Patients with persistent pain after breast cancer surgery show both delayed and enhanced cortical stimulus processing

Abstract

Background: Women who undergo breast cancer surgery have a high risk of developing persistent pain. We investigated brain processing of painful stimuli using electroencephalograms (EEG) to identify event-related potentials (ERPs) in patients with persistent pain after breast cancer treatment.

Methods: Nineteen patients (eight women with persistent pain, eleven without persistent pain), who were surgically treated more than 1 year previously for breast cancer (mastectomy, lumpectomy, and axillary lymph node dissection) and/or had chemoradiotherapy, were recruited and compared with eleven healthy female volunteers. A block of 20 painful electrical stimuli was applied to the calf, somatopically remote from the initially injured or painful area. Simultaneously an EEG was recorded, and a visual analog scale (VAS) pain rating obtained.

Results: In comparison with healthy volunteers, breast cancer treatment without persistent pain is associated with accelerated stimulus processing (reduced P260 latency) and shows a tendency to be less intense (lower P260 amplitude). In comparison to patients without persistent pain, persistent pain after breast cancer treatment is associated with stimulus processing that is both delayed (ie, increased latency of the ERP positivity between 250-310 ms [P260]), and enhanced (ie, enhanced P260 amplitude).

Conclusion: These results show that treatment and persistent pain have opposite effects on cortical responsiveness.

Keywords: breast cancer surgery; event-related potentials; nerve injury; pain processing; persistent pain.

Figure 1

Area of tactile hypoesthesia (numbness). Notes: This figure shows the topographical map of areas of tactile hypoesthesia (numbness) drawn by the patients without pain and with pain. The scale of percentages shown in the legend represents the number of patients (converted to percentages) who marked that area as hypoesthetic.

Figure 2

Grand average global field power (GFP) and corresponding topographic representations. (A) Grand average GFP (N = 30). The dotted lines indicate peak latency of the different event-related potential (ERP) components. Four different components can be identified: a negative voltage between 110–180 ms, maximal at FCz, labeled as N150, a positive voltage between 190–230 ms, maximal at Cz, labeled as P200, a positive voltage between 250–310 ms, maximal at FCz, labeled as P260, and a positive voltage between 310–380 ms, maximal at Cz and labeled as P350. (B) Topographic representations of the ERP components at the ERP latencies (N = 30). To best illustrate the maximal activity in each representation, we adjusted the scale to its maximal absolute values (for increases and decreases in voltages). As a result the scale differs between the different representations and is therefore left out.

Figure 3

Group-specific topographic representations. Shown are the topographic representations of the different event-related potential (ERP) components at the different ERP latencies (Figure 2). Notes: To best illustrate the maximal activity in each representation, we adjusted the scale to its maximal absolute values (for increases and decreases in voltages). As a result the scale differs between the different representations and is therefore left out.

Figure 4

Event-related potential (ERP) waveforms. Grand average ERPs observed from FCz showing the P260 differences (A) effect of treatment, (B) effect of pain. Notes: Upward deflection is positive charge and downward is negative charge. Representations of ERPs are with respect to common reference.