Taxane-Induced Neuropathy and Its Ocular Effects-A Longitudinal Follow-up Study in Breast Cancer Patients

Abstract

A common severe neurotoxic side effect of breast cancer (BC) therapy is chemotherapy-induced peripheral neuropathy (CIPN) and intervention is highly needed for the detection, prevention, and treatment of CIPN at an early stage. As the eye is susceptible to neurotoxic stimuli, the present study aims to determine whether CIPN signs in paclitaxel-treated BC patients correlate with ocular changes by applying advanced non-invasive biophotonic in vivo imaging. Patients (n = 14, 10 controls) underwent monitoring sessions after diagnosis, during, and after therapy (T0-T3). Monitoring sessions included general anamnesis, assessment of their quality of life, neurological scores, ophthalmological status, macular optical coherence tomography (OCT), and imaging of their subbasal nerve plexus (SNP) by large-area confocal laser-scanning microscopy (CLSM). At T0, no significant differences were detected between patients and controls. During treatment, patients' scores significantly changed while the greatest differences were found between T0 and T3. None of the patients developed severe CIPN but retinal thickenings could be detected. CLSM revealed large SNP mosaics with identical areas while corneal nerves remained stable. The study represents the first longitudinal study combining oncological examinations with advanced biophotonic imaging techniques, demonstrating a powerful tool for the objective assessment of the severity of neurotoxic events with ocular structures acting as potential biomarkers.

Keywords: breast cancer therapy; corneal nerves; neurotoxic events; polyneuropathy; retinal layers; taxanes.

Figure 1

Study flow chart and cohort. A total of 14 patients with newly diagnosed breast cancer and 10 healthy controls were recruited. Of 14 recruited patients, 2 were excluded due to health reasons, therefore, 12 patients and 10 control subjects were assessed. All patients underwent monitoring sessions after diagnosis and before (T0), during (T1 and T2), and after cessation (T3) of the paclitaxel therapy regimen. The monitoring sessions included: the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ)-CIPN20 and Functional Assessment of Cancer Therapy (FACT)-Taxane questionnaire, Neuropathy Disability Score (NDS), Neuropathy Symptom Score (NSS), ophthalmological status, retinal Optical Coherence Tomography (OCT) of the macula, and corneal Confocal Laser Scanning Microscopy (CLSM) of the corneal subbasal nerve plexus. Because of different therapy durations and individual health conditions, not all patients could be examined at all ophthalmological monitoring sessions. One individual underwent paclitaxel, trastuzumab, and pertuzumab treatment. Controls underwent one monitoring session including general anamneses, EORTC QLQ-CIPN20, FACT-Taxane, NDS, and NSS.

Figure 2

Generation of deviation maps. Point-wise aggregated thickness values were computed per retinal layer and time point from the segmented OCT data of all considered patients. Deviation maps were derived by a point-by-point comparison of the aggregated thickness values of each follow-up time point (T1 to T3) with the baseline time point (T0). The color scale explains the thickness differences of the retinal layers compared via deviation maps. Dark colors indicate values outside the baseline interval, a white color indicates values near the baseline mean, and intermediate colors denote either thinning (blue) or thickening (red). Adjacent points with significant differences (p < 0.05) are combined and highlighted by black outlines in the deviation maps.

Figure 3

Overview of the imaging and mosaicking process. The combination of EyeGuidance (guided eye movement with spiral pattern), RCM 2.0 (oscillating focal plane), and HRT (continuous image acquisition) captures a volumetric dataset. The final SNP mosaic is created after image registration and tissue classification.

Figure 4

Curves for the patient-reported outcomes (PROs) concerning quality of life (QoL) and paclitaxel-induced neuropathy. (A) European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire—Chemotherapy Induced Peripheral Neuropathy (EORTC QLQ-CIPN)20; (B) Taxane Subscale Score; (C) Trial Outcome Index; (D) Functional Assessment of Cancer Therapy (FACT) Taxane Total Score; (E) Neuropathy Disability Score (NDS); and (F) Neuropathy Symptom Score (NSS).

Figure 5

Deviation maps showing differences in the thickness of retinal layers (total retina, retinal nerve fiber layer (RNFL), ganglion cell layer (GCL), inner nuclear layer (INL), and inner plexiform layer (IPL)) between the follow-up time points (T1 to T3) and the baseline time point (T0). The ETDRS grid was used to mark three anatomically distinct areas of the retina (center of the fovea (A) and the parafoveal (B) and perifoveal (C) rings). The rows from top to bottom (T1–T0, T2–T0, and T3–T0) each show a relatively greater number and extent of areas with darker red color, indicating an increase in retinal thickness over time. The black outlines mark areas with significant differences (paired T-Test, p < 0.05). This is most evident in the maps for total retinal thickness (first column) and RNFL thickness (second column).

Figure 6

Boxplots showing differences in the thickness of retinal layers (total retina, retinal nerve fiber layer (RNFL), ganglion cell layer (GCL), inner nuclear layer (INL), and inner plexiform layer (IPL)) between baseline (T0) and follow-up (T1 to T3) time points for the center of the fovea and the parafoveal and perifoveal rings of the ETDRS grid. Statistically significant differences (paired T-Test) are marked with *, **, and *** for p < 0.05, p < 0.01, and p < 0.001, respectively. The plots demonstrate a steady increase in the median retinal thickness over time, affecting the measured areas in almost all layers. As with the deviation maps (Figure 5), this is most apparent for total retinal thickness and RNFL thickness.

Figure 7

Example of large-area SNP mosaics of one patient acquired before (T0), during (T1 and T2), and after completion of therapy (T3). Please note, the areas outlined in red denote locally identical regions within the SNP that were used for morphological characterization.

Figure 8

Mean values and 95% confidence intervals (CI) of (A)—corneal nerve fiber length per area (CNFL), (B)—corneal total nerve fiber density (CTNFD), (C)—corneal nerve branching point density (CNBPD), and (D)—corneal nerve fiber tortuosity (CNFTo) of the eight patients. Repeated analysis of variance (ANOVA) measures showed no significant differences.