Keywords lung neoplasms - neoplasm metastasis - histology
Introduction
Lung cancer is one of the three most prevalent malignancies and the leading cause
of death by cancer.[1 ] The natural history of the disease includes aggressive evolution and reduced survival.[2 ]
[3 ] In addition, metastatic carcinoma is the most common malignant bone tumor, occurring
in approximately 15–40% of patients with lung cancer.[4 ] Necropsy data show that the prevalence of bone metastases reaches 85% when the primary
site is the lung.[5 ] In such context, bones are among the most frequent sites of metastatic lung cancer,
resulting in high morbidity and reduced quality of life for these patients.[6 ]
[7 ]
[8 ]
Approximately 40% of patients with non-small cell lung cancer (NSCLC) develop bone
metastases, and adenocarcinoma is the most frequent histological subtype.[4 ]
[9 ]
[10 ]
[11 ]
[12 ]
[13 ] In a previous study, we showed that lung cancer histology influences the clinicopathological
features of bone metastases.[14 ] However, this study aimed to assess whether the risk of bone metastases development
is related to the histological subtype of lung carcinoma. These findings may be useful
in guiding early surveillance for bone metastases detection or interventions in high-risk
groups to improve quality of life and patient survival.
Material and Methods
Population
This study retrospectively evaluated the medical records of 413 patients diagnosed
with primary malignant lung tumors at our institution between 2003 and 2012. The study
design was approved by our Research Ethics Committee. The inclusion criteria were
histopathological diagnoses of NSCLC or small cell lung cancer (SCLC) with complete
tumor staging data. Lung carcinoma was classified according to the histological subtype
per the World Health Organization criteria: adenocarcinoma, squamous cell carcinoma
(SCC), large cell carcinoma (LCC), unspecified NSCLC (NSCLC, unsp) and SCLC.[15 ]
The events of interest were occurrence of bone metastases and death. Bone metastases
were diagnosed by histopathological examination of bone biopsy samples or bone scintigraphy
with Tc99 plus two additional imaging tests (radiography, computed tomography or magnetic
resonance imaging). These metastases were classified according to the time of diagnosis
of the first bone metastasis in relation to the time of lung carcinoma diagnosis.
Synchronous metastases were defined as bone metastases present at the time of lung
carcinoma diagnosis, while metachronous metastases were defined as bone metastases
occurring after the diagnosis of the primary tumor. To accurately assess overall survival,
patients diagnosed with a second primary malignant tumor and those with unknown death
dates were excluded. The minimum follow-up period was 24 months for patients who developed
bone metastases (including cases of death in less than 24 months).
Statistical Analysis
The chi-square test was used to compare the proportions of histological subtypes among
patients who developed bone metastases or not. The relationship between histological
subtype and bone metastases occurrence was also evaluated by odds ratio (OR) and 95%
confidence interval (95% CI) determinations. Overall survival was assessed using the
Kaplan-Meier method. All analyzes were performed in the SPSS for Windows, version
10.0 (SPSS Inc., Chicago, IL, USA), and a p value < 0.05 was considered statistically significant.
Results
Patients Characteristics
In total, 407 patients met the inclusion criteria and were eligible for analysis.
The cohort diagram is shown in [Figure 1 ], and the demographic characteristics of the patients are shown in [Table 1 ]. Average age was 63.4 years-old, and 61.4% of the patients were men. Approximately
13.5% of the patients never smoked; smokers were characterized per a smoking load
of < 40 packs/year (23.6%) or ≥ 40 packs/year (62.9%). Adenocarcinoma was the most
common subtype (n = 190, 46.7%), followed by SCC (n = 102, 25.1%), NSCLC, unsp (n = 56, 13.7%), SCLC (n = 48, 11.8%), and LCC (n = 11, 2.7%) ([Table 1 ]); A total of 115 patients presented bone metastases, and the characteristics of
patients with (n = 115) or without bone metastases (n = 292) are shown in [Table 1 ].
Table 1
Characteristic
Patients with lung carcinoma
Patients with lung carcinoma
Patients with lung carcinoma
General cohort
Without bone metastasis
With bone metastasis
n = 407
n = 292
n = 115
n
%
n
%
n
%
Average age (range)
63.4 years-old (32–87)
65.7 years-old (42–84)
62.6 years-old (32–87)
Gender
Male
250
61.4
175
59.3
75
65.2
Female
157
38.6
117
40.7
40
34.8
Smoking
Never smoke
55
13.5
30
10.3
25
21.7
< 40 packs/year
96
23.6
63
21.6
33
28.7
≥ 40 packs/year
256
62.9
199
68.1
57
49.6
Histology
Adenocarcinoma
190
46.7
123
42.1
67
58.3
Squamous cell carcinoma
102
25.1
87
29.8
15
13
Large cell lung cancer
11
2.7
9
3.1
2
1.8
Non-small cell lung cancer, unspecified
56
13.7
40
13.7
16
13.9
Small cell lung cancer
48
11.8
33
11.3
15
13
Fig. 1 Cohort diagram. Patients with pulmonary carcinoma according to the stage at diagnosis
and the presence of metastasis at diagnosis or during the follow-up period.
Bone Metastasis Prevalence
The overall prevalence of bone metastases was 28.2% (n = 115). The prevalence of bone metastases according to histology is shown in [Figure 2 ]. Bone metastases were more prevalent among patients with adenocarcinoma (n = 67; 35.3%). Bone involvement was also observed in patients with SCC (n = 15, 14.7%), LCC (n = 2, 18.1%), NSCLC, unsp (n = 16, 28.6%), and SCLC (n = 15, 31.2%). Synchronous metastases were significantly more frequent than metachronous
metastases (n = 81, 70.4% versus n = 34, 29.6%, p = 0.0021).
Fig. 2 Bone metastasis prevalence according to histology.
Metastasis Number and Location
Among the 115 patients with bone metastases, there were 305 tumors (approximately
2.65 per patient). The most frequent sites were the spine (98 metastases, 32.1%),
the pelvic girdle (53 metastases, 17.4%), the proximal femur and humerus (52 metastases,
17.1%), and the thoracic wall (46 metastases; 15.1%) ([Fig. 3 ]).
Fig. 3 Bone metastasis number and location.
Bone Metastasis Risk
The analysis of the proportions of patients who developed bone metastases or not according
to histology showed that adenocarcinoma was the most frequent histological subtype
in both groups, although the frequency of adenocarcinoma was significantly higher
among patients with metastases (n = 67; 58.3%) compared to patients without bone metastases (n = 123; 42.1%) (p = 0.003). The frequency of SCC was significantly lower among patients who developed
bone metastases (n = 15, 13.0% versus n = 87, 29.8%, p = 0.0004). However, there was no statistical difference between the
other histological subgroups: LCC (n = 2, 1.7% versus n = 9, 3.1%, p = 0.451), NSCLC, unsp (n = 16, 13.9% versus n = 40, 13.7, p = 0.954), and SCLC (n = 15, 13.0% versus n = 33, 11.3%, p = 0.623) ([Fig. 4 ]).
Fig. 4 Comparison of histological subtypes frequency in groups of patients with lung carcinoma
that developed bone metastasis or not. *p -value < 0.05.
The associations between histology and bone metastases occurrence were evaluated through
ORs and 95% CI. The ORs for bone metastases development were 1.92 (95% CI 1.29–2.97)
for adenocarcinoma, 0.35 (95% CI 0.19-0.64) for SCC, 0.55 (95% CI 0.12–2.61) for LCC,
1.01 (95% CI 0.54–1.90) for NSCLC, and 1.17 (95% CI 0.61–2.26) for SCLC ([Table 2 ]).
Table 2
Histology
Odds ratio
95% confidence interval
Adenocarcinoma
1.92
1.29–2.97
Squamous cell carcinoma
0.35
0.19–0.64
Large cell lung cancer
0.55
0.12–2.61
Non-small cell lung cancer, unspecified
1.01
0.54–1.90
Small cell lung cancer
1.17
0.61–2.26
Survival
The survival results were calculated after the exclusion of 20 patients diagnosed
with a second primary malignant tumor (n = 12) and those with unknown date of death (n = 8). Median survival after bone metastases
diagnosis was 4 months. Median survival for adenocarcinoma, SCC, LCC, NSCLC, and SCLC
was 3.0 months, 4.5 months, 9.0 months, 2.5 months, and 4.5 months, respectively.
Discussion
In this cohort study of patients with lung carcinoma treated at the same university
general hospital in Brazil, the prevalence of secondary bone involvement was approximately
28%. This result is similar to previously obtained values, demonstrating that bone
tissue is one of the main sites for lung carcinoma metastasis.[16 ] Studies comparing the frequency of metastatic NSCLC sites showed that the frequency
of bone involvement (20–40%) is comparable to that of the liver (25–30%) and the contralateral
lung (40–50%).[7 ]
[10 ]
[17 ] We believe that assessment methods influence the incidence of bone metastases, since
studies from the 1990s using bone scintigraphy demonstrated that the incidence was < 20%.[18 ]
[19 ]
[20 ]
[21 ] On the other hand, more recent studies using positron emission tomography revealed
a frequency ranging from 20 to 40%.[9 ]
[22 ]
[23 ] Tsuyia et al[10 ] identified 70 (30.4%) patients with bone metastases from lung cancer using scintigraphy,
radiography, and magnetic resonance imaging. Their methods of detection and the frequency
of metastases were similar to those from this study. However, our sample is larger
due to the inclusion of a patient with SCLC.
The incidence of adenocarcinoma increased over time and surpassed SCC as the most
prevalent lung neoplasm subtype.[24 ] In this study, adenocarcinoma was the most prevalent histologic subtype among patients
with or without bone metastases. Our results agree with international epidemiological
reports demonstrating that adenocarcinoma accounts for > 45% of all cases.[25 ] However, studies in the Brazilian population have shown that SCC remains the most
prevalent subtype in some areas of the country.[26 ] We found that the prevalence of bone metastases varied according to histology, with
a higher prevalence for adenocarcinoma (35.3%; n = 67) and SCLC (31.2%; n = 15), and a lower prevalence for SCC (14.7%; n = 15). These results indicate that the probability of bone dissemination may vary
according to the histology of lung tumor. Thus, we compared the frequencies of histological
subtypes between patients with and without bone metastases to test this hypothesis.
In such analysis, the frequency of bone metastases was significantly and positively
associated with adenocarcinoma (58.3% versus 42.1%, p = 0.003), and it was significantly and negatively associated with SCC (29.8% versus
13.0%, p = 0.0004). A recent cohort study assessed the entire Danish population and reported
a higher frequency of adenocarcinoma among patients with bone metastases (50.3%) compared
to all patients with lung neoplasms (37.9%), and a lower frequency of SCC (13.0%)
compared to all patients with lung neoplasms (24.6%).[11 ] However, such study did not evaluate the possible association between histology
and frequency of metastases. Lorusso et al[12 ] also reported that adenocarcinoma was the most frequent histological subtype, affecting
78% of the patients in their sample. Similarly, Kagohashi et al[13 ] reported that 67% of 24 patients with bone metastases had the adenocarcinoma subtype.
In addition, in the initial study sample from Tsuya et al,[10 ] adenocarcinoma was observed in 61% of 259 patients with NSCLC, although the frequency
among the 70 patients who developed bone metastases was not assessed. Prior to our
publication of an earlier study suggesting that the histological subtype of lung carcinoma
could influence the clinicopathological characteristics of bone metastases, we had
found only one paper suggesting that adenocarcinoma has a particular tropism for bone
dissemination, and that study reported that 45% of adenocarcinomas resulted in bone
metastases.[21 ] The prevalence of bone metastases in the current study was a little lower (35%),
although we have observed significantly higher and lower risks of bone metastases
development from adenocarcinomas and SCCs, respectively.
Many studies have evaluated bone involvement in lung neoplasms without including SCLC
patients, which was the second most frequent cancer subtype in the current study.
Thus, it remains unclear whether the frequency and other characteristics of bone metastases
for this histologic subtype differ from those of other subtypes. However, the relatively
high frequency of bone metastases is consistent with the more aggressive nature of
this cancer.[27 ] Populational-based studies from Cetin et al[11 ] and Sathiakumar et al[28 ] reported higher rates of bone dissemination for NSCLC. Nevertheless, these publications
were among the few that evaluated SCLC, diagnosed in 16.7% and 13% of patients with
bone metastases, respectively, secondary to lung carcinoma. These incidences are similar
to the 13% we observed in this study. On the other hand, the 31.2% prevalence of bone
metastases in the SCLC patients in our study was higher when compared to the prevalence
of 5.8% observed by Cetin et al[11 ] and 23.3% noted by Sathiakumar et al.[28 ] These results indicate the need for further studies carefully evaluating the characteristics
of bone metastases in these individuals.
In this study, most patients with bone metastases exhibited bone involvement at the
time of the diagnosis of the lung neoplasm (70.4%). These results may indicate the
spreading ability of lung neoplasm in the early stages of the disease.[29 ] However, our frequency might also have been overestimated as a result of a possible
selection bias, since our institution became a reference for patients with bone tumors
in 2010, resulting in a significant increase in the number of individuals with bone
metastases from primary sites not determined at their hospital of origin.
The present study revealed that survival after bone metastases diagnosis is reduced,
and some other studies have determined that the presence of bone metastases indicates
a poor prognosis for lung cancer.[25 ]
[27 ]
[30 ] Hansen et al[30 ] published a report about the effect of bone metastases on survival in several types
of carcinomas and showed that patients with lung cancer had an average survival time
of 3 months after the diagnosis of these metastases. The median survival time in the
current study, of 4 months, was lower when compared to that of most previous international
studies and may be the result from the relatively high frequency of patients initially
diagnosed in stage IV.
Our results are closer to those from a retrospective French study, which estimated
a mean survival time of 5.8 months. However, the design of our study does not allow
the comparison of survival after diagnosis among patients with or without bone metastases
because our goal was to evaluate survival among individuals with these metastases.
In addition, the primary lung site has been reported as the main prognostic factor
in patients with carcinomatous bone metastases.[25 ]
[27 ] Thus, extreme low survival in current studies confirms three assumptions: 1) Lung
neoplasia survival is low; 2) Bone metastases are a predictor of poor prognosis for
any carcinoma and 3) The main prognostic factor is lung tumor location in patients
with metastatic carcinomas. However, we did not observe a statistically significant
difference in survival time according to histology.
The advantages of this study are the larger Brazilian population sample and the inclusion
of SCLC in analyzing the association between bone metastases and lung carcinoma histology.
However, our findings point to the need for further studies comparing clinicopathological
features according to histology. The main limitations of this study include a possible
selection bias, as the institution became a reference hospital for bone tumors in
2010 (which increased the frequency of patients with synchronous metastases).
Conclusion
Adenocarcinoma was the most common cancer subtype in patients with lung carcinoma
with or without bone metastases. In addition, adenocarcinoma was associated with a
higher risk of developing bone metastases, while SCC was associated with lower risk.
These data suggest that patients with adenocarcinoma may benefit from a more cautious
surveillance program aimed at the early detection and treatment of bone metastases.
