Tehran University of Medical Sciences

Science Communicator Platform

Stay connected! Follow us on X network (Twitter):
Share By
Pitfalls in Machine Learning-Based Assessment of Tumor-Infiltrating Lymphocytes in Breast Cancer: A Report of the International Immuno-Oncology Biomarker Working Group Publisher Pubmed



Thagaard J1, 2 ; Broeckx G3, 4 ; Page DB5 ; Jahangir CA6 ; Verbandt S7 ; Kos Z8 ; Gupta R9 ; Khiroya R10 ; Abduljabbar K11 ; Acosta Haab G12 ; Acs B13, 14 ; Akturk G15 ; Almeida JS16 ; Alvaradocabrero I17 Show All Authors
Authors
  1. Thagaard J1, 2
  2. Broeckx G3, 4
  3. Page DB5
  4. Jahangir CA6
  5. Verbandt S7
  6. Kos Z8
  7. Gupta R9
  8. Khiroya R10
  9. Abduljabbar K11
  10. Acosta Haab G12
  11. Acs B13, 14
  12. Akturk G15
  13. Almeida JS16
  14. Alvaradocabrero I17
  15. Amgad M18
  16. Azmoudehardalan F19
  17. Badve S20
  18. Baharun NB21
  19. Balslev E22
  20. Bellolio ER23
  21. Bheemaraju V24
  22. Blenman KRM25, 26
  23. Botinelly Mendonca Fujimoto L27
  24. Bouchmaa N28
  25. Burgues O29
  26. Chardas A30
  27. Chon U Cheang M31
  28. Ciompi F32
  29. Cooper LAD33
  30. Coosemans A34
  31. Corredor G35
  32. Dahl AB1
  33. Dantas Portela FL36
  34. Deman F3
  35. Demaria S37, 38
  36. Dore Hansen J2
  37. Dudgeon SN39
  38. Ebstrup T2
  39. Elghazawy M40, 41
  40. Fernandezmartin C42
  41. Fox SB43
  42. Gallagher WM6
  43. Giltnane JM44
  44. Gnjatic S45
  45. Gonzalezericsson PI46
  46. Grigoriadis A47, 48
  47. Halama N49
  48. Hanna MG50
  49. Harbhajanka A51
  50. Hart SN52
  51. Hartman J13, 14
  52. Hauberg S1
  53. Hewitt S53
  54. Hida AI54
  55. Horlings HM55
  56. Husain Z56
  57. Hytopoulos E57
  58. Irshad S58
  59. Janssen EAM59, 60
  60. Kahila M61
  61. Kataoka TR62
  62. Kawaguchi K63
  63. Kharidehal D24
  64. Khramtsov AI64
  65. Kiraz U59, 60
  66. Kirtani P65
  67. Kodach LL66
  68. Korski K67
  69. Kovacs A68, 69
  70. Laenkholm AV70, 71
  71. Langschwarz C72
  72. Larsimont D73
  73. Lennerz JK74
  74. Lerousseau M75, 76, 77
  75. Li X78
  76. Ly A79
  77. Madabhushi A80
  78. Maley SK81
  79. Manur Narasimhamurthy V82
  80. Marks DK83
  81. Mcdonald ES84
  82. Mehrotra R85, 86
  83. Michiels S87
  84. Minhas FUAA88
  85. Mittal S89
  86. Moore DA90
  87. Mushtaq S91
  88. Nighat H92
  89. Papathomas T93, 94
  90. Penaultllorca F95
  91. Perera RD96, 97
  92. Pinard CJ98, 99, 100, 101
  93. Pintocardenas JC102
  94. Pruneri G103, 104
  95. Pusztai L105, 106
  96. Rahman A6
  97. Rajpoot NM107
  98. Rapoport BL108, 109
  99. Rau TT110
  100. Reisfilho JS111
  101. Ribeiro JM112
  102. Rimm D113, 114
  103. Roslind A22
  104. Vincentsalomon A115
  105. Saltotellez M116, 117
  106. Saltz J9
  107. Sayed S118
  108. Scott E119
  109. Siziopikou KP120
  110. Sotiriou C121, 122
  111. Stenzinger A123, 124
  112. Sughayer MA125
  113. Sur D126
  114. Fineberg S127, 128
  115. Symmans F129
  116. Tanaka S130
  117. Taxter T131
  118. Tejpar S7
  119. Teuwen J132
  120. Thompson EA133
  121. Tramm T134, 135
  122. Tran WT136
  123. Van Der Laak J137
  124. Van Diest PJ138, 139
  125. Verghese GE47, 48
  126. Viale G140, 141
  127. Vieth M72
  128. Wahab N142
  129. Walter T75, 76, 77
  130. Waumans Y143
  131. Wen HY50
  132. Yang W144
  133. Yuan Y145
  134. Zin RM146
  135. Adams S83, 147
  136. Bartlett J148
  137. Loibl S149
  138. Denkert C150
  139. Savas P97, 151
  140. Loi S97, 151
  141. Salgado R3, 97
  142. Specht Stovgaard E22, 152

Source: Journal of Pathology Published:2023


Abstract

The clinical significance of the tumor-immune interaction in breast cancer is now established, and tumor-infiltrating lymphocytes (TILs) have emerged as predictive and prognostic biomarkers for patients with triple-negative (estrogen receptor, progesterone receptor, and HER2-negative) breast cancer and HER2-positive breast cancer. How computational assessments of TILs might complement manual TIL assessment in trial and daily practices is currently debated. Recent efforts to use machine learning (ML) to automatically evaluate TILs have shown promising results. We review state-of-the-art approaches and identify pitfalls and challenges of automated TIL evaluation by studying the root cause of ML discordances in comparison to manual TIL quantification. We categorize our findings into four main topics: (1) technical slide issues, (2) ML and image analysis aspects, (3) data challenges, and (4) validation issues. The main reason for discordant assessments is the inclusion of false-positive areas or cells identified by performance on certain tissue patterns or design choices in the computational implementation. To aid the adoption of ML for TIL assessment, we provide an in-depth discussion of ML and image analysis, including validation issues that need to be considered before reliable computational reporting of TILs can be incorporated into the trial and routine clinical management of patients with triple-negative breast cancer. © 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland. © 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.
Other Related Docs
11. Immunoinformatics of Cancers, Immunoinformatics of Cancers: Practical Machine Learning Approaches Using R (2022)
17. Breast Cancer Immunotherapy: Current and Novel Approaches, International Immunopharmacology (2021)