CRP, Lung Function Predict Outcomes After Solid-Tumor Cell Therapy
Key Takeaways
- Among solid-tumor cellular-therapy recipients in MD Anderson’s Houston phase I program, higher baseline CRP was associated with shorter OS, while higher infused cell dose was associated with longer PFS.
- In the subgroup with pulmonary function testing, higher zDLCO was associated with longer OS and PFS.
- CRS affected more than half of recipients, and higher baseline ANC was associated with lower adjusted odds of both any-grade and grade 2 or higher CRS.
- ICANS was less frequent than CRS after cellular therapy for solid tumors.
Investigators retrospectively analyzed 122 solid-tumor recipients treated from January 2016 through July 2025 in MD Anderson’s phase I program in Houston, as detailed in Lim and colleagues' solid-tumor cellular therapy study in Frontiers in Immunology. Recipients received T-cell or natural killer cell products, predominantly T-cell therapies, for varied tumors. Bacteremia cases were excluded because infection could not reliably be distinguished from cytokine release syndrome (CRS). Investigators assessed preinfusion C-reactive protein (CRP), absolute neutrophil count (ANC), and base-10 log-transformed infused cell dose; 63 recipients had discretionary pulmonary function testing (PFT) within 12 months before infusion, so the measurements were not necessarily obtained immediately before treatment. Overall survival (OS) was measured from infusion. The authors defined progression-free survival (PFS) as time from infusion to documented disease progression or last follow-up; their stated definition does not specify death as a PFS event. Inpatient teams monitored CRS and immune effector cell–associated neurotoxicity syndrome (ICANS) prospectively and graded them under American Society for Transplantation and Cellular Therapy criteria. Adjusted survival models used complete cases, whereas adjusted toxicity models imputed missing covariates.
In the overall cohort, higher baseline CRP was associated with shorter OS after adjustment (hazard ratio
, 1.009). In the complete-case adjusted PFS model, which included 75 of 122 recipients, higher log-transformed infused cell dose was associated with longer PFS (HR, 0.636). Among recipients with PFT, higher diffusing capacity for carbon monoxide z-score (zDLCO) was associated with longer OS (HR, 0.794) and also tracked with longer PFS. The methods describe Global Lung Function Initiative–derived pulmonary z-scores accounting for age, sex, height, and ethnicity; the authors’ discussion separately describes the z-scores as race-agnostic. The pulmonary associations apply to those tested, and the survival models do not establish that changing inflammation or cell dose would change outcomes.
CRS occurred in 57.4% of recipients, while ICANS occurred in 9%. CRS was observed in 70 of 103 T-cell recipients (68.0%) and none of the 19 natural killer cell recipients; the small natural killer cell group does not establish that natural killer cell therapy prevents CRS. Most CRS was grade 1 or 2, although some recipients required intensive care. Per 1,000 cells/µL higher baseline ANC, adjusted odds were lower for both any-grade and grade 2 or higher CRS. Infused cell dose was associated with CRS before adjustment, but that association was not statistically significant in the adjusted analysis.
Retrospective, single-center associations cannot establish causation, demonstrate a benefit from changing cell dose, or define treatment-selection thresholds. PFT ordering was discretionary; those tested differed from untested recipients in treatment history, performance status, interleukin-2 agonist receipt, and CRS frequency. Pulmonary findings therefore apply only to the tested subgroup. The cohort also combined different tumors, cell products, and interleukin-2 regimens, while people with severe cardiopulmonary disease were excluded from treatment. Too few grade 3 or higher CRS and ICANS events occurred for reliable adjusted modeling; age and cardiopulmonary-comorbidity signals for ICANS came only from univariate analyses.
Lim and colleagues called for prospective evaluation of baseline prognostic and toxicity associations in larger, more homogeneous solid-tumor cellular-therapy cohorts. Their findings leave the clinical role of these pretreatment measures to be determined.
Clinician Questions
Did CRS occur after T-cell and natural killer cell therapy for solid tumors?
Among solid-tumor recipients given adoptive cellular therapy at MD Anderson, CRS was observed only among T-cell recipients, not natural killer cell recipients. Median OS and PFS did not differ significantly between the product groups; the comparison does not establish that either product prevents CRS or confers a survival advantage.
How was zDLCO standardized before cellular therapy for solid tumors?
zDLCO is the z-score for lung diffusing capacity for carbon monoxide assessed before cellular therapy for solid tumors. Investigators derived pulmonary z-scores from Global Lung Function Initiative reference equations accounting for age, sex, height, and ethnicity; this standardizes the measurement rather than establishing a validated treatment-selection threshold.