Tumor-to-tumor metastasis from lung cancer: a clinicopathological postmortem study
Susumu Matsukuma · Takako Kono · Hiroaki Takeo · Yusuke Hamakawa · Kimiya Sato
Received: 16 May 2013 /Revised: 17 June 2013 / Accepted: 9 July 2013 C Springer-Verlag Berlin Heidelberg 2013
Abstract This study examined 47 cases of lung cancer concomitant with other tumors and found eight cases (17 %) with nine foci of tumor-to-tumor metastasis, defined as metastasis of lung cancer into another tumor. Donor lung cancers were four adenocarcinomas, two squamous cell car- cinomas, and two small cell carcinomas. Tumor-to-tumor metastasis was found in five of six renal cell carcinomas (83 %), one of eight thyroid papillary carcinomas (13 %), one of three adrenocortical adenomas (33 %), one of three pancreatic endocrine microadenomas (33 %), and another lung cancer (one of six cases of multiple lung cancers, 17 %). The higher recipient incidence in renal cell carcinoma was statistically significant compared with prostatic carcinoma (0/16, P<0.001), colorectal carcinoma (0/7, P=0.005), and gastric carcinoma (0/5, P=0.015). Generalized metastases were found in 88 % of the tumor-to-tumor metastasis cases. The total clinical course of patients with tumor-to-tumor metastasis was shorter than that of the patients without tumor-to-tumor metastasis (mean, 5.4 versus 18.8 months; P=0.046). Tumor-to-tumor metastasis sometimes mimicked undifferentiated recipient tumor cells. Immunostains for thy- roid transcription factor 1 (TTF-1), Napsin A, cytokeratin 7 (CK7), and CK5/6 were useful to confirm tumor-to-tumor metastasis. However, TTF-1-, Napsin A-, and/or CK7- negative lung adenocarcinoma components metastasized to renal cell carcinoma in three cases, and recipient renal cell carcinomas were focally Napsin A+ (two cases) or CK7+ (two cases). Tumor-to-tumor metastasis can occur as a result
S. Matsukuma · T. Kono . H. Takeo . Y. Hamakawa · K. Sato Department of Pathology, Japan Self-Defense Forces Central Hospital, 1-2-24 Ikejiri, Setagaya-ku, 154-0001 Tokyo, Japan
Health Care Center, Japan Self-Defense Forces Central Hospital, 1-2-24 Ikejiri, Setagaya-ku, 154-0001 Tokyo, Japan
e-mail: skuma@cocoa.plala.or.jp
of metastasis from lung cancer with more aggressive behav- ior. Tumor-to-tumor metastasis should be carefully distin- guished from undifferentiated recipient tumor cells.
Keywords Tumor-to-tumor metastasis · Cancer-to-cancer metastasis · Lung cancer · Renal cell carcinoma · Multiple lung cancers · Endocrine microadenoma · Immunohistochemistry
Introduction
“Tumor-to-tumor metastasis” is a rare phenomenon in which one tumor metastasizes into another tumor [1-22] and in- cludes “cancer-to-cancer metastasis” [1-14] and “cancer-to- benign tumor metastasis” [15-22]. Four criteria have been traditionally used to determine tumor-to-tumor metastasis [2-6, 11, 22]: (1) the presence of two or more distinct neoplasms, (2) the exception of a “collision tumor,” defined as two histologically distinct tumors simultaneously involv- ing the same location with an intermediate transitional zone, (3) the presence of extravascular metastasis, and (4) the exception of neoplasms metastasizing to the lymphatic sys- tem which were already involved by generalized lymphatic or hematological malignancy. Previous articles [3-6, 8, 10, 11, 20] have described that lung cancer is the most frequent donor tumor. However, most of the articles were sporadic case reports, and the true incidence of tumor-to-tumor me- tastasis from lung cancer remains unknown. The recognition of tumor-to-tumor metastasis can be challenging in cases with no information regarding prior or concomitant malig- nancy [9, 10]. To elucidate the incidence and clinicopatho- logical characteristics of tumor-to-tumor metastasis from lung cancer, we investigated autopsy cases of lung cancer with additional lung cancers and/or extrapulmonary tumors.
Materials and methods
Forty-seven cases of lung cancer, in which autopsy disclosed other distinct lung or extrapulmonary tumors, were retrieved from the autopsy files of the Department of Pathology, Japan Self-Defense Forces Central Hospital (from 1975 to 2012), Tokyo. Hematological cancer cases and post-therapeutic extrapulmonary cancer cases with no residual cancer cells were excluded. Medical charts, autopsy request forms, and surgical pathology files were reviewed for clinical informa- tion and pathological findings of surgically removed speci- mens. Tumors were diagnosed and classified according to the WHO classification and/or published textbooks [23-27]. The clinicopathological features are summarized in Table 1. Six patients had multiple lung cancers. A total of 55 lung cancers and 64 extrapulmonary tumors were examined for tumor-to-tumor metastasis satisfying four criteria described in the “Introduction”; the former included thirty-four adeno- carcinomas, fourteen squamous cell carcinomas, five small cell carcinomas, one pleomorphic carcinoma, and one carci- nosarcoma, and the latter consisted of sixteen prostatic ade- nocarcinomas, eight thyroid papillary carcinomas, eight co- lorectal adenocarcinomas, six renal cell carcinomas (RCCs), five gastric adenocarcinomas, four renal papillary adenomas, three adrenocortical adenomas, three pancreatic endocrine microadenomas, three renal angiomyolipomas, and eight other tumors. The postmortem interval ranged from 0.9 to 38.9 h (mean, 9.8 h). The representative tissues sampled from generalized organs for histological evaluation ranged from 24 to 79 sections (mean, 50.3 sections) in 36 cases without central nervous system examination and ranged from 32 to 77 sections (mean, 58.7 sections) in 11 cases with central nervous system examination. The representative tis- sues of lung cancers and extrapulmonary tumors ranged from 1 to 12 sections (mean, 3.5 sections). Gross photographs that were available were retrospectively examined. Representa- tive tissues of donor lung cancer and tumor-to-tumor metas- tasis were cut and immunostained with thyroid transcription factor 1 (TTF-1; SPT24; Novocastra, Newcastle, UK), Napsin A (IP64; Novocastra), cytokeratin 7 (CK7; OV-TL 12/30; Dako, Glostrup, Denmark), CK5/6 (D5/16 B14; Dako), vimentin (Vim; V9; Nichirei Corporation, Tokyo, Japan), and D2-40 (Dako). Positivity was assessed at 5- 100 % of positive cells. Modifying the models proposed by Edeline et al. [28], we immunostained the representative sections of all six RCCs with CD34 (QBEnd10; Dako) and attempted to divide them into high angiogenic phenotype and low angiogenic phenotype; the former was composed of richly distributed vasculature, and the latter consisted of scattered or poorly distributed vessels. We statistically ex- amined the recipient incidence of tumor-to-tumor metastasis among the coexisting tumors and the relationship between tumor-to-tumor metastasis and clinicopathological features
using Fisher’s exact test, the chi-square test, unpaired ttest, and Mann-Whitney U test. Statistical significance was set at P<0.05.
Results
In eight (17.0 %) of the forty-seven patients with lung cancer, nine tumor-to-tumor metastases were identified. The clini- copathological findings are shown in Table 2. The donor lung cancers were four papillary adenocarcinomas with or without large foci of poorly differentiated components, two squamous cell carcinomas, and two small cell carcinomas, and their size ranged from 2.5 to 8.2 cm (mean, 4.8 cm). The nine recipient tumors included one thyroid papillary carci- noma, one lung adenocarcinoma, five RCCs, one adrenocor- tical adenoma, and one pancreatic endocrine microadenoma, and the recipient incidences of tumor-to-tumor metastasis were 12.5 %, 16.7 %, 66.7 %, 33.3 %, and 33.3 %, respec- tively. In seven patients, generalized metastases were found, but in the other (Case 7), cancerous spreading other than metastasis to adrenocortical adenoma was localized in the upper thoracic and cervical regions. Metastases to non- neoplastic recipient tissues adjacent to tumor-to-tumor me- tastasis were found in four but not in the other five. Com- pared with the sites of donor tumors, tumor-to-tumor metas- tases were ipsilateral (two lesions, 22.2 %) or contra-lateral (seven lesions including a pancreatic tail lesion regarded as “left-sided,” 77.8 %). Metastasis of the recipient tumor itself was observed only in Case 8, which was limited to the ipsilateral regional lymph nodes. All patients died of lung cancer within 12 months from the time of onset.
Tumor-to-tumor metastases showed a scattered distribu- tion in minute nests and/or clusters without necrosis (Figs. 1 and 2a-d) or small nodular features with or without necrosis (Figs. 2e and 3). The largest metastatic deposit was found in a nodular metastasis to adrenocortical adenoma (1.2 cm in the longest dimension). In Case 8, a whitish small nodule, corresponding to tumor-to-tumor metastasis, could be mac- roscopically pointed out (Fig. 3b). There were no other macroscopic findings indicating T-T-M in any cases. Tumor-to-tumor metastatic cells were histologically identi- cal to the respective lung cancer cells. In most cases, there were histological abrupt borders between metastatic lung cancer cells and recipient tumor cells. Therefore, compared with donor tumor histology, tumor-to-tumor metastases were detectable. However, looking at the recipient tumors only, the recognition of tumor-to-tumor metastasis was difficult in some cases. A few scattered tumor-to-tumor metastases to RCC and pancreatic endocrine microadenoma mimicked undifferentiation of recipient tumor cells (Figs. 1f and 2b). Small nodular metastasis of squamous cell carcinoma
| Demographics | Histology | ||
|---|---|---|---|
| Age (years) | 48-94 (mean, | 75.1) Histology of lung cancers (n=55) | |
| Gender, male/female | 33/14 | Adenocarcinoma | 34 |
| Therapy for lung cancer | Squamous cell carcinoma | 14 | |
| Operation with or without chemo/radiotherapy | 10 | Small cell carcinoma | 5 |
| Chemo/radiotherapy only | 20 | Pleomorphic carcinoma | 1 |
| None (conservative therapy only) | 17 | Carcinosarcoma | 1 |
| Cause of death | Histology of coexisting Extrapulm tumors (n=64) | ||
| Died of lung cancer | 37 | Malignant (n=49) | |
| Died of other disease | 10 | Prostatic adenocarcinoma | 16 (all latent) |
| Other cancers | 4ª | Thyroid papillary carcinoma | 8 (all latent) |
| Pneumonia | 1 | Colorectal adenocarcinoma | 8e (5 latent/3 post-therapyf) |
| Myocardial infarction | 1 | RCC | 6 (all latent) |
| Pulmonary infarction | 1 | Gastric adenocarcinoma | 5 (4 latent/1 no therapy) |
| Sudden death after operation | 1 | Panc adenocarcinoma | 2 (1 latent/1 post-therapyf) |
| Multiplicity of lung cancer and other tumors | Hepatocellular carcinoma | 1 (post-therapyf) | |
| 1 lung cancer (with other tumors) | 41b | Gallbladder adenocarcinoma | 1 (latent) |
| +1 or 2 Extrapulm benign tumors | 7 | Cholangiocellular carcinoma | 1 (latent) |
| +1 Extrapulm Ca only | 22 | Undifferentiated carcinoma of the pleuroperitoneal cavity | 1 (post-therapyf) |
| +1 Extrapulm Ca+ 1 or 2 Extrapulm benign tumors | 3 | Benign (n=15) | |
| +2 Extrapulm Ca+ 0 or 2 Extrapulm benign tumors | 7 | Renal papillary adenoma | 4 (latent) |
| +3 Extrapulm Ca only | 2 | Adrenocortical adenoma | 3 (latent) |
| 2 lung cancers | 5℃ | Renal angiomyolipoma | 3 (latent) |
| Without Extrapulm tumors | 2 | Panc endocrine microadenoma | 3 (latent) |
| +1 Extrapulm Ca only | 2 | Adrenal pheochromocytoma | 1 (latent) |
| +2 Extrapulm Ca only | 1 | Panc serous cystadenoma | 1 (latent) |
| 4 lung carcinomas only (without other tumors) | 1ª | ||
Ca cancer, Extrapulm extrapulmonary, Panc pancreatic, RCC renal cell carcinoma
a Including one colon adenocarcinoma, one pancreatic adenocarcinoma, one hepatocellular carcinoma, and one undifferentiated carcinoma of the pleuroperitoneal cavity
b Including five latent lung cancers
” Including two latent lung cancers
d Including two latent lung cancers
e Including one with double colorectal carcinomas
f Residual primary and/or metastatic cancer cells were present at autopsy
somewhat resembled focal solid growth of the recipient lung adenocarcinoma (Fig. 3c, d).
Table 3 summarizes the immunohistochemical findings of the tumor-to-tumor metastasis cases. In Case 6, metastatic foci disappeared in the additionally cut sections, and further immunostained features were unknown. Donor (primary) lung adenocarcinomas were Napsin A+/CK7+ in four (100 %), TTF-1+ in two (50 %), and Vim+ in two (50 %). However, TTF-1-, Napsin A-, and CK7- poorly differentiat- ed components metastasized to RCC in three cases, three cases, and one case, respectively. Therefore, TTF-1+, Napsin A+, and CK7+ were observed in only one (25 %), one (25 %), and three (75 %) of four tumor-to-tumor metastases
from lung adenocarcinoma, respectively. Donor lung squa- mous cell carcinomas were CK/5/6+ in two (100 %), CK7+ in one (50 %), CD10+ in one (50 %), Vim+ in one (50 %), and D2-40+ in one (50 %). Tumor-to-tumor metastasis from squamous cell carcinoma showed similar features to donor tumors, including D2-40+ metastasis to adrenocortical ade- noma (Fig. 2e). Donor small cell carcinomas were CK7+ in two (100 %) and TTF-1+ in one (50 %). Tumor-to-tumor metastasis from lung small cell carcinoma examined was weak CK7+ in two (100 %) and CD10+ in one (50 %). Recipient RCCs were Vim+ in five (100 %), CD10+ in four (80 %), focal Napsin A+ in two (40 %, Fig. 1h), and focal CK7+ in two (40 %). Recipient adrenocortical adenoma was
| Case | Age/sex | Donor lung cancers (n=8) | Recipient tumors (n=9) | Tumor-to-tumor metastases (n=9) | Total clinical course (months) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Location (lung) | Size | (cm) Histology | Metastasis in non-neoplastic areas of recipient organ | Location | Size | (cm) Histology | Metastasis of recipient tumor | Growth features | Necrosis | |||
| 1 | 88/M | Rt lower lobe | 8.2 | Adenocaª | Noned | Lt kidney | 1.3 | Clear cell RCCf | None | Scatteredh | None | 0.7 |
| 2 | 69/M | Rt upper lobe | 2.5 | Adenocab | Presentd | Rt kidney | 3 | Clear cell RCC | None | Scattered | None | 7 |
| 3 | 72/M | Lt upper lobe | 6.5 | Adenocaª | Presentd | Rt kidney | 8 | Pap RCC | None | Scatteredh | None | 1 |
| 4 | 48/M | Lt upper lobe | 2.6 | Adenocaª | Noned | Lt kidney | 1.2 | Clear cell RCC | None | Scatteredh | None | 12 |
| 5 | 82/M | Rt lower lobe | 6 | Small cell | Noned | Lt kidney | 2 | Clear cell RCC | None | Small nodule | Present | 8 |
| carcinoma | Presentd | Pancreas (tail portion) | 0.3 | Endocrine microadenoma | None | Scattered | None | |||||
| 6 | 67/F | Lt lower lobe | 3.5 | Small cell carcinoma | Noned | Thyroid, (right lobe) | 0.5 | Pap carcinoma | None | Scattered | None | 9 |
| 7 | 82/F | Rt upper lobe | 4.5 | SqCCº | Nonee | Lt adrenal gland | 1.5 | Adrenocortical adenoma | None | Small nodule | Present | 1.6 |
| 8 | 66/M | Lt upper lobe | 4.8 | SqCCC | Presentd | Rt lung (lower lobe) | 2.5 | Adenoca | Lymph nodes& | Small nodule | None | 4 |
Adenoca adenocarcinoma, Lt left, Pap papillary, RCC renal cell carcinoma, Rt right, SqCC squamous cell carcinoma
a Papillary adenocarcinoma with large foci of poorly differentiated components
b Papillary adenocarcinoma
” Moderately differentiated
d Other distant metastases also seen in generalized organs
e Other direct invasion to mediastinum and lymph node metastases (right pulmonary hilar and cervical) only seen
“Chiefly composed of granular cell type
§ Right pulmonary hilar lymph nodes only
h Metastasis of poorly differentiated adenocarcinomatous components only
TTF-1
C
Napsin A
d
CK7
a
b
e
CK7
Napsin A
f
g
h
Vim+ only, and thyroid papillary carcinoma was TTF-1+ /CK7+ and focally CK5/6+. Recipient pancreatic endocrine microadenoma was negative for any antibodies examined.
admixed with scattered tumor-to-tumor metastasis (f, arrows) in Case 3, and high-power view of tumor-to-tumor metastasis (f, inset)of poorly differentiated adenocarcinoma, mimicking undifferentiated RCC cells. Tumor-to-tumor metastatic cells highlighted by CK7+ (g and g, inset) (f ×100; f, inset ×450; g ×100; g, inset ×450). h RCC cells focally showing Napsin A+ in Case 3 (×450)
Recipient lung adenocarcinoma was TTF-1+/Napsin A+ /CK7+. Practically, tumor-to-tumor metastasis from lung ad- enocarcinomas could be discriminated from RCC by TTF-1+,
CK7
*
*
*
*
*
b
*
*
a
C
Necrosis
d
e
a
b
c
CK5/6
e
T-T-M
TTF-1
T-T-M
d
f
Napsin A+, and/or CK7+ (Fig. 1c-e, g). Focal CK7+ features in two RCCs showed weakly cytoplasmic membranous stain- ing, which was different from the diffuse cytoplasmic CK7+ in metastatic lung adenocarcinoma cells. However, non-neoplastic renal tubules were always Napsin A+/CK7+, and some of them were focally entrapped in RCC. The CD10+ in RCC cells was variable and did not always highlight CD10 tumor-to-tumor metastasis. Metastatic small cell carcinoma could be discrimi- nated from pancreatic endocrine microadenoma and RCC by weak CK7+ (Fig. 2c). Metastatic squamous cell carcinoma was distinguishable from adrenocortical adenoma by CK5/6+/D2- 40+/Vim- and could be discriminated from recipient adenocar- cinoma by CK5/6+/CD10+ (Fig. 3e) and TTF-1/Napsin A- (Fig. 3f).
A statistically significant difference in the recipient inci- dence of tumor-to-tumor metastasis was found between RCC (5/6) and prostatic carcinoma (0/16, P<0.001), colorectal car- cinoma (0/7, P=0.005), gastric carcinoma (0/5, P=0.015), renal papillary adenoma (0/4, P=0.048), and renal angiomyolipoma (0/3, P=0.048). The mean size of these RCCs, prostatic carci- nomas, colorectal carcinomas, gastric carcinomas, renal papil- lary adenomas, and renal angiomyolipomas were 2.8 (range,
metastasis) intermingled with recipient adenocarcinoma cells, resem- bling solid growth of adenocarcinoma cells (x100). e CK5/6+ highlighting metastatic squamous cell carcinoma (arrows) in CK5/6- adenocarcinoma cells. Compare with HE-stained features of the same area shown in Fig. 3c (×20). f TTF-1-negative tumor-to-tumor metas- tasis is distinguishable from TTF-1+ recipient lung adenocarcinoma cells (×200).
1.2-8), 0.89 (range, 0.2-3), 3.4 (range, 0.5-10), 2.7 (range, 1.9-4), 0.42 (range, 0.07-1), and 0.47 cm (range, 0.4-0.6 cm), respectively. Among other coexisting tumors, there were no significant differences in the recipient incidence of tumor-to- tumor metastasis. Table 4 is a summary of the relationship between tumor-to-tumor metastasis and clinicopathological features. There was a significant difference in the number of non-neoplastic organs involved by metastatic lung cancer be- tween patients with and without tumor-to-tumor metastasis (mean, 10.3 versus 4.6, P=0.003). Tumor-to-tumor metastasis was associated with metastases to the non-neoplastic thyroid gland (P=0.004), liver (P=0.022), pancreas (P=0.003), left or right adrenal gland (P=0.010), and both adrenal glands (P=0.002). The total clinical course of eight patients with tumor-to-tumor metastasis was shorter than that of 29 patients without tumor-to-tumor metastasis who died of lung carcinoma (mean, 5.4 versus 18.8 months; P=0.046). Tumor-to-tumor metastasis was not associated with age, gender, donor tumor size, recipient tumor size, or metastases to the lungs, kidneys, spleen, small intestines, bone, or lymph nodes. CD34+ vascu- larity of five recipient RCCs could be divided into three with high angiogenic phenotype and two with low angiogenic
| TTF-1, n (%) | Napsin A, n (%) | CK7, n (%) | CK5/6, n (%) | CD10, n (%) | Vim, n (%) | D2-40, n (%) | |
|---|---|---|---|---|---|---|---|
| Immunoexpression of donor (primary) lung cancers | |||||||
| Adenocarcinoma (n=4) | 2ª (50.0) | 4b (100) | 4º (100) | 0ª (0.0) | 0 (0.0) | 2 (50.0) | 0 (0.0) |
| Squamous cell carcinoma (n=2) | 0 (0.0) | 0f (0.0) | 1 (50.0) | 2 (100) | 1ª (50.0) | 1ª (50.0) | 1 (50.0) |
| Small cell carcinoma (n=2) | 1 (50.0) | 0 (0.0) | 2 (100) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Immunoexpression of tumor-to-tumor metastases | |||||||
| Adenocarcinoma (n=4) | 1ª (25.0) | 1ª (25.0) | 3f (75.0) | 0° (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Squamous cell carcinoma (n=2) | 0 (0.0) | 0 (0.0) | 1 (50.0) | 2 (100) | 1 (50.0) | 1 (50.0) | 1 (50.0) |
| Small cell carcinoma (n=2)g | 0 (0.0) | 0 (0.0) | 2 (100) | 0 (0.0) | 1 (50.0) | 0 (0.0) | 0 (0.0) |
| Immunoexpression of recipient tumors | |||||||
| RCC (n=5) | 0 (0.0) | 2h (40.0) | 2ª (40.0) | 0 (0.0) | 4ª (80.0) | 5 (100) | 0 (0.0) |
| Adrenocortical adenoma (n=1) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (100) | 0 (0.0) |
| Thyroid papillary carcinoma (n=1) | 1 (100) | 0 (0.0) | 1 (100) | 1 (100) | 0ª (0.0) | 1 (100) | 0 (0.0) |
| Pancreatic endocrine microadenoma (n=1) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Lung adenocarcinoma (n=1) | 1 (100) | 1 (100) | 1 (100) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
RCC renal cell carcinoma, TTF-1 thyroid transcription factor 1, Vim vimentin
a Multifocally containing TTF-1-negative poorly differentiated areas in one tumor (Case 3)
b Multifocally containing Napsin A-negative poorly differentiated areas in three tumors (Cases 1, 3, and 4)
” Multifocally containing CK7-negative poorly differentiated areas in one tumor (Case 1)
d Excluding tumors <5% of positive cells
e One tumor in Case 2; tumor-to-tumor metastases in the other three cases showing metastases of TTF-1-negative and Napsin A-negative poorly differentiated components only
Three tumors (Cases 2-4); tumor-to-tumor metastases in Case 1 showing metastases of CK7-negative poorly differentiated components only
” Metastatic foci in Case 6 was disappeared in the additionally cut sections and a total number of tumor-to-tumor metastases from lung small cell carcinoma examined was two in Case 5.
h Positive cells, 10-20 %
i Weakly positive cells (cytoplasmic membrane), 5-10 %
phenotype. The vascularity of the other RCC without tumor-to- tumor metastasis was high angiogenic phenotype. There was no statistically significant association of tumor-to-tumor metas- tasis with high angiogenic phenotype of RCC.
Discussion
Previous studies of tumor-to-tumor metastasis from lung can- cer in larger series are rare. Ottoson et al. [1] investigated 478 lung cancers, 265 renal cancers, and 112 small intestinal cancers and found tumor-to-tumor metastasis from lung can- cer to RCC (3 cases) and to an ileal carcinoid tumor (1 case). However, in their study, the number of cases with coexistence of lung cancers and these tumors was unknown. Onuigbo [15] reported that 15 (0.21 %) of 7,238 lung cancers metastasized to adrenocortical adenoma, but he did not state the coexistent incidence of adrenocortical adenoma either. Moriya et al. [17] reported that tumor-to-tumor metastasis to adrenocortical ad- enoma was found in 1 (0.63 %) of 160 autopsy cases of lung
cancer, but the number of coexisting adrenocortical adenoma was limited to only 1. In the current study, we examined 47 lung cancer cases concomitant with other distinct tumors and found 8 tumor-to-tumor metastasis cases (17.0 %), in 14.5 % of 55 lung carcinomas examined. This may be the true inci- dence of tumor-to-tumor metastasis from lung cancer. Seven (88 %) of eight patients with tumor-to-tumor metastasis had widespread other metastases of lung cancer. The number of organs metastasized by lung cancer in patients with tumor-to- tumor metastasis was significantly higher than that in those without tumor-to-tumor metastasis. The total clinical course of the patients with tumor-to-tumor metastasis was significantly shorter than that of the patients without tumor-to-tumor me- tastasis. These findings suggest that tumor-to-tumor metasta- sis is a result of metastasis of lung cancer with more aggres- sive behavior. Our results also imply that tumor-to-tumor metastasis was not associated with the size of donor lung cancers or size of recipient tumors.
In the current study, the coexistent incidence of RCC in lung cancer patients was not high (13 %), while the recipient
| Tumor-to-tumor metastasis (+) (n=8) | Tumor-to-tumor metastasis (-) (n=39) | P | |
|---|---|---|---|
| Age (years) | 48-88 (mean, 71.8) | 54-94 (mean, 75.8) | 0.543f |
| Gender, male/female | 6/2 | 27/12 | 0.921ª |
| Size of donor lung cancer (cm) | 2.5-8.2 (mean, 4.8) | 1-10.5 (mean, 4.0) [46 lung cancers]ª | 0.387f |
| Size of recipient (coexisting) tumor (cm) | 0.3-8 (mean, 2.3) | 0.05-10 (mean, 2.0) [66 tumors] b | 0.786f |
| The number of non-neoplastic organs involved by lung cancer ℃ | 2-16 (mean, 10.3) | 0-16 (mean, 4.6) | 0.003f i |
| Other metastatic sites (non-neoplastic)d | |||
| Lung (right or left) | 7 | 28 | 0.629g |
| Both lungs | 6 | 19 | 0.333ª |
| Thyroid gland | 4 | 2 | 0.004g, i |
| Myocardium | 3 | 4 | 0.154g |
| Liver | 7 | 14 | 0.022&, i |
| Spleen | 1 | 5 | 0.578g |
| Pancreas | 5 | 4 | 0.003g, i |
| Small intestine | 3 | 4 | 0.154g |
| Adrenal gland (left or right) | 7 | 12 | 0.010%, i |
| Both adrenal glands | 7 | 9 | 0.002g, i |
| Kidney (left or right) | 3 | 6 | 0.227g |
| Both kidneys | 3 | 4 | 0.154g |
| Bone | 5 | 16 | 0.470g |
| Lymph nodes | 8 | 25 | 0.110g |
| Total clinical course (months) | 0.7-12 (mean, 5.4) | 1-104 (mean, 18.8) [n=29]* | 0.046h, i |
a Forty-seven lung cancers in which tumor size was known. In ten post-therapeutic cases (eight post-operative cases and two post-chemotherapeutic cases), the tumor size before treatment was calculated because the post-chemotherapeutic size was markedly decreased
b Fifty-five extrapulmonary tumors without tumor-to-tumor metastases, in which tumor size was known, and 11 coexisting lung cancers without tumor-to-tumor metastasis, which were recognized at autopsy
” The number of organs involved by metastatic lung cancer other than tumor-to-tumor metastases, such as brain, left lung, right lung, pleura, pericardium, myocardium, liver, spleen, pancreas, small intestine, left adrenal gland, right adrenal gland, left kidney, right kidney, bone, lymph node, and skin
d Metastatic lesions of main lung cancer in each case, excluding tumor-to-tumor metastases (non-neoplastic tissues only)
e Twenty-nine patients died of lung cancer (excluding ten patients who died of other diseases)
៛Unpaired ttest, two-tailed
” Chi-square test with Yates’ correction
h Mann-Whitney U test, two-tailed
Statistically significant
incidence of tumor-to-tumor metastasis to RCC was the highest (83 %). On the other hand, prostatic adenocarcinoma was the most common coexisting tumor (34 % of cases) but did not receive metastatic lung carcinoma in any cases. No tumor-to-tumor metastasis was observed in coexisting colo- rectal adenocarcinoma and gastric carcinoma, although their coexisting incidences (15 % and 11 %) were almost equal to that of RCC. Statistically, there were significant differences in the recipient incidence of tumor-to-tumor metastasis be- tween RCC and these three tumors. Kidney is more proximal from lung than prostatic gland, and the uniquely high inci- dence of tumor-to-tumor metastases to RCC raises the ques- tion of whether anatomical proximity plays a role. However, stomach, in which tumor-to-tumor metastasis was not found
either, is closer to lung than RCC. Interestingly, moreover, 78 % of tumor-to-tumor metastasis was present in the con- tralateral site of the respective donor lung cancer. These findings suggest that lung cancer has a preference for recip- ient tumors in its related tumor-to-tumor metastasis, which would not be associated with the anatomical proximity. In our review of the English literature, 80 cases of tumor-to- tumor metastasis from lung carcinoma were found (Table 5) [1-3, 6-8, 10-17, 19-21]; the most common recipient tumor was RCC (25 %), followed by adrenocortical adenoma (22.5 %), meningioma (20 %), thyroid carcinoma (5 %), and intracranial schwannoma (5 %). Reported cases of tumor-to-tumor metastasis from lung cancer to prostatic car- cinoma, colorectal carcinoma, and gastric carcinoma were
| Recipient tumor | n (%) | References |
|---|---|---|
| Cancer-to-cancer metastasis (n=34, 42.5 %) | ||
| Renal cell carcinoma | 20 (25.0) | [1-3, 6, 8, 11, 14] |
| Thyroid carcinoma | 4 (5.0) | [6, 7, 10, 21] |
| Prostatic carcinoma | 3 (3.8) | [2,6,14] |
| Colorectal carcinoma | 2 (2.5) | [12, 13] |
| Pancreatic carcinoma | 1 (1.3) | [2] |
| Ileal carcinoid | 1 (1.3) | [1] |
| Testicular seminoma | 1 (1.3) | [6] |
| Melanoma of the skin | 1 (1.3) | [6] |
| Renal angiomyoliposarcoma | 1 (1.3) | [2] |
| Cancer-to-benign tumor metastasis (n=46, 57.5 %) | ||
| Adrenocortical adenoma | 18 (22.5) | [6, 15, 17, 20] |
| Meningioma | 16 (20.0) | [6, 16] |
| Schwannoma | 4 (5.0) | [11, 16] |
| Thyroid adenoma | 3 (3.8) | [6, 21] |
| Renal oncocytoma | 2 (2.5) | [4, 6] |
| Pituitary adenoma | 1 (1.3) | [6] |
| Renal angiomyolipoma | 1 (1.3) | [19] |
| Fibrous histiocytoma of soft tissue | 1 (1.3) | [20] |
rare or not found (three cases, two cases, or not found, respectively) [2, 6, 12-14]. These features were consistent with our results, although our study could not detect tumor- to-tumor metastasis to meningioma and schwannoma, pos- sibly due to limited autopsy cases with central nervous system examination. To our knowledge, the current study is the first to describe tumor-to-tumor metastasis between mul- tiple lung cancer and metastasis from lung cancer to pancre- atic endocrine microadenoma.
The reason for tumor-to-tumor metastasis favoring spe- cific tumors is still unknown. To explain this, previous articles have repeatedly discussed the hypervascularity of recipient tumors and the so-called seed and soil theory; the former represents more common receipt of circulating tumor emboli, and the latter supposes a close relationship between metastatic tumor cells (the seed) and a fertile or suitable environment of recipient tumor (the soil) [2-6, 8, 9, 11]. In our study, the close association among tumor-to-tumor me- tastasis and high angiogenic phenotype of recipient RCCs could not be determined. Tumor-to-tumor metastasis was statistically associated with other metastases to the non- neoplastic thyroid gland, liver, pancreas, and adrenal gland, but not with metastases to other blood vessel-rich organs, such as the lungs, kidneys, or spleen. In five (55.6 %) of nine cases of tumor-to-tumor metastasis, metastases to non- neoplastic vasculature-rich tissues (three kidneys, one thy- roid gland, and one adrenal gland) were not found. There- fore, tumor-to-tumor metastasis would not be influenced
only by a vasculature-rich environment. In 14 reported cases of tumor-to-tumor metastasis to RCC from lung cancer of which the histology was known [1-4, 6, 8, 11], lung adeno- carcinoma and small cell carcinoma accounted for 86 % (adenocarcinoma, 57 %; small cell carcinoma, 29 %). Also in our series, donor lung cancers metastasizing to RCC were adenocarcinoma (80 %) and small cell carcinoma (20 %). These findings support the “seed and soil” combination be- tween metastasizing lung adenocarcinoma/small cell carcino- ma and RCC. However, the current study could not reveal the distinct mechanism explaining such a combination.
In the current study, metastatic lung small cell carcinoma was histologically different from thyroid papillary carcinoma and clear RCC. Metastatic lung adenocarcinoma cells show- ing eosinophilic cells with swollen nuclei were dissimilar to clear RCC cells with small nuclei. In addition, there were abrupt borders between tumor-to-tumor metastatic cells and the adjacent recipient tumor cells in most of our cases. These histomorphological features could raise the possibility of tumor-to-tumor metastasis rather than diversity of recipient tu- mor cells. On the other hand, tumor-to-tumor metastases may be sometimes misdiagnosed as diversity or undifferentiation of recipient tumor because of minute size of metastatic deposit or histological mimicry between metastatic cancer cells and recip- ient tumor cells. The current study demonstrated the immuno- histochemical utility for confirming tumor-to-tumor metastasis, which was chiefly attributable to TTF-1+, Napsin A+, and CK7+ in lung adenocarcinoma [29-31] and CK5/6+ and CK7+ in lung squamous cell carcinoma [29, 30]. Recipient RCC was focally CK7+ in two cases, but their staining patterns were different from that in metastatic lung adenocarcinoma, which could be distinguished. However, our observation re- vealed a low incidence of TTF-1+ and Napsin A+ in tumor-to- tumor metastasis because TTF-1- and Napsin A- poorly differ- entiated components metastasized into RCCs in three cases. A decrease in TTF-1 and Napsin A expression is known in poorly differentiated lung adenocarcinoma [30], and such components may carry a risk for the development of tumor-to-tumor metas- tasis. In addition, one recipient RCC was focally Napsin A+, which can occur in 10-43 % of clear cell RCC and in 75-83 % of papillary RCC [32]. Furthermore, non-neoplastic renal tu- bules were CK7+ and Napsin A+ and were focally entrapped in RCC. These unexpected Napsin A+/CK7+ cases should be discriminated from tumor-to-tumor metastasis.
In Case 8, CD10+ metastatic squamous cell carcinoma was distinguished from CD10-negative lung adenocarcinoma. However, CD10 positivity can be found in other histological types of lung cancer [33], and immunostaining for CD10 is not useful for differentiating between types of lung cancer. In Case 7, squamous cell carcinoma cells in both the donor tumor and tumor-to-tumor metastasis showed focal D2-40+, and distant metastases were limited to the lymph nodes and adre- nocortical adenoma. Recently, D2-40 expression of penile
squamous cell carcinoma has been reported as a marker for lymph node metastases [34]. Hence, D2-40+ tumor-to-tumor metastasis indicates lymphatic spread, and D2-40+ lung can- cer may predict the development of lymphatic spreading.
In conclusion, lung cancer with more aggressive behavior can infrequently cause tumor-to-tumor metastasis to coexisting tumors. The possible diagnosis of undifferentiated recipient tumor cells should be carefully ruled out in assessing tumor- to-tumor metastasis histologically and immunohistochemically.
Acknowledgments The authors thank Kenji Okada for excellent technical assistance and Daniel Mrozek for editing the manuscript.
Conflict of interest statement We declare that we have no conflict of interest.
References
1. Ottosson L, Berge T (1968) Metastasis from carcinoma to carcino- ma. Acta Pathol Microbiol Scand 73:481-488
2. Campbell LV Jr, Gilbert E, Chamberlain CR Jr, Watne AL (1968) Metastases of cancer to cancer. Cancer 22:635-643
3. Ichijima K, Yamabe H, Kobashi Y, Iwata T (1980) Metastasis of cancer to cancer. Acta Pathol Jpn 30:293-300
4. Sella A, Ro JY (1987) Renal cell cancer: best recipient of tumor-to- tumor metastasis. Urology 30:35-38
5. Singh EO, Benson RC, Wold LE (1984) Cancer-to-cancer metas- tasis. J Urol 132:340-342
6. Petraki C, Vaslamatzis M, Argyrakos T et al (2003) Tumor to tumor metastasis. Report of two cases and review of the literature. Int J Surg Pathol 11:127-135
7. Mori K, Kitazawa R, Kondo T, Kitazawa S (2008) Lung adenocar- cinoma with micropapillary component presenting with metastatic scrotum tumor and cancer-to-cancer metastasis: a case report. Cases J 1:162
8. Sawada T, Takahashi H, Hasatani K et al (2009) Tumor-to-tumor metastasis: report of an autopsy case of lung adenocarcinoma metastasizing to renal cell carcinoma. Int Med 48:1525-1529
9. Sakai Y, Kanomata N, Itami H, Kajimoto K, Sakuma T, Ohbayashi C (2010) Signet-ring cell carcinoma of the stomach metastasizing to renal cell carcinoma: a case report and review of the literature. Kobe J Med Sci 55:E122-131
10. Hashimoto K, Yamamoto H, Nakano T et al (2011) Tumor-to-tumor metastasis: lung adenocarcinoma metastasizing to a follicular var- iant of papillary thyroid carcinoma. Pathol Int 61:435-441
11. Aggarwal N, Amin RM, Chung D, Parwani AV (2012) Tumor-to- tumor metastasis: case report of a pulmonary adenocarcinoma metastatic to a clear cell renal cell carcinoma. Pathol Res Pract 208:50-52
12. Jackson JG, Symmers WSC (1951) Coexistence at one site of two neoplasms, one of local origin and one metastatic. Brit J Cancer 5:38-44
13. Fried BM (1958) Primary multiple cancers. Arch Surg 77:730-741
14. Maloney JD, Immergut M (1967) The kidney as host in cancer-to- cancer metastasis. J Urol 98:657-660
15. Onuigbo WIB (1963) Lung cancer metastasis to adrenal cortical adenomas. J Pathol Bacteriol 86:541-543
16. Chambers PW, Davis RL, Buck FS (1980) Arch Pathol Lab Med 104:350-354
17. Moriya T, Manabe T, Yamashita K, Arita S (1988) Lung cancer metastasis to adrenocortical adenomas. A chance occurrence or a predilected phenomenon? Arch Pathol Lab Med 112:286-289
18. Honma K, Hara K, Sawai T (1989) Tumour-to-tumour metastasis. A report of two unusual autopsy cases. Virchows Archiv A Pathol Anat 416:153-157
19. Ricketts R, Tamboli P, Czerniak B, Guo CC (2008) Tumor-to- tumor metastasis: report of 2 cases of metastatic carcinoma to angiomyolipoma of the kidney. Arch Pathol Lab Med 132:1016- 1020
20. Martin JT, Alkhoury F, Helton S, Fiedler P, Sakharova O, Yood S (2009) Metastatic adenocarcinoma within a functioning adrenal adenoma: a case report. Cases J 2:7965
21. Stevens TM, Richards AT, Bewtra C, Sharma P (2011) Tumors metastatic to thyroid neoplasms: a case report and review of the literature. Patholog Res Int 31:238693
22. Moody P, Murtagh K, Piduru S, Brem S, Murtagh R, Rojiani AM (2012) Tumor-to-tumor metastasis: pathology and neuroimaging considerations. Int J Clin Exp Pathol 5:367-373
23. Travis WD, Brambilla E, Müller-Hermelink HK, Harris CC (2004) World Health Organization classification of tumours. Pathology and genetics of tumours of the lung, pleura, thymus and heart, 3rd edn. IARC Press, Lyon
24. Eble JN, Sauter G, Epstein JI, Sesterhenn IA (2004) World Health Organization classification of tumours. Pathology and genetics of tumours of the urinary system and male genital organs, 3rd edn. IARC Press, Lyon
25. Hruban RH, Pitman MB, Klimstra DS (2007) Tumors of the pan- creas. In: Silverberg SG, Sobin LH (eds) AFIP atlas of tumor pathology, 4th series, fascicle 6. American Registry of Pathology, Washington
26. Lack EE (2007) Tumors of the adrenal glands and extraadrenal paraganglia. In: Silverberg SG, Sobin LH (eds) AFIP atlas of tumor pathology, 4th series, fascicle 8. American Registry of Pathology, Washington
27. Lloyd RV, Douglas BR, Young WF Jr (2002) Endocrine diseases. In: King DW, Sobin L, Stocker JT, Wagner B (eds) Atlas of nontumor pathology, 1st series, fascicle 1. American Registry of Pathology, Washington
28. Edeline J, Mottier S, Vigneau C et al (2012) Description of 2 angiogenic phenotypes in clear cell renal cell carcinoma. Hum Pathol 43:1982-1990
29. Travis WD, Brambilla E, Noguchi M et al (2011) International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society international multidisciplin- ary classification of lung adenocarcinoma. J Thorac Oncol 6:244- 285
30. Whithaus K, Fukuoka J, Prihoda TJ, Jagirdar J (2012) Evaluation of Napsin A, cytokeratin 5/6, p63, and thyroid transcription factor 1 in adenocarcinoma versus squamous cell carcinoma of the lung. Arch Pathol Lab Med 36:155-162
31. Turner BM, Cagle PT, Sainz IM, Fukuoka J, Shen SS, Jagirdar J (2012) Napsin A, a new marker for lung adenocarcinoma, is com- plementary and more sensitive and specific than thyroid transcrip- tion factor in the differential diagnosis of primary pulmonary car- cinoma: evaluation of 1674 cases by tissue microarray. Arch Pathol Lab Med 136:163-171
32. Ordóñez NG (2012) Napsin A expression in lung and kidney neoplasia: a review and update. Adv Anat Pathol 19:66-73
33. Kristiansen G, Schlüns K, Yongwei Y, Daniel M, Petersen I (2002) CD10 expression in non-small cell lung cancer. Anal Cell Pathol 24:41-46
34. Minardi D, d’Anzeo G, Lucarini G et al (2011) D2-40 immunore- activity in penile squamous cell carcinoma: a marker of aggressive- ness. Hum Pathol 42:1596-1602