The terms diplegia and quadriplegia should not be abandoned

Heterogeneity has been implicitly incorporated into historical formulations that attempt to define cerebral palsy (CP) as a diagnostic entity. This intrinsic heterogeneity with respect to such aspects as presentation, causation, and natural history has been rendered explicit in CP’s most recent consensus definition.¹ Heterogeneity challenges and complicates our varied attempts to provide classification schemes that are meaningful, yet heterogeneity also enhances the importance of our efforts to classify this disorder pragmatically so as to provide insight and understanding. Inherent heterogeneity also implies that no single classification approach will be all encompassing and indeed each may be seen as complimentary in terms of the information they provide.²

The oldest, and probably most widely utilized, classification approach to CP is one employed in the clinic, based on a combination of observation and the standard process of the neurological examination.³ The first aspect to this classification approach highlights the predominant quality of the individual’s with movement abnormality. This may be predominantly spastic, dyskinetic (i.e. dystonic, athetotic, or choreic), ataxic-hypotonic, or mixed (i.e. usually a combination of spasticity and dyskinesia). The spastic variant, which predominates in the population of individuals with CP as a whole (approximately 80% of total cases), is further subdivided by its apparent topographic distribution. Quadriplegia requires involvement of all four limbs, which need not be symmetrical on both sides of the body, with the upper limbs often, but not always, more involved than the lower. Hemiplegia refers to spasticity restricted to one side of the body, once again with asymmetry permitted between the upper and lower limbs. While diplegia strictly speaking refers to involvement of any two limbs, by convention it has been taken to refer to spasticity of both lower limbs far in excess of any perceptible upper limb involvement. Spasticity’s presence, and its distribution, is determined by observation and formal testing of the passive resistance to stretch (i.e. tone), motor strength, stretch reflexes, and the preservation of primitive reflexes (i.e. plantar or Babinski) that reflects the absence of descending inhibitory influences.

Careful clinical pathology studies, since the advent of classical neurology in the mid-19th century, have consistently indicated that the finding of spasticity clinically correlates with objective pathology in the pyramidal tract (i.e. cortical motor areas [precentral gyrus, supplementary motor area] and descending fiber bundles [internal capsule, cerebral peduncles, pyramids, lateral and ventral corticospinal tracts]) that project predominantly onto the contralateral anterior horn cells in the ventral gray matter of the spinal cord. Collectively, all aspects of this pathway rostral to the anterior horn cell are considered the upper motor neuron system. Complex basal ganglia (e.g. caudate, putamen, globus pallidus, substantia nigra), cerebellar and peripheral sensory (i.e. proprioceptive) inputs act to modulate the output of this pathway.

While the topographically driven distinction of the spastic variant of CP has extensive historical roots, some authors and CP registries have recently called for the amalgamation of the quadriplegia and diplegia subtypes into a single entity labelled ‘bilateral spastic cerebral palsy’.⁴ Proponents of such a change highlight the lack of a clear objective distinction or ‘dividing line’ between quadriplegia and diplegia. They further rightly point out that the assignment of neurological subtypes proceeds without clear objective markers or operational features. This has precluded establishing this schema’s reliability and interrater consistency in both a cross-sectional or longitudinal manner. Recent emphasis on function within the context of CP, both at a gross motor (Gross Motor Function Classification System [GMFCS]) and fine motor (Manual Ability Classification System [MACS]) level, raises the possibility that functional impairments may be an additional modality to be considered, beyond the standard neurological examination, in assigning a spasticity subtype designation.⁵

The foregoing are valid points to be considered when evaluating the traditional classification scheme. However, an additional, and perhaps more important point to be considered, is the schema’s utility. This refers to its ability to be employed and its meaningfulness in providing important insights and guiding what can be expected in several dimensions, especially from a clinical perspective. In CP, such expectations can be applied to causation, functional status (and by extension rehabilitation needs and future prognosis), the results of laboratory investigations, and the comorbidities that may indeed be the major burdens of illness and healthcare efforts.

At a fundamental level, the traditional distinction of spastic CP into its currently recognized variants is undertaken because clinicians generally possess the skills and resources to do so within the clinical setting. However, ease alone would not justify its retention. While there may not exist a clear ‘dividing line’ between quadriplegia and diplegia, a clinician-driven distinction has consistently yielded considerably different implications between these entities. The etiological profile for these respective subtypes are vastly different with intrapartum asphyxia, periventricular leukomalacia, and cerebral dysgenesis the top three causes for spastic quadriplegia; whereas for spastic diplegia, periventricular leukomalacia, intracranial hemorrhage, and intrapartum asphyxia, in descending order, are the most common three causes.⁶ Further, roughly 10% of spastic quadriplegia cases remain etiologically unknown given present evaluation modalities, whereas in 40% of cases of spastic diplegia, an underlying etiology cannot currently be found.

If the etiological spectrum between these two subtypes is distinct, it is not surprising that radiological findings also differ.⁷ In spastic quadriplegia, a slight majority of cases will have imaging findings consistent with a diffuse encephalopathy, cerebral malformation, or the sequelae of infection, while in spastic diplegia roughly one half of cases are accounted for by periventricular white matter injury alone. Roughly 10% of children with spastic quadriplegiç’ CP will have a normal imaging study, while a quarter of children with spastic diplegia CP will have normal scans.

Dramatic differences in gross motor functional status is apparent between spastic quadriplegia and spastic diplegia.⁸ Roughly three-quarters of children with spastic quadriplegic CP are non-ambulatory (GMFCS level IV or V), while such a non-ambulatory status is rarely encountered (2%) in children found to have spastic diplegic CP. Since neurological subtype is frequently determined before final motor ablement status in relation to ambulation, stratification of the topographic distribution between spastic quadriplegia and spastic diplegia has considerable prognostic importance to families. Similarly, the difference between these groups in terms of ambulation capability has rehabilitation implications. Studies correlating spastic subtype variants and fine motor skills are not yet apparent given the relatively recent introduction of the MACS. However, it is reasonably safe to surmise that obvious differences (given their definitive formulations) will exist between children with spastic quadriplegia and spastic diplegia evaluated with respect to fine motor skills that will also have rehabilitation implications.

A final area of discriminative capacity provided by the distinction of subtype status between diplegia and quadriplegia is in the area of comorbidity burden.⁹ Children with spastic quadriplegia consistently experience a higher frequency (2.5–10-fold depending on the comorbidity studied) of burdensome comorbidities such as sensory impairment (cortical blindness, sensorineural hearing loss), non-verbal communication skills, a requirement for gavage feeding, and concurrent epileptic seizures. While up to half of children with spastic quadriplegia may experience one of these comorbidities, no single comorbidity, except perhaps learning difficulties, occurs at a frequency greater than 10% in children with spastic diplegia. Further, the collective comorbidity burden in spastic quadriplegic CP has been documented as being six-fold that of spastic diplegic CP.

The documentation of the differences between the spastic quadriplegia and spastic diplegia variants of CP strongly suggests that the traditional classification scheme is meaningful and yields information to the clinician beyond the descriptive. This is a strong argument in favour of the retention of our existing terminology and approach. However, it is readily apparent that this approach provides but one means of characterizing the child with CP. The simultaneous application of complimentary classification approaches provides the best hope of accurately capturing the heterogeneity that is CP in each child. Further, the addition of functional approaches to classification (i.e. GMFCS and MACS) are likely to serve to reinforce the certainty of the clinical impression of whether a child has diplegia or quadriplegia. Clinical utilization of such a multi-modal approach should be strongly supported and be a definite goal of our future care efforts.