Histological/morphological classification of cancers based on ICD-O morphology axis. Applies only to neoplastic diseases. Single-inheritance hierarchy.
ICD-O morphology axis classification for cancer subtypes. Values link to NCI Thesaurus for formal definitions.
This is a coarse histogenetic vocabulary, not a slot for four-digit ICD-O codes. Two rules govern it.
Granularity (the lump/split rule): a value names a morphology *family*, not an individual tumour entity. A family earns its own value when the knowledge base holds neoplastic entries that no existing value can hold correctly, and when it is a top-level morphology group in ICD-O / the WHO classification. A sub-family is split out of its parent only when it dominates curation practice — which is why ``Adenocarcinoma`` and ``Squamous Cell Carcinoma`` sit beside ``Carcinoma``, and ``Multiple Myeloma`` beside ``Plasma Cell Neoplasm``, while single entities (glomus tumour, chordoma, GIST) are held by their family rather than given a value of their own.
Behaviour: this axis names the morphology family, and most values are behaviour-neutral — ``Nerve Sheath Neoplasm`` and ``Pericytic Neoplasm`` cover benign and malignant members alike. Where ICD-O itself splits a family on behaviour, the values follow it (``Adenoma`` vs ``Adenocarcinoma``). Do not read malignancy into a value that does not assert it. A first-class behaviour slot (the ICD-O ``/0``-``/3`` digit) and a place for the four-digit code itself remain open on monarch-initiative/dismech#7548.
When no value fits, omit ``icdo_morphology`` and record why in the entry's ``notes`` or a ``CURATION_TODO`` discussion. There is deliberately no ``Other`` value: the omissions are the signal that tells us which family to add next, and this expansion was driven by exactly those notes.