Keywords
metastasectomy - lung function - bilateral metastasectomy - functional impairment
Introduction
The resection of lung metastases is a well-accepted method to treat oligometastases
from a wide range of primary tumors. The postoperative loss of pulmonary function
has been described for wedge resections[1]
[2] and anatomical resections[3]
[4] and can be predicted to a certain extent. Postoperative chemotherapy and bilateral
surgery were recently reported to be a risk factor for impaired function,[5] but the extent of functional loss in the bilateral surgery group remains unclear.
Knowledge of functional loss is of particular importance because many patients have
impaired cardiopulmonary reserve after several cycles of chemotherapy, due to increased
age or comorbidities. Furthermore, only 25 to 35% of patients receiving metastasectomy
can expect long-term survival; thus, metastasectomy should not be debilitating and
sufficient pulmonary function should remain for an adequate quality of life.[6]
[7]
[8]
[9] In a previous analysis,[1] we found that 5% of the pulmonary functional impairment may be attributed to thoracotomy.
Therefore, one may hypothesize that bilateral thoracotomy causes a significant deterioration
in pulmonary function, which is twice the loss following unilateral intervention.
To prove this hypothesis, we performed a post-hoc subanalysis of the previously analyzed
collective of 117 prospectively evaluated patients[1] and compared standardized spirometric values and Dlco before and after metastasectomy of the unilateral with the bilateral resected patients.
Methods
This study was performed as a post-hoc subanalysis of a prospective observational
case series that was previously published.[5] The subset of patients with bilateral metastasectomy was further analyzed and a
second comparative analysis of the unilateral versus bilateral operated patients was
performed. From April 2008 to April 2010, patients from the Department of Thoracic
Surgery who were scheduled for pulmonary metastasectomy were prospectively included
in this study. Patients with a planned complete removal of lung metastases were asked
for written consent the day before surgery. The study protocol was approved by our
hospital's review board. The Eastern Cooperative Oncology Group (ECOG) performance
state of all patients was 0 or 1 and no significant cardiorespiratory limitations
were noted preoperatively. All metastasectomies were performed via an anteroaxillary
muscle-saving thoracotomy with thoracotomy being used for the systematic palpation
of the entire lung. Bilateral metastases were removed simultaneously or staged within
6 weeks. Lung resection was performed at the surgeon's discretion with staplers, electrocautery,
or laser, and was supplemented by lymph node sampling. One or two drains were placed
before closing the thoracotomy. All patients received analgesia based on the hospital's
standard protocol, which included peridural anesthesia whenever possible.[10] A standard physiotherapy program was used to support expiratory flow, lung expansion,
and coughing.
Pulmonary Function Testing
Standard pulmonary function tests (PFTs) and Dlco were measured preoperatively (t1) and postoperatively the day before discharge (t2).
Patients with a sequential bilateral metastasectomy were measured again after the
second intervention (t3). All cases had a planned follow-up evaluation 3 months after
the last resection (t4). Spirometry tests were performed using a ZAN 500 Body Plethysmograph
equipped with a ZAN 200 SB CO-Diffusion module (Spire Health Ltd., Hertford, UK) according
to the criteria of the American Thoracic Society. The measurements were taken after
administering a bronchodilator. The Dlco was measured using the single-breath method. The FVC, TLC, FEV1, Dlco, and diffusion coefficient for carbon monoxide (KCO) were expressed as percentages
of the predicted values for age, gender, and height, according to the European Community
for Steel and Coal (ECCS) prediction equations .[11]
Evaluation
Lung function parameters, blood gas values, and the percentage of carbon monoxide
diffusion capacity before and after metastasectomy and at follow-up were entered into
a database along with the number of resections, the type of resection, and the laterality
of the resection.
Statistical Analysis
All variables were analyzed by the methods of descriptive statistics (frequency, mean ± standard
deviation, range). To test differences between means, the Student t-test was applied in the case of normal distribution; otherwise, the Mann–Whitney
U test was used. To discriminate between normal and non-normal distributed variables,
the Kolmogorov–Smirnov test was applied. Levene test was used to assess the equality
of variance in different samples. We applied the chi-square test or the Fisher exact
test to compare frequencies and proportions in two groups. We considered p < 0.05 to be statistically significant. Statistical analysis was performed with SPSS
20.0 (SPSS Inc., Chicago, United States).
Results
Altogether, 117 patients completed the follow-up visit and could be further analyzed;
86 of these patients had unilateral and 31 had bilateral resections. Twenty one of
the bilaterally operated patients were males. Fifteen of these 31 patients had complete
PFTs at t1 and t4 and after both sequential operations (t2, t3) before discharge from
hospital. Information about PFTs at t2 or t3 was incomplete for 16 patients. The overall
mortality of the initially enrolled study population corresponding to the whole study
period was 3/162 (1.8%) and was published earlier.[5]
Characteristics of the Bilateral Group
Altogether, 271 nodules (median 7, mean 8.2) were removed from the 31 patients with
lobectomy in 2, segmentectomy in 8, and multiple wedge resection in 21 patients, with
this being the largest resection. Postoperative complications potentially affecting
pulmonary function were noticed in two cases (6.5%) with diaphragmatic elevation and
two cases (6.5%) receiving pleurodesis for prolonged air leak. The mean time between
the two operations was 28 days; four patients had a simultaneous bilateral metastasectomy.
The mean time from the date of first operation to the follow-up visit was 4.4 months.
The mean loss of pulmonary function at the follow-up visit compared with preoperative
values was significant for all volume measures and for Dlco between 13 and 16% ([Table 1]). No changes were found between preoperative and follow-up values for FEV1/FVC, Po
2, Pco
2, and KCO. The stepwise deterioration of PFT values was calculated using the complete
dataset of the 15 patients who had PFTs after each metastasectomy before discharge
from hospital and is depicted in [Fig. 1] with the corresponding values shown in [Table 2]. All volume measures demonstrated a significant stepwise deterioration after the
first and second interventions with a significant reduction in Po
2 noticed after each operation (− 4.0 mm Hg [t2] and further − 5.7% mm Hg [t3], p = 0.01), such that Po
2 was almost 10 mm Hg lower after the second intervention compared with preoperative
Po
2. KCO and Pco
2 remained unchanged at all times. Partial recovery was registered at follow-up for
all volume measures and Dlco compared with t3 measures, along with complete normalization of Po
2.
Table 1
Mean values of pulmonary function tests before and 3 months after bilateral pulmonary
metastasectomy
|
Parameter
|
n =
|
Preoperative
|
Follow-up
|
Difference
|
Significance
|
|
t1 (min–max)
|
t4
|
(t4 − t1) (SD)
|
(t4–t1)
|
|
FVC (%)
|
31
|
97.8 (70–126)
|
82.7
|
− 15.1 (13.1)
|
< 0.001
|
|
TLC (%)
|
31
|
96.3 (73–130)
|
82.4
|
− 13.9 (14.8)
|
< 0.001
|
|
FEV1 (%)
|
31
|
89.7 (59–120)
|
73.5
|
− 16.3 (11.5)
|
< 0.001
|
|
FEV1/FVC
|
31
|
72.8 (48–92)
|
70.8
|
− 2.1 (9.2)
|
0.856
|
|
Dlco (%)
|
24
|
83.6 (36–115)
|
70.2
|
− 10.7 (14.6)
|
< 0.001
|
|
KCO (%)
|
24
|
72.3 (39–100)
|
72.7
|
0.4 (10.3)
|
0.407
|
|
Po
2 mm Hg
|
31
|
79.3 (65–97)
|
81.2
|
1.1 (11.9)
|
0.347
|
|
Pco
2 mm Hg
|
31
|
38.1 (29–44)
|
38.7
|
0.6 (3.8)
|
0.446
|
|
AaDo
2 mm Hg
|
31
|
21.9 (4–48)
|
20.3
|
1.6 (8.9)
|
0.416
|
Abbreviations: AaDO2, alveoloarterial oxygen tension difference; Dlco, diffusing capacity of lung for carbon monoxide FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; SD, standard deviation;
TLC, total lung capacity.
Notes: In 7 of 31 patients, Dlco and KCO were not available either preoperatively or postoperatively so the difference
could not be calculated. Significances were calculated for the differences between
follow-up pulmonary function values and preoperative values.
Table 2
Spirometry values at different times with significance of changes (n = 15)
|
Parameter
|
Preoperative
|
Postoperative 1
|
Postoperative 2
|
Follow-up
|
|
t1
|
t2 (sig 1)
|
t3 (sig 2)
|
t4 (sig 4)
|
|
FVC (%)
|
97.8
|
75.6 (< 0.001)
|
51.3 (0.001)
|
82.7 (< 0.001)
|
|
TLC (%)
|
96.3
|
79.7 (< 0.001)
|
58.5 (0.001)
|
82.4(< 0.001)
|
|
FEV1 (%)
|
89.7
|
66.9 (< 0.001)
|
46.5 (0.001)
|
73.5 (< 0.001)
|
|
Dlco (%)
|
83.6
|
62.8 (0.008)
|
45.6 (0.007)
|
70.2 (< 0.001)
|
|
KCO (%)
|
72.3
|
68.0 (0.23)
|
69.1 (0.721)
|
72.9 (0.407)
|
|
Po
2 mm Hg
|
79.3
|
75.3 (0.01)
|
69.6 (0.014)
|
81.2 (0.347)
|
|
Pco
2 mm Hg
|
38.1
|
37.7(0.20)
|
37.8 (0.905)
|
38.7 (0.446)
|
Abbreviations: Dlco, diffusing capacity of lung for carbon monoxide; FVC, forced vital capacity; FEV1, forced expiratory volume in 1 second; KCO, diffusion coefficient for carbon monoxide;
TLC, total lung capacity.
Notes: Of 31 individuals with PFTs preoperatively and at follow-up, 15 had a complete set
of values after the first and second metastasectomies. The statistical significance
of differences between t1–t2 (sig 1), t2–t3 (sig 2), and t1–t4 (sig 4) was calculated.
Fig. 1 Predicted pulmonary function values of bilaterally operated patients at different
times. Legend 1: Values were taken from the 15 patients with complete PFTs at t1–t4.
The y-axis represents percentage (%) of the corresponding PFT value. Dlco, diffusing capacity of lung for carbon monoxide FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; TLC, total lung
capacity.
Characteristics of the Unilateral Group
Altogether, 266 nodules (median 2, mean 3.1) were removed from the 86 patients, with
lobectomy in 11, segmentectomy in 17, and wedge excision in 58 patients, with this
being the largest resection. The mean time from the date of first operation to the
follow-up visit was 3.8 months. Postoperative complications potentially affecting
pulmonary function were noticed in six cases (7.0%) with diaphragmatic elevation and
two cases (2.3%) receiving pleurodesis for prolonged air leak. A small subgroup of
13 (15.1%) patients had FEV1 values less than 70% and were thus impaired preoperatively. The loss of FEV1 at the follow-up visit was a mean of 5.5% (− 20 to + 10%). The other characteristics
of the two groups are depicted in [Table 3].
Table 3
Comparison of treatment-related factors in the two groups
|
Parameter
|
Unilateral group (n = 86)
|
Bilateral group (n = 31)
|
Significance
|
|
Age
|
Mean 60.3 (SD 12.1)
|
Mean 57.1 (SD 14.8)
|
0.331
|
|
Specimen
|
Mean 2.7 (SD 3.6)
|
Mean 8.2 (SD 5.4)
|
< 0.001
|
|
Preoperative chemo
|
16 (18.6%)
|
9 (29%)
|
0.306
|
|
Presence of adhesions
|
37 (43%)
|
9 (29%)
|
0.202
|
|
Lobectomy
|
11 (12.8%)
|
2 (6.5%)
|
0.509
|
|
PTS
|
11 (12.8%)
|
2 (6.5%)
|
0.509
|
|
Continued pain med.
|
13 (15.3%)
|
5 (16.1%)
|
1.000
|
|
Postoperative chemo
|
15 (17.6%)
|
6 (19.4%)
|
1.000
|
|
Tumor recurrence
|
10 (11.8%)
|
6 (19.4%)
|
0.362
|
Abbreviations: PTS, postthoracotomy pain syndrome; SD, standard deviation.
Note: Significance = bilateral significance with Pearson chi-square test, tumor recurrence
evaluated at the follow-up visit after 3 months.
Comparison of the Uni- and Bilateral Groups
The two groups did not differ in terms of age (p = 0.331), preoperative chemotherapy (p = 0.306), number of segmentectomies (p = 0.177), preoperative spirometry values (FVC [p = 0.930], TLC [p = 0.200], FEV1 [p = 0.398], Dlco [p = 0.916], KCO [p = 0.596], Po
2 [p = 0.868], and Pco
2 [p = 0.459]), or the frequency of post-thoracotomy pain syndrome (unilateral vs. bilateral,
6.5 vs. 12.8%), or detected tumor recurrence (unilateral vs. bilateral, 11.8 vs. 19.4%)
at follow-up. Similar rates were found in both groups for continuous use of pain medication
(unilateral vs. bilateral, 15.3 vs. 16.1%) and postoperative chemotherapy (unilateral
vs. bilateral, 17.6 vs. 19.4%) at follow-up. There was a trend for more adhesions
to be removed in the unilateral group, with 29.1 versus 12.9% in the bilateral group
(p = 0.091) ([Table 3]). The only terms that differed significantly between the groups were the cumulative
number of resected specimens (mean 2.7 vs. 8.2; p < 0.001) ([Table 3]) and the relative functional loss of FVC (%) (p = 0.004), TLC (%) (0.004), and FEV1 (%) (p = 0.001) at follow-up. Dlco and Po
2 went back to normal in both groups after 3 months ([Table 4]).
Table 4
Mean differences of pulmonary function tests (follow-up—preoperative value) between
the unilateral (n = 86) and the bilateral (n = 31) group (Δt4 − t1)
|
Parameter
|
Unilateral (SD)
|
Bilateral (SD)
|
Significance (p)
|
|
FVC (%)
|
− 8.0 (10.7)
|
− 15.1 (13.1)
|
0.004
|
|
TLC (%)
|
− 7.0 (10.9)
|
− 13.8 (14.8)
|
0.004
|
|
FEV1 (%)
|
− 9.0 (9.7)
|
− 16.3 (11.5)
|
0.001
|
|
FEV1/FVC
|
− 1.7 (5.6)
|
− 2.1 (9.2)
|
0.556
|
|
Dlco (%)
|
− 9.5 (10.0)
|
− 10.3 (14.6)
|
0.868
|
|
KCO (%)
|
− 0.7 (8.8)
|
0.4 (10.3)
|
0.502
|
|
Po
2 (mm Hg)
|
0.1 (10.1)
|
1.1 (11.9)
|
0.550
|
|
Pco
2 (mm Hg)
|
0.6 (7.0)
|
0.6 (3.8)
|
0.346
|
|
AaDo
2 (mm Hg)
|
− 0.6 (9.0)
|
− 1.5 (8.9)
|
0.416
|
Abbreviations: AaDO2, alveoloarterial oxygen tension difference; Dlco, diffusing capacity of lung for carbon monoxide; FVC, forced vital capacity; FEV1, forced expiratory volume in 1 second; KCO, diffusion coefficient for carbon monoxide;
SD, standard deviation; TLC, total lung capacity.
Note: The significance was calculated as the differences in Δt4 − t1 between the two groups.
Discussion
Bilateral simultaneous or staged metastasectomy is a routine procedure to treat oligometastases
from various primaries.[8]
[12] Unfortunately, more than 60% of patients have a tumor relapse and only 36% will
survive for more than 5 years.[6]
[9]
[13] Metastasectomy is usually performed as a parenchyma-sparing precision resection,
but can also include anatomical resections. Current selection criteria include sufficient
pulmonary function to tolerate the planned resection.[8]
[14] This tolerance is usually expressed when predicted postoperative function exceeds
a certain lower limit. A significantly increased operative risk for complications
was found when ppo (predicted post-operative) FEV1 was less than 1.0 L,[15] or when preoperative Dlco was reduced.[16]
[17] Current guidelines advise excluding candidates from anatomic lung resection when
ppo FEV1 and Dlco are < 30 to 40% of their predicted values.[17]
[18]
[19] Thus, the prediction of postoperative preserved pulmonary function seems essential,
even for metastasectomy candidates, especially those who have cardiopulmonary impairment.[15]
[20] In terms of metastasectomy, patients are currently accepted for or denied these
procedures based on the personal experience of the surgeon, rather than on the calculation
of ppo pulmonary function. The loss of pulmonary function after metastasectomy is
a topic of current interest,[2]
[5]
[21] but many details remain unclear, such as the functional loss after bilateral operations.
Because pulmonary metastasectomy is noncurative in two-thirds of the patients, it
is important not to look solely at operative risk, but to also include the goal of
a mobile life free from oxygen dependence and therefore calculate and predict preserved
postoperative pulmonary function.
Current Study
This post-hoc subanalysis was based on prospectively evaluated patients and was performed
to define the average functional loss after bilateral metastasectomy, with the aim
of enabling prediction of the function remaining after such interventions in the future.
To the best of our knowledge, this study is the first detailed report about midterm
loss of pulmonary function after bilateral metastasectomy and the first comparison
of lung function changes after unilateral versus bilateral open pulmonary metastasectomy
([Table 4]).
Current Results
Our results conflicted with our initially formulated hypothesis: first, the mean number
of resected specimens per patient was 2.7 in the unilateral group and significantly
more (8.2) in the bilateral group, which implies that bilateral manifestation is associated
with a higher number of lesions per side. Furthermore, there was a trend for fewer
adhesions in the group with bilateral metastasectomies. Second, whereas the functional
impairment of the volume parameters FVC, TLC, and FEV1 in the bilateral group was significantly higher, almost double that of the unilateral
group after metastasectomy, the changes in Dlco did not differ significantly: − 9.5% in the unilateral versus − 10.7% in the bilateral
group (p = 0.868). The fact that Dlco after 3 months was higher than ppo Dlco was reported previously for lobectomy (+ 10%) and pneumonectomy (+ 17%) without pathophysiological
explanation.[22] The volume-adjusted diffusion index KCO did not change at all in either group, meaning
that the effects of chemotherapy, infections, or other lung injuries cannot be used
to explain the difference in Dclo. In both groups, blood gases returned to normal values after 3 months. Others also
found unchanged KCO after resection.[22]
Operative Access
The influence of operative access was not investigated in our study because all patients
had the same anteroaxillary incision. However, it is well known that operative access
is an important predictor of postoperative function, with thoracoscopic interventions
causing less impairment than sternotomy, and sternotomy causing less impairment than
posterolateral thoracotomy.[21] The effect of anteroaxillary thoracotomy may be responsible for a 5% restrictive
disturbance of volume parameters.[1]
Limitations of the Study
This study has some limitations. It was conducted in a single center, where metastasectomy
is a routine procedure. The sample size is rather small for a substantiated message;
this is most evident at the high values of standard deviations. The operation was
performed by different surgeons with different preferences for the use of laser, electrocautery,
or stapler. As this might reflect the normal spectrum in the thoracic surgery community,
however, it might not be a disadvantage. The amount of lung tissue removed was not
quantified. Besides the fact that patients with lobectomies and segmentectomies were
included into the analysis, it is unquestionable that small superficial nodules have
been removed with less tissue than a deep wedge resection for a large nodule. To our
understanding, this bias cannot be solved, but we think that the analysis of a heterogeneous
population reflects the daily situation and can offer a certain impression of the
results of bilateral resection for a patient. The study protocol with PFTs at t1,
t2, t3, and t4 was completed in only 15 patients in the bilateral group, which is
undoubtedly a limiting factor for further interpretation. However, the most interesting
question, that is, which of the two interventions in the bilateral group is more harmful
to pulmonary function, was addressed with the analysis of this small subgroup. We
found a linear reduction of all volume parameters from preoperative (t1) to t2 and
from t2 to t3, which can be attributed to the similar incision on both sides and comparable
amounts of lung tissue resected on both sides. Nonetheless, the second intervention
was worse for the patient, leading to the lowest values of volume parameters, at around
40% less than preoperative values, and a reduction in Po
2 of 10 mm Hg. Subsequent pain management in this situation is even more important
to reduce stress to a minimum and avoid complications. Within 3 months after t3, significant
functional recovery occurred, back to higher values than measured after the first
metastasectomy. However, the remaining functional impairment after 3 months exceeded
that known for lobectomy in lung cancer treatment.[22]
[23] Therefore, one can postulate that midterm pulmonary function changes after bilateral
metastasectomy are less severe than the early changes after the first operation.
Despite the fact that the sample size was small, the statistical differences between
the groups and different time points were highly significant and thus relevant. Also,
because bilateral metastasectomy is a rare procedure, we feel that it is difficult
to collect larger sample sizes within an adequate time period. Long-term results are
difficult to obtain because many patients have tumor relapse and need further treatment,
which affects pulmonary function and thus superimposes the effect of surgery. We believe
that midterm results after 3 months are sufficient to describe the effect of metastasectomy
on pulmonary function. One can assume that functional recovery occurs within 3 to
6 months postoperatively. Other authors found that pulmonary function recovery after
lung resection reaches a plateau after 3 to 6 months.[22]
[24] Bolliger et al[23] investigated patients after lobectomy and found a functional loss for FVC, FEV1, TLC, and Dlco of − 10.7, − 10.9, − 12.4, and − 8.5%, respectively, at 3 months with a further recovery
at 6 months to − 7.3, − 8.8, − 10.2, and − 4.0%. This underlines our hypothesis that
only minor functional improvement occurs after 3 months.
Conclusion
Because more than two-thirds of patients will develop tumor recurrence after pulmonary
metastasectomy, a thorough consultation should include the expected functional impairment
of the procedure. Bilateral interventions are associated with a three- to fourfold
higher number of metastases which need to be resected. Impairment of pulmonary volume
parameters is around 40% early after the second operation, but recovers to a midterm
loss of 15% ( ± 13) at 3 months when using anteroaxillary thoracotomy on both sides.
Oxygen partial pressure is decreased by 10 mm Hg after the second operation but normalizes
completely after 3 months. However, to predict the postoperative function of a specific
patient, one should bear in mind that the amount of tissue to be removed is different
in every individual. Mean values from this study should only be used as an indication.
