Article Text
PDF
Abstract
Aims Evaluation of ‘alternative’ vascularisation in human cancer is considered an important prognostic parameter; the 2022 WHO classification of parathyroid tumours despite progresses in clinical triaging of patients strongly emphasises new histopathological parameters to properly stratify these lesions. ‘Alternative’ and ‘classic’ vessels were here investigated for the first time in parathyroid tumours for their possible histopathological and clinical relevance during progression.
Methods Using a double CD31/PAS staining, microvessel density (MVD, ‘classic’ CD31+ vessels), mosaic vessel density (MoVD, ‘alternative’ CD31+/−vessels) and vessel mimicry density (VMD, ‘alternative’ CD31−/PAS+ vessels) were evaluated in 4 normal parathyroid glands (N), 50 Adenomas (A), 35 Atypical Tumours (AT) and 10 Carcinomas (K).
Results Compared with N, MVD significantly increased in A (p=0.012) and decreased in K (p=0.013) with vessel counts lower than in AT and A (p<0.001). MoVs and VMs, absent in normal tissue, were documented in non-benign parathyroid lesions (AT, K) (p<0.001), with MoVs and VMs most represented in AT and K, respectively (p<0.001), in peripheral growing areas. Vessel distribution was correlated to neoplastic progression (r=−0.541 MVD; r=+0.760 MoVD, r=+0.733 VMD), with MVD decrease in AT and K inversely related to MoVD and VMD increase (r=−0.503 and r=−0.456).
Conclusions ‘Alternative’ vessel identification in parathyroid tumours is crucial because it: (1) explains the paradox of non-angiogenic tumours, consisting in a new bloody non-endothelial vessel network and (2) helps pathologists to unmask worrisome lesions. Furthermore, detection of alternative vascular systems in human tumours might explain the limited success of antiangiogenic therapies and encourage new oncological studies.
- ANGIOGENESIS
- Neovascularization, Pathologic
- Parathyroid Diseases
- CARCINOMA
- BLOOD VESSELS
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
A reduction of microvascular density has been documented in the parathyroid adenoma to carcinoma transition. Deeper insights about new vascular mechanisms supporting neoplastic growth and metastasis are necessary.
WHAT THIS STUDY ADDS
The evidence of mosaic and vascular mimicry vessels in neoplastic progression as compensation for the reduction of ‘classic’ vessels.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Morphological detection of ‘alternative’ vessels might be a useful diagnostic parameter to distinguish benign from non-benign lesions and it adds important knowledge to rationalise new ‘angiogenic’ therapies.
Introduction
Parathyroid neoplasms are a heterogeneous group of tumours affecting 0.1%–0.3% of population, usually sporadic and associated with primary hypercalcemia.1 In the last decade, increased surgery and detailed immune phenotypical and molecular characterisation of parathyroid neoplasms has led to a better understanding of their pathogenesis and to important advances in clinical triaging of ill patients, as reported in the 2022 WHO histopathological classification.2 However, the lack of morphological parameters beyond invasion to better define worrisome borderline lesions with potential metastatic capacity, grouped in the endorsed term of ‘Atypical Tumour’ (AT), is still an unmet pathological need.3–5
Similarly to most tumours, parathyroid neoplasms require an adequate blood supply to grow, invade and metastasise.6 7 Understanding mechanisms involved in new vessel formation is one of the core aims of oncologic therapies.8 9
For years, angiogenesis, or the endothelial-dependent formation of new vessels, was believed the only means of blood supply in cancer,10–12 and targeting endothelial cells was the major focus of antiangiogenic therapies.13–15 The increasing evidence about limited efficacy of these therapies in some cancers now redefined ‘non-angiogenic’16–18 has recently found explanation in new scientific evidence about cancer perfusion, documenting temporally and spatially heterogeneous ‘alternative’ vascular networks, like vessel co-option, vessel mimicry and mosaic vessels, during cancer progression.9
These new blood supplies are gaining relevance in human cancer as critical factors related to cancer aggressiveness, increased risk of metastasis, resistance to chemotherapy and to patients’ low survival and poor prognosis.19–24 Thus, a deep comprehension of biomolecular pathways weaving these new vessel networks and finding new targets beyond endothelial cells have become crucial to gain cancer control.9
Vascularisation of parathyroid tumours was investigated only in two works, about 20 years ago, by Viacava et al and Garcia de la Torre et al, who evaluated endothelial-dependent microvessels density.25 26 Though their series excluded AT and included some patients with MEN syndromes, the authors reported some critical observations: (1) adenomas were highly hypervascularised as compared with control tissue; (2) carcinomas showed an ‘apparent paradox’ with a very low angiogenic profile similar to other endocrine carcinomas with aggressive behaviour and (3) a subset of adenomas had a vessel density similar to the one of carcinomas and was supposed to represent lesions with potential aggressive behaviour.25
In consideration of recent findings about the biological relevance of new neoplastic vessel networks in human cancer, the aims of the current study were: (1) to assess a comprehensive evaluation of ‘classic’ and ‘alternative’ endothelial-independent blood supplies in all parathyroid neoplastic categories reported in the last 2022 WHO classification, (2) to assess crucial information in the process of parathyroid malignant transformation and (3) to provide new data with possible relevance in both decisional diagnostic processes in pathology and in clinical patients’ management.
Materials and methods
The study group comprised 99 patients: 28 men (28%), 71 women (72%), age: 20–85 years (median age: 59 years) diagnosed at the Division of Pathology of San Paolo Hospital, University of Milan, Italy, from 2012 to 2022. None of the patients suffered from oncologic pathologies or received chemotherapy and/or radiation therapy before surgery. Specimens were recovered immediately after excision, routinely fixed in 10% buffered neutral formalin and processed for conventional histopathology; 4 normal parathyroid glands from normocalcemic patients accidentally recovered during thyroid surgery were used as controls. Standard 3μ thick tissue sections were stained with H&E and examined by light microscopy.
Histological definition of parathyroid tumours was performed according to the 2022 WHO histopathological classification.2 Briefly, adenomas were benign well-circumscribed masses composed of an admixture of chief, oncocytic, transitional or water-clear cells, with absent or sparse stromal fat.
Atypical Tumours were worrisome neoplastic growths sharing with Carcinomas atypical cytological and architectural features and increased mitotic activity exceeding 5 mitoses/50 high power fields, but lacking unequivocal capsular, vascular or perineural invasion or invasion into adjacent structures or metastases. Our series comprised 50 (53%) Adenomas (A), 35 (37%) Atypical Tumours (AT) and 10 (10%) Carcinomas (K).
For each case, lesion dimension (range: 1.9–2.8 cm, mean: 2.3 cm),parathyroid hormone (PTH) (range: 272.2–620.2 pg/mL, mean: 446.2 pg/mL) and calcium (range: 9.8–12.2 mg/dL, mean: 11 mg/dL) serum levels were also considered.
Patients with AT and K underwent follow-up programmes. After pathological diagnosis PTH and CA serum levels were measured at 1, 6 and 12 months, and then for 5 years for AT annually; in patients with K, ultrasound examinations in the neck region were performed at 6, 12 and 24 months with blood tests at 1, 3, 6 and 12 months and then annually indefinitely.
Immunohistochemistry and histochemistry
Immunohistochemical evaluation of neoplastic vessels was performed on 3 μm-thick tissue sections using the DAKO Omnis immunostainer (Agilent Santa Clara, CA 95051). Vessels were evaluated with a CD31/Periodic acid–Schiff (PAS) double staining performed as follows: all CD31 stained sections were scanned with the Nanozoomer Hamamatsu (Hamamatsu Photonics K.K.) and analysed with the NDP.view software (NDP.view2 Viewing software U12388-01). Once removed the coverglass, all slides were stained with PAS and rescanned. ‘Classic’ vessels were channels with continuous CD31+ luminal lining, mosaic vessels showed intermittent CD31 immunoreactivity of inner vascular walls, and vessels of the vascular mimicry appeared as CD31−/PAS+ tubules surrounded by neoplastic cells.27–29
Quantitative evaluation of ‘classic’ CD31 vessels was performed according to Viacava et al25: briefly, the most CD31 immunoreactive areas (hot spots) at X10 and X20 were identified, and 8 fields at X40, corresponding to a 1 mm2 area, were selected. The identified area was ‘captured’ and CD31+ vessels counted (microvessel density, MVD) using the ‘Image J’ support. Density of mosaic vessels (MoVD, ‘alternative’ CD31+/−vessels) and of channels of the vascular mimicry (VMD, ‘alternative’ CD31−/PAS+ vessels) was assessed in the same way.
Vessel distribution (whether uniform in the whole section without differences in central and in peripheral growing areas, or rarefied in some areas) was evaluated and recorded. Vessel distribution and count are represented in online supplemental figure S1.
Supplemental material
To mark intraluminal erythrocytes30 and document blood perfusion in the different vascular networks,28 31 32 Glycophorin A/PAS double staining was done using the same methodology as CD31/PAS dual staining.
In presence of difficult or equivocal interpretation of data, slides were re-evaluated and discussed till interobserver concordance.
Statistical analysis
The results were expressed as mean±SD Comparison among the four groups was performed by Kruskal-Wallis test. When two groups were compared, the Mann-Whitney test was used. Furthermore, the Spearman’s rank correlation coefficient was used to measure the strength of association between two ranked variables. Statistical and correlation analyses were performed using IBM SPSS Statistics V.28.0. P values <0.05 were considered statistically significant.
Results
Microvessel density
All normal parathyroid tissues had CD31+ microvessels regularly distributed around glandular nests (figure 1A).
Vascularization of normal and neoplastic parathyroid tissue (A-F: CD31/PAS double staining; G-H: Glycophorin A/PAS) (A) Normal parathyroid tissue. Vascularisation is represented by ‘classic’ CD31+ endothelised vessels only. (B) Adenomas. A high number of ‘classic’ CD31+ vessels is homogeneously distributed in adenomatous parenchyma. (C) Atypical Tumours. CD31 vessel immunoreactivity is decreased in neoplastic parenchyma both for CD31 intensity reduction in ‘classic’ vessels (black arrow) and for the rise of mosaic vessels with intermittent CD31 immunostaining (white arrow). (D) Atypical Tumour. A high magnification of a neoplastic area full in mosaic vessels. (E) Carcinoma. Neoplastic nests are surrounded by CD31−/PAS+ channels of the vascular mimicry phenomenon. Tumour cells display granular CD31 cytoplasmic deposits (black arrows). (F) Carcinoma. A high magnification of a vessel mimicry channel. (G) Carcinoma. Glycophorin A red blood cells are documented in the lumen of PAS+vessel mimicry network (arrows). (H) Carcinoma. Glycophorin A immunoreactivity is appreciable in the cytoplasm of neoplastic cells (white arrows) as well as in red blood cells (black arrow) circulating in PAS+VM vessels. PAS, Periodic acid–Schiff.
Compared with normal glands, MVD markedly increased in A (p=0.012) (figure 1B) and decreased in K (p=0.013). Also, comparing A with AT (p<0.001) (figure 1C,D) and with K (p<0.001) or AT with K (p<0.001), a decrease in the number of ‘classic’ vessels was found.
Therefore, the progressive decrease in MVD in the four groups (N, A, AT, K) was statistically significant (p<0.001) and appeared to have an inverse correlation with the neoplastic evolution (r=−0.541, p<0.001).
Vessels had a homogeneous distribution in A, while in borderline and malignant lesions, they were irregularly distributed (p<0.001), with peripheral vascular rarefaction and irregular lumina.
Mosaic vessel density
Mosaic vessels were absent in normal parathyroid tissue. Isolated and peripheral, very small mosaic vessels were observed in 8 (16%) cases of A.
In comparison to normal counterparts, this type of vessels showed significant increases in AT (p=0.014) and in K (p=0.012) (figure 1D). MoVD increase in AT and K compared with A was statistically significant (p<0.001).
The progressive increase in MoVD in the four groups (N, A, AT and K) was statistically significant (p<0.001) and seemed to be directly correlated with neoplastic progression (r=0.760, p<0.001). Furthermore, the number of MoVs is higher in non-benign than in benign lesions (AT+K vs N+A: p<0.001).
MoVs in borderline and malignant lesions displayed a predominantly peripheral distribution.
Vascular mimicry density
VM vessels were absent in normal parathyroid tissue and in A; they were found in 16 cases (46%) of AT and in all K (100%) (figure 1E,F).
VMD was higher in K than in normal glands (p=0.012) and in AT (p<0.001), with a significant increase observed in AT and K as compared with A (p<0.001).
The VMD progressive increment in the four groups (N, A, AT and K) was statistically significant (p<0.001) and seemed to be directly correlated to neoplastic progression (r=0.733, p<0.001). Furthermore, VM vessels in non-benign lesions outnumbered those in benign tumours (AT + K vs N + A: p<0.001).
VM vessels in borderline and malignant lesions displayed a predominantly peripheral distribution.
Finally, the MVD decrease in AT and K was significantly inversely related to MoVD and VMD increase (r=−0.503, p<0.001 and r=−0.456, p<0.001, respectively).
No statistical correlations were found among different vessels (MV, MoV and VM), lesion dimensions, serum PTH and calcium levels or patients’ characteristics.
Glycophorin A+ erythrocytes were documented in both ‘classic’ and in ‘alternative’ vessels in all cases (figure 1G).
Additional findings
Clinical findings
The median follow-up was 67.7 (range: 20.2–264.3) months for AT and 55.90 (range 33.1–164.5) months for K; none of them developed recurrences and/or distant metastasis.
Immunohistochemical findings
CD31 cytoplasmic and/or membrane immunostaining was detected in neoplastic cells of all categories investigated: very faint and finely granular in A, CD31 expression increased in AT and in K in which it appeared as cytoplasmic coarse granular deposits (figure 1E).
Glycophorin A cytoplasmic immunoreactivity was detected in 3 (33%) K (figure 1H).
Statistical results are detailed in tables 1 and 2 and figure 2.
Microvessel density (MVD), mosaic vessel density (MoVD) and vascular mimicry density (VMD) in normal glands, adenomas, atypical tumours and carcinomas compared by Kruskal-Wallis test. Mean (± SD).
Pairwise comparison between normal glands and parathyroid tumours according to vascular category by Mann-Whitney test
Mean distribution and correlation index of ‘classic’ and ‘alternative’ vessels into four categories (N, normal glands; A, adenomas; AT, atypical tumours; K, carcinomas). MoVD, mosaic vessel density; MVD, microvessel density; VMD, vascular mimicry density. Created with IBM SPSS Statistics 28.0.
Discussion
In this study, for the first time, a comprehensive evaluation of ‘classic’ and ‘alternative’ vessel networks (mosaic vessels and channels of the vascular mimicry) has been performed in all categories of parathyroid neoplastic lesions reported in the last 2022 WHO classification2 and, for the first time, heterogeneity in vascular architecture has been documented during neoplastic progression.
Regarding ‘classic’ vascularisation, our results are in line with previous data in the literature25 26: as compared with normal control tissues, a high homogeneous hypervascularisation in adenomas and an evident drop of vessels counts (even a bit beyond normal tissue) in carcinomas (r=−0.541, p<0.001) were documented; furthermore, the vascular behaviour of atypical tumours of our series was comparable to that of the subset of adenomas described by Viacava et al25 with blood perfusion similar to carcinomas, thus confirming their borderline nature.
Interestingly, in our work, benign lesions had an almost exclusive ‘classic’ CD31+ vessel network (only isolated, very small peripheral mosaic vessels were observed in 16% of adenomas), while in atypical tumours and in carcinomas, in the peripheral growing areas, two new vascular channels characterised by reduced or absent CD31 expression (MoV and VM), with abnormal lumina and courses, were documented. Intraluminal circulating glycophorin A+ erythrocytes were observed in all kinds of vessels, thus documenting blood perfusion in the whole neoplastic network, even in ‘non-angiogenic’ areas lacking CD31 immunoreactivity. This finding has never been described before in parathyroid lesions and might have a dual relevance. First, it could explain that ‘the paradox’ described by Viacava et al25 in carcinomas is actually only apparent; second, ‘alternative’ vascularisation appears to be a noteworthy event, necessary in non-benign parathyroid tumour growth (MoVD: p<0.001; VMD: p<0.001). A further peculiar finding of our study is that distribution of ‘alternative’ vessels in borderline tumours and in parathyroid cancer was not casual: mosaic vessels were the most represented channels in atypical tumours (p<0.001), and vascular mimicry channels were the principal sustenance blood network in carcinomas (p<0.001). This finding might suggest that CD31-dependent vascularisation progressively disappears during parathyroid neoplastic evolution to cancer (MoV: r=−0.503; VM: r=−0.456), with intermittent CD31+ mosaic vessels replacing ‘classic’ vascularisation before leaving the way to the CD31−/PAS+ channels of vascular mimicry in carcinomas. Different theories reported in the literature could explain this finding in non-benign lesions: some authors hypothesised a progressive endothelial detachment with formation of luminal ‘holes’, or an endothelial growth insufficient to guarantee a complete lining of new vessels in expanding neoplasia; alternatively, the invasion of migrating neoplastic cells with intraluminal drop of adjacent endothelial cells in vessel abnormal walls was proposed.33 34 Actually, the ultrastructural morphological defects in the neoplastic vessel walls found both in the lamina propria and in the endothelial cells,27 33 35 and our findings about the irregular distribution of new vessels with aberrant lumina in peripheral growing areas would be in line with this theory. Modifications in the endothelial and the epithelial cell morphology, immunophenotype and functionality, such as motility, with acquisition of mesenchymal features configuring the endothelial/epithelial-mesenchymal transformation,36 37 described in some human cancers where VM has been reported, would further sustain this hypothesis.24 36–38 The ‘CD31 staining switch’, usually triggered by hypoxic conditions,39–42 is one of these effects: neoplastic growing cells not in close proximity of vessels acquire endothelial features as documented by CD31 immunoreactivity acquisition in their cytoplasm.43 The progressive cytoplasmic CD31+ deposition in neoplastic cells from adenoma to carcinoma documented in this study, as well as glycophorin A accumulation in cancer cells,44 suggests that relevant complex biomolecular processes are involved in the origin and organisation of the new neoplastic parathyroid perfusion. However, the pathogenesis of mosaic and VM vessels, and whether mosaic vessels anticipate the channels of the vascular mimicry during neoplastic progression to cancer is still a matter of debate in the literature.45–47 Anyway, the biological relevance of our findings in atypical tumours is that in mosaic vessels, some tumour cells are directly exposed to blood flow with possibility to disseminate at distance,27 34 48 49 thus explaining the (though low) metastatic potential of these lesions; moreover, vessel of the vascular mimicry in parathyroid carcinomas can elucidate how ‘low angiogenic’ cancers can metastasise.
In conclusion, this study gave important new insights into the comprehension of mechanisms involved in human cancer progression and resistance to angiogenic oncologic strategies,50–52 documenting that a new vessel organisation, with channels characterised by progressive loss in CD31 immunoreactivity (mosaic vessels and vessels of the vascular mimicry), is necessary to tumour growth in borderline and malignant parathyroid lesions.
Identification of ‘alternative’ vessels in routine histopathological evaluation of parathyroid lesions might be of crucial relevance because (1) it explains that the paradox of non-angiogenic tumours consists in a new bloody non-endothelial vessel network and (2) it might be an additional parameter for pathologists to evaluate morphological worrisome lesions.
In addition, detection of alternative vascular systems in human cancer adds new information explaining, at least in part, the limited success of current antiangiogenic therapies and it encourages new studies for future oncological treatments.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Handling editor Munita Bal.
X @Teo92984267
MF and MDL contributed equally.
Contributors DT and UG designed the research study. MF and MDL wrote the paper. DT, GG, LDP, AMS and GF performed the research. DT contributed essential reagents, tools and prepared figures. MF, MDL analysed the data. UG is responsible for the overall content as guarantor. Data curation: all authors.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer-reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.