Metabolomic Insight into Implications of Induction Chemotherapy Followed by Concomitant Chemoradiotherapy in Locally Advanced Head and Neck Cancer

Abstract

The present study compares two groups of locally advanced patients with head and neck squamous cell carcinoma (LA-HNSCC) undergoing concurrent chemoradiotherapy (cCHRT), specifically those for whom it is a first-line treatment and those who have previously received induction chemotherapy (iCHT). The crucial question is whether iCHT is a serious burden during subsequent treatment for LA-HNSCC and how iCHT affects the tolerance to cCHRT. Of the 107 LA-HNSCC patients, 54 received cisplatin-based iCHT prior to cCHRT. The patients were clinically monitored at weekly intervals from the day before until the completion of the cCHRT. The 843 blood samples were collected and divided into two aliquots: for laboratory blood tests and for nuclear magnetic resonance (NMR) spectroscopy (a Bruker 400 MHz spectrometer). The NMR metabolites and the clinical parameters from the laboratory blood tests were analyzed using orthogonal partial least squares analysis (OPLS) and the Mann-Whitney U test (MWU). After iCHT, the patients begin cCHRT with significantly (MWU p-value < 0.05) elevated blood serum lipids, betaine, glycine, phosphocholine, and reticulocyte count, as well as significantly lowered NMR inflammatory markers, serine, hematocrit, neutrophile, monocyte, red blood cells, hemoglobin, and CRP. During cCHRT, a significant increase in albumin and psychological distress was observed, as well as a significant decrease in platelet, N-acetyl-cysteine, tyrosine, and phenylalanine, in patients who received iCHT. Importantly, all clinical symptoms (except the decreased platelets) and most metabolic alterations (except for betaine, serine, tyrosine, glucose, and phosphocholine) resolve until the completion of cCHRT. In conclusion, iCHT results in hematological toxicity, altered lipids, and one-carbon metabolism, as well as downregulated inflammation, as observed at the beginning and during cCHRT. However, these complications are temporary, and most of them resolve at the end of the treatment. This suggests that iCHT prior to cCHRT does not pose a significant burden and should be considered as a safe treatment option for LA-HNSCC.

Keywords: LA-HNSCC; NMR spectroscopy; chemoradiotherapy; head and neck cancer; induction chemotherapy; metabolomics; treatment toxicity.

Figure 1

OPLS coefficient overview plot showing how iCHT and the cCHRT duration affect the blood serum metabolites. A total of 843 blood serum samples collected on a day before and during cCHRT were used in the model. The OPLS coefficients are calculated for the two analyzed property variables (duration of CHRT in days—a continuous variable; iCHT—a discrete variable (1 or 0)) normalized to the range [−1, 1] and displayed as the vertical bars. The height of the bar determines how strongly the analyzed property variable (iCHT, cCHRT duration) affects a given descriptive variable (blood serum metabolite), and the positive and negative values determine the increase or decrease in the value of a given metabolite under the influence of the analyzed property variable. The error whiskers represent 95% confidence interval. The results take into account the resonance signals of metabolites at different chemical shifts, e.g., isoleucine at 0.95 and 1.02 ppm. Legend: 3HB—3-hydroxybutyrate; NAG—N-acetyl-glycoprotein; AceAce—acetoacetate; Glu—glutamate; NAC—N-acetylcysteine.

Figure 2

The box plot representation of the relative changes during the course of cCHRT in the metabolites identified by OPLS as significantly affected by the induction treatment. The lines connect the median points, and the boxes represent 25–75% percentiles. The differences between the iCHT + cCHRT and cCHRT alone groups were assessed using the MWU test, and the statistically significant time points are indicated with a triangle (Δ). The relative concentrations of the metabolites are provided in arbitrary units (a.u.) that do not correspond to the physiological values.

Figure 3

OPLS coefficient overview plot showing how the duration of cCHRT and iCHT affect the clinical parameters. A total of 843 blood serum samples collected on a day before and during cCHRT were used in the model. The OPLS coefficients are calculated for the two analyzed property variables (duration of CHRT in days—a continuous variable; iCHT—a discrete variable (1 or 0)) normalized to the range [−1, 1] and displayed as the vertical bars. The height of the bar determines how strongly the analyzed property variable (iCHT, cCHRT duration) affects a given descriptive variable (clinical parameter), and the positive/negative values determine the increase/decrease in the value of a given clinical parameter under the influence of the analyzed property variable. The error whiskers represent 95% confidence interval. HCT—hematocrit; ALB—albumin; CRP—C-reactive protein; RBC—red blood cell count; HGB—hemoglobin; Retic#—reticulocyte count; HGB REC—hemoglobin concentration in reticulocytes; PLT—platelet; MPV—mean platelet volume.

Figure 4

The box plot representation of the relative changes during the cCHRT course in the clinical parameters identified by OPLS as significantly affected by the induction treatment. The lines connect the median points, and the boxes represent 25–75% percentiles. The differences between the iCHT + cCHRT and cCHRT groups were assessed using the MWU test, and the statistically significant time points are indicated with a triangle. Retic#—reticulocyte count; HCT—hematocrit; CRP—C-reactive protein; RBC—red blood cell count; HGB—hemoglobin; PLT—platelet; MPV—mean platelet volume.