Home » News

Gradual Response To Initial Treatment May Be Sign Of Better Prognosis In Multiple Myeloma

10 Comments By
Published: Apr 23, 2018 3:25 pm

Results of a retro­spec­tive­ study conducted at the Mayo Clinic indicate that multiple myeloma patients who respond more gradually to their initial treat­ment may have better over­all survival.

Specifically, the authors of the new study find that newly diag­nosed patients who required more than 120 days to achieve their best response to initial treat­ment had better pro­gres­sion-free and over­all survival than patients who achieved their best response in 120 days or less.

The five-year survival rate was 77 per­cent for patients who achieved their best response to initial treat­ment in more than 120 days, and 56 per­cent for patients who achieved their best response in 120 days or less.

The pos­i­tive impact on survival of a more gradual response to treat­ment was seen both in patients who re­ceived a stem cell trans­plant as part of their initial treat­ment and in patients who did not.

The impact also was present when the researchers controlled for a variety of other factors that could affect prognosis, such as the patient’s age, their best response to treat­ment, and whether or not they had high-risk chromosomal ab­nor­mal­i­ties.

The study is based on data for more than 1,000 multiple myeloma patients who were diag­nosed be­tween 2005 and 2015 at the Mayo Clinic in Rochester, Minnesota.

Based on their findings, the study authors suggest that patients who respond more gradually to initial treat­ment may rep­re­sent a subgroup of patients whose disease is less likely to progress rapidly and less likely to develop resistance to treat­ment.

Background

Overall survival in multiple myeloma patients has in­­creased markedly since the in­tro­duc­tion of novel ther­a­pies such as thalido­mide, Revlimid (lena­lido­mide), and Velcade (bor­tez­o­mib) in the last 20 years (see related Beacon news article).

The new treat­ments have given physicians more options with which to treat multiple myeloma, extending the time each patient has until treat­ment options are exhausted.

In addi­tion, the new treat­ments have made it possible to achieve deeper responses to treat­ment, which generally lengthens the time until a patient’s disease progresses and a new treat­ment must be tried.

Yet a deep response to initial treat­ment does not always guar­an­tee long over­all survival. The authors of the new study note that “survival analyses in both trans­plant and non-transplant pop­u­la­tions have dem­onstrated that up to 20 per­cent of patients achieving a [complete response to initial treat­ment] will die within four years.”

Researchers there­fore are trying to identify addi­tional factors to better predict survival in myeloma patients. Knowledge of such factors could make it easier for doctors to customize ther­apy based on a patient’s prognosis.

Extensive research has dem­onstrated that both the duration of a patient’s initial response to treat­ment, and their depth of response, are strong (but not perfect) predictors of patient survival.

There has been conflicting evi­dence, how­ever, about whether the speed of a patient’s response to initial treat­ment pos­i­tively or neg­a­tively affects patient prognosis.

Studies done before the in­tro­duc­tion of novel myeloma ther­a­pies suggested that a rapid response to initial treat­ment was asso­ci­ated with poorer survival out­comes. Studies involving patients who have received treat­ment with novel myeloma ther­a­pies have had more mixed results. For example:

  • Relapsed patients who received treatment with Velcade-based regimens in a clinical trial had longer time to progression if their M-spike decreased fast after the start of treatment (reference)
  • Newly diagnosed patients in a combined analysis of results from three European clinical trials had survival outcomes that were unaffected by whether or not the patients achieved a complete response before or after six months (reference)
  • Newly diagnosed patients in a single center U.S. clinical trial had lower overall survival if their serum free light chain level dropped rapidly after the start of treatment (reference)
  • Relapsed patients in an international trial testing the efficacy of a Ninlaro (ixazomib)-based regimen had longer progression free survival when they took longer to achieve their best response to treatment (reference)

Given the lack of consensus in studies investigating speed of response in the age of novel ther­a­pies, the authors of the new study decided to in­ves­ti­gate the issue using data from patients at their own institution.

Study Design

The study authors retro­spec­tive­ly analyzed data from all 1,099 multiple myeloma patients who were diag­nosed at the Mayo Clinic's Rochester, Minnesota location be­tween 2005 and 2015, received initial treat­ment with novel agents, and achieved at least a very good partial response to initial treat­ment.

The median patient age at diag­nosis was 63 years old.

The most common initial treat­ments the patients received in­cluded Revlimid and dexa­meth­a­sone (33 per­cent of patients), Velcade, cyclophosphamide, and dexa­meth­a­sone (24 per­cent), and Velcade, Revlimid, and dexa­meth­a­sone (16 per­cent). One third of the patients received a stem cell trans­plant as part of their initial treat­ment, which was defined as receiving a trans­plant within 12 months of starting initial treat­ment.

The median follow-up time was 3.8 years. The median length of initial ther­apy across all patients in the sample was 1.8 years, the authors told The Beacon. First-line ther­apy varied in length across patients depending, for example, on whether patients had an up­front trans­plant and whether they underwent main­te­nance ther­apy.

The median over­all survival from initial diag­nosis was 8.8 years.

Study Results

The study authors found that median time to best response among the patients in their dataset was 4.9 months, and the median duration of best response was 1.8 years.

More im­por­tantly, the researchers found that the time it took for patients to achieve their best response to initial treat­ment affected both their pro­gres­sion-free survival and over­all survival.

The researchers divided the patients in the study into two groups:

  1. Patients who achieved their best response to treatment in 120 days or less
  2. Patients who took more than 120 days to reach their best response.

Median pro­gres­sion-free survival was 2.1 years for patients in the first (fast-responding) group, and 3.3 years in the second (slow-responding) group.

Median over­all survival was about six years in the first (fast-responding) group, and has not yet been reached in the second (slow-responding) group.

The five-year survival rates are 56 per­cent and 77 per­cent for the fast- and slow-responding groups, re­spec­tive­ly.

These dif­fer­ences in survival be­tween the two patient groups are statistically very sig­nif­i­cant.

Just as im­por­tantly, the dif­fer­ences were found to persist even when the authors controlled for other factors that could affect a patient’s prognosis.

For example, the researchers divided patients into two dif­fer­en­t age groups, with 65 years of age as the dividing line. In both age groups, time to best response still had a sig­nif­i­cant impact on survival. Patients who reached their best response in longer than 120 days had longer pro­gres­sion-free survival and over­all survival than patients in the same age group who had more rapid responses to treat­ment.

The researchers got the same result when they divided patients into two groups based on whether or not they had an up­front trans­plant as part of their initial ther­apy. Once again, the patients in both these groups who took more than 120 days to reach their best response had the better survival out­comes.

The study authors also esti­mated several dif­fer­en­t statistical models to test further whether other factors, such as gender, best response to initial treat­ment, type of initial treat­ment, size of initial M-spike, stage at diag­nosis, or presence of high-risk chromosomal ab­nor­mal­i­ties could explain away the impact of time to best response.

In all models, how­ever, time to best response was still a sig­nif­i­cant factor.

Given these results, the authors write that “it is possible that multiple myeloma responding more gradually to initial treat­ment portends some biologic advantage that is independent of, and not simply a surrogate for,” a patient’s disease stage or chromosomal ab­nor­mal­i­ties. Myeloma patients who respond more gradually to treat­ment, the authors con­tinue, may be “less prone to devel­op­ing treat­ment resistance and sub­se­quent disease pro­gres­sion.”

Modified vs. Standard Survival Metrics

In most of their analyses, the study authors focus on “modified” measures of pro­gres­sion-free survival and over­all survival. The measures use as their starting point the time when a patient reaches their best response to treat­ment, rather than the usual starting point (either time of diag­nosis or time when treat­ment starts). Using modified survival measures avoids statistical problems that could occur with the kind of analyses the authors do in this study.

This summary by The Beacon of the researcher’s work, how­ever, reports standard survival metrics because they are easier to interpret.

For more in­­for­ma­tion, please see the study by Mellors, P.W. et al., “Time to plateau as a predictor of survival in newly diag­nosed multiple myeloma,” American Journal of Hematology, April 16, 2018 (abstract).

Photo by Robert S. Donovan on Flickr - some rights reserved.
Tags: , , , ,


Related Articles:

10 Comments »

  • Joyce E. said:

    I enjoyed reading this article. It has some interesting information. I am glad to see the Beacon again publishing research-related articles in addition to the patient and caregiver columns.

  • Marjorie Smith said:

    Thank you very much for this informative article. I had heard of this before but had not seen a report of a detailed study. As I am in the slow responding group, it gives me hope! I remember well when my Consultant told me that he was not concerned about my slow response, and it seems that he was well informed.

  • Craig said:

    I am confused by this article, which should come as no surprise to those who know me. The engineer in me wants to define "best response" as a low or minimum in some measureable quantitative value, such as M-spike. In calculus terms, it's an inflection point where the slope of the response goes from negative (downward) to positive (upward), i.e. the slope is zero (horizontal).

    By definition, once you reach your best, you can only hold your own, or get worse.

    It only seems logical that if you have bottomed out before 120 days and are already increasing because you are no longer at your 'best' or minimum that you have already relapsed. In this case, I am defining remission in the most general terms of 'disease stable and not progressing' and relapse as when 'you are no longer in remission.'

    The article also associates the rate of response with best results. While there might be a strong correlation between the two, they aren't the same.

    What I take from the article, 'The longer your numbers keep improving, the better your odds, regardless of how fast initially.'

  • Coachhoke said:

    I agree with Craig. The authors' conclusions don't seem to be valid, based on those facts.

  • Cheryl G said:

    Craig,

    I'm not sure what the reason for your confusion is.

    Normally, we've been conditioned to think that if "Patient A" achieves a complete response (CR) as her best response, and "Patient B" achieves a very good partial response (VGPR) as his best response, then Patient A should have the better prognosis.

    What this study shows is that if Patient A achieves her CR in 2 months, and Patient B achieves his VGPR in 6 months, then there's a good chance Patient B has the better prognosis.

    That's an interesting result.

    Sure, deeper responses may take longer times to achieve, so the example I just gave may not happen too often. That's why, according to the article, the researchers checked whether the effect of time to best response was still significant even when you control for the response patients achieve. It is.

    Apparently, given the authors' results, if you take two patients, both of whom reached a CR as their best response, the patient who took longer to reach the CR is likely to have a better survival prognosis.

    Again, that's an interesting result. It's different than you might expect, but that's one reason it's so interesting.

  • Craig said:

    Cheryl G,

    Maybe my problem is with the use of the word "best", which in my mind means the lowest cancer burden, or however you'd like to describe it in a quantitative term. Values which you could plot in a graph. You appear to be looking at it in terms of steps, VGPR, PR, CR, sCR, and the like. I am of the mindset you attack the cancer cells with as much as you can tolerate. I doubt it would be wise to decrease initial treatment to delay optimal or best response to occur after 120 days.

  • Andrew said:

    I do not have access to the entire article, so I am not sure if my question is answered there, but I wonder about the selection of 120 days as the dividing point for the study. It seems arbitrary unless there is some evidence that the results divide at that point in time. Perhaps I am missing something here.

  • Nancy Shamanna said:

    I suppose that if a patient reached their best response in fewer than 120 days, then the cycles of initial therapy might be stopped. For example, if the initial therapy was to produce a really good response before a stem cell transplant, then the patients who did not get a really good response by 120 days might have been given more cycles of treatment. So what the findings might be saying is that having a longer time of being given treatment might be beneficial. Even if the fast responders were in remission, if they had been given more cycles of treatment before moving on to other treatments, that might have helped. Of course, this assumes that the best responders actually were given less cycles of drugs.

  • Paul said:

    I also had issues with this concept. It appears that if the "best" one does is measured in the first 120 days (let's use day 110 for this example), then it appears obvious after day 110 you are by definition doing "worse" and are progressing to a worst state. It seems to me that you are starting to go down hill after day 110.

    But let us say the better measurement occurs after 120, or as I think of it as 120+ (plus). Your best day may be day 120, or 240, or 360, 1000, etc. All these numbers are a lot "better than getting "worse" at day 110. Some of the people that start at 110 may never see 120, or 240, etc.

    Or am I looking at this incorrectly?

  • Mike F. said:

    Hi Craig,

    I went and took a look at the article and I think you're expecting a more quantitative measurement of response than what they made. They looked at how long it took a patient to get to their best response, with response being defined as either partial response, very good partial response, or complete response. Then they looked at how long they held that level of response, which they defined as their response plateau (thus the title, "Time to plateau as a predictor of survival in newly diagnosed multiple myeloma"). So really what they're saying is that those who took >120 days to get to their best response (PR, VGPR, or CR) held that response for a significantly longer time than those who took <120 days to get to their best response. It makes sense given that there are no numbers involved in their measurement of response.