Dr. Harry W. Snyder, Jr., Ph.D., and his wife, Renee Peugeot, appeal their convictions and the sentences imposed after they were found guilty of mail fraud, making false statements to the Food and Drug Administration, and conspiring to commit offenses against the United States. Snyder was employed as a Vice-President of Clinical Development at BioCryst Pharmaceuticals, Inc., a small, publicly-traded pharmaceutical company concerned with the development of synthetic drugs. Specifically, BioCryst was involved in the development of BCX-34 for the treatment of psoriasis and cutaneous T-cell lymphoma (CTCL), a relatively rare and potentially fatal form of skin. cancer. Peugeot was employed as the Study Coordinator and Sub-Investigator for Dr. W. Mitchell Sams, the principal Investigator of the BCX-34 clinical studies at the University of Alabama at Birmingham. Snyder and Peugeot’s indictments and convictions stem from their participation in a scheme to falsify and misrepresent data from the clinical studies in order to show the drug’s effectiveness at treating CTCL. Due to Snyder’s stock options, defendants had a substantial personal interest in BioCryst. After the false BCX-34 results were announced,’ BioCryst stock rose to as much as' $13.75 per share. The stock fell dramatically after the fraud was announced.
Trial and Conviction
Defendants first contend certain comments made by the trial judge to the jury through the course of the trial denied them a fair trial by distracting the jury from the legal principles they were to apply in deciding guilt or innocence. Many of the comments referenced by the defendants are clearly innocuous. For instance, the judge’s inquiry as to the medicinal effects of vitamin C — for the purpose of emphasizing the importance of staying healthy after two alternate jurors had been dismissed — is certainly not improper. Even the potentially problematic comments following the judge’s charge to the jury1 cannot be considered “so prejudicial *1294as to amount to the denial of a fair trial.” United States v. Ramos, 933 F.2d 968, 973 (11th Cir.1991). The trial judge’s comments clearly do not warrant reversal.
Defendants also argue for reversal based upon a question by the prosecutor concerning Snyder’s indictment. To prevail on a claim of prosecutorial misconduct, a defendant must not only show improper comments by the prosecutor, but also show those comments prejudicially affected the substantial rights of the defendant. United States v. Castro, 89 F.3d 1443, 1450 (11th Cir.1996). The district court’s denial of defendants’ motion for mistrial is reviewed for abuse of discretion. United States v. Perez, 30 F.3d 1407, 1410 (11th Cir.1994). The comment complained of came on redirect examination by the prosecutor after the defense attorney asked a witness whether misconduct would nullify the severance agreement as it related to Snyder’s stock options. The prosecutor then asked whether an indictment would have been considered a form of misconduct, precluding Snyder’s entitlement to the stock options. The trial judge sustained defense counsel’s objection and instructed the jury to ignore the question. Out of the presence of the jury, the trial judge explained it was unnecessary to give an additional curative instruction because the issue — that an indictment is not evidence of guilt — had been covered “no less than four times” previously. Furthermore, there is no evidence the jury misunderstood the earlier instructions, nor is there a reasonable probability that “but for the offending remarks, the defendant would not have been convicted.” United States v. Calderon, 127 F.3d 1314, 1335 (11th Cir.1997) (quoting United States v. Rodgers, 981 F.2d 497, 499 (11th Cir.1993)). We fail to see any prejudice resulting from the alleged improper comment and hold the trial court’s denial of defendants’ motion for a mistrial was not an abuse of discretion.2
Sentencing Guidelines
Initially, defendants claim the trial court erred in awarding a two-level enhancement for knowingly endangering the clinical-trial volunteers.3 Defendants argue the trial court failed to make explicit findings respecting the enhancement and the government failed to prove their behavior created a risk of serious bodily injury. See U.S. Sentencing Guidelines Manual § 2Fl.l(b)(6)(A) (1998).4 The *1295Guidelines, however, do not'require that the victim actually suffer serious bodily injury. Rather, the question is whether the defendant placed the victim at such a risk. CTCL is a rare and potentially fatal form of cancer. Treatment is essential in slowing the disease’s progress to later, and more deadly, stages. Volunteers testing the CTCL treatments at issue here were required to forego other treatments. Two patients testified they volunteered for the second, six-month study based upon the falsified results of the first, six-week study. CTCL meets the definition of serious bodily injury, and defendants’ conduct qualifies as creating a conscious or reckless risk of such injury. Thus, we hold application of. the two-level enhancement was not erroneous.
The' final issue presented concerns the appropriate measure of the victims’ loss under Sentencing Guideline § 2Fl.l(b)(l). Prior to sentencing, the government obtained an estimated victim loss calculation from an accounting report prepared by Dr. Russell M. Barefield, CPA, Ph.D., at the direction of the U.S. Attorney’s Office. Barefield’s calculations were based on the amount BioCryst stockholders were deemed to have lost due to defendants’ fraud. Barefield concluded the amount of the victims’ loss was approximately $34.4 million. Pursuant to the Sentencing Guidelines, this would require enhancing the base level of six (for crimes involving fraud and deceit) by sixteen levels. U.S. Sentencing Guidelines Manual § 2Fl.l(b)(l)(Q). The trial court, however, found a reasonable estimate of the victims’ loss was not feasible:
To ascertain'whom the individual investors were that suffered financial losses would be exhaustive; indeed highly improbable. It would require significant expenditures of time and resources to determine the large amount of detailed information and no such information is before this Court.
Transcript of Sentencing Hearing at 152-53. The court then found, sua sponte, the better calculation of the victims’ loss would be to use the intended or potential gain attributable to the defendants. Using this method, the court concluded the amount of loss attributable to defendants was between $200,000 and $350,000. Thus, the court awarded an eight-level increase pursuant to Sentencing Guideline § 2Fl.l(b)(l)(I). Defendants argue the trial court erred in calculating the amount of their gain. The government concedes the court erred, but contends the error was the choice of methodology rather than the calculation. In its cross-appeal, the government argues the court should have enhanced the defendants’ sentences based upon the actual loss incurred by the BioC-ryst stockholders.
We review the application and interpretation of the Sentencing Guidelines de novo, and we review findings of fact for clear error. United States v. Harness, 180 F.3d 1232, 1234 (11th Cir.1999). When calculating the monetary value of the victims’ loss under Sentencing Guideline § 2Fl.l(b)(l), substitution of defendants’ gain is not the preferred method because it ordinarily underestimates the loss. U.S. Sentencing Guidelines Manual § 2F1.1 cmt. n. 9; United States v. Orton, 73 F.3d 331, 334 (11th Cir.1996). Where precise figures are not ascertainable, it is preferable to calculate the victims’ loss by determining “the approximate number of victims and an estimate of the average loss to each victimU.S. Sentencing Guidelines Manual § 2F1.1 cmt. n. 9.
In the present case, the trial court erred when it found calculating the loss to victims was not feasible. “All that is required is that the court ‘make a reasonable estimate of the loss, given the available information.’” Orton, 73 F.3d at 335 (quoting U.S. Sentencing Guidelines Manual § 2F1.1 cmt. n. 9). The court was cor*1296rect, though, in resisting Dr. Barefield’s ultimate conclusions. While presenting a thorough analysis, Dr. Barefield uses the number of outstanding shares as a proxy for the number of victims. This overestimates the actual loss because, unlike the circumstances presented in United States v. Hedges, 175 F.3d 1312, 1314 n. 6 (11th Cir.1999), the stock here was not totally worthless after the conspiracy was discovered. Thus, not every shareholder suffered a loss. We understand the trial court was attempting to follow our decision in Orton, where this court held the “loss to losing victim” approach is not required in every case. Orton, 73 F.3d at 334. Just because each individual’s precise loss cannot be ascertained does not mean that the district court should abandon a loss calculation altogether. Dr. Barefield’s research suggests a reasonable estimate of the victims’ loss based on existing information is feasible. Therefore, the trial court will need to recalculate the monetary loss under Sentencing Guideline § 2Fl.l(b) by estimating the loss caused by defendants’ fraud.5
Defendants claim there is no actual loss to the innocent (i.e., non-insider) BioCryst stockholders because the price of BioCryst stock was higher after disclosure of the fraud than its average price during the life of the fraud. While this method of valuing loss — subtracting the stock price after, the fraud from the average stock price over the life of the fraud — may be appropriate in many cases involving securities fraud, it is certainly not the only method available to the court. Keeping in mind the need to achieve only a “reasonable estimate” of the loss to the victims, the court may need employ a slightly different, but nonetheless proper, methodology. For instance, it may be most effective to focus on the period between the May 31, 1995, press release announcing BCX-34 was “effective” in treating CTCL and the days immediately following the announcement of the fraud on June 16, 1995.6 It is fair to say that, as a result of the fraud, a large number of individuals and institutions were induced to purchase BioCryst stock at an artificially inflated price. Their losses are substantially undervalued by utilizing the “gain to defendants” approach.
Accordingly, on the issue of calculating loss under Sentencing Guideline § 2Fl.l(b), we REVERSE and REMAND for further proceedings consistent with this opinion. We AFFIRM the district court’s rulings with respect to all other issues raised.