’Original article A training program for primary care physicians improves the effectiveness of ultrasound surveillance of hepatocellular carcinoma Paolo Del Poggioa, Stefano Olmia, Francesca Ciccaresea, Marzio Mazzolenid, Michele Jazzettid, Carlo Jamolettib, Mario Mattielloc, Anna del Poggioc, Vanessa Portugalic and Tommaso Stroffolinie Background and aims Cirrhosis is the main risk factor of hepatocellular carcinoma (HCC), but only a minority of cirrhotic patients are referred to the hepatologist by primary care physicians (PCP) and receive regular ultrasound surveillance. The aim of this study was to determine whether a training program targeted to PCP could enhance the effectiveness of surveillance in a reallife setting. Patients and methods A total of 120 PCP in an Italian area with a high incidence of HCC were trained to identify cirrhotic patients, to refer them to the Hepatology Centers of the area, and to start regular ultrasound surveillance. Clinical characteristics, outcome of treatments, and survival of 190 consecutive HCC patients in the same centers after training were compared with 244 HCC referred from the same area before training, and to 232 HCC referred by untrained PCP from other areas. Results Trained PCP referred significantly more HCC patients detected under surveillance and at an early stage (Barcelona Clinic Liver Cancer-A), suitable for radical treatments. In the intervention area, the 3 and 5-year survival of HCC patients increased after training from 35 to 48% and from 20 to 40%, respectively (P < 0.05). In contrast, survival was unchanged in the other areas. At multivariate analysis, independent predictors adversely affecting survival were Child–Pugh B-C, α-fetoprotein more than 10 ng/ dl, nonviral etiology, intermediate/advanced Barcelona Clinic Liver Cancer stage, and referral by an untrained PCP. Conclusion Specific training of PCP aimed at the identification and referral of cirrhotic patients efficiently improves HCC survival. Eur J Gastroenterol Hepatol 27:1103–1108 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Introduction

Hepatocellular carcinoma (HCC) is one of the most common solid malignancies worldwide and the main cause of mortality of cirrhotic patients [1]. Despite advances in medical technology, its prognosis remains poor because the majority of HCC are diagnosed at a late stage and in patients with advanced liver disease. Several cohort studies and one randomized case–control study have shown that regular ultrasound surveillance in high-risk populations can detect the tumor at an early and curable stage and increase survival [2–6]. For these reasons, all international guidelines recommend starting a regular semi-annual or European Journal of Gastroenterology & Hepatology 2015, 27:1103–1108 Keywords: cirrhosis, hepatocellular carcinoma, primary healthcare, ultrasound surveillance a Hepatology Unit, Policlinico S. Marco, Zingonia, bHepatology Unit, Azienda Ospedaliera di Treviglio, cRadiology Department, Azienda Ospedaliera di Treviglio, d Primary Care Physician, ‘Bassa Bergamasca Area’ Health District, Bergamo and e Department of Infectious and Tropical Diseases, Policlinico Umberto Primo, University of Rome, Rome, Italy

Correspondence to Paolo Del Poggio, MD, Unità di Epatologia, Policlinico San Marco, Zingonia, Corso Europa Unita 7, 24040 Zingonia-Osio Sotto, Bergamo, Italy Tel: + 39 335409229; fax: + 39 0363352889; e-mail: [email protected] Received 14 November 2014 Accepted 1 May 2015 Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's website (www.eurojgh.com).

annual ultrasound surveillance in these patients [1,7]. Cirrhosis, irrespective of etiology, is the main risk factor for the development of HCC. The identification of cirrhotic patients is therefore a prerequisite for the successful outcome of surveillance programs at the population level. Primary care physicians (PCP) play a pivotal role in the identification of cirrhotic patients and in supervising their adherence to surveillance. However, the clinical diagnosis of cirrhosis is difficult as the majority of patients are asymptomatic [8] and even after cirrhosis has been diagnosed, only a minority of patients are referred to a gastroenterologist/hepatologist and receive regular ultrasound surveillance [9]. The performance of PCPs can be ameliorated by training interventions [10] and by participation in specialist networks, as shown by the French experience in hepatitis C [11]. The aim of this study was to determine whether a specific training program targeted to PCP and directed at the identification, referral, and surveillance of cirrhotic patients in an area at high mortality for HCC [12] could lead to cancer detection at an early stage and improve patient survival. Patients and methods Area of intervention

The study was carried out in the southern part of the Bergamo Province (northern Italy), namely, in the Health District ‘Bassa Bergamasca’, a mixed rural and industrial

0954-691X Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

DOI: 10.1097/MEG.0000000000000404

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

1103

1104

September 2015 • Volume 27 • Number 9

European Journal of Gastroenterology & Hepatology

area of 200 000 inhabitants with two middle-sized hospitals (Policlinico S. Marco, Zingonia and Azienda Ospedaliera di Treviglio) and 120 PCP working in that area. Each hospital has an outpatient liver clinic and radiological plus surgical facilities to provide comprehensive care for HCC patients, although more complex cases are referred to tertiary care hospitals. Since 1998, the two liver clinics have established a network of collaboration with PCP of this area for the management and referral of patients with liver disease (for details, see Supplementary File 1, Supplementary digital content 1, http://links.lww. com/EJGH/A39). In addition to patients referred by PCP of the same area, the two clinics also provide care for patients referred from the nearby provinces, where no specific network for the management of liver disease has been implemented. The area where the study was carried out and the nearby areas are shown in Supplementary Fig. 3 (Supplementary digital content 2, http://links.lww. com/EJGH/A40). Training program of primary care physicians

In May 2006, we started a training program targeted at the 120 GPs of the area, aimed at the identification of cirrhotic patients and their surveillance for HCC. The program is described in detail in the supporting document n.1. Basically, PCP were instructed to perform an opportunistic screening of all patients at risk for liver cirrhosis, to refer them to the hepatologist to confirm the diagnosis, and to start regular ultrasound surveillance. The flow of the patients and the criteria for referral and follow-up by PCP are presented in Fig. 1. The workshops and the annual symposia were considered as required hours of continuous medical education by the local Health Authority and therefore all PCP who participated in the project. Evaluation of the effectiveness of the program

We analyzed all consecutive newly diagnosed HCC cases observed at the two liver clinics (Policlinico S. Marco, Zingonia and Azienda Ospedaliera di Treviglio) from January 1994 through June 2006 and from July 2006 through December 2013. January 1996 was chosen because from that date onwards, all cases of HCC were recorded in the two Hepatology Centers to be entered into a National Database. July 2006 was chosen because at that time, the official project had been presented to all PCP of our area in the first two workshops. We analyzed all consecutive HCC according to the referring physician: (1) Trained PCP if they were referred by PCP from the intervention area. (2) Untrained PCP if they were referred by PCP from other areas. The clinical characteristics, type of treatment, and outcome of HCC referred by PCP in the intervention area after implementation of the program were compared with: (1) All HCC referred by PCP of the same area before the implementation of the training program (internal control group).

(2) All HCC referred by PCP outside our area after implementation of the training program (external control group).

HCC diagnosis and treatment

The diagnosis of HCC was made on the basis of histology or by imaging according to the different EASL/AASLD guidelines published over time. For the purpose of this study, HCC was staged according to the Barcelona Clinic Liver Cancer (BCLC) classification [13] as follows: very early stage (stage 0): single nodule < 2 cm in a Child A patient – early stage (stage A): single nodule < 5 cm or 3 nodules ≤ 3 cm diameter in a Child A-B patient – intermediate stage (stage B), multinodular tumor in a Child A-B patient – advanced stage (stage C): portal invasion or distant metastases in a Child A-B patient – terminal stage (stage D): any tumor in a Child C patient and/or very poor performance status. According to the 2012 EASL-EORTC guidelines [1] and the more recent ECMO-ESMO guidelines [14], unifocal tumors larger than 5 cm were no longer considered stage A, but included in stage B. We defined the modality of diagnosis of HCC as follows: it was considered diagnosed under surveillance if an ultrasound examination with a normal result was performed in the 6–12 months preceding the diagnosis of HCC. It was considered diagnosed by symptoms if the diagnostic examination was triggered by tumor-related symptoms and occasionally in the remaining cases. The efficacy of treatment was evaluated after the first treatment at diagnosis according to the mRECIST criteria – that is, complete resection of the tumor or absence of contrast enhancement on computed tomography or MRI imaging performed 1 month after ablation or transcatheter arterial chemoembolization [15]. Statistical analysis

The methodology used in statistical analysis is reported in additional file 2, Supplementary digital content 3 (http:// links.lww.com/EJGH/A41). Results

Characteristics of the cirrhotic patients who developed HCC. A total of 566 cirrhotic patients were newly diagnosed with HCC at the two referral hospitals of the area from January 1996 through December 2013. In all, 434 were referred by PCP of the same area and 132 from PCP of other areas. The demographic and clinical characteristics of the patients are presented in Table 1. The majority of the patients were males with Child A cirrhosis of viral etiology. Cirrhosis was confirmed histologically in almost 30% of the cases and clinically or radiologically in the others. Esophageal varices and comorbidities were present in one-third and half of the cases, respectively. No significant differences existed in the baseline characteristics of cirrhotic patients who developed HCC between the two areas (Table 1).

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

Training family physicians improves HCC survival Del Poggio et al.

www.eurojgh.com

1105

Patients at risk for liver cirrhosis∗

Screened by PCP with criteria (a—c): (a) Clinical signs of cirrhosis (b) Two US signs of cirrhosis∗∗ (c) One US sign + platelet count < 130 000/dl

At least 1 criteria present

Hepatologist referral

Sent back to PCP for clinical follow-up and US surveillance

No criteria present

Continues follow-up by PCP according to local guidelines, repeat screening test every year

Cirrhosis confirmed if: - Liver surface nodularity on US + APRI > 2 - Liver stiffness > 12 kPa on transient elastography - Liver biopsy Fig. 1. Flow of the patients and follow-up criteria according to the training program implemented in the intervention area. *Patients were considered at risk for liver cirrhosis in the presence of one or more of the following risk factors: HCV infection, HBsAg positivity, alcohol abuse (>60 g/day in women and > 80 g/day in men), BMI > 30 kg/m2. **Ultrasound (US) signs of liver cirrhosis: liver surface nodularity, nonhomogeneous liver texture, left lobe or caudate lobe hypertrophy, distorted hepatic veins, spleen longitudinal diameter more than 12 cm. APRI, aspartate aminotransferase-to-platelet ratio index; HBsAg, hepatitis B surface antigen; HCV, hepatitis C virus; PCP, primary care physicians.

Table 1. Demographic and clinical characteristics of 566 consecutive cirrhotic patients diagnosed with HCC at the two referral hospital of the area (1994–2013) Training area

Median age (range) (years) Mean age (± SD) (years) Sex [n (%)] Male Female Etiology of cirrhosis [n (%)] HBV HCV Alcohol NAFLD Combined Others Comorbid illness Histological diagnosis of cirrhosis Antiviral therapya Histological diagnosis of HCC Child–Pugh [n (%)] A B C Esophageal varices [n (%)] Platelet count × 103 (median and range)

Other areas

Group A (n = 244)

Group B (n = 90)

Group C (n = 81)

Group D (n = 51)

69 (47–85) 68 ± 8.7

72 (25–88) 71 ± 6.7

67 (28–88) 67 ± 9.6

72 (53–87) 71 ± 7.8

187 (77) 57 (23)

136 (71) 54 (29)

65 (80) 16 (20)

32 (63) 19 (37)

16 112 38 8 58 12 109 54 37 147

(6) (46) (15) (3) (23) (5) (45) (23) (15) (60)

21 92 41 8 25 3 109 81 49 31

(11) (48) (21) (4) (13) (1) (57) (43) (26) (16)

5 43 10 1 17 5 44 26 11 33

6 25 7 3 9 1 27 11 12 11

184 49 11 79 126

(75) (20) (5) (32) (23–517)

160 23 7 52 131

(84) (12) (4) (27) (20–441)

57 16 8 26 127

(6) (53) (12) (2) (20) (6) (54) (32) (13) (40) (70) (19) (11) (32) (26–323)

41 8 2 16 141

(11) (49) (13) (5) (17) (2) (53) (21) (23) (21) (80) (16) (4) (31) (38–346)

Groups A and C: HCC referred before training, groups B and D: HCC referred after training. HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; NAFLD, nonalcoholic fatty liver disease; PCP, primary care physicians. a Under or past treatment with interferon + ribavirin or nucelos(t)ide analoges.

Modality of diagnosis and stage of HCC

The modality of HCC diagnosis in the area of intervention and in the other areas before and after implementation of the project is presented in Supplementary Fig. 4 (Supplementary digital content 4, http://links.lww.com/ EJGH/A42). In the intervention area, 85 (34.8%) out of 244 HCC were diagnosed during surveillance before training and 105 (55%) out of 190 after training (Δ =

+ 20.5%, P < 0.001), whereas in the other areas, 21 (25.9%) out of 81 were diagnosed during surveillance in the first time period and 20 (39%) out of 51 in the second time period (Δ = + 13.1%, P = 0.11, nonsignificant). After implementation of the program, a significantly higher proportion of HCC referred by trained PCP was diagnosed during surveillance compared with untrained PCP (55 vs. 39%, P < 0.04). Table 2 shows the stage of the tumors at

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

1106

September 2015 • Volume 27 • Number 9

European Journal of Gastroenterology & Hepatology

Table 2. Stage of HCC at diagnosis in the intervention and other areas before (groups A and C) and after the implementation of training (groups B and D) Training area [n (%)]

BCLC stage Stage 0–A (single nodule < 5 cm, less than three nodules < 3 cm) Very early (stage 0) Early (stage A) Stage B (multifocal) Stage C (advanced) Stage D (terminal)

Other areas [n (%)]

Group A (n = 244)

Group B (n = 190)

Group C (n = 81)

Group D (n = 51)

116 (47.5)

122 (64.2)

31 (38.2)

22 (43.1)

21 95 76 23 29

(8.6) (39) (31.1) (9.4) (11.8)

29 93 39 17 12

(15.2) (48.9) (20.5) (8.9) (6.3)

Table 3. Type of treatment of HCC according to BCLC stage in the intervention and other areas before (A and C) and after implementation of the training program (B and D)

31 17 13 20

0 (38.2) (20.9) (16) (24.6)

8 14 19 4 6

(15.6) (27.4) (37.2) (7.8) (11.7)

the time of diagnosis. The proportion of HCC diagnosed at an early stage (BCLC-A) increased from 47.5% (116 out of 244) to 64.2% (122 out of 190) in cases referred by trained PCP (Δ = + 16.7%, P < 0.001) and from 38.2% (31 out of 81) to 43.1% (22 out of 51) in cases referred by untrained PCP (Δ = + 4.9%, P = 0.7, nonsignificant). In the second time period, a significantly higher proportion of HCC referred by trained PCP was diagnosed at an early stage compared with untrained PCP (64.2 vs. 43.1%, P < 0.02). Conversely, the proportion of cases diagnosed as intermediate stage (BCLC-B) decreased from 31.1% (76 out of 244) to 20.5% (39 out of 190) in case of referral by trained PCP (Δ = − 10.6%, P < 0.02), whereas it paradoxically increased from 20.9% (17/81) to 37.2% (19/51) if the patient was referred by untrained PCP (Δ = + 16.3%, P < 0.02). The percentage of stage C and D tumors in this latter group decreased from 40.6% (33/81) to 19.5 (10/51, Δ = + 21.1%, P < 0.02) in the second time period. Treatments

The type and results of first treatments of newly diagnosed HCC are presented in Table 3. The treatment strategy in the two referral centers before and after the training period and for each stage of HCC is reported in Table 3. No significant differences in treatment strategy were observed between the two areas. The proportion of cases receiving potentially curative treatments – for example, surgical resection, thermal ablation, liver transplantation, and percutaneous ethanol injection – increased from 44.6% (109/244) to 60.5% (115/190) in case of HCC referred by trained PCP (Δ = + 16%, P = 0.0015) and from 40.7% (33/81) to 47% (24/51) if the patient was referred by untrained PCP (Δ = + 6.6%, P = 0.59, nonsignificant). The complete response rate after treatment, as defined by modified RECIST criteria, increased from 33% (81/244) to 50% (96/190) after implementation of the program (Δ = + 23%, P < 0.001) in case of HCC referred by trained PCP, whereas it remained stable – for example 32% (26/81) and 37% (19/51) – in case of HCC referred by untrained PCP (Δ = + 5%, P = 0.67, nonsignificant). Survival

The overall survival curves of HCC patients are presented in Fig. 2. Progression of HCC was the most common cause

Training area [n (%)]

BCLC stage BCLC stage 0–A (n) Surgical resection Thermal ablation (radiofrequency/ microwave) Liver transplantation Percutaneous alcohol injection (PEI) Transarterial chemoembolization (TACE) Combined PEI + TACE Systemic treatments** No treatment/dropout BCLC stage B (n) Surgical resection Thermal ablation TACE/radioembolization Combined PEI + TACE PEI Systemic treatments** No treatment/dropout BCLC stage C–D (n) Surgical resection TACE/radioembolization PEI Thermal ablation Systemic treatments** No treatment/palliation

Other areas [n (%)]

Group A (N = 244)

Group B (N = 190)

Group C (N = 81)

Group D (N = 51)

116 12 (10) 41 (35)

122 30 (24) 54 (44)

31 8 (25) 14 (45)

22 4 (18) 12 (54)

1 (0.8) 23 (20)

20 (16)

4 (13)

4 (18)

14 (12)

7 (5.7)

4 (13)

15 (13) 7 (6) 3 (2) 76 6 (7.8) 17 8 3 37 5

(22) (10) (4) (48) (6.5) 52 1 (1.9)

1 (4) 11 (9) 39 7 (18) 1 (2.5) 17 (43) 3 (7.6) 7 (18) 4 (10) 29 2 (7)

4 (7.6) 32 (61) 15 (29)

2 (7) 23 (79) 2 (7)

17 2 (12) 1 (6) 3 (18) 2 (12) 1 (6) 7 (11) 1 (6) 33 1 (3) 1 (3) 1 (3) 23 (69) 7 (21)

2 (9) 19 3 (15) 1 (5) 8 (42)

6 (31) 1 (5) 10

7 (70) 3 (30)

BCLC, Barcelona Clinic Liver Cancer; PEI, percutaneous ethanol injection. **sorafenib, tamoxifene, megestrol or palliative care.

of death (348/404: 86%); 28 (7%) died because of gastroesophageal variceal bleeding or hepatic failure and 28 died (7%) because of other diseases. Three-year and 5-year survival of HCC patients referred by trained PCP increased after training, respectively, from 35 to 48% and from 20 to 40% (P < 0.05 log-rank test), whereas survival was unchanged in cases referred by untrained PCP [3-year survival 29% in the first time period, 31% in the second (P = 0.87, nonsignificant), 5-year survival 20% in both time periods]. In the first time period, no difference in survival was detected between HCC referred by trained and untrained PCP (P = 0.44, log-rank test), whereas after implementation of the training program, survival of HCC patients referred by PCP in the intervention area was significantly higher compared with HCC referred by PCP from other areas (P < 0.04 log-rank test). The percentage of patients who were lost at follow-up after 3 years was low: in the first time period, 0.4% (9/244) in the intervention area versus 6.1% (5/81) in the other areas (χ2: P < 0.05), in the second time period, 9% in the intervention area (17/190) versus 17% (9/51) in the other areas (χ2: P = 0.12, nonsignificant). The dropout rate was not calculated for 5-year survival because a significant number of patients – that is, those diagnosed after 2008 – had a shorter follow-up. At univariate analysis (Supplementary Table 4, Supplementary digital content 5, http://links.lww.com/ EJGH/A43), male sex, nonviral etiology, low platelet count, Child–Pugh B-C, α-fetoprotein levels more than 10 ng/dl, presence of comorbidities, diagnosis outside

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

Training family physicians improves HCC survival Del Poggio et al.

(a)

(b)

First time period (January 1994—June 2006) Survival probability (%)

www.eurojgh.com

100

Second time period (July 2006—December 2013) 100 80

80 Group 1 2

60 40 20

Group 1 2

60 40 20

0

0 0

50

100

150

200

0

20

Time Group 1 Group 2

1107

244 81

54 17

10 6

40

60

80

100

16 3

3 0

0 0

Time 3 0

0 0

Group 1 190 Group 2 51

84 13

43 5

Fig. 2. Kaplan–Meier survival curves of HCC patients. (a) In the first time period, no difference in survival was detected between HCC referred by PCP of the

same area (group 1) and of other areas (group 2), P = 0.44, log-rank test. (b) In the second time period – for example, after implementation of the training program – survival of HCC patients referred by PCP of the same area (group 1) was significantly better compared with HCC referred by PCP of other areas (group 2), P < 0.05, log-rank test. Short bars in the graphs indicate censored data. HCC, hepatocellular carcinoma; PCP, primary care physicians.

ultrasound surveillance, intermediate/advanced BCLC stage, and referral by an untrained PCP were significantly associated with lower overall survival. After adjustment for the confounding effect of the considered variables by multivariate Cox model analysis, Child–Pugh B-C, α-fetoprotein more than 10 ng/dl, nonviral etiology, intermediate/advanced BCLC stage, and referral by untrained PCP were all independent predictors of the likelihood of lower survival. Male sex, low platelet count, presence of comorbidities, and diagnosis outside ultrasound surveillance were no longer associated. Discussion

This is a real-life study on the effectiveness of surveillance of HCC at the primary care level. Although all international guidelines released from 2001 onwards recommend regular ultrasound surveillance of cirrhotic patients for the detection of HCC at an early stage, uptake of surveillance at the population level is low [16,17]. In one study from 11 cancer registries of the USA during 1996–2002, only 29% of 541 patients with a recorded diagnosis of cirrhosis for 3 or more years before HCC were receiving regular ultrasound surveillance [18]. This figure is similar to that observed in our study, in which only 31.6% out of 335 HCC diagnosed before the implementation of the training program were found during ultrasound surveillance. One main finding of our study is that after training, the proportion of HCC cases detected during surveillance increased to 55% in the area of intervention, but only to 39% in the other areas. It is intuitive to ascribe the modest increase observed in the latter areas to a better knowledge of the guidelines on liver disease derived from continuous medical education and personal reading. It is nonetheless important to underline that by means of targeted training, we could achieve a much better rate of ultrasound surveillance utilization than that obtained with conventional medical education. This better performance translated into an earlier stage migration and more effective treatment of detected HCC, resulting in improved overall survival. It is clear that all studies on surveillance are affected by the so-called ‘lead time bias’, which is an overestimation of survival in the screened group because of detection of the

tumor at an earlier stage. However, it has been shown in a large cohort study that the survival benefit is real and is still significant even after statistical correction for this bias [5]. Our data further underscore the need to expand the utilization of ultrasound surveillance in the population at risk. Cirrhosis is the condition at highest risk of developing HCC compared with hepatitis and fatty liver disease [1] and the application of surveillance in this setting can be considered cost-effective [19]. The effectiveness of surveillance of HCC can be ameliorated by improving the surveillance tool by increasing patient adherence and by fostering the identification of atrisk individuals. Ultrasound is, to date, the only approved tool by international guidelines as α-fetoprotein is no longer endorsed by the last EASL-EORTC guidelines [1] and the addition of the new serum oncomarkers or the variation of α-fetoprotein over time are still under study [20,21]. Recent data from a Spanish hospital network showed good adherence to ultrasound surveillance, with the exception of patients with alcohol-related and drugrelated problems [22]. In the study by Davila et al. [18], patient groups that were more likely to receive regular surveillance included women, younger people, Asians, and those living in urban or higher income areas. In our study, we did not address the issue of adherence or of the surveillance tool, but focused on improving the detection of patients at risk by providing guidance to PCP to increase both detection of cirrhosis and referral of the patient for specialist evaluation. Our endeavor aimed at establishing a network in which cirrhotic patients are identified by PCP, referred to the specialist, and sent back to PCP for surveillance and supervision. A limitation of our study is the possibility that referral bias may have altered our results. It is likely that PCP may have referred younger and fitter HCC patients to tertiary care hospitals for transplant evaluation, thus bypassing the two centers of the area, as suggested by the negligible proportion of transplanted patients. It is unlikely, however, that PCP working in the other areas may have preferentially referred advanced HCC to our centers, first, because the other areas are included within the basin of our two hospitals, second because no differences were found in the baseline characteristics of HCC detected in

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

1108

September 2015 • Volume 27 • Number 9

European Journal of Gastroenterology & Hepatology

the two areas, and third because the 3-year survival of HCC referred from the other areas was similar to those observed throughout the country in a large multicentric Italian real-life study [23]. It is noteworthy that in the same study, the proportion of transplanted patients was as low as in our study, making our results generalizable to a reallife population. We nonetheless took care to correct for referral bias by comparing the characteristics of HCC detected in the intervention area not only with other areas but also with HCC detected in the same area before the implementation of the training program. We found that the diagnostic and clinical skills of PCP improved much more in the intervention area as the proportion of HCC diagnosed under surveillance increased significantly more in the former (Δ = + 20.5%, P < 0.001) than in the latter areas (Δ = + 13.1%, P = 0.11, nonsignificant). We can therefore conclude that, if any referral bias was operating, it did not affect the main findings of our study, being unlikely that the referral behavior of PCP may have changed over time.

4

5

6

7

8

9

10

11

Conclusion

We showed that improving the detection rate of cirrhotic patients improves HCC survival. Establishing a collaborative network between specialists and PCP to identify, refer, and follow at-risk individuals is a crucial step to detect HCC at an early stage and improve patient survival.

12

13 14

Acknowledgements

The authors thank all primary care physicians of the ‘Area Bassa Bergamasca’ who participated in the project and Mr Stan Ajello for language revision of the manuscript. Author contributions: study concept and design: Paolo Del Poggio, Marzio Mazzoleni, Tommaso Stroffolini; acquisition of data: all authors; analysis and interpretation of data: all authors; drafting of the manuscript: Paolo Del Poggio; critical revision of the manuscript for important intellectual contents: Tommaso Stroffolini; statistical analysis: Francesca Ciccarese, Anna Del Poggio; study supervision: Tommaso Stroffolini.

15 16

17

18

19

Conflicts of interest

There are no conflicts of interest. 20

References 1 European Association For The Study Of The Liver; European Organisation For Research And Treatment Of Cancer. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol 2012; 56:908–943. 2 Wong LL, Limm WM, Severino R, Wong LM. Improved survival with screening for hepatocellular carcinoma. Liver Transpl 2000; 6:320–325. 3 Bolondi L, Sofia S, Siringo S, Gaiani S, Casali A, Zironi G, et al. Surveillance programme of cirrhotic patients for early diagnosis and treatment of hepatocellular carcinoma: a cost effectiveness analysis. Gut 2011; 48:251–259.

21

22

23

Pateron D, Ganne N, Trinchet JC, Aurousseau MH, Mal F, Meicler C, et al. Prospective study of screening for hepatocellular carcinoma in Caucasian patients with cirrhosis. J Hepatol 1994; 20:65–71. Trevisani F, de Notariis S, Rapaccini G, Farinati F, Benvegnù L, Zoli M, et al. Semiannual and annual surveillance of cirrhotic patients for hepatocellular carcinoma: effects on cancer stage and patient survival (Italian experience). Am J Gastroenterol 2002; 97:734–744. Zhang BH, Yang BH, Tang ZY. Randomized control trial of screening for hepatocellular carcinoma. J Cancer Res Clinical Oncol 2004; 130:417–422. Bruix J, Sherman M. American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update. Hepatology 2011; 53:1020–1022. Graudal N, Leth P, Marbjerg L, Galloe AM. Characteristics of cirrhosis undiagnosed during life: a comparative analysis of 73 undiagnosed cases and 149 diagnosed cases of cirrhosis, detected in 4929 consecutive autopsies. J Intern Med 1991; 230:165–171. El-Serag HB, Kramer JR, Chen GJ, Duan Z, Richardson PA, Davila JA. Effectiveness of AFP and ultrasound tests on hepatocellular carcinoma mortality in HCV-infected patients in the USA. Gut 2011; 60:992–997. Cozzolongo R, Cuppone R, Petruzzi J, Stroffolini T, MAnghisi O. Approach of primary care physicians to hepatitis C: an educational survey from a Southern Italian area. J Infect 2005; 51:396–400. Delarocque-Astagneau E, Meffre C, Dubois F, Pioche C, Le Strat Y, Roudot-Thoraval F, et al. The impact of the prevention program of hepatitis C over more than a decade: the French experience. J Viral Hepatitis 2010; 17:435–443. Meroni G, Silini E. Primary liver cancer is presently the main cause of death for liver disease in Bergamo, Italy. J Hepatol 1999; 30 (Suppl 1):104. Llovet JM, Brù C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis 1999; 19:329–338. Verslype C, Rosmorduc O, Rougier P. on behalf of the ESMO Guidelines Working Group on. Hepatocellular carcinoma: ESMO–ESDO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol 2012; 23 (Suppl 7):41–48. Lencioni R, Llovet JM. Modified RECIST (mRECIST) assessment for hepatocellular carcinoma. Semin Liver Dis 2010; 30:52–60. Davila JA, Henderson L, Kramer JR, Kanwal F, Richardson PA, Duan Z, El-Serag HB. Utilization of surveillance for hepatocellular carcinoma among hepatitis C virus-infected veterans in the United States. Ann Intern Med 2011; 154:85–93. Singal AG, Yopp A, Skinner C, Packer M, Lee WM, Tiro JA. Utilization of hepatocellular carcinoma surveillance among American patients: a systematic review. J Gen Intern Med 2012; 27:861–867. Davila JA, Morgan RO, Richardson PA, Du XL, Mc Glynn KA, El Serag HB. Utilization of surveillance for hepatocellular carcinoma among patients with cirrhosis in the United States. Hepatology 2010; 52:132–141. Thompson Coon J, Rogers G, Hewson P, Wright D, Anderson R, Cramp M, et al. Surveillance of cirrhosis for hepatocellular carcinoma: systematic review and economic analysis. Health Technol Assess 2007; 11:1–206. Lee E, Edward S, Singal A, Lavieri MS, Volk M. Improving screening for hepatocellular carcinoma by incorporating data on levels of alphafetoprotein over time. Clin Gastroenterol Hepatol 2013; 11:437–440. El Serag H, Kanwal F. Alpha-fetoprotein in hepatocellular carcinoma surveillance: mend it but do not end it. Clin Gastroenterol Hepatol 2013; 11:441–443. Gonzalez-Diéguez ML, Mancebo A, Cadahia V, Perez R, Varela M, Navascues C, Rodriguez M. Analysis of adherence to a hepatocellular carcinoma surveillance programme based on biannual controls. J Hepatol 2014; 60:S259. Santi V, Buccione D, di Micoli A, Fatti G, Frigerio M, Farinati F, et al. The changing scenario of hepatocellular carcinoma over the last two decades in Italy. J Hepatol 2012; 56:397–405.

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

A training program for primary care physicians improves the effectiveness of ultrasound surveillance of hepatocellular carcinoma.

Cirrhosis is the main risk factor of hepatocellular carcinoma (HCC), but only a minority of cirrhotic patients are referred to the hepatologist by pri...
237KB Sizes 0 Downloads 12 Views