Typhoid With Two Drugs Left: Extensively Drug-Resistant Salmonella Typhi From Pakistan, 2016–2018

The typhoid strain behind an outbreak in Pakistan, with cases from November 2016, resists chloramphenicol, ampicillin, trimethoprim-sulfamethoxazole, fluoroquinolones and third-generation cephalosporins — every antibiotic long used as a first choice against the disease. Only azithromycin and carbapenems still work. It is called extensively drug-resistant (XDR) Salmonella Typhi, and by 2018 it had reached U.S. patients.

How typhoid ran out of first-line drugs

Typhoid fever is a systemic febrile illness that needs prompt antibiotic treatment. It causes about 12–27 million cases a year worldwide; the United States reports about 350 culture-confirmed cases to CDC each year, most in people who traveled abroad in the previous 30 days.

The outbreak in Pakistan

It was reported in February 2018. From November 2016 to September 2017, 339 XDR cases were reported, mostly in Karachi and Hyderabad, plus one travel-associated case in the United Kingdom. As surveillance was strengthened, the count rose to 5,372 cases during 2016–2018.

What U.S. surveillance found

In March 2018, CDC asked state and local health departments to ask every typhoid patient about travel to or from Pakistan and to rush their isolates to CDC's National Antimicrobial Resistance Monitoring System (NARMS), which tests each against 14 antimicrobials.

Before the outbreak, 2006–2015, 3,538 culture-confirmed typhoid patients were reported (a median of 338 a year), 244 (7%) of whom had traveled only to Pakistan. Of 3,598 isolates tested, 65% were fluoroquinolone nonsusceptible, 12% were MDR and none resisted ceftriaxone. Fluoroquinolone nonsusceptibility rose from 55% in 2006 to 66% in 2015. Among isolates from travelers to Pakistan, the rates were higher: 79% fluoroquinolone nonsusceptible and 50% MDR.

During 2016–2018, 29 U.S. patients had traveled to or from Pakistan (counted through October 12, 2018):

Why it matters beyond Pakistan

About 250,000 trips were made from the United States to Pakistan in 2017. The strain could spread to neighboring countries such as India, to which Americans made about 2.4 million trips that year — and returning travelers from India typically account for 57%–69% of the typhoid cases reported to CDC.

For clinicians

• Test properly: take a travel history, blood and stool cultures and susceptibility testing. Serologic tests should not be used to diagnose typhoid — they yield no bacterial isolate to test.
• Don't start with a fluoroquinolone, especially for travelers back from South Asia: most U.S. infections are nonsusceptible, and that has been linked to treatment failure or slow response.
• After travel to or from Pakistan: treat uncomplicated typhoid with azithromycin, at a higher dose than for routine uses; severe or complicated typhoid — encephalopathy, intestinal perforation, peritonitis, intestinal hemorrhage, or bacteremia with sepsis or shock — may need a carbapenem. Adjust once culture results arrive.
• Expect relapses: a similar, often milder illness can return 1–3 weeks after improvement, even with appropriate treatment.
• Report confirmed cases to the local health department, and send every isolate to NARMS.

For travelers

Vaccination plus safe food and water — drink only disinfected or bottled water, and wash hands before eating — gives the best protection. Two vaccines are available in the United States, each protecting 50%–80% of recipients:

Anyone who falls ill while abroad or after returning should seek medical care.

Limits. Surveillance sees only culture-confirmed infections, a fraction of the total, and travel and resistance data were missing for some patients.

Sources

Based on Chatham-Stephens K, Medalla F, Hughes M, et al., "Emergence of Extensively Drug-Resistant Salmonella Typhi Infections Among Travelers to or from Pakistan — United States, 2016–2018," MMWR Morbidity and Mortality Weekly Report volume 68, number 1, Centers for Disease Control and Prevention; a work of the United States government in the public domain. The imported copy was cut short; the missing text is from the report's full text in PubMed Central (PMC6342547).