HBsAg subtyping was performed in 5,337 sera from chronic carriers who originated from 54 different countries of the 5 continents. Nine subtypes were defined: ayw1, ayw2, ayw3, ayw4, ayr, adw2, adw4, adr q- and adr q+. The repartition of these subtypes, according to the country of origin of the carriers, enhances the previous results and supplies new data: ayw1 is prevalent only in Vietnam (51%); ayw2 in Mediterranean countries (73%); ayw3 in Greece and Yugoslavia (54%) along with ayw2 (41%); ayw4 in West Africa (82%) and Central Africa (42%) along with ayw2; ayr was only found in Vietnam (3.4%); adw2 is prevalent in North and Central Europe (70%), East and South Africa (95%), India (55%), along with ayw3 (35%) in northern South American (74%), and in the Antilles (82%); adw4 is widespread in French Polynesia (45%)--with a 100% frequency in the Marquesas archipelago--as well as in Argentina (42%); adr q--was found only in Ocean: French Polynesia (34%) with a 69% frequency in the Australes, New Caledonia (3 out of 3 HBsAg carriers); adr q+ is the prevalent subtype in South-East Asia if we exclude Vietnam (61%). These results show that a precise geographical distribution of HBsAg subtypes needs more than 'four main subtypes' generally used. Enlarging from 4 subtypes to 9 is a requirement for valuable epidemiologic studies, as well as for the specification of anti-HBs antibodies produced by hybrids or induced by synthetic peptide. The geographical distribution of these 9 HBsAg subtypes and the serological relationship between some determinants suggest a genetic recombination of viral DNA.

In addition to the 10 HBs Ag subtypes already described, 2 new subtypes were defined by using the q determinant. Exceptions to the rule generally accepted were found in that the q determinant was only lacking in HBs Ag/adw4. These exceptions occurred in adw and adr categories. These 2 new subtypes are adw q positive and adr q negative. Out of 98 HBs Ag/adw4 from silent carries and patients from different parts of the world, mainly from France (79), 6 were found q positive. 3 out of these 6 cases came from Montpellier (South of France) and another 3 from Germany. The 92 other cases were found q negative. Further studies will be necessary to better know the location of this new subtype adw4 q positive, but it seems to be present only in certain parts of Europe. Out of 86 HBs Ag/adr from silent carriers from Asia (58), Oceania (17), France (10, most of them contaminated in Asia) and Réunion (1), 10 were found q negative. All these 10 cases were detected in Oceania, 2 out of 2 in carriers from New Caledonia and 8 out of 13 from French Polynesia. The new subtype adr q negative seems localized in Melanesia and Polynesia and absent from Asia. These 2 new markers of hepatitis B virus will allow better epidemiological and geographical studies.

605 HB antigens from silent carriers of different geographic areas were subtyped. Guinea pigs and goats were immunized with selected HB antigens. Subtyping was performed by CIEP using monospecific or duospecific antisera, and by ID with unabsorbed antisera if the antigen was strong enough. Eight categories of HB antigens previously described were confirmed: a1y(w), a2ly(w), a23y(w), a3y(w), ayr, a21d(w2) a3d(w) and adr. The relations between the four categories of a subdeterminants and ‘w’ are analysed. All the antigens with ‘w’ determinant are either a1 or a2l or a23 or a3. The ‘w' determinant can be subdivided into the system of a subdeterminants which are mutually exclusive among themselves, r appears to be allelic with the group of the four a subdeterminants. It cannot yet be stated if r may be considered as a fifth subdeterminant of a or if it is independant of a subdeterminant. For the present time we use the following terminology a1(w), a21l(w2), a23(w3), a3(w4) and r. By combining a(w) subdeterminants, r, d and y determinants, it is possible to greatly increase our knowledge about the geographical distribution of HB antigen: ay subtype is predominant in Africa and in Mediterranean countries but three areas can be distinguished by means of a subdeterminants : a3(w4)y being mainly found in West and Central Africa (87%) and a31(w2)y in Mediterranean countries (86%) except for Greece where a same ratio of a21(w2)y and a23(w3)y is encountered. This a23y subtype is the most frequent of ay subtypes in France and in Hungary. In these two countries, the subtypes ad(w) are predominant: 67% of a21(w2)d and 1.3% of a3(w4)d in France, adr is widespread in Asia and especially in the Thai area (Thailand and Laos), but we have found a1(w1)y in Vietnam. Many examples are given which show that the subtype of HB antigen is related to the country of origin and does not seem to depend of the host. These findings illustrate the importance of HB antigen subtyping for epidemiology and for study of populations.

Abstract. 261 HB antigens were subtyped for d and y. In the healthy carriers, 63 were found to be ad and 75 were ay. In the patients with acute hepatitis, 24 were ad and 33 ay (6 were ad and 2 ay in chronic hepatitis). Among haemodialyzed patients, 50 were ay and only 6 were ad. In 2 of these renal patients, we found a true association of ad and ay (ady). The w and r determinants were detected in 63 of these 261 HB antigens. 61 were w (33 ayw, 26 adw and 2 adyw), only 2 were w negative (I ayr and 1 adr). In the course of this study, several abnormalities: reactions of some ay with anti‐d sera, peculiar behaviour of an antiserum from Dakar which was responsible for the formation of spurs between some ay (and three ad) over all other HB antigens could not be explained by the w and r determinants. This and other observations (especially selected absorptions) led us to subdivide the ay into three categories (1, 2 and 3) and the ad into two categories (4 and 5). A sixth category consisted of the genuine association of ad and ay (ady). The observation that ay of category 3 and ad of category 5 behaved similarly in absorption studies and with the antiserum ‘Dakar’, led us to conclude that the common a determinant was heterogeneous and included 3 subdeterminants: al, a2 and a3, of decreasing antigenic strength. The five categories observed were interpreted as follows: (1) aly, healthy carriers from Vietnam; (2) a2y, healthy carriers, niainly Caucasians, and all the ay patients; (3) a3y, healthy carriers, mainly Africans; (4) a2d the great majority of ad carriers and patients, and (5) a3d, only 3 cases (see text). These new subdeterminants are discussed and compared with the determinants already described by other investigators.

Screening for e antigen and anti-e was performed in 517 HBs antigen chronic carriers, 329 blood donors and 188 hemodialyzed patients, e antigen was detected in 35 blood donors (10.6%): 88% had a high titre of HBs Ag, and 80% had a disturbed liver function. No difference was noted regarding sex and age of the carrier. Anti-e was detected in 33.7% of the blood donors. A significant difference (p<0.01) was noted between males (29%) and females (46.6%). Anti-e was found more prevalent in young people between 21 and 34 than in older people (p<0.05). Anti-e was rarely found when the titre of HBs Ag was low or high. 88% of the anti-e carriers had an intermediate titre from 1/1 to 1/32 by CEP. Both e Ag and anti-e were more prevalent in ad subtype than in ay subtype (p<0.02 for both). Inside the ad subtype, anti-e is less frequent in adr (6%) than in adw2 (42%) and in adw(4) (57%); pcO.Ol. The difference between the prevalence of e Ag in adr (31%) and in adw(2) (17%) is not significant but the healthy carriers with e antigen are more numerous in adr than in adw(2) subtype. e antigen was detected in 53.7% of HBs Ag-hemodialyzed patients and anti-e in 3.7% of these patients. These results are neither correlated with the liver function nor with the state of chronic carrier (27% of e Ag in transient HBs Ag carrier).