To date cancer immunotherapies such as ICT are broadly available for multiple tumor entities. However, the diagnostic capabilities to differentiate responders from non-responders are still insufficient and available imaging modalities are limited in predicting therapy response or irAE. [18F]FDG PET/CT is widely used as a very sensitive method for initial staging and monitoring of various cancer types as well as response assessment to therapy [21, 22]. However, this method lacks in specificity, especially with regard to its application to ICT. Reinfeld et al. recently revealed that a large proportion of the glucose metabolism as measured by [18F]FDG PET is not exclusively related to viable tumor cells but also to activated immune cells [23].
Although the exact immunological processes in ICT treated patients remain elusive, CD8+ T cells are critically involved in the ICT-induced anti-tumoral immune response [24]. In recent years several efficient immune cell tracking approaches have been developed but the translatability of many preclinical approaches into clinical application is problematic [25, 26]. The radiolabeled antibody fragment (minibody) [89Zr]Zr-Df-IAB22M2C has been engineered to improve the disadvantages of full antibodies for in vivo imaging in particular by reducing the long plasma half-life. To avoid interactions with Fc receptors and the associated immune activation, the Fc region was replaced by a pharmacologically inert domain.
Here, we present first clinical experiences with [89Zr]Zr-Df-IAB22M2C PET/MRI in eight patients under or eligible for ICT. Within our small retrospective patient cohort we detected an elevated CD8 dependent [89Zr]Zr-Df-IAB22M2C uptake in two metastases of two patients. Unfortunately, the two [89Zr]Zr-Df-IAB22M2C accumulating lesions were not eligible for resection. Thus, we were unable to cross validate the elevated [89Zr]Zr-Df-IAB22M2C tracer uptake by CD8 immunohistochemistry. Nevertheless, two of the metastases without enhanced [89Zr]Zr-Df-IAB22M2C uptake were resected: the pararenal metastasis of Patient 1 and the brain metastasis of Patient 3. CD8 immunohistochemistry of the pararenal metastasis revealed a very faint CD8+ T cell infiltrate in the center of the metastasis, thus well cross-validating the lack of [89Zr]Zr-Df-IAB22M2C uptake. Strikingly, some areas with a dense infiltration of CD8+ T cells were found at the margins of the metastasis. The immunohistochemistry staining for CD8 in Fig. 1F is demonstrating that the dense infiltration of CD8+ T cells are located in a very thin layer at the margin of the lesion. As this layer has only a thickness in the micrometer scale and only few CD8+ T cells were located in the tumor center, partial volume effect could explain why this infiltration of CD8+ T cells at the margin was not reflected by the [89Zr]Zr-Df-IAB22M2C PET images.
In the case of the brain metastasis of Patient 3 the breakdown of the blood brain barrier in melanoma metastasis should enable a sufficient delivery of the tracer into the malignant tissue. Since there are only very few CD8+ T cells located in this metastasis according to immunohistochemistry (Supplementary Fig. 1) no relevant uptake was observed in the [89Zr]Zr-Df-IAB22M2C PET.
In recent years, it became evident that the amount and the localization of T cell infiltration in the tumor is a major predictor of the patient outcome [27–29]. The distinction of a) immunological „hot“ tumors with dense immune infiltrates, b) immune „deserts“ without relevant immune infiltrate as well as c) immune excluded tumors with an immune infiltrate at the tumor margins has been proposed by different authors [30, 31]. The CD8 immunohistochemistry of the pararenal metastasis of patient 1 correlated very well with the non-invasive CD8 PET/MR imaging and would classify the lesion into the so called immune „excluded“ phenotype. Nevertheless, larger scale prospective trials with histological cross-validation are needed to validate whether [89Zr]Zr-Df-IAB22M2C PET is applicable to distinguish immunological „hot“ tumors with a dense immune infiltrate from immune excluded or immune deserted tumors.
In addition, it has to be considered that the presence of CD8+ T cells does not guarantee the functionality of the CD8+ T cells. Consequently, additional characterization of the tumor microenvironment like metabolic parameters (glucose metabolism, lactate production), or acidity (pH) etc. might be necessary to further evaluate the effectivity of the T cell infiltrate in regard of an efficient anti tumoral immune response.
Upon ICT the primary and secondary lymphatic organs are considered to be highly involved in the elicited systemic immune response in cancer patients.
Recently, our group performed preclinical translational as well as prospective clinical studies with [18F]FDG PET for the assessment of metabolic changes in primary and secondary lymphoid organs before and upon ICT in order to visualize and quantify the systemic response [21, 32]. These studies showed that an elevated splenic glucose metabolism was related to responsiveness to ICT [32] and may serve as an early treatment response marker [21]. In addition, we observed that the glucose metabolism in the bone marrow, a primary lymphatic organ, was increased in patients responsive to ICT. Interestingly, an elevated glucose metabolism in the bone marrow even before start of ICT was applicable to predict treatment response [21, 32]. Nevertheless, it has to be considered that the increased, but probably relatively unspecific glucose metabolism in the primary and secondary lymphatic organs, might be a feature which can be found also in a variety of both physiological and pathological conditions.
In our patient cohort a high interpatient variability of the [89Zr]Zr-Df-IAB22M2C uptake within the primary and secondary lymphoid organs was observed. Non-malignant lymph nodes with a high [89Zr]Zr-Df-IAB22M2C uptake were found in the majority of the patients especially in the cervical and thoracic region. Interestingly, presence or absence of ICT treatment didn’t impact the elevated [89Zr]Zr-Df-IAB22M2C uptake in these lymph nodes.
Generally CD8+ T cells are critically involved in adaptive immune responses including the elimination of malignant cells but also e.g. in the removal of virus infected cells [17]. Therefore, obviously a lot of physiologic and pathologic processes within the primary and secondary lymphatic organs might influence the [89Zr]Zr-Df-IAB22M2C biodistribution. Thus, patient specific physiological (aged immune system etc.) and pathological processes (viral infections etc.) might interfere with the holistic [89Zr]Zr-Df-IAB22M2C PET signature.
Interestingly, 4 out of 6 patients with tumor progression upon ICT tended towards a lower spleen to liver ratio in comparison to the two ICT responsive patients which were in remission. These findings are in line with preclinical data, where ICT-induced glucose metabolism of the spleen was associated to an elevated number of neutrophils and a reduced number of infiltrating T cells [32]. In accordance to this, our prospective study mentioned above revealed an elevated [18F]FDG uptake in the spleen of most ICT responsive patients, while non-responsive patients did not exhibit any significant differences. Moreover, multiple reasons for dysfunctional CD8+ T cells within the tumor microenvironment have been discovered which are major hurdles for ICT efficacy [13].
Because [89Zr]Zr-Df-IAB22M2C PET alone might be not applicable to evaluate the functionality of CD8+ T cells. Thus, dual tracer approaches might be required for identification of efficient CD8+ T cell functioning and potentially enable the detection of therapy response. As second tracer beside [89Zr]Zr-Df-IAB22M2C newly developed tracers like [18F]-arabinosyl guanine [33] and the granzyme B targeting tracer [68Ga]-NOTA-GZP [34] might provide additional information on T cell functionality. However, also conventional [18F]FDG could potentially give evidence towards T cell activation as we were able to demonstrate previously that the distribution of [18F]FDG in primary and secondary lymphoid organs is accompanied with successful cancer immune responses [32].
Subsequent prospective clinical studies are needed to explore how reliable these patterns are and whether they allow a treatment stratification in order to improve the therapeutic outcome of patients with metastasized cancer.