Distinct Impact of Platinum Chemotherapies on the Fitness and Genome of Hematopoietic Stem Cells and the Risk of Therapy-Related Myeloid Neoplasms

The widespread use of platinum-based chemotherapy has been associated with an increased incidence of therapy-related myeloid neoplasms (t-MNs). Among the three most commonly used platinum—cisplatin, carboplatin, and oxaliplatin—oxaliplatin presents a distinctive clinical usage and toxicity. This distinction is supported by a previous study, which postulated that oxaliplatin-induced cytotoxicity primarily results from ribosomal stress as opposed to a DNA-damage response. Additionally, anecdotal data suggest a lower incidence of t-MNs associated with oxaliplatin.

To investigate the differential risk of t-MNs among the various platinum chemotherapies, we analyzed data from 52,179 patients with diverse cancer types who received one of the platinum chemotherapies at our hospital. Of these patients, 22,652 (44%) received carboplatin alone, 15,358 (30%) received cisplatin alone, 8,805 (17%) received oxaliplatin alone, and 5,364 (9%) received multiple combinations of platinums. A total of 171 patients developed t-MNs, with only 6% of these patients having been exposed to oxaliplatin. There was no significant difference in the prevalence of TP53 mutations among t-MNs associated with different platinum exposures (56%, 60%, and 60% for cisplatin, carboplatin, and oxaliplatin, respectively). The cumulative incidence risk of t-MNs, using the Fine and Gray model, was significantly lower for patients treated with oxaliplatin alone compared to those receiving cisplatin alone (HR 0.39 [95% CI: 0.2-0.75], P = 0.005) or carboplatin alone (HR 0.42 [95% CI: 0.22-0.78], P = 0.01). Multivariate Fine and Gray models adjusting for other chemotherapy exposures also indicated that oxaliplatin treatment was associated with a significantly lower risk of t-MNs (HR 0.49 [95% CI: 0.26-0.9], P = 0.02).

To explore the mechanisms underlying the protective effect of oxaliplatin on t-MN risk, we hypothesized that oxaliplatin treatment does not confer a selective advantage to mutant hematopoietic stem cells (HSCs), given that the selective expansion of mutant HSCs (such as TP53 or PPM1D) under chemotherapy has been implicated in the development of t-MNs. To test this hypothesis, we generated a chimeric mouse transplant model with Trp53-/- or Ppm1d truncating mutant cells (Ppm1d Tr/+) transplanted with wild-type cells in a 1:9 ratio. Recipient mice were treated with either vehicle, cisplatin, or oxaliplatin. Contrary to our hypothesis, both Trp53-/- and Ppm1d Tr/+ cells showed significant clonal expansion after both cisplatin and oxaliplatin treatment compared to vehicle. Thus, the reduced risk of t-MNs with oxaliplatin treatment cannot be attributed to reduced selective pressure of oxaliplatin compared to other platinums.

We then hypothesized that oxaliplatin exerts less genotoxic effect on HSCs compared to other platinums. To test this hypothesis, we generated single-cell-derived HSC colonies from the recipient mice described above and performed whole-genome sequencing of individual colonies to measure platinum-induced somatic mutations in HSCs. Oxaliplatin-treated HSCs exhibited significantly fewer total somatic mutations compared to cisplatin-treated HSCs (mean: 524 vs. 891 mutations, P < 0.0001) and were comparable to vehicle-treated HSCs (mean: 524 vs. 485 mutations, P = 0.49). Mutation signature analysis revealed that a median of 74% (range: 52-100) of somatic mutations were attributed to known platinum signatures (SBS31 and SBS35) in cisplatin-treated HSCs, whereas a median of 34% (range: 25-41) of mutations in oxaliplatin-treated HSCs were platinum-associated (P < 0.001).

In summary, while oxaliplatin treatment imposed a similar degree of selective pressure and promoted the expansion of TP53 and PPM1D mutated HSCs compared to other platinums, it induced significantly fewer somatic mutations in HSCs. These data align with clinical findings that the risk of t-MNs is significantly lower for patients treated with oxaliplatin. Nevertheless, t-MNs that do develop after oxaliplatin treatment still exhibit a high prevalence of TP53 mutations, consistent with our findings that oxaliplatin confers a similar selective advantage to TP53-mutated HSCs. These insights enhance our understanding of t-MN pathogenesis and aid in the development of strategies to mitigate this risk.