Introduction
A cancer of unknown primary origin (CUP) is a
metastatic tumor for which a primary site cannot be identified
despite examination using appropriate techniques, such as imaging
with CT or MRI, endoscopic evaluations and investigating various
tumor markers (1). CUP is a rare
cancer, accounting for 2-5% of all types of cancer globally and
patients with CUP have a poor prognosis, with a 1-year survival
rate of ~20% (2-4).
Furthermore, ~15% of patients with CUP respond to specific
therapies and achieve long-term survival, with a prognostic
analysis of 1,000 cases demonstrating a median survival of 11
months (2). Favorable histological
entities of CUP represent a distinct minority of patients and
include well-defined clinicopathological subsets such as isolated
unilateral axillary lymph node metastases with breast cancer-like
features, squamous-cell carcinoma involving cervical lymph nodes,
poorly differentiated carcinoma with midline distribution, women
with papillary serous adenocarcinoma of the peritoneum and men with
osteoblastic bone metastases and elevated serum prostate-specific
antigen (1). These favorable
subsets have outcomes comparable with those of corresponding known
primary tumors and are associated with improved prognoses. By
contrast, unfavorable subsets of CUP are defined as patients with
disseminated visceral disease such as adenocarcinoma metastatic to
the liver, non-papillary malignant ascites, multiple lung or
pleural metastases, multiple brain metastases, or lytic bone
disease (1,2). Previous randomized global trials in
patients with unfavorable subsets of CUP report a median overall
survival (OS) of ~12 months from the initiation of first-line
treatment, with the longest median OS being 13.6 months (5,6).
CUP is typically diagnosed at an advanced metastatic
stage where the primary site remains undetermined despite extensive
evaluation (7). Consequently,
surgical intervention is generally limited to diagnostic or
palliative purposes. Furthermore, curative surgery for CUP is rare,
and systemic therapy, including chemotherapy, is the cornerstone of
treatment (7).
The treatment approach for CUP involves identifying
the most probable primary site based on its unique clinical
presentation, histopathological findings and genetic profile
(4). The European Society for
Medical Oncology recommends that therapy is selected based on the
provisional primary site identified through the gene expression
profiling of collected cancer tissue (site-specific regimen)
(4). For unfavorable subsets with
no histological organ-specific findings, a platinum-doublet regimen
(empirical regimen) (4,8) is recommended; carboplatin-paclitaxel
is often used; however, there is not an established advantage
compared with other regimens. Findings from recent studies suggest
that CUP shares similar immunological characteristics with solid
tumors that are responsive to immune checkpoint inhibitors,
including higher levels of tumor-infiltrating lymphocytes and an
increased expression of PD-L1 (9-11).
Therefore, administering immune checkpoint inhibitors, such as
nivolumab, to patients with CUP yields a certain degree of
efficacy, which makes it a possible treatment option (11,12).
Previous studies report that the poor performance
status (PS), high levels of lactate dehydrogenase, high number of
organs with metastasis and the presence of liver metastases are
predictive factors for a poor prognosis and reduced probability of
survival in patients with CUP (13,14).
However, useful prognostic biomarkers are yet to be elucidated.
Furthermore, CUP is a rare cancer with a high level of
heterogeneity and a poor prognosis.
In recent years, the nutritional status of patients
with cancer has been recognized as a factor influencing treatment
response and survival duration (15). The nutritional status of patients
with advanced cancer is a useful prognostic factor. The most common
nutritional marker is the albumin level. Other nutritional markers
include the neutrophil-to-lymphocyte ratio (NLR), prognostic
nutritional index (PNI), modified Glasgow prognostic score (GPS;
mGPS) and controlling nutritional status (CONUT) score (12). Previous studies demonstrate the
usefulness of the aforementioned nutritional markers in estimating
the prognosis, especially in the perioperative period and during
chemotherapy, of patients with gastrointestinal types of cancer
(16,17). By contrast, the association between
nutritional status and survival outcomes in patients with CUP is
yet to be fully elucidated. Therefore, the present study aimed to
investigate the effect of nutritional status on the prognosis of
patients with CUP who were treated with chemotherapy.
Materials and methods
Study design and patient
selection
In the present study, the data of 25 patients with
CUP who received chemotherapy at Kurume University Hospital
(Kurume, Japan) between July 2011 and December 2023 was
retrospectively evaluated. All patients were diagnosed with CUP
because of the difficulty in identifying the primary tumor site
after close examination using imaging and histological techniques,
including CT and MRI scans, endoscopic evaluations, tumor marker
assessments, and tissue biopsies of metastatic lesions. Tissue
samples collected from the patients were subjected to histological
evaluation and immunohistochemical analysis to predict the primary
site. Clinical data were collected from the medical records,
including age, sex, PS, body mass index (BMI), type of pathology
(adenocarcinoma, squamous cell carcinoma or undifferentiated
carcinoma), number of organs with metastasis, laboratory results
(blood count and chemistry profiles), NLR, PNI, mGPS and CONUT
scores.
The present study was conducted in accordance with
the principles of the Declaration of Helsinki and approved by the
Ethics Committee of Kurume University (Kurume, Japan; approval no.
23229). An opt-out approach was used to obtain informed consent
from the patients, and personal information was protected during
data collection.
Calculations of the nutritional scores
and definitions of the cut-off values
Various nutritional scores were calculated before
chemotherapy and their relevance to prognosis was assessed. NLR was
calculated by dividing the total neutrophil count by the total
lymphocyte count. PNI was calculated as [10 x serum albumin (g/dl)]
+ [0.005 x total lymphocyte count (ml)]. mGPS was calculated based
on the levels of serum C-reactive protein and albumin [C-reactive
protein, <0.5 mg/dl and albumin, ≥3.5 g/dl (0 points);
C-reactive protein, <0.5 mg/dl or albumin, ≥3.5 g/dl (1 point);
or C-reactive protein, ≥0.5 mg/dl and albumin, <3.5 g/dl (2
points)]. The CONUT score was calculated based on the serum albumin
content [≥3.5 g/dl (0 points); 3.0-3.4 g/dl (2 points); 2.5-2.9
g/dl (4 points); or <2.5 g/dl (6 points)], total lymphocyte
count [≥1,600 cells/µl (0 points); 1,200-1,599 cells/µl (1 point);
800-1,199 cells/µl (2 points); or <800 cells/µl (3 points)] and
serum levels of total cholesterol [≥180 mg/dl (0 points); 140-179
mg/dl (1 point); 100-139 mg/dl (2 points); or <100 mg/dl (3
points)]. The nutritional scores defined in the present study were
based on the cut-off values used in previous studies (low NLR,
<3; high PNI, >40; low mGPS, ≤1 point; and low CONUT score,
≤3 points) (18-20).
Regimen selection
All patients underwent a biopsy of the tumor tissue
for histological and immunological diagnoses. To estimate the
location of the primary tumor immunohistochemistry was used. Tissue
preparation and staining were carried out as follows:
Paraffin-embedded tissue samples (samples treated with 10% neutral
buffered formalin at room temperature for 6-72 h) were cut 5 µm
thick and examined on a coated glass slide (cat. no. SCRE-13;
Matsunami Glass Ind., Ltd.). Each slide was heat-treated at 99˚C
using Ventana CC1 retrieval solution (Roche Tissue Diagnostics) for
64 min. Subsequently, samples were labeled with primary antibodies
using the BenchMark ULTRA (Roche Tissue Diagnostics) and Bond-III
autostainer (Leica Microsystems, Ltd.). This automated system used
a biotin-free polymer detection method. Specifically, positive
immunoreactive signals were visualized using the UltraView
Universal DAB Detection kit (cat. no. 760-500; Roche Diagnostics
KK), which used the UltraView Universal HRP Multimer containing a
cocktail of HRP-labeled secondary antibodies. Blocking reagents
were not used with the antibodies in the present study. Although
exogenous blocking reagents were not applied separately, the
automated staining system used in the present study used optimized
buffer formulations and detection kits designed to minimize
non-specific background. In the present study, no notable
non-specific binding or background interference was observed that
would confound the interpretation of the biopsies. The primary
antibodies used were: Cytokeratin (CK)7 (undiluted; cat. no.
05986818001; Roche Diagnostics, Ltd.) and CK20 (undiluted; cat. no.
05587760001; Roche Diagnostics, Ltd.), at room temperature for 16
min; synaptophysin (undiluted; cat. no. 413831; Nichirei
Biosciences, Inc.) and chromogranin (undiluted; cat. no. M0869;
Dako; Agilent Technologies, Inc.) at room temperature for 16 min,
with heat-induced antigen retrieval at 95˚C with epitope retrieval
solution 2 (pH 9.0) for 20 min and a Refine polymer detection
system (Leica Microsystems, Ltd.); and S100 protein (undiluted;
cat. no. IR504; Dako; Agilent Technologies, Inc.) and HMB-45
(1:100; cat. no. M0634; Dako; Agilent Technologies, Inc.) at room
temperature for 16 min using the Refine polymer detection system
without the heat-induced antigen retrieval and epitope retrieval
solution. Positive immunoreactive signals [UltraView Universal HRP
Multimer containing a cocktail of HRP-labeled antibodies (goat
anti-mouse IgG, goat anti-mouse IgM and goat anti-rabbit;
undiluted) in a buffer containing protein with ProClin 300, a
preservative, at 36˚C for 8 min] and chromogen detection [UltraView
Universal DAB H2O2 was used containing 0.04%
hydrogen peroxide in phosphate buffer solution, then UltraView
Universal DAB Copper was used in an acetic acid buffer solution
containing copper sulfate (5 g/l) with a proprietary preservative
formulation] were visualized using the UltraView Universal DAB
Detection kit (cat. no. 760-500; Roche Diagnostics KK). All images
of the stained sections were obtained with an all-in-one
fluorescence microscope (model, BZ-X700; Keyence Corporation).
If the location of a primary tumor site could be
estimated based on only the histological findings (without using
next-generation sequencing), a site-estimated treatment regimen was
used (Table SI). Patients were
divided into groups according to whether the location of the
primary tumor could be estimated or not; this was carried out based
on the differences in the treatment regimens of the patients. The
group with an estimated location of the primary tumor (but without
a tumor found at the estimated location) underwent site-specific
treatment. If no clear histological features were present, the
patients were diagnosed with undifferentiated carcinoma, and
carboplatin-paclitaxel was used as the empirical regimen (21). The patients underwent treatment in
21-day cycles, with up to six cycles administered. The dose of
carboplatin was 6 mg/ml per min on day 1 based on an area under the
concentration-time curve, while paclitaxel was administered at 200
mg/m2 on day 1, with dosage adjustments made at the
discretion of the attending physician.
Assessment of outcomes
Response to treatment was evaluated at 8-week
intervals using dynamic computed tomography according to the
Response Evaluation Criteria in Solid Tumors version 1.1(22). Progression-free survival (PFS) was
defined as the duration between the enrollment date and the
occurrence of progressive disease or mortality due to any cause. OS
was defined as the duration between the date of enrollment and
mortality due to any cause.
Statistical analysis
Univariate analyses were conducted using the Cox
regression model to identify risk factors associated with OS or
PFS. Multivariate analyses were used to assess the impact of
independent variables on survival. For variable selection, all
factors were evaluated using P<0.10 in univariate analyses. OS
and PFS were calculated using the Kaplan-Meier method and assessed
using the log-rank test. Data analysis was carried out using JMP
Pro version 17.0 (SAS Institute, Inc.). P<0.05 was considered to
indicate a statistically significant difference.
Results
Patient characteristics
Table I presents
the characteristics of the patients with CUP that were included in
the present study. In total, 25 patients were diagnosed with CUP
and treated with chemotherapy at Kurume University Hospital
(Kurume, Japan) between July 2011 and December 2023. Overall, 9
patients (36%) were aged <65 years, and 16 patients (64%) were
women. The PS score was 0 or 1 in 20 patients (80%). The
pathological types of tumor diagnosed in the present cohort were
adenocarcinoma [12 patients (48%)], squamous cell carcinoma [6
patients (24%)] and undifferentiated carcinoma [7 patients (28%)].
Metastasis to <3 organs was identified in 16 patients (64%). In
total, 14 patients (56%) were treated with organ-specific regimens.
Regarding the nutritional scores, 11 patients (44%) had low NLR, 12
patients (48%) had high PNI, 16 patients (64%) had low mGPS and 14
patients (56%) had low CONUT scores.
 | Table IPatient characteristics. |
Table I
Patient characteristics.
| Characteristic | Number of patients, N
(%) |
|---|
| Age, years | |
|
<65 | 9(36) |
|
≥65 | 16(64) |
| Sex | |
|
Male | 9(36) |
|
Female | 16(64) |
| Performance status
score | |
|
0 or 1 | 20(80) |
|
≥2 | 5(20) |
| BMI,
kg/m2 | |
|
<22 | 11(44) |
|
≥22 | 14(56) |
| Type of
pathology | |
|
Adenocarcinoma | 12(48) |
|
Squamous
cell carcinoma | 6(24) |
|
Undifferentiated
carcinoma | 7(28) |
| Number of organs
with metastasis | |
|
<3 | 16(64) |
|
≥3 | 9(36) |
| Regimen | |
|
Site-estimated | 14(56) |
|
Empirical | 11(44) |
| NLR | |
|
Low
(<3) | 11(44) |
|
High
(≥3) | 14(56) |
| PNI | |
|
Low
(≤40) | 13(52) |
|
High
(>40) | 12(48) |
| mGPS | |
|
Low (0 or
1) | 16(64) |
|
High
(2) | 9(36) |
| CONUT score | |
|
Low
(≤3) | 14(56) |
|
High
(>3) | 11(44) |
OS and PFS in patients with CUP
To evaluate the prognosis of patients with CUP in
the present study, the OS and PFS was assessed in all cases. The
Kaplan-Meier survival curves for OS and PFS in patients receiving
chemotherapy for CUP are shown in Fig.
1. The median OS was 17.2 months [95% confidence interval (CI),
12.5-26.9 months) (Fig. 1A), and
the median PFS was 6.3 months (95% CI, 2.9-8.2 months) (Fig. 1B).
Prognostic factors for OS and PFS in
patients with CUP
Univariate and multivariate analyses were carried
out to evaluate factors affecting the prognosis of patients with
CUP. The results of the univariate analyses of the prognostic
factors for OS and PFS are presented in Tables II and III, respectively. The univariate
analysis demonstrated that the site-estimated regimen was
significantly associated with poor OS, while metastasis in three or
fewer organs, a low mGPS and a low CONUT score were significantly
associated with good prognoses. There were no significant
differences in OS based on age, sex, PS, BMI, type of pathology,
albumin levels, low NLR or high PNI scores.
| Table IIUnivariate and multivariate analyses
of prognostic factors for overall survival. |
Table II
Univariate and multivariate analyses
of prognostic factors for overall survival.
| | Univariate
analysis | Multivariate
analysis |
|---|
| Characteristic | HR | 95% CI | P-value | HR | 95% CI | P-value |
|---|
| Age, ≤65 years | 1.13 | 0.39-3.28 | 0.823 | | | |
| Sex, male | 2.17 | 0.75-6.34 | 0.155 | | | |
| Performance status
score, 0-1 | 0.67 | 0.18-2.41 | 0.535 | | | |
| BMI, >22
kg/m2 | 1.25 | 0.43-3.65 | 0.693 | | | |
| Type of pathology,
adenocarcinoma | 0.75 | 0.26-2.18 | 0.601 | | | |
| Number of organs
with metastasis, ≤3 | 0.20 | 0.06-0.66 | 0.008 | 0.28 | 0.06-1.43 | 0.127 |
| Regimen,
site-estimated | 3.91 | 1.19-12.84 | 0.024 | 4.06 | 0.77-21.49 | 0.100 |
| Albumin level,
>3.5 g/dl | 0.52 | 0.17-1.58 | 0.247 | | | |
| Low NLR, ≤3 | 0.52 | 0.17-1.58 | 0.245 | | | |
| High PNI,
>40 | 0.61 | 0.20-1.82 | 0.370 | | | |
| Low mGPS, 0 or
1 | 0.27 | 0.09-0.80 | 0.017 | 0.21 | 0.05-0.94 | 0.041 |
| Low CONUT score,
≤3 | 0.23 | 0.07-0.75 | 0.015 | 0.20 | 0.05-0.90 | 0.035 |
| Table IIIUnivariate and multivariate analyses
of prognostic factors for progression-free survival. |
Table III
Univariate and multivariate analyses
of prognostic factors for progression-free survival.
| | Univariate
analysis | Multivariate
analysis |
|---|
| Characteristic | HR | 95% CI | P-value | HR | 95% CI | P-value |
|---|
| Age, ≤65 years | 0.90 | 0.37-2.14 | 0.803 | | | |
| Sex, male | 1.85 | 0.78-4.41 | 0.165 | | | |
| Performance status
score, 0-1 | 1.04 | 0.34-3.14 | 0.951 | | | |
| BMI, >22
kg/m2 | 1.21 | 0.49-2.99 | 0.673 | | | |
| Type of pathology,
adenocarcinoma | 0.67 | 0.29-1.59 | 0.365 | | | |
| Number of organs
with metastasis, ≤2 | 0.50 | 0.21-1.20 | 0.119 | | | |
| Regimen,
site-estimated | 1.02 | 0.44-2.38 | 0.963 | | | |
| Albumin level,
>3.5 g/dl | 0.55 | 0.23-1.29 | 0.168 | | | |
| Low NLR, ≤3 | 0.48 | 0.20-1.14 | 0.095 | 1.65 | 0.41-6.72 | 0.483 |
| High PNI,
>40 | 0.53 | 0.23-1.25 | 0.146 | | | |
| Low mGPS, 0 or
1 | 0.21 | 0.08-0.61 | 0.004 | 0.20 | 0.06-0.71 | 0.012 |
| Low CONUT score,
≤3 | 0.45 | 0.19-1.06 | 0.067 | 0.44 | 0.13-1.53 | 0.197 |
Univariate analysis of the prognostic factors for
PFS demonstrated that low mGPS and low NLR were significantly
associated with a good prognosis. There were no significant
differences in PFS based on age, sex, PS, BMI, pathology type,
number of organs with metastases, use of a site-specific regimen,
albumin level, high PNI or low CONUT score.
Furthermore, patients were divided into two groups
based on whether the location of the primary tumor could be
estimated. The univariate analysis results for the OS and PFS are
presented in Tables SII and
SIII, respectively. In the group
with an estimated location of the primary tumor (but without a
tumor found at the presumed primary site), the PS, a high PNI, low
mGPS and a low CONUT score were associated with a significant
increase in the OS. There were no significant differences in the OS
in the group in which the location of the primary tumor could not
be estimated.
In the group with an estimated location of the
primary tumor, the PS, albumin level, a high PNI and a low mGPS
were associated with a significant increase in the PFS. There were
no significant differences in the PFS across all factors in the
group without an estimated location of the primary tumor.
The results of the multivariate analyses of the
prognostic factors for OS and PFS are presented in Tables II and III, respectively. Low mGPS and CONUT
scores were independent good prognostic factors for OS. No
significant differences in OS were observed for the number of
organs with metastasis or administration of site-specific regimens.
A low mGPS was an independent good prognostic factor for PFS. No
significant differences in PFS were observed for the low NLR nor
the low CONUT score. These results indicated that low mGPS and low
CONUT scores may be associated with prolonged survival.
Clinical outcomes based on the
nutrition score
Univariate and multivariate analyses suggested that
low mGPS and low CONUT nutritional scores may contribute to an
improved prognosis. Therefore, these factors were further
investigated using survival curves. The Kaplan-Meier survival
curves for OS and PFS stratified by mGPS and CONUT scores are shown
in Fig. 2. The median OS was
significantly longer in the low-mGPS group [25.7 months (95% CI,
12.5-80.5 months)] compared with the high-mGPS group [7.2 months
(95% CI, 0.4-17.2 months)] (Fig.
2A). In addition, the median PFS was significantly longer in
the low-mGPS group [8.2 months (95% CI, 4.0-11.3 months)] compared
with the high-mGPS group [1.9 months (95% CI, 0.4-7.0 months)]
(Fig. 2B).
Additionally, the median OS was significantly longer
in the low-CONUT group [26.1 months (95% CI, 12.5-80.5 months)]
compared with the high-CONUT group [14.3 months (95% CI, 1.0-22.1
months)] (Fig. 2C). The median PFS
tended to be longer in the low-CONUT group [7.5 months (95% CI,
4.7-11.3 months)] compared with the high-CONUT group [2.4 months
(95% CI, 0.6-7.0 months)]; however, the difference was not
statistically significant (Fig.
2D). By contrast, the median OS was significantly shorter in
the high-mGPS and high-CONUT group [5.4 months (95% CI, 0.4-14.3
months)] compared with the low-mGPS and/or low-CONUT group [25.7
months (95% CI, 16.0-80.5 months)] (Fig. S1A). Furthermore, the median PFS
was significantly shorter in the high-mGPS and high-CONUT group
[1.3 months (95% CI, 0.4-5.4 months)] compared with the low-mGPS
and/or low-CONUT group [7.5 months (95% CI, 4.6-10.2 months)]
(Fig. S1B). These results
indicated that patients with low-mGPS and/or low-CONUT scores may
have a prolonged survival duration.
Discussion
To the best of our knowledge, the present study was
the first to indicate that nutritional scores, such as the mGPS and
CONUT scores, may be independent prognostic factors in patients
undergoing chemotherapy for CUP. The results of the present study
demonstrated that OS and PFS were significantly longer in the
low-mGPS and/or low-CONUT groups compared with the respective
high-mGPS and high-CONUT groups. The multivariate analysis of
prognostic factors for OS indicated that low mGPS and low CONUT
scores were independent good prognostic factors. Taken together,
the results of these analyses suggested that good pretreatment
nutritional status may be associated with increased treatment
efficacy.
In previous clinical studies investigating CUP, the
median patient age is 60-65 years (23-26),
and the percentage of PS scores of 0 and 1 is >90% (25,27).
In addition, adenocarcinoma accounts for ~50% of tumor histology,
whereas undifferentiated carcinoma, which is difficult to classify,
accounts for ~20% (23,25). Furthermore, the number of
metastatic organs is <3 in 60% of the previous cases (23,27).
In the present study, 64% of the patients with CUP were aged ≥65
years, which was higher compared with previously reported
proportions in which 40-50% of patients were aged ≥60 years
(3,26). However, the PS score, tumor
histology and number of metastatic organs were similar to those
reported previously. Therefore, the data of the present study
suggested that advanced age may be a characteristic finding in
Japanese patients with CUP.
There are various indicators of nutritional status.
In the present study, the mGPS and CONUT scores were identified as
prognostic factors, while the PNI and NLR did not affect prognosis.
An ancillary study of the EXPAND trial (a phase III clinical
trial), which investigates the efficacy of first-line chemotherapy
for unresectable advanced gastric cancer, demonstrates that the
low-mGPS group has a notably prolonged OS compared with the
high-mGPS group (17,28). In another study, the low-CONUT
group showed a markedly higher 5-year OS and 5-year recurrence-free
survival compared with the high-CONUT group when assessing
preoperative nutritional status for resectable gastric cancer
(16). This indicates that a low
CONUT score may be a useful indicator during the perioperative
period. In the aforementioned study, the cut-off value of the CONUT
score is set at ≤3, which was also used in the present study.
However, the aforementioned study indicates that NLR and mGPS do
not affect OS (16). Therefore, in
gastric cancer, mGPS and CONUT scores may be useful prognostic
indicators in unresectable cases treated with chemotherapy and
resectable cases treated with surgery, respectively. Moreover, the
association between mGPS/GPS and OS is highlighted as a strong
prognostic factor in meta-analyses of multiple carcinomas (29). In the present study, multivariate
analyses highlighted that a low mGPS was a possible independent
favorable prognostic factor for PFS. Additionally, a low mGPS and
low CONUT score were highlighted to be possible independent
favorable prognostic factors for OS in multivariate analyses. These
results supported the findings of previous reports (16,17,29).
Furthermore, when patients were stratified according to whether the
location of the primary tumor could be estimated or not, univariate
analysis demonstrated that in cases where the location of the
primary tumor could be estimated, low-mGPS was an independent
favorable prognostic factor for PFS. In addition, a low mGPS and a
low CONUT score were independent favorable prognostic factors for
OS in cases where the location of the primary tumor could be
estimated. However, in cases where the location of the primary
tumor could not be estimated, a low mGPS and a low CONUT scores did
not affect prognosis in terms of PFS and OS. This may be because,
in the present study, the number of cases without an estimated
location of the primary tumor was small, and the high clinical
heterogeneity may have prevented an impact on prognosis. Therefore,
future investigations should include a larger sample size. These
results suggest that nutritional scores (low-mGPS and low-CONUT
score) may serve as prognostic predictors for CUP before
chemotherapy.
The present study suggested that the empirical
regimen may be an improved prognostic factor compared with a
site-estimated regimen in the univariate analysis of OS, excluding
nutritional indices. In a previous report that estimates the
primary tumor using DNA methylation profiles and reverse
transcription PCR, treatment with a site-specific regimen notably
prolonged OS compared with treatment with an empirical regimen
(30,31). By contrast, a prospective
randomized phase II clinical trial reports no notable difference in
efficacy between empirical and site-estimated regimens (25). In general, site-specific regimens
are based on gene expression profiles (25). However, in the present study, a
site-estimated regimen was selected based exclusively on
histological and immunological diagnoses; therefore, the regimens
did not overlap. Despite this difference, the results of the
present study did not support the efficacy of the site-estimated
regimen, as the site-estimated regimen was not retained as an
independent prognostic factor for OS in the multivariate analysis.
However, the number of patients in the present study was small,
which may have caused bias due to their high nutritional status.
Therefore, additional studies should be conducted in the
future.
The present study had a number of limitations.
Firstly, the small sample size was a limitation, which may have
potentially led to bias. Secondly, the present study was
retrospective and was conducted at a single institution in Japan.
Thirdly, although nutritional scores (such as the low-mGPS and
low-CONUT score) were highlighted as potentially favorable
prognostic factors in the present study, the cut-off values were
determined based on prior research and standard cut-off values are
yet to be established. Finally, a period of ~12 years was used to
include patients into the present study, which is longer compared
with the period of time used in the majority of cohort studies.
During the long period of time used to include patients in the
present study, the recommendations may have changed, especially in
terms of site-estimated regimens, which may have affected survival
outcomes.
In conclusion, the present study, to the best of our
knowledge, indicated for the first time that nutritional scores,
such as the mGPS and CONUT scores, are independent prognostic
factors in patients undergoing chemotherapy for CUP. These
nutritional scores may be useful biomarkers for estimating
prognosis before chemotherapy.
Supplementary Material
Kaplan-Meier survival analysis of the
OS and PFS stratified by the nutrition score. (A) OS and (B) PFS in
patients with CUP and a low-mGPS and/or low-CONUT score (n=18; blue
line) or a high mGPS and high CONUT score (n=7; red line). OS,
overall survival; PFS, progression-free survival; CI, confidence
interval; CUP, cancer of unknown primary origin; mGPS, modified
Glasgow prognostic score; CONUT, controlling nutritional
status.
Estimated location of the primary
tumor and the treatment regimen used in the present study.
Univariate analysis of prognostic
factors for overall survival stratified by patients with and
without estimated primary tumor sites.
Univariate analysis of prognostic
factors for progression-free survival stratified by patients with
and without estimated primary tumor sites.
Acknowledgements
The authors would like to thank Dr. Hideyuki Abe and
Dr. Akihiko Kawahara of the Department of Diagnostic Pathology,
Kurume University Hospital, Kurume, Japan, for their guidance and
support in diagnosing CUP during the present study.
Funding
Funding: The present study was partly supported by an award from
the Michel Mitsuo Yokoyama Foundation for the Promotion of
Immunology Science.
Availability of data and materials
The data generated in the present study may be
requested from the corresponding author.
Authors' contributions
SY and TT conceived of the present study and drafted
the manuscript. SN and YS collected clinical data and made
substantial contributions to the conception and design of the
present study. KMu, FF and KMi were involved in raw data analysis.
MF and TK contributed to the analysis and interpretation of data
and critically revised the manuscript for important intellectual
content. TK supervised, wrote, reviewed and edited the manuscript.
SY and TT confirm the authenticity of all the raw data. All authors
read and approved the final version of the manuscript.
Ethics approval and consent to
participate
The present study was conducted in accordance with
the principles in the Declaration of Helsinki and reviewed and
approved by the Ethics Committee of Kurume University School of
Medicine (approval no. 23229; Kurume, Japan). An opt-out approach
was used to obtain informed consent from patients, and personal
information was protected during data collection.
Patient consent for publication
Not applicable.
Competing interests
All authors declare that they have no competing
interests.
References
|
1
|
Pavlidis N and Pentheroudakis G: Cancer of
unknown primary site. Lancet. 379:1428–1435. 2012.PubMed/NCBI View Article : Google Scholar
|
|
2
|
Hess KR, Abbruzzese MC, Lenzi R, Raber MN
and Abbruzzese JL: Classification and regression tree analysis of
1000 consecutive patients with unknown primary carcinoma. Clin
Cancer Res. 5:3403–3410. 1999.PubMed/NCBI
|
|
3
|
Spurgeon L, Mitchell C, Cook N and Conway
AM: Cancer of unknown primary: The hunt for its elusive
tissue-of-origin - is it time to call off the search? Br J Cancer.
133:733–742. 2025.PubMed/NCBI View Article : Google Scholar
|
|
4
|
Kramer A, Bochtler T, Pauli C, Baciarello
G, Delorme S, Hemminki K, Mileshkin L, Moch H, Oien K, Olivier T,
et al: Cancer of unknown primary: ESMO clinical practice guideline
for diagnosis, treatment and follow-up. Ann Oncol. 34:228–246.
2023.PubMed/NCBI View Article : Google Scholar
|
|
5
|
Olivier T, Fernandez E, Labidi-Galy I,
Dietrich PY, Rodriguez-Bravo V, Baciarello G, Fizazi K and
Patrikidou A: Redefining cancer of unknown primary: Is precision
medicine really shifting the paradigm? Cancer Treat Rev.
97(102204)2021.PubMed/NCBI View Article : Google Scholar
|
|
6
|
Palmeri S, Lorusso V, Palmeri L, Vaglica
M, Porta C, Nortilli R, Ferraú F, Comella G, Massidda B and Danova
M: Carcinomas of Unknown Primary Italian Study Group. Cisplatin and
gemcitabine with either vinorelbine or paclitaxel in the treatment
of carcinomas of unknown primary site: Results of an Italian
multicenter, randomized, phase II study. Cancer. 107:2898–2905.
2006.PubMed/NCBI View Article : Google Scholar
|
|
7
|
Raghav K: Cancer of unknown primary site.
N Engl J Med. 392:2035–2047. 2025.PubMed/NCBI View Article : Google Scholar
|
|
8
|
Briasoulis E, Kalofonos H, Bafaloukos D,
Samantas E, Fountzilas G, Xiros N, Skarlos D, Christodoulou C,
Kosmidis P and Pavlidis N: Carboplatin plus paclitaxel in unknown
primary carcinoma: A phase II Hellenic cooperative oncology group
study. J Clin Oncol. 18:3101–3107. 2000.PubMed/NCBI View Article : Google Scholar
|
|
9
|
Haratani K, Hayashi H, Takahama T,
Nakamura Y, Tomida S, Yoshida T, Chiba Y, Sawada T, Sakai K, Fujita
Y, et al: Clinical and immune profiling for cancer of unknown
primary site. J Immunother Cancer. 7(251)2019.PubMed/NCBI View Article : Google Scholar
|
|
10
|
Abu Sabbah T, Theurer S, Baba HA, Lueong
S, Rashidi-Alavijeh J, Hilser T, Zaun Y, Metzenmacher M,
Pogorzelski M, Virchow I, et al: PD-L1 expression associates with
favorable survival of patients with cancer of unknown primary (CUP)
not treated with checkpoint inhibitors. Eur J Cancer.
210(114268)2024.PubMed/NCBI View Article : Google Scholar
|
|
11
|
Tanizaki J, Yonemori K, Akiyoshi K, Minami
H, Ueda H, Takiguchi Y, Miura Y, Segawa Y, Takahashi S, Iwamoto Y,
et al: Open-label phase II study of the efficacy of nivolumab for
cancer of unknown primary. Ann Oncol. 33:216–226. 2022.PubMed/NCBI View Article : Google Scholar
|
|
12
|
Raghav KP, Stephen B, Karp DD, Piha-Paul
SA, Hong DS, Jain D, Chudy Onwugaje DO, Abonofal A, Willett AF,
Overman M, et al: Efficacy of pembrolizumab in patients with
advanced cancer of unknown primary (CUP): A phase 2 non-randomized
clinical trial. J Immunother Cancer. 10(e004822)2022.PubMed/NCBI View Article : Google Scholar
|
|
13
|
Culine S, Kramar A, Saghatchian M, Bugat
R, Lesimple T, Lortholary A, Merrouche Y, Laplanche A and Fizazi K:
French Study Group on Carcinomas of Unknown Primary. Development
and validation of a prognostic model to predict the length of
survival in patients with carcinomas of an unknown primary site. J
Clin Oncol. 20:4679–4683. 2002.PubMed/NCBI View Article : Google Scholar
|
|
14
|
Seve P, Ray-Coquard I, Trillet-Lenoir V,
Sawyer M, Hanson J, Broussolle C, Negrier S, Dumontet C and Mackey
JR: Low serum albumin levels and liver metastasis are powerful
prognostic markers for survival in patients with carcinomas of
unknown primary site. Cancer. 107:2698–2705. 2006.PubMed/NCBI View Article : Google Scholar
|
|
15
|
Parsons HM, Forte ML, Abdi HI, Brandt S,
Claussen AM, Wilt T, Klein M, Ester E, Landsteiner A, Shaukut A, et
al: Nutrition as prevention for improved cancer health outcomes: A
systematic literature review. JNCI Cancer Spectr.
7(pkad035)2023.PubMed/NCBI View Article : Google Scholar
|
|
16
|
Kuroda D, Sawayama H, Kurashige J,
Iwatsuki M, Eto T, Tokunaga R, Kitano Y, Yamamura K, Ouchi M,
Nakamura K, et al: Controlling nutritional status (CONUT) score is
a prognostic marker for gastric cancer patients after curative
resection. Gastric Cancer. 21:204–212. 2018.PubMed/NCBI View Article : Google Scholar
|
|
17
|
Hacker UT, Hasenclever D, Baber R, Linder
N, Busse H, Obermannova R, Zdrazilova-Dubska L, Valik D and Lordick
F: Modified Glasgow prognostic score (mGPS) is correlated with
sarcopenia and dominates the prognostic role of baseline body
composition parameters in advanced gastric and esophagogastric
junction cancer patients undergoing first-line treatment from the
phase III EXPAND trial. Ann Oncol. 33:685–692. 2022.PubMed/NCBI View Article : Google Scholar
|
|
18
|
Zhang W, Tan Y, Li Y and Liu J: Neutrophil
to Lymphocyte ratio as a predictor for immune-related adverse
events in cancer patients treated with immune checkpoint
inhibitors: A systematic review and meta-analysis. Front Immunol.
14(1234142)2023.PubMed/NCBI View Article : Google Scholar
|
|
19
|
Watanabe H, Yamada T, Komori K, Hara K,
Kano K, Takahashi K, Kumazu Y, Fujikawa H, Numata M, Aoyama T, et
al: Effect of prognostic nutrition index in gastric or
gastro-oesophageal junction cancer patients undergoing nivolumab
monotherapy. In Vivo. 35:563–569. 2021.PubMed/NCBI View Article : Google Scholar
|
|
20
|
Ikuta Y, Takamori H, Sakamoto Y, Hashimoto
D, Chikamoto A, Kuroki H, Sakata K, Sakamoto K, Hayashi H, Imai K,
et al: The modified Glasgow Prognostic Score (mGPS) is a good
predictor of indication for palliative bypass surgery in patients
with unresectable pancreatic and biliary cancers. Int J Clin Oncol.
19:629–633. 2014.PubMed/NCBI View Article : Google Scholar
|
|
21
|
Huebner G, Link H, Kohne CH, Stahl M,
Kretzschmar A, Steinbach S, Folprecht G, Bernhard H, Al-Batran SE,
Schoffski P, et al: Paclitaxel and carboplatin vs gemcitabine and
vinorelbine in patients with adeno- or undifferentiated carcinoma
of unknown primary: A randomised prospective phase II trial. Br J
Cancer. 100:44–49. 2009.PubMed/NCBI View Article : Google Scholar
|
|
22
|
Eisenhauer EA, Therasse P, Bogaerts J,
Schwartz LH, Sargent D, Ford R, Dancey J, Arbuck S, Gwyther S,
Mooney M, et al: New response evaluation criteria in solid tumours:
Revised RECIST guideline (version 1.1). Eur J Cancer. 45:228–247.
2009.PubMed/NCBI View Article : Google Scholar
|
|
23
|
Yang H, He F, Xu W and Cao Z: Clinical
features of cancer with unknown primary site (clinical features,
treatment, prognosis of cancer with unknown primary site). BMC
Cancer. 22(1372)2022.PubMed/NCBI View Article : Google Scholar
|
|
24
|
Fusco MJ, Knepper TC, Balliu J, Del Cueto
A, Laborde JM, Hooda SM, Brohl AS, Bui MM and Hicks JK: Evaluation
of targeted next-generation sequencing for the management of
patients diagnosed with a cancer of unknown primary. Oncologist.
27:e9–e17. 2022.PubMed/NCBI View Article : Google Scholar
|
|
25
|
Hayashi H, Kurata T, Takiguchi Y, Arai M,
Takeda K, Akiyoshi K, Matsumoto K, Onoe T, Mukai H, Matsubara N, et
al: Randomized phase II trial comparing site-specific treatment
based on gene expression profiling with carboplatin and paclitaxel
for patients with cancer of unknown primary site. J Clin Oncol.
37:570–579. 2019.PubMed/NCBI View Article : Google Scholar
|
|
26
|
Kang S, Jeong JH, Yoon S, Yoo C, Kim KP,
Cho H, Ryoo BY, Jung J and Kim JE: Real-world data analysis of
patients with cancer of unknown primary. Sci Rep.
11(23074)2021.PubMed/NCBI View Article : Google Scholar
|
|
27
|
Takamizawa S, Shimoi T, Yoshida M, Tokura
M, Yazaki S, Mizoguchi C, Saito A, Kita S, Yamamoto K, Kojima Y, et
al: Diagnostic value of tumor markers in identifying favorable or
unfavorable subsets in patients with cancer of unknown primary: A
retrospective study. BMC Cancer. 22(412)2022.PubMed/NCBI View Article : Google Scholar
|
|
28
|
Lordick F, Kang YK, Chung HC, Salman P, Oh
SC, Bodoky G, Kurteva G, Volovat C, Moiseyenko VM, Gorbunova V, et
al: Capecitabine and cisplatin with or without cetuximab for
patients with previously untreated advanced gastric cancer
(EXPAND): A randomised, open-label phase 3 trial. Lancet Oncol.
14:490–499. 2013.PubMed/NCBI View Article : Google Scholar
|
|
29
|
Dolan RD, McSorley ST, Horgan PG, Laird B
and McMillan DC: The role of the systemic inflammatory response in
predicting outcomes in patients with advanced inoperable cancer:
Systematic review and meta-analysis. Crit Rev Oncol Hematol.
116:134–146. 2017.PubMed/NCBI View Article : Google Scholar
|
|
30
|
Moran S, Martinez-Cardus A, Sayols S,
Musulén E, Balañá C, Estival-Gonzalez A, Moutinho C, Heyn H,
Diaz-Lagares A, de Moura MC, et al: Epigenetic profiling to
classify cancer of unknown primary: A multicentre, retrospective
analysis. Lancet Oncol. 17:1386–1395. 2016.PubMed/NCBI View Article : Google Scholar
|
|
31
|
Hainsworth JD, Rubin MS, Spigel DR, Boccia
RV, Raby S, Quinn R and Greco FA: Molecular gene expression
profiling to predict the tissue of origin and direct site-specific
therapy in patients with carcinoma of unknown primary site: A
prospective trial of the Sarah Cannon research institute. J Clin
Oncol. 31:217–223. 2013.PubMed/NCBI View Article : Google Scholar
|
