Kaposi's sarcoma (KS) is a life-threatening multifocal neoplasm that is frequent in HIV-infected individuals . It primarily involves the skin and mucous membranes, although the lungs, lymph nodes, and gastrointestinal tract can also be affected . A recently discovered virus, human herpes virus 8 (HHV-8), is likely to be the causative agent of KS . In patients with early and non-progressive KS, no therapy or local destructive treatments are appropriate. Systemic therapies for more aggressive forms of KS are interferon (IFN)-α and various cytotoxic drug regimens . In patients with CD4 cell counts ≥ 200 × 106/l, IFN-α induces a partial response in 10–50% of patients, with a few complete responses. Single- or multiple-agent cytotoxic chemotherapy induces partial response in 26–88% of patients treated, with few complete or long-lasting responses. Even with the most active regimens, such as adriamycin-bleomycin-vincristine (ABV), liposomal anthracyclins or paclitaxel, the relapse of KS usually occurs a few months after treatment initiation. In addition, all of these drugs induce dose-limiting toxicities. Hence, therapy of AIDS–KS remains largely unsatisfactory.
The incidence of new cases of KS , as well as most AIDS-defining opportunistic infections , has declined dramatically since 1996, coinciding with the increasing use of potent antiretroviral agents, such as HIV-1 protease inhibitors (PI) in combination with two nucleoside analogues [also known as highly active antiretroviral therapy (HAART)]. HAART can be expected to impair HIV-1 replication dramatically and to raise the CD4 cell count by a mean of 150 × 106/l over 12 months . The efficacy of HAART in patients with ongoing KS is documented less well; a few isolated case reports, as well as short series of patients [7–11], suggest that it may induce regression of KS in some patients. These series, however, include few patients, and most publications are retrospective and/or have a patient follow-up of < 1 year. In some publications, no validated criteria for response were used. Furthermore, the exact role of HAART in KS regression is not well established because most patients were simultaneously receiving systemic therapy for KS. Finally, worsening of KS after introduction of HAART has also been reported in two patients [12,13].
To investigate the role of HAART in the treatment of established KS, we prospectively constituted a large cohort of AIDS–KS patients initiating HAART and followed them for 24 months. Because many patients were not simultaneously receiving any systemic KS treatment, we were able to investigate the role of HAART itself. As a large number of patients were enrolled, we were able to perform a multivariate analysis to identify predictors of complete regression of KS after 2 years on HAART.
Materials and methods
HIV-seropositive males and females, 18–70 years old, with biopsy-proven KS lesions of any stage of the TIS classification of AIDS Clinical Trials Group (ACTG) , could be included in two overlapping cohorts of patients. They had to have active, measurable KS lesions at inclusion and to be starting HAART including at least one PI and two nucleosides analogues. Main criteria for exclusion were ongoing opportunistic infection or lymphoma, Karnovsky index < 70, and previous treatment with, or contraindication to, an HIV1-protease inhibitor drug.
The first cohort consisted of the 27 patients with KS included in the CISIH 92 (centre d'information et de soins sur l'immunodéficience humaine) cohort and 263 patients starting HAART including at least one PI between March 1996 (when PI became available for prescription in AIDS patients in France) and December 1996: the clinicians of the CISIH 92 – a government-founded association of all clinicians caring for HIV-infected patients in the hospitals of the West Paris metropolitan area (two million inhabitants) – entered all of their 263 patients in a prospective cohort; 12 of these patients were also referred for evaluation and treatment of KS to the A Paré Hospital Department of Dermatology, which belongs to the CISIH 92.
The second cohort included 12 AIDS–KS patients referred during the same period to this Department of Dermatology from hospitals outside the CISIH 92 organization. These 24 `dermatology' patients (from either the CISIH 92 or the Department of Dermatology cohort) originated from a series of 40 AIDS–KS patients followed at that time in this department (main causes for non-inclusion were absence of PI, and, for four patients, impossibility of long-term follow-up). All patients gave written informed consent and were treated as outpatients.
When initiating HAART, the choice of PI and nucleoside analogues was free, but the use of previously unused nucleoside analogues was recommended. HAART had to remain unmodified during the course of the study unless drug intolerance occurred, CD4 cell counts failed to improve, or plasma HIV RNA increased. Therapy of KS was also freely chosen, but a recommendation to delay initiation of cytotoxic chemotherapy and to use, if required, local treatments, was made for patients with slowly progressing non-life-threatening T0 KS. Treatments of KS were classified as local (cryosurgery, radiotherapy), minor chemotherapy (bleomycin, vinblastin), major chemotherapy (ABV, liposomal anthracyclin or paclitaxel) and miscellaneous (IFN-α, oral all-trans-retinoic acid ). A progressive tapering of KS treatment was suggested for responding KS patients. Neither patients nor clinicians were blinded to the antiretroviral or KS therapies.
KS was evaluated at inclusion and at least every 6 months thereafter by estimating the total number of KS lesions and by monitoring representative lesions. The 24 `dermatology' patients were evaluated by trained clinicians involved in clinical trials on AIDS–KS. KS in the remaining 15 patients was evaluated by the CISIH 92 clinicians. Changes in and side-effects of HAART and all medical events were recorded at each evaluation. Chest radiography was performed at inclusion and at least every 6 months for patients with abnormal initial findings or with non-responding KS. For pulmonary KS, computed tomography scans were performed to assess response. The effect of HAART on HIV-infection was studied by monitoring the number of CD4 lymphocytes and plasma HIV-1 RNA by branched-DNA signal amplification-based hybridization (Quantiplex HIV-RNA assay kit; Chiron Corp, Emeryville, California, USA; lower limit of detection 500 copies/ml) or, for four patients, by reverse transcriptase (RT)–PCR (Roche Amplicor; Neuilly, France; lower limit of detection 20 copies/ml). Because such measurements became available routinely only during the summer of 1996, baseline HIV-1 RNA plasma levels were not assayed in the first patients enrolled.
KS response to therapy was classified as complete response, partial response, progression, or stabilization according to slightly modified ACTG recommendations [14,15]. Absence of any residual KS lesion on clinical examination was required for clinical complete response , with skin biopsy required only in doubtful cases. Partial response was defined as no new lesion and one of the following: > 50% decrease in the total number of lesions and/or in the size of the index lesions; marked flattening of KS lesions; clinical complete response but with residual oedema. Progression was defined as an increase of > 25% in the size and/or number of KS lesions, the appearance of new sites of disease and/or change of > 25% in the skin or oral lesions from macular to plaque-like or from plaque to nodular lesions. Stabilization was defined as any response not meeting the criteria for complete response, partial response or progression. For the 12 patients belonging to both cohorts, 40 evaluations previously made by the CISIH 92 clinicians were duplicated by the Department of Dermatology, and 90% of them gave identical results, suggesting a similar mode of evaluation in the two cohorts. The four discordant pairs of evaluations were slightly less optimistic (i.e. partial response instead of complete response) and were performed by the Department of Dermatology. The latter evaluations were used for this study. Modifications of KS treatment were also assessed.
Descriptive data were expressed as mean ± SD unless otherwise stated (median). Between-group differences in continuous variables were examined using Student's t test or the Wilcoxon matched pairs non-parametric test. Associations between nominal variables were tested by the chi-square test with Yates's correction or by Fisher's exact test when necessary. CD4 cell count increases during HAART were analysed using a multivariate analysis of variance. Factors predictive of complete response after 2 years on HAART were assessed among several variables: prior KS treatment at inclusion, T of the TIS staging of KS, CD4 cell counts at inclusion and month 12, plasma HIV RNA < 500 copies/ml at month 12, increase in the CD4 cell count from inclusion to month 12 (ΔCD40−−12) > 150 × 106/l. Firstly, association with complete response was tested with the univariate tests previously described. Secondly, variables found to be significantly associated with complete response (P < 0.2) in the univariate analysis were entered into a logistic regression model, which allows simultaneous checking of multiple factors. A stepwise procedure was used, as well as tests of interactions between significant risk factors in the multivariate model. P values < 0.05 were considered significant in multivariate analyses. The adjusted odds ratio (OR) of complete response and 95% confidence interval (CI) were calculated for all independent significant predictors of complete response. Statistical analysis were performed with the SAS 6.2 software package (SAS Institute Inc, Cary North Carolina, USA).
The characteristics of the 39 patients at inclusion are listed in Table 1. Fifteen patients had previously experienced another AIDS-defining illness. All tested patients had high plasma HIV-RNA levels. Indinavir was generally the PI initially prescribed. The associated nucleoside analogues were usually zidovudine, stavudine, and lamivudine. Only five patients were naive to antiretroviral therapy with nucleoside analogue(s). All but five patients had continuous prophylaxis against Pneumocystis carinii pneumonia with cotrimoxazole or aerosolized pentamidine. Two patients were receiving secondary prophylaxis for cytomegalovirus retinitis with ganciclovir. The severity of KS ranged from the more benign T0I0S0 form to the most severe T1I1S1 form. Eight patients with pulmonary KS and 10 with oedematous skin or non-plane mucous membrane lesions had T1 KS. Previous KS treatments before HAART were at least one systemic therapy in 19 patients and local therapy only in six. At inclusion, 19 patients were not receiving any general KS treatment (Table 2). There were no statistically significant differences between the CISIH 92 and Department of Dermatology cohorts in terms of age, TIS classification of KS at inclusion, type and number of previous KS therapies, CD4 cell count and plasma HIV RNA at inclusion, type of HAART used, or length of follow-up.
At the closing date of 31 December 1998, median duration of follow-up was 24 months (range, 5–28 months). Three patients died during follow-up, one from progressive pulmonary KS after 6 months on HAART despite major therapy with ABV, and two from unrelated causes (ongoing cardiac failure and biopsy-proven lymphoma diagnosed during the fourth month on HAART) after 5 and 15 months on HAART, respectively. Of these two, one was rated as partial response and the other as stabilized. One patient was lost to follow-up after 12 months.
Response of HIV infection to HAART
CD4 cell counts increased significantly from 85 ± 110 × 106/l at inclusion to 198 ± 156 × 106/l, 258 ± 203 × 106/l, 298 ± 263 × 106/l and 316 ± 296 × 106/l at months 6, 12, 18 and 24, respectively (P < 0.0001). The percentage of patients with plasma HIV-RNA < 500 copies/ml was 43%, 49%, 46%, and 44% at months 6, 12, 18, and 24, respectively. Only three new AIDS-defining events occurred during the study. Only one patient had long-lasting therapy with foscarnet initiated during follow-up. At month 12, HAART had been modified in 15 patients for persistently elevated plasma HIV RNA. Those modifications induced no sustained effect on plasma HIV-RNA, expect in one patient. CD4 cell counts improved markedly in three of these patients. PI was replaced because of intolerance in five additional patients.
Outcome of KS
Progressive improvement in KS was observed (Table 3), with complete response or partial response in 49%, 74%, 77%, and 74% of patients at months 6, 12, 18, and 24, respectively. No new pulmonary KS developed. At month 24, the 29 responding patients (complete response + partial response), when compared with non-responding patients, had significantly higher CD4 cell counts (360 ± 300 versus 113 ± 185 copies/ml;P < 0.05), greater increases in ΔCD40−−24 (270 ± 265 versus 26 ± 54 × 106/l;P < 10−4), and HIV-1 RNA plasma levels more frequently < 500 copies/ml (57% versus 0%;P < 0.05). At month 24, the 18 patients achieving complete response had significantly higher CD4 cell counts (465 ± 343 versus 185 ± 167 × 106/l;P < 0.01) and ΔCD40−−24 (386 ± 295 versus 85 ± 89 × 106/l;P < 0.0001) than patients with partial response, stabilization or progression, but the proportions of patients with HIV-1 RNA < 500 copies/ml did not differ significantly between the two groups.
No recurrence of KS was observed during subsequent follow-up once complete response had been achieved. Only five of the 30 patients rated as showing partial response at any time-point during follow-up were later rated as progressive. When rated as progressive all of these patients had plasma HIV-1 RNA > 2 × 104 copies/ml; three of them had CD4 cell counts < 10 × 106/l, and two > 200 copies/ml. A final rating of complete response, partial response or stabilization, respectively, was made in five, nine, and one of the 20 patients rated progressive at any time-point.
The effect of HAART itself on KS outcome (without the confounding factor of systemic KS therapy) was assessed in the 19 patients without any systemic KS treatment at inclusion (Table 4). Five had T1 KS, and 14 T0 KS. Complete response or partial response was achieved in 64%, 69%, 79%, and 74% of included patients at months 6, 12, 18, and 24, respectively. A progressive increase in complete response rate was observed, with 11%, 32%, 42%, and 58% of included patients achieving complete response at the same respective time-points. Systemic KS treatment was introduced in only four patients with progressive KS during follow-up. All complete response and all but one partial response were achieved without the introduction of any new local or general KS treatment.
Most of the 20 patients (13 with T1 KS, and seven with T0 KS) on minor (n = 13) or major (n = 7) chemotherapy at inclusion experienced improvement of KS, with 35%, 70%, 75%, and 75% achieving complete response or partial response at months 6, 12, 18, and 24, respectively. Complete response rate progressively increased, with 0%, 15%, 30%, and 35% of patients achieving complete response at the same respective time-points. All patients with complete response and five out of eight patients with partial response had no general KS treatment at month 24. Response of KS to HAART was also assessed by a progressive tapering of KS treatment (Table 2). Minor chemotherapy was, however, upgraded to major in four and two patients at months 6 and 12, respectively.
Progress was initially mainly favourable in the eight patients with pulmonary KS receiving HAART and (in all but one patient) cytotoxic chemotherapy, with one progressive and seven partial responders at month 12. Some of these partial responders later relapsed, and complete response, partial response, stabilization and progression were observed at month 24 in two, two, none and three patients, respectively. Three of the four pulmonary KS patients who responded were no longer receiving chemotherapy at month 24.
Predictors of complete response at month 24
Logistic regression revealed only two factors predictive of achieving complete response at month 24: ΔCD40−−12 > 150 × 106/l (OR, 13.4; 95% CI, 2–82) and T0 of TIS classification at inclusion: (OR, 7; 95% CI, 1.1–42). No interaction was found between these factors. ΔCD40−−12 > 150 × 106/l was still a predictor when logistic regression was performed only on patients receiving systemic chemotherapy at inclusion.
This prospective series of AIDS–KS patients treated with HAART – to the best of our knowledge the largest ever published – clearly demonstrated that KS responded to HAART in most patients. Unlike systemic treatments of KS , HAART resulted in complete response in 46% of the patients. Predictors of complete response after 2 years on HAART were: T0 KS at inclusion and CD4 cell count increase > 150 × 106/l in the first year on HAART.
Our results agree with those of previous smaller cohorts of AIDS–KS patients treated with HAART. Among the 41 patients included in the four published series that included more than eight patients [7,9–11], 32%, 37%, 17% and 12% achieved complete response, partial response, stabilization, and progression, respectively. We also confirm, with a larger number of patients, that systemic therapies of KS can be interrupted in responding patients , even in those with pulmonary KS, where median survival after diagnosis was 4–10 months before the era of HAART.
Our study yields further information. Response to therapy was slow but long-standing. No recurrence was observed after complete response had been obtained. Once partial response had been achieved, KS recurred in only 20% of patients. Because nearly half of our patients were not receiving any systemic KS treatment, we were able to show that HAART itself can induce complete response in > 50% of patients. Logistic regression identified two predictors of complete response after 2 years on HAART. Although our study included only 39 patients, the OR was high for both predictors, which may be useful in daily clinical practice. Some predictors of poor prognosis in KS patients pre-dating the era of HAART, such as low CD4 cell count at inclusion  and failure of previous systemic KS treatment, had no prognostic value in our series. Because few enrolled patients were naive to previous antiretroviral treatment, we were unable to demonstrate any predictive value for KS response of this parameter, which usually predicts better efficacy of HAART on the inhibition of HIV replication.
Because only three out of 39 patients received long-term ganciclovir or foscarnet, we believe that these drugs, which inhibit HHV-8 DNA synthesis in in vitro systems of lytic infection , did not explain our results. Moreover, in vivo, they did not affect the HHV-8 DNA load within the leukocytes in two independent studies [19,20] and, in vitro, did not inhibit episomal HHV-8 DNA synthesis. Consequently, no benefit can be expected from such drugs during latency or in KS . We did not investigate HHV-8 DNA in the leukocytes of our patients with HAART because its predictive value was uncertain when we planned our study. It has been reported positive in only half of untreated KS patients , with unexplained fluctuations over time. Some [7,9], but not all , small series of KS patients treated with HAART have later suggested that response of KS may be associated with a reversal in HHV-8 viraemia.
We have designed our cohort study to lessen the usual flaws of such a design. All patients entered the cohort when they first received HAART. To avoid bias due to referral patterns, we entered eligible patients from all major hospitals caring for HIV-infected patients within the West Metropolitan Paris area. It is likely that almost all KS patients from these hospitals who initiated HAART during the study period entered the cohort because HAART was at that time delivered only by hospital pharmacists and with the written permission of the CISIH 92 members. Follow-up was complete for all but one patient and validated criteria for response assessments were used. All previously established prognostic factors of KS were taken into account in the statistical analysis. However, our conclusions do not emanate from a carefully designed placebo-controlled randomized study. Thus, proof of efficacy of HAART on KS has not been formally obtained, and probably never will be: performing such studies is not ethical because of the obvious benefits of HAART for other manifestations of HIV infection.
Several factors may explain improvement in KS during HAART. Direct action of PI on HHV-8 replication seems unlikely . The reduction in HIV Tat production due to impaired HIV replication with HAART is probably not the main mechanism involved. Tat stimulates the growth and invasion of spindle cells derived from KS lesions . The percentages of patients with HIV-1 RNA levels < 500 copies/ml did not differ significantly between patients achieving complete response and others. We feel that our results are explained largely by the recovery of the immune system during HAART, a continuous, at best partial, process which continues long after inhibition of HIV replication has been achieved . The early first-phase cellular restoration appears to be primarily a redistribution of cells originally present in lymphoid tissues. The second-phase increase seems to reflect maturation of newly generated T cells, as evidenced by diversification of the CD4 cell repertoire: it is characterized by a shallower increase in circulating naive CD4 and CD8 cells and a decrease in circulating memory CD8 cells . The slow improvement in KS during HAART temporarily paralleled the immune restoration, and when HAART substantially raised CD4 counts within the first year, this was predictive of later complete response. Such a threshold effect in the restoration of CD4 counts suggests that control of KS is linked to immune restoration to a level at which it is able to control HHV-8 infection. This is a reminder of what is observed in transplanted patients, in whom KS frequently disappears on reduction of immunosuppressive treatments .
In patients with AIDS–KS, the efficacy of HAART on KS surpassed that obtained previously with systemic KS treatments. Our study has important practical implications for clinicians. To avoid delaying the restoration of immune system with cytotoxic chemotherapy, we suggest avoiding its use unless required by aggressive forms of KS. Instead, conservative strategies (local treatments, well-tolerated systemic therapies such as oral all-trans-retinoic acid , or even camouflage) should be encouraged until KS responds to HAART, and patients should be reassured using our predictors of response. On the other hand, clinicians should now be highly cautious when analysing therapeutic trials of KS treatments. The quality of the antiretroviral treatments used may be a major confounding factor. Further work is requested, such as study of newer antiretroviral strategies with combinations of nucleoside analogues and non-nucleoside inhibitors of HIV-reverse transcriptase.
The authors thank C. Michon, (CISIH 92 medical co-ordinator, Hôpital L Mourier, Colombes) for evaluating the patients included in the trial. We also thank D. Marsh for editing our English.
1. Brodt HR, Kamps BS, Helm EB, Schofer H, Mitrou P. Kaposi's sarcoma in HIV infection: impact on opportunistic infections and survival
. AIDS 1998, 12: 1475 –1481.
2. Schulz TF, Boshoff CH, Weiss RA. HIV infection and neoplasia
. Lancet 1996, 348: 587 –591.
3. Murphy PM. Pirated genes in Kaposi's sarcoma
. Nature 1997, 385: 296 –299.
4. Yarchoan R. Therapy for Kaposi's sarcoma: recent advances and experimental approaches
. J Acquir Immune Defic Syndr 1999, 21 (suppl 1): S66 –S73.
5. Jacobson LP, Yamashita TE, Detels R. et al. Impact of potent antiretroviral therapy on the incidence of Kaposi's sarcoma and non-Hodgkin's lymphomas among HIV-1-infected individuals. Multicenter AIDS Cohort Study
. J Acquir Immune Defic Syndr 1999, 21 (suppl 1): S34 –S41.
6. Powderly WG, Landay A, Lederman MM. Recovery of the immune system with antiretroviral therapy.
:The end of opportunism?
JAMA 1998, 280: 72 –77.
7. Dupin N, Rubin de Cervens V, Gorin I. et al. The influence of highly active antiretroviral therapy on AIDS-associated Kaposi's sarcoma
. Br J Dermatol 1999, 140: 875 –881.
8. De Milito A, Catucci M, Venturi G. et al. Antiretroviral therapy with protease inhibitors in human immunodeficiency virus type 1- and human herpesvirus 8-coinfected patients
. J Med Virol 1999, 57: 140 –144.
9. Lebbé C, Blum L, Pellet C. et al. Clinical and biological impact of antiretroviral therapy with protease inhibitors on HIV-related Kaposi's sarcoma
. AIDS 1998, 12: F45 –F49.
10. Krischer J, Rutschmann O, Hirschel B, Vollenweider-Roten S, Saurat JH, Pechere M. Regression of Kaposi's sarcoma during therapy with HIV-1 protease inhibitors: a prospective pilot study
. J Am Acad Dermatol 1998, 38: 594 –598.
11. Tavio M, Nasti G, Spina M, Errante D, Vaccher E, Tirelli U. Highly active antiretroviral therapy in HIV-related Kaposi's sarcoma
. Ann Oncol 1998, 9: 923. 923.
12. Harindra V, Foley E. Highly active antiretroviral therapy
. Lancet 1998, 351: 1058 –1059.
13. Weir A, Wansbrough-Jones M. Mucosal Kaposi's sarcoma following protease inhibitor therapy in an HIV- infected patient
. AIDS 1997, 11: 1895 –1896.
14. Krown SE, Metroka C, Wernz JC. Kaposi's sarcoma in the acquired immune deficiency syndrome: proposal for uniform evaluation, response, and staging criteria
. J Clin Oncol 1989, 7: 1201 –1207.
15. Saiag P, Pavlovic M, Clerici T. et al. Treatment of early AIDS-related Kaposi's sarcoma with oral all-trans- retinoic acid: results of a sequential non-randomized phase II trial. Kaposi's Sarcoma ANRS Study Group. Agence Nationale de Recherches sur le SIDA
. AIDS 1998, 12: 2169 –2176.
16. Little RF, Pluda JM, Feigal E, Yarchoan R. The challenge of designing clinical trials for AIDS-related Kaposi's sarcoma
. Oncology (Huntingt) 1998, 12: 871 –884.
17. Krown SE, Testa MA, Huang J. AIDS-related Kaposi's sarcoma: prospective validation of the AIDS Clinical Trials Group staging classification. AIDS Clinical Trials Group Oncology Committee
. J Clin Oncol 1997, 15: 3085 –3092.
18. Medveczky MM, Horvath E, Lund T, Medveczky PG. In vitro antiviral drug sensitivity of the Kaposi's sarcoma-associated herpesvirus
. AIDS 1997, 11: 1327 –1332.
19. Boivin G, Gaudreau A, Toma E. et al. Human herpesvirus 8 DNA load in leukocytes of human immunodeficiency virus-infected subjects: correlation with the presence of Kaposi's sarcoma and response to anticytomegalovirus therapy
. Antimicrob Agents Chemother 1999, 43: 377 –380.
20. Humphrey RW, O'Brien TR, Newcomb FM. et al. Kaposi's sarcoma (KS)-associated herpesvirus-like DNA sequences in peripheral blood mononuclear cells: association with KS and persistence in patients receiving anti-herpesvirus drugs
. Blood 1996, 88: 297 –301.
21. Whitby D, Howard MR, Tenant-Flowers M. et al. Detection of Kaposi sarcoma associated herpesvirus in peripheral blood of HIV-infected individuals and progression to Kaposi's sarcoma
. Lancet 1995, 346: 799 –802.
22. Kedes DH, Ganem D. Sensitivity of Kaposi's sarcoma-associated herpesvirus replication to antiviral drugs. Implications for potential therapy
. J Clin Invest 1997, 99: 2082 –2086.
23. Fiorelli V, Barillari G, Toschi E. et al. IFN-gamma induces endothelial cells to proliferate and to invade the extracellular matrix in response to the HIV-1 Tat protein: implications for AIDS-Kaposi's sarcoma pathogenesis
. J Immunol 1999, 162: 1165 –1170.
24. Penn I. Kaposi's sarcoma in transplant recipients
. Transplantation 1997, 64: 669 –673.
© 2000 Lippincott Williams & Wilkins, Inc.