Antituberculous drug resistance in western Canada ( 1993 to 1994 )

RICHARD LONG BSc MD FRCPC FCCP, ANNE FANNING MD FRCPC, ROBERT COWIE MD, VERNON HOEPPNER MD FRCPC, MARK FITZGERALD MB FRCP(I) FRCPC, AND THE WESTERN CANADA TUBERCULOSIS GROUP* University of Manitoba, Winnipeg, Manitoba; University of Alberta, Edmonton, Alberta; University of Calgary, Calgary, Alberta; University of Saskatchewan, Saskatoon, Saskatchewan; and University of British Columbia, Vancouver, British Columbia ORIGINAL ARTICLE


PRINCIPALES MESURES DES RÉSULTATS: Les tests de susceptibilité aux médicaments ont été pratiqués dans tous les cas voir page suivante
Can Respir J Vol 4 No 2 March/April 1997 I n a recent (1980 to 1989) retrospective survey from the province of Manitoba, 7.1% of patients with culture-positive tuberculosis (TB) had disease caused by organisms that were resistant to antituberculous drugs.Risk of drug resistance was found to be significantly higher among foreignborn patients, especially if they developed TB within the first five years of arrival in Canada (1).Although aboriginal persons (Canadian native Indian and Inuit) were not at increased risk of drug resistance, they were over-represented in a poorly compliant subgroup that became increasingly drugresistant over time.In the past decade the proportion of TB cases in foreign-born patients has increased in Canada, while the proportion in the aboriginal population has remained constant and the proportion in the Canadian-born nonaboriginal population has decreased (2).As well, regional differences in the proportion of TB cases in these three ethnic groups have developed; for example, in western Canada the majority of TB cases in British Columbia are found in the foreign born, while the majority of cases in Saskatchewan are found in aboriginal persons (2).In the present prospective survey we sought to determine the antituberculous drug resistance rate in western Canada, to confirm that the risk of drug resistance was higher in the foreign born compared with other ethnic groups and to ascertain whether regional differences in the ethnicity of TB patients might result from regional differences in drug resistance.
In the United States the strongest risk factor for human immunodeficiency virus (HIV)-associated multidrug-resistant TB (MDRTB) has been found to be place of residence, with most cases coming from New York City (3).If there was a problem with HIV-associated MDRTB in western Canada, one might expect to find it in Vancouver, British Columbia, where AIDS-related TB has been reported (4).In the present survey TB patients at risk for HIV were serotested.

METHODS
Demographic, clinical and mycobacteriological data were gathered on all cases of active TB diagnosed in residents of British Columbia, Alberta, Saskatchewan and Manitoba.Data were collected between February 1, 1993 and January 31, 1994 and were recorded prospectively by an on-site study nurse.Drug-resistant cases were identified through the records of the provincial mycobacteriology laboratories.Demographic data: Age when diagnosed with TB, sex and ethnic status -foreign-born, aboriginal (registered [Treaty] Indians, Métis and Inuit) and Canadian-born nonaboriginalwere recorded.Treaty Indians are those registered with the Department of Indian and Northern Affairs according to the Indian Act of Canada.Canadians who were not Aboriginal were defined as 'others'.For foreign-born patients the country of origin and the date of arrival in Canada were noted.Clinical data: For each case the type of disease (respiratory, nonrespiratory, both) and disease status (new active disease, reactivation) were recorded according to the Canadian Tuberculosis Reporting System, Statistics Canada.A new active case is a case not previously reported.A reactivation is a recurrence of active disease in the same patient after a known period of inactivity.Patients from whom drug-resistant organisms were isolated were classified as having 'initial' drug resistance (no history of antituberculous drug use) or 'acquired' drug resistance (received antituberculous drugs in the past including prophylaxis for three months or more).Resistance was classified as 'unknown' in patients whose drug use history was unknown.
Resistance in patients with no history of prior antituberculous drug use was reported as 'initial' rather than 'primary' because it was anticipated that most of the drug-resistant patients would be foreign-born, precluding corroboration of their drug use history through records (5).'Initial' resistance in Canadian-born Aboriginals and non-Aboriginals may be considered 'primary' (ie, resistance in a patient with no prior drug use with the information on drug use being collected in a standardized fashion and corroborated) because their drug use histories, in addition to being taken prospectively by a study nurse, were further corroborated through the records of the provincial tuberculosis registries.Laboratory data: Isolates were processed with the radiometric system (BACTEC 460, Becton-Dickinson Diagnostic Instrument Systems, Maryland).Details of this method have been described previously (6).In each provincial mycobacterialogy laboratory the BACTEC system had been validated against the proportion method used by the Laboratory Centre for Disease Control, Ottawa, Ontario.Susceptibilities to four (isoniazid [INH], rifampin [RIF], ethambutol [EMB], and streptomycin [SM]) and, in some cases, five (pyrazinamide [PZA]) drugs were performed in duplicate with the radiometric method.Drug concentrations used were as follows: INH 0.1 µg/mL, RIF 2.0 µg/mL, EMB 7.5 µg/mL, SM 6.0 µg/mL and PZA 100 µg/mL.
Patients identified at risk for HIV infection (when this Odds ratios (ORs) and 95% confidence intervals (CIs) were estimated (8).

RESULTS
Over the study period 534 cases of culture-positive tuberculosis were identified in residents of western Canada.Drug resistance occurred in isolates from 37 patients (6.9%; 29 foreign-born, six aboriginal and two others).
Table 1 shows the results of culture and drug resistance by age, sex, ethnic status, years since immigration, disease status and type of disease.It also shows ORs and 95% CIs to indicate the likelihood that a culture-positive case of tuberculosis would be resistant to one or more antituberculous drugs.
ORs indicated that drug resistance was not influenced by age, time since arrival in Canada or type of disease.On the other hand, in a univariate analysis the risk of antituberculous drug resistance was significantly higher among foreign-born patients than Aboriginals or others (OR 3.2, 95% CI 1.3 to 7.8).A borderline elevated risk of antituberculous drug resistance was seen between females versus males (OR 2.0, 95% CI 1.0 to 4.0) and among those with reactivation versus new disease status (OR 2.4, 95% CI 1.0 to 5.5).When a multiple logistic regression was performed on the same data, foreign-born ethnicity and reactivation disease status were confirmed to be independent risk factors for drug resistance.
The proportion of cases that were drug-resistant in each province was 1.1% in Manitoba, 8.1% in Saskatchewan, 8.3% in Alberta and 8.1% in British Columbia; the proportion in Manitoba was significantly less than in the other three provinces (P<0.05)(Table 2).The most western provinces (Alberta and British Columbia) had the largest proportion of foreign-born TB patients (62%), and the most eastern provinces (Manitoba and Saskatchewan) had the largest proportion of TB patients who were aboriginal (64%).Drug resistance tended to be higher in British Columbia and Alberta (31 of 380, 8.2%) than in Saskatchewan and Manitoba (six of 154, 3.9%).
Of the foreign-born patients, 68% (179 of 263) emigrated from five countries: China, India, Vietnam, Philippines and Hong Kong (Table 3).Most foreign-born patients who were drug-resistant (26 of 29, 90%) were from these countries.Rates of initial resistance among foreign-born TB patients  More patients with drug resistance were classified as having initial resistance (76%) than acquired resistance (24%) (Table 4).There was no significant difference among the three ethnic groups as to the type of resistance.Patients who had received antituberculous drugs in the past were significantly more likely to be drug-resistant than those who had never received them (15% versus 6%, P<0.05).
Table 5 shows the distribution of resistance by individual drugs and combinations of drugs.Most common was resistance to INH (25 [67.6%] of the patients with drug-resistant disease) and SM (18 [48.6%]).In the majority of the patients with drug-resistant organisms (25 [67.6%]), the isolates were resistant to only one drug.In four of the 12 cases (25%) in which the isolates were resistant to two or more drugs, the isolates were resistant to both INH and RIF.Only two patients, both foreign-born, were shedding organisms resistant to three or more drugs.Both denied having received antituberculous drugs in the past.Of the 465 patients with culturepositive new active tuberculosis, initial resistance was found to the following drugs: INH 19 (4.1%), SM 13 (2.8%),EMB four (0.9%) and RIF one (0.2%).Drug susceptibility to PZA was performed on only 210 isolates, all from Manitoba and Alberta.One patient of the 210 tested was found to be resistant (0.5%).Of patients who were serotested for HIV, 14 were found to be positive, 11 from British Columbia and three from the Prairie provinces.Only one, a foreign-born patient from Manitoba, was drug-resistant.

DISCUSSION
Over the study period we found that 6.9% of the patients in western Canada with culture-positive TB had disease that was resistant to antituberculous drugs.This proportion is very similar to the rate of resistance last reported in Manitoba (7.1% in 1993) (1) and in the Canadian population overall (6.3% reported in 1975) (9).Relative risk of drug resistance was 2.4 times greater among those with a reactivation than among those with new disease and 3.2 times greater among foreign-born patients than among Canadian-born.Previous surveys of antituberculous drug resistance in Canada suggested that the earlier TB developed in the foreign-born after their arrival in Canada the more likely the disease was to be drug-resistant (1,9).In the present study there was a tendency for the risk of drug resistance to be higher among foreignborn patients who developed TB within their first two years in Canada than in those in whom it developed more than two years after immigration.However, the difference was not statistically significant, probably because of the smaller numbers in this survey.
Among foreign-born TB patients, 68% were from five   Asian countries (China, India, Vietnam, Philippines and Hong Kong), reflecting the recent pattern of immigration to Canada, with increasing proportions of immigrants coming from Asian countries where TB morbidity rates are high (10).The majority of drug-resistant foreign-born patients (90%) was from these five countries.Together the rate of initial resistance among them was 14%, up from initial resistance rates of 7.6% (10) and 6.8% (11) reported in Asian immigrants to British Columbia in the early 1970s and 1980s, respectively.Because 87% of the patients who developed drug-resistant TB after immigration to Canada denied being treated (initial resistance), we must assume that they have been infected with a drug-resistant strain in their country of origin or that their history was unreliable.That a substantial number of the foreign-born patients with disease resistant to the first-line antituberculous drugs denied previous treatment must be taken into account in the planning of antituberculous regimens for this group.It is recommended that foreign-born patients, particularly recent arrivals, be treated with four drugs until susceptibility testing allows a more individualized regimen.In most, but not all, this will ensure that at least two of the drugs will be effective when the bacterial population is large.Even this regimen would have been inadequate for two foreign-born patients who denied previous antituberculous drug use and yet were resistant to three or more first-line drugs.These cases illustrate the need for drug susceptibility testing of all isolates, particularly those from the foreign born.
On the other hand, according to the most recent American Thoracic Society/Centers for Disease Control and Prevention guidelines (12), rates of resistance in Canadian-born, aboriginal (3.8%) and nonaboriginal (2.0%) patients were sufficiently low (less than 4%) to warrant the use of only three drugs in the initial treatment regimen, INH, RIF and PZA.
British Columbia and Alberta had a greater proportion of foreign-born TB cases (62%), whereas Saskatchewan and Manitoba had a greater proportion of aboriginal TB cases (64%).This regional difference in the ethnicity of TB patients may have contributed to the higher, though not statistically significant, rate of resistance in British Columbia and Alberta (8.2%) compared with Saskatchewan and Manitoba (3.9%).A more likely explanation for this difference in resistance rates between regions was the unusually low rate of resistance in Manitoba in the study year (1.1%); a rate significantly less than that reported on average in the decade 1980 to 1989 (7.1%), but nevertheless consistent with the variability in the rate of resistance seen over that period (4% to 9%, unpublished data).
Twelve (32%) of the 37 patients with drug-resistant disease had isolates resistant to two or more first-line drugs.This high proportion of multiple-drug resistance supports a policy of treating all TB patients who are from a country with a high prevalence of drug resistance, have had previous treatment with antituberculous medications or have had known exposure to a drug-resistant case with four antituberculous drugs until the drug susceptibility test results are known [12].Four (11%) of the 37 drug-resistant patients were resistant to both INH and RIF.Outbreaks of disease resistant to these two drugs have been reported in association with HIV infection in the United States (3).These cases may be particularly hard to treat (14,15).No MDRTB was found in association with HIV infection in our survey; only one HIV seropositive TB patient was drug-resistant and he was foreign-born, consistent with an earlier report of TB in the HIV infected in British Columbia (4), in which almost all were were Canadian-born nonaboriginal patients and none were drug-resistant.

TABLE 1 Culture results and resistance to antituberculous drugs among patients in western Canada with tuberculosis (TB) diagnosed from February 1, 1993 to January 31, 1994 Variable Total patients
*First group for each variable is the reference category for calculating odds ratios.† Canadian-born nonaboriginal patients

TABLE 2 Resistance to antituberculous drugs among tuberculosis patients from western Canada according to province and ethnic status
*Refers to Canadian-born nonaboriginal patientsAntituberculous drug resistance in western CanadaCan Respir J Vol 4 No 2 March/April 1997

TABLE 3 Resistance to antituberculous drugs among foreign-born tuberculosis patients from western Canada according to country of origin
*Figures in brackets refer to those not treated in the past