HL7 Europe Common Cancer Model
0.1.0 - ci-build 150

HL7 Europe Common Cancer Model, published by HL7 Europe. This guide is not an authorized publication; it is the continuous build for version 0.1.0 built by the FHIR (HL7® FHIR® Standard) CI Build. This version is based on the current content of https://github.com/hl7-eu/cancer-common/ and changes regularly. See the Directory of published versions

Glossary

Concept Overview

Table 1 provides for each Concept: a description of what it represents, some notes (e.g., exclusions), other terms usually adopted to indicate it, and the relationship with other Concepts.

On dates and temporality. To rebuild the cancer journey, dates are fundamental. It is important to define which concepts or attributes can evolve during the journey to understand when to update the same instance versus when to create another one.

Table 1: Concepts, descriptions, notes, synonyms, and connections/dependencies.
Concept Description Note Also known as Connection / Dependency
CancerConditionAtDiagnosis

Represents the Cancer Condition at the First Diagnosis, which is the start of the cancer journey. All attributes never change. It must have:

  • An HistologyBehaviour: histology describes the cell type of the neoplasm and the behaviour describes whether a tumour is malignant or benign, whether it is a carcinoma in situ or a neoplasm of uncertain malignancy. As the definitive histology and behaviour are usually established through Biopsy performed after Surgery, this information may not be available at the early stages of the diagnostic process; however, it is expected to be present once the diagnostic workup has been completed.
  • A BodySite: represents the first location where Cancer Condition was detected (typically with an Imaging).
  • A TumourGrade: describes a tumour in terms of how abnormal the tumour cells are. This is compared to normal cells. It also describes how abnormal the tissues look under a microscope. It describes the biological aggressiveness of the tumour. The grade provides some idea of how the cancer might behave. A low-grade cancer is likely to grow more slowly and be less likely to spread than a high grade one.

It must include at least one of these dates, as they represent the key dates in the diagnostic process:

  • VisitDate: the date of the clinical visit that diagnosed the CancerCondition.
  • BiopsyDate: the date of the biopsy that diagnosed the CancerCondition.
  • LabReportDate: the date of the laboratory report, including biomarker tests when applicable, that diagnosed the CancerCondition.
  • ImagingDate: the date of the diagnostic imaging that diagnosed the CancerCondition.

Each time a patient has a new cancer (so not related and independent from the first one), they have a new Cancer Condition. In case in time, there is an updated of the Cancer Condition (related to the same cell type, so depending on the first cancer), it will be reflected in a new ClinicalCancerProgression instance.

If a cancer is metastatic at the First diagnosis, the information on the metastasis is collected within the CancerStage not in the ClinicalCancerProgression instance.

The couple Histology Behaviour and Body Site identifies the Cancer Condition.

Cancer Condition: Primary Cancer, Secondary Primary Condition
Histology: Morphology
Body Site: location, tumor side, topography

Connected with:

  • A CancerPatient, which is the Subject
  • One or two CancerStage (it must always have one CancerStage of type Clinical and it may have a CancerStage of type Pathological)
  • One or more CancerTreatment
  • One or more OverallCancerTreatmentResponse
  • One or more ClinicalCancerProgression

Basing on the availability in EHR, it can be directly or indirectly connected with some other concepts as:

  • One or more Evidence (on which the diagnosis is based)
  • A Surgery (during which the Biopsy specimen is collected)
  • A Biopsy (on which the TumorGrade and the HistologyBehaviour are defined)
  • An Imaging (on which the BodySite is defined)
CancerPatient

Represents the subject that is affected by one or more Cancer Conditions. It has:

  • Demographics that represents demographic information (e.g. the sex, birth date (or only year)). It can evolve during the time.

It can have:

  • one or more ComorbiditiesAtDiagnosis that represent the active chronic conditions present at the time of the CancerConditionAtDiagnosis.

Subject

Directly connected with:

  • One or more CancerConditionAtDiagnosis
  • One or more CancerTreatment

Indirectly connected with other concepts through CancerConditionAtDiagnosis.

If available in EHR, connected to all possible Procedures and Reports as Subject (not reported in the model).

CancerStage

Represents the stage of the Cancer Condition at the First Diagnosis (CancerConditionAtDiagnosis) and describes the tumour extent. The stage can be Clinical (is based on Imaging) or Pathological (is based on the main Surgery). A CancerConditionAtDiagnosis must always have one CancerStage of type Clinical, which is defined based on Imaging. It may also be associated with a CancerStage of type Pathological, which is defined based on Surgery.

It must have:

  • A Type: indicates whether the Cancer Stage is Clinical or Pathological.
  • A ClassificationType: indicates the name of the staging system used (e.g., TNM (Tumor Linphonode Metastasis)).
  • A Value: indicates the stage values as defined by the applied classification.

In case in time, there is an extension of the Cancer Condition (related to the same cell type, so depending on the first cancer), it will be reflected in a new ClinicalCancerProgression instance with Disease Status (as Progression or Recurrence), while the CancerStage will not change.

If a cancer is metastatic at the First diagnosis, the information on the metastasis is collected within the CancerStage not in the ClinicalCancerProgression instance.

If a cancer is metastatic at the diagnosis, the information on the metastasis is collected within the CancerStage not in the ClinicalCancerProgression.

Connected with:

  • A CancerConditionAtDiagnosis
  • One or more Imaging (in case it is Clinical)
  • A Surgery (in case it is Pathological)
Imaging

Represents the Imaging information used to define the Clinical CancerStage.

It must have:

  • A Type: the type of imaging procedure performed (e.g. thoracic MRI).
  • One or more BodySite: represents the body site on which the procedure was performed.

Connected with a CancerStage (of type Clinical).

Basing on the availability in EHR it can be directly or indirectly connected with an Imaging Report.

CancerTreatment (Surgery, ActiveSurveillance, Radiotherapy, SystemicTreatment)

Represents either a single treatment or a multimodal set of procedures and therapies performed to treat the CancerCondition (i.e., the CancerCondition (AtDiagnosis) and, where applicable, the ClinicalCancerProgression).

It can be divided in 4 main types:

  • Surgery: it is characterized by:
    • A Date: indicates the date the procedure was performed.
    • One or more BodySite: represents the anatomical site where the procedure was performed.
    • An Intent: indicates if it is curative or not (so only to decrease the symptoms) for the Cancer Condition. It can be a choice from: Definitive and Palliative.
  • ActiveSurveillance:
    • It is characterized by:
      • A Start Date: indicates the starting date of the surveillance.
    • can have:
      • An End Date: indicates the date on which the surveillance was concluded.
  • Radiotherapy:
    • It is characterized by:
      • A Start Date: indicates the starting date of the procedure.
      • An End Date: indicates the date on which the procedure was concluded.
      • One or more BodySite: represents the anatomical site where the procedure was performed.
      • An Intent: indicates if it is curative or not (so only to decrease the symptoms) for the Cancer Condition. It can be a choice from: Definitive and Palliative.
    • can have:
      • A Setting: indicates the strategic timing of the CancerTreatments. It can be:
        • Alone: if it is used as the primary treatment, without any type of other CancerTreatments.
        • Preoperative/Neoadjuvant: if is administered before Surgery or SystemicTreatment.
        • PostOperative/Adjuvant: if is delivered after the primary treatment (usually Surgery or SystemicTreatment).
        • Concomitant: if is administered simultaneously with another CancerTreatment.
  • SystemicTreatment:
    • It is characterized by:
      • A Start Date: indicates the starting date of the procedure.
      • A possible End Date: indicates the date on which the procedure was concluded. It should be present if the therapy is not Ongoing.
      • A possible Ongoing: indicates if the therapy is ongoing (in case of immunotherapy or hormone therapy).
      • An Intent: indicates if it is curative or not (so only to decrease the symptoms) for the Cancer Condition. It can be a choice from: Definitive and Palliative.
      • A Type: specifies the category of the treatment (e.g. Chemotherapy, Immunotherapy, Targeted Therapy, or Hormonal Therapy)
    • can have:
      • A Setting: indicates the strategic timing of the CancerTreatments. It can be:
        • Alone: if it is used as the primary treatment, without any type of other CancerTreatments.
        • Preoperative/Neoadjuvant: if is administered before Surgery or SystemicTreatment.
        • PostOperative/Adjuvant: if is delivered after the primary treatment (usually Surgery or SystemicTreatment).
        • Concomitant: if is administered simultaneously with another CancerTreatment.

Typically it is very difficult to define if the target is the CancerCondition (AtDiagnosis) or to a specific ClinicalCancerProgression because the target is the overall CancerCondition.

Also Diet or Exercise can be prescribed Treatments but are not considered for the first period.

SystemicTreatment excludes drugs administered for other pathologies, such as diabetes or hypertension. Should it be necessary to manage administrations unrelated to the CancerCondition, a separate entity would need to be introduced.

In the logical model:

  • Additional attributes can be defined for each type of treatment
  • The concept CancerTreatment will not be represented, since the different types of treatment have highly specific and significantly different attributes that must be defined individually. Furthermore, both in OMOP and in FHIR it is not possible to represent the concept of a combination of treatments of different types. For this reason, it has been decided that the logical model will represent individual treatments.

Treatment: Anti-cancer treatment

Definitive: curative

Connected to:

  • One CancerConditions(AtDiagnosis), as Target/Reason
  • A possible or more ClinicalCancerProgression
  • A CancerPatient, as Subject

Basing on the availability in EHR it can be connected with some other concepts as:

  • A Report (in case the Treatment is a Surgery)
  • a Treatment Plan/Prescription (in case of Radiotherapy or SystemicTreatment)
Last Follow Up

Represents the patient’s health status information collected during the most recent follow-up visit. It has:

  • VitalStatus: that represents the evolving life status of the patient verified during a follow-up Visit, after the first treatment, to indicate if the patient is alive or dead, therefore can be:
    • Alive
    • Dead
  • Date that represent the date of the follow-up Visit.

In case of a living CancerPatient, it must have:

  • EvidenceOfDisease that indicates indicates whether, in the case of a living patient, there is evidence of the presence of disease. in case the CancerPatient is alive it must be provided and can be set to:
    • Yes when the disease is present
    • No when it does not appear to be present.

In case of a dead CancerPatient, it must have:

  • CauseOfDeath that specifies the reason for death.
  • DeathDate that records the date on which the patient died.

Directly connected with:

  • One CancerPatient

Indirectly with other concepts through CancerPatient.

Basing on the availability in EHR it can be connected with some other concepts as:

  • A Visit (when the follow-up was performed).
OverallCancerTreatmentResponse

Represents the standard way to measure how well a CancerPatient responds to a all performed Treatments on the Cancer Condition. It is based on whether tumours shrink, stay the same, or get bigger. It is evaluated by clinician during a follow-up Visit based on an Evidence. It can be one of the following:

  • Progression (in case the cancer is grown respect the previous Visit)
  • Stable Disease (in case the cancer is not changed respect the previous Visit)
  • Partial Remission (in case the cancer shrunk respect the previous Visit)
  • Complete Remission (in case the cancer completely disappeared).

It must have:

  • A Date: it depends on the data source availability and so it is not possible to define a common date to consider as response evaluation date. It could be the Visit date or the date of the Evidence. The common aspect is that it must correspond to one documented report (e.g. Visit, Imaging, Biomarker, Lab Report).

It could seem like a repetition of the Disease Status, but it doesn’t. It is important to have OverallCancerTreatmentResponse and Disease Status separated because they have different purposes. The OverallCancerTreatmentResponse is used to evaluate the effects of the Treatments while Disease Status is to represent the evolution of the Cancer Condition over time.

It is the Disease Status that evolves over the time not the OverallCancerTreatmentResponse.

It is difficult to associate a response with a specific individual treatment. Instead, the response represents the cumulative effect of all interventions performed up to that point to treat the Cancer Condition.

Connected to:

  • One CancerConditions(AtDiagnosis)
  • A possible or more ClinicalCancerProgression

Basing on the availability in EHR it can be connected with some other concepts as:

  • One or more Evidence (on which it the response is evaluated)
  • A Visit (when the response is evaluated).
ClinicalCancerProgression

Represents the evolution over the time and over the space of the Cancer Condition. Every time that there is a new evaluation, there will be a new instance of ClinicalCancerProgression that must have:

  • A Disease Status: represents the status of the Cancer Condition evaluated during a Visit. It could be one of the following:
    • Progression: in case the cancer is grown respect the previous Visit
    • Stable Disease: in case the cancer is not grown or shrunk respect the previous Visit
    • Partial Remission: in case the cancer shrunk respect the previous Visit
    • Complete Remission: in case the cancer completely disappeared
    • Recurrence: in case the cancer after complete remission re-appeared or cancer that has recurred (come back), usually after a period of time during which the cancer could not be detected.
  • An Asserted Date: represents the date when evolution is recorded. It depends on the data source availability and so it is not possible to define a common date to consider as asserted date. It could be the Visit date or the date of the Evidence. The common aspect is that it must correspond to one documented report (e.g. Visit, Imaging, Biomarker, Lab Report).

and it can have attributes to represent the location where the cancer is extended at the time of the assertion (typically evaluated with an Imaging) that can be:

  • One or more Extent type that can be:
    • Local: in case the site is the primary site (the site of the ClinicalCancerProgression is the same of CancerConditionAtDiagnosis)
    • Loco-regional: in case the site is the primary site and the surrounding anatomical region (also to the regional lymph nodes but with no distant metastases)
    • Metastatic: in case there are also distant metastases.
  • Can have one or more:
    • Loco-regional site: the specific site of the surrounding anatomical region involved (it must be present if the the ExtentType is Loco-regional)
    • Metastatic site: the specific site of the metastases (It must be present if the ExtentType is Metastatic).
  • A Tumour Grade: the new tumor grade, in case it is changed respect the first diagnosis (it usually doesn't change).

After the First Diagnosis, every time that there is an evaluation of the Cancer Condition during a Visit, a new instance of ClinicalCancerProgression is created.

If a cancer is metastatic at the First diagnosis, the information on the metastasis is collected within the CancerStage not in the ClinicalCancerProgression.

Please note that, although the term ClinicalCancerProgression includes the word progression, it is intended to describe the evolution of the disease over time, which does not necessarily imply a worsening or progression of the disease.

Connected with:

  • A CancerConditionAtDiagnosis
  • One or more CancerTreatment
  • One or more OverallCancerTreatmentResponse

Basing on the availability in EHR it can be directly or indirectly connected with some other concepts as:

  • One or more Evidence (on which it the diagnosis is based)
  • An Imaging (on which the ExtentType and the possible Loco-regional sites or Metastatic sites are defined)
  • One or more Visit (when the disease evolution is evaluated).
  • a Biopsy (on which the Grading and the Histology Behaviour is defined)

Other concepts not included in this version

Table 2: Additional concepts considered for future use.
Concept Notes
Treatment Plan

It could be useful to record not only the actual treatment, but also what was planned. In fact, the Treatment can change based on the Treatment Response. It could be also useful to consider guidelines and if a patient has agreed or not.

Clinical Trials

It could be useful to consider that a patient can participate in trials and represent its purposes and objectives.

Quality of Life (QoL)/ Patient-Reported Outcome Measures (PROMs) ect..

It could be useful to consider questionnaires to evaluate the quality of life and outcomes, but not with priority. They could be recorded during a Visit.

Adverse Events / Late Effects / Toxicities

It could be useful, but is difficult to have and can be biased, it requires hospitalization.

Risk Factors / Environmental Factors / Genomic Predisposition / Familiarity

It could be useful to have to cover specific research, but not with priority.

Genomics / Biomarker

They could be important, but their scope and usage needs specific discussions, so they are not included in the version 1, but are the next candidate future model extensions.