Hepatocellular carcinoma screening in NAFLD: the paradox of nearly half the cases arising in non-cirrhotic low risk patients

Edi o ial Hepa ocellula ca cinoma sc eening in NAFLD: The pa adox o nea ly hal he cases a ising in non-ci ho ic low isk pa ien s Li e cance is in he op h ee leading causes o dea h om cance wo ldwide. Hepa ocellula ca cinoma (HCC) accoun s o 85% o p i- ma y li e cance s and de elops in pa ien s wi h ch onic li e disease (CLD). The p og essi e inc ease in he p e alence o nonalcoholic a y li e disease (NAFLD), which ecen ly eached up o 30% o he global popula ion [1], and he ecen e ec i e ea men o hepa i is C, esul s in NAFLD being he mos apidly inc easing e iology o HCC. Since he beginning o he millennium, he p opo ion o NAFLD-a ibu able HCC inc eased up o 8- old, cu en ly accoun ing o 15% o HCC cases [2]. NAFLD-associa ed HCC has he pa icula i y o being 5 imes mo e equen in he p e-ci ho ic phase o he disease, as compa ed wi h o he e iologies o CLD [3]. Ac ually, 2 ou o 5 pa ien s wi h NAFLD- associa ed HCC do no ha e ci hosis [3]. Impo an ly, he absence o ci hosis does no seem o independen ly impac su i al, which a he is dependen on he cance s age and ea men p o ided. F om ano he pe spec i e, conside ing he isk o NAFLD pa ien s de eloping HCC, hose wi hou ci hosis ha e 100 imes lowe isk: he HCC annual incidence a e is a ound 3.8% in pa ien s wi h ci ho- sis (simila o o he e iologies) and 0.03% in pa ien s in he p e-ci - ho ic s a e [4]. NAFLD-associa ed HCC, compa ed o o he e iologies, ends o occu la e in li e, in pa ien s wi h me abolic and ca dio ascula como bidi ies [2]. This may help explain why, in pa ien s wi h HCC and li e ci hosis, NAFLD pa ien s, as compa ed o o he e iologies, p esen a wo se p ognosis [2]. The ea men o hose pa ien s is no only jeopa dized by olde age and como bidi ies, NAFLD-associa ed HCC seems o be less esponsi e o immuno he apy [5]. Impo an ly, we a e ailing o sc een hese pa ien s, since mo e e- quen ly han in o he e iologies, NAFLD-associa ed HCC is de ec ed ou - side specific su eillance [2]. This canno be jus ified only by non- ci ho ic HCC, which would all ou o sc eening p og ams because i also happens in he con ex o ci hosis. Indeed, pa ien s wi h hepa i is C i us-associa ed ci hosis a e 2 imes mo e likely o be en olled in HCC sc eening p og ams han pa ien s wi h NAFLD-associa ed ci hosis [6]. Cu en sc eening ools pe o m wo se in pa ien s wi h NAFLD. Fo example, ul asound, he basis o he biannual p oposed HCC sc eening p o ocol, seems o be 3 imes mo e inaccu a e o de ec HCC in pa ien s wi h NAFLD-associa ed ci hosis compa ed o o he o ms o li e ci hosis [7], as s ea osis inc eases ul asound a enua- ion impai ing he de ec ion o deep li e nodules. Also, he e is a dose-dependen dec ease in ul asound accu acy wi h inc easing BMI [7]. Ul asound epo s should illus a e he possible limi a ions o isualiza ion acco ding o he US LI-RADS algo i hm, which s a i- fies in o minimal, mode a e, and se e e limi a ions. Pa ien s wi h low ul asound sco es would p obably benefi om o he imaging echniques such as CT and MRI, o he mos ecen ly p oposed abb e- ia ed MRI p o ocols. The combina ion o he ul asound wi h alpha- e op o ein (AFP) seems o inc ease by 20% he sensi i i y o HCC sc eening. Howe e , in non-hepa i is C i us ci hosis, a cu o o 11ng/mL may ou pe o m he classic 20ng/mL cu o [8]. Clinical-lab- o a o y sco es may ou pe o m AFP. One such sco e, al eady e alu- a ed in phase 2 s udies in NAFLD-associa ed ci hosis, is he GALAD ha inco po a es sex, age, and umo ma ke s, wi h an AUROC o 0.90. GALAD sco e may also be use ul o iden i y pa ien s a isk o de eloping HCC ha would benefi om being en olled in sc eening p og ams since high sco es ha e been de ec ed e en 1.5 yea s be o e he de elopmen o HCC [9]. Taking all in o conside a ion, when deciding o sc een pa ien s wi h HCC, we ace he pa adox o up o 40% o pa ien s wi h NAFLD- associa ed HCC no p esen ing li e ci hosis, while, he de elopmen o HCC in a pa ien wi h NAFLD wi hou ci hosis is a e y a e e en . Pa ien s wi h NAFLD-associa ed ci hosis, ha is, wi h li e s i ness measu emen (LSM) highe han 15kPa, should undoub edly be con- side ed o HCC sc eening, since i s annual incidence is highe han he 1.5% cu o o HCC sc eening cos -e ec i eness in pa ien s wi h ci hosis. No iceably, o he ac o s mus be aken in o conside a ion when en olling pa ien s in HCC sc eening p og ams, such as unc- ional s a us, o e all heal h and app op ia eness o HCC ea men i HCC is ound. Sc eening may be mo e expensi e in NAFLD-associa ed ci hosis, wi h a highe need o mo e sensi i e echniques such as abb e ia ed MRI due o less accu acy o ul asound in his se . As such, isk-s a ifica ion models such as he hcc isk [10] ha in eg a- es age, gende , BMI, diabe es-melli us, pla ele s coun , amino ans- e ases, and se um albumin, may help iden i y low isk pa ien s ha would no benefi om sc eening and high- isk pa ien s ha would benefi om mo e in ensi e sc eening s a egies. Rega ding p e-ci ho ic HCC, uni e sal sc eening is no cos -e ec- i e, he challenge being he iden ifica ion o high- isk popula ions. Indeed, non-ci ho ic NAFLD-a ibu ed HCC co esponds o 6% o HCC [2], and conside ing an es ima ed global incidence o o e 1 million cases by 2025, excluding non-ci ho ic NAFLD pa ien s om sc eening, would esul in 60000 HCC cases pe yea being missed om sc eening p og ams. The mos impo an isk ac o o HCC de elopmen in pa ien s wi h non-ci ho ic NAFLD is he p esence and se e i y o li e fib osis (assessed by his ology, non-in asi e sco es o LSM) [11].Indeed, cu en guidelines by Eu opean, Ame ican and Japanese socie ies o he s udy o he li e , al eady ecommend sc eening in pa ien s wi h F3 fib osis. O he isk ac o s a e olde age (being exceedingly a e in hose younge han 65 yea s old), he p esence o diabe es-melli us, pa icu- la ly hose wi h e hinopa hy [12], and inc eased amino ans e ase le - els [3]. Alcohol in ake, e en in he social ange, is a s ong isk ac o in h ps://doi.o g/10.1016/j.aohep.2023.101101 1665-2681/© 2023 Fundación Clínica Médica Su , A.C. Published by Else ie España, S.L.U. This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/) Annals o Hepa ology 28 (2023) 101101 Con en s lis s a ailable a ScienceDi ec Annals o Hepa ology jou nal homepage: www.else ie .es/annalso hepa ology ci ho ic pa ien s, whe eas in non-ci ho ic is con o e sial, albei he e is a known syne gism be ween alcohol and inc easing BMI [13].As such, p obably pa ien s sui ed o sc eening would be olde han 65 yea s old, wi h diabe ic e inopa hy, inc eased amino ans e ases, and some li e fib osis. Polygenic [14] and ansc i p ome [15] isk sco es ha e shown p omising esul s in s a i ying NAFLD pa ien s o HCC isk, e en in he p e-ci ho ic s a e, wi h he o me showing high pe o mance o selec ing pa ien s o sc eening, and he la e o excluding hem. In conclusion, HCC sc eening in NAFLD pa ien s is ailing o 3 main easons: 1) as onishingly unde -diagnosis o NAFLD-associa ed ci hosis o ad anced fib osis in he gene al popula ion, 2) imp essi e unde - sc eening o HCC in pa ien s al eady known o ha e NAFLD-associa ed ci hosis, and 3) a high p opo ion o HCC diagnosed in p e-ci ho ic pa ien s. The fi s p emise could be o e come wi h an ac i e sea ch o ad anced li e fib osis in high- isk popula ions such as obese and pa ien s wi h diabe es-melli us, as al eady p oposed in AASLD guide- lines. The second p emise wa an s highe awa eness om physicians ha ake ca e o hese pa ien s. Rega ding non-ci ho ic NAFLD pa ien s, he e is s ill a need o be e s a ifica ion ools such as clinical and polygenic sco es ha accu a ely iden i y pa ien s a isk o HCC, so we do no miss 1 in e e y 20 HCC cases wo ldwide. Decla a ion o in e es None. Re e ences [1] Hen y L, Paik J, Younossi ZM. Re iew a icle: he epidemiologic bu den o non- alcoholic a y li e disease ac oss he wo ld. Alimen Pha macol The 2022;56:942–56. h ps://doi.o g/10.1111/ap .17158. [2] Tan DJH, Ng CH, Lin SY, Pan XH, Tay P, Lim WH, e al. Clinical cha ac e is ics, su - eillance, ea men alloca ion, and ou comes o non-alcoholic a y li e disease- ela ed hepa ocellula ca cinoma: a sys ema ic e iew and me a-analysis. Lance Oncol 2022;32:521–30. h ps://doi.o g/10.1016/S1470-2045(22)00078-X. [3] Kanwal F, K ame JR, Mapakshi S, Na a ajan Y, Chayanupa kul M, Richa dson PA, e al. Risk o Hepa ocellula Cance in Pa ien s Wi h Non-Alcoholic Fa y Li e Disease. Gas oen e ology 2018;155:1828–37. h ps://doi.o g/10.1053/j.gas o.2018.08.024. [4] O ci LA, Sanduzzi-Zampa elli M, Caballol B, Sapena V, Colucci N, To es F, e al. Incidence o Hepa ocellula Ca cinoma in Pa ien s Wi h Nonalcoholic Fa y Li e Disease: A Sys ema ic Re iew, Me a-analysis, and Me a- eg ession. Clin Gas o- en e ol Hepa ol 2022;20:2830292. h ps://doi.o g/10.1016/j.cgh.2021.05.002. [5] Pfis e D, N u~ nez NG, Pinyol R, Go ae e O, Pin e M, Szydlowska M, e al. NASH limi s an i- umou su eillance in immuno he apy- ea ed HCC. Na u e 2021;592:450–6. h ps://doi.o g/10.1038/s41586-021-03362-0. [6] Wol E, Rich NE, Ma e o JA, Pa ikh ND, Singal AG. Use o Hepa ocellula Ca ci- noma Su eillance in Pa ien s Wi h Ci hosis: A Sys ema ic Re iew and Me a- Analysis. Hepa ology 2021;73:713–25. h ps://doi.o g/10.1002/hep.31309. [7] Simmons O, Fe ze DT, Yokoo T, Ma e o JA, Yopp A, Kono Y, e al. P edic o s o adequa e ul asound quali y o hepa ocellula ca cinoma su eillance in pa ien s wi h ci hosis. Alimen Pha macol The 2017;45:169–77. h ps://doi.o g/ 10.1111/ap .13841. [8] Gopal P, Yopp AC, Waljee AK, Chiang J, Neh a M, Kandunoo i P, e al. Fac o s ha a ec accu acy o a- e op o ein es in de ec ion o hepa ocellula ca cinoma in pa ien s wi h ci hosis. Clin Gas oen e ol Hepa ol 2014;12:870–7. h ps://doi. o g/10.1016/j.cgh.2013.09.053. [9] Bes J, Bechmann LP, Sowa JP, Sydo S, Dech^ ene A, Pflanz K, e al. GALAD Sco e De ec s Ea ly Hepa ocellula Ca cinoma in an In e na ional Coho o Pa ien s Wi h Nonalcoholic S ea ohepa i is. Clin Gas oen e ol Hepa ol 2020;18:728–35. h ps://doi.o g/10.1016/j.cgh.2019.11.012. [10] Ioannou GN, G een P, Ke KF, Be y K. Models es ima ing isk o hepa ocellula ca cinoma in pa ien s wi h alcohol o NAFLD- ela ed ci hosis o isk s a ifica- ion. J Hepa ol 2019;71:523–33. [11] Pons M, Ri e a-Es eban J, Manzano R, e al. Non-In asi e Tes s o Li e Fib osis Help in P edic ing he De elopmen o Hepa ocellula Ca cinoma among Pa ien s wi h NAFLD. J Clin Med 2022;11:2466. h ps://doi.o g/ 10.1016/j.jhep.2019.05.008. [12] Azuma S, Asahina Y, Kakinuma S, Azuma K, Miyoshi M, Inoue E, Tsunoda T, e al. Diabe ic Re inopa hy as a Risk Fac o Associa ed wi h he De elopmen o Hepa- ocellula Ca cinoma in Nonalcoholic Fa y Li e Disease. Dig Dis 2019;37:247– 54. h ps://doi.o g/10.1159/000493580. [13] Loomba R, Yang HI, Su J, B enne D, Ba e -Conno E, Iloeje U, e al. Syne gism be ween obesi y and alcohol in inc easing he isk o hepa ocellula ca cinoma: a p ospec i e coho s udy. Am J Epidemiol 2013;177:333–42. h ps://doi.o g/ 10.1093/aje/kws252. [14] Bianco C, Jamialahmadi O, Pelusi S, Baselli G, Dongio anni P, Zanoni I, e al. Non- in asi e s a ifica ion o hepa ocellula ca cinoma isk in non-alcoholic a y li e using polygenic isk sco es. J Hepa ol 2021;74:775–82. h ps://doi.o g/10.1016/j. jhep.2020.11.024. [15] Fujiwa a N, Kubo a N, C ouche E, Kone u B, Ma quez CA, Jajo iya AK, e al. Molec- ula signa u es o long- e m hepa ocellula ca cinoma isk in nonalcoholic a y li e disease. Sci T ansl Med 2022;14:eabo4406. h ps://doi.o g/10.1126/sci- anslmed.abo4474. Ma iana Ve delho Machado* Se i¸co de Gas en e ologia, Hospi al de Vila F anca de Xi a, Lisboa, Po ugal Faculdade de Medicina, Uni e sidade de Lisboa, Lisboa, Po ugal *Co esponding au ho . E-mail add ess: [email p o ec ed] M.V. Machado Annals o Hepa ology 28 (2023) 101101 2