Disease severity in chronic graft-versus-host disease: doctors’ gut feeling versus biostatistics?

Chronic graft-versus-host disease (GVHD) is a frequent and potentially life-threatening complication of allogeneic hematopoietic stem cell transplantation. An increase in transplants in older patients and the more frequent use of unrelated donors have led to greater numbers of patients with this painful complication. Recent advances have been made in understanding the pathophysiology of chronic GVHD as well as in establishing precise criteria for the diagnosis and classification of disease manifestations. These advances will, it is hoped, pave the way to improving both the prophylaxis and treatment of chronic GVHD. We recently reviewed current issues in chronic GVHD with Dr. Ritz and readers of Haematologica interested in this field can find detailed information in this article.1

In the past, chronic GVHD included any clinical manifestations of GVHD that occurred beyond 100 days after transplantation. This definition was clearly imprecise and became inadequate. In 2005 a group of experts met under the auspices of the National Institutes of Health (NIH), USA in a consensus meeting. The goals of this NIH consensus working group on the diagnosis and staging of GVHD were: (i) to establish criteria for diagnosis of the disease, emphasizing the distinction between acute and chronic GVHD; (ii) to define criteria for scoring the severity of clinical manifestations in affected organs; and (iii) to propose categories describing the overall severity of the disease and the indications for treatment.2

The NIH consensus conference recognized two main categories of GVHD, each with two subcategories. The broad category of acute GVHD includes classic acute GVHD (maculopapular erythematous rash, gastrointestinal symptoms, or cholestatic hepatitis). The broad category of chronic GVHD includes classic chronic GVHD, presenting with manifestations that can be ascribed only to chronic GVHD. Chronic GVHD also includes an overlap syndrome, which has diagnostic or distinctive manifestations of chronic GVHD together with features typical of acute GVHD.

Numerous prognostic indices in chronic GVHD have been described.1 Thrombocytopenia (platelet count <100×10/L) is the first reported and most reproducible prognostic factor even when using NIH criteria. Other factors, such as diarrhea, might be prognostic only due to the older definition of the disease or to the worse prognosis of the overlap syndrome. The NIH consensus conference proposed a new global chronic severity score establishing mild, moderate and severe forms of chronic GVHD based on a numerical scoring system for individual organs to calculate a summary scale.2 Although the NIH global score was developed through expert opinion, several studies have shown that the global score at onset of chronic GVHD is associated with risk of subsequent mortality.63 However as nicely described by Inamoto et al. in this issue of Haematologica,7 several issues remain to be elucidated. Firstly, since the NIH global score was based on expert opinion and was not originally intended to predict mortality, does this score provide an optimal model for predicting mortality risk in patients with chronic GVHD¿ The authors hypothesized that empirically derived estimates of overall mortality risk incorporating the relative importance of different organ involvement might be more accurate than estimates derived from the opinion-based global score. Secondly, does the NIH global score predict mortality risk when it is applied at time points after the onset of treatment for chronic GVHD¿ Thirdly, does the NIH global score correlate with risk of recurrent malignancy¿ To address these important clinical questions, Inamoto et al. performed an impressive set of analyses on data collected in a prospective, multicenter, longitudinal, observational study of patients with chronic GVHD. The authors analyzed 574 adult patients with chronic GVHD, using multivariate time-varying analysis accounting for serial changes in severity of disease in eight individual organ sites over time. They randomly divided the cohort into a training set (482 patients who underwent 1602 visits) and a validation set (including 150 patients with 600 visit ratings). In the training set, severity of skin, mouth, gastrointestinal tract, liver and lung involvement were independently associated with the risk of non-relapse mortality. Weighted mortality points were assigned to individual organs based on the hazard ratios. The population was divided into three risk groups based on the total mortality points. The three new risk groups were validated in an independent validation set, but did not show better discriminative performance than the NIH global score. As compared to moderate or mild global score, severe global score was associated with increased risks of non-relapse mortality and overall mortality across time but not with a decreased risk of recurrent malignancy. Thus at a first glance one might have the impression that doctors’ gut feeling (expert opinion) might discriminate overall chronic GVHD severity better than sophisticated biostatistics! Is this true¿ The response, as usual in good medical science, is both yes and no!

First the statistical analysis performed by Inamoto et al. must be applauded. In my view it is one of very few examples of how a statistical analysis can reproduce the true life in chronic GVHD. Every physician with expertise in the field will easily recognize that one of the hallmarks of chronic GVHD is its evolution; with flares and waves of manifestations increasing and decreasing sometimes due to new treatment but sometimes also in the absence of treatment modification (which is what PJ Martin often refers to as the waxing and waning of chronic GVHD symptoms). Some years ago Dr. Martin represented this evolution graphically and drew a cartoon (Figure 1A) which he kindly allowed to be published here in Haematologica. The real tour de force of the work by Inamoto et al. was to take advantage of a large cohort of patients who underwent hundreds of visit ratings to model the erratic evolution of chronic GVHD (Figure 1B). The only other study that aimed to model this evolution was one reporting the long-term results from a randomized trial on GVHD prophylaxis with or without anti-T-cell globulin8 in which Dr. Schmoor used a similar (but not identical) methodology. Both of these models87 will be of major interest in the analysis of future long-term studies of chronic GVHD.

Figure 1.The waxing and waning manifestations of chronic GVHD.

The second issue is the predominance of clinical manifestations in estimating disease severity both in single organs and overall. Once again, it should be remembered that the current scoring system for each organ and the consequent overall grading was derived from experts’ discussion and not from analysis of real patients’ data. In the cohort analyzed in this paper, the patients were basically scored by the same senior authors who described the scoring system in 2005. Furthermore the inherent caveat regarding the Chronic GVHD Consortium analysis has always been that it is a mixture of incident cases (for which organ severity can be studied prospectively without problem) and of prevalent cases (for which prognostic studies could be the subject of controversies since by definition patients have to be alive long enough to be able to be scored after months, or even years of evolution).

It should be noted that patients with chronic GVHD now tend to survive for quite a long time and the median follow-up time of survivors in this study is rather short (for chronic GVHD): i.e. just a little longer than 3 years. Thus, longer follow-up of prevalent cases and studies including larger numbers of patients might lead to different conclusions.

Finally, the authors found no relationship between chronic GVHD severity and relapse; this might be of interest but does not rule out that the graft-versus-leukemia effect is mainly restricted to NIH-defined chronic GVHD (and not to acute GVHD).9

Footnotes

• Gerard Socie, MD, PhD, is Head of the Hematology-Immunology-Oncology Division at the Hospital St Louis, Paris; Professor of Hematology at the University Paris 7; and current President of the French Society of Hematology
• Financial and other disclosures provided by the author using the ICMJE (www.icmje.org) Uniform Format for Disclosure of Competing Interests are available with the full text of this paper at www.haematologica.org.