June 2022

This work seeking to understand the mechanisms underlying T cell exhaustion in cancer used a melanoma mouse model to find that tumor-associated macrophages in the tumor microenvironment drive exhaustion programs in a manner that is exacerbated by hypoxic conditions to ultimately protect tumor growth - but could be a target of future therapeutics. 

Summary: Myeloid cells (macrophages, dendritic cells, monocytes etc) within the tumor microenvironment (TME) are key regulators of immune mediated cancer survival. Of these cells, tumor associated macrophages (TAMs) have been widely studied and shown to interact with CD8 T cell. TAMs can control CD8 T cells response, however, little is known about the exact mechanism. Kersten et al used an OVA expressing melanoma mouse model and spatiotemporal analysis to show that exhausted CD8 T cells recruit TAMs using CSF1 to the TME which further promotes T cell exhaustion in the context of hypoxia.

First, the authors show that CD8 T cell exhaustion correlates with the number of TAMs in the TME. Adoptive transfer of OT-1 CD8 T cells (genetically modified CD8 T cells with TCR specific to OVA) into a mouse bearing melanoma expressing OVA tumors (B78ChOVA) lead to high levels of exhaustion by day 14. Depletion of TAMs using the anti-CSF1R antibody reduced the amount of OT-1 CD8 T cell exhaustion and improved pro-inflammatory cytokine production such as IFNg and TNFa.

Bulk RNA sequencing of OT-1 CD8 T cells within the tumors of mice bearing B78ChOVA at day 4 and 14 or from the spleen showed that there was an upregulation of myeloid recruiting genes such as CSF1, CCL3, CCL4 and CCL5 in T cells that were exhausted compared to naïve. In an ex-vivo experiment, the upregulation of these genes in exhausted T cells led to increased migration of monocytes across a trans well system compared to effector or naïve T cells. Interestingly, this also led to an upregulation of proteins that increased antigen presentation potential (MHC I and II) suggesting that the recruited monocytes would form stronger interactions with the exhausted T cells. Depletion of CD8 T cells using anti-CD8 in their melanoma mouse model confirmed that there were less TAMs within tumors and lower expression of MHC I/II compared to IgG controls or CD4 depleted mice. The migration of TAMs and increased expression of MHC I/II was shown to be mediated by CSF1 by using a transgenic CSF1 mutated Rag-/- mouse. All together these results suggest that TAMs promote exhausted T cells which recruit more TAMs to the TME using CSF1.

To further investigate the physical interaction of TAMs and exhausted CD8 T cells, the authors used microscopy to show that exhausted OT-1 CD8 T cells where physically bound to TAMs longer than dendritic cells loaded with the OVA peptide. The TAM bound CD8 T cells had more surface expression of their TCRs in the region closest to the TAM but interestingly less Ca2+ influx compared to CD8 T cells interacting with dendritic cells. This led to decreased proliferation of the CD8 T cells indicating that the TAM exhausted CD8 T cell interaction is long lasting but not activating and likely potentiates further exhaustion.

Lastly, the authors show that CD8 T cells co-cultured with TAMs in hypoxic conditions had higher levels of dysfunction that increased with the abundance of TAMs compared to normoxic conditions. To confirm this finding, the authors used ZipSeq in their melanoma mouse model. ZipSeq is a method that uses barcoding directly onto cells which allows for gene expression mapping at the single cell level based on the cell’s location within the tissue. The results showed that there was increased exhaustion of CD8 T cells towards the hypoxic core of tumors which correlated with terminally differentiated TAM states as macrophages moved from the outer to inner regions of a tumor. CellChat analysis confirmed that CSF1-CSF1R ligand receptor interactions were enriched at the center of tumors compared to outer regions.

Bottom line: Exhausted CD8 T cells can enter a positive feedback loop with tumor associated macrophages in the hypoxic cores of tumors. Blocking the ligand CSF1 or its receptor CSF1R could be a therapeutic target to break the cycle of TAM induced CD8 T cell exhaustion.

This secondary analysis of a randomized clinical trial investigated the ideal total neoadjuvant therapy sequence in locally advanced rectal cancer and found that chemoradiotherapy followed by chemotherapy resulted in a superior pathological response without sacrificing survival or quality of life.

Summary: Total neoadjuvant therapy (TNT), which involves administration of all chemoradiotherapy in the neoadjuvant setting, has been increasingly adopted for multimodal rectal cancer treatment. TNT involves either short-course radiotherapy and chemotherapy or long-course chemoradiotherapy with additional chemotherapy. TNT has many advantages over the traditional treatment paradigm for locally advanced rectal cancer (LARC) including earlier administration of effective systemic therapy to treat micrometastatic disease, preoperative downsizing/downstaging of the tumor, higher rates of pathologic complete response (pCR) following total mesorectal excision (TME), and better toleration/compliance with chemotherapy and radiation regimens. The chemotherapy portion of TNT can be administered by one of two basic strategies: either prior to radiation (termed induction chemotherapy), or after radiation therapy (termed consolidation chemotherapy). Two major trials have demonstrated the benefits of TNT compared to traditional management of LARC: the PRODIGE23 trial, which used an induction chemotherapy sequence, and the RAPIDO trial, which used a consolidation sequence. However, the optimal sequence approach remains unknown. To address this knowledge gap, CAO/ARO/AIO-12 trial compared induction vs. consolidation sequences for patients with LARC, the long-term results of which are now published.

All eligible patients in the trial had LARC and received TNT. Patients were randomized to receive either induction chemotherapy before chemoradiotherapy, or to receive consolidation chemotherapy after chemoradiotherapy. The primary end point was pCR, and secondary endpoints included disease-free survival (DFS), cumulative incidence of locoregional recurrence and distant metastases, overall survival (OS), and chronic toxicity development. Of the 306 eligible patients who underwent TME, 156 were randomized to induction chemotherapy, and 150 were randomized to consolidation chemotherapy. The rate of pCR was significantly higher for those in the consolidation group as compared to the induction group (25% vs. 17%, p<0.001). The duration between completion of radiotherapy and TME performance was longer for those in the consolidation group than the induction group (median duration 90 vs. 45 days). Three-year DFS rates (73% consolidation vs. 73% induction, p=0.82), locoregional recurrence rates (5% consolidation vs. 6% induction, p=0.67), rates of distant metastasis (16% consolidation vs. 18% induction), and OS rates (92% consolidation vs. 92% induction, p=0.81) did not differ between groups. At 36 months, similar rates of chronic toxicity were seen between both groups (9.9% in the consolidation group and 11.8% in the induction group).

This is the first published trial comparing the two TNT sequences head-to-head. The higher pCR rate for the sequence with consolidation therapy is probably due to the difference in time between completion of chemoradiotherapy and surgery. We similarly known from anal cancer that radiation therapy continues to provide benefit as time progresses from completion of treatment. Interestingly, despite these improved rates of pCR, neither sequence provided benefit in regard to recurrence or survival metrics. However, a major benefit of TNT is its ability to allow patients to potentially develop a complete clinical response and forego surgery, thus allowing for organ preservation. We also know from the treatment of anal cancer that earlier evaluation for clinical response may lead to unnecessary surgery, and that sufficient time is needed following completion of chemoradiotherapy to make this assessment. This trial did not evaluate organ preservation, and it remains to be seen which sequence is optimal for this endpoint, although the longer duration between completion of radiotherapy and reevaluation for clinical response may ultimately prove to support a consolidation sequence in this regard. Although the full results are not yet published, preliminary results from the Organ Preservation in Rectal Adenocarcinoma (OPRA) have found that a consolidation approach may result in higher rates of organ preservation for those who develop a complete or near-complete clinical response and undergo a watch and wait strategy (79% consolidation vs. 52% induction, p<0.001), and may also lead to improved DFS (84% consolidation vs. 76% induction, p<0.001) (OPRA results)

Most of the trauma centers that opened in the U.S. between 2014 and 2019 were opened in areas that were already within a 30-minute drive to a trauma care. However, new Level 3, 4, and 5 trauma centers have helped expand access to underserved populations. 

Summary: This is a retrospective study of American trauma centers (TCs) registered with the American Trauma Society (ATS) in 2014 and 2019, examining geographic data to determine whether new centers tend to open in areas that are already well-served or in underserved areas. The country was broken down into census tracts, and a tract was considered “served” if its geographic center was within a 30-minute drive time to a TC. Data were obtained from the census, CDC, American Community Survey (ACS), ATS, and the University of Minnesota’s Integrated Public Use MicroData Set (IPUMS). The authors found that 87.3% of census tracts served by new centers were already served. 82% of the 256 new TCs over the study period were level 3, 4, or 5 centers. Of the newly served tracts, those served by level 3, 4, or 5 centers had a higher proportion of the population living in poverty than those newly served by level 1 or 2 centers (15.7% vs 13.2%, p<0.05).

Overall, the authors conclude that the proportion of tracts served in 2019 was 78.1%, up from 75.3% in 2014 – translating to an increase in population served from 76.0% to 79.4%. While any increase seems good at face value, this marginal increase in access would seem discordant with the number of new centers opened in this 5-year period.

This is an interesting follow up to a landmark JAMA paper by Branas et al (2005) that similarly mapped American access to trauma centers. The authors provide an interesting discussion that incorporates evidence from this study and others demonstrating an overabundance of TCs in some areas, with pockets of the country having no immediate access at all. As they point out, despite evidence that increased access to timely care is associated with lower mortality, the maintenance of a TC is expensive, and opening one in an already-saturated area may dilute resources, personnel, and patient volume. The opening of a new TC should only be done after careful consideration of the existing resources and coordination with surrounding centers.