pattern occurred in Connecticut and Georgia. In Connecticut, the relative and absolute increases were greatest among white heterosexual males. In Georgia, increases occurred only among white and black males, and a substantial portion of the increase appeared to be among homosexual/bisexual males.

In the five states reporting decreases, the only exception to the overall pattern of decrease occurred among white females. The number of reported cases increased by 51% (20 cases) in this group.

The pattern of increase differed among reporting areas. In some areas, such as Philadelphia and Los Angeles, the increase appears to have plateaued in the middle of 1987. However, in other areas, such as NYC, Florida, and Oregon, the increase continued to climb. In still others, such as Pennsylvania (excluding Philadelphia), the increase began during this period.

Reported by: RG Sharrar, MD, M Goldberg, Philadelphia Dept of Public Health. Participating City and State Health Depts and STD Control Programs. Div of Sexually Transmitted Diseases, Center for Prevention Svcs, CDC.

Editorial note: These increases in infectious syphilis not only reverse the downward trend of the past 4 years, they also suggest an important shift in the epidemiology of the disease in the United States. As infectious syphilis has decreased among homosexual and bisexual males, largely because of changes in sexual behavior due to AIDS, a sizeable increase has occurred among heterosexuals. A similar shift was documented earlier in two small outbreaks (2,3).

While the cause of this increase is unknown, several hypotheses have been proposed. First, anecdotal reports from persons interviewing syphilis patients and their sexual partners indicate that prostitution in which nonintravenous drugs (especially "crack" cocaine) are exchanged for sex may be partially responsible for outbreaks of syphilis as well as other sexually transmitted diseases. A review of TABLE 1. Reporting areas with the largest absolute increases in infectious syphilis United States, weeks 1-46, 1987

records of interviews in Philadelphia showed that the proportion of patients associated with both prostitution and drug use increased significantly between 1985 and 1987 (4).

Second, some investigators have suggested that routine use of spectinomycin (which does not appear to cure incubating syphilis [5,6]) in areas where a sizeable proportion of gonorrhea infections are caused by ß-lactamase-producing organisms may explain the increase in infectious syphilis.* Events in NYC, Florida, and Los Angeles are compatible with this theory; however, for several other areas with sizeable increases in reported syphilis, spectinomycin was not in common use before the increases began. While this mechanism may play a role in some areas, it alone cannot account for the nationwide increase.

Third, a decrease in the resources available for syphilis control programs has been suggested as a contributing factor. Twenty reporting areas provided data on the number of staff available for syphilis control during 1985 and 1986. Ten of these areas

*Parenteral penicillin regimens used to treat gonorrhea have been shown to cure incubating syphilis acquired at the same time as gonorrhea infection (7).

*Arizona, Baltimore, Connecticut, North Carolina, Oregon, and Philadelphia.

TABLE 2. Cases of infectious syphilis from 14 reporting areas,* by race, sex, and sexual preference - United States, January-August, 1987

*Arizona, California, Connecticut, Florida, Georgia, Maryland, Massachusetts, Mississippi, North Carolina, Oregon, Pennsylvania, South Carolina, Tennessee, and New York City. Data for California (other than Los Angeles and San Francisco) are for the first 6 months only. *Males naming at least one male sexual partner were classified as "homosexual/bisexual"; those not naming any were classified as "heterosexual." Overall, 87% of males were interviewed in 1986 and 85%, in 1987. Over 80% of males were interviewed in all reporting areas except New York City, where 55% were interviewed in 1986 and 45%, in 1987.

reported increases in the number of persons interviewing patients with early syphilis between 1985 and 1986; four reported no change; and six reported decreases. Areas reporting increases in total syphilis morbidity were somewhat more likely to report a decrease in the number of interviewers; however, the association was not statistically significant.

The increases in infectious syphilis among females and heterosexuals are disturbing for three reasons. First, an increase in the number of females with syphilis will likely be followed by increased morbidity and mortality from congenital syphilis. Second, the marked increase among inner-city, heterosexual minority groups suggests that high-risk sexual activity is increasing in these groups despite the risk of HIV infection, which is already elevated because of the high prevalence of intravenous drug abuse. Third, studies in Africa and in the United States suggest that genital ulcer diseases such as primary syphilis increase the risk of HIV transmission (8,9). References

1. Landrum S, Beck-Sague C, Kraus S. Racial trends in syphilis among men with same-sex partners in Atlanta, Georgia. Am J Public Health 1988;78:66-7.

2. Centers for Disease Control. Early syphilis - Broward County, Florida. MMWR 1987;36:221-3. 3. Lee CB, Brunham RC, Sherman E, Harding GKM. Epidemiology of an outbreak of infectious syphilis in Manitoba. Am J Epidemiol 1987;125:277-83.

4. Rolfs RT, Goldberg M, Sharrar RG. Outbreak of early syphilis in Philadelphia. Presented at the 115th annual meeting of the American Public Health Association and related organizations, New Orleans, Louisiana, October 18-22, 1987.

5. Petzoldt D. Effect of spectinomycin on T. pallidum in incubating experimental syphilis. Br J Vener Dis 1975;51:305-6.

6. Rein MF. Biopharmacology of syphilotherapy. J Am Vener Dis Assoc 1976;3:109-27.

7. Schroeter AL, Turner RH, Lucas JB, Brown WJ. Therapy for incubating syphilis: effectiveness of gonorrhea treatment. JAMA 1971;218:711-3.

8. Quinn TC, Glasser D, Matuszak DL, et al. Screening for human immunodeficiency virus (HIV) infection in patients attending sexually transmitted diseases clinics: risk factors and correlates of infection. Presented at the International Society for STD Research, Atlanta, Georgia, August 2-5, 1987.

9. Cameron DW, Plummer FA, Simonsen JN, et al. Female to male heterosexual transmission of HIV infection in Nairobi. Presented at the International Society for STD Research, Atlanta, Georgia, August 2-5, 1987.

Current Trends

Antigenic Variation of Recent Influenza A(H3N2) Viruses

Analysis of recent influenza A(H3N2) viruses indicates antigenic drift from the previously prevalent strains A/Mississippi/1/85 and A/Leningrad/360/86. One reference variant virus strain, A/Victoria/7/87, was first isolated in Australia in April of this year. A second reference variant, A/Sichuan/2/87, was first isolated in China, also in April. In hemagglutination inhibition tests with antiserum from infected ferrets, antibody to A/Victoria/7/87 reacts poorly with other strains, even though the virus itself is inhibited well by antiserum to A/Mississippi/1/85 (Table 1). Antiserum to A/Sichuan/2/87 reacts at lower titers with viruses such as A/Mississippi/1/85 and A/Leningrad/360/86, which circulated earlier, than it does with A/Sichuan/2/87 anti

gen. Also, A/Sichuan/2/87 is inhibited poorly by antisera to all of the viruses that circulated earlier. Analysis of about 50 recently isolated A(H3N2) viruses from Asia, Oceania, and the United States indicates a spectrum of antigenic specificity, with many isolates having reaction patterns intermediate between A/Leningrad/360/86 and A/Sichuan/2/87.

The antibody response induced by the current type A(H3N2) vaccine component is greater toward the homologous A/Leningrad/360/86 virus than toward the reference variants A/Victoria/7/87 and A/Sichuan/2/87. This response confirms the existence of antigenic variation in recent virus isolates. Vaccinees in all age groups developed titers of 40 or more to A/Leningrad/360/86 with greater frequency than they did to the new antigenic variants (Table 2). In addition, the geometric mean titers were higher to the homologous A/Leningrad/360/86 antigen than to the antigenic variants A/Sichuan/2/87 or A/Victoria/7/87.

TABLE 1. Hemagglutination-inhibition reactions* of influenza type A(H3N2) viruses

*Titers are the reciprocal of antiserum dilutions; homologous titers appear in bold type. When comparing reactions of sera with different antigens, fourfold or greater differences are considered significant.

TABLE 2. Hemagglutination-inhibition serum antibody response to influenza vaccine in immunized* children and adults

United States, fall 1988

*Volunteers received trivalent influenza vaccine containing 15 μg each of hemagglutinin of A/Leningrad/360/86(H3N2), A/Taiwan/1/86(H1N1), and B/Ann Arbor/1/86 viruses. *Geometric mean titer.

Reported by: P Graves, G Meiklejohn, MD, School of Medicine, Univ of Colorado Health Sciences Center, Denver, Colorado. F Ruben, MD, Univ of Pittsburg, Pittsburg, Pennsylvania. P Palmer, K Edwards, MD, Vanderbilt Univ, Nashville, Tennessee. Influenza Research Center, Baylor College of Medicine, Houston, Texas. Participating State and Territorial Epidemiologists and State Laboratory Directors. Sentinel Physicians of the American Academy of Family Physicians. WHO Collaborating Laboratories. WHO Collaborating Center for Influenza, Influenza Br, Div of Viral Diseases, Center for Infectious Diseases, CDC.

Editorial Note: In 1987, the World Health Organization Collaborating Centers for Influenza (Atlanta and London), in conjunction with National Influenza Centers in several countries in Asia and Oceania, detected antigenic variants of influenza A(H3N2). Evidence is accumulating that these viruses are infecting persons of all age groups, including high-risk elderly persons (1). These variants are associated with the reappearance of influenza A(H3N2) viruses after a period of quiescence during the winter of 1986/87.

Antigenic variation has always complicated influenza vaccine formulation. The occurrence of viruses that exhibit antigenic drift from the vaccine strain has on

TABLE II. Notifiable diseases of low frequency, United States

2

11

21

33

31

17

74

59

25

3

72

43

19

14

2

16

7

55

55

*One of the 54 reported cases for this week was imported from a foreign country or can be directly traceable to a known internationally imported case within two generations.