanced lung cancer to achieve better cancer clearance. Pneumonectomy has been reported to have high complication rates, with substantial postoperative morbidity and mortality of 3.3% to 13%; however, more recent studies have shown that operative mortality has reduced significantly. Licker and colleagues ¹⁰ reported 9.3% 30-day mortality in 193 consecutive pneumonectomies, and Shapiro and colleagues ¹² described 5.6% mortality utilizing the Society for Thoracic Surgeons General Thoracic Surgery Database. Similarly, Ludwig and colleagues ¹² reported 4.6% operative mortality, whereas Damhuis and colleagues ¹³ described 9.4% postoperative mortality in octogenarians. In our study cohort, 306 patients underwent pneumonectomy over a 15-years period, 98% (300 patients) for lung cancer, and 2% (6 patients) for nonmalignant pathology; 10% of our patients had an extrapleural pneumonectomy for malignant mesothelioma. Our national mortality for pneumonectomy reported by the 2011 National Thoracic Surgery Activity & Outcomes Report was 6.5%, whereas our predicted mortality by Thoracoscore was calculated to be 8%. ⁹ However, the overall observed mortality in this cohort was 4.5% (14 patients) while it was 3.7% (11 patients) in lung cancer patients. There was no postoperative death in the mesothelioma group. Among those who died, 7/14 were ≥70 years, 11/14 were male, and 9/14 had a right pneumonectomy, while 3/6 patients died postoperatively following an unplanned urgent pneumonectomy for nonmalignant diseases.

In our study group, we observed that age ≥70 years, a right pneumonectomy, and urgent non-cancer lung surgery were the main risk factors for hospital mortality. Various studies have identified age ≥70 years, right-sided pneumonectomy, current smoking, cardiac diseases, and reduced preoperative forced expiratory volume in 1 s as major risk factors for postoperative mortality. Alexiou and colleagues ⁵ suggested that old age and bronchopleural fistula are major risk factors for poor hospital outcomes. ⁵ Ramnath and colleagues ⁶ found that American Society of Anesthesiology score less than 3, current smoker status, and pathological stage were important predictors of survival. Leo and colleagues ⁷ reported higher risks of mortality (11.4%) and morbidity (54.2%) in patients aged >70 years and those with a forced expiratory volume in 1 s of 70%. Bernard and colleagues ⁸ found in a retrospective study of 693 patients operated on in the Mayo Clinic that adverse prognostic factors were age, male sex, associated respiratory and cardiovascular diseases, cigarette smoking, decreased vital capacity, forced expiratory volume in 1 s, diffusion capacity, preoperative radiation, right-sided pneumonectomy, crystalloid and blood transfusion, low hemoglobin preoperatively, bronchial stump reinforcement, and hematological disease.

We experienced high mortality in our non-cancerous pneumonectomies, although this was a small group of 6 patients. The underlying pathology was botryomycosis, bronchiectasis, pneumoconiosis, tuberculosis empyema, lung abscess with empyema, and inflammatory myofibroblastic tumor. Three of these patients died in the postoperative period. Decisions for surgery were made due to uncontrolled sepsis despite antibiotic treatment, and increasing respiratory distress with a continuously increasing oxygen requirement. Surgery was not planned for pneumonectomy in these patients, but it was the only option because most of the normal lung tissues were distorted and damaged due to severe disease and infection. Pneumonectomy was the only choice intraoperatively to achieve complete excision of infective material and damaged lung, to give a better chance of survival. One of these patients had a myocardial infarction on the 2nd postoperative day and 2 died due to uncontrolled sepsis and very poor lung compliance. We included these patients in our study to show that pneumonectomy is a rare choice of operation for nonmalignant pathology. However, it is a very high-risk procedure in uncontrolled sepsis and unplanned circumstances.

Long-term outcome has been reported as 1-, 3-, and 5-year overall survival after pneumonectomy for lung cancer. Ferguson and colleagues ⁴ reported overall 1-, 3-, and 5-year survival of 52%, 28%, and 21%, respectively. The 1-, 3-, and 5-year survival was 66%, 38%, and 38% in a cohort of 36 patients in stage III reported by Saha and colleagues. ¹⁴ However, there are scant published data regarding 10-year survival in these patients. In our cohort, overall 1-, 2-, 3-, 5-, and 10-year survival rates were 70%, 54%, 42%, 32%, and 20%. Median and mean survival was significantly better in females, patients who had a left pneumonectomy, and those with squamous cell lung cancer. Although median survival was slightly better in patients <70 vs. ≥70 years of age, the difference was not significant. Similarly, long-term outcome in terms of 5- and 10-year survival was also significantly better in females, patients who had a left pneumonectomy, and those with squamous cell lung cancer. Although the 5-year survival was not significantly better in patients <70 years old, there was a significant survival difference at 10 years (23% vs. 9%). However, this could be related to their comorbidities or older age. The surgeon performing the operation did not influence the short- and long-term outcomes.

Our low operative mortality following pneumonectomy is multifactorial. In the preoperative stage, we are very particular that the patient is fully prepared and worked up according to British Thoracic Society guidelines, with a recent computed tomography scan or positron-emission tomography to avoid any surprise intraoperatively. Our surgical technique was a standard pneumonectomy as described in the literature, but all of our patients were admitted to the high-dependency unit for overnight care, which is part of a dedicated cardiothoracic intensive care unit and managed by cardiothoracic intensivists. We earnestly stress appropriate postoperative fluid management and physiotherapy. Fluid management is crucial in these patients, especially in the first 48 h. We avoid fluid overload and maintain hypovolemic conditions. We also have a very dedicated and well-trained thoracic physiotherapist who provides regular physiotherapy at least twice a day. We believe that these small improvements in every stage of the management pathway enhanced the final outcomes.

This retrospective single-institution study has shown that our mortality is 50% less than the national mortality for pneumonectomy and significantly lower than that predicted by Thoracoscore. Although we agree with the recommendations for lung-sparing resections such as lobectomy or bilobectomy, with or without sleeve resection, for invasive lung cancer, performing a pneumonectomy to achieve clear resection margins is sometimes unavoidable. We have shown that pneumonectomy is still an effective and valuable modality for the treatment of advanced lung cancer, and better outcomes can be achieved, including low operative mortality and good long-term survival, especially in females, younger patients, those with squamous cell lung cancer, and after a left pneumonectomy. However, we must be cautious in patients >70 years of age and those who require a right pneumonectomy or urgent surgery for nonmalignant disease.