Progressive vaccinia was confirmed [14
] on 3 March, and a request for release of vaccinia immune globulin intravenous (human) (VIGIV) from the Department of Defense and the Strategic National Stockpile (SNS) at the CDC was submitted along with an EIND application for the use of oral and topical ST-246 [9
]. Infusions of VIGIV from the SNS began on 4 March, while ST-246 oral and topical treatment began on 5 and 6 March, respectively. On 7 March, the patient became septic, required stress-dose corticosteroids, and was transferred to the intensive care unit. ST-246 was withheld for 24 hours; Pseudomonas aeruginosa
bacteremia was identified as the cause of his deteriorating condition. Subsequent ST-246 doses were periodically increased following identification of subtarget plasma drug concentrations (target concentrations determined by efficacy studies in nonhuman primates), compared with those seen in healthy volunteers. Further VIGIV was administered at 6000 IU/kg.
Initially, the vaccination site began to respond to treatment, with drying, flattening, and cessation of enlargement, until 19 March, when satellite lesions were noted (Figure C
). During this time, corticosteroid therapy was being slowly tapered (24 March was the final day), and periodic granulocyte colony-stimulating factor was administered. Additional VIGIV was administered to achieve a stable IgG level (Figure ). The dosing of ST-246 was escalated to achieve target plasma concentrations, and an EIND was filed for an additional antiviral, CMX001, a prodrug of cidofovir. The patient received the first dose of CMX001 (200 mg) on 26 March. High-dose VIGIV (24 000 IU/kg) was administered in an attempt to increase the level of immunoglobulin at the site of vaccinia infection [9
Figure 2. Anti-orthopoxvirus immunoglobulin G (IgG) and immunoglobulin M (IgM) levels. Arrows correspond to administration of vaccinia immune globulin (human) (VIGIV). VIGIV contains only gamma globulin (IgG) and therefore has no effect on increasing IgM levels. (more ...)
The patient was intermittently neutropenic until 21 April and received granulocyte colony-stimulating factor periodically to maintain an absolute neutrophil count of 1000 neutrophils/mm3. Additionally, he was noted to be lymphopenic, with absolute lymphocyte counts of <200 cells/m3 until 1 April, then stabilizing at >500 cells/mm3 by 21 April (Figure A–C). On 10 April, the vaccine site, which had been gradually improving, had a raised border once more; bacterial cultures revealed methicillin-resistant Staphylococcus aureus supporting superinfection of his vaccination site, rather than virologic progression. Treatment with intravenous vancomycin was initiated, and the vaccine site resumed healing. The eschars were manually debrided on 21 April with minimal bleeding, and healing continued steadily (Figure D). PCR failed to detect vaccinia virus DNA from swabs of the vaccine site collected on 27 April. A single-punch biopsy was performed to look for evidence of virus at the leading edge of the apparently regressing lesion. Immunohistochemical staining and PCR were both negative for orthopoxvirus. The treatment regimen was successfully de-escalated by discontinuing one agent at a time while observing the patient for any clinical indication of relapse. His entire treatment course consisted of 341 vials of VIGIV, 73 days of oral ST-246 (nearly 75 g), 68 days of topical ST-246, and 6 weekly doses of CMX001 (totaling 700 mg) (Figure ).
Figure 3. White blood cell markers, viable virus and viral DNA load detected from the vaccine site, and viral DNA load in EDTA blood. A, White blood cell count. B, Absolute neutrophil count. C, Absolute lymphocyte count. D, Plaque-forming units per swab of the (more ...)
Figure 4. Time line of the patient's clinical course. Important landmarks in the progression of disease are found along the top portion of the time line. Medications used to treat progressive vaccinia and declining white blood cell counts can be found directly (more ...)
PCR evidence of virus levels above the limit of detection was documented in ethylenediaminetetraacetic acid–exposed blood samples from 4–8 and 22–26 March. The protocol of using DNase I pretreatment to measure the protected DNA (whole virus) versus unprotected DNA (extra-viral DNA) was used on blood samples. All samples treated with DNase prior to DNA extraction were negative by PCR (Figure F), with the exception of a 22 March sample (concomitant with the appearance of satellite lesions), consistent with free viral DNA (rather than encapsulated viral DNA) in the blood. In addition, oropharyngeal samples (from 5 March) and rectal swab samples (from 6 March) were negative for orthopoxvirus DNA. On 2 April, the lesion was free of viable virus.
No adverse events were clearly attributable to any investigational therapy. The patient experienced emesis and 1 bout of bloody diarrhea within hours after the first dose of CMX001, but he was thrombocytopenic at the time, and he did not experience any emesis or bleeding with subsequent doses.
Flow cytometry–based analysis was done on 2 April, 22 April, and 28 May to analyze circulating lymphocytes. Initial analysis on 2 April showed that overall the absolute numbers of CD4+
T cells, CD8+
T cells, and natural killer (NK) cells were well below the range present in healthy individuals, while CD19+
B cells were completely absent (Table ). This likely reflected the effects of chemotherapy. Analysis also showed that despite lower absolute numbers, both CD4+
T cells were proliferating, as evidenced by expression of the cell cycle marker Ki-67, and expressed high levels of the activation markers HLA-DR and CD38. Together with the decreased expression of the anti-apoptotic protein Bcl-2, these characteristics are indicative of a potent, short-lived effector T-cell response (Table ). Although the frequency of activated cells was comparable to that seen in an effector response in a healthy Dryvax recipient [12
], who developed immunity and was of the same age group, sex, and race as the patient (Figure B
.), we estimate that the absolute number of activated CD8+
T cells (approximately 32 cells/μL of blood) was 3–10-fold lower than that in the healthy Dryvax control. Analysis on 22 April and 28 May showed that absolute cell numbers of T, B, and NK cells had increased, although they were still below the range seen in healthy adults. T-cell proliferation and the activated T-cell phenotype were reduced by 22 April but still much higher than baseline levels seen in any healthy adult (<1% of CD4+
Absolute Cell Counts in Whole-Blood Specimens From the Patient With Progressive Vaccinia and a Healthy Donor Control
Figure 5. A, T-cell response to vaccinia infection. Boxed numbers represent the percentage of cells specific to the vaccinia cytotoxic T-lymphocyte epitope. Effector response is measured approximately 2 weeks after vaccination. During this time, vaccinia virus–specific (more ...)
Because the patient was HLA-A2 positive, we were able to track the CD8 T cells recognizing a vaccinia virus cytotoxic T-lymphocyte epitope, using HLA-A2 restricted VVCLT
]. Initial VVCLT
tetramer frequencies were high (1% of total CD8+
T cells) and decreased (0.2% of total CD8+ T cells) by 28 May. The VVCLT
T cells on 2 April were also highly activated (Ki-67+
, and Granzyme B+
) and lost this phenotype by 28 May. This differentiation was similar to that of the VVCLT
-specific effector and memory CD8+
T cells in the healthy Dryvax vaccinee. In summary, the initial activation was consistent with a vigorous virus-induced T-cell response that seemed to wane as the virus was controlled (Figure A
In addition to the hospital course as delineated above, the patient also developed gangrene as a consequence of pseudomonal sepsis and underwent below-the-knee amputations of both legs on 14 April. He was discharged in stable condition on 28 May. During follow-up, the former site of PV continued to remain quiescent and well healed (Figure E and 1F). On 24 July, he underwent allogeneic stem cell transplantation and to date remains free of vaccinia infection and of AML. His anti-orthopoxvirus IgM and IgG titers were negligible when evaluated 6 months following his allogeneic stem cell transplantation.
The apparent ST-246 EC50
of virus samplings from the vaccination site lesion from 23 March (EC50
= 0.95 µM) and March 31 (EC50
= 3.55 µM) were 13.5-fold and 50-fold, respectively, greater than that seen with the virus isolated prior to institution of ST-246 therapy on 2 March (EC50
= 0.07 µM). Minimal changes in apparent ST-246 EC50
were observed in evaluation of virus samplings from satellite lesions from 23 March (EC50
= 0.06 µM) or 31 March (EC50
= 0.14 µM). No resistance to CMX001 was observed in comparing virus samplings from 23 and 31 March (3 days prior to and 5 days after initiation of CMX001). Ultradeep sequencing at >20 000-fold coverage of a region of the F13 gene (nucleotides 778–944), the only known target of ST-246 (unpublished data, Siga Technologies), and analysis using the GS Amplicon Variant Analyzer were performed. An extremely low level of single-nucleotide polymorphism (SNP) variation, 0.28%–0.43% of the population amplified, was identified in numerous SNPs across that region sequenced. Random SNP variation was present in vaccination site and satellite lesion samples from 2, 29, and 31 March without significant change in frequency of appearance of any of these individual SNPs in these samples (Table ). All changes were SNPs, and these low-frequency-level changes were all transitions. Although a G277C amino acid change resulting from a SNP at nucleotide 831 has been previously reported to be associated with resistance to ST-246 [17
], analysis of the predicted protein sequence in these isolates did not reveal evidence of the G277C mutation. However, 2 specific predicted amino acid changes (from alanine to valine [amino acid 290] and from leucine to methionine [amino acid 315]) were observed in increasing frequency in the population between the material sequenced from the vaccination site on 23 March (3% and 7%, respectively) and 31 March (31% and 10%, respectively). These changes were not observed at levels above the limit of analysis (0.1% of the population) in the predicted protein sequences from the vaccination site samples sequenced from 2 March, nor from the satellite lesions sequenced from 23 or 31 March.
Percentage Nucleotide Variability and Change at a Partial Nucleotide Within the Vaccinia Virus F13L Gene