Supplementary Digital Content Utility of Nuclear Grading System in Epithelioid Malignant Pleural Mesothelioma in Biopsy-heavy setting: An External Validation Study of 563 Cases

1

Supplementary Digital Content

Utility of Nuclear Grading System in Epithelioid Malignant Pleural Mesothelioma in Biopsy-heavy setting: An External Validation Study of 563 Cases

Yu Zhi Zhang ^1,2, Cecilia Brambilla ², Philip L Molyneaux ^3,4, Alexandra Rice ^2,3, Jan Lukas Robertus

2,3, Simon Jordan ⁵, Eric Lim ^3,5, Loic Lang-Lazdunski ⁶, Sofina Begum ⁵, Michael Dusmet ⁵, Vladimir Anikin ^5,7, Emma Beddow ⁵, Jonathan Finch ⁵, Nizar Asadi ⁵, Sanjay Popat ^3,8, William OC Cookson

1,3, Miriam F Moffatt ^1,3, Andrew G Nicholson ^2,3

Supplementary Table 1 Assessment criteria for nuclear atypia 2

Supplementary Table 2 Detection of lymphatic and vascular invasion by type of procedure 3

Supplementary Table 3 Association between nuclear grade and mitosis-necrosis score 4

Supplementary Table 4 Comparison of demographic and clinicopathologic characteristics with

previous studies 5

Supplementary Figure 1 Overall survival by clinicopathologic variables 6

Supplementary Figure 2 Overall survival by nuclear features 7

Supplementary Figure 3 Mitotic activity by nuclear grade 8

Supplementary Figure 4 3-tier nuclear grade and mitosis-necrosis score in single site, ≤10mm

setting 9

Supplementary Figure 5 Association between hazard ratio/p values (2-tier nuclear grade) and

number of sites biopsied/maximum tissue dimension 10

2 Supplementary Table 1 Assessment criteria for nuclear atypia

Nuclear Atypia Score 1 (Mild) Score 2 (Moderate) Score 3 (Severe)

Nuclei Small

Uniform Round

Intermediate Limited anisonucleosis Limited pleomorphism

Large Anisonucleosis

Pleomorphism Multinucleation

Nucleoli Inconspicuous Conspicuous Prominent

Multi-nucleoli

Chromatin Quality Fine granular Coarse Coarse

Other Features - - Intra-nuclear inclusion

Irregular nuclear contour Tumor giant cells

Cell-in-cell

3

Supplementary Table 2 Detection of lymphatic and vascular invasion by type of procedure

Procedure Patients (%)

No Invasion

Lymphatic Invasion

Vascular Invasion

Both

p

Biopsy 490 (87.0) 468 (95.5) 7 (1.4) 10 (2.0) 5 (1.0)

<0.001

(Lymphatic)

<0.001

(Vascular)

Resection 73 (13.0) 30 (41.1) 16 (21.9) 8 (11.0) 19 (26.0)

4

Supplementary Table 3 Association between nuclear grade and mitosis-necrosis score

Patients (%) Nuclear Grade

Grade I Grade II Grade III

N (+) N (-) N (+) N (-) Mitosis-Necrosis Score

0 290 159 (54.8) 131 (45.2)

0 (0.0) 0 (0.0) 159

(54.8) 0 (0.0) 131 (45.2)

I 155 18 (11.6) 111 (71.6)

26 (16.8) 18

(11.6) 0 (0.0) 53 (34.2)

58 (37.4)

II 118 0 (0.0) 50 (42.4)

68 (57.6) 0 (0.0) 0 (0.0) 50

(42.4) 0 (0.0)

N, necrosis; (+), present; (-), absent.

5

Supplementary Table 4 Comparison of demographic and clinicopathologic characteristics with previous studies

RBHT Kadota et al., 2012 Rosen et al., 2018

n

563 232 776

Age (Mean, Range) 69 (32-91) 64 (29-85) 65 (29-91)

Sex (M:F) 75:25 72:28 79:21

Procedure

Biopsy 490/563 (87.0%) 11/232 (4.8%) 206/776 (26.5%) PD or EPD 65/563 (11.6%) 91/232 (39.2%) 269/776 (34.7%) EPP 5/563 (0.9%) 130/232 (56.0%) 253/776 (32.6%)

Others 3/563 (0.5%) 0/232 (0%) 48/776 (6.2%)

No. of H&E Sections Reviewed

3.9 (1-47) 9 (1-43) 3 (2-20)

Solid-predominant Growth Pattern

213/563 (37.8%) n/a 275/776 (35.4%)

Necrosis 190/563 (33.7%) n/a 235/776 (30.3%)

Nuclear Grade, Median OS (Months)

Grade I 24.7 28 27

Grade II 12.7 14 14

Grade III 7.2 5 8

Mitosis-Necrosis Score, Median OS (Months)

0 19.8 - 22

I 12.0 - 14

II 8.4 - 10

RBHT, Royal Brompton And Harefield NHS Foundation Trust; M, male; F, female; PD, pleurectomy and decortication; EPD, extended pleurectomy and decortication; EPP, extrapleural pneumonectomy;

H&E, hematoxylin & eosin; OS, overall survival.

6 Supplementary Figure 1 Overall survival by clinicopathologic variables. In univariate analysis, (A) age≤65 years, (B) surgical resection, (C) non solid-predominant growth pattern and (D) absence of necrosis were favorable prognostic variables (all p<0.001). CI, confidence interval; OS, overall survival.

7 Supplementary Figure 2 Overall survival by nuclear features. (A) Nuclear atypia score of 1 was associated with the most favorable prognosis followed by score 2 and 3. (B) Mitotic count of 0-1 per 10 HPF was associated with the most favorable prognosis followed by 2-4 and ≥5. (C) The absence of atypical mitosis was associated with more favorable prognosis (all p<0.001). CI, confidence interval;

HPF, high power field; OS, overall survival.

8 Supplementary Figure 3 Mitotic activity by nuclear grade. No significant difference was found comparing the mitotic activity between nuclear grade III and pleomorphic malignant pleural mesothelioma (p=0.547, Kruskal-Wallis test).

9 Supplementary Figure 4 3-tier nuclear grade and mitosis-necrosis score in single site, ≤10mm

setting. (A) 3-tier nuclear grade discriminates survival difference between grade I and II/III, but not between grade II and III. (B) M-N score discriminates survival difference between score 0 and I/II, but not between grade I and II. CI, confidence interval; OS, overall survival.

10 Supplementary Figure 5 Association between hazard ratio/p values (2-tier nuclear grade) and

maximum tissue dimension /number of sites sampled. (A) Hazard ratio reached plateau with maximum tissue dimension of 20mm. (B) Hazard ratio reached plateau with three or more sites being sampled. Error bar denotes 95% confidence interval and the dotted orange line denotes p=0.05.