We were very interested to read the recent publication by Vu et al. (1) on the use of spatial statistics in cancer tissue research. We agree with the authors that this will be a crucial area of study in the future, particularly as advanced in situ biomarker detection methods become more common and generate more complex data.

In 2021, we published research on the use of the R-package "Spatstat" for analyzing the tumor microenvironment (TME) (2). In our study, we discussed various summary statistics that can be used for such analyses and highlighted some of the challenges in the spatial analysis of tumor tissue that need to be taken into account. Using a cohort of head and neck cancer specimens, we were able to demonstrate the inhibitory effect of intratumoral hypoxia on CD8-positive lymphocytes (CTLs). However, we found that it can be difficult to reduce complex spatial summary statistics into simplified measures that can be used for patient risk stratification.

We were excited to see that Vu et al. (1) have proposed the Additive Functional Cox Model (AFCM) as a potentially valuable tool for addressing this issue. However, we believe that there are a few points that deserve further consideration.

First, the use of standard spatial summary functions, such as the mark connection function (mcf) used by Vu et al. (1), assumes an isotropic distribution. In reality, the distribution of cells in cancer tissue is often more complex and influenced by a variety of factors. As we demonstrated in our previous study, it is important to try to correct for this inhomogeneity inherent to cancer tissue when applying spatial statistics to single cell-based point patterns.

Second, the correct segmentation, labeling, and export of single-cell data is crucial for the statistical analysis of point patterns. For this reason, we feel that the data presented in Figure 5B of Vu et al. (1) would be more persuasive if the authors provided a depiction of the original MIBI image. An example of our approach can be seen in the app we provided (http://apps.math.aau.dk/s...).

Third, we suggest that the lack of statistical significance in the results of Vu et al. (1) may be due to the need to specify the biological correlates of the cell populations being analyzed more clearly. While the results from the NSCLC data set suggested that increased proximity between cancer and stromal (immune) cells is associated with a better outcome, the opposite was observed for triple negative breast cancer (TNBC). This could be a biological phenomenon, but we believe it is more likely to reflect differences in the compositions of the compartments being studied. We think that the analysis of more clearly defined subpopulations might have produced more relevant results. For example, previous studies in head and neck cancer have shown the importance of tumor-infiltrating CD8+CTLs specifically (3,4).

In conclusion, we are excited to see growing interest in the use of spatial statistics in cancer research. The method proposed by Vu et al. (1) shows great promise for translating the implementation of spatial summary statistics into patient risk stratification. We look forward to applying it to our clinical datasets.

Sincerely,

J. Kaufmann & A. Mayer



References:

1. Vu T, Wrobel J, Bitler BG, Schenk EL, Jordan KR, Ghosh D. SPF: A spatial and functional data analytic approach to cell imaging data. Meier-Schellersheim M, editor. PLOS Comput Biol. 2022 Jun 15;18(6):e1009486.

2. Kaufmann J, Biscio CAN, Bankhead P, Zimmer S, Schmidberger H, Rubak E, et al. Using the R Package Spatstat to Assess Inhibitory Effects of Microregional Hypoxia on the Infiltration of Cancers of the Head and Neck Region by Cytotoxic T Lymphocytes. Cancers. 2021 Apr 16;13(8):1924.

3. Balermpas P, Rödel F, Rödel C, Krause M, Linge A, Lohaus F, et al. CD8+ tumour-infiltrating lymphocytes in relation to HPV status and clinical outcome in patients with head and neck cancer after postoperative chemoradiotherapy: A multicentre study of the German cancer consortium radiation oncology group (DKTK-ROG). Int J Cancer. 2016 Jan 1;138(1):171–81.

4. Kawaguchi T, Ono T, Sato F, Kawahara A, Kakuma T, Akiba J, et al. CD8+ T Cell Infiltration Predicts Chemoradiosensitivity in Nasopharyngeal or Oropharyngeal Cancer. Laryngoscope. 2021 Apr;131(4):E1179–89.