Clinical characteristics and prognosis of lung metastases from unknown primary cancer sites

Introduction

The lung ranks as the second most frequent site for tumor metastasis following the liver [1]. It is estimated that approximately 30–40 % of malignant tumors undergo lung metastasis [2]. Lung metastases typically signify an advanced disease stage, and if not promptly detected and treated, can precipitate rapid deterioration of the patient’s condition, potentially culminating in severe cases in rapid death. Most patients with lung metastasis can identify the primary tumor site. Common cancers that lead to lung metastasis include colorectal cancer, breast cancer, choriocarcinoma, and head and neck cancer [3], [4], [5], [6]. However, there is a subset of patients who present with tumor metastases confined solely to the lungs. Despite comprehensive investigations, including a detailed history, thorough physical examination, imaging, and laboratory tests, the primary tumor site remains elusive in these cases. This condition is termed “lung metastasis with unknown primary site” (CUPL).

Metastatic cancer of unknown primary focus (CUP) constitutes approximately 3–5 % of all cancer cases [7]. Despite its relatively low incidence, it ranks as the fourth leading cause of cancer-related deaths globally due to its elevated mortality rate [8]. Recent years have witnessed a decline in CUP incidence compared to previous periods, attributed to advancements in tumor detection technology, resulting in improved success rates of primary tumor localization [9]. Nevertheless, the overall prognosis for CUP remains bleak, with a reported 1-year survival rate of less than 20 %, and nearly half of the patients succumb within three months of diagnosis [10]. Previous reports indicate that CUP predominantly affects older individuals, with a median age of 60 years, and is exceedingly rare in patients under 40 [11, 12].

To date, there has been a lack of comprehensive large-scale studies on the clinical characteristics and prognostic factors of lung metastases with unknown primary sites. Moreover, our understanding of the disease traits among young CUPL patients remains limited. Although CUPL is not an uncommon occurrence in clinical practice, its existence presents a formidable challenge to diagnosis and treatment. Consequently, an in-depth examination of CUPL holds significant theoretical value and clinical importance. To address this gap, we conducted a retrospective study utilizing the Surveillance, Epidemiology, and End Results (SEER) database from the National Cancer Institute. Our study aimed to elucidate the clinical characteristics of CUPL patients, explore prognostic factors influencing CUPL, and compare prognostic traits across different age groups. Ultimately, this endeavor aims to provide clinicians with a deeper understanding of this disease, enabling the refinement of treatment strategies.

Methods

Patient selection

We identified patients with lung metastasis of unknown primary cancer sites from the SEER database (https://seer.cancer.gov/), spanning the period 2010 to 2020, utilizing SEER*Stat version 8.4.2. The inclusion criteria comprised cases falling under the classification of (1) “C80.9-Unknown primary site” and (2) SEER Combined Mets at DX-lung (2010+) indicating “YES”. Exclusion criteria were applied, excluding cases lacking pathologic confirmation of diagnosis (n=1,118) and those without survival time data (n=4). From the initial pool, consisting of 3,596 patients with lung metastases of unknown primary cancer sites, 1,122 cases with incomplete information were excluded. Ultimately, 2,474 cases met the criteria and were included in this study. The flowchart is delineated in Figure 1. Ethical approval was considered unnecessary since the SEER database is openly accessible to researchers worldwide.

Figure 1:

Flow chart of the study.

Results

Patient characteristics

The study comprised a total of 2,474 patients with lung metastases of unknown primary site, of whom 1,261 (51 %) were males and 1,213 (49 %) were females, demonstrating a slight male predominance. The patients exhibited a median survival time of 1 month, with a mean of 4.4 months. The median survival time for young patients was 3.5 months, with a mean of 12.8 months. For middle-aged patients, the median survival time was 1 month, averaging 4.7 months. Elderly patients had a median survival time of 1 month as well, but their mean survival time was slightly shorter at 3.6 months. Urban residency was more prevalent, with 85.3 % of patients residing in urban areas compared to 14.7 % in rural areas. Regarding income distribution, 42.1 % were classified as high-income, 56.7 % as middle-income, and only 1.2 % as low-income. The median age of patients was 70 years, and the distribution across age groups revealed 3 % as young patients (<40 years old), 50.2 % as middle-aged patients (40–70 years old), and 46.8 % as elderly patients (>70 years old) (Figure 2).

Figure 2:

Proportion of patients with lung metastases of unknown origin in different age groups.

In terms of race, 80.6 % of CUPL patients were White, with 2.5 % young patients, 49.3 % middle-aged patients, and 48.2 % elderly patients. Adenocarcinoma was the predominant pathology in CUPL patients, constituting 40.7 %, with 2.2 % in young patients, 49.7 % in middle-aged patients, and 48.1 % in elderly patients. Regarding other metastatic sites, 58.3 % of CUPL patients presented with concurrent liver metastases, 41.0 % with concomitant bone metastases, and only 13.3 % with concurrent brain metastases. This trend was generally consistent across young patients, middle-aged patients, and elderly patients. Concerning chemotherapy, only 27.2 % of patients with lung metastases of unknown primary site underwent chemotherapy, with 5.9 % of young patients, 63.1 % of middle-aged patients, and 31.0 % of elderly patients opting for chemotherapeutic interventions. As for radiotherapy, only 17.3 % of patients with lung metastases of unknown origin received radiotherapy, with 3.0 % of young patients, 56.4 % of middle-aged patients, and 40.6 % of elderly patients choosing this treatment modality. Refer to Table 1 for details.

Table 1:

Baseline of characteristics.

Characteristics	Total	<40 years	40–70 years	>70 years	p-Value
n (%)	2,474	72 (3.0)	1,243 (50.2)	1,159 (46.8)
Race, n (%)
Others	209 (8.4)	11 (15.3)	102 (8.2)	96 (8.3)	0.008
Black	256 (10.3)	11 (15.3)	149 (12.0)	96 (8.3)
Unknown	14 (0.6)	0 (0.0)	9 (0.7)	5 (0.4)
White	1,995 (80.6)	50 (69.4)	983 (79.1)	962 (83.0)
Sex, n (%)
Female	1,213 (49.0)	37 (51.4)	596 (47.9)	580 (50.0)	0.544
Male	1,261 (51.0)	35 (48.6)	647 (52.1)	579 (50.0)
Marital, n (%)
Unknown	106 (4.3)	3 (4.2)	51 (4.1)	52 (4.5)	0.001
Married	1,164 (47.0)	20 (27.8)	622 (50.0)	522 (45.0)
Others	1,204 (48.7)	49 (68.1)	570 (45.9)	585 (50.5)
Income, n (%)
High	1,042 (42.1)	31 (43.1)	497 (40.0)	514 (44.3)	0.256
Median	1,403 (56.7)	40 (55.6)	729 (58.6)	634 (54.7)
Low	29 (1.2)	1 (1.4)	17 (1.4)	11 (0.9)
Year of diagnosis, n (%)
2010–2015	762 (30.8)	27 (37.5)	391 (31.5)	344 (29.7)	0.294
2016–2020	1,712 (69.2)	45 (62.5)	852 (68.5)	815 (70.3)
Rural/urban, n (%)
Urban	2,110 (85.3)	63 (87.5)	1,049 (84.4)	998 (86.1)	0.428
Rural	364 (14.7)	9 (12.5)	194 (15.6)	161 (13.9)
Histology, n (%)
Others	646 (26.1)	33 (45.8)	326 (26.2)	287 (24.8)	0.002
Adenocarcinoma	1,008 (40.7)	22 (30.6)	501 (40.3)	485 (41.8)
Carcinoma, NOS	492 (19.9)	7 (9.7)	245 (19.7)	240 (20.7)
Neuroendocrine carcinoma	165 (6.7)	8 (11.1)	91 (7.3)	66 (5.7)
Squamous cell carcinoma	163 (6.6)	2 (2.8)	80 (6.4)	81 (7.0)
Grade, n (%)
Unknown	2,291 (92.6)	64 (88.9)	1,146 (92.2)	1,081 (93.3)	0.228
Moderately	36 (1.5)	1 (1.4)	16 (1.3)	19 (1.6)
Poorly	119 (4.8)	4 (5.6)	64 (5.1)	51 (4.4)
Undifferentiated	16 (0.6)	2 (2.8)	9 (0.7)	5 (0.4)
Well	12 (0.5)	1 (1.4)	8 (0.6)	3 (0.3)
Liver metastases, n (%)
No	875 (35.4)	18 (25.0)	404 (32.5)	453 (39.1)	<0.001
Unknown	157 (6.3)	5 (6.9)	67 (5.4)	85 (7.3)
Yes	1,442 (58.3)	49 (68.1)	772 (62.1)	621 (53.6)
Bone metastases, n (%)
No	1,227 (49.6)	34 (47.2)	584 (47.0)	609 (52.5)	<0.001
Unknown	233 (9.4)	9 (12.5)	98 (7.9)	126 (10.9)
Yes	1,014 (41.0)	29 (40.3)	561 (45.1)	424 (36.6)
Brain metastases, n (%)
No	1,844 (74.5)	46 (63.9)	908 (73.0)	890 (76.8)	<0.001
Unknown	302 (12.2)	16 (22.2)	134 (10.8)	152 (13.1)
Yes	328 (13.3)	10 (13.9)	201 (16.2)	117 (10.1)
Chemotherapy, n (%)
No/unknown	1,800 (72.8)	32 (44.4)	818 (65.8)	950 (82.0)	<0.001
Yes	674 (27.2)	40 (55.6)	425 (34.2)	209 (18.0)
Radiotherapy, n (%)
No/unknown	2,045 (82.7)	59 (81.9)	1,001 (80.5)	985 (85.0)	0.015
Yes	429 (17.3)	13 (18.1)	242 (19.5)	174 (15.0)

Survival and prognosis analysis

The univariate Cox regression analysis indicated that several factors, including age, marital status, rural/urban residency, histology, grade, presence of liver, bone, and brain metastases, as well as receipt of chemotherapy and radiotherapy, were prognostic factors impacting the overall survival of patients with lung metastasis of an unknown primary site. Multifactorial Cox regression results revealed that age ≤70, being married, having well-differentiated tumors, and undergoing chemotherapy and radiotherapy were associated with improved overall survival (Table 2). In terms of cancer-specific survival, the univariate competitive analysis identified age, year of diagnosis, histology, grade, liver metastasis, bone metastasis, brain metastasis, chemotherapy, and radiotherapy as influential prognostic factors. Multifactorial competitive analysis indicated that an age range of 40–70 years, chemotherapy, and radiotherapy were associated with improved CSS (Table 3).

Table 2:

Univariable and multivariable Cox regression of factors associated with overall survival in patients with lung metastases of unknown origin.

Characteristics	Univariate		Multivariate
	HR (95 % CI)	p-Value	HR (95 % CI)	p-Value
Age, years
>70	Reference		Reference
40–70	0.79 (0.73–0.86)	<0.001	0.90 (0.83–0.98)	0.018
<40	0.50 (0.39–0.66)	<0.001	0.71 (0.54–0.93)	0.014
Race
Others	Reference
Black	0.91 (0.75–1.10)	0.343
Unknown	1.11 (0.60–2.04)	0.737
White	1.01 (0.87–1.18)	0.877
Sex
Female	Reference
Male	1.01 (0.93–1.10)	0.784
Marital, n (%)
Unknown	Reference		Reference
Married	0.66 (0.54–0.82)	<0.001	0.77 (0.63–0.95)	0.016
Others	0.76 (0.62–0.94)	0.011	0.82 (0.67–1.02)	0.069
Income
High	Reference
Median	1.06 (0.97–1.15)	0.173
Low	0.87 (0.58–1.29)	0.492
Year of diagnosis
2010–2015	Reference
2016–2020	0.93 (0.85–1.02)	0.109
Rural/urban
Urban	Reference		Reference
Rural	1.12 (1.00–1.26)	0.049	1.15 (1.02–1.29)	0.023
Histology
Others	Reference		Reference
Adenocarcinoma	1.36 (1.23–1.52)	<0.001	1.40 (1.26–1.56)	<0.001
Carcinoma, NOS	1.45 (1.28–1.65)	<0.001	1.48 (1.31–1.68)	<0.001
Neuroendocrine carcinoma	0.98 (0.82–1.17)	0.819	1.05 (0.88–1.27)	0.574
Squamous cell carcinoma	1.06 (0.89–1.28)	0.508	1.22 (1.00–1.46)	0.050
Grade
Unknown	Reference		Reference
Moderately	1.02 (0.73–1.44)	0.897	0.96 (0.68–1.35)	0.802
Poorly	1.23 (1.02–1.48)	0.031	1.32 (1.09–1.59)	0.004
Undifferentiated	1.59 (0.96–2.64)	0.074	1.66 (0.99–2.77)	0.054
Well	0.53 (0.29–0.99)	0.046	0.45 (0.24–0.85)	0.014
Liver metastases
No	Reference		Reference
Unknown	1.39 (1.17–1.66)	<0.001	1.11 (0.89–1.38)	0.371
Yes	1.29 (1.18–1.41)	<0.001	1.33 (1.21–1.46)	<0.001
Bone metastases
No	Reference		Reference
Unknown	1.16 (1.00–1.34)	0.044	0.95 (0.77–1.17)	0.620
Yes	1.04 (0.95–1.13)	0.389	1.10 (1.00–1.21)	0.034
Brain metastases
No	Reference		Reference
Unknown	1.27 (1.12–1.44)	<0.001	1.15 (0.95–1.39)	0.157
Yes	1.15 (1.02–1.30)	0.026	1.32 (1.15–1.50)	<0.001
Chemotherapy
No/unknown	Reference		Reference
Yes	0.41 (0.37–0.45)	<0.001	0.41 (0.37–0.46)	<0.001
Radiotherapy
No/unknown	Reference		Reference
Yes	0.65 (0.58–0.72)	<0.001	0.66 (0.59–0.75)	<0.001

Table 3:

Univariable and multivariable competing risk regression of factors associated with cancer-specific survival in patients with lung metastases of unknown origin.

Characteristics	Univariate		Multivariate
	HR (95 % CI)	p-Value	HR (95 % CI)	p-Value
Age, years
>70	Reference		Reference
40–70	0.86 (0.80–0.93)	<0.001	0.91 (0.84–0.99)	0.032
<40	0.67 (0.53–0.85)	<0.001	0.78 (0.61–1.01)	0.055
Race
Others	Reference
Black	0.86 (0.72–1.02)	0.083
Unknown	0.85 (0.43–1.70)	0.650
White	0.98 (0.86–1.12)	0.790
Sex
Female	Reference
Male	1.01 (0.94–1.09)	0.770
Marital, n (%)
Unknown	Reference
Married	0.94 (0.76–1.17)	0.600
Others	0.98 (0.79–1.22)	0.880
Income
High	Reference
Median	1.06 (0.98–1.15)	0.120
Low	0.86 (0.58–1.29)	0.470
Year of diagnosis
2010–2015	Reference		Reference
2016–2020	0.92 (0.85–1.00)	0.037	0.93 (0.85–1.01)	0.100
Rural/urban
Urban	Reference
Rural	1.10 (0.99–1.22)	0.077
Histology
Others	Reference		Reference
Adenocarcinoma	1.26 (1.14–1.39)	<0.001	1.25 (1.13–1.39)	<0.001
Carcinoma, NOS	1.29 (1.15–1.46)	<0.001	1.27 (1.12–1.45)	<0.001
Neuroendocrine carcinoma	0.98 (0.82–1.15)	0.770	1.00 (0.84–1.18)	0.960
Squamous cell carcinoma	1.05 (0.88–1.25)	0.570	1.11 (0.92–1.33)	0.270
Grade
Unknown	Reference		Reference
Moderately	1.19 (0.91–1.54)	0.210	1.12 (0.86–1.46)	0.410
Poorly	1.23 (1.05–1.44)	0.009	1.23 (1.04–1.44)	0.014
Undifferentiated	1.44 (0.86–2.41)	0.170	1.51 (0.86–2.65)	0.160
Well	0.77 (0.48–1.26)	0.300	0.75 (0.47–1.20)	0.220
Liver metastases
No	Reference		Reference
Unknown	1.25 (1.05–1.48)	0.011	1.03 (0.83–1.27)	0.800
Yes	1.22 (1.12–1.32)	<0.001	1.21 (1.11–1.32)	<0.001
Bone metastases
No	Reference		Reference
Unknown	1.18 (1.02–1.35)	0.023	1.02 (0.83–1.23)	0.880
Yes	1.11 (1.03–1.21)	0.008	1.16 (1.07–1.27)	<0.001
Brain metastases
No	Reference		Reference
Unknown	1.23 (1.09–1.38)	<0.001	1.16 (0.99–1.37)	0.074
Yes	1.06 (0.95–1.19)	0.280	1.10 (0.97–1.25)	0.150
Chemotherapy
No/unknown	Reference		Reference
Yes	0.62 (0.57–0.67)	<0.001	0.63 (0.58–0.69)	<0.001
Radiotherapy
No/unknown	Reference		Reference
Yes	0.80 (0.73–0.87)	<0.001	0.83 (0.75–0.92)	<0.001

Patients with lung metastases of unknown primary site exhibited a 1-year overall survival rate of 11 % and a 1-year cancer-specific survival rate of 13.8 %. Survival disparities were observed in overall survival and cancer-specific survival among patients in different age groups. Despite constituting only 3 % of the cohort, young patients demonstrated significantly better overall and cancer-specific survival than middle-aged and elderly patients (p<0.0001 and p<0.001) (Figure 3). Subgroup analyses revealed that middle-aged and elderly patients who received both chemotherapy and radiotherapy experienced enhanced overall survival and cancer-specific survival compared to those who did not undergo these treatments. Conversely, among young patients, chemotherapy alone yielded benefits, while the addition of radiotherapy did not confer any additional advantage (refer to Figure 4). Further scrutiny unveiled that, among treated patients, young individuals demonstrated superior overall survival compared to their middle-aged and elderly counterparts, albeit without a significant difference in cancer-specific survival (see Figure 5).

Figure 3:

Kaplan–Meier and Cumulative Incidences Function comparison of survival differences in patients in different age groups. (A) Overall survival was significantly better in the group aged under 40 years compared to the groups aged 40–70 years and over 70 years; (B) patients under 40 years of age had significantly fewer cancer-specific mortality compared to those aged 40–70 years and those over 70 years.

Figure 4:

Comparison of overall survival (A–F) and cancer-specific mortality (G–L) with and without chemotherapy and radiation therapy in patients of different age groups.

Figure 5:

Comparison of survival differences between patients receiving chemotherapy and radiation in different age groups. Overall survival was better in patients under 40 years of age receiving chemotherapy (A) or radiotherapy (C) compared to those aged 40–70 years and those over 70 years; cancer-specific deaths in patients under 40 years of age receiving chemotherapy (B) or radiotherapy (D) did not differ from those in the 40–70 years and over 70 years groups.

Discussion

Metastatic cancer of an unknown primary site represents a distinct form of malignancy, characterized by metastasis in the absence of an identifiable primary tumor site, despite comprehensive clinical and pathological evaluations [13]. The international community has not reached a consensus on the pathogenesis of CUP, with experts attributing it to various factors, such as the body’s immune system targeting and suppressing the primary foci, subtle primary foci eluding detection by current examination instruments, the specific location of primary foci impeding detection, and the delayed manifestation or disappearance of primary foci during treatment [14, 15]. CUP poses significant clinical challenges due to the rapid progression of the disease, often resulting in patient demise before the primary foci can be identified.

Research on lung metastases from an unknown primary site is limited, and to our knowledge, this represents the first extensive retrospective study on CUPL. Our results indicate a bleak prognosis for CUPL patients, with an average survival time of 4.4 months, a one-year overall survival rate of just 11 %, and a one-year cancer-specific survival rate of 13.8 %. Tumors metastasizing to the lungs typically incur a diminished five-year survival rate; however, the lack of information on the primary tumor in CUPL hinders the development of effective treatment strategies, exacerbating survival prognosis challenges. Therefore, gaining a deeper understanding of the prognostic risk factors associated with CUPL could potentially improve survival rates and inform clinicians’ treatment decisions.

Age was identified as an independent prognostic factor for CUPL patients, aligning with the common knowledge that tumors tend to manifest in the elderly due to multifaceted genetic, environmental, and cumulative factors [16]. Our study revealed a rise in CUPL incidence with advancing age, noting a median age of 70 years among patients. Despite young patients constituting only 3 % of the cohort, they exhibited significantly better overall and cancer-specific survival than middle-aged and elderly patients, especially among those receiving radiotherapy and chemotherapy. The poorer prognosis observed in elderly patients may be attributed to declining physical functions, reducing their ability to respond to cancer treatments, as well as the presence of comorbid chronic diseases that complicate the treatment regimen [17].

Histological typing and tumor grading proved to be other essential independent prognostic factors for CUPL patients. Adenocarcinoma was the most prevalent pathological type, consistent with previous reports [18], and both adenocarcinoma and carcinoma pathologies were linked to an elevated risk of mortality in CUPL patients (p<0.001). Additionally, while a hyperdifferentiated tumor grade positively influenced overall survival, poorly differentiated tumors significantly increased the risk of death in CUPL patients (p<0.05). Distinct histopathological features of CUPL are associated with varying survival prognoses for patients. This correlation can be attributed to the unique biological traits inherent to each tumor type, as well as their differential responses to treatment regimens [19]. Specifically, well-differentiated adenocarcinomas often demonstrate a more favorable response to therapy, whereas poorly differentiated or undifferentiated malignancies tend to behave more aggressively, leading to a less optimistic prognosis.

No standard treatment regimen exists for CUP, with chemotherapy being the primary modality, supplemented by surgery or radiotherapy tailored to the specific characteristics of the cancer [20], [21], [22]. In recent years, immunotherapy, targeted therapy, and site-specific therapy have gained increasing application in the treatment of patients with cancer of unknown primary. However, there is controversy surrounding these therapeutic approaches, and the current consensus among scholars favors a combination therapy approach [23], [24], [25]. For CUPL patients, the absence of a primary tumor site renders traditional staging and treatment approaches less applicable, complicating treatment decisions. Our study indicated that surgical interventions were infrequent, possibly due to patients being in the advanced stages of the disease, where surgery might hasten their demise [26]. Only a minority of CUPL patients received chemotherapy (27.2 %) or radiotherapy (17.3 %). Fortunately, the survival prognosis of patients across various age groups who received treatment was superior to those who did not undergo radiotherapy and chemotherapy. It is noteworthy that, particularly within the young cohort, radiotherapy did not notably impact overall survival or cancer-specific survival. This observation may be attributed to the highly heterogeneous nature of cancer of unknown primary; nevertheless, given the limited number of young CUPL patients in this study, further investigation is warranted. Despite the restricted availability of targeted therapeutic options for CUPL, our findings advocate for a dual approach: firstly, patients should strive to identify the primary tumor site through comprehensive imaging, aiming for precise localization of the primary foci. Secondly, treatment strategies should be tailored to individual patient circumstances. For patients in good physical condition, more aggressive treatments, such as chemotherapy-based systemic therapy, may be appropriate. Conversely, for those in poorer physical health, priority should be accorded to supportive therapy and symptom management.

Our study has limitations, primarily stemming from its retrospective nature and the partial lack of data in the SEER database, potentially introducing selection bias. Additionally, our study lacks detailed treatment information crucial for evaluating prognosis, such as specific radiotherapy and chemotherapy regimens or the use of immunotherapy and targeted therapy, due to the unavailability of the SEER database. Despite the findings of our study, numerous questions concerning CUPL remain unanswered, underscoring the need for further investigation. Future research endeavors should prioritize the exploration of novel biomarkers aimed at enhancing the precision of primary lesion localization. Additionally, the development of innovative therapeutic approaches and medications represents a crucial avenue for future studies.

In conclusion, our study emphasizes the difficult prognosis faced by patients with lung metastases of unknown primary sites. This condition predominantly affects the middle-aged and the elderly, often manifesting as metastatic adenocarcinoma. Factors such as age, tumor histological typing and grading, as well as the presence of liver and bone metastases, along with the administration of radiotherapy and chemotherapy, collectively influence patients’ overall and cancer-specific survival. Importantly, our results indicate that radiotherapy and chemotherapy enhance survival outcomes for CUPL patients.