Introduction

The development of medical technologies, progress in the early diagnosis of malignancies and the emergence of new treatments has led to a gradual increase in the survival times of patients with cancer. Although lung cancer is the leading cause of death related to malignant tumors in China [1], improved survival rates have been reported [2]. As a result, the incidence of MPMN has also increased due to the development in survivors of malignant tumors. In a recent study of 24,859 patients registered with cancer, 1 in 8 cases had developed second primary neoplasms [3]. It has also been reported that the number of cancer survivors is increasing by about 2% each year [4].

In addition, the risks of contracting second primary cancers were higher in cancer survivors compared with the general population with a 3.8% higher incidence of metachronous second primary cancers within a median follow-up time of 2.5 years; what’s more, the estimated 10-year cumulative risk of second primary cancers for patients who were firstly diagnosed with cancer within age 60 to 69 was as high as 13% [5]. In the present study, we collected the clinical data of MPMN patients with lung cancer and analyzed their clinical characteristics and prognoses. The aim of our investigation was to determine the incidence of MPMN with lung cancer, the most common concomitant malignancies, the most common onset time interval of the two cancers, and the factors that affect the survival of MPMN patients with lung cancer.

Methods

Patients

The clinical and pathological data of 14,528 lung cancer patients, who were diagnosed and treated in the tumor hospital affiliated with Zhengzhou University from January 2008 to August 2015 were retrospectively analyzed.

Diagnostic criteria and staging

MPMN was based on the diagnostic criteria proposed by Warren and Gates [6]: (1) Malignancy of each type of tumor must be confirmed by histocytology; (2) Each malignancy occurs in different regions or organs; (3) Must confirm that the new emergent cancer is non-metastatic cancer; (4) Each malignancy has its own pathological characteristics. There are two MPMN categories according to the time interval between the two tumor occurrences. Synchronous MPMN (SMPMN) occurring within ≤ 6 months and metachronous MPMN (MMPMN) occurring at > 6 months after the initial cancer diagnosis. Lung cancer first (LCF) comprised the cases in which the lung cancer proceeded the second primary malignancy whereas in the other cancer first (OCF) group, lung cancer was a second primary malignancy. Lung cancers were staged using the 7th edition of lung cancer TNM staging criteria developed by the Union for International Cancer Control (UICC) [7].

Collection of data

Gender, age, smoking history, disease history, family history of malignant tumors, pathological lung cancer types and the staging of lung cancer, as well as histological types, location and the onset time of MPMN and the time interval of diagnosis between the two primary malignancies in the SMPMN and MMPMN, LCF and OCF groups were compared. The primary endpoint of the study was OS, defined as time interval between first primary cancer diagnosis and patient’s death. Censored survival time data were from the pathological diagnosis to the last follow-up.

A follow-up was performed during outpatient visits, during hospitalization or via telephone on 350 patients diagnosed as MPMN with lung cancer. Patients who were not followed-up in hospital, denied follow-up on the phone or who could not be contacted due to phone number changes were recorded as lost. The follow-up ended on September 30, 2015. The median follow-up times of both groups were the same being 5.37 years (1 month–31 years). At the end of follow-up, 35 patients (10.0%) were lost, 256 patients (73.1%) died, and 59 patients (16.9%) survived; the follow-up rate was 90%.

Statistical analysis

All statistical analyses were performed with SPSS for Windows (Version 16.0. Chicago, SPSS Inc.). Data are expressed as the mean ± SD for normally distributed data or median interquartile range (IQR) for non-normally distributed data. A χ2 or fisher’ exact test was used for qualitative data. Mann–Whitney U tests were used for comparison of means of quantitative variables. The Kaplan–Meier method was used for survival analysis and the log-rank test was used to compare differences of survival rates between different groups. Univariate and multivariate COX proportional hazard regression model analyses were used to determine the factors associated with OS. A P-value < 0.05 was considered to be statistically significant.

Considering 90% power for detection of an expected overall survival rate of 3 years in 42.0% of the SMPMN and 65% of the MMPMN patients, based on a 1:4 ratio of SMPMN to MMPMN cases and a significance level of 0.05, 300 patients in total were needed with 60 cases in the SMPMN and 240 in the MMPMN group. Considering 15% losses during follow-up at least 353 subjects were needed for the study initiation.

Results

General characteristics of MPM patients

There were 364 MPMN patients (2.5%) who met the diagnostic criteria with complete follow-up data including 350 cases of primary malignancies diagnosed twice, 13 cases of primary malignancies diagnosed three times and one case of primary malignancies diagnosed four times (Fig. 1).

Fig. 1
figure 1

Flow chart of the present study

Full size image

Comparison of general pathologic characteristics between SMPMN and MMPMN as well as LCF and OCF patients

As shown in Table 1, 74 patients were SMPMN cases (21.1%), 276 patients were MMPMN cases (78.9%) and 93 patients belonged to the LCF (26.6%) group, whereas 257 patients belonged to the OCF (73.4%) group. Among the 350 cases of primary malignancies diagnosed twice, 224 patients were male (224/350, 64.0%).

Table 1 General characteristics of 350 MPM patients (n/%)
Full size table

The histological type of lung cancer was mainly adenocarcinoma. The median diagnostic age of the first and second primary cancers was 62 years in SMPMN patients. The median diagnostic age of the first primary cancer was 56 years and the second primary cancer 63 years in MMPMN patients.

The median time interval between the onset of the two primary malignancies in MMPM patients was 5 years (range, 6 months–30 years) and in SMPMN patients 0 month (range 0–5 months).

SMPMN and MMPMN cases were more common in males and mainly in male smokers. Between the SMPMN and MMPMN groups, the pathological types of lung cancers were significantly different (P < 0.04). Both the proportions of male patients and smokers among male patients in the LCF group, were higher than that in the OCF group. The NSCLC staging was worse in the LCF group compared to the OCF group.

The composition of the histopathological types of lung cancer were not significantly different in LCF and OCF patients (Table 1).

Comparison of primary malignancy onsets LCF and OCF groups

The median age of first primary cancer diagnosed in the LCF group was 61 years and 57 years in the OCF group (P = 0.008), while there was no significant difference in the median diagnostic age of primary cancer diagnosed second (P = 0.544) (Table 2).

Table 2 The onset age of 350 cases of primary lung cancer diagnosed twice
Full size table

Distribution of concomitant malignancies in MPMN patients with lung cancer

The five most common concomitant malignancies of MPMN lung cancer cases were esophageal cancer (51, 14.6%), breast cancer (41, 11.7%), gastric cancer (35, 10.0%), colorectal cancer (32, 9.1%), and second primary lung cancer (30, 8.6%). The five most concomitant malignancies of patients in the LCF group were lung cancer (30, 32.3%), esophageal cancer (11, 11.8%), gastric cancer (9, 9.7%) as well as thyroid (7, 7.5%), and breast cancer (7, 7.5%). The five most concomitant malignancies of patients in the OCF group were esophageal cancer (40, 15.6%), breast cancer (34, 13.2%), colorectal cancer (29, 11.3%), cervical cancer (26, 10.1%), and gastric cancer (26, 10.1%). The most concomitant malignancies of male patients were esophageal cancer (46, 20.6%), gastric cancer (30, 13.5%), second primary lung cancer (26, 11.7%), colorectal cancer (25, 11.2%), and laryngeal cancer (16, 7.2%), whereas the most common concomitant malignancies in females were breast cancer (39, 30.7%), cervical cancer (27, 21.3%), and thyroid cancer (14, 11.0%) (Supplementary Table 1).

Analysis of prognostic factors

The median survival time of the 350 patients was 6.83 years, the 1-year, 3-year, and 5-year survival rates were 94.8%, 78.4%, and 63.4%, respectively (Fig. 2a). After the diagnosis of the first primary cancer, the overall survival in the OCF group was superior to that in the LCF group (P < 0.001) (Fig. 2b). There was no significant difference in the overall survival time between the LCF group and the OCF group after the diagnosis of the second primary cancer (P = 0.120) (Fig. 2c).

Fig. 2
figure 2

Overall survival time curves for a all 350 patients with primary cancer diagnosed twice; b LCF and OCF patients after first primary cancer diagnosis; c LCF and OCF patients after second primary cancer diagnosis; d SMPM and MMPM patients; e LCF SMPM and MMPM patients; f OCF SMPM and MMPM patients; g for different lung cancer treatment modes; h for age of the first primary cancer diagnosis

Full size image

MMPMN patients had a better prognosis than SMPMN patients (Fig. 2d, P < 0.001) in both the LCF (Fig. 2e, P < 0.001) and OCF (Fig. 2f, P < 0.001) groups.

The median survival times of MMPMN and SMPMN patients in the LCF group were 6.83 years and 2.42 years, respectively (Fig. 2d). The 1-year, 3-year, and 5-year survival rates of LCF SMPMN patients were 75.0%, 25.1%, and 25.1%, and the 1-year, 3-year, and 5-year survival rates of LCF MMPM patients were 98.3%, 73.6%, and 56.8%, respectively (Fig. 2e). The median survival times of OCF MMPMN and SMPMN patients were 8.67 years and 2.25 years, respectively. The 1-year, 3-year, and 5-year survival rates of OCF SMPMN patients were 76.6%, 37.3%, and 18.7%, and the 1-year, 3-year, and 5-year survival rates of OCF MMPMN were 100%, 90.7%, and 75.0%, respectively (Fig. 2f).

The overall survival of patients < 60 years old at the time of primary cancer diagnosis was significantly better than that of patients ≥ 60 years old (P < 0.001) (Fig. 2g).

Univariate and multivariate analysis were merely the statistical methods used for the determination of prognostic factors

An univariate Cox proportional hazard regression analysis showed that age of primary cancer diagnosed first, the onset order of lung cancer, the time interval between the onset of two primary malignancies, other histological type of lung cancer than adenocarcinoma, squamous cell carcinoma or small cell carcinoma and NSCLC staging >II were factors which had significant effects on survival time of MPM patients (Table 3). A multivariate Cox proportional hazard regression analysis showed that the onset age of the first primary cancer diagnosis, onset order of lung cancer, time interval of the two primary malignancies (metachronous vs. synchronous), histological type of lung cancer and the NSCLC staging were independent factors that affected the prognosis of lung cancer-related MPM (Table 4).

Table 3 Univariate COX proportional hazard regression model analysis of factors associated with survival in 350 MPM patients with lung cancer
Full size table
Table 4 Multivariate COX proportional hazard regression model analysis of factors associated with survival in 350 MPM patients with lung cancer
Full size table

Discussion

In this study, we found the incidence rate of MPMN with lung cancer was 2.5%, which is in line with a previous Chinese study that reported a rate of 3.4% [8].

In the present study, we found that SMPM accounted for 21.1% (74/350), and MMPM accounted for 78.9% (276/350) of cases, which is in agreement with a previous report from China [9]. Clinically, MPMN must be distinguished from cancer recurrence or metastasis, since MPMN refers to the occurrence of a new primary malignancy with completely different characteristics to the initial primary tumor [10]. The most concomitant malignancies of male patients were esophageal cancer, gastric cancer, second primary lung cancer, colorectal cancer and laryngeal cancer, whereas the most common concomitant malignancies in females were breast cancer, cervical cancer, and thyroid cancer. In contrast, Fujita and colleagues [11] reported that the top three most common concomitant malignancies in male lung cancer-related MPMN patients were gastric cancer, prostate cancer, and colon cancer, while in female patients they were breast cancer, thyroid cancer, and colon cancer. However, the incidence of gastric cancer is high in Japan [12] and the present study was conducted in Henan, which has the highest incidence rate of esophageal cancer in the world [13], which might explain the differences in males. In addition, esophageal cancer can easily be detected at an early stage due to the occurrence of swallowing obstruction and other symptoms, and thus can be treated early with the prospect of longer survival times. Prostate cancer has been found less frequently in our study. Possible causes include the lower prostate incidence in Asian compared to African and Caucasian males [14], and that in our study, the patients under observation were relatively young. The median age of the first primary cancer diagnosis was < 60 years, while the risk of prostate cancer is supposed to be high in patients above 65 years. In addition, a previous study noted, that in contrast to breast cancer multiple diagnosed cancers in prostate cancer patients are rare [15]. However, since the follow-up rate in our study was 90%, secondary primary cancer diagnoses in other medical centers or unrecognized cases cannot be entirely excluded.

For breast cancer in females, innovative diagnostic and treatment technologies have led to early diagnosis and better therapy, but the carcinogenic effects of radiotherapy might play a role in promoting cancer development, since previous studies indicated that radiotherapy for breast cancer could increase the risk of developing second primary cancers [16,17,18].

Our study showed that lung cancer staging was lower in patients of the LCF compared to the OCF group, since I–III staged patients accounted for 80.1% in the LCF and 58.5% in the OCF group. An explanation could be that the earlier detection in the LCF patients led to a better survival chance and time, which reflected in a higher incidence of secondary primary cancer, which also has been proposed in other similar studies with similar results [19, 20]. A recent paper reported that lung cancer staging affected the prognosis of MPMN in patients with lung cancer, which is in line with our findings that NSLC staging >II was an independent risk factor for survival [8]. We also found that the median diagnostic age of the primary cancer diagnosed first in the LCF group was significantly higher than that in OCF group, but there was no significant difference in the median diagnostic age of the primary cancer diagnosed second between patients in the two groups, which indicated that although lung cancer was diagnosed at an earlier stage than for other cancers, the survival rate was worse and the time until second primary malignancy was shorter than in the OCF group. When patients < 60 years old were diagnosed with the first primary cancer they had a better prognosis than patients ≥ 60 years old, though the age of the second primary cancer diagnosis had no effect on the survival times. Limitations of the study were its retrospective design and that no molecular information of the enrolled patients were included.

Conclusion

Of the 350 MPMN patients, 21.1% were SMPMN and 78.9% MMPMN cases with 26.6% LCF and 73.4% OCF malignancy occurrences. Detections of first primary neoplasms were at an earlier stage for LCF patients and the age of their first lung cancer diagnosed was 59.3 years vs. 55.4 years in the OCF group (P = 0.008). The factors that negatively influenced the prognosis of MPMN lung cancer patients were age (≥ 60 years old) at the first primary cancer diagnosis and whether lung cancer was diagnosed prior to the second primary malignant neoplasm. In addition, MMPMN cases had better prognoses than SMPM lung cancer patients and NSLC staging also had an effect on survival rates.