Indian Journal of Pathology and Microbiology
Home About us Instructions Submission Subscribe Advertise Contact e-Alerts Ahead Of Print Login 
Users Online: 1354
Print this page  Email this page Bookmark this page Small font sizeDefault font sizeIncrease font size

  Table of Contents    
Year : 2020  |  Volume : 63  |  Issue : 1  |  Page : 98-99
Ki 67: Are we counting it right?

Department of Pathology, AFMC, Pune, Maharashtra, India

Click here for correspondence address and email

Date of Web Publication31-Jan-2020

How to cite this article:
Kinra P, Malik A. Ki 67: Are we counting it right?. Indian J Pathol Microbiol 2020;63:98-9

How to cite this URL:
Kinra P, Malik A. Ki 67: Are we counting it right?. Indian J Pathol Microbiol [serial online] 2020 [cited 2022 Dec 9];63:98-9. Available from:

Ki67 is a nuclear protein that is expressed in all phases of cell cycle. This nuclear protein is best detected during the interphase of the cell cycle within the nucleus of the actively dividing cells. During mitosis, Ki67 migrates to the surface of the chromosome and the cellular content peaks during the synthetic phase. The analysis of the IHC stained cells involves detection of the cells with high Ki67 content as they are positively stained (dark brown) than the rest (blue).[1] In breast carcinoma, particularly the Ki67 is an established independent risk factor determining the disease-free survival. Study by Inwald et al. showed that the 5-year disease-free survival (DFS) rate was 86.7% in patients with a Ki-67 value ≤15% compared to 75.8% in patients with a Ki-67 value >45%. Based on the data from a large cohort of a clinical cancer registry, it was demonstrated that Ki-67 is frequently determined in routine clinical work.[2] Similarly in neuroendocrine tumors, the grading of tumor is based categorically on ki67 count. The Ki67 IHC is immensely altered by preanalytical variables such as type of biopsy, cold ischemia time, type of fixative, and fixation time. To standardize the same ideally, IHC should be carried out on whole section instead of trucut biopsy; cold ischemia time should be <30 min, formaline is the best fixative (avoid Ki67 IHC on frozen tissue), and ideally the fixation should last between 6 and 72 h. The analytical variables include clone of antibody, type of antigen retrieval, and blocking agents. We recommend utilization of MiB1 antibody in dilution validated by each laboratory; antigen retrieval should ideally be done by microwaving and blocking should be done by 5% serum.[3]

The analysis involves broadly two methods of counting the Ki67 marked cells, one is by visually counting the cells within the field of maximum staining, i.e., hotspot, and the other is by automated digital analysis of the same field of view by a computer-based software.

The correct way of visual calculating Ki 67 is by counting the total number of positive-staining tumor cells in each image/field and count the total number of tumor cells in each image to calculate the Ki67 index. Calculation of the Ki67 index = No. of positive tumor cells/total No. of tumor cells ×100. The pathologist should be clear on how to discern positive tumor cells, negative tumor cells, and non-tumor cells. Their brown nuclear stain identifies positive tumor cells. The intensity of brown color will vary from cell to cell. Any degree of brown nuclear stain should be considered as a positive cell. Cytoplasmic brown stain should not be counted as a positive tumor cell. Lymphocyte nuclei will often stain positive should not be counted with the positive tumor cells. The easiest way to recognize a lymphocyte versus a tumor cell is to note the small size of the lymphocyte. It may also help to take into consideration the location of the cell. The tumor cells tend to be in sheets, nests, or groups. Lymphocytes will often be seen outside of these groups as single cells. The negative nuclei will be stained blue. It is important to distinguish negative tumor cells from stromal cells, which should not be counted. Stromal cells are typically smaller, are spindled, and generally surround the clustered groups and nests of tumor cells. Many pathologists tend to do eyeballing of the tumor; however, this is very inaccurate method and should not be followed. Easiest way of calculation is to pickup hotspots by eyeballing technique. These areas should be assessed and a minimum of 500 cells should be counted to assess the Ki67 index. In case there is a discordance of Ki67 and mitotic index, the higher of the two is accepted.[4]

The method of visually counting the cells is the oldest and the conventionally followed method, which is now being largely replaced by the automated digital method. The shift from the conventional to the automated method comes in light of various studies reporting inter-laboratory variation of Ki67 score. Focke et al. conducted a study wherein they found a substantial difference between the Ki67 labeling index of 30 different labs reporting on the same set of breast cancer patients with variation ranging from 17% to 57%, P < 0.0001.[5] A study was conducted by Niazi et al. for standardizing the methods of Ki67 analysis to prevent limitation of inter-laboratory reproducibility and be largely absorbed in clinical assessment of breast cancer. In this study, they visualized and compared the analysis of 75 cases (images); both visually analyzed as well as automated and assessed the error. Their study did not involve counting the cells but was based on a real analysis (eyeballing technique) based on variable staining of the malignant cells and hence differentiating them from the normally stained cells. The proposed automated method resulted in an average root mean square (RMS) error of 3.34, while the analysis by a pathologist accounted for an RMS error of 9.98. This study utilized both tumor and non-tumor tissue and involved approximation.[6],[7]

The commonly adopted procedure of image analysis is appended below [Figure 1]
Figure 1: 4 steps in computational analysis of Ki67

Click here to view

  • Step 1. Uploading an image onto the software
  • Step 2. Selecting the hotspot region (area of maximum staining) on the micrograph
  • >
  • Step 3. Analysis of the dark and light stained cells
  • Step 4. Obtaining results of the digital analysis, i.e., the Ki67 score for the uploaded IHC stained tissue micrograph.

[Figure 1] shows the 4 steps of automated Ki67 calculated value.

One of the significant aspects of Ki67 estimation is the standardization and validation of the test, be it visual or with digital imaging. The software has to ascertain the size of tumor cells, how to negate lymphocytes which are small in size, how to negate the false-positive spindled cells. The non-cellular area is also negated by artificial intelligence. The deep neural networks inbuilt into system algorithmically continue to do this task. Once standardization is done, an automated system can perform a 5000-cell count in 2 ms with an accuracy and reproducibility of 99.9%.

The most important benefit of the automated method is that the process is highly accurate in counting the cells and is less time-consuming. Automated digital image analysis helps to reduce the human effort and removes the most important source of error, i.e., counting error. The similar method can be used to check the proliferation, extent of spread, and most importantly the dynamic response to the chemotherapy.[8]

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

   References Top

Scholzen T, Gerdes J. The Ki-67 protein: From the known and the unknown. J Cell Physiol 2000;182:311-22.  Back to cited text no. 1
Inwald EC, Klinkhammer-Schalke M, Hofstädter F, Zeman F, Koller M, Gerstenhauer M, et al. Ki-67 is a prognostic parameter in breast cancer patients: Results of a large population-based cohort of a cancer registry. Breast Cancer Res Treat 2013;139:539-52.  Back to cited text no. 2
Dowsett M, Nielsen TO, A'Hern R, Bartlett J, Coombes RC, Cuzick J, et al. Assessment of Ki67 in breast cancer: Recommendations from the International Ki67 in Breast Cancer working group. J Natl Cancer Inst 2011;103:1656-64.  Back to cited text no. 3
Cottenden J, Filter ER, Cottreau J, Moore D, Bullock M, Huang WY, et al. Validation of a cytotechnologist manual counting service for the Ki67 index in neuroendocrine tumors of the pancreas and gastrointestinal tract. Arch Pathol Lab Med 2018;142:402-7.  Back to cited text no. 4
Focke CM, Bürger H, van Diest PJ, Finsterbusch K, Gläser D, Korsching E, et al. German breast screening pathology initiative. Interlaboratory variability of Ki67 staining in breast cancer. Eur J Cancer 2017;84:219-27.  Back to cited text no. 5
Niazi MK, Senaras C, Pennell M, Arole V, Tozbikian G, Gurcan MN. Relationship between the Ki67 index and its area based approximation in breast cancer. BMC Cancer 2018;18:867.  Back to cited text no. 6
Koopman T, Buikema HJ, Hollema H, de Bock GH, van der Vegt B. Digital image analysis of Ki67 proliferation index in breast cancer using virtual dual staining on whole tissue sections: Clinical validation and inter-platform agreement. Breast Cancer Res Treat 2018;169:33-42.  Back to cited text no. 7
Saha M, Chakraborty C, Arun I, Ahmed R, Chatterjee S. An advanced deep learning approach for ki-67 stained hotspot detection and proliferation rate scoring for prognostic evaluation of breast cancer. Sci Rep 2017;7:3213.  Back to cited text no. 8

Correspondence Address:
Prateek Kinra
Department of Pathology, AFMC, Pune - 411 040, Maharashtra
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/IJPM.IJPM_770_19

Rights and Permissions


  [Figure 1]

This article has been cited by
1 No differences in histopathological degenerative changes found in acute, trauma-related rotator cuff tears compared with chronic, nontraumatic tears
Knut E. Aagaard, Hanna Cecilia Björnsson Hallgren, Karl Lunsjö, Richard Frobell
Knee Surgery, Sports Traumatology, Arthroscopy. 2022;
[Pubmed] | [DOI]
2 Cynara Scolymus (Artichoke) Improves Liver Regeneration after Partial Liver Resection in Rats
Clinical and Experimental Health Sciences. 2022;
[Pubmed] | [DOI]
3 Unusual bilateral kidney and duodenal plasmablastic lymphoma presentation in an elderly patient: A case report
Yi-Chen Liu, Yu-Te Su, Chih-Kang Huang, Yu-Chi Tsai, Yeu-Chin Chen, Peng-Fei Li
Molecular and Clinical Oncology. 2022; 17(1)
[Pubmed] | [DOI]
4 Development of a Machine learning image segmentation-based algorithm for the determination of the adequacy of Gram-stained sputum smear images
Manraj Sirohi, Mahima Lall, Swapna Yenishetti, Lakshmi Panat, Ajai Kumar
Medical Journal Armed Forces India. 2021;
[Pubmed] | [DOI]
5 The Challenge of Evaluating Response to Peptide Receptor Radionuclide Therapy in Gastroenteropancreatic Neuroendocrine Tumors: The Present and the Future
Virginia Liberini, Martin W. Huellner, Serena Grimaldi, Monica Finessi, Philippe Thuillier, Alfredo Muni, Riccardo E. Pellerito, Mauro G. Papotti, Alessandro Piovesan, Emanuela Arvat, Désirée Deandreis
Diagnostics. 2020; 10(12): 1083
[Pubmed] | [DOI]


    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
    Email Alert *
    Add to My List *
* Registration required (free)  

    Article Figures

 Article Access Statistics
    PDF Downloaded213    
    Comments [Add]    
    Cited by others 5    

Recommend this journal