Surveillance for Early Diagnosis of Hepatocellular Carcinoma in Intermediate and Advanced Cirrhosis

1
Surveillance for Early Diagnosis of
Hepatocellular Carcinoma in Intermediate and
Advanced Cirrhosis

• Narges Farahi, MD
• March 12, 2008

2
Background Mr.. C

• 44 year-old man with h/o HBV and Child-Pugh Class
  C cirrhosis
• Screened with AFP -
• 8/04 20, 3/05 10, 6/05 7, 11/07 1355
• (lost to follow-up 2005-2007)
• Followed with imaging -
• MRI abdomen after elevated AFP in 2004, 2005,
  2007
• 12/07 Lesions consistent with multifocal HCC
• Diagnosed with Hepatocellular Carcinoma -
• by AFP and imaging criteria

3
Background Screening

• Screening
• A test performed on asymptomatic individuals
  that allows for early detection, therapeutic
  intervention, and decreased mortality from the
  disease
• Surveillance is the repeated application of
  screening test

4
Background Screening

• Principles of Screening
• Significant burden of disease in population
• Preclinical stage is detectable and prevalent
• Effective treatment available for disease
• Early detection and treatment improves outcome
  (mortality) with acceptable morbidity
• Screening tests are acceptable to population,
  inexpensive, and sufficiently accurate

5
Background HCC

• Fourth most common cause of cancer death
  worldwide
• Increasing incidence in the United States
• Poor prognosis 5-year survival rate 5 in the
  U.S.
• Diagnosis is often made at an advanced stage
• Small HCC can be cured
• 5-year disease-free survival gt 50 reported for
  both resection and liver transplantation

6
Background The Guidelines(American Association
for the Study of Liver Disease)

• Who to Screen?
• Patients at high risk (gt1.5) for HCC (LOE I)
• Note Specific groups listed based expert
  opinion (LOE III)

7
Background The Guidelines(American Association
for the Study of Liver Disease)

• Who to Screen?
• Patients on liver transplant list (LOE III)
• In the U.S., development of HCC gives increased
  priority fo OLT

8
Background The Guidelines(American Association
for the Study of Liver Disease)

• How to Screen?
• Ultrasound (LOE II)
• Better than serologic tests
• Sensitivity 65-80, Specificity gt90
• AFP should not be used alone (LOE II)
• Poor screening test
• 20ng/mL 60 sensitivity, 41.5 PPV if 5
  prevalence
• Screening interval 6 to 12 months (LOE II)

9
Background The Evidence

• One large RCT showed that surveillance prolongs
  survival in patients with Hepatitis B
• 2 cohort studies found that surveillance
• - prolonged survival in patients with
  Child-Pugh Class A cirrhosis
• - did not improve survival in Class C
• - role in Class B unclear
• Advances in both management of cirrhosis and
  treatments for HCC may increase the benefit of
  surveillance for patients with advanced cirrhosis

10
Background Classifications
Child-Pugh Classification Parameter Points
assigned 1 2 3 Ascites Absent Slight
Mod Bilirubin (mg/dL) lt2 2-3 gt3 Albumin
(g/dL) gt3.5 2.8-3.5 lt2.8 Prothrombin time
Seconds over control lt4 4-6 gt6
INR lt1.7 1.7-2.3 gt2.3 Encephalopathy None
Grade 1-2 Gr 3-4 Grade A Score 5-6, Grade B
7-9, Grade C 10-15
MELD 3.8Ln serum bilirubin (mg/dL) 11.2Ln
INR 9.6Ln serum creatinine (mg/dL) 6.4
11
The Study Question
Does surveillance for HCC improve survival for
patients with Child-Pugh Class B or C cirrhosis?
12
Methods The Patients

• The Italian Liver Cancer (ITA.LI.CA) database has
  prospective data on 1,834 HCC patients, 1987-2004
• Of these, 608 patients were retrospectively
  selected
• 468 Class B
• 140 Class C
• Excluded Patients
• Class A cirrhosis (1,084 patients)
• Class unreported (59 patients)
• Surveillance interval unspecified (83 patients)
  

13
Methods The Patients

• Patients categorized into two cohorts
• Group 1 (Surveillance, 252 patients)
• HCC detected on regular surveillance with
  ultrasound and AFP every 6 or 12 months
• Group 2 (Not Surveillance, 356 patients)
• HCC detected because patient symptomatic or
  incidentally (outside any programmed
  surveillance or during examination for other
  diseases)

14
Methods The Patients

• To minimize length bias, patients under
  surveillance who received US secondary to
  symptoms (41 patients) remained in Group 1
• Length bias apparent improvement in survival
  due to the preferential detection of prevalent,
  slower growing cancers by screening
• Most (80) Group 1 patients were under the care
  of a ITA.LI.CA group clinician. Most Group 2
  cases were referred from outside physicians.

15
MethodsEtiology and Diagnosis of Cirrhosis

• Classification of etiology of liver disease
• HBV, HCV, Alcoholic
• Multietiology
• Others hemochromatosis, primary biliary
  cirrhosis, cyptogenic
• Diagnosis of cirrhosis
• Most based on clinical (radiologic/endoscopic)
  and laboratory features
• Histology in 168 patients
• Laparotomy/laparoscopy in 10 patients

16
MethodsDiagnosis of HCC

• Histology or cytology in 42 patients
• In the others, diagnosis made by
• 1) Diagnostic AFP (gt200 ng/mL) with a typical
  lesion on one imaging technique
• OR
• 2) Typical-appearing lesion on two imaging
  techniques in the absence of diagnostic AFP

17
MethodsStaging of HCC

• By both US and CT scan or MRI
• Macroscopic classification
• unifocal, paucifocal (lt or 3 nodules),
  multifocal (gt3 nodules), infiltrating, massive
• Scored by latest United Network for Organ Sharing
  (UNOS) TNM and Cancer of the Liver Italian
  Program (CLIP) systems

18
Methods Statistical Analysis

• Primary outcome was all-cause mortality
• Survival time was adjusted for lead time bias
• No multivariate analysis was conducted for the
  primary outcome
• The following variables were tested as predictors
  of survival by univariate analysis
• Age, sex, etiology, time of diagnosis,
  comorbidity, surveillance, ALT, AFP level, gross
  pathology, cancer size, portal vein/caval
  thrombosis, metastases, treatment
• Those significantly associated, were tested by
  multivariate analysis.

19
Results

20
Results

• In Child-Pugh C patients, no significant
  difference between groups in most demographic and
  clinical characteristics, except AFP levels

21
Results

• Tumor Features
• Group 1 with more favorable features in Class B
  and C
• More single nodules
• Smaller tumor diameter and less vascular invasion
  
• Earlier stage cancer (62 T1 or T2 stage vs. 26
  in group 2 in Class B, 53 vs. 26 in Class C)
• Treatment
• More surgery or percutaneous treatment in group 1
• More systemic treatment and palliation in group 2
• Difference more marked in Class B patients

22
Results

• Class B

Class C
23
Results Survival

• Child-Pugh Class B
• Prolonged survival in surveillance group
• Median survival 17.1 months vs. 12.0 months
• Survival adjusted for estimated lead time

24
Results Survival

• Child-Pugh Class B
• Subset analysis excluding transplanted patients
  showed that no statistically significant
  difference existed between groups
• No other subset analyses were completed
• Child-Pugh Class C
• No statistically significant difference in
  survival between groups

25
Results Predictors of Survival

Independent prognostic factors, proven by
multivariate analysis
26
Study Strengths

• Study addressed an important clinical question
  that affects a large number of patients with
  intermediate and advanced cirrhosis
• Measures were taken to reduce length bias and
  lead time bias in the evaluation of the efficacy
  of screening

27
Study Limitations

• Selection bias
• Non-randomized allocation of patients to
  surveillance led to significant differences among
  groups
• - academic center vs. referring hospitals
• - etiology of cirrhosis (viral, alcoholic)
• These differences may have affected the
  therapeutic options available to the patients
• Example patients in academic centers may have
  greater access to OLT

28
Study Limitations

• No true control arm
• Some patients in group 2 did receive screening,
  although not in surveillance program
• It remains unclear who will benefit from
  surveillance benefit only shown in patients
  with Child-Pugh Class B cirrhosis who underwent
  OLT, but no other sub-group analyses reported

29
Bottom Line

• This study supports surveillance for patients
  with Child-Pugh Class B Cirrhosis using
  ultrasound and AFP. This applies to those
  patients who would be candidates for treatment if
  HCC is detected.
• It does not support surveillance for Child-Pugh
  Class C Cirrhosis, which is consistent with prior
  studies
• This does not significantly impact the group of
  patients who should be offered screening, since
  practice guidelines in the United States do not
  differentiate who should receive screening based
  on severity of cirrhosis

30
Reference

• Trevisani F, Santi V, Gramenzi A, et al.
  Surveillance for Early Diagnosis of
  Hepatocellular Carcinoma Is It Effective in
  Intermediate/Advanced Cirrhosis? Am J
  Gastroenterol 20071022448-2457.

31
Other References

• 1. Bruix J, Sherman M. Practice Guidelines
  Committee, American Association for the Study of
  Liver Diseases. Management of hepatocellular
  carcinoma. Hepatology 2005421208-36.
• 2. Yuen MF, Cheng CC, Lauder IJ, et al. Early
  detection of hepatocellular carcinoma increases
  the chance of treatment Hong Kong experience.
  Hepatology 200031330-5.
• 3. Zhang BH, Yang BH, Tang ZY. Randomized
  controlled trial of screening for hepatocellular
  carcinoma. J Cancer Res Clin Oncol
  200412735-50.

32
Acknowledgements

• Dave Thom, MD, PhD