PEDIATRICS Vol. 107 No. 3 March 2001, pp. 611-612

The Science and Fiction of Pertussis Vaccines

To the Editor.

In the December 1999 issue of Pediatrics, Cherry and Olin criticized the Munich study, in which we investigated the efficacy of the 2-component Biken acellular pertussis vaccine using a nested case-control study design.¹ On the basis of the results of this and previously published studies, the Biken acellular pertussis vaccine has been licensed in 1995 in Germany and in 1996 in the United States as the first acellular pertussis vaccine for use in infants. Cherry and Olin held that the efficacy values were inflated "due to the study design (case-control), the lack of serological diagnosis, the lack of a case definition, which included mild illness and the almost certain observer bias, because there was no prospective surveillance." We regret that Cherry and Olin misread some of the basic facts of the study design, and we feel that the readers of Pediatrics deserve a rebuttal to their errors of interpretation.

1. The Munich study used prospective surveillance with strictly defined diagnostic procedures in all infants presenting with cough 7 days in the pediatrician's office. The possibility of observer bias, as observed by Cherry et al² in another pertussis vaccine trial, was controlled by using a cough log at each physician's practice to verify that each child with 7 days of cough had been cultured. Ninety-eight percent of the children were cultured as requested in the protocol. Diagnostic workup with nasopharyngeal swabs was performed at a median of 9 days after the cough had started in Biken DTacP recipients, similar to a median of 9.5 days in DT recipients. Moreover, 76% of cultures were performed in children receiving Biken DTacP study vaccine, a proportion that corresponds almost exactly to the 74.6% of infants overall who had received the study vaccine. Clearly, cultures were performed both independently of the vaccine status and the cough symptomatic of the child.
2. The use of a case-control design does not per se result in bias. There are numerous examples of efficacy studies using case-control design, which have contributed essentially to progress in vaccine development.³ The Munich study recruited a cohort of 16 780 infants, in whom there was prospective case surveillance and for whom all the relevant study data were recorded before the occurrence of any disease. The case-control element of the design was simply a matter of efficient transfer of recorded data to computerized files.
3. The objective of the Munich case-control study was to investigate the efficacy of the Biken DTacP and a comparative whole cell vaccine against typical pertussis disease with paroxysmal cough of >21 days and against milder pertussis disease with any cough >21 days. Cherry and Olin correctly point out that efficacy depends on the case definition used, and that all currently licensed pertussis vaccines are less efficacious against mild pertussis and pertussis infection than against typical pertussis. We believe, as do most clinicians and epidemiologists, that protection against symptomatic disease is vastly more relevant to child health than is protection against infection in minimally symptomatic patients. Classification of currently licensed acellular pertussis vaccines as "superior" and "inferior" vaccines according to their protectiveness against atypical disease is, we believe, arbitrary and counterproductive. The efficacy estimates of the Munich study are comparable to the results of a placebo-controlled, randomized Swedish trial of 1986 in which the same Biken acP vaccine yielded an efficacy of 80% against culture-confirmed pertussis with any cough of >30 days, and an efficacy of 87% against typical pertussis with paroxysmal cough.^4,5 The validity of the Munich case-control study is further confirmed by the efficacy estimates found for the DTwcP vaccine, which were almost identical to the estimates found for the same vaccine in another independent study done in the same population.⁶

An important factor not mentioned in the commentary is herd immunity, which contributes to effectiveness of acellular pertussis vaccines once they are introduced in the target population. This factor has not been investigated in any of the published efficacy studies. We may expect that an increasing herd immunity due to increasing vaccination coverage decreases circulation of Bordetella pertussis in the population and might be a much more important factor in the prevention of pertussis than the ongoing discussions on relatively small differences in efficacy estimates. In Germany the introduction of combination vaccines containing 2- or 3-component acellular pertussis vaccines has already resulted in a large documented increase of pertussis vaccination coverage from an estimated 20% to 30% in 1991 to 80% to 90% in 1998. We are currently evaluating the long-term efficacy of four doses of Biken acellular pertussis vaccine to define the optimal time interval for further booster doses.

Johannes G. Liese
Bernd H. Belohradsky
University Childrens Hospital
Ludwig-Maximilians-Universität
D-80337 Munich, Germany

Alexander M. Walker
Harvard School of Public Health
Department of Epidemiology
Boston, MA 02115

REFERENCES

1. Liese JG, Meschievitz CK, Harzer E, Efficacy of a two-component acellular pertussis vaccine in infants. Pediatr Infect Dis J. 1997; 16:1038-1044 [CrossRef][Medline]
2. Cherry JD, Heininger U, Stehr K, Christenson P The effect of investigator compliance (observer bias) on calculated efficacy in a pertussis vaccine trial. Pediatrics. 1998; 102:909-912 [Abstract/Free Full Text]
3. Comstock GW Evaluating vaccination effectiveness and vaccine efficacy by means of case-control studies. Epidemiol Rev. 1994; 16:77-89 [Free Full Text]
4. Ad Hoc Group for the Study of Pertussis Vaccines Placebo-controlled trial of two acellular pertussis vaccines in Swedenprotective efficacy and adverse events. Lancet. 1988; 1:995-960
5. Storsaeter J, Hallander H, Farrington CP, Olin P, Möllby R, Miller E Secondary analyses of the efficacy of two acellular pertussis vaccines evaluated in a Swedish phase III trial. Vaccine. 1990; 8:457-461 [CrossRef][Medline]
6. Schmitt HJ, Wirsing von König CH, Neiss A, Efficacy of acellular pertussis vaccine in early childhood after household exposure. JAMA. 1996; 275:37-41 [Abstract/Free Full Text]

In Reply.

We have carefully reviewed the comments of Liese and Walker about our analysis of their study in our recent commentary.¹ However, their presentation does not alter our conclusions. The efficacy that they determined for their DTaP vaccine was inflated because of study design,^2,3 lack of serologic diagnosis,^4-6 and observer bias.⁷

Their letter itemizes 3 points that we will address. They first imply that they controlled observer bias by various physician procedures. However, because there was no prospective monitoring of DTaP, DTP, or DT recipients, it was the parent observers who decided which children to bring to the physician for evaluation. Because disease is more severe in DT recipients than DTaP recipients and the parents expected DTaP to work, it is easy for observers (parents) to overlook mild cases in DTaP vaccinees and therefore inflate efficacy.

Their second point is that case-control design does not per se result in bias. This is true and case-control methodology was extremely useful in evaluating Haemophilus influenzae type B vaccine efficacy trials where the endpoint was severe disease.⁸ However, it has been noted by Fine and Clarkson³ and Fine² that the efficacy estimates in pertussis vaccine trials are considerably higher in case-control studies than in household contact studies or cohort trials.

Liese and Walker's third point is rather vague but suggests:

1. That we shouldn't be concerned about preventing mild disease but only typical disease. We disagree with this because patients with mild disease can also spread disease to susceptibles. In fact, mild cases (and severe cases) among immunized children are often not recognized as pertussis, and, therefore, pertussis in these children can paradoxically be a greater risk to susceptibles than typical cases among unimmunized children because appropriate preventive measures may not be taken.
2. That the high efficacy estimates reported in their trial were comparable with those for the Biken vaccine in the randomized, controlled trial in Sweden 1986.⁴ What they don't seem to realize is that differential sensitivity of culture-confirmation (and of serological confirmation of cases as done in the early Swedish trial)^4,5 gave spuriously high efficacy estimates for the Biken acellular vaccine in both Munich and Sweden. The Swedish investigators recognized these biases and relied on a more careful study design in a later study⁹ with collection of serum samples at 12 months and 2 to 21/2 years that were used as diagnostic preexposure samples, thus increasing the yield of serologically confirmed cases in the immunized groups, resulting in overall lower efficacy estimates, as indicated by Fine² in his review.

James D. Cherry
Department of Pediatrics
UCLA School of Medicine
Los Angeles, CA 90095

Patrick Olin
Swedish Institute for Infectious Disease Control
S-105 21, Stockholm, Sweden

REFERENCES

1. Cherry JD, Olin P The science and fiction of pertussis vaccines. Pediatrics. 1999; 104:1381-1384 [Free Full Text]
2. Fine PEM. Implications of different study designs for the evaluation of acellular pertussis vaccines. In: Brown F, Greco D, Mastrantonio P, Salmaso S, Wassilak S, eds. Pertussis Vaccine Trials Dev Biol Stand. Basel, Switzerland: Karger; 1997;89:123-133
3. Fine PEM, Clarkson JA Reflections on the efficacy of pertussis vaccines. Rev Infect Dis. 1987; 9:866-883 [Medline]
4. Ad Hoc Group for The Study of Pertussis Vaccines Placebo-controlled trial of two acellular pertussis vaccines in Swedenprotective efficacy and adverse events. Lancet. 1988; i:955-960
5. Storesaeter J, Hallander HO, Farrington CP, Secondary analyses of the efficacy of two acellular pertussis vaccines evaluated in a Swedish phase III trial. Vaccine. 1990; 8:457-461
6. Storesaeter J, Blackwelder WC, Hallander HO Pertussis antibodies, protection, and vaccine efficacy after household exposure. Am J Dis Child. 1992; 146:167-172 [Abstract/Free Full Text]
7. Cherry JD, Heininger U, Stehr K, Christenson P The effect of investigator compliance (observer bias) on calculated efficacy in pertussis vaccine trial. Pediatrics. 1998; 102:909-912
8. Greenberg DP, Vadheim C, Bordenave N, Protective Efficacy of Haemophilus influenzae type b polysaccharide and conjugate vaccines in children 18 months of age and older. JAMA. 1991; 265:987-992 [Abstract/Free Full Text]
9. Gustafsson L, Hallander HO, Olin P, Reizenstein E, Storsaeter J A placebo-controlled trial of a two-component acellular, a five-component acellular and a U.S. licensed whole-cell vaccine. N Engl J Med. 1996; 334:349-355 [Abstract/Free Full Text]