The diffusing capacity of the lung for carbon monoxide is associated with the histopathological aggressiveness of lung adenocarcinoma†

Abstract

OBJECTIVES

The diffusing capacity of the lung for carbon monoxide (DL_CO) is an indicator of lung damage. We sought to determine whether DL_CO is associated with the aggressiveness of lung adenocarcinoma using histopathological indexes, such as tumour differentiation, scar grade, nuclear atypia and the mitotic index.

METHODS

Fifty-seven patients with low DL_CO (≤80% of predicted) and 466 patients with normal DL_CO (>80% of predicted) who underwent R0 resection of lung adenocarcinoma between 2005 and 2012 were retrospectively reviewed. The relationships between the DL_CO status and each histopathological index as well as the overall survival were evaluated.

RESULTS

Low DL_CO had significant relationships with moderate/poor differentiation (79% vs 57% [low DL_CO vs normal DL_CO]), scar grade 3/4 (37% vs 18%), nuclear atypia 3 (65% vs 30%) and the mitotic index 3 (26% vs 8%). After adjusting for the age, sex, forced expiratory volume in 1 s, smoking status and tumour size, a low DL_CO still showed a significant correlation with the histopathological indexes. These histopathological indexes were all significant factors for the overall survival on log-rank tests. In a multivariable Cox regression analysis with 13 clinicopathological variables, moderate/poor differentiation and nuclear atypia Grade 3 were significant histopathological factors for the overall survival (hazard ratios: 2.16 and 1.84; 95% confidence intervals: 1.10–4.51 and 1.06-3.21; P = 0.024 and 0.029, respectively).

CONCLUSIONS

Our findings regarding the relationship between DL_CO and the histopathological indexes of lung adenocarcinoma suggest that lung damage may be associated with carcinogenesis and progression.

INTRODUCTION

The diffusing capacity for carbon monoxide (DL_CO) is a surrogate marker for lung damage, as found in patients with emphysema, pulmonary fibrosis and similar lung diseases, and it is useful during lung cancer screening in patients with chronic obstructive pulmonary disease [1, 2]. In addition, DL_CO has been reported to be strongly related to the postoperative overall survival (OS) in patients with non-small-cell lung cancer [3–8]. We can expect a shorter OS in patients with poor DL_CO due to the influence of pulmonary complications. However, the relationship between the aggressiveness of lung adenocarcinoma and DL_CO is unclear.

Lung adenocarcinoma is the most common type of lung cancer at our institution. Several histopathological indexes have been reported to be associated with the aggressiveness of lung adenocarcinoma. The tumour differentiation, scar grade, nuclear atypia and mitotic index are histopathological indexes reported to be prognostic factors [9–14]. The scar grade is based on the degree of fibrosis, nuclear atypia is based on the nuclear diameter of tumour, and the mitotic index is based on the number of mitotic cells per 10 high-power fields.

In this study, we investigated patients with lung adenocarcinoma to evaluate the histopathological aggressiveness. We hypothesized that more aggressive adenocarcinoma develops in more severely damaged lung and that DL_CO has some relationship with the histopathological aggressiveness. We conducted this study to clarify the relationship between DL_CO and the histopathological indexes of lung adenocarcinoma, such as tumour differentiation, scar grade, nuclear atypia and the mitotic index and to determine its prognostic value.

MATERIALS AND METHODS

This study was conducted with the approval of the Institutional Review Board of Nagoya University Hospital. We extracted the clinicopathological data of patients who underwent R0 resection for pathological Stage IA to IIIA lung adenocarcinoma without preoperative therapy at Nagoya University Hospital between 2005 and 2012. Using these criteria, 523 patients with full data available were enrolled in this study. All patients were considered for surgery after the assessment of their general condition and comorbidities by a preoperative interview, physical examinations, blood tests, urine tests, cultures, electrocardiography, spirometry and further management (if required). All patients received regular postoperative follow-up examinations by thoracic surgeons and/or respirologists. The follow-up data were updated annually (last follow-up: 19 November 2015).

The postoperative patients were scheduled for follow-up every 1–3 months for 2 years and every 6 months thereafter. They were then surveyed by physical examinations, chest roentgenograms and measurements of their serum carcinoembryonic antigen levels to detect recurrence [15]. At a minimum, computed tomography of the chest and abdomen was performed at 2 and 5 years after surgery in patients with Stage I tumours and every 12 months after surgery in patients with Stage II and III tumours, in accordance with the physician’s decision.

The seventh edition of the tumour–node–metastasis classification [16] was applied in this cohort. The pathological diagnosis of the tumour was made based on the definition of the World Health Organization classification [17]. The scar grade, nuclear atypia and mitotic index were assessed as described previously [9, 11]. We adopted the modified scar grade previously reported by Maeshima et al. [11]. The histopathological diagnosis of the tumour was made by a number of pathologists. The cut-off values of DL_CO and FEV_1.0 were set to 80%, based on the clinical guidelines [3, 18].

Statistical analyses

Fisher’s exact test and Student’s t-test were used to compare the distribution of categorical and continuous values between the 2 groups, respectively. A multivariable logistic regression analysis was performed to estimate the odds ratios (ORs) and 95% confidence intervals for each of the histopathological indexes using the following clinical variables: the DL_CO status_, the forced expiratory volume in 1 s (FEV_1.0) status, age, sex, smoking status and tumour size. These clinical variables were selected because they are known to be common prognostic factors and might be related to tumour aggressiveness. The DL_CO status, sex, smoking status and tumour size were found to be significant predictive factors for each of the histopathological indexes in a univariable analysis (P < 0.05). The FEV_1.0 status was found to be a significant predictive factor for tumour differentiation (P < 0.05) and a marginal predictive factor for nuclear atypia and the mitotic index (P < 0.10).

The OS was defined as the time from surgery to death due to any cause. The cancer-specific survival (CSS) was defined as the time from surgery to death due to lung cancer, and 34 patients who died due to reasons other than lung cancer were censored. The Kaplan–Meier method was used to estimate the OS and CSS, and the log-rank test was used to compare the survival curves. A multivariable Cox regression analysis was performed to estimate the hazard ratios (HRs) and 95% confidence intervals for the OS using the DL_CO status, FEV_1.0 status, age, sex, smoking status, tumour size, pathological N status, lymphatic permeation, vascular invasion and histopathological indexes—clinicopathological variables that are known to be common prognostic factors. With the exception of age (P = 0.20), all of these variables had been found to be significant prognostic factors for the OS in a univariable analysis (P < 0.05).

The proportional hazard assumptions for OS and CCS were verified using the log of the negative log of the estimated survivor functions for the DL_CO status. Statistical significance was defined as P < 0.05. All analyses were conducted using the JMP software program (version 11.0.0; SAS institute Inc., Cary, NC, USA).

RESULTS

Clinicopathological characteristics

Table 1 shows the clinicopathological characteristics of the 57 patients with low DL_CO (≤80% of predicted) and 466 patients with normal DL_CO (>80% of predicted). The low DL_CO patient subgroup was characterized by a lower FEV_1.0 status, a greater proportion of males, a higher number of current or former smokers, a larger tumour size and a greater proportion of histopathological N1/2 disease, lymphatic permeation, moderate/poor tumour differentiation, scar grade 3/4, nuclear atypia 3 and mitotic index 3 than the normal DL_CO group. Six (11%) patients with low DL_CO and 22 (5%) patients with normal DL_CO were identified as having pulmonary fibrosis (P = 0.066). After adjusting for the clinical variables (the DL_co status, FEV_1.0 status, age, sex, smoking status and tumour size) in a multivariable logistic regression analysis, low DL_CO remained a significant predictor of moderate/poor tumour differentiation [OR: 2.00 (95% CI: 1.02–4.18); P = 0.045], scar grade 3/4 [OR: 2.19 (95% CI: 1.15–4.08); P = 0.017], nuclear atypia 3 [OR: 3.58 (95% CI: 1.95–6.76); P < 0.001] and mitotic index 3 [OR: 2.57 (95% CI: 1.19–5.37); P = 0.017] (Table 2). Despite the association between the DL_CO status and the histopathological indexes, the FEV_1.0 status had no significant relationship with the histopathological indexes.

The association between non-smokers and the development of well-differentiated adenocarcinoma with a better prognosis is already known. To eliminate this influence, we assessed the data of 307 former or current smokers, whose clinicopathological characteristics are shown in Table 3. As mentioned earlier, the low DL_CO patient subgroup was also characterized by a greater proportion of moderate/poor tumour differentiation, scar grade 3/4, nuclear atypia 3 and mitotic index 3 than the normal DL_CO group. Similarly, after adjusting for the clinical variables (the DL_CO status_, FEV_1.0 status, age, sex and tumour size), low DL_CO still showed a significant correlation with moderate/poor tumour differentiation [OR: 3.18 (95% CI: 1.36–8.75); P = 0.006], scar grade 3/4 [OR: 2.40 (95% CI: 1.19–4.79); P = 0.016], nuclear atypia 3 [OR: 5.23 (95% CI: 2.56–11.46); P < 0.001] and mitotic index 3 [OR: 2.80 (95% CI: 1.28–5.96); P = 0.011].

In addition, we demonstrated the analysis using the data of 450 patients with pN0 tumours. The tumours of the Low DL_CO patients had a greater proportion of moderate/poor tumour differentiation (P = 0.007), scar grade 3/4 (P < 0.001), nuclear atypia 3 (P < 0.001) and mitotic index 3 (P < 0.001). After adjusting for the clinical variables (the DL_CO status_, FEV_1.0 status, age, sex, smoking status and tumour size), the low DL_CO status showed a marginal correlation with moderate/poor tumour differentiation [OR: 1.83 (95%CI: 0.90–3.93); P = 0.096] and a significant correlation with scar grade 3/4 [OR: 2.84 (95%CI: 1.36–5.80); P = 0.006], nuclear atypia 3 [OR: 3.99 (95%CI: 2.02–8.07); P < 0.001] and mitotic index 3 [OR: 3.06 (95%CI: 1.19–7.47); P = 0.021].

Overall survival and cancer-specific survival curves

The median follow-up period was 49 months (range: 2–123 months). The OS of the low-DL_CO patients was significantly shorter than that of the normal-DL_CO patients (5-year OS rates: 64% vs 87%; P < 0.001; Fig. 1A). The CSS of the low-DL_CO patients was also significantly shorter than that of the normal-DL_CO patients (5-year CSS rates: 75% vs 92%; P = 0.001; Fig. 1B).

Figure 2 shows the OS curves classified by tumour differentiation, scar grade, nuclear atypia and the mitotic index. These histopathological indexes were all significant factors for the OS on log-rank tests.

A multivariable Cox regression analysis of the histopathological indexes for the overall survival

Table 4 shows the results of a multivariable Cox regression analysis of the clinicopathological variables. Moderate/poor differentiation [HR: 2.16 (95% CI: 1.10–4.51); P = 0.024] and nuclear atypia Grade 3 [HR: 1.84 (95% CI: 1.06–3.21); P = 0.029] were found to be significant factors for the OS, while scar grade 3/4 [HR: 1.06 (95% CI: 0.63–1.74); P = 0.82] and mitotic index 3 [HR: 1.19 (95% CI: 0.62–2.21); P = 0.60] were not significant factors.

DISCUSSION

To our knowledge, this study is the first to elucidate the relationship between DL_CO and the histopathological aggressiveness of lung adenocarcinoma. Furthermore, the prognostic value of tumour differentiation, scar grade, nuclear atypia and the mitotic index were investigated in a larger number of patients (limited to those with adenocarcinoma) than in previously published studies.

Several studies have referenced the influence of tumour differentiation, scar grade, nuclear atypia and the mitotic index of lung adenocarcinoma on the survival [9–14]. Shimizu et al. also reported the univariable prognostic impact of the histopathological indexes of tumour differentiation, scar grade, nuclear atypia and the mitotic index on the OS in 1074 patients with non-small-cell lung cancer [14]. In the present study, we performed a multivariable Cox regression analysis of 523 patients with lung adenocarcinoma, and our results suggest that tumour differentiation and nuclear atypia are the strongest prognostic factors predicting the OS.

To date, no detailed study has examined the relationship between the pulmonary function and histopathological aggressiveness of lung adenocarcinoma. In this study, we examined the relationship between the clinical variables (including DL_CO and FEV_1.0) and tumour aggressiveness using a multivariable logistic regression analysis (Table 2). The DL_CO status and tumour size were found to be significantly correlated with all of the histopathological indexes, while sex and smoking status showed a limited correlation with the histopathological indices. The FEV_1.0 status had no significant correlation with the histopathological indexes, which might mean that it did not exactly reflect the histocytological changes in the damaged lung. From the results above, we hypothesized that the difference in the condition of the lung (the DL_CO status) is associated with the carcinogenesis and progression of lung adenocarcinoma.

Lung adenocarcinoma is considered to develop in a mostly linear multistep progressive manner: atypical adenomatous hyperplasia to adenocarcinoma in situ, followed by invasive adenocarcinoma. However, Yatabe et al. reported that lung adenocarcinoma can be considered as a subset of cancers that arise from different molecular pathways and that it is possible that not all lung adenocarcinomas show linear progression [19]. Several factors, including—but not limited to—the mutational status, smoking status, oestrogen and air pollution have been reported to be associated with carcinogenesis and the progression of lung adenocarcinoma [20–23]. Dacic et al. reported the correlation between the epidermal growth factor receptor mutational status and sex, smoking status and tumour differentiation, but showed no correlation with scarring and vascular–lymphatic invasion [20]. Maeshima et al. reported the correlation between smoking history and the degree of histopathological aggressiveness, including the scar grade [21]. They indicated the effects of smoking on carcinogenesis. We hypothesized that differences in the carcinogenesis associated with these factors might be responsible for the difference in the histopathological aggressiveness of lung adenocarcinoma. In this study, the DL_CO status had a significant correlation with all of the histopathological indexes; however, the detailed mechanism underlying this phenomenon remains to be elucidated.

Limitations

Several limitations associated with the present study warrant mention. First, several pathologists made the pathological diagnosis, and the evaluation criteria of histopathological aggressiveness were not always consistent. Second, details regarding the genetic status, such as driver mutation status, were not available.

CONCLUSION

In conclusion, our findings regarding the relationships between DL_CO and the histopathological indexes of lung adenocarcinoma suggest that damaged lung with low DL_CO is associated with carcinogenesis and progression.

Conflict of interest: none declared.

REFERENCES

Author notes