SEARCH

SEARCH BY CITATION

Keywords:

  • hepatocellular carcinoma;
  • iodine-131;
  • Lipiodol;
  • multifocal

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

BACKGROUND:

Hepatic artery administration of iodine-131-Lipiodol serves as a modality that delivers targeted radiation therapy to hepatocellular carcinoma. Its efficacy has been promising according to trials conducted in the adjuvant setting after hepatic resection. Further investigation of its role in the palliative setting is warranted.

METHODS:

A retrospective review of 72 patients with unresectable hepatocellular carcinoma treated with iodine-131-Lipiodol and followed up by the St. George Hospital Sydney's hepatobiliary service was conducted. Efficacy of treatment was determined based on progression-free and overall survival as the endpoints using the Kaplan-Meier method.

RESULTS:

Sixty men and 12 women with a mean age of 65 years (standard deviation = 11) underwent iodine-131-Lipiodol treatment. Chronic viral hepatitis was present in 29 (41%) patients. Fifty (69%) patients were Child-Pugh class A. Median progression-free survival was 6 months, and overall survival was 14 months; the 1-, 2-, and 3-year survival rates were 52%, 33% and 20%, respectively. Factors associated with survival include the American Joint Committee on Cancer stage (P = .03), Barcelona Clinic Liver Cancer stage (P = .05), Cancer of the Liver Italian Program score (P = .008), tumor size (P = .01), extrahepatic disease (P < .001), previous surgery (P = .02), and response to treatment (P < .001). The response to treatment was identified through a multivariate analysis as the single independent predictor for survival (hazard ratio, 3.5; 95% confidence interval, 2.2-5.4; P < .001).

CONCLUSIONS:

Encouraging survival outcomes may be derived through administration of iodine-131-Lipiodol in patients with unresectable hepatocellular carcinoma. The overall success of treatment may be determined by the response to treatment. Cancer 2010. © 2010 American Cancer Society.

Hepatocellular carcinoma (HCC) is an increasingly common cancer of the liver in developed countries owing to chronic viral hepatitis and nonalcoholic fatty steatohepatitis. In developing countries, because of the lack of access to healthcare and the endemic presence of chronic viral hepatitis, HCC is common, and tumors are often diagnosed at an advanced stage.1 Only a small proportion of patients will be candidates for curative therapies, whether liver transplant or liver resection. It is estimated that only about 5% of HCC patients are suitable for transplantation,2 and surgical resection rates vary between 9% and 27%.3, 4 In a population-based study of treatment and outcomes of 2963 patients with HCC diagnosed between 1992 and 1999 in the United States, only 27 (1%) patients underwent liver transplantation; 243 (8%) patients underwent surgical resection.5

For the majority of patients with unresectable disease, a bridge toward curative therapies using preoperative therapies may be of value if adequate down-staging can render a feasible and safe hepatic resection. However, the majority of patients have unresectable disease that is generally considered incurable, for which the direction of treatment is palliative. In this regard, the median survival of patients treated with transarterial chemoembolization (TACE) is 324 days; survival without treatment is 82 days.5 TACE is currently widely regarded as the mainstay of treatment owing to data from randomized trials and a meta-analysis.6

With knowledge of the radiosensitivity in hepatoma cell models,7 for which a demonstrable dose-response relationship through local radiotherapy may be achieved,8 recent technological and conceptual developments have paved the way toward a tumor-targeted approach of delivering radiation therapy. Such is the use of Lipiodol-labeled iodine-131 (Lipiodol Laboratoire, Guerbert, France), which is administered intra-arterially. The atom-atom exchange reaction allows tagging of radioactivity to Lipiodol (iodine-131-Lipiodol) for which emission of gamma radiation with energy of 374 KeV penetrates up to 0.4 mm, having a physical half-life of 8 days.9 As Lipiodol is selectively taken up by HCC tumor cells, this allows targeted dose-intensified radiation therapy to be delivered. Despite its reported efficacy in a variety of settings in the treatment of HCC, its use in routine clinical practice is limited, and additional data will contribute toward the overall evaluation of this therapeutic strategy.10 In this study, we report our institution's experience with iodine-131-Lipiodol as a palliative treatment where tumors are considered unresectable and nontransplantable, representing the largest reported experience of iodine-131-Lipiodol in this setting.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Patient Selection

Seventy-two patients treated over a 14-year period between 1996 and 2009 with follow-up information were selected for a retrospective review. Patients had HCC diagnosed either via a tissue biopsy or by a correlative clinical history, alpha fetoprotein levels, and imaging characteristics on a triple-phase computed tomography (CT) scan. Patients with primary HCC or recurrent HCC were included if tumors were deemed unresectable and were Lipiodol avid. Follow-up data were obtained from referring physicians and from telephone calls and/or e-mails from the patients at the time of analysis (October 2009).

Iodine-131-Lipiodol

Iodine-131-Lipiodol was administered intra-arterially by both a nuclear medicine specialist and an interventional radiologist. This treatment protocol involves an initial iodine-131-Lipiodol transmission scan to determine body/lung density before therapy administration, and a 200-MBq (made up to 2.5 mL with cold Lipiodol) ultra-fluid as a tracer dose administration to determine the distribution of iodine-131-Lipiodol (including extrahepatic activity). After this, a tracer scan is performed with analysis to estimate the patient's lung dose and liver dose per unit administered activity. Calculation is done by the Co-57 transmission scan, smoothed using a 10-mm (full width half maximal) Gaussian filter and extrapolated to I-131 energy using the formula:

  • equation image

in which Ix is the ratio of transmitted intensity to free air intensity for the gamma energy x, with linear attenuation coefficient (in H2O) of μx.

The anterior-posterior whole body emission images are smoothed using the same filter as above and combined to form a geometric mean (GM) image, which is then corrected for attenuation using the extrapolated transmission image in accordance with the formula:

  • equation image

in which IA and IP are the anterior and posterior projections, respectively, IS is a measure of free air intensity, and IT is a measure of transmitted intensity.

Two region of interests (ROIs) are drawn on the attenuation-corrected GM image, 1 around the left lung and 1 around the liver. The count rate in each ROI and in the whole image is measured. The single-photon emission CT (SPECT) scan is reconstructed using a quantitative algorithm incorporating CT for attenuation correction. A ROI analysis is performed to estimate the ratio of tumor uptake to liver uptake. If SPECT is not available, or is not amenable to analysis, an uptake ratio of 4:1 is assumed.

The CT scan is analyzed using a ROI to estimate the volume of healthy and diseased tissue in the liver (excluding ascites). If the tumor mass cannot be described using ROI(s), then this can be estimated by eye as percentage liver volume. The therapy dose is planned to limit the total (lifetime) lung dose to no more than 12 grays (Gy) and the total liver dose to no more than 35 Gy. A maximum of 2 GBq (including the tracer dose) will be administered in any case.

The therapeutic dose is administered by the radiologist under the supervision of the nuclear medicine physician intra-arterially through the hepatic artery. After the procedure, the patient is monitored in a radio-protection isolation room for up to 7 days. All patients were administered with 1 treatment. A second dose was administered selectively only in patients who responded to treatment but yet had remnant tumor present and could tolerate a second dose of treatment. In total, 63 (88%) patients received a single treatment, and 9 (13%) patients received 2 treatments. One (1%) patient received a dose of 0.5 Gbq, 66 (92%) patients received a dose of 1.0 Gbq, 1 (1%) patient received a dose of 1.5 Gbq, and 4 (6%) patients received a dose of 2.0 Gbq.

Data Collection

The following data were collected for each patient: demographics—age, sex, presence and cause of chronic hepatitis, ethnicity, and Child-Pugh score; clinical characteristics of the tumor—classification according to the American Joint Committee on Cancer staging (AJCC), Barcelona Clinic Liver Cancer stage, Cancer of the Liver Italian Program score, number of lesions, maximum size of the largest lesion, focality of the tumor, and presence of extrahepatic disease; and treatment-related information—previous surgery or previous ablation treatment. Response to iodine-131-Lipiodol treatment was evaluated retrospectively through examination of CT scan and alpha fetoprotein levels on the first imaging scan and tumor marker testing performed before (in patients with progressive disease) or at 3 months after treatment on the basis of a multidisciplinary tumor board review. A favorable response was classified as when there was evident shrinkage in tumor size, with a correlated decline in alpha fetoprotein levels. An unfavorable response was classified as when there was no evident shrinkage of tumor and a stable or rising alpha fetoprotein level. Response Evaluation Criteria in Solid Tumors were not applied to avoid biasness in a retrospective and nonclinical trial setting.

Statistical Analysis

The data collected were analyzed using SPSS for Windows version 15.0 (SPSS, Munich, Germany). The patient characteristics were reported using frequency and descriptive analyses. The Kaplan-Meier method was used to analyze survival. Univariate analysis (log-rank) was performed to determine the clinical demographic, tumor-related characteristics and treatment-related factors affecting survival. Multivariate analysis was performed on all factors with P < .10 using the Cox proportional hazards regression model. Categorical variables were compared using the chi-square test and multivariate analysis using the logistic regression model. The median time to death was defined as the time when 50% of patients have died. Follow-up was calculated from the date of treatment of carcinomatosis to the date of last follow-up. P ≤ .05 was considered statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Patient Characteristics

Seventy-two patients underwent iodine-131-Lipiodol treatment. There were 60 men and 12 women. The mean age was 65 years (standard deviation = 11). Chronic hepatitis was related to hepatitis B infection in 20 (28%) patients, hepatitis C infection in 9 (13%) patients, and hemachromatosis in 6 (8%) patients. There were 54 (75%) Caucasians, 17 (24%) Asians, and 1 (1%) Middle Eastern patient. Fifty (69%) patients were scored as having Child-Pugh class A, 21 (29%) had Child-Pugh class B, and 1 (1%) had Child-Pugh class C. When classified according to the AJCC, 4 patients were stage I, 29 patients were stage II, 33 patients were stage III, and 6 patients were stage IV. When classified according to the Barcelona Clinic Liver Cancer stage; 6 patients were stage A, 50 patients were stage B, and 16 patients were stage C. When classified according to the Cancer of the Liver Italian Program score; 4 patients had a score of 0, 28 patients had a score of 1, 23 patients had a score of 2, 13 patients had a score of 3, 3 patients had a score of 4, and 1 patient had a score of 5.

Analysis of Treatment Safety

The median length of hospital stay was 4 days (range, 2-14 days). In-hospital complications occurred in 5 (7%) patients, of whom 1 (1%) patient developed acute liver failure post-treatment and died in hospital at Day 14. The remaining 4 patients developed minor complications, including fever in 2 patients, pneumonia in 1 patient, and inguinal hernia in 1 patient. At 3 months follow-up, 47 (65%) patients derived symptomatic benefit, and 24 (33%) patients had stable or worsening symptoms. Two (8%) of 24 patients developed delayed symptoms as a result of delayed treatment-related complications; 1 with radiation-induced pneumonitis and 1 with hepatic vein thrombosis. Nineteen (79%) of 24 patients who were symptomatic at 3 months had an unfavorable response to treatment. The most common symptom occurring after 3 months in these symptomatic patients was ascites secondary to hepatic decompensation from treatment, observed in 12 (50%) patients.

Survival and Follow-up

The median follow-up period in the whole cohort was 9 months (range, 1-122 months). There were 52 (72%) deaths. Disease progression was observed in 67 (93%) patients. The median progression-free survival was 6 months (95% confidence interval [CI], 3.2-9.4) (Fig. 1). The median overall survival after treatment was 14 months (95% CI, 7.4-19.7). The 1-, 2-, and 3-year survival rates were 52%, 33%, and 20%, respectively (Fig. 2).3

thumbnail image

Figure 1. Kaplan-Meier progression-free survival analysis is shown in 72 patients with unresectable hepatocellular carcinoma who were treated with intra-arterial iodine-131-Lipiodol.

Download figure to PowerPoint

thumbnail image

Figure 2. Kaplan-Meier survival analysis is shown in 72 patients with unresectable hepatocellular carcinoma who were treated with intra-arterial iodine-131-Lipiodol.

Download figure to PowerPoint

thumbnail image

Figure 3. Log-rank analysis of response to treatment is shown in 72 patients with unresectable hepatocellular carcinoma who were treated with intra-arterial iodine-131-Lipiodol.

Download figure to PowerPoint

Factors Associated With Survival

Clinical demographic data, tumor-related characteristics, and treatment-related variables were subjected to a univariate analysis to determine relevant associations with survival (Table 1). Factors associated with survival include the AJCC stage (1-2 vs 3-4; 21 vs 4 months; P = .03), Barcelona Clinic Liver Cancer stage (A-B vs C; 15 vs 3 months; P = .05), Cancer of the Liver Italian Program score (≤2 vs >2; 17 vs 3 months; P = .008), maximum tumor size (≤4 vs >4 cm; 20 vs 3 months; P = .01), extrahepatic disease (no vs yes; 14 vs 2 months; P < .001), previous surgery (no vs yes; 7 vs 21 months; P = .02), and response to treatment (favorable vs unfavorable; 26 vs 3 months; P < .001).

Table 1. Univariate Analysis of Clinicopathological Factors Associated With Survival
FactorNo. of PatientsMedian Survival, moP
  • HBV indicates hepatitis B virus; HCV, hepatitis C virus; HFE, hemochromatosis; AJCC, American Joint Committee on Cancer; BCLC, Barcelona Clinic Liver Cancer; CLIP, Cancer of the Liver Italian Program.

  • a

    Statistically significant.

Total72  
Age, y  .94
 <653514 
 ≥65379 
Sex  .29
 Male604 
 Female1214 
Hepatitis  .81
 Nil3714 
 HBV2014 
 HCV94 
 HFE621 
Ethnicity  .46
 Caucasian5411 
 Non-Caucasian (Asian/others)1814 
Child-Pugh class  .54
 A5014 
 B-C225 
AJCC stage  .03a
 I-II3321 
 III-IV394 
BCLC stage  .05a
 A-B5615 
 C163 
CLIP score  .008a
 ≤25517 
 >2173 
α-Fetoprotein  .21
 ≤4004218 
 >400305 
No. of lesions  .29
 ≤4489 
 >42414 
Focality  .85
 Unifocal135 
 Multifocal5914 
Size, cm  .01a
 ≤44320 
 >4294 
Extrahepatic disease  <.001a
 No6614 
 Yes62 
Previous ablation  .13
 No649 
 Yes830 
Previous surgery  .02a
 No567 
 Yes1621 
Response  <.001a
 Favorable4226 
 Unfavorable303 

When variables P < .10 in the univariate analysis were subjected to a Cox proportional hazards regression model for a multivariate analysis, the response to treatment emerged through the multivariate analysis as the single independent predictor for survival (hazard ratio, 3.5; 95% CI, 2.2-5.4; P < .001) (Fig. 3).

Factors Associated With Favorable Response

At the 3-month follow-up, an alpha-fetoprotein reduction of ≥50% was observed in 30 (71%) of 42 patients who were assessed as having a favorable response. At last follow-up, 37 (88%) of 42 patients developed disease progression. Best supportive care was delivered in 13 patients, 8 patients received systemic chemotherapy, 3 patients received repeat iodine-131-Lipiodol, 9 patients received transarterial chemoembolization, 1 patient received radiotherapy, 1 patient received selective internal radiation therapy, 1 patient underwent ablation, 3 patients were subjected to surgery, and 1 patient underwent transplantation.

Favorable response examined using univariate analysis was associated with AJCC stage 1-2 (P < .001), Barcelona Clinic Liver Cancer Stage A-B (P = .003), Cancer of the Liver Italian Program score ≤2 (P < .001), tumor size ≤4 cm (P = .001), absence of extrahepatic disease (P = .004), previous ablation treatment (P = .02), and previous surgery (P = .001) (Table 2). Multivariate analysis using a logistic regression model of variables P < .10 in the univariate analysis identified AJCC stage 1-2 (hazard ratio, 6.0; 95% CI, 1.8-20.7; P = .004), and Cancer of the Liver Italian Program score ≤2 (hazard ratio, 5.5; 95% CI, 1.3-24.0; P = .03) as independent predictors for a favorable response.

Table 2. Univariate Analysis of Clinicopathological Factors Associated With Favorable Response
FactorFavorable ResponseUnfavorable ResponseP
  • AJCC indicates American Joint Committee on Cancer; BCLC, Barcelona Clinic Liver Cancer; CLIP, Cancer of the Liver Italian Program.

  • a

    Statistically significant.

Age, y  1.00
 <652015 
 ≥652215 
Sex  .11
 Male48 
 Female3822 
Viral hepatitis  .47
 No2716 
 Yes1514 
Ethnicity  .58
 Caucasian3024 
 Non-Caucasian (Asian/others)126 
Child-Pugh class  .07
 A3317 
 B-C913 
AJCC stage  <.001a
 I-II285 
 III-IV1425 
BCLC stage  .003a
 A/B3818 
 C412 
CLIP score  <.001a
 ≤23916 
 >2314 
α-Fetoprotein  .48
 ≤4002616 
 >4001614 
No. of lesions  .80
 ≤42721 
 >4159 
Focality  .37
 Unifocal67 
 Multifocal3623 
Size, cm  .001a
 ≤43211 
 >41019 
Extrahepatic disease  .004a
 No4224 
 Yes06 
Previous ablation  .02a
 No3430 
 Yes80 
Previous surgery  .001a
 No2729 
 Yes151 

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Current options for patients with unresectable HCC are limited. Although HCC tumor cells are radiosensitive, the role of external beam radiotherapy is limited by the inability to deliver sufficiently high tumoricidal dose that is tolerable by the liver. It is estimated that the maximum tolerable external radiotherapy dose to the entire liver is approximately 30 Gy; however, up to 4× this maximum tolerable dose is required to achieve tumoricidal effects.9 Hence, external radiotherapy, although it may be useful in livers with limited tumor replacement, is unlikely to be feasible in multifocal HCC, where tumors are often widespread in the entirety of the liver. Percutaneous ethanol injection, radiofrequency ablation, and transarterial chemoembolization tend to be more effective in smaller tumors (about 2 cm) and are used when there is limited hepatic involvement. Where there are multifocal tumors and/or vascular involvement, the role of these options becomes more limited. Most recently, interest has been directed toward sorafenib, a tyrosine-kinase inhibitor that was administered in patients with advanced HCC (defined as patients who were not eligible for or had disease progression after surgical or locoregional therapies) in the SHARP trial. In the treatment arm, progression-free survival and overall survival of 5.5 and 10.7 months (compared with 2.8 and 7.9 months for the placebo group) were achieved, respectively.11 In a US Food and Drug Administration-analyzed independent trial of 602 patients with unresectable HCC treated with sorafenib, a statistically significant survival advantage was demonstrated in the sorafenib arm compared with the placebo arm (median survival, 10.7 vs 7.9 months; P = .00058).12

In our study, the cohort of 72 patients with unresectable HCC of were mainly Child-Pugh class A or B. The majority of patients had multifocal HCC, with some having a single dominant large lesion and several other smaller satellite lesions, and others having a single large HCC deemed unresectable and unsuitable for ablation. Six patients had extrahepatic disease and were AJCC stage IV. These 6 patients were selected for inclusion in the study as they had asymptomatic extrahepatic disease and were treated primarily because of symptomatic intrahepatic disease. After iodine-131-Lipiodol treatment, the median survival time was 14 months, with 1-, 2-, and 3-year survival rates of 52%, 33%, and 20%, respectively. Disease progression occurred after a median of 6 months. In patients who were favorable responders, this opened up further therapeutic opportunities for which resection, transplantation, or ablation were offered to some patients that led to long-term survival. The 3 staging systems applied to this cohort—AJCC, Barcelona Clinic Liver Cancer, and Cancer of the Liver Italian Program—all predicted for survival. It was evident from our experience that patients with larger tumors (>4 cm) have poorer survival after treatment, suggesting that response of large tumors to iodine-131-Lipiodol treatment is less effective. In addition, the 6 patients with extrahepatic disease survived a median of 2 months. The poor outcome reiterates that this liver-specific locoregional treatment should be restricted to patients with intrahepatic disease. For patients with disseminated metastases, an alternative treatment administered systemically might prove to be more appropriate. To gauge the overall outcome from treatment, the response to treatment was identified to be the best predictor. Favorable responders did significantly better than patients who had unfavorable responses. Where favorable responses are achieved, this may open the option of repeated treatment, if there is adequate lung shunting, or for pursuit of other therapies. It appeared that AJCC stage 1-2 and Cancer of the Liver Italian Program score ≤2 were independent predictors for a favorable response. It must be emphasized that delivering radiation therapy to the liver may have potentially detrimental effects if patients (especially those with cirrhosis) are not judiciously selected. In our series, there was 1 treatment-related mortality from hepatic failure, 1 case of hepatic thrombosis after 3 months, and 12 patients who had symptoms of hepatic decompensation manifested by symptomatic ascites.

Four recent studies reported outcomes after iodine-131-Lipiodol in unresectable HCC with median survival ranging between 7 and 27 months.13-16 The heterogenous survival results reflect the selection of patients for treatment. In the study by Boucher et al.,14 a significant proportion of patients underwent 2 treatments, had a Cancer of the Liver Italian Program score of 0-1, and had 1 or 2 tumors. This highly selected group of patients had favorable characteristics, hence a median survival of 27 months was achieved in this study. In the studies by Borbath et al.13 and Kanhere et al.,15 the treated population appeared to have baseline characteristics more similar to those of the patients reported in our study. Their outcomes were median survival of 7 months and 15 months, respectively. The poorer survival outcome in the study by Borbath et al. was likely to be attributed to the finding that a third of patients in that study group had evidence of portal vein thrombosis. As shown in the Kaplan-Meier plot of the results of our current study, the steep decline in the initial follow-up period reflects the group of patients who despite treatment developed progressive disease or had stable disease. It was likely that in this group, which had poor prognostic factors identified through the univariate analysis, a more appropriate endpoint might perhaps have been symptom palliation rather than survival.

Iodine-131-Lipiodol is effective and has been shown in a small randomized trial of patients with portal vein thrombosis to be superior over best supportive care, achieving a 6-month survival rate of 48% in the treatment arm compared with 0% in the best supportive care arm, hence establishing itself as an effective palliative treatment.17 In a randomized comparative trial of iodine-131-Lipiodol versus chemoembolization, the 3-year survival rate in the iodine-131-Lipiodol arm was 14% and in the chemoembolization arm was 3%, without any statistical difference in actuarial survival outcomes. However, tolerance appeared to be significantly better in the iodine-131-Lipiodol group, with fewer patients suffering from severe side effects and vascular complications.18 Marelli et al., in a retrospective comparative trial, also demonstrated similar survival outcomes, but again showed that iodine-131-Lipiodol was superior to TACE in patients with advanced clinical staging or portal vein thrombosis.16 However, the constraints surrounding the need for a period of hospital admission after the procedure in a radioisolation facility has made TACE a more attractive and evidence-based therapeutic option, for which it has been further evaluated in 3 randomized trials comparing against best supportive/medical care.19-21 It is unlikely that a comparative trial between iodine-131-Lipiodol and TACE is feasible in this present day.

In conclusion, iodine-131 Lipiodol is an appropriate palliative treatment with encouraging survival outcomes for patients with unresectable exclusively intrahepatic HCC. The best achieved therapeutic will likely be observed in patients with multiple small tumors and who have good liver function. In the future, an effective palliative therapy such as iodine-131-Lipiodol should be evaluated in a clinical trial that incorporates molecular-targeted therapies such as sorafenib.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Dr. Terence Chua is a surgical oncology research scholar funded by the St. George Medical Research Foundation.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES