Cholangiocarcinoma Risk stratification - Prognostic factors related to the patient survival.pptx

Gian Luca Grazi
Professor of Surgery
HepatoBiliaryPancreatic Surgery, AUO Careggi, Florence, Italy
Risk stratification: Prognostic
factors related to the patient
survival

2
Risk stratification - Gian Luca Grazi – HepatoBiliaryPancreatic Surgery, AOU Careggi, Florence, Italy
Gravely AK, J Hepatol. 2022;77(3):865-867. doi: 10.1016/j.jhep.2022.01.004

3

4
Tsilimigras DI, Ann Surg Oncol. 2020;27(4):1110-1119
Factors associated with survival among
patients with multifocal tumors (n = 210)
• CA19-9, UI/mL >200
• AJCC 8th edition N stage N1
• Morphologic type PI, MF + PI
• Grade Poor/undifferentiated
The CART model identified
• tumor number and
• size,
• ALBI grade and
• preoperative LN status
as the most important factors to consider in
the preoperative setting.

5
Tsilimigras DI, Ann Surg Oncol. 2020;27(4):1110-1119
Classification and Regression Tree (CART) model depicting the hierarchical
association of preoperative factors relative to 5-year OS among patients
who underwent surgery for ICC

6
Wiggers JK, J Am Coll Surg. 2016;223(2):321-331.e1. doi: 10.1016/j.jamcollsurg.2016.03.035

7

8
Risks Predicted by the Mortality Risk Score
Developed Mortality Risk Score to Predict 90-Day
Postoperative Mortality after Liver Resection for Perihilar
Cholangiocarcinoma

9
Bird NTE, HPB (Oxford). 2019 ;21(10):1376-1384. doi: 10.1016/j.hpb.2019.02.014

10
Kaplan-Meier curve demonstrating effect of serum CA
19-9 on OS
Kaplan-Meier curve demonstrating effect of
radiological arterial involvement on OS

11
Multivariate modelling for OS

12
Sahara K, J Surg Oncol. 2019;120(2):223-230. doi: 10.1002/jso.25480

13
Cox regression analysis of clinicopathological
factors associated with survival

14
Illustration of the online calculator with principal
characteristics (A) and additional information (B)

15
Illustration of the resultant predicted 5‐year OS (C, D). ICC,
intrahepatic cholangiocarcinoma; OS, overall survival

16
Kaplan‐Meier curve showing the actual OS of patients stratified
by classes of predicted OS

17
Büttner S, HPB 2021, 23, 25–36 doi: 10.1016/j.hpb.2020.07.007

18
Büttner S, HPB 2021, 23, 25–36 doi: 10.1016/j.hpb.2020.07.007
Prisma flow diagram of the VALIDATE effort, to find studies
that validate prognostic models after resection of HPB cancer

19
Choi WJ, Ann Surg Open. 2023 ;4(3):e328. doi: 10.1097/AS9.0000000000000328

20
PRISMA 2020 flow diagram

21
The study characteristics of the included studies
• Twenty-six studiesdeveloped their own original prediction models and 5
studies conducted external validations on previously published models.
 Eighteen studies (69%) published nomograms,
 7 studies (27%) published scoring systems, and
 1 study (4%) published an online calculator.
• All 26 original models were developed using a retrospective cohort, and
9 studies (35%) used single-center data to develop the models.

22
Characteristics of Studies That Have Conducted External Validation Only

23
Characteristics of Studies That Have Conducted External
Validation Only

24
The ICC-Metroticket by Sahara et al. was selected as the overall best
model, and external validation using
• institution’s data estimated Harrell’s and Uno’s C-statistics of 0.67
(95% CI:0.56– 0.77),
• Uno’s time-dependent AUC of 0.71
(95% CI:0.53–0.88) at 5 years,
• Brier score of 0.20
(95% CI:0.15–0.26), and
• good moderate calibration plots.

25
All but one of these models were found to have moderate to high risk
of bias, primarily
• due to insufficient reporting on handling missing data, the use of P
value based forward selection for predictor selection, and
• a lack of reporting on key study characteristics such as median
follow-up or time-zero definitions.
None of these studies reported their S0(t) or baseline risk function,
which is an essential component when validating a Cox prognostic
model in an external dataset.

26
Reasons to be unable to identify a model with
better performance.
• the heterogeneous nature of iCCA poses challenges
for accurate prediction.
• variations in study populations and methodologies
can introduce bias, overfitting, or generalizability
issues.

27
We believe that the main reasons why doctors reject published prognostic
models are lack of clinical credibility and lack of evidence that a prognostic
model can support decisions about patient care

28
The main reasons why doctors reject published prognostic models are
lack of clinical credibility and lack of evidence that a prognostic model can
support decisions about patient care
Clinical credibility
of the model
However accurate a
model is in statistical
terms, doctors will be
reluctant to use it to
inform their patient
management
decisions unless they
believe in the model
and its predictions.
Wyatt JW, BMJ 1995; 311: 1539 doi: 10.1136/bmj.311.7019.1539
Evidence
of
accuracy
Evidence
of
generality
Evidence
of
clinical
effectiveness

29
Evidence of accuracy
A prognostic model is
unlikely to be useful
unless its predictions
are at least as
accurate as those of
the doctors who
would use it.
Clinical
credibility
of
the
model
Evidence
of
generality
Evidence
of
clinical
effectiveness

30
Evidence of generality
Some doctors believe
that no prognostic
model derived from
one population can be
generalised to patients
drawn from another, in
the same way that
some deny that clinical
trials or overviews can
inform individual
decisions about
treatment.
Clinical
credibility
of
the
model
Evidence
of
accuracy
Evidence
of
clinical
effectiveness

31
Evidence of clinical
effectiveness
Even in the case of an
accurate credible
prognostic model,
doctors should
demand empirical
evidence from well
conducted clinical
trials that the model
is clinically effective.
Clinical
credibility
of
the
model
Evidence
of
accuracy
Evidence
of
generality

33

34
Banales JM, Nat Rev Gastroenterol Hepatol. 2020;17(9):557-588. doi: 10.1038/s41575-020-0310-z
Staging cholangiocarcinoma

35
There is no widely used staging system for CCA,
although it can be staged according to the American
Joint Committee on Cancer (AJCC) TNM system.
Despite providing a clinically meaningful
classification correlated with prognosis, the current
TNM classification has some limitations.

36
It has limited discriminatory ability between T2 and T3
tumours in surgically resected iCCAs.
T2 tumours include multifocal disease or disease with
intrahepatic vascular invasion that probably reflect
disseminated disease and the OS in patients with these
tumours does not differ from the OS in patients with
extrahepatic metastatic disease.

37
There is also evidence supporting the negative effect of the
presence of multifocal iCCA (iCCA with liver metastases; T2)
on prognosis (OS) when compared with other early stages,
which might require consideration in future versions of the
AJCC TNM classifications.

38
Although size has been included for the first time as a
prognostic factor for iCCA in the eighth edition of the
AJCC Cancer Staging Manual, the only cut- off size
considered is 5 cm in T1 tumours.
Several authors have shown that a 2 cm cut- off value
might identify very early tumours with very low
likelihood of dissemination and potential long- term
survival with low recurrence rates.

39
The TNM classification misses relevant prognostic factors
such as the presence of cancer- related symptoms (such as
abdominal pain or malaise) or the degree of liver function
impairment.

40

41
Jun SY, Cancer Res Treat. 2019;51(1):98-111

42
Jun SY, Cancer Res Treat. 2019;51(1):98-111

43

44
Clinicopathological and molecular features of cholangiocarcinoma

45
Mechanisms of chemoresistance in cholangiocarcinoma

46
Long‐term outcomes of patients undergoing curative intent resection for
ICC were relatively poor with a 5‐year OS of approximately 40%.
Prognosis of patients was associated with factors such as:
• tumor size and
• number,
• serum CA19‐9 levels,
• lymph node status,
• margin status, and
• tumor differentiation,
• major and microscopic vascular invasion as well as
• receipt of adjuvant chemotherapy.
TAKE HOME MESSAGE (1)

47
TAKE HOME MESSAGE (2)
 The art of prediction in the field of cholangiocarcinoma
has produced a number of proposals, of which none,
however, has managed to find widespread and lasting
application in clinical practice.
 The TNM system, despite its limitations, still remains a
reference in the staging and evaluation of the results of
surgical therapy for cholangiocarcinoma.

Risk stratification
Gian Luca Grazi
Professor of Surgery – HepatoBiliaryPancreatic Surgery AOU Careggi
gianluca.grazi@unifi.it
www.chirurgiadelfegato.it

Cholangiocarcinoma Risk stratification - Prognostic factors related to the patient survival.pptx

More Related Content

Similar to Cholangiocarcinoma Risk stratification - Prognostic factors related to the patient survival.pptx (16)

More from Gian Luca Grazi (20)

Recently uploaded (20)

Cholangiocarcinoma Risk stratification - Prognostic factors related to the patient survival.pptx