Context

Loose tumor cells and tumor cell clusters can be recognized in the lumen of intratumoral pulmonary arteries of resected non–small cell lung cancer specimens. It is unclear whether these should be considered tumor-emboli, and as such could predict a worsened prognosis.

Objective

To investigate the nature and prognostic impact of pulmonary artery intraluminal tumor cells.

Design

This multicenter study involved an exploratory pilot study and a validation study from 3 institutions. For the exploratory pilot study, a retrospective pulmonary resection cohort of primary adenocarcinomas, diagnosed between November 2007 and November 2010, were scored for the presence of tumor cells, as well as potentially other cells in the intravascular spaces, using hematoxylin-eosin and cytokeratin 7 (CK7) stains. In the validation part, 2 retrospective cohorts of resected pulmonary adenocarcinomas, between January 2011 and December 2016, were included. Recurrence-free survival (RFS) and overall survival (OS) data were collected.

Results

In the pilot study, CK7+ intravascular cells, mainly tumor cells, were present in 23 of 33 patients (69.7%). The 5-year OS for patients with intravascular tumor cells was 61%, compared with 40% for patients without intravascular tumor cells (P = .19). In the validation study, CK7+ intravascular tumor cells were present in 41 of 70 patients (58.6%). The 5-year RFS for patients with intravascular tumor cells was 80.0%, compared with 80.6% in patients without intravascular tumor cells (P = .52). The 5-year OS rates were, respectively, 82.8% and 71.6% (P = .16).

Conclusions

Loose tumor cells in pulmonary arterial lumina were found in most non–small cell lung cancer resection specimens and were not associated with a worse RFS or OS. Therefore, most probably they represent an artifact.

Lung cancer is associated with a high incidence of pulmonary embolism, caused by thrombotic emboli, as a paraneoplastic phenomenon, but also caused by real tumor-emboli.1  In general, pulmonary tumor embolism can be categorized by size, into large, proximal emboli and smaller emboli in the microvasculature. Tumor-emboli occurring in large branches of the pulmonary artery are associated with an acute deterioration of clinical performance. Pulmonary tumor embolism localized in smaller branches of the pulmonary arteries is described by 2 different entities: pulmonary tumor microembolism and pulmonary tumor thrombotic microangiopathy.24  These 2 conditions are likely to represent a spectrum of the same disease and may be associated with pulmonary hypertension.2,5 

Although pulmonary tumor-emboli are a known specific entity and can be diagnosed in tumor biopsies or tumor resection specimens, it is unclear whether all intravascular tumor cells in pathology slides can be considered tumor-emboli. In our experience, a substantial incidence of tumor cells can be recognized in the lumen of intratumoral pulmonary arteries in many resected non–small cell lung cancer (NSCLC) specimens, albeit with a morphologic aspect, distinct from regular tumor-emboli. We hypothesized that these “common” loose intravascular tumor cells may be caused by technical artifacts through displacement in the preparation process of the slides. Therefore, we aimed to investigate the prognostic impact of these intravascular tumor cells and to substantiate the claim that these intravascular tumor cells may be caused by iatrogenic artifacts.

Design

This study involved an exploratory pilot study and a validation study. In a cohort of patients with adenocarcinoma of the lung resected between November 2007 and November 2010, we observed the presence of intravascular tumor cells in a high number of cases. We conducted an exploratory pilot study on a subset of this cohort (n = 33) to confirm our findings. The selection of participants was based on the availability of slides and paraffin blocks for the additional staining of resections performed at the Vrije Universiteit Medical Center in Amsterdam, the Netherlands. Additionally, participants were included in the study based on the availability of clinical follow-up. Hematoxylin-eosin (H&E)–stained slides were retrieved from the archive, and for each resected tumor, the 1 or 2 most representative tumor slides were selected. Cytokeratin 7 (CK7) staining was performed as described below on the sections of the same paraffin block(s). The presence of loose tumor cells and other cells in pulmonary arteries was scored in H&E- and CK7-stained slides by 2 experienced pulmonary pathologists. Follow-up data were retrieved from patient files.

Because the exploratory study contained a relatively small number of patients and consisted of a not uniformly defined set of patients, we expanded the study with 2 retrospective separate cohorts in search of validation of the initial findings: patients with resected pulmonary adenocarcinoma diagnosed between January 1, 2011, and December 31, 2016 in (1) Onze Lieve Vrouwe Gasthuis (OLVG) Hospital, Amsterdam, the Netherlands, and (2) San Raffaele Scientific Institute, Milan, Italy.

The inclusion criteria for both of these cohorts were different from the exploratory cohort and included resection specimens with a primary pulmonary adenocarcinoma of a pathologic tumor diameter 3 cm or less and available follow-up information. The exclusion criteria were the presence of nodal or hematogenous metastases at the time of resection, treatment with neoadjuvant chemotherapy, multiple nodules in the same or other lobes, synchronous previous lung carcinoma, and invasive mucinous adenocarcinoma or other special type patterns (intestinal and fetal adenocarcinoma). Freshly cut sections of the selected paraffin blocks from both institutes were stained for H&E, elastica von Gieson (EvG or elastin), and CK7 immunohistochemistry at the department of pathology of the Vrije Universiteit Medical Center. Both institutes used neutral buffered formaldehyde for the fixation of their resection specimens. It was estimated that a total of 70 to 80 cases would be sufficient to validate the initial finding. Therefore, after determining suitable cases in both institutes, a minimum of 40 cases in each were randomly selected.

Age, sex, total tumor size, tumor location, pathologic stage adjusted to the 8th edition of the Union for International Cancer Control/American Joint Committee on Cancer (UICC/AJCC) TNM classification system for NSCLC,6  time to recurrence, death, and cause of death were determined by retrospective chart review.

In the OLVG in Amsterdam, 218 cases of resected adenocarcinoma in the lung between January 1, 2011, and December 31, 2016 were retrieved from the pathology archives. A total of 158 cases were excluded because they did not meet inclusion criteria (47 with nodal metastases, 26 with metastases from other organ [mainly colorectal carcinoma], 20 with multiple nodules, 2 subjected to neoadjuvant therapy, 47 with a diameter >3 cm, 4 with previous or synchronous lung cancer, 12 mucinous or other type). The remaining 60 cases were included for random selection.

A total of 365 cases of resected primary lung adenocarcinoma between January 1, 2011, and December 31, 2016 were retrieved from the San Raffaele archive in Milan. A total of 223 were firstly excluded because they did not meet the inclusion criteria of the study (122 with nodal metastases, 12 without lymph node histologic examination, 6 subjected to neoadjuvant treatment, 8 with multiple lung nodules at the time of resection, 18 mucinous and 1 enteric adenocarcinoma, 18 mixed mucinous and nonmucinous adenocarcinomas, and 38 cases with tumoral diameter >3cm). Of the remaining 142 cases, 48 had short or no accessible follow-up information, for 2 cases cause of death was unknown, and for 17 cases material was not available or insufficient. The remaining 75 cases from the San Raffaele Scientific Institute were included for randomization.

After randomization, 80 cases were included. Histologic review of these cases was performed by 3 pathologists. For each case, 1 representative tumor block was selected and digitized. The digital images were uploaded on the collaborative platform for online teaching, training, and quality assurance in pathology, Pathogate (https://pathogate.net).

After technical review of the cases, 10 were excluded because of insufficient quality of the slide or out-of-focus areas after scanning. Therefore, 70 cases, 35 from each institute, were analyzed. Because in the original pathology reports subtyping of the adenocarcinomas was rarely reported, this was done by 2 pathologists in the categories of adenocarcinoma in situ (AIS), lepidic predominant adenocarcinoma, low-grade adenocarcinoma (acinar, papillary) with or without more than 10% high-grade component, and high-grade adenocarcinoma (micropapillary, solid). Cases with discrepancies (dominant pattern, presence of high-grade component) were reviewed, reaching consensus.

Cytokeratin 7

The immunohistochemical stain for CK7 (clone OVTL12/30, catalog No. M701801, Agilent/Dako, Glostrup, Denmark) was performed in a Roche/Ventana benchmark Ultra (Roche, Basel Switzerland) with 3-μm tissue slides mounted on TOMO-glass slides (Roche). Antigen retrieval was applied with high pH-buffer CC1 (32 minutes at 100°C), and the antibody was diluted 1:100 (incubated for 32 minutes at 36°C) and detected with the Optiview DAB kit under standard conditions. Sections were dehydrated with ethanol 100%, cleared with xylene, and coverslipped with Tissue-Tek coverslip film on the Sakura coverslipper (Sakura Finetek Europe BV, Alphen aan de Rijn, the Netherlands). Internal positive controls were pneumocytes and respiratory epithelial cells. Internal negative controls were, for example, smooth muscle cells in vessel walls and bronchioli. Occasional CK7+ endothelial cells were not scored as luminal positive cells.

Statistical Analysis

The presence of tumor cells within the lumen of pulmonary arteries was evaluated using H&E and CK7. Continuous variables were described by mean and SD, categoric variables by frequency and percentage. For recurrence-free survival (RFS) and overall survival (OS) analysis Kaplan-Meier curves were estimated and compared using the log-rank test. Statistical analysis was performed in SPSS version 26 (IBM Corp, Armonk, New York). P values <.05 were considered statistically significant.

Exploratory Pilot Study

Clinicopathologic Characteristics

In the pilot study from a total of 33 patients, 74 histologic sections of 40 tumors were analyzed. The clinicopathologic characteristics of the patients are presented in Table 1. A total of 7 patients had 2 tumors; in 4 of them they were located in 2 separate wedge excisions, 1 patient had them in a lobectomy and a wedge excision, and 2 patients had a double tumor in 1 lobe. The histologic diagnosis was invasive carcinoma for most tumors, except for 3 with a diagnosis of AIS, of which 1 was in combination with a separate invasive carcinoma.

Table 1

Clinicopathologic Variables of Pilot Study Patients

Clinicopathologic Variables of Pilot Study Patients
Clinicopathologic Variables of Pilot Study Patients

Microscopic Findings

All the histologic sections contained alveolar walls covered with tumor cells located near branches of the pulmonary artery. Frequently, in the routine H&E stain, a few clusters of tumor cells, focal remaining erythrocytes, and rare alveolar macrophages were observed. The cytonuclear appearance of the tumor cells in the arterial lumen was similar to the adjacent alveolar wall lining tumor cells.

Immunohistochemical staining with CK7, which is reactive to the cytoplasm of most pulmonary adenocarcinomas, revealed immunohistochemistry-positive single tumor cells or small clusters in the “looking like air” partially empty lumen of the larger pulmonary arteries, without fibrin (Figure 1). This finding was present in 41 of 74 histologic sections (55%), 26 of 40 tumors (65%), and 23 of 33 patients (70%). In the smallest branches of the pulmonary arteries, tumor cells were not discerned. Intravascular tumor cells were found in 2 of the 3 tumors without invasion (AIS; 67%) and in 19 of the 37 tumors with invasion (51%; P = .61). No tumor cells were observed in venous or lymphatic vascular structures.

Figure 1

A, Example of partially collapsed peripheral lung in which alveoli are covered with epithelial tumor cells. Classified as noninvasive, adenocarcinoma in situ. Centrally, a pulmonary artery with clusters and isolated tumor cells. B, Same area as A. C, Example of another pulmonary artery with clusters of tumor cells (blue arrowheads), alveolar macrophages (blue arrows), and some erythrocytes. D, Same artery as C (hematoxylin-eosin, original magnifications ×200 [A] and ×400 [C]; cytokeratin 7 immunohistochemical stain, original magnifications ×200 [B] and ×400 [D]).

Figure 1

A, Example of partially collapsed peripheral lung in which alveoli are covered with epithelial tumor cells. Classified as noninvasive, adenocarcinoma in situ. Centrally, a pulmonary artery with clusters and isolated tumor cells. B, Same area as A. C, Example of another pulmonary artery with clusters of tumor cells (blue arrowheads), alveolar macrophages (blue arrows), and some erythrocytes. D, Same artery as C (hematoxylin-eosin, original magnifications ×200 [A] and ×400 [C]; cytokeratin 7 immunohistochemical stain, original magnifications ×200 [B] and ×400 [D]).

Close modal

Validation Study

Clinicopathologic Characteristics

In the validation study a total of 70 tumor slides of 70 patients were analyzed. The clinicopathologic characteristics of the patients are presented in Table 2. The histologic diagnosis was invasive carcinoma for 60 tumors. A total of 10 cases were classified as noninvasive/AIS (14.3%). Although there were significant differences between the 2 cohorts with respect to age (Milan mean, 67.2 years; OLVG, 61.5 years; P = .009); sex (Milan, 25 male and 10 female patients; OLVG, 12 male and 23female patients; P = .002), and follow-up time (Milan mean, 61 months; OLVG, 81 months; P = .005), the RFS time in cases with a relapse between the 2 cohorts did not differ (Milan mean, 26.7 months [n = 7]; OLVG mean, 34.9months [n = 8]; P = .49), nor did the dominant tumor type (P = .66). Therefore, for the analysis of the presence of intravascular cells, the 2 cohorts were combined.

Table 2

Clinicopathologic Variables in the 2 Validation Cohorts

Clinicopathologic Variables in the 2 Validation Cohorts
Clinicopathologic Variables in the 2 Validation Cohorts

Microscopic Findings

The presence of intravascular cells in the combined cohorts for both invasive and noninvasive (AIS) cases is shown in Table 3. In a total of 41 of 70 patients (58.6%), intravascular tumor cells, tumor cell clusters, and/or macrophages were found. These were also found in 8 of the 10 tumors without invasion (AIS; 80.0%) and in 33 of the 60 invasive tumors (55.0%; P = .18). No tumor cells were observed, neither in venous nor in lymphatic vascular structures.

Table 3

Presence of Intravascular Cells (Cytokeratin 7–Positive [CK7+] and/or Macrophages) in Both Validation Cohorts, in Invasive and Adenocarcinoma In Situ (AIS) Cases

Presence of Intravascular Cells (Cytokeratin 7–Positive [CK7+] and/or Macrophages) in Both Validation Cohorts, in Invasive and Adenocarcinoma In Situ (AIS) Cases
Presence of Intravascular Cells (Cytokeratin 7–Positive [CK7+] and/or Macrophages) in Both Validation Cohorts, in Invasive and Adenocarcinoma In Situ (AIS) Cases

Survival Outcomes

The median follow-up time in the pilot study was 58 months (range, 1–112 months). The OS for intra-arterial tumor cells is shown in the Supplemental Figure (see supplemental digital content at https://meridian.allenpress.com/aplm in the May 2024 table of contents). In the exploration cohort, the mean 5-year OS rate for the patients with intravascular tumor cells was 61%, compared with 40% for the patients without intravascular tumor cells (P = .19).

In the cohorts of the validation study, the median follow-up time was 61 months (range, 11–113 months) for the Milan cases and 81 months (range, 28–132 months) for the OLVG cases.

In the Milan cohort, the 22 cases with intravascular tumor cells had a 5-year OS of 85.6%, compared with 41.5% for the13 cases without intravascular tumor cells (P = .002). In the OLVG cases, the 5-year OS for the 19 cases with and the 16cases without intravascular tumor cells was 78.9% and 93.8%, respectively (P = .35). In both cohorts combined, the 5-year OS for cases with intravascular tumor cells was 82.5% versus 71.6% for cases without (P = .16; Figure 2). The 5-year RFS data for both cohorts showed no significant differences, neither for the separate cohorts nor for the combined one (80.0% with and 80.9% without intravascular tumor cells; P = .52).

Figure 2

The Kaplan-Meier curves for overall survival in the validation cohorts are shown for cytokeratin 7–positive tumor cells and/or macrophages in lumen of the pulmonary arteries (P = .16).

Figure 2

The Kaplan-Meier curves for overall survival in the validation cohorts are shown for cytokeratin 7–positive tumor cells and/or macrophages in lumen of the pulmonary arteries (P = .16).

Close modal

In this multicenter retrospective study, we conducted an exploratory pilot study and a validation study in 2 separate cohorts. Our findings showed that intravascular tumor cells were present in a surprisingly high number of patients with resected early-stage NSCLC. In the pilot study, individual tumor cells, small clusters of tumor cells, and/or macrophages were found in 70% of the 33 patients and in 65% of the 40 resected tumors. This finding was confirmed in the validation cohort, where intravascular tumor cells were found in 41 of 70 cases (58.6%). Interestingly, the presence of arterial tumor cells was associated with a seemingly better survival outcome in the pilot study and in 1 of the validation cohorts. However, this finding was contradictory to the expected biologic phenomenon. We also observed the presence of intravascular tumor cells in cases without invasive carcinoma (AIS) in both the pilot study and the validation cohort. The morphologic diagnosis of tumor cells within pulmonary arteries, as we found in our study, can be distinguished from 5 other pathologic entities, as discussed below:

First, direct invasive growth through the arterial wall of tumor cells. In none of the cases in our study was direct tumor invasion in the adventitia, media, or intima layer observed, either in the H&E stain or in the CK7 stain.

Second, tumor cells in smaller branches with or without occlusive fibrointimal remodeling in small pulmonary arteries, which is usually a sign of metastases from other organs in the form of pulmonary tumor microembolism and pulmonary tumor thrombotic microangiopathy. In the pilot study, only 4 of the patients had a history of cancer, of whom only 1 had a colonic carcinoma (Table 1), which is a tumor type that is infrequently CK7+. In the validation cohort, the presence of a previous malignancy was an exclusion criterion. Additionally, none of the cases with luminal tumor cells showed intimal fibro-remodeling.

Third, the morphologic features in our study differ from tumor emboli in larger vessels from metastases of other organs in several ways. (1) In our cases, the alveolar walls surrounding the pulmonary artery were covered with tumor cells extending from the adjacent tumor, with a similar cytonuclear appearance as in the lumen of the artery. In metastatic emboli, the adjacent alveolar walls do not contain tumor, and the cytonuclear details of the metastases have a similarity with the primary tumor, which is usually different from a primary lung adenocarcinoma. (2) The amount of tumor cells was small (ie, isolated tumor cells or small clusters) and without surrounding fibrin or other blood components, as discussed above in point 2. In cases of metastasis, the lumen of the larger pulmonary artery is usually slightly extended by the filling of the tumor embolus, whereas in our study, a large part of the occasionally indented lumen was “empty.” In patients with clinical suspicion of pulmonary emboli on preoperative imaging, the differential diagnosis includes sarcoma of the pulmonary artery,7  which was not present in our patients.

Furthermore, hydrophilic coats on medical interventional devices may dissociate from the device surface during endovascular manipulation and give rise to hydrophilic polymer embolism.810  These are morphologically characterized by intraluminal foreign body material surrounded by thrombosis, with or without inflammatory response, and are usually located in smaller arteries. Foreign body material was not present in our study.

A last distinction can be made from circulating tumor cells that may appear similar, but these cells in the pulmonary artery of the resection specimen should then be derived from a primary source elsewhere in the body and flow via the heart to the lung. Morphologically, the circulating tumor cells should be surrounded by blood components, whereas in our study, the arterial lumen looked “empty,” with occasional tumor cells without a lot of blood components. Moreover, in most patients, a primary tumor of other origin than the lungs was not present at the time of resection and during follow-up (median time >5 years). In addition, circulating tumor cells are rare in the bloodstream. For analysis of circulating tumor cells, enrichment and isolation methods have been developed,11  whereas in our study there was a relatively high frequency in histologic sections of both the pilot study and the validation cohorts.

In the lung, a pilot study of blood sampling from tumor-draining pulmonary veins at the time of tumor resection was previously performed, with the hypothesis that patients with detectable circulating tumor cells in the pulmonary vein would have a higher risk for disease recurrence.12  Although the concept of venous drainage from a pulmonary carcinoma is plausible, involvement of arterial tumor cells in this context is not.

The morphologic and clinical diagnostic aspects discussed above support the idea that our findings are different from clinical entities described so far, except for the report by Pechet and colleagues,13  who described arterial invasion in stage I NSCLC. Pechet and colleagues13  defined the presence of tumor within the lumen of a muscular vessel with a clearly defined internal or external elastic lamina in at least 2anatomically distinct vessels as a positive score for arterial invasion. They found that patients without arterial invasion had significantly better survival than those with arterial invasion (73% versus 38%; P < .001). However, their study did not include histopathologic images for comparison. In our study, we specifically searched for tumor cells using CK7 immunohistochemical staining in at least 1 pulmonary artery, whereas the study by Pechet et al13  required at least 2 vessels (ie, 39% of the patients). Complete data on recurrence in their study were available for 64 patients, of whom 16 had arterial invasion. Remarkably, 6 of the 16 were alive at 5 years (2 with and 4 without recurrence). Although the selection criteria and follow-up in the study by Pechet et al13  were different from ours, this does not exclude the possibility of an iatrogenic artifact in their study.

Several findings from our study are remarkable. One is the high incidence of intratumorally intravascular tumor cells compared with previous literature (25%–50%).14,15  Another is that we only found intra-arterial vascular tumor cells and no intravenous or lymphovascular invasive tumor cells. A third notable finding is that we also found intra-arterial tumor cells in 3 of the 33 cases in the pilot study and also in8 of the 10 cases in the validation cohorts that were classified as AIS and also on revision showed no other criteria of invasiveness. This supports the idea that at least some of the intravascular tumor cells we found were not a true representation of vascular invasion, but most likely an artifact. We believe that isolated tumor cells in pulmonary arteries with adjacent abundant tumor cells are easily overlooked in routine H&E staining, but are more clearly recognized in CK7 immunohistochemical staining. This raised the question of whether our observation is a clinically relevant reality or an artifact. Based on the following arguments, we strongly favor the idea that our findings are the result of an iatrogenic artifact during gross handling rather than a clinically relevant reality: (1) the lack of morphologic similarities with known clinical entities; (2) the lack of clinical signs of metastases in our patients (also in the 5-year follow-up period); (3) the association with a counterintuitive, better prognosis for patients with intra-arterial tumor cells than those without; (4) the presence of occasional “alveolar macrophages” in the lumen of pulmonary arteries adjacent to tumor cells; (5) the detection in resection specimen; and (6) its presence in cases of AIS.

Moreover, for loose tumor cells in alveolar spaces, criteria have been mentioned that might be associated with artifacts, such as jagged edges.16  These characteristics were not present in the tumor cells located in the pulmonary artery. Of note, rounded clusters/isolated tumor cells can also be an artifact as well. A possible explanation may be that during cutting of the resection specimen a knife is used. As during the surgical collapse, the blood and lymph volume in the vessels is reduced, and the vessel walls are forced to a collapsed state and remain in this way by the surgical ligation until the gross examination in the pathology laboratory. During cutting of the resection specimen with the gross knife, the pulmonary artery wall is interrupted and the shape of the wall may change somewhat to alleviate the local compressed situation, leading to at least a slight enlargement of the almost empty lumen. In parallel, the front of the knife blade, with a width of the knife around 10 to 15 times bigger than that of tumor cells,17  induces a tremendous force, with mechanical displacement of tumor (and normal) cells at the cutting edge. Apparently, the “empty” arterial vascular lumen forms during the decompression moment a niche that is fillable, probably due to the still flexible structure of the pulmonary arterial wall compared to veins and lymph vessels. The example of a peripheral pulmonary tissue fragment in the lumen of a pulmonary artery strongly supports this explanation (see figure 5, B through D, in Thunnissen et al17).

We do not have the intention to call the isolated tumor cells in the pulmonary arteries of resection specimens “spread through vascular space,” but rather “spread through a knife surface” as described in a study as 1 of the 4 ex vivo artifacts in pulmonary resection specimen.17  The possibility of artifacts has been described in several organs as well, such as in thyroid resections, in which the artifactual displacement of adenoma and nonmalignant epithelial cells has been reported.18  But also in resection specimens for prostate cancer it could be a pitfall.19  In colon cancer, extramural venous invasion was associated with a worse prognosis,20,21  in contrast to the presence of intramural venous invasion, which revealed a similar prognosis when compared to no venous invasion.20  In laparoscopic abdominal hysterectomy specimens, vascular pseudoinvasion of malignant and benign cells was reported.22,23  In one of these studies a suggestion was made that pathologists may be generating postoperative pseudoinvasion by mechanically transporting tumor into vascular spaces during the grossing process.23 

Recently, Metovic and colleagues24  published an interesting article stating that “STAS is without any doubt, no artifact that can be induced by gross specimen handling.” The authors are to be complimented for their carefully designed methodology. Nevertheless, the effect of cutting (including the manual pressure) combined with the tendency of dissociation of tumor cells remains open for discussion.25 

In short, there are no defined morphologic criteria to distinguish true invasive tumor cells or small clusters from mechanically dragged tumor particles. Taking the morphologic, clinical, and literature information, including from other organs, into account, we interpret the presence of tumor cells floating in the arterial lumen detached from the vessel wall and without associated thrombus in resection specimens of primary pulmonary adenocarcinomas in the presented cases as an artifact, although true vascular invasion may occur.

The study's limitations include its retrospective design and the use of available tumor blocks, which may not have included the entire tumor for histopathologic examination, even for small tumors. Additionally, because all cases had a low stage and nonmucinous adenocarcinomas, resection alone may lead to a cure. The study did not conduct multivariate analysis to account for potential clinical confounders, which requires at least 20 events for RFS and OS. However, our study did not fulfill the minimum requirements, with only 14 RFS events and 18 OS events. Further larger prospective studies also examining higher stages are warranted.

In most NSCLC resection specimens, CK7+ tumor cells were detected in the lumen of pulmonary arteries. The presence of these tumor cells was not associated with worse OS. The arguments discussed are strongly in favor of an iatrogenic artifact, probably created during gross handling of the resection specimen. It is important to be aware of this artifact as a frequent cause of intravascular tumor cells observed in resection specimens.

1.
Cha
SI,
Shin
KM,
Lim
JK,
et al.
Pulmonary embolism concurrent with lung cancer and central emboli predict mortality in patients with lung cancer and pulmonary embolism
.
J Thorac Dis
.
2018
;
10
(
1
):
262
272
.
2.
Price
LC,
Seckl
MJ,
Dorfmüller
P,
Wort
SJ.
Tumoral pulmonary hypertension
.
Eur Respir Rev
.
2019
;
28
(
151
):
1
15
.
3.
Miyano
S,
Izumi
S,
Takeda
Y,
et al.
Pulmonary tumor thrombotic microangiopathy
.
J Clin Oncol
.
2007
;
25
(
5
):
597
599
.
4.
Price
LC,
Wells
AU,
Wort
SJ.
Pulmonary tumour thrombotic microangiopathy
.
Curr Opin Pulm Med
.
2016
;
22
(
5
):
421
428
.
5.
Mainardi
AS,
Trow
T.
Tumor emboli: a rare cause of acute pulmonary hypertension
.
Am J Med
.
2017
;
130
(
4
):
e137
e139
.
6.
Brierley
JD,
Gospodarowicz
M,
Wittekind
C.
TNM Classification of Malignant Tumors. 8th ed
.
Hoboken, NJ
:
John Wiley & Sons;
2017
.
7.
Al-Mehisen
R,
Al-Halees
Z,
Alnemri
K,
Al-Hemayed
W,
Al-Mohaissen
M.
Primary pulmonary artery sarcoma: a rare and overlooked differential diagnosis of pulmonary embolism: clues to diagnosis
.
Int J Surg Case Rep
.
2019
;
65
:
15
19
.
8.
Mehta
RI,
Mehta
RI,
Choi
JM,
Mukherjee
A,
Castellani
RJ.
Hydrophilic polymer embolism and associated vasculopathy of the lung: prevalence in a retrospective autopsy study
.
Hum Pathol
.
2015
;
46
(
2
):
191
201
.
9.
Miyaoka
M,
Hatanaka
K,
Uekusa
T,
Nakamura
N.
Clinicopathological features of hydrophilic polymer emboli in Japanese autopsy cases
.
Apmis
.
2018
;
126
(
11
):
838
841
.
10.
Hickey
TB,
Honig
A,
Ostry
AJ,
et al.
Iatrogenic embolization following cardiac intervention: postmortem analysis of 110 cases
.
Cardiovasc Pathol
.
2019
;
40
:
12
18
.
11.
Maly
V,
Maly
O,
Kolostova
K,
Bobek
V.
Circulating tumor cells in diagnosis and treatment of lung cancer
.
In Vivo (Brooklyn)
.
2019
;
33
(
4
):
1027
1037
.
12.
Crosbie
PAJ,
Shah
R,
Krysiak
P,
et al.
Circulating tumor cells detected in the tumor-draining pulmonary vein are associated with disease recurrence after surgical resection of NSCLC
.
J Thorac Oncol
.
2016
;
11
(
10
):
1793
1797
.
13.
Pechet
TTV,
Carr
SR,
Collins
JE,
Cohn
HE,
Farber
JL.
Arterial invasion predicts early mortality in stage I non-small cell lung cancer
.
Ann Thorac Surg
.
2004
;
78
(
5
):
1748
1753
.
14.
Al-Alao
BS,
Gately
K,
Nicholson
S,
McGovern
E,
Young
VK,
O'Byrne
KJ.
Prognostic impact of vascular and lymphovascular invasion in early lung cancer
.
Asian Cardiovasc Thorac Ann
.
2014
;
22
(
1
):
55
64
.
15.
Maeda
R,
Yoshida
J,
Ishii
G,
Hishida
T,
Nishimura
M,
Nagai
K.
Prognostic impact of intratumoral vascular invasion in non-small cell lung cancer patients
.
Thorax
.
2010
;
65
(
12
):
1092
1098
.
16.
Kadota
K,
Nitadori
J,
Sima
CS,
et al.
Tumor spread through air spaces is an important pattern of invasion and impacts the frequency and location of recurrences after limited resection for small stage I lung adenocarcinomas
.
JThorac Oncol
.
2015
;
10
(
5
):
806
814
.
17.
Thunnissen
E,
Blaauwgeers
HJLG,
De Cuba
EMV,
Yick
CY,
Flieder
DB.
Ex vivo artifacts and histopathologic pitfalls in the lung
.
Arch Pathol Lab Med
.
2016
;
140
(
3
):
212
220
.
18.
Mete
O,
Asa
SL.
Pathological definition and clinical significance of vascular invasion in thyroid carcinomas of follicular epithelial derivation
.
Mod Pathol
.
2011
;
24
(
12
):
1545
1552
.
19.
Kryvenko
ON,
Epstein
JI.
Histologic criteria and pitfalls in the diagnosis of lymphovascular invasion in radical prostatectomy specimens
.
Am J Surg Pathol
.
2012
;
36
(
12
):
1865
1873
.
20.
Leijssen
LGJ,
Dinaux
AM,
Amri
R,
et al.
Impact of intramural and extramural vascular invasion on stage II-III colon cancer outcomes
.
J Surg Oncol
.
2019
;
119
(
6
):
749
757
.
21.
Van Eeghen
EE,
Flens
MJ,
Mulder
MMR,
Loffeld
RJLF.
Extramural venous invasion as prognostic factor of recurrence in stage 1 and 2 colon cancer
.
Gastroenterol Res Pract
.
2017
;
2017
:
1
6
.
22.
Logani
S,
Herdman
AV,
Little
JV,
Moller
KA.
Vascular “pseudo invasion” in laparoscopic hysterectomy specimens: a diagnostic pitfall
.
Am J Surg Pathol
.
2008
;
32
(
4
):
560
565
.
23.
Kitahara
S,
Walsh
C,
Frumovitz
M,
Malpica
A,
Silva
EG.
Vascular pseudoinvasion in laparoscopic hysterectomy specimens for endometrial carcinoma: a grossing artifact?
Am J Surg Pathol
.
2009
;
33
(
2
):
298
303
.
24.
Metovic
J,
Falco
EC,
Vissio
E,
et al.
Gross specimen handling procedures do not impact the occurrence of spread through air spaces (STAS) in lung cancer
.
Am J Surg Pathol
.
2021
;
45
(
2
):
215
222
.
25.
Blaauwgeers
H,
Thunnissen
E.
Spread through air spaces (STAS): can an artifact really be excluded?
Am J Surg Pathol
.
2021
;
145
(
10
):
1439
.

Author notes

Supplemental digital content is available for this article at https://meridian.allenpress.com/aplm in the May 2024 table of contents.

Competing Interests

The authors have no relevant financial interest in the products or companies described in this article.

Supplementary data