In human medicine, electroencephalography (EEG), and particularly continuous EEG (cEEG) has become an essential part of monitoring brain function in critically ill patients, both with and without known brain disease (1, 2). First described in human EEG over 60 years ago (3, 4), periodic discharges are noted with varying frequency depending on the patient population, however, they are being documented with increasing frequency in the continuous critical care EEG recordings (2, 5–7).

The first reported periodic pattern was termed PLEDs (periodic lateralized epileptiform discharges), and these discharges were often noted in association with seizures and acute structural brain disease in humans, particularly stroke (3, 8–11). In 2005, the American Clinical Neurophysiology Society (ACNS) recommended removing the word “epileptiform” from the terminology for periodic discharges to be more strictly descriptive of the waveforms, and to avoid clinical connotations or to suggest etiology (12). Subsequent terminology guidelines upheld and elaborated on standardized descriptions for periodic discharges to facilitate consistency (13–15) (Box 1). This alteration in terminology reflects the research over the years supporting that periodic discharges (PDs) are not always related to seizures (i.e. not “epileptiform”), but are sequelae to a focally or globally injured or otherwise dysfunctional cerebrum (10, 11, 16, 17) from a wide variety of causes, including seizures/status epilepticus (6), systemic metabolic disease (1, 18), hypothermia (19), and administration of (20, 21) or withdrawal from (7) anesthetic drugs.

The concept of the “ictal-interictal continuum” (IIC) recognizes that many patients with status epilepticus or acute brain injury from any cause may alternate between ictal and interictal—but encephalopathic—states, and it is within the IIC where many of the rhythmic and periodic patterns lie (2, 10, 22–24).

Veterinary medicine has few reports on EEG beyond describing seizures and ictal patterns (25, 26). In spite of EEG abnormalities determining the highest tier of confidence for diagnosis of idiopathic epilepsy in dogs (27), its use is currently uncommon beyond emergency referral or university practices (28–31). With increased access to digital EEG, its use in monitoring and managing veterinary patients with status epilepticus (SE) or non-convulsive status epilepticus (NCSE) is also increasing (28–30).

This descriptive paper presents visual examples of PDs in veterinary EEG paralleling those described in human EEG, and reviews the human literature on what is known and still unknown about PDs. The goal of this report is to increase recognition and understanding of these periodic patterns on veterinary EEG, especially for monitoring in the critical setting.

Canine and feline EEG cases from two veterinary facilities (UC Davis, Bush Veterinary Neurology Services—multiple locations) from 2003 to 2021 were retrospectively identified by authors (DCW, WWB, MFK) for presence of periodic discharges on the recording as defined in Box 1. Subdermal needle electrodes were used in all cases, and recording montages and electrode placement for each facility were slightly different (Supplementary Figures 1, 2), with concurrent electrocardiogram (ECG). Recordings were reviewed primarily in each institution's bipolar montage (Supplementary Figures 1, 2) (29, 32); remontaging to view a referential montage was also done to aid localization of events (33). The referential montage utilized ‘linked ears' (A1 & A2; Supplementary Figure 1) as the second input.

Cases were chosen exclusively on the visual identification of PDs occurring for 25% or more of the EEG recording. All cases with PDs were clinical patients, for which the EEG was done at the request of the primary clinician for one of three reasons: documenting ictal or interictal abnormalities in a patient with known seizures, evaluating for possible non-convulsive status epilepticus (NCSE), or assessment of cortical function in a patient with stupor or coma. Specific sedation drugs to obtain the EEG were used as clinically needed to assist in electrode placement and dampen muscle artifact. This ranged from no medications (obtunded/stuporous patients), to only whatever antiseizure medications (ASMs) the patient was receiving for management, or specific intravenous (IV) administration of various sedative drugs to obtain a diagnostic EEG recording. Characteristics of the PDs noted were waveform morphology, frequency of the discharges, fluctuation or evolution of either morphology or frequency, and the localization of the discharge, either generalized (GPD) or lateralized (LPD) (Box 1).

Even with hundreds of published cases of periodic discharges on EEG, with corresponding imaging, diagnosis, and outcome, there is still much debate in human medicine about the origin and clinical implications of various PDs, and particularly whether or not treatment with anti-seizure medication is indicated (9–11, 16, 17, 24, 34–36). In human medicine, PD frequency is known to change—either increasing or decreasing (8–11)—sliding up or down the IIC; PD waveform morphology can change—either spontaneously (11) or in response to intervention like surgery (37) or drugs (20, 21); and PDs are frequently documented to resolve with resolution or improvement of the underlying process (7, 8, 11, 19, 21). The case examples in this report demonstrated all of these characteristics. It is perhaps not surprising that dogs and cats can exhibit pathological periodic patterns comparable to those in people, since these species have been investigated for decades for similarities to humans in functional neuroanatomy (38–40).

Advances in invasive and non-invasive multimodal monitoring have not only increased detection of PDs in critically ill people, they have provided new information about the generators and pathophysiological processes associated with periodic discharges. Periodic discharges were originally thought to be secondary to white matter disease only, but human necropsy studies of patients with PDs on EEG showed cortical and/or subcortical gray matter lesions almost exclusively (41).

The origin of PDs is currently theorized as subsequent to interneuron exhaustion from repeated depolarization. The prolonged recovery of interneurons secondary to a structural lesion or frequent, intense depolarization (e.g., SE) results in disinhibition of adjacent local field potentials, permitting synchronization and generation of large-scale discharges, likely detected as PDs on scalp EEG (24, 42). Recent neural network modeling with disinhibition of synaptic connections as a parameter, reliably reproduces GPDs (43), and altering the patterns of disinhibition or the time delay in the model simulates GPDs with varying morphology (polyphasic, multiple waves). This is interesting pathophysiology to consider, given the variation in the morphology of the PDs both between different patients (Figures 1–3, 5, 6), and changes in PD morphology within the same patient, either spontaneously (Figures 4, 10) or with time and continued treatment (Figure 5). It is likely that the changes in PD morphology over time indicate re-establishment of more normal synaptic connections, and that the patient is “moving in the right direction,” down the IIC. In human EEG, the “+F” and “+R” superimposed additional rhythms (Figure 7) impart a higher likelihood of seizures, and the PDs with these additional features are more likely to be on the ictal side of the IIC (15). Therefore, simpler PDs likely indicate a more stable or less epileptogenic cortex, and are likely better for the patient. In case series of human patients with acute causes for PDs, over time, PD frequency decreased and the waveform morphology became less complex and longer in duration (wider waveform) before disappearing (8, 11). Similarly, in our cases where repeat EEGs were done, “simplification” or complete resolution of the PDs was noted (two cases illustrated in Figure 5) as the patient gradually improved clinically.

Visual examples of GPDs with triphasic morphology were specifically highlighted (Figure 8), since the origin, pathophysiology, and naming of this waveform morphology is one of the most debated of all the periodic patterns (44, 45). In human EEG, these previously named “triphasic waves” (TWs) were initially considered pathognomonic for hepatic encephalopathy (HE) (46). Over the years, however, many exceptions were noted, where TWs were present and the underlying process was not HE [toxins/medications (7, 20), immune-mediated encephalitis (47), among many others (44)], and there is no longer an assumption of a metabolic encephalopathy when TWs are noted. A human EEG study (48) evaluating inter-rater agreement on identification of GPDs and TWs found excellent inter-rater agreement for presence of GPDs, but only fair agreement for triphasic morphology. Interestingly, cases with TWs in this human study were less likely to have toxic/metabolic causes for their encephalopathy than those without TWs (48). Similar to the other adjustments in EEG terminology to avoid implying etiology, the current nomenclature describes these waveforms as “GPDs with triphasic morphology” (15), for which a metabolic encephalopathy would be one potential underlying cause (18, 44, 45). In the four cases of TWs illustrated in Figure 8, only one had HE (Figure 8D); the other cases had no systemic metabolic derangements, and seizures only. Conversely, the absence of GPDs with triphasic morphology does not exclude metabolic encephalopathy as an underlying disease: Figure 6 is an example of a dog with HE, seizures, and LPDs with a spike-wave morphology and not the prototypical triphasic morphology.

DW, WB, MK, and KT contributed to the conception of the study. DW, WB, and MK selected EEG cases. KT, MK, and WB organized clinical information. MK wrote the first draft of the manuscript. All authors contributed to manuscript revision, read, and approved the submitted version.