D ow nloaded from http://journals.lw w .com /jaids by BhD M f5ePH Kav1zEoum 1tQ fN 4a+kJLhEZgbsIH o4XM i0hC yw C X1AW nYQ p/IlQ rH D 3i3D 0O dR yi7TvSFl4C f3VC 4/O AVpD D a8KKG KV0Ym y+78= on 01/28/2022 Downloadedfromhttp://journals.lww.com/jaidsbyBhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78=on01/28/2022 Delayed Sample Processing Leads to Marked Decreases in Measured Plasma IL-7 Levels To the Editor: Interleukin-7 (IL-7) is a cytokine that plays a critical role in lymphocyte homeostasis. Produced by stromal cells, it stimulates thymopoiesis by providing survival signals to early thymocyte pro- genitors1 and also stimulates the pro- liferation of lymphocytes in the periphery.2,3 Several investigators have demonstrated increased IL-7 levels in CD4 lymphopenic states, such as during HIV infection or after chemotherapy.4,5 IL-7 levels have been reported in both plasma and serum, but it is not clear whether the levels from these 2 sources differ or how the different processing techniques may affect them. We compared serum and plasma IL-7 levels from healthy volunteers. Samples were either processed immedi- ately (within 15 minutes of blood draw) or 2 to 4 hours after blood draw. Blood for plasma was collected in tubes containing EDTA and was centrifuged; plasma was removed, aliquoted, and frozen atj20-C. Blood for serum was collected in serum collection tubes and allowed to clot on ice either immediately or 2 to 4 hours after blood draw. After clotting overnight, the tubes were centrifuged; serum was re- moved, aliquoted, and frozen at j20- C. Blood was collected from 10 healthy volunteers who presented consec- utively over 3 days to the National Institutes of Allergy and Infectious Dis- eases intramural AIDS research clinic. All participants were enrolled on institutional review boardYapproved study protocols and gave informed consent. IL-7 levels were measured in plasma and serum using a high-sensitivity enzyme-linked immu- nosorbent assay (ELISA; R&D Systems, Inc, Minneapolis, MN) according to manufacturer`s recommendations. Mean values were compared using paired t tests. As shown in Figure 1, the mean IL-7 levels from immediately processed plasma (10.75 pg/mL) and serum (8.05 pg/mL) were not significantly different (P = 0.17); however, they were significantly different if processing was delayed (mean, 0.75 vs. 9.31 pg/mL, respectively; P G 0.001). The measured level of IL-7 in plasma was significantly lower if processing was delayed (P G 0.001), whereas serum levels did not change significantly (P = 0.51). These results show that although there is no significant difference between plasma and serum IL-7 levels when sam- ples are processed immediately, there is a marked difference in these levels when there is a delay in processing. The difference that results from delayed processing seems only to affect plasma levels of IL-7, whereas levels in serum are stable. Both serum and plasma levels from samples that were processed im- mediately were higher than the mean values reported by the ELISA kit manufacturer for healthy volunteers (serum: mean, 2.2 (range, 0.27Y8.7) pg/mL; EDTA plasma: mean, 2.2 (range, 0.66Y9.2) pg/mL.6 We cannot exclude the possibility that these higher levels may be related to the method of processing of the samples. There are several possible mech- anisms for the decreased levels seen in plasma samples after delayed proces- sing. There may be degradation or mod- ification of IL-7 due to the presence of proteins found in plasma, but not in serum, leading to a lack of recognition by the antibodies present in the IL-7 ELISA. Likewise, there may be binding of IL-7 to proteins or to cellular receptors. It is important to note that these experiments were performed with plasma samples collected with EDTA and with the use of a single type of ELISA kit; these results cannot be extrapolated to the use of other anticoagulants or ELISA kits. These results clearly show that plasma and serum levels cannot be used interchangeably (e.g., when making com- parisons between groups), unless samples have been processed immediately after collection and appropriate comparisons for the specific laboratory and method- ology have been performed. It is also important to standardize the collection techniques when IL-7 levels are being measured, particularly in multicenter clinical trials. Finally, it is equally important to provide details of collection and processing techniques when IL-7 levels are reported in a publication. FIGURE 1. Mean plasma and serum IL-7 levels. The samples underwent either immediate processing (within 30 minutes of blood draw) or delayed processing (within 2 to 4 hours of blood draw). Error bars indicate SD. Sarah W. Read* Adam Rupert† Randy Stevens† Angeline O’Shea* Irini Sereti* *National Institute of Allergy and Infectious Diseases National Institutes of Health Bethesda, MD †AIDS Monitoring Laboratory Clinical Services Program SAIC-Frederick, Inc., NCI-Frederick Frederick, MD LETTERS TO THE EDITOR J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 511 Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. REFERENCES 1. Pallard C, Stegmann AP, van Kleffens T, et al. Distinct roles of the phophatidylinositol 3-kinase and STAT5 pathways in IL-7-mediated development of human thymocyte precursors. Immunity. 1999;10:525 Y 535. 2. Tan JT, Dudl E, LeRoy E, et al. IL-7 is critical for homeostatic proliferation and survival of naBve T cells. Proc Natl Acad Sc U S A. 2001;98:8732 Y 8737. 3. Kondrack RM, Harbertson J, Tan JT, et al. Interleukin 7 regulates the survival and generation of memory CD4 cells. J Exp Med. 2003;198:1797Y1806. 4. Fry TJ, Connick E, Falloon J, et al. A potential role for interleukin-7 in T-cell homeostasis. Blood. 2001;97:2983 Y 2990. 5. Napolitano LA, Grant RM, Deeks SG, et al. Increased production of IL-7 accompanies HIV-1-mediated T-cell depletion: implication for T-cell homeostasis. Nat Med. 2001;7:73 Y 79. 6. Quantikine HS, Human IL-7 & Immunoassay [package insert]. Minneapolis, MN: R&D Systems. Polymorphism of the Protease and Reverse Transcriptase and Drug Resistance Mutation Patterns of HIV-1 Subtype B Prevailing in China To the Editor: HIV replication can be markedly inhibited by highly active antiretroviral therapy, whereas some mutations in the genes of reverse tran- scriptase (RT) and protease, which endow the virus with drug resistance ability, are considered a major cause of treatment failure. Although many studies have focused on characterization of the polymorphisms or drug resistance muta- tions of protease and RT gene, HIV-1 subtype B isolates of North America and Europe were studied in most cases. In this study, we analyzed the polymor- phism of protease and RT gene in HIV-1 subtype B prevailing in China. The treatment-associated resistance mutation spectrum of HIV-1 RT gene during RT inhibitors (RTIs) therapy was also studied. The subjects of this survey were from 4 midland provinces of China (Anhui, Henan, Shanxi, and Sichuan) where the HIV-1 subtype B is mostly prevalent; all of the samples were collected in 2004. RNA was extracted from plasma samples, and the sequences of protease and RT were obtained by reverse transcription nested polymerase chain reaction and sequencing. All sequences belonging to subtype B were compared with the subtype B consensus sequence in Stanford HIV Drug Resis- tance Database (http://hivdb.stanford. edu) using the HIVdb software (version, 4.1.2). The W 2 tests or the Fisher exact tests were used to compare the differ- ences of the proportion of the amino acid substitution at each position between different groups or to determine the correlations between different mutation positions. One hundred seventy-four se- quences of protease and 197 sequences of RT from 197 subjects infected with HIV-1 subtype B were obtained. All of the 197 participants were of Han nation- ality, predominantly aged from 30 to 50 years (81.7%). The percentage of men was 50.8%. The large majority of patients indicated that they were most likely infected with HIV through blood donation (89.8%). All of the patients were treatment-naive of protease inhibi- tors (PIs). Seventy-one patients (36.0%) never received antiretroviral therapy; 126 other patients (64.0%) experienced RTIs therapy; 55 patients (27.9%) used a regimen of azidothymidine (AZT), di- danosine (ddI), and nevirapine (NVP); 66 patients (33.5%) used a regimen of stavudine (d4T), ddI, and NVP; 4 patients (2.0%) used a regimen of AZT, lamivudine (3TC), and NVP; and 1 TABLE 1. The Incidence of Amino Acid Substitution in Protease of HIV-1 Subtype B Isolated From 174 HIV/AIDS Patients Never Treated With PIs Position The Amino Acid of Clade B Consensus Subtype B in China Substituted Amino Acid and Proportion (%) Total Substitution Proportion (%) 5 L P (0.6), X (0.6) 1.1 10 L I (3.4) 3.4 12 T A (4.6), K (0.6), P (8.6), S (1.1) 14.9 13 I V (4.0) 4.0 14 K R (7.5) 7.5 19 L E (0.6), I (1.7), V (8.0) 10.3 31 T P (0.6), S (0.6) 1.1 33 L I (0.6), V (0.6) 1.1 35 E D (79.9), K (1.1) 81.0 37 N D (0.6), H (0.6), S (1.1) 2.3 41 R K (10.9) 10.9 43 K R (1.7) 1.7 51 G E (0.6), V (0.6) 1.1 57 R K (10.3) 10.3 60 D E (1.7) 1.7 61 Q E (2.9), H (0.6), N (0.6), S (0.6) 4.6 62 I V (20.7) 20.7 63 L C (0.6), P (90.2), S (6.9), T (2.3) 100.0 64 I L (1.7), M (1.7), V (1.7) 5.2 67 C S (0.6), W (0.6), Y (0.6) 1.7 69 H P (1.1), N (0.6) 1.7 70 K N (0.6), R (1.7) 2.3 71 A I (0.6), T (20.7), V (10.9) 32.2 72 I E (17.2), V (77.0) 94.3 76 L F (0.6), I (0.6) 1.1 77 V I (96.6) 96.6 93 I L (91.4), V (0.6) 92.0 Letters to the Editor J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 512 * 2006 Lippincott Williams & Wilkins Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. patient (0.5%) changed the regimen from d4T, ddI, and NVP to AZT, 3TC, and efavirenz (EFV). In this study, polymorphism is the variation from the consensus sequence within the subtype B, arbitrarily defined as position-specific amino acid differ- ences occurring in more than 1% of protease or RT sequences isolated from those patients who are treatment-naive of PIs or RTIs. If the proportion of an amino acid is more than 50%, then this amino acid is considered to be the inherent amino acid at this position. We found that polymorphisms oc- curred at 27 codons (27.3%) of the total 99 amino acids of protease (Table 1); the proportions of amino acid substitution at positions 63 (100.0%), 77 (96.6%), 72 (94.3%), 93 (92.0%), and 35 (81.0%) were all more than 50%; the L63P (90.2%), V77I (96.6%), I72V (77.0%), I93L (91.4%), and E35D (79.9%) were the inherent amino acid of the protease of the HIV-1 subtype B strains prevailing in China compared with the strains prevail- ing in America and Europe. In addition, the proportions of amino acid substitution were between 20% and 50% at the positions 71 (32.2%) and 62 (20.7%). Positions 63, 71, 77, and 93 are also the polymorphic positions in HIV-1 subtype B prevailing in America and Europe, and these positions are consid- ered as the secondary drug resistance mutation positions. Previous studies in- dicated that, in America and Europe, about 55% of isolates have amino acid substitution (amino acid other than L) at positions 63, about 25% at positions 77 and 93, and about 5% at position 71 in untreated persons; in heavily treated patients, those proportions reached 90% at position 63, 40% at positions 77 and 93, and 70% at position 71.1,2 In this study, the proportions of L63P, V77I, and I93L in untreated patients were even higher than those in heavily treated patients described previously. In addi- tion, the proportion of amino acid substi- tutions at protease position 71 was higher than that of subtype B prevailing in America and Europe. Although secondary mutations do not significantly alter the susceptibility of the PIs, these mutations, which exist before the emergence of a primary mu- tation, may alter the rate of development of drug resistance or the subsequent evolution of mutations.3 It is reported that a relatively poorer response to ther- apy was associated with high-baseline protease polymorphism and with the pres- ence of I93L and A71V/T at baseline.4 Therefore, the high frequencies of baseline amino acid substitutions at positions 63, 71, 77, and 93 of protease in HIV-1 subtype B prevailing in China may have potential influence on the curative effects of PIs that will be introduced in antiretroviral therapy of China hereafter. In this study, 71 RT sequences were obtained from untreated patients. Seventy-four (24.7%) of the first 300 amino acid positions of RT were found to be polymorphic. The proportions of amino acid substitution at positions 278 (100.0%), 162 (95.8%), 200 (90.1%), 135 (83.1%), and 277 (74.2%) were more than 50%; the Q278E (100%), S162C (93.0%), T200A (81.7%), I135V (69.0%), and K277R (74.2%) were the inherent amino acid of RT. In addition, relatively high proportions of amino acid substitution were found at positions 211 (31.0%) and 293 (24.2%). All amino acid substitutions previously mentioned have not been reported to be associated with the RTIs resistance, except for po- sition 135. Previous study indicated that those strains containing the I135M/L/T substitution in combination with the L283I substitution showed decreases in susceptibility to all 3 of the nonnucleo- side reverse transcriptase inhibitors (NNRTIs).5 In our study, although the total proportion of I135L/T is 9.8%, the polymorphism at position 135 would not influence the susceptibility of HIV- 1 to drugs because there was no L238I existing together with I135L/T. In this study, only 3 isolates from 71 treat- ment-naive patients harbored 3 nucleo- side reverse transcriptase inhibitors (NRTIs) resistanceYassociated muta- tions (A62V, Y115F, and T215A); no NNRTIs resistance mutation was found. These results demonstrated that the frequency of baseline RTIs resistance mutations was very low among these treatment-naive patients. In the RT sequences from 126 patients undergoing RTIs therapy, we found 6 kinds of NRTIs resistance mu- tations at 4 amino acid positions: K65R (3.2%), D67G (1.6%), D67N (1.6%), L74V (4.8%), T215F (0.8%), and T215Y (2.4%). In contrast, we found 10 kinds of NNRTIs resistance mutations at 7 amino acid positions: K101E (2.4%), K103N (34.1%), K103S (1.6%), V106A (4.0%), Y181C (23.8%), Y188C (0.8%), Y188L (0.8%), G190A (15.1%), G190S (1.6%), and K238T (1.6%). Notably, the propor- tions of K103N, Y181C, and G190A mutations were significantly higher in treated patients than in untreated patients (P = 0.000, 0.000, and 0.001, respective- ly). These results indicated that the emergence of NNRTIs resistance muta- tions is also an important problem during antiretroviral therapy in China. To find whether there were novel mutations in RT associated with anti- HIV therapy, the proportions of amino acid substitution at each mutation position were compared between RTI- treated and untreated groups. We found that position 221 was highly conserved in HIV-1 isolates from untreated group (proportion, 0/71), whereas amino acid substitution at this position occurred in treated patients with higher proportion (14/126). Moreover, as the most fre- quent amino acid substitution of this position, the proportion of H221Y was significantly higher in RTI-treated patients (13/126 [10.3%]) than that in untreated patients (P = 0.012), which indicated that H221Y mutation was associated with RTIs treatment. Furthermore, we compared the proportions of H221Y between 2 groups with and without K103N or Y181C mutationsVthe 2 most frequent drug resistance mutations in RT after RTIs treatment in our study. The proportion of H221Y in the group with Y181C was significantly higher than that in the group without Y181C (40.0% vs 1.0%, P = 0.000). However, there was no significant difference in proportion of H221Y between 2 groups with and without K103N (14.0% vs 8.4%, P = 0.511). H221Y is a novel therapy-related mutation first reported by Gonzales6 in 2003. Gonzales et al proposed that H221Y was associated with NRTI ther- apy. But this proposition is discrepant with the results of a recent study in which the H221Y was indicated to be associated with NNRTI resistance.7 In our study, the H221Y mutation was found to be correlated with Y181C mutation (L = 0.545), whereas Y181C J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 Letters to the Editor * 2006 Lippincott Williams & Wilkins 513 Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. has been reported to be NNRTIs resis- tance mutation; thus, we speculate that the H221Y mutation is more likely associated with the NNRTIs resistance. In summary, there are inherent amino acids at some positions of prote- ase and RT of HIV-1 subtype B strains prevailing in China compared with the strains in America and Europe. As the secondary PIs resistance muta- tions, the high frequencies of L63P, V77I, and I93L may have potential influence on the curative effect of PIs that will be used in China. The muta- tions K103N, Y181C, and G190A were the 3 most prevalent drug resistance mutations after antiretroviral therapy in China, and H221Y was a novel mutation more likely associated with the NNRTI therapy. REFERENCES 1. Bossi P, Mouroux M, Yvon A, et al. Polymorphism of the human immunodeficiency virus type 1 (HIV-1) protease gene and response of HIV-1-infected patients to a protease inhibitor. J Clin Microbiol. 1999;37:2910Y2912. 2. Wu TD, Schiffer CA, Gonzales MJ, et al. Mutation patterns and structural correlates in human immunodeficiency virus type 1 protease following different protease inhibitor treatments. J Virol. 2003;77:4836Y4847. 3. Kozal MJ, Shah N, Shen N, et al. Extensive polymorphisms observed in HIV-1 clade B protease gene high-density oligonucleotide arrays. Nat Med. 1996;7:753Y759. 4. Servais J, Lambert C, Fontaine E, et al. Variant human immunodeficiency virus type 1 proteases and response to combination therapy including a protease inhibitor. Antimicrob Agents Chemother. 2001;45:893Y900. 5. Brown AJ, Precious HM, Whitcomb JM, et al. Reduced susceptibility of human mmunodeficiency virus type 1 (HIV-1) from patients with primary HIV infection to nonnucleoside reverse transcriptase inhibitors is associated with variation at novel amino acid sites. J Virol. 2000;74:10269Y10273. 6. Gonzales MJ, Wu TD, Taylor J, et al. Extended spectrum of HIV-1 reverse transcriptase mutations in patients receiving multiple nucleoside analog inhibitors. AIDS. 2003;17: 791Y799. 7. Saracino A, Monno L, Scudeller L, et al. Impact of unreported HIV-1 reverse transcriptase mutations on phenotypic resistance to nucleoside and non-nucleoside inhibitors. J Med Virol. 2006;78:9Y17. Depressive Symptoms Predict Increased Incidence of Neuropsychiatric Side Effects in Patients Treated With Efavirenz To the Editor: Efavirenz (EFV), a nonnucleoside reverse transcriptase inhibitor (NNRTI) of HIV-1, is associated with neuropsy- chiatric central nervous system side ef- fects not always seen with other members of the NNRTI class. Although the mech- anism of action responsible for these side effects has not been elucidated, EFV has a 60% homology with antidepressants, with autonomic and cognitive adverse events.1,2 The neuropsychiatric side effects secondary to EFV include dizzi- ness, confusion, stupor, impaired con- centration, amnesia, hallucinations, abnormal dreams, and insomnia.3 Most of the EFV-related neuropsychiatric side effects occur within the first several days of treatment and tend to resolve within the first 6 to 10 weeks of therapy; how- ever, they may be serious and can in- clude anxiety, depression, and suicidal ideation.4,5 We were particularly inter- ested in evaluating neuropsychiatric side effects in patients able to tolerate an EFV-containing regimen for more than 4 weeks while still experiencing neuropsychological side effects. By identifying an association between de- pression history and neuropsychological symptoms, we hoped to better predict the development and progression of long- term EFV-related neuropsychiatric side effects, which could negatively affect treatment adherence.6,7 Of all patients in the clinic over the set period, medical records of 110 eligible patients receiving EFV-based therapy from 6 participating private HIV practices in the San Francisco Bay area were reviewed to examine the number of neuropsychiatric side effects and the number of patients discontinuing EFV because of these effects. Baseline demographics, CD4+ cell counts, and viral loads were recorded along with concur- rent medications. Patients with docu- mented or suspected use of cocaine, methamphetamines, ecstasy, ketamine, F-hydroxybutyrate, or heroin were excluded from this statistical analysis. Patients not receiving EFV-based anti- retroviral therapy (ART) for more than 4 weeks were also excluded. A history of depression was defined as one or a com- bination of the following factors diag- nosed and resolved more than 6 months before starting an EFV-containing reg- imen: (a) patient history of depressive symptoms, (b) patient under the care of a psychiatrist/therapist, and (c) patient under antidepressive treatment. The length of time the patient remained on EFV before discontinuation was noted, as was the viral load and CD4+ count at the time of drug discontinuation. Cur- rent depression was defined as docu- mented depression in the patient’s chart within 6 months of starting EFV. Charts were further examined to identify the occurrence and magnitude of EFV- related depressive symptoms during EFV treatment. Using a specialized eva- luation questionnaire as per the Sustiva (Bristol-Myers Squibb, New York, NY) package insert, developed to standardize data collection, EFV-related neuropsy- chiatric symptoms were graded on a scale of 1 to 5 by frequency of occur- rence: grade 1, never experienced; grade 2, rarely (0Y1 day a week); grade 3, sometimes (2Y4 days a week); grade 4, very often (4Y5 days a week); and grade 5, most of the time (6Y7 days a week). For the primary analysis, the number of discontinuations for each group was tabulated, and a relative risk (RR) and 95% confidence interval for the odds ratio were calculated. The P value for the difference between odds ratio was then determined. For the secondary depres- sion subcategory analysis, median scores Shulin Jiang, PhD Hui Xing, MSc Xuefeng Si, MSc Yu Wang, MSc Yiming Shao, PhD Division of Research on Virology and Immunology, State Key Laboratory for Infectious Disease Prevention and Control (Chinese Center for Disease Control and Prevention), National Center for AIDS/STD Control and Prevention, Chinese Center for Disease Control and Prevention, Beijing People’s Republic of China Letters to the Editor J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 514 * 2006 Lippincott Williams & Wilkins Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. for each of the previously defined groups were calculated, with the difference in the median values assessed by a Wil- coxon rank sum test. The charts of 110 patients were screened. Most of the patients were men having sex with men (99%), with a me- dian age of 43 years. All statistical analyses were performed on 101 re- maining patients unless otherwise stated. The median baseline CD4+ count was 336 cells/mm3, and the median viral load was 25,000 copies/mL, both with wide ranges. Fifty patients (50%) had a history of depression; of these, 56% were cur- rently depressed. History of depression was further divided into subcategories: reported by patient (n = 5), depression requiring medication (n = 30), and depres- sion requiring psychotherapy (n = 19). Four patients reported depression requir- ing both medication and psychotherapy in the past; one discontinued EFV-based ART. Relative risk for current depression was significantly higher for patients with any history of depression than for patients without such a history (RR, 31.2; 95% CI, 6.81Y142.6, P G 0.0001). Eighteen per- cent of patients with a history of depres- sion discontinued EFV because of treatment-related symptoms of depression (RR, 11; 95% CI, 1.3Y90.2, P = 0.0079), whereas only 2% of patients who did not have a history of depression at any time discontinued EFV (RR, 10.0, 95% CI, 1.2Y82.6, P = 0.016). All patients with a history of depression and/or current depression developed at least one subcat- egory depressive symptom of grade 2 severity or higher. All depression sub- categories associated with EFV, exclud- ing the Bunintentional weight loss of more than 5%^ subcategory and the Bsuicidal ideation/attempt^ subcategory, were sta- tistically significantly higher (P e 0.0009) in patients with a history of depression than in patients without such a history. Our study revealed an overall inci- dence of neuropsychiatric symptoms of any kind while on an EFV-containing regimen of 30%, with a 10% discontinu- ation rate due to EFV-associated neuro- psychiatric side effects at or after 4 weeks of treatment. Disparity exists between registration/investigational studies and clinical studies performed within primary HIV practices. It has been suggested that in Breal world^ settings, there is a much greater likelihood that these symptoms will occur because of unrecognized addi- tional pressures, patient risk-taking, and comorbidities that could negatively affect a patient’s outcome.8 Discontinuation of therapy did not appear to be related to virological failure, as 40% of the patients discontinuing EFV therapy after 4 weeks continued to have viral loads less than 50 copies/mL. If these results are confirmed, this compo- nent of treatment failure could poten- tially be anticipated and adjusted for. Our results tend to support the complex nature of Btreatment success^ and show clearly that one can continue to Bsucceed^ virologically while Bfailing^ in some other necessary requirement of ART. It should be noted that the use of alcohol and marijuana was high in our study, but no association was found between their use and the incidence of depression or treatment discontinuation. The preva- lence of other documented recreational drug use was too low to allow analysis of any potential associations. If depressive symptoms cannot be tolerated by the patient or ameliorated by psychotherapy, alteration of the ART regimen may be warranted. Other NNRTIs such as nevirapine and protease inhibitors such as lopinavir/ritonavir have not been associated with neuropsy- chiatric effects of the same severity as those seen with EFV and yet have been shown to have comparable efficacy.9,10 Limitations of this study include the fact that this was a retrospective chart review, and diagnosing past depression based on chart documentation may either overestimate or underestimate rates of depression. Moreover, it failed to clarify specific diagnoses, such as major depres- sion versus bipolar depression versus dysthymia, etc, and men having sex with men was the only subpopulation well represented in this sample. Future research that controls for other variables and has a comparison group of patients not taking EFV would help establish whether EFV is the direct cause of new depressive symptoms. ACKNOWLEDGMENTS The authors gratefully acknowl- edge the editorial assistance of Michael Craig and Benedetta Granelli. Larry Boly Virginia Cafaro Toby Dyner Private Practice, Shared Perspectives on Therapies San Francisco, CA REFERENCES 1. Carvalhal AS, de Abreu PB, Spode A, et al. An open trial of reboxetine in HIV-seropositive outpatients with major depressive disorder. J Clin Psychiatry. 2003;64:421Y 424. 2. Boly L, Dyner T, Cafaro V. A history of depressive symptoms is predictive of an increased incidence of central nervous system (CNS) side effects in patients treated with efavirenz (EFV). In: Program and abstracts of the 8th European Conference on Clinical Aspects and Treatment of HIV-Infection; October 28Y31, 2001; Athens, Greece. Abstract 149. 3. Richman DD, Fischl MA, Grieco MH, et al. The toxicity of azidothymidine (AZT) in the treatment of patients with AIDS and AIDS- related complex. A double-blind, placebo- controlled trial. N Engl J Med. 1987;317:192Y197. 4. Raines C, Radcliffe O, Treisman GJ. Neurologic and psychiatric complications of antiretroviral agents. J Assoc Nurses AIDS Care. 2005;16(5):35Y48. 5. Lochet P, Peyriere H, Lotthe A, et al. Long- term assessment of neuropsychiatric adverse reactions associated with efavirenz. HIV Med. 2003;4:62 Y 66. 6. Starace F, Ammassari A, Trotta MP, et al. Depression is a risk factor for suboptimal adherence to highly active antiretroviral therapy. J Acquir Immune Defic Syndr. 2002;31:S136 Y S139. 7. Tucker JS, Burnam MA, Sherbourne CD, et al. Substance use and mental health correlates of nonadherence to antiretroviral medications in a sample of patients with human immunodeficiency virus infection. Am J Med. 2003;114:573 Y 580. 8. Guest JL, DeSilva KE, Flaherty TD, et al. Incidence of CNS effects in patients with a history of mental health disorders and/or substance abuse starting an efavirenz- containing regimen: data from HAVACS. In: Program and abstracts of the XIV International AIDS Conference; July 7Y12, 2002; Barcelona, Spain. Abstract/poster TuPeB4502. 9. Ward DJ, Curtin JM. Switch from efavirenz to nevirapine associated with continued viral suppression as well as resolution of efavirenz- related neuropsychiatric adverse events and improvements in lipid profiles. AIDS Patient Care and STDs. 2005: In press. 10. Rockstroh J, Shen Y, Bortolozzi R, et al. Assessment of depression in patients after substitution of their PI/NNRTI with lopinavir/ritonavir (LPV/r). Antivir Ther. 2003;8(suppl 1):S392. Abstract 549. J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 Letters to the Editor * 2006 Lippincott Williams & Wilkins 515 Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Further Information on the Administration of H2-Receptor Antagonists With Atazanavir To the Editor: Further data on the use of ataza- navir (ATV) with H2-receptor antago- nists and proton pump inhibitors (PPIs) have become available since the letter published by Drs. Homayoun Khanlou and Charles Farthing.1 Previously, ATV and PPI interaction studies demonstrat- ed that ATV exposures were reduced by more than 70%, leading to a recommen- dation that PPIs should not be used concomitantly with ATV.2,3 Studies to evaluate other dosing strategies with ATV and PPIs are planned. More recently, data have become available on the interaction between ATV and H2-receptor antagonists, indicating that reduced exposures to ATV resulting from an interaction with H2-receptor antagonists may be attenuated by con- current use of low-dose ritonavir (RTV) or by temporal separation from an H2-receptor antagonist.4 In 2 drug interaction studies of 108 HIV-negative subjects, the effect of the commonly used H2-receptor antagonist famotidine at 40 mg twice daily on ATV at 400 mg and ATV/RTV at 300/100 mg was evaluated.4 Relative to a control arm of ATV at 400 mg alone, coadministra- tion of famotidine at 40 mg twice daily decreased ATV exposures by approxi- mately 40%. Temporal separation (dos- ing ATV 10 hours after and 2 hours before the famotidine dose) attenuated the reduction in ATV exposures. Fur- thermore, ATV at 300 mg coadminis- tered with RTV at 100 mg led to a 79% and 4.5-fold increase in area under the curve and Cmin, respectively, with Cmax values similar to ATV at 400 mg alone. Relative to a control arm of ATV/ RTV at 300/100 mg alone, coadminis- tration with famotidine at 40 mg twice daily modestly reduced ATV exposures by 18%, 14%, and 28% for area under the curve, Cmax, and Cmin, respectively. These exposures were similar to an antiretroviral regimen of ATV/RTV with 2 nucleosides including tenofovir, the efficacy of which has been estab- lished in a well-controlled study in HIV- infected subjects relative to a standard of care regimen.5 Thus, ATV/RTV at 300/100 mg may be given concomitantly in combination with H2-receptor antago- nists. Atazanavir/RTV at 300/100 mg is recommended with H2-receptor antago- nists, although the reductions in ATV exposures with maximal doses of famo- tidine could be overcome by increasing the dose of ATV to 400 mg while keep- ing the dose of RTV steady at 100 mg. Atazanavir/RTV at 400/100 mg is not recommended because there is a potential for increased risk of ATV intolerability, resulting from increased ATV exposures that could occur with less potent, lower doses, or sporadic dosing of H2-receptor antagonists in combination with higher doses of ATV. To further minimize any reduction in ATV exposures, an alterna- tive to simultaneous dosing of H2-recep- tor antagonists with ATV/RTV at 300/100 mg is temporal separation, which may be attained by administering ATV/RTV at 300/100 mg at least 10 hours after and at least 2 hours before an H2-receptor antagonist; temporal separa- tion should be considered in treatment- experienced patients. In summary, although H2-recep- tor antagonists increase intragastric pH and have been shown to reduce ATV absorption, H2-receptor antagonists are less potent than PPIs in suppressing intragastric acid secretion and may be used with ATV. Atazanavir/RTV at 300/100 mg may be administered concomitantly with H2-receptor antago- nists. Alternatively, ATV at 400 mg or ATV/RTV at 300/100 mg may be temporally separated by administering ATV at least 10 hours after and at least 2 hours before an H2-receptor antago- nist to minimize the interaction; tempo- ral separation of ATV/RTV at 300/100 mg should be considered in treatment- experienced patients. Local prescribing information may vary; therefore, clini- cians using ATV with H2-receptor antago- nists should consult their local package inserts for specific information. Additional studies to evaluate alternative dosing strategies with ATV and H2-receptor antagonists are ongoing. REFERENCES 1. Khanlou H, Farthing C. Co-Administration of atazanavir with proton-pump inhibitors and H2 blockers. J Acquir Immune Defic Syndr. 2005;39:503. 2. Agarwala S, Gray K, Eley T, et al. Pharmacokinetic interaction between atazanavir and omeprazole in healthy subjects. Presented at: 3rd IAS Conference on HIV Pathogenesis and Treatment; July 24Y27, 2005; Rio de Janeiro, Brazil. Abstract WePe3-3C08. Available at: http://ww2.aegis.org/conferences/ iashivpt/2005/WePe3-3C08.html. Accessed February 24, 1006. 3. Agarwala S, Gray K, Wang Y, et al. Pharmacokinetic effect of omeprazole on atazanavir co-administered with ritonavir in healthy subjects. Presented at: 12th Conference on Retroviruses and Opportunistic Infections; February 22Y25, 2005; Boston, MA. Abstract 658. Available at: http://www.retroconference. org/2005/CD/Abstracts/24122.htm. Accessed February 24, 2006. 4. Agarwala S, Eley T, Child M, et al. Pharmacokinetic effect of famotidine on atazanavir with and without ritonavir in healthy subjects. Presented at: 6th International Workshop on Clinical Pharmacology of HIV Therapy; April 28Y30, 2005; Quebec City, Quebec, Canada. Abstract 11. 5. Johnson M, Grinsztejn B, Rodriguez C, et al. Atazanavir plus ritonavir or saquinavir, and lopinavir/ritonavir in patients experiencing multiple virological failures. AIDS. 2005;19:685Y694. The Correlation Between Prevalence of Oral Manifestations of HIV and CD4+ Lymphocyte Counts Weakens With Time To the Editor: Oral manifestations of HIV (OMHs) have been well documented since the onset of the HIV pandemic as early markers of HIV infection and Sangeeta Agarwala, PhD* Gary Thal, MD† Richard Nettles, MD* Richard Bertz, PhD* *Bristol-Myers Squibb Co, Princeton, NJ and †Bristol-Myers Squibb Co, Plainsboro, NJ sangeeta.agarwala@bms.com Letters to the Editor J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 516 * 2006 Lippincott Williams & Wilkins Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. progression.1 OMHs are commonly di- agnosed and therefore contribute signif- icantly to HIV-related morbidity. Oral candidiasis (OC) and oral hairy leuko- plakia, 2 lesions associated with fungal and viral pathogens, respectively, are the most frequently occurring OMHs. Other OMHs, such as human papillomavirus- related warts, aphthous-like ulcers, and Kaposi sarcoma, have also been exten- sively reported. The risk of developing an OMH correlates significantly with a decreas- ing peripheral CD4 count, a marker of HIV disease progression.2 Therefore, although the treatment of specific OMHs has been reasonably successful, it is evident that the most successful therapeutic approach is to prevent or reverse the underlying primary immu- nodeficiency disease and thereby pre- vent OMHs. Unsurprisingly, the introduction of highly active antiretroviral therapy (HAART) has resulted in a significant decrease in the prevalence of opportu- nistic diseases including that of OMHs. Seemingly in contradiction with those findings, despite a marked improvement in CD4 cell counts, an increased prev- alence in oral warts has been noted by some investigators.3 Taken together, the epidemiology of OMHs in the HAART era seems to indicate that OMHs are, as a whole, less frequent, but that new and poorly understood paradoxes are emerg- ing. It has been suggested that those apparent paradoxes could be explained at least partially by an incomplete functional reconstitution of the immune system as a result of HAART4 or by a transient deterioration of immune func- tion referred to as the immune reconsti- tution syndrome.5 The quantification of peripheral CD4 T lymphocytes would then not accurately reflect the functional status of the immune system against all opportunistic pathogens and diseases. We hypothesized that in patients with less than 10 years of diagnosed HIV infection, a lower CD4 count would be associated with an increased risk for OMHs, whereas this association may not exist for the subjects with a longer history of diagnosed HIV infection. The target population for our study was the Ryan White Dental Clinic at the Univer- sity of Illinois at Chicago. We stratified the subjects into 2 groups, a group with a history of HIV infection shorter than 10 years (group 1) who benefited from HAART throughout their HIV infection and a group with a history of HIV infection longer than 10 years who bene- fited from a later HAART-induced immune reconstitution (group 2). The study was conducted with the approval of the institutional review board at the University of Illinois at Chicago. The presumptive oral lesions were diagnosed according to the criteria described by Greenspan and Greenspan6 A normal examination was defined as a subject in whom no HIV-associated lesion was diagnosed during the clinical oral evaluation. Patients also completed a self-report questionnaire which includes demographic variables, such as age and ethnicity, as well as psychosocial vari- ables, such as having dental insurance and receiving counseling. Group 1 was made up of 151 individuals, with a mean age of 42.77 years (range, 18Y79 years), the male/ female ratio was 4.35, and the mean time since seroconversion was 62.25 months (T32.91). Group 2 was made up of 180 individuals, with a mean age of 46.71 years (range, 11Y68 years), the male/female ratio was 4.55, and the mean time since seroconversion was 178.55 months (T44.81). Forty-seven percent of the sub- jects in group 1 had at least one OMH with an overall average of 1.41 T 0.55 per subjects with at least one lesion. Forty-nine percent of the subjects in group 2 had at least one OMH with an overall average of 1.58 T 0.76 per subjects with at least one lesion. The most common lesions for both groups were OC and hairy leukoplakia which together represented more than two thirds of all diagnosed oral lesions. To determine which variables might be confounding the relationship between normal exam and CD4 count, zero-order correlations were analyzed for the following variables: smoking, number of cigarettes smoked per day, HAART, sex, ethnicity, and dental insurance coverage. Confounding vari- ables examined were cigarette smoking, due to the statistically significant rela- tionship to normal examination, and HIV medication, based on theoretical rationale and previous reports associat- ing a lower risk for OMH and HAART independently of CD4 counts.3 Whether a patient reported smok- ing cigarettes did not impact the rela- tionship between normal examination and CD4 count or between the number of lesions a person has and CD4 count (Table 1). For group 1, the partial correlation between normal examination TABLE 1. Partial Correlations (Controls for Variable 1); Pearson Correlations (Variables 2 and 3) Variable 1 Variable 2 Variable 3 Partial r (sig) Correlation (sig) Subjects with a documented HIV infection shorter or equal than 10 years (group 1) Smokes Normal examination CD4 0.263 (0.007) 0.270 (0.006) Smokes No. lesions CD4 j0.280 (0.004) j0.286 (0.003) No. cigarettes smoked per day No. lesions CD4 j0.290 (0.003) HAART Normal examination CD4 0.206 (0.061) Subjects with a documented HIV infection longer than 10 years (group 2) Smokes Normal examination CD4 0.086 (0.324) 0.093 (0.470) Smokes No. lesions CD4 j0.097 (0.264) j0.104 (0.231) No. cigarettes smoked per day No. lesions CD4 j0.111 (0.201) HAART Normal examination CD4 0.144 (0.174) J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 Letters to the Editor * 2006 Lippincott Williams & Wilkins 517 Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. and CD4 count remained significant (zero-order r = 0.270, P = 0.006; partial r = 0.263, P = 0.007) and between number of lesions a person has and CD4 count remained significant (zero-order r = j0.286, P = 0.003; partial r = j0.28, P = 0.004) when controlled for smoking. Furthermore, the relationship between number of lesions a person has and CD4 count remained significant when the reported number of cigarettes smoked per day was controlled for group 1 (zero- order r = j.286, P = 0.003; partial r =j.29,P = 0.003). Conversely, all zero- order and partial correlations for these variables were nonsignificant for group 2. In addition to controlling for smoking behaviors, whether the person is currently on HAART was also con- trolled for in both groups. For group 1, the relationship between normal exam- ination and CD4 count still approached significance after controlling for HAART (zero-order r = 0.270, P = 0.006; partial r = 0.206, P = 0.061). Although the zero-order correlation is significant and the partial correlation only approaches significance, the magnitude of these correlations supports a relationship be- tween normal examination and CD4 otherwise not explained by whether the person is taking HAART. Again, the zero- order and partial correlations for these variables were nonsignificant for group 2. Overall, CD4 counts are associat- ed with oral lesions when controlling for smoking behaviors and HAART for people who contracted HIV within the last 10 years. The same pattern was not true for people who contracted HIV more than 10 years ago. However, these 2 groups did not differ on the other test variables (data not shown). This study indicates that oral lesions remain a common morbidity in this population. Although our data, which are not longitudinal, do not allow testing whether OMHs are less frequent since the introduction of HAART, they clearly indicate that OMHs remain a common problem in HIV-infected patients in the HAART era. In our study population, close to 50% of the subjects had at least one OMH and more than a third of the subjects were diagnosed with OC, a marker of poor HIV/AIDS prognosis.7 The predominant OMHs diagnosed in this study were OC followed by oral hairy leukoplakia, a traditional distribu- tion which does not support a greater proportional weight for oral warts noted by some in association with HAART.8 Finally, this study presents evi- dence that the commonly described correlation between CD4 count and OMHs is not seen after more than 10 years of HIV infection. We attribute this finding to incomplete immune reconsti- tution associated with HAART. Then again, this finding could be an integral part of HIV natural history independent- ly on the type of medications a patient may benefit from. Separately of this line of reasoning, we contend that although CD4 counts are merely surrogate mar- kers for HIV disease progression, OMHs may reflect a more accurate representa- tion of immune health. If so, this study may indicate that traditional methods assessing HIV progression and treat- ment response, often heavily based on stable or increasing CD4 counts, may fall short of accurately measuring im- mune health in patients with a history of HIV disease longer than 10 years. Future studies will address this issue by testing whether OMHs in this population is linked to disease progression and treat- ment failure independently of CD4 counts. REFERENCES 1. Greenspan D, Greenspan JS, Overby G, et al. Risk factors for rapid progression from hairy leukoplakia to AIDS: a nested case-control study. J Acquir Immune Defic Syndr. 1991;4(7):652Y658. 2. Adurogbangba MI, Aderinokun GA, Odaibo GN, et al. Oro-facial lesions and CD4 counts associated with HIV/AIDS in an adult population in Oyo State, Nigeria. Oral Dis. 2004;10(6):319Y326. 3. Greenspan D, Canchola AJ, MacPhail LA, et al. Effect of highly active antiretroviral therapy on frequency of oral warts. Lancet. 2001;357(9266):1411Y1412. 4. Lange CG, Valdez H, Medvik K, et al. CD4+ T-lymphocyte nadir and the effect of highly active antiretroviral therapy on phenotypic and functional immune restoration in HIV-1 infection. Clin Immunol. 2002;102(2): 154Y161. 5. Lipman M, Breen R. Immune reconstitution inflammatory syndrome in HIV. Curr Opin Infect Dis. 2006;19(1):20Y25. 6. Greenspan D, Greenspan JS. Oral manifestations of HIV infection. AIDS Clin Care. 1997;9(4):29Y33. 7. Dodd CL, Greenspan D, Katz MH, et al. Oral candidiasis in HIV infection: pseudomembranous and erythematous candidiasis show similar rates of progression to AIDS. AIDS. 1991;5(11):1339Y1343. 8. Cameron JE, Mercante D, O’Brien M, et al. The impact of highly active antiretroviral therapy and immunodeficiency on human papillomavirus infection of the oral cavity of human immunodeficiency virus-seropositive adults. Sex Transm Dis. 2005;32(11): 703Y709. Comparison of HIV-1 Mother-to-Child Transmission After Single-Dose Nevirapine Prophylaxis Among African Women With Subtypes A, C, and D To the Editor: The HIVNET 012 trial in Uganda showed that mother-to-child transmis- sion (MTCT) of HIV-1 can be prevented by providing pregnant women and their infants with a single dose (SD) of the antiretroviral drug, nevirapine (NVP).1,2 Safety and efficacy of 1- or 2-dose NVP prophylaxis for prevention of MTCT have been documented in other studies. We have shown that NVP resistance emerges in some women after SD NVP prophylaxis3 and that the portion of women with NVP resistance is influ- enced by HIV-1 subtype.4 At 6 to 8 weeks after SD NVP, NVP resistance was more common in women with sub- type C (69.2%) than in women with subtype D (36.1%, P G 0.0001) or subtype A (19.4%, P G 0.0001).4 Selec- tion of NVP-resistant HIV-1 variants Herve Y. Sroussi, DMD, PhD* Dana Villines, MA† Joel Epstein, DMD, MSD, FRCD(c)* Marilda, C. F. Alves† Mario E. A. F. Alves, DDS, MS, DSc† *Department of Oral Medicine and Diagnostic Sciences University of Illinois College of Dentistry and †Department of Periodontics University of Illinois College of Dentistry Letters to the Editor J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 518 * 2006 Lippincott Williams & Wilkins Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. in women after NVP dosing could theoretically lower the efficacy of NVP prophylaxis for prevention of HIV transmission by breast-feeding in the first few weeks after birth. In the HIVNET 012 trial, most women were infected with HIV-1 subtype A or D. Risk of MTCT was slightly (but not statistically) higher in women with subtype D.5 In this report, we com- bined data from the HIVNET 012 and NVAZ trials6 to compare the risk of MTCT in women with subtype C to the risk of MTCT in women with subtypes A and D in the setting of SD NVP prophylaxis. METHODS Diagnosis of HIV-1 Infection in Infants Methods used for diagnosis of HIV-1 infection in infants have been described previously for the HIVNET 0121 and NVAZ6 trials. Viral Load Assays HIV viral load testing of maternal plasma was performed using the Roche Amplicor Monitor Test Kit (Branchburg, NJ). Version 1.5 of the test kit was used in the NVAZ trial and in the HIVNET 012 trial after November 1998; before November 1998, samples in HIVNET 012 were tested with version 1.0 of the kit with additional primers. HIV Subtyping HIV subtyping was performed by phylogenetic analysis of pol region sequences, as previously described.3 This study included 270 (88%) of the 306 Ugandan women who received SD NVP in HIVNET 012 (158 with subtype A, 7 with subtype C, and 105 with sub- type D). Thirty women with intersubtype recombinant HIV-1 and 6 women with indeterminate subtypes were not included in this study. This study also included 445 Malawian women who received SD NVP in the NVAZ trial.6 HIV-1 subtypes were obtained for 67 (15%) of those women; all had subtype C HIV-1 with no evidence of intersubtype recombination. HIV-1 subtypes were also determined for 91 women enrolled in a related study of HIV-1 MTCT performed at the same clinical site (the NVAZ late presenter trial)7; all of those women also had subtype C HIV-1. Identification of sub- type C in all 158 of the Malawian women tested provides 95% confidence that at least 370 (98%) of the 378 additional women from NVAZ who were included in this report but who did not have subtyping performed also had subtype C. Based on these calculations, we assumed that all 445 of the Malawian women in NVAZ had subtype C infection. Statistical Methods Unadjusted associations of HIV-1 subtype with transmission at birth and at 6 to 8 weeks were estimated using exact tests, and adjusted associations were estimated using logistic regression. Con- fidence intervals (CIs) for MTCT were calculated using exact methods. The multivariate model included all 3 sub- types, with subtype C as a reference. SAS version 9.1 (SAS Institute, Inc., Cary, NC) performed these analyses. RESULTS We analyzed MTCT in 715 Afri- can women who received the HIVNET 012 regimen (SD NVP to the mother in labor and SD NVP to the infant shortly after birth). This cohort included 158 with subtype A, 452 with subtype C, and 105 with subtype D (see Methods). The mean log10 delivery viral load was similar among women with subtypes A (median, 4.28), C (median, 4.45), and D (median, 4.51; P = 0.46, Kruskal-Wallis test) and was similar among women in HIVNET 012 (median, 4.39, all sub- types) versus NVAZ (median, 4.46; P = 0.23 [Wilcoxon rank test]). Data on MTCT were available from HIVNET 012 to infant age 18 months and from NVAZ to infant age 24 months. At 6 to 8 weeks, breast-feeding was almost universal in HIVNET 012 and NVAZ (282/283 [99.7%] and 338/341 [99.1%], respectively; P = 0.63 [Fisher exact test]). In contrast, at 6 months, the portion of women who were breast-feeding was significantly higher in NVAZ (299/305 [98.0%]) than in HIV- NET 012 (219/258 [84.9%]; P G 0.0001 [Fisher exact test]). For this reason, we confined our analysis to transmission at birth and at 6 to 8 weeks of age. In univariate analysis, there was no difference in the risk of MTCT among women with subtypes A, C, and D at birth or 6 to 8 weeks (for A vs C vs D: for infants infected at birth, P = 0.28; for infants uninfected at birth but in- fected at 6Y8 weeks, P = 0.58; and for infants infected at 6Y8 weeks, P = 0.16 [Fisher exact test]; Table 1). In a multivariate model including log10 HIV-1 maternal viral load at delivery, age, parity, and HIV-1 subtype, MTCT was associated with maternal viral load for infants infected at birth (odds ratio [OR], 2.01; 95% CI, 1.36 Y2.96; P = 0.0004), for infants uninfected TABLE 1. Risk of HIV-1 MTCT Among African Women Who Received Single-Dose Nevirapine Prophylaxis HIV Subtype* OR [95% CI] A (n = 158) C (n = 452)† D (n = 105) R (n = 30) C (Reference) A D Infected at birth‡ 8/156 (5.1%) [2.2%Y9.9%] 41/452 (9.1%) [6.6%Y12.1%] 9/104 (8.7%) [4.0%Y15.8%] 5/30 (16.7%) [5.6%Y34.7%] 1.00 0.54 [0.21Y1.21] 0.95 [0.39Y2.07] Uninfected at birth but infected at 6Y8 wk‡ 7/145 (4.8%) [2.0%Y9.7%] 22/357 (6.2%) [3.9%Y9.2%] 3/93 (3.2%) [0.7%Y9.1%] 1/24 (4.2%) [0.1%Y21.1%] 1.00 0.77 [0.27Y1.93] 0.51 [0.10Y1.75] Infected at 6Y8 wk‡ 15/153 (9.8%) [5.6%Y15.7%] 63/398 (15.8%) [12.4%Y19.8%] 12/102 (11.8%) [6.2%Y19.7%] 6/29 (20.7%) [8.0%Y39.7%] 1.00 0.58 [0.30Y1.07] 0.71 [0.33Y1.40] *Six women had unknown HIV subtype. R indicates intersubtype recombinant. †Subtype C includes 7 women from HIVNET 012 and 445 women from NVAZ. ‡Number of infected infants/number of infants available at each visit (% infected) [95% CI]. J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 Letters to the Editor * 2006 Lippincott Williams & Wilkins 519 Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. at birth but infected at 6 to 8 weeks (OR, 2.49; 95% CI, 1.53Y4.04;P= 0.0002), and for infants infected at 6 to 8 weeks (OR, 2.01; 95% CI, 1.46Y2.78;P G 0.0001), but was not associated with age, parity, or HIV-1 subtype. DISCUSSION This report extends our analysis of MTCT and NVP resistance in women enrolled in the HIVNET 012 and NVAZ trials. In a previous report, we demon- strated that women with subtype C HIV-1 were more likely to have NVP- resistant HIV-1 variants detected 6 to 8 weeks after SD NVP than women with either subtype A or D.4 In this report, we find no significant difference in the risk of MTCT among women with these 3 subtypes. This provides empiric evi- dence that selection of NVP-resistant HIV-1 variants in women after SD NVP dosing does not significantly increase the risk of MTCT during the first few weeks of life. However, even in this large study of 715 women, a difference in MTCT risk may not be apparent because of low transmission rates in the setting of SD NVP prophylaxis. Based on the CIs for the ORs comparing MTCT in the 3 subtypes, we cannot exclude the possibility that subtypes A and D, which have less resistance, also have as much as a 50% lower risk of MTCT than subtype C. Other smaller studies have exam- ined the association of env, gag, or LTR subtypes with MTCT in the absence of antiretroviral prophylaxis. Among 28 women from Tanzania, there was no significant difference in MTCT risk among women with env and LTR subtypes A, C, and D or intersubtype recombinant HIV-1.8 In contrast, anoth- er study of 51 matched transmitting and nontransmitting women from Tanzania found a higher risk of MTCT in women with gag and env regions that were either both subtype C or recombinant (different combinations of subtypes A, C, and D), compared with women with gag and env regions that were both subtype D.9 Among 253 Tanzanian infants, a higher rate of in utero HIV-1 infection was observed with env subtype C compared with env subtype A or D.10 In another study of 414 Kenyan women, there was a higher risk of MTCT for women with gag and env regions that were either both subtype D or recombi- nant (D/A or A/D) compared with women with gag and env regions that were both subtype A.11 In this study, HIV-1 subtype was determined using HIV-1 pol (protease and reverse transcriptase) sequences. This region of the HIV-1 genome is the most relevant to use of antiretroviral drugs for prevention of MTCT. We recently observed an association be- tween HIV-1 pol region replication capacity and MTCT among women with subtype C,12 suggesting that pol region determinants may influence transmission risk. Our inability to detect a difference in the risk of MTCT among women with pol region subtypes A, C, and D who received SD NVP prophy- laxis may be influenced by several factors. First, the power to see differ- ences in transmission among different subtypes may be reduced because the number of transmission events is re- duced by NVP prophylaxis. Second, the SD NVP regimen could also lower MTCT to a greater degree in women with certain subtypes (eg, reflecting subtype-based differences in drug sus- ceptibility). Finally, in countries such as Uganda where multiple HIV-1 subtypes cocirculate, the subtype of different regions of the viral genome may be different because of intersubtype recom- bination.13 Therefore, it is possible that associations between HIV-1 subtype and MTCT risk might be observed in women in HIVNET 012 and NVAZ, if subtyp- ing were based on a different region of the HIV-1 genome. Further studies are needed to identify HIV-1 determinants for MTCT and to examine the associa- tion of HIV-1 subtype and MTCT in the presence and absence of antiretroviral prophylaxis. ACKNOWLEDGMENTS The authors thank staff in Uganda, Malawi, and at the Johns Hopkins Uni- versity for their assistance with sample processing. They also acknowledge the as- sistance of clinical trials specialistMelissa Allen (Family Health International). REFERENCES 1. Guay LA, Musoke P, Fleming T, et al. Intrapartum and neonatal single-dose nevirapine compared with zidovudine for prevention of mother-to-infant transmission of HIV-1 in Kampala, Uganda: HIVNET-012 randomised trial. Lancet. 1999;354:795Y802. 2. Jackson JB, Musoke P, Fleming T, et al. Intrapartum and neonatal single-dose nevirapine compared with zidovudine for prevention of mother-to-child transmission of HIV-1 in Kampala, Uganda: 18 months follow-up of the HIVNET 012 randomised trial. Lancet. 2003;362:859Y868. 3. Eshleman SH, Mracna M, Guay LA, et al. Selection and fading of resistance mutations in women and infants receiving nevirapine to prevent HIV-1 vertical transmission (HIVNET 012). AIDS. 2001;15:1951Y1957. 4. Eshleman SH, Hoover DR, Chen S, et al. Nevirapine (NVP) resistance in women with subtype C, compared to subtypes A and D, after administration of single-dose NVP. J Infect Dis. 2005;192:30Y36. Susan H. Eshleman, MD, PhD* Jessica D.Church, BS* Shu Chen, MS† Laura A. Guay, MD* Anthony Mwatha, MS‡ Susan A. Fiscus, PhD§ Francis Mmiro, MBChB|| Philippa Musoke, MBChB¶ NewtonKumwenda,MSc,MPH, PhD† J. Brooks Jackson, MBA, MD* Taha E. Taha, MD, PhD† Donald R. Hoover, MPH, PhD** *Department of Pathology, The Johns Hopkins Medical Institutions, †Department of Epidemiology, The Johns Hopkins Bloomberg School of Public Health, Baltimore, MD; ‡Statistical Center for HIV/AIDS Research and Prevention, Fred Hutchinson Cancer Research Center, Seattle, WA; §Department of Microbiology and Immunology, University of North Carolina at Chapel Hill, Chapel Hill, NC; ||Departments of Obstetrics and Gynaecology; and ¶Department of Paediatrics, Makerere University, Kampala, Uganda; and **Department of Statistics and Institute for Health, Health Care Policy and Aging Research, Rutgers University, Piscataway, NJ seshlem@jhmi.edu Letters to the Editor J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 520 * 2006 Lippincott Williams & Wilkins Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. 5. Eshleman SH, Guay LA, Mwatha A, et al. Comparison of mother-to-child transmission rates in Ugandan women with subtype A vs D HIV-1 who received single dose NVP prophylaxis: HIVNET 012. J Acquir Immune Defic Syndr. 2005;39:593Y597. 6. Taha TE, Kumwenda NI, Hoover DR, et al. Nevirapine and zidovudine at birth to reduce perinatal transmission of HIV in an African setting: a randomized controlled trial. JAMA. 2004;292:202Y209. 7. Taha TE, Kumwenda NI, Gibbons A, et al. Short postexposure prophylaxis in newborn babies to reduce mother-to-child transmission of HIV-1: NVAZ randomised clinical trial. Lancet. 2003;362:1171Y1177. 8. Tapia N, Franco S, Puig-Basagoiti F, et al. Influence of human immunodeficiency virus type 1 subtype on mother-to-child transmission. J Gen Virol. 2003;84:607Y613. 9. Renjifo B, Fawzi W, Mwakagile D, et al. Differences in perinatal transmission among human immunodeficiency virus type 1 genotypes. J Hum Virol. 2001; 4:16Y25. 10. Renjifo B, Gilbert P, Chaplin B, et al. Preferential in-utero transmission of HIV-1 subtype C as compared to HIV-1 subtype A or D. AIDS. 2004;18:1629Y1636. 11. Yang C, Li M, Newman RD, et al. Genetic diversity of HIV-1 in western Kenya: subtype-specific differences in mother-to-child transmission. AIDS. 2003; 17:1667Y1674. 12. Eshleman SH, Lie Y, Hoover DR, et al. Association of HIV-1 replication capacity with HIV-1 mother-to-child transmission among antiretroviral drug naive Malawian women. J Infect Dis. 2006. In press. 13. Eshleman SH, Gonzales MJ, Becker-Pergola G, et al. Identification of Ugandan HIV-1 variants with unique patterns of recombination in pol involving subtypes A and D. AIDS Res Hum Retroviruses. 2002;18: 507Y511. ERRATA In the article by Spacek et al. appearing in the Journal of Acquired Immune Deficiency Syndromes, Vol. 41, No. 5, pp. 607Y610, entitled BEvaluation of a Low-Cost Method, the Guava EasyCD4 Assay, to Enumerate CD4-Positive Lymphocyte Counts in HIV-Infected Patients in the United States and Uganda,[ the list of authors was incorrect. It should have included Moses Kamya, MBChB, MMed, MPH (affiliated with Makerere University, Kampala, Uganda), as the sixth author in the list, to read: BLisa A. Spacek, MD, PhD, Hasan M. Shihab, MBcHB, Fred Lutwama, MBcHB, Jean Summerton, BA, Harriet Mayanja, MD, Moses Kamya, MBChB, MMed, MPH, Allan Ronald, MD, Joseph B. Margolick, MD, PhD, Tricia L. Nilles, BS, and Thomas C. Quinn, MD[ BNonnegligible Increasing Temporal Trends in Unprotected Anal Intercourse Among Men Who Have Sexual Relations With Other Men in Montreal[ by Clemon George, PhD, Michel Alary, MD, PhD, Joanne Otis, PhD, Eric Demers, MSc, Benoı̂t Mâsse, PhD, René Lavoie, MSc, Robert S. Remis, MPH, MD, Bruno Turmel, MD, Jean Vincelette, MD, Raymond Parent, MSc, Roger LeClerc, and the Omega Study Group, Omega Cohort, Montreal, Québec, Canada, which appeared in the Journal of Acquired Immune Deficiency Syndromes, volume 42, number 2, pages 207Y212, is a duplication of the article that appeared in the Journal of Acquired Immune Deficiency Syndromes, volume 41, number 3, pages 365Y370. When citing, please use the volume 41, number 3 reference. We apologize to the authors of the article and the readers of JAIDS for this error. J Acquir Immune Defic Syndr & Volume 42, Number 4, August 1, 2006 Letters to the Editor * 2006 Lippincott Williams & Wilkins 521 Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.