Mycobacterium avium complex lung disease: more patients, multiple bacteria

Eric Houpt MDProfessor and Chief

Division of Infectious Diseases and International HealthUniversity of Virginia

Outline• Epidemiology• Clinical presentation• Diagnosis and treatment• UVA research findings

Epidemiology• Nontuberculous lung disease• Medicare: 8% increase in prevalence per year• 20K-80K cases annually in the US (9K TB)• 1.4x Female predominance• Higher in certain geographies, such as SE US

l \ SKIN-TEST RESULTS FOR STATES 103 102 ATLAS OF SENSITIVITY TO TUBERCULIN, PPD-B, AND HISTOPLASMIN IN U.S.

} PPD-S ) VIRGINIA VIRGINIA (0.0001mg) }

) STATE INDURATION (IN GROUPS OF TWO MILLIMETERS) PERCENT ECONOMIC TOTAL 22 24 26 28 30+ REACTORS 3,280 NAVY RECRUITS ) AREA TESTED 0-1 2 4 6 8 10 12 14 16 18 20

< PPD-S tO+MM

WHITE MALES 17-21 YEARS STATE TOTAL 3,280 3,013 39 44 28 17 28 33 19 22 16 12 5 4 4.2 LIFETIME ONE- COUNTY RESIDENTS ( AREA l 203 187 l l 2 2 2 2 2 5.4 PERCENT REACTIONS 10+ mm 2 326 307 4 7 2 2 2 1.8

TESTED 1958-1965 Oo.0-1.9 Im! 6.0-7.9 3 248 238 2 3 l 3 l 1.6 i 4 393 359 6 3 l l 4 5 4 3 2 3 5. 9 Q2.o-3.9 Ill 8.0-9.9 ) 5 172 157 l 3 2 l 3 2 2 5.2

[::l 4.0-5. 9 11110.0+ 6 211 195 4 l l 2 2 4 l 4.3 ) 7 305 278 5 5 2 4 4 l 3 3 4.9 8 91 75 2 3 4 l l l l 2 7. 7 ) 9 34 33 l 2. 9

10 96 88 2 4.2

HISTOPL. PPD-B J A 202 183 2 3 2 2 l 3 3 5. 0 l B 213 200 3 2 2 3 2 l 1.4 (H42 1:100) (0.0001mg) C 363 325 5 8 7 l 10 2 l 2 4. 7 D 185 169 2 5 l l 4 l l 4.3 ~, E 72 61 2 2 2 2 l 8.3 l F 166 158 l l l 2 3.6

iD PPD-B 4+MM PERCENT REACTIONS 4+ mm PERCENT REACTIONS 4+mm STATE TOTAL 3,280 1,989 100 291 324 241 178 109 30 9 5 2 2 36.3

D o- 4 ~20-29 D 0-14 30-39 AREA l 203 142 5 13 17 13 4 5 2 27, 6 Q ~30-39 D 15-19 40-49 2 326 222 14 35 25 17 8 3 27 .6 5- 9

3 248 172 9 24 20 15 3 5 27 .o I0-19 11140+ EZi 20-29 l!ll 5o+ 4 393 273 14 30 27 24 15 7 3 27 .o

5 172 99 4 13 24 18 4 8 l 40. l

6 211 104 8 24 24 20 17 7 4 2 46.9 STATE STATE STATE 7 305 153 14 26 33 26 31 13 5 3 1 45.2 STATE STATE

8 91 47 14 13 6 4 4 2 1 48,4 ECONOMIC COUNTY ECONOMIC COUNTY ECONOMIC COUNTX ECONOMIC COUNTY ECONOMIC COUNTY 9 34 18 4 6 4 2 47. l AREA

10 96 43 2 9 10 10 12 9 53, l Buchanan ( 3) Clifton Forge* Amelia 8 Caroline A Roanoke A 202 130 1 19 20 15 7 9 35; l Dickenson Radford* Appomattox Charles City Roanoke* B 213 155 3 14 14 9 10 8 25.8 Lee Covington* Bedford Essex C 363 189 15 25 46 30 32 15 8 2 43.8 Taze1·1el l Buckingham Gloucester B Arlington D 185 94 8 21 21 15 16 6 3 l 44.9 ,Ii se 4 Augusta Cumberland Hanover Fairfax E 72 39 l 8 7 7 6 4 44.4 Norton* Clarke Di nwi ddi e James City Alexandria* F 166 109 2 12 17 12 7 6 33, l Frederick Fluvanna King and Queen Fa 11 s Church*

2 Bl and Page Goochland King George Carroll Racki ngham King William C Chesterfield Grayson Shenandoah Louisa Henri co HISTOPLASMIN 4+MM Russell Warren Nelson Lancaster Richmond* Scott Harrisonburg* Nottoway Mathev1s STATE TOTAL 3,280 2,607 40 23 40 102 153 157 94 46 14 2 2 19.3 Smyth Staunton* Powhatan Middlesex D Norfolk* Washington l·li nchester* Pri nee Edward New Kent Princess Anne AREA 1 203 145 l 3 7 17 12 8 6 3 28.6 ,lythe Waynesboro* Petersburg* Northumberland Norfolk 2 326 220 4 5 2 17 29 22 16 9 1 31.3 Bristol* Colonial Richmond Portsmouth* 3 248 174 4 2 6 10 16 19 9 5 3 28. 2 Gal ax* Albemarle Heights* l•lestmorel and South 4 393 198 3 5 13 37 38 46 36 11 5 48,9 Culpeper vlilliamsburg* Norfolk* 5 172 94 3 2 5 10 20 23 7 7 43.6 Alleghany Fauquier 7 Brunswick Vi rgi ni a Bath Greene Charlotte 9 Accomack Beach* 6 211 191 4 3 3 6 4 7.6 Botetourt Loudoun Franklin Northampton 7 305 295 2 2 2 2 2 2.6 Craig Madison Halifax E York 8 91 84 1 1 1 1 2 6.6 Floyd Orange Henry 10 Greens vi 11 e Hampton* 9 34 30 2 1 11.8 Gil es Lunenburg Isle of ,!right Ne1·1port News* 10 96 92 1 1 3.1 Pri nee Willi am Nansemond Highland Rappahannock Mee kl enburg Pri nee George F Amherst A 202 174 4 4 4 5 7 2 l 11.9 Montgomery Spotsylvania Patrick Southampton Campbell B 213 174 5 l l 6 11 7 4 4 16.0 Pulaski Stafford Pittsylvania Surry Lynchburg* C 363 346 6 4 1 2 3 1 3,0 Rockbridge Cha rl ottesvi 11 e* Danville* Sussex D 185 172 2 3 5 2 1 7.0 Buena Vista* Fredericksburg* Martinsvil 1 e* Hopewell* E 72 70 1 1 2,8 South Boston* Suffo 1 k* F 166 148 3 4 4 4 9.0 *Indicates Independent City

Epidemiology• UVA

2004 2006 2008 2010 2012 2014 2016

Coun

t 2010

“Nontuberculous Mycobacteria” on CT scan reports

2018

0

10

20

30

40

50

60

2001 2002 2003 2004 2005 2006 2007 2008

Num

ber o

f inf

ectio

ns

Abscessus

MAC

Chelonae

Fortuitum

Kansasii

Other

Satyanarayana et al., BMC ID 2011

Epidemiology• UVA

2004 2006 2008 2010 2012 2014 2016

Coun

t 2010

“Nontuberculous Mycobacteria” on CT scan reports

2018

0

10

20

30

40

50

60

2001 2002 2003 2004 2005 2006 2007 2008

Num

ber o

f inf

ectio

ns

Abscessus

MAC

Chelonae

Fortuitum

Kansasii

Other

Satyanarayana et al., BMC ID 2011

2016N=142

Epidemiology – NTM lung disease• Morbidity/Mortality

• 1.4-2x higher mortality than similarly aged patients without NTM (Adjemian, AJRCCM 2012)

• Esp males, elderly

• May relate to comorbidities (die with it > from it)

• Generally a slowly progressive disease, with progressive symptoms and lung function decline

Clinical phenotype 1: male smoker cavitary lung disease• 50 y.o. male former smoker

• Therapy with azithromycin, ethambutol, rifampin (Rif) for over a year, plus amikacin and clofazimine, yet experienced symptomatic, radiographic, and microbiologic failure.• Required lobectomy which demonstrated caseating necrosis

with abundant extracellular MAC.• Died from the infection.

Classic phenotype 2 (more common): thin female with nodular bronchiectasis• 71 y.o. female• Cough for years prior to diagnosis• Consistently worsening dyspnea, severe disease on chest

imaging, and persistently positive sputum cultures.

• Has been on therapy for years, including standard triple therapy (Azi,Rif,Emb) plus adjunctive IV Amik (with hearing loss), followed by inhaled liposomal Amik, followed by Clofazimine, with essentially no improvement.

2013 2014 2015 2016 2017 2018Treatment Clari,Rif,Emb

+IV StreptomycinOff Azi,Rif,Emb

+IV AmikAzi,Rif,Emb

+inhaled liposomal-Amik

Azi,Rif,Emb+Clofaz

Azi500qd,Rif,Emb+Clofaz

MAC culture results + + - - - - - + + + + - - + + + + +Susceptibility testing Clarithro S Clarithro S Clarithro S Clarithro S Clarithro S Clarithro S

71 y.o. female timeline

2013 2014 2015 2016 2017 2018Treatment Clari,Rif,Emb

+IV StreptomycinOff Azi,Rif,Emb

+IV AmikAzi,Rif,Emb

+inhaled liposomal-Amik

Azi,Rif,Emb+Clofaz

Azi500qd,Rif,Emb+Clofaz

MAC culture results + + - - - - - + + + + - - + + + + +Susceptibility testing Clarithro S Clarithro S Clarithro S Clarithro S Clarithro S Clarithro S

• Only Clarithromycin matters• Only drug with in vitro/in vivo

correlation• Generally ignore the others• Clarithro (S) means use standard

Azi/Clari regimens• Clarithro (R) rare

71 y.o. female timeline

Clinical phenotype 2: thin female with nodular bronchiectasis

Reich et al., Chest 1992 …we offer the hypothesis that habitual voluntary suppression of cough may have led to the development of nonspecific inflammatory processes in these poorly draining lung regions (L lingula, RML), upon which MAC engrafted. We offer the term, Lady Windermere's syndrome, to describe this pattern among elderly women and to suggest that their fastidiousness may be its root cause….

Has been described for a long time

Phenotype 2b: thin female with nodular bronchiectasis with minimal or no symptoms

Often CT scan performed for other reasons, often with subtle changes years prior

ATS/IDSA guidelines for NTM• Dx NTM lung disease = • Pulm symptoms (cough, sputum, fatigue, weight loss)

and/or consistent imaging • Exclusion of other diagnoses• Positive sputum cultures x 2 (or 1 bronchoscopic)

• First question with any patient: Dx Y or N

Am J Respir Crit Care Med 2007

ATS/IDSA guidelines for NTM

• If Yes, Question 2: To treat or not to treat?• Minimal symptoms:• May observe.• “…one with minimal symptoms and radiographic findings

such that the treatment seems worse than the disease, follow closely with collecting respiratory specimens for AFB analysis as well as follow-up radiographic studies, usually HRCT scans, over a long period of time as the MAC disease will likely progress at some time and the patient’s symptoms and chest radiographs will likely worsen.”

Am J Respir Crit Care Med 2007

ATS/IDSA guidelines for NTM

• If treatment – at least 12 months (post-culture conversion)• Azithromycin (or clarithromycin) – if ”Clari S”• Plus rifampin (or rifabutin) – regardless of Rif S/R result• Plus ethambutol – regardless of EMB S/R result• M-W-F if nodular bronchiectasis• Daily, plus IV amikacin if cavitary (if can tolerate)• Long term efficacy about 50% - high relapse/reinfection

rate, drug tolerability an issue

Am J Respir Crit Care Med 2007

Relapse or reinfection?

• Known to occur in 15-50% of patients (especially with bronchiectasis form of disease)

• Molecular analysis (PFGE, etc) of strains traditionally has shown more reinfection (new strain – 80%) than relapse (same strain – 20%)

Wallace, Zhang, Brown, et al.: MAC Infections in Nodular Bronchiectasis 1239

dominant genotype (i.e., the genotype from each patient iden-tified in the greatest number of cultures) from 15 of 17 pa-tients (88% ).

The findings among patients with cavitary disease were al-most identical. Of the 11 genotypes identified among patientswith cavitary disease, 8 (73% ) were M. intracellulare, 2 (18% )were M. avium , and 1 (9% ) was MAC-X. Mycobacterium in-tracellulare was the predominant genotype in 8 of 9 (89% ) ofthe patients.

Complete agreement was evident among the DNA probes,the multiplex PCR, PRA for all isolates identified as M. aviumor M. intracellulare. Agreement was also present with the DNAprobes and multiplex PCR for the 4 genotypes of MAC-X. Norecognized PRA pattern has yet been determined for thisgroup with PRA.

SerotypingOf the 49 genotypes identified among patients with nodularbronchiectasis, 42 were subjected to serotyping (see Table 1).Of the 30 genotypes of M. intracellulare, 16 (53% ) were eithernontypable or rough. There was no predominant serovaramong M. intracellulare that were serotypable, with only sero-vars 7, 19, and 22 appearing more than once. Serovars thatwere encountered were 7, 12, 13, 16, 19, 22, 24, and 42. Of the9 genotypes of M. avium , 8 (89% ) were serotypable with 5 of 8being serovar 8.

Of the patients with cavitary disease, 4 of the 7 genotypesof M. intracellulare and 2 of 2 genotypes of M. avium wererough or nontypable, results that were comparable to patientswith nodular bronchiectasis.

A surprising finding was that four patients had differentgenotypes (by the definition of Tenover and coworkers [14])that were of the same serovar: one patient had two genotypesof serovar 8 (M. avium ) (Patient 6, Table 4); one patient had

three genotypes of M. intracellulare that were rough (Patient7); one patient had two genotypes of M. intracellulare serovar19 (Patient 11); and one patient had two genotypes of M. in-tracellulare serovar 22 (Patient 12). Review of the PFGE andserotyping data demonstrated these findings to be correct. Asummary of the large restriction fragment band differences forthe genotypes of these four patients with DraI and X baI isshown in Table 5. A comparison of the PFGE patterns forthree of the patients is shown in Figure 6. The three roughstrains from Patient 8 shared few common bands and ap-peared unrelated (different). The other three pairs of isolatesdiffered by 9 to 16 bands with either DraI or X baI but shared11 to 20 bands in common, suggesting they may be related.The most unusual were the two isolates of serovar 22 from Pa-tient 12 which were identical except for the presence of 11 ex-tra bands in one of the strains, suggesting a large deletion fromone of the strains. Thus identical serovars with different geno-types (by current definitions) (14) was seen with 4 of 16 (25% )of patients in both disease groups with multiple genotypes thathad been subjected to serotyping.

DISCUSSIONPrevious studies of disseminated M. avium in HIV-seroposi-tive patients have demonstrated that polymicrobial infectionoccurs in this setting. In 1993 Arbeit and coworkers (18) re-ported PFGE results on 28 cultures from blood, stool, and

Figure 1. Example of multiple PFGE types (genotypes) in the samesputum specimen. DraI PFGE patterns of six sputum cultures froma patient with nodular bronchiectasis over a period of 3 yr and 10mo. Pattern a is present in lanes 2 and 3, pattern b is present inlane 6, while the cultures in lanes 1, 4, and 5 are a mixture of pat-terns a and b; lane 7, DNA standards.

Figure 2. Example of multiple PFGE patterns (genotypes) in differ-ent sputum specimens from the same patient. DraI (A) and XbaI(B) PFGE patterns of nine genotypes recovered from a single pa-tient with nodular bronchiectasis. Lanes 1 to 9, genotypes recov-ered, received in a time period of 3 yr and 3 mo from a single pa-tient; lane 10, DNA standards.

Wallace, AJRCCM 1998

Clinical microbiology of NTM

DNA probe to idendify”M. Avium complex”

vs. M. Tuberculosis

Storage

MAC = Mycobacterium avium complex

Tortoli Clin. Microbiol. Rev. 2014; doi:10.1128/CMR.00035-14

UVA study: 2010-2017

• 35 patients that met criteria for MAC LD with multiple longitudinal isolates available (n=95)• Mostly nodular bronchiectasis

seen in both patients with cavitary (6/8, 75%) and patients withnodular bronchiectatic (19/27, 70%) lung disease. Strainvariation was not associated with a longer period between isolates(data not shown). Antibiotic susceptibility testing was performedfor clarithromycin, rifampin, ethambutol, amikacin,moxifloxacin, and linezolid by broth microdilution, according toClinical and Laboratory Standards Institute methodology (4),and different strains were more often of a different antibiogramthan identical strains (data not shown). M. avium isolates hadhigher minimum inhibitory concentrations of rifampin andlinezolid than predominant M. intracellulare and M. chimaeraisolates (P, 0.05); otherwise, there was no difference inantibiotic resistance between the species clusters.

DiscussionClinically, we generally assume that longitudinal MAC isolatesfrom patients with nontuberculous mycobacterial lung diseaserepresent the same infection. Using WGS, we found that this isfrequently not the case, as there is great diversity within the MAC.The high rate of MAC diversity within a patient over time wasrobust to multiple bioinformatic approaches, and was supportedby the antibiogram data. Moreover, the phenomena were notconcentrated in patients who had clinical relapse, as wehypothesized, nor was it associated with antibiotic pressure. Wecan imagine several explanations for this diversity. First, somepatients with MAC lung disease may have polymicrobialinfections, and certain subtypes are sampled or preferentiallygrow over time. Such infection with multiple strains has beennoted, particularly with nodular bronchiectatic disease (5).Second, there could be a high force of reinfection from theenvironment. Third, some variation could be spurious andreflect transient colonization as a result of specimencontamination, for example, from drinking water (6), making itdifficult to know which isolates are clinically significant andwhich are not. Further study is needed to assess these clinicalimplications and whether certain species or subtypes carrygreater prognostic significance.

Next, we found that although the M. avium and M. chimaeraisolates were largely pure, at 94–100% of reads at the sequencelevel, M. intracellulare isolates were more heterogeneous,often with genomic content that mapped substantially toM. avium,M. chimaera, or other MAC species (7). This heterogeneitymay suggest, again, that certain M. intracellulare infectionsare polymicrobial, or simply that available M. intracellularereference genomes are poorly representative. Notably, mostof our MAC isolates were predominantly M. intracellulare(43%), followed by M. avium (25%), then M. chimaera (15%), andthen a mixture of others (16%). The high prevalence ofM. intracellulare is similar to the studies from Texas (8), but differsfrom an M. avium predominance seen in other studies (9). Asubstantial prevalence of M. chimaera has been seen in Illinois (9).

Taken together, there appears to be abundant interspeciesand intraspecies MAC diversity in patients over time withMAC lung disease. As a starting point, we believe clinicalmicrobiology laboratories should routinely speciate isolates frompatients with MAC lung disease, for instance, using matrix-assisted laser desorption/ionization–time-of-flight massspectroscopy methods (10), to enable further clarification ofthis phenomenon. n

Author disclosures are available with the text of this letter atwww.atsjournals.org.

Darwin J. Operario, Ph.D., M.P.H.Suporn Pholwat, M.S.Alex F. Koeppel, Ph.D.Alyson Prorock, M.S.Yongde Bao, Ph.D.Katia Sol-Church, Ph.D.Michele Scheurenbrand, B.S.Melinda Poulter, Ph.D.Stephen Turner, Ph.D.Hardik I. Parikh, Ph.D.Amy Mathers, M.D.Eric R. Houpt, M.D.*University of VirginiaCharlottesville, Virginia

*Corresponding author (e-mail: erh6k@hscmail.mcc.virginia.edu).

Table 1. Clinical and Demographic Data

All (N=35)Off Therapy

(n=16)On Therapy

(n= 11)Relapse/Reinfection

(n= 8) P Value

Sex, M 17 (49%) 8 (50%) 4 (36%) 5 (63%) NSMedian age at first culture (range), yr 60 (10–87) 35 (10–80) 66 (34–87) 62 (31–86) ,0.05Median BMI (IQR) 24 (20–27) 22 (21–28) 21 (20–23) 26 (24–27) NSSmoking history ,0.05Current or former 14 (40%) 2 (13%) 7 (64%) 5 (63%)Never 21 (60%) 14 (87%) 4 (36%) 3 (37%)

MAC disease type NSCavity present 8 (23%) 3 (19%) 4 (36%) 1 (13%)Noncavitary-nodular bronchiectasis 27 (77%) 13 (81%) 7 (64%) 7 (87%)

Underlying conditions ,0.05None 11 (31%) 1 (6%) 6 (55%) 4 (50%)CF 12 (34%) 9 (56%) 2 (18%) 1 (13%)COPD 5 (14%) 2 (13%) 3 (27%) 0Other lung disease 6 (17%) 4 (25%) 0 3 (38%)

Definition of abbreviations: BMI = body mass index; CF= cystic fibrosis; COPD=chronic obstructive pulmonary disease; IQR= interquartile range;MAC=Mycobacterium avium complex; NS=not significant.

CORRESPONDENCE

394 American Journal of Respiratory and Critical Care Medicine Volume 200 Number 3 | August 1 2019

Whole genome sequencing

Loman et al., 2015

MAC ~ 5MB

Isolate #Patients #1 through 35

Off Rx On Rx RelapseSpecies compositionCluster

95 iso

lates

Isolate #Patients #1 through 35

Off Rx On Rx RelapseSpecies compositionCluster

M. Avium

M. intracellulare

M. chimaera

Other MAC, mixtures

Isolate #Patients #1 through 35

Off Rx On Rx RelapseSpecies compositionCluster

M. Avium

M. intracellulare

M. chimaera

Other MAC, mixtures

Prior studies of MAC LD:Mostly M. avium in Japan/Korea70% M. intracellulare in Texas25% of MAC LD in Illinois

Isolate #Patients #1 through 35

Species compositionCluster

M. Avium

M. intracellulare

M. chimaera

Other MAC, mixtures

Isolate #Patients #1 through 35

Species compositionCluster

Hypothesis: patients not being treated (minimal dz), early on rx, or with relapse, would have the same strain

M. Avium

M. intracellulare

M. chimaera

Other MAC, mixtures

Isolate #Patients #1 through 35

Species compositionCluster

Those with re-infection would have a distant strain

M. Avium

M. intracellulare

M. chimaera

Other MAC, mixtures

Hypothesis: patients not being treated (minimal dz), early on rx, or with relapse, would have the same strain

Isolate #Patients #1 through 35

Off Rx On Rx RelapseSpecies compositionCluster

Generally false:All groups had substantial MAC isolate diversity over time

Am J Respir Crit Care Med. 2019 Aug 1;200(3):393-396.

M. Avium

M. intracellulare

M. chimaera

Other MAC, mixtures

Hypothesis: patients not being treated (minimal dz), early on rx, or with relapse, would have the same strainThose with re-infection would have a distant strain

Isolate #Patients #1 through 35

Off Rx On Rx RelapseSpecies compositionCluster

Mixtures and changes over time has been noted before:

Am J Respir Crit Care Med. 2019 Aug 1;200(3):393-396.

M. Avium

M. intracellulare

M. chimaera

Other MAC, mixtures

Wallace, Zhang, Brown, et al.: MAC Infections in Nodular Bronchiectasis 1239

dominant genotype (i.e., the genotype from each patient iden-tified in the greatest number of cultures) from 15 of 17 pa-tients (88% ).

The findings among patients with cavitary disease were al-most identical. Of the 11 genotypes identified among patientswith cavitary disease, 8 (73% ) were M. intracellulare, 2 (18% )were M. avium , and 1 (9% ) was MAC-X. Mycobacterium in-tracellulare was the predominant genotype in 8 of 9 (89% ) ofthe patients.

Complete agreement was evident among the DNA probes,the multiplex PCR, PRA for all isolates identified as M. aviumor M. intracellulare. Agreement was also present with the DNAprobes and multiplex PCR for the 4 genotypes of MAC-X. Norecognized PRA pattern has yet been determined for thisgroup with PRA.

SerotypingOf the 49 genotypes identified among patients with nodularbronchiectasis, 42 were subjected to serotyping (see Table 1).Of the 30 genotypes of M. intracellulare, 16 (53% ) were eithernontypable or rough. There was no predominant serovaramong M. intracellulare that were serotypable, with only sero-vars 7, 19, and 22 appearing more than once. Serovars thatwere encountered were 7, 12, 13, 16, 19, 22, 24, and 42. Of the9 genotypes of M. avium , 8 (89% ) were serotypable with 5 of 8being serovar 8.

Of the patients with cavitary disease, 4 of the 7 genotypesof M. intracellulare and 2 of 2 genotypes of M. avium wererough or nontypable, results that were comparable to patientswith nodular bronchiectasis.

A surprising finding was that four patients had differentgenotypes (by the definition of Tenover and coworkers [14])that were of the same serovar: one patient had two genotypesof serovar 8 (M. avium ) (Patient 6, Table 4); one patient had

three genotypes of M. intracellulare that were rough (Patient7); one patient had two genotypes of M. intracellulare serovar19 (Patient 11); and one patient had two genotypes of M. in-tracellulare serovar 22 (Patient 12). Review of the PFGE andserotyping data demonstrated these findings to be correct. Asummary of the large restriction fragment band differences forthe genotypes of these four patients with DraI and X baI isshown in Table 5. A comparison of the PFGE patterns forthree of the patients is shown in Figure 6. The three roughstrains from Patient 8 shared few common bands and ap-peared unrelated (different). The other three pairs of isolatesdiffered by 9 to 16 bands with either DraI or X baI but shared11 to 20 bands in common, suggesting they may be related.The most unusual were the two isolates of serovar 22 from Pa-tient 12 which were identical except for the presence of 11 ex-tra bands in one of the strains, suggesting a large deletion fromone of the strains. Thus identical serovars with different geno-types (by current definitions) (14) was seen with 4 of 16 (25% )of patients in both disease groups with multiple genotypes thathad been subjected to serotyping.

DISCUSSIONPrevious studies of disseminated M. avium in HIV-seroposi-tive patients have demonstrated that polymicrobial infectionoccurs in this setting. In 1993 Arbeit and coworkers (18) re-ported PFGE results on 28 cultures from blood, stool, and

Figure 1. Example of multiple PFGE types (genotypes) in the samesputum specimen. DraI PFGE patterns of six sputum cultures froma patient with nodular bronchiectasis over a period of 3 yr and 10mo. Pattern a is present in lanes 2 and 3, pattern b is present inlane 6, while the cultures in lanes 1, 4, and 5 are a mixture of pat-terns a and b; lane 7, DNA standards.

Figure 2. Example of multiple PFGE patterns (genotypes) in differ-ent sputum specimens from the same patient. DraI (A) and XbaI(B) PFGE patterns of nine genotypes recovered from a single pa-tient with nodular bronchiectasis. Lanes 1 to 9, genotypes recov-ered, received in a time period of 3 yr and 3 mo from a single pa-tient; lane 10, DNA standards.

Wallace, AJRCCM 1998

Some possible explanations

• MAC LD often reflects a polymicrobial infection, with shifts in bacteria or different sampling among sputa over time. “Biofilm”• There is an extremely high reinfection rate from the

environment in these susceptible hosts

MAC in the environment• M. avium and chimaera are found in drinking water, more than M. intracellulare

• However, M. intracellulare (and avium, and others) are found in plumbing biofilms

• Soil:

• Thus, some recommend water treatments to patients:

• Drain and refill the hot water heater every 2 wk.

• Raise hot water heater temperatures (> 130F).

• Remove and clean showerheads (full- strength household bleach for 30 min).

• Replace showerhead with one that produces streams (holes . 1 mm diameter) and not a fine mist.

• Reduce aerosol exposures in bathrooms (fan and window).

• Install shower and tap filters that remove bacteria (> 0.45 μm pore size).

• Replace granular activated carbon filters every 2 wk.

• Get rid of any and all humidifiers.

• Avoid dusts from potting soils (wet potting soil).

Falkinham, AEM, 2001

R01: “Mycobacterial Lung Diseases across Virginia: sequencing and clinical determinants of relapse and outcome”

Aim 1• All ~200 Virginia MAC LD cases/yr = ~600 overall• WGS to discern relapse vs. reinfection (n ~120)

• WGS to discern environment-patient MAC are similar or distinct (n ~ 500)

Aim 2• All ~100 NEW Virginia MAC LD cases/yr = ~300 overall• ~200 will be treated and complete 2 yr follow-up• Follow for clinical outcomes and in vitro/in vivo

correlations (MIC, PK/PD, Biofilm, MAC species)

NTM

LD ca

ses (

micr

obio

logi

cal c

riter

ia)

NTM

lab

repo

rts

VA Dept Health

Environmental testing

Natural history, determinants ofclinical outcome

Summary

• MAC lung disease is a growing problem at UVA and the US, particularly in thin white post-menopausal women• Most patients warrant treatment but a few do not –

depends on symptoms, radiography, and microbiology• Treatment moderately effective but is long and difficult

and often not durable - relapse/reinfection is common• MAC lung disease not simply one infection “MAC” lung

disease – speciation probably important• Appears to reflect polymicrobial infection or frequent

environmental reinfection into a bronchiectatic lung (which has poor clearance) : “Biofilm”• We need better treatments

1 new drug - Liposomal Inhaled Amikacin (Arikayce)• Dose: One vial (590 mg) daily

• Indication: treatment refractory MAC LD

• Primary endpoint was microbiological reduction

over 12 weeks: not achieved.

• However, a greater proportion of the LAI group

demonstrated at least one negative sputum culture

32% vs. 9%; and improvement in 6-minute-walk

test at week 12. > FDA approval

• 39% cough, 16% stopped rx

Lifecycle of Research

• Careful descriptive epidemiology of an important clinical problem

• Apply new thinking and new research tools to the leading questions

• Interventional study

Thank you• Darwin J. Operario• Suporn Pholwat• Alex F. Koeppel• Alyson Prorock• Yongde Bao• Katia Sol-Church• Michele Scheurenbrand• Melinda Poulter• Stephen Turner• Hardik I. Parikh• Amy Mathers• Scott Heysell